experimental study depression

Stanford depression treatment nearly 80% effective

Stanford neuromodulation therapy (SNT), an experimental, accelerated version of magnetic pulse brain stimulation developed at Stanford, may provide a “revolutionary” treatment for severe depression, according to researchers at Stanford’s School of Medicine.

The researchers’ optimism comes on the heels of an Oct. 29 study in which the treatment was found to have induced remission in nearly 80% of participants. The study tested the efficacy of SNT on patients with long-term, moderate to severe depression who were also “treatment resistant,” meaning that they had unsuccessfully tried several other conventional forms of treatment. For many of these patients, SNT was the first treatment that successfully alleviated their depression in a significant, lasting way, researchers said.

“My brain has been rebooted,” said one patient treated with a modified, clinical adaptation of SNT who asked to remain anonymous due to concerns regarding the stigmatization of mental health treatment. “It’s like this cloud of depression has been lifted from me.”

SNT is an adaptation of an already existing, noninvasive form of brain stimulation called transcranial magnetic stimulation (TMS). TMS delivers magnetic pulses to specific locations within the brain, activating neural circuits that show decreased activity during depressive episodes. According to researchers, while TMS has been approved for use by the Food and Drug Administration (FDA) since 2008, its efficacy is limited, and it can take weeks for patients to show improvement. 

SNT differs from conventional TMS in several key ways. First, SNT relies on functional neuroimaging to map out the brain, allowing researchers to pinpoint where to deliver magnetic pulses. Second, SNT deploys an especially efficient stimulation pattern that relies on fewer pulses to change neural circuits. Third — and most crucially — SNT dramatically increases the rate at which patients receive brain stimulation. While TMS involves one brain stimulation session per day for 36 weekdays, SNT requires only 10 stimulation sessions per day over a five-day period.

“There have been a lot of innovations in the TMS technology space,” said Nolan Williams, the lead investigator in the Oct. 29 study. “The question we tried to answer was, ‘how do we take what we now understand and reengineer TMS?’ And that’s what we did.”

Williams, who is the Director of the Stanford Brain Stimulation Lab and an assistant professor of psychiatry and behavioral sciences, said that the research team plans to submit the data to the FDA, which has already deemed SNT a “ breakthrough therapy .” Williams said that the timeline for the treatment’s deployment will depend on whether or not it receives FDA approval, although he is optimistic about its potential to treat a broad range of neurological disorders.

“One application [of SNT] is for people who are in psychiatric emergencies,” Williams said. “We’ve been doing this in obsessive compulsive disorder and borderline personality disorder. We think that this is a brain tool.”

SNT’s versatility makes it a much more compelling alternative to other treatments, said Kristin Raj, the co-chief of Stanford Mood Disorders and chief of the Stanford Bipolar Clinic. As one of the researchers in the study, Raj said that the study’s results demonstrate that SNT is far more effective than medication, which often produces diminishing returns after the first few doses. Moreover, SNT has none of the side effects that come with other popular treatments for depression like electroconvulsive shock therapy or ketamine therapy, which are still viewed as controversial.

“I’ve had many patients tell me how much hope it gives them to hear about SNT,” Raj said.

Still, despite overwhelmingly positive results, SNT is not a miracle cure. 

Researchers have said that not all patients respond in the same way to the treatment — while some patients are still in remission years after being treated, others have reported relapse after only a few weeks. 

Most individuals fall somewhere in the middle, said psychiatry and behavioral sciences professor Brandon Bentzley, who offered one hypothesis as to why responses to SNT might vary. 

“What we’re doing in SNT is trying to move the brain from the default mode network, which is dominant during depression, to the central executive network,” Bentzley said, another researcher in the study. “My speculation is that different people have a different propensity to shift back to the default mode network.”

The question of how to extend the positive effects of SNT to all patients, Bentzley said, remains an active research focus.

David Carreon, another researcher for the study and a clinical assistant professor of psychiatry and behavioral sciences, said that he has already adapted a modified version of SNT to his private practice, albeit without the functional neuroimaging techniques used in the Oct. 29 study. While the treatment is not as effective as the full version of SNT, it is still showing success in patients he said.

While SNT might offer hope for depression treatment, the stigma that hangs over mental health issues remains dangerous, according to the same patient who requested anonymity. The patient, who is a physician and was treated by Carreon, said that this stigma is especially prevalent within the healthcare industry, which takes an almost “militaristic” approach to denying the existence of mental health issues in its workforce. 

“You would think that the healthcare system, since it’s so good at treating others, would be better at providing treatment for physicians,” the patient said. “But it’s more like, ‘Everyone else has problems, but we don’t.’”

Carreon said that many of his patients, including those who have gone into remission, have refused interviews with media outlets looking to report on SNT due to this stigma.

“People don’t want to be on national TV as the guy who was depressed,” Carreon said. “Even if it was in the past.”

The stigma surrounding mental health has real-world dangers; according to Enas Dakwar, a clinical psychologist at the Vaden Health Center, social stigma prevents many individuals from seeking support for mental health. This is exacerbated by the fact that many individuals with depression are high-functioning, Dakwar said, which can lead them to try to “fix” their depressive state on their own. Without proper support, these individuals often fail, “taking them deeper into the rabbit hole” of depression, Dakwar said.

This article has been updated to reflect that the treatment one of the patients received was a clinical adaptation of SNT and to correct the spelling of David Carreon’s name. The Daily regrets this error.

Brandon Kim '25 is a Desk Editor for Campus Life and the Diversity, Equity and Inclusion Chair. He was previously a Managing Editor for The Grind from Vol. 262-263. Now studying philosophy, he has probably tried out every major here. Ask him about baseball, hiking very tall mountains and old-school Korean pop.

Login or create an account

In Sight: The Biological Diagnosis of Depression and Anxiety

The science of psychiatry is gaining on the daunting complexity of the brain..

Updated June 20, 2024 | Reviewed by Hara Estroff Marano

• What Is Depression?
• Find a therapist to overcome depression
• Diagnostic models in psychiatry are largely based on clinical experience, with some statistical modeling.
• Scientific biomedical models are necessary to advance understanding of mental illness.
• Understanding the biology of mental illness will allow development of better treatments.
• Personalized analysis of brain networks holds promise for people suffering from depression and anxiety.

According to the World Health Organization (WHO), clinical depression affects nearly 300 million people worldwide. The Centers for Disease Control (CDC) estimates that 20 million or more people in the U.S. have depression at any given time, while more than 18 percent of U.S. adults report depression at some point in their lives and .more than 12 percent of adults report significant feelings of anxiety .

Treatment for depression is of limited effectiveness; only 30-40 percent of those initially treated experience full resolution of symptoms, or remission. What's more, studies show, successive efforts to achieve remission are less and less effective.

Understanding the underlying biology of depression, anxiety, and related conditions such as post- traumatic stress disorder ( PTSD ) is necessary for making correct diagnosis and planning effective treatment. But especially in psychiatry, given the complexities of the brain, diagnosis and treatment are not yet well-grounded in biological understanding.

Medical treatment is ideally based on a number of factors, including knowledge of the disease process, the ability to make accurate diagnoses, and an understanding of how individual factors affect treatment planning and outcome. The National Institutes of Health started the BRAIN Initiative (Brain Research Through Advancing Innovative Neurotechnologies) in 2013, calling for neuroscience -based models of disease and health. Understanding the causal factors of disease suggests the levers clinicians can manipulate to provide the most effective treatment possible.

Toward a More Scientific Psychiatry

Psychiatric diagnosis in the United States is currently based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). Although efforts have been made to improve its approach with more specific criteria for mental illness based on statistics and available data, the DSM is not firmly scientifically based. For the vast majority of illnesses described, the diagnostic criteria say little to nothing about the cause of the disease; instead, they primarily reflect long-observed clinical patterns, rendering the DSM a work in progress and of less-than-desirable utility for diagnosis and treatment.

Not all causes of psychiatric disease are strictly biological. For example, inheriting a genetic predisposition to depression—and many genes contribute–does not invariably lead to depression. It’s highly likely that social and family factors can also bring on depression—not being productive or social, for example. Current treatments for depression often address biological and social factors, but there is as yet no standard biological testing for depression nor any clear framework for scientifically based treatment.

DSM 5 identifies several subtypes of depression, all based on clinical observation and statistical analysis. With unipolar depression (as contrasted with bipolar disorder ), subtypes include atypical and melancholic, and specifiers include severity of symptoms as well as presence or absence of psychotic features. Diagnosis is made by reviewing clinical presentation and history, whether informally through clinical interview or formally via structured, guided interview and review of accompanying information. The same approach applies to anxiety disorders, including generalized anxiety, social anxiety , panic disorder, and obsessive-compulsive disorder, as well as to stress-related disorders such as PTSD. The disorders may be divided into subtypes by specifiers, but a true medical-scientific framework is lacking.

Biotyping Depression and Anxiety

A recent study of brain networks in depression and anxiety reported in Nature Medicine (2024) is an important step toward establishing an empirical model of "biotypes",; it echoes prior work on depression 3 and brain-based personality research 4 . Researchers Tozzi and colleagues used functional magnetic resonance imaging (fMRI) to look at brain activity in more than 1,000 people with depression and anxiety, measuring “task-free” brain activity. They repeated imaging in a subset of patients who received psychotherapy or pharmacotherapy or underwent a variety of activities. During the imaging, subjects were shown a variety of stimuli—such as sad, threatening, or happy faces—and were asked to perform a variety of cognitive and attention tasks. Cluster analysis was used to identify underlying biotypes based on brain circuit dysfunction, sometimes referred to as “dysconnectivity” in brain networks. Treatment would, therefore, restore "euconnectivity".

Notably, the study was transdiagnostic: Given how much depression an anxiety overlap, the study didn't assume they are separate disorders. The analysis was unbiased by preconceived diagnostic models. Participants included those diagnosed with various conventional disorders, including major depression, generalized anxiety, panic disorder, social anxiety, PTSD, obsessive-compulsive disorder,. Some patients met criteria for more than one diagnosis.

Six underlying biotypes of depression and anxiety were identified among participants with clinically-significant symptoms. Their labels are complicated, based on activity levels in key brain networks: default mode, or resting state (D); salience, or what stands out as important (S); and attentional (A). In addition to connectivity patterns, research looked at such key factors as negative emotional circuitry in response to sadness and threat (conscious and unconscious ), positive emotion circuits, and cognitive circuits. There were no sex differences in response patterns, and minimal differences in age.

• Biotype D C+ S C+ A C+ . This cluster showed hyperconnectivity among all three networks. This biotype had slow responses identifying sad faces and increased errors in executive function tasks. The response to behavior coaching for wellness was strong.
• Biotype A C− . This cluster had less connectivity in the attention network, less severe stress compared with other biotypes, and relatively little dysfunction in cognitive control, with faster responses on cognitive measures but more errors, as well as faster priming when viewing threatening faces. Response to behavioral coaching was less robust.
• Biotype NS A+ P A+ . This cluster had elevated activity during conscious processing of emotions, with sadness evoking greater negative circuit activity and happiness more positive. This group also experienced more severe anhedonia —inability to enjoy things—and greater ruminative brooding.
• Biotype C A+ . This group showed increased cognitive control under specific conditions. This cluster also showed greater anhedonia, greater anxious arousal, negative bias , and dysregulation in response to threat. Cognitive errors were high, especially with sustained attention. This biotype had a better response to the antidepressant venlafaxine, a serotonin-norepinephrine reuptake inhibitor (SNRI) commonly prescribed.
• Biotype NTC C- C A− . This small group was characterized by loss of functional connectivity in negative emotion circuits during conscious processing of threatening faces and by reduced activity in cognitive control circuits. This cluster had less ruminative brooding and faster reaction times to sad faces.
• Biotype D X S X A X N X P X C X . This small group did not show significant circuit dysfunction. There were slower reaction times to implicit threat.

Implications

While not ready for standard clinical use, the study results build on prior work demonstrating that brain network analysis holds promise for developing biologically based diagnostic testing for depression, anxiety, and stress-related disorders. The study also provides initial proof of concept that psychiatric biotyping could be used in the selection of treatments, with some biotypes responding better to medication and others to psychotherapeutic interventions.

Clearly, more work is needed before such models of illness can underpin diagnosis and treatment. Given the complexity of the human experience, it's important to recognize that many of the causes of mental illness are likely to be social or circumstantial, external to the individual. More debatable is how to distinguish psychology from neuroscience, mind from brain, without becoming neuroreductionistic. Personalized scientific approaches to psychiatry on a par with other medical disciplines remain largely aspirational, but current approaches are likely to move the needle.

1. An important clarification about how the word “causal” is being used here–it is being used to refer to the causes of the problems in the present moment, the precise factors in the complex system which maintain the status quo of health and illness. Notably, we are not necessarily talking about the historical causes–what started the process in motion may not be what is currently causing it to persist. This is a mathematic definition of causality.

2. Causal discovery has been used to look at PTSD among police officers using a process called Protocol for Computation Causal Discovery in Psychiatry (PCCDP). Saxe and colleagues (2020) reviewed a large data set from over 200 police officers. They identified 83 causal pathways with 5 causes: changes (single-nucleotide polymorphisms–SNPs) in histidine decarboxylase and mineralocorticoid receptor genes involved with stress-response, acoustic startle to low perceived threat during training, peritraumatic distress to incident exposure in the first year of service, and general symptom severity during training after one year of service. This study is a proof-of-concept for using causal discovery to identify points for intervention, and clearly could be used preventively–for example, identifying trainees with those features and responding accordingly.

3. Four Biotypes of Depression

4. Brainprint of Basic Mental Activity

WHO Depression Fact Sheet

CDC Depression Prevalence 2020

CDC Anxiety

NIH Brain Initiative

Saxe GN, Bickman L, Ma S, Aliferis C. Mental health progress requires causal diagnostic nosology and scalable causal discovery. Front Psychiatry. 2022 Nov 15;13:898789. doi: 10.3389/fpsyt.2022.898789. PMID: 36458123; PMCID: PMC9705733.

Saxe GN, Ma S, Morales LJ, Galatzer-Levy IR, Aliferis C, Marmar CR. Computational causal discovery for post-traumatic stress in police officers. Transl Psychiatry. 2020 Aug 11;10(1):233. doi: 10.1038/s41398-020-00910-6. PMID: 32778671; PMCID: PMC7417525.

Tozzi, L., Zhang, X., Pines, A. et al. Personalized brain circuit scores identify clinically distinct biotypes in depression and anxiety. Nat Med (2024). https://doi.org/10.1038/s41591-024-03057-9

Grant Hilary Brenner, M.D., a psychiatrist and psychoanalyst, helps adults with mood and anxiety conditions, and works on many levels to help unleash their full capacities and live and love well.

• Find a Therapist
• Find a Treatment Center
• Find a Psychiatrist
• Find a Support Group
• Find Online Therapy
• United States
• Brooklyn, NY
• Chicago, IL
• Houston, TX
• Los Angeles, CA
• New York, NY
• Portland, OR
• San Diego, CA
• San Francisco, CA
• Seattle, WA
• Washington, DC
• Asperger's
• Bipolar Disorder
• Chronic Pain
• Eating Disorders
• Passive Aggression
• Personality
• Goal Setting
• Positive Psychology
• Stopping Smoking
• Low Sexual Desire
• Relationships
• Child Development
• Self Tests NEW
• Therapy Center
• Diagnosis Dictionary
• Types of Therapy

At any moment, someone’s aggravating behavior or our own bad luck can set us off on an emotional spiral that could derail our entire day. Here’s how we can face triggers with less reactivity and get on with our lives.

• Emotional Intelligence
• Gaslighting
• Affective Forecasting
• Neuroscience

ORIGINAL RESEARCH article

Insensitivity to success and failure: an experimental study of performance-based feedback in depression.

• 1 Medical Psychological Center, The Second Xiangya Hospital, Central South University, Changsha, China
• 2 Medical Psychological Institute of Central South University, Changsha, China
• 3 National Clinical Research Center for Mental Disorders, Changsha, China

Objective: This experimental study set out to examine the effects of performance feedback (success or failure) on depressed emotions and self-serving attribution bias in inpatients suffering from major depressive disorder (MDD).

Methods: The study was based on a 2 × 2 experimental design in which 71 MDD patients and 59 healthy controls participated. Both groups (MDD and controls) were randomly assigned to two conditions: success or failure in the performance feedback. A section of Raven’s Standard Progressive Matrices (SPM) was used as a bogus test of the participants’ reasoning abilities, and the Core Depressive Factor of the Zung Self-Rating Depression Scale was used to measure changes in depressed emotion in the subjects following the performance feedback. Participants then rated the accuracy of the SPM as a measure of their reasoning capacity.

Results: The levels of depressed emotions in patients with MDD did not differ significantly under the two feedback conditions. In contrast, depressed emotion levels increased significantly in healthy individuals in response to failure feedback but did not change in response to success feedback. With regard to the ratings of SPM accuracy, there was no significant difference across the two feedback conditions for depressed patients; however, the accuracy ratings were higher in the success condition than in the failure condition for the controls.

Conclusion: Individuals with MDD exhibit blunted emotional reactivity when experiencing new positive or negative social stimuli, supporting the theory of Emotion Context Insensitivity. In addition, self-serving attribution bias does not occur in MDD, which is consistent with the theory of learned helplessness in depression.

Introduction

Major depressive disorder (MDD) is the most prevalent mood disorder, characterized by persistent and severe low moods, apathy, and associations with obvious suffering and functional impairment ( Bora and Berk, 2016 ). According to the Diagnostic and Statistical Manual of Mental Disorders (fifth edition), one of the diagnostic criteria of MDD is that depressive symptoms must occur for at least 2 weeks and include deficient positive affects (e.g., anhedonia) and/or excessive negative affects (e.g., guilt, sadness). Clark et al. (1994) also suggested that patients diagnosed with MDD reported low positive and high negative affects on various questionnaires and interview measures. Durable mood disturbance is therefore considered one of the most distinguishable features of MDD ( Clark et al., 1994 ).

The cognitive theory of depression proposed by the American psychologist Beck posited that painful childhood experiences of failure or abandonment might lead people to form a negative cognitive schema ( Beck, 1967 ). This relatively stable and potential cognitive structure acts as a filter, allowing people to selectively notice and memorize stimuli consistent with their schema, while inconsistent information is unconsciously ignored. Bower (1981) further conceptualized the mood-congruent effect as “the enhanced effect of materials which is congruent with ongoing mood on the process of encoding and/or retrieval,” which was subsequently confirmed by a number of experimental studies ( Bower, 1981 ; Natoli et al., 2016 ; Rygula and Popik, 2016 ). From this perspective, stimuli with a negative valence that match the persistent and low mood states of MDD patients may increase their reactivity regarding their depression level. However, Rottenberg et al. (2005) proposed the Emotion Context Insensitivity (ECI) hypothesis based on the observation that depressed inpatients exhibited very few changes in terms of expression and behavior in response to a range of environmental events. Specifically, the hypothesis suggests that mood states in MDD greatly reduce enthusiasm for activities and lead to social withdrawal behavior and a reduced emotional reactivity to new positive or negative stimuli ( Rottenberg et al., 2005 ; Rottenberg, 2007 ). Further, Steele et al. (2007) found no change in reaction times to feedback information (“win” or “lose”) in depressive illness. Despite previous studies, however, the relationship between depressed emotions and positive/negative stimuli remains unclear.

Researchers have adopted the method of experimental ethology to study various emotions, and a variety of stimuli that elicit emotional responses have been proved to be a validated procedure, such as viewing pictures or videos, listening to radio programs, and engaging in certain laboratory tasks with success or failure feedback on their performance ( Kayikcioglu et al., 2016 ; Brinkmann and Brixius, 2017 ). In the social sciences, feedback on performance is called performance appraisal, which reflects one aspect of social feedback. As one of the most important types of social information, Ruff and Ernst (2014) defined social feedback as comments made by others that are opinions on our personality traits or beliefs about their preference, satisfaction, and willingness to interact with us. They proposed that this feedback engages three psychological processes: anticipation, consumption, and emotion regulation. Researchers have focused more on the consumption aspect of social feedback and pointed to a lack of understanding regarding anticipation and emotion regulation ( Kupferberg et al., 2016 ).

To examine the effect of success and failure feedback on emotion and self-recognition, a bogus test was conducted by Linville (1985) ; the test was described as an analytical task related to certain aspects of intelligence. Further studies of self-recognition have shown that Raven’s Standard Progressive Matrices (SPM), an achievement test in terms of intelligence without cultural or linguistic restrictions, could be used as an efficient tool in the progress of performance evaluation ( Cai and Yang, 2003 ; Xi et al., 2007 ). In the current study, participants completed an SPM test and received personal feedback regarding inferior or superior performance, which was manipulated experimentally by varying the difficulty of the test. This allowed us to examine the effects of experiences with different valences on negative emotion in MDD patients and to clarify the characteristics of emotional regulation and reactivity among these patients in terms of performance-based social feedback.

Cognitive theories of emotion have asserted that individuals’ emotional responses to success and failure are governed by their beliefs about the causes of their performance. Further, the learned helplessness theory of depression ( Seligman, 1975 ) posited that self-serving attribution bias plays an important role in mood disorders, and the suggestion has been supported by many empirical studies ( Morris, 2010 ; Jonas et al., 2014 ); however, results have been contradictory in different cultural contexts ( Guo et al., 2011 ). Self-serving attribution bias is the term used to describe the tendency to give credit to ourselves for success but attribute failure to external sources. To examine the effect of success and failure feedback on causal attribution, Dutton and Brown (1997) asked subjects to rate on a 9-point scale the validity of an accomplished achievement task in assessing integrative orientation ability. Based on their experimental design, we compared differences in self-service attribution bias across feedback conditions (success or failure) within each group (MDD or healthy controls) to explore the relationships between depressed emotion and causal attribution. Generally, our research could provide a clinically meaningful reference for identifying the best time during the course of MDD for psychotherapeutic intervention (i.e., supportive psychotherapy and cognitive behavior therapy) and for helping to stabilize patients’ conditions and promote their physical and psychological health.

Two hypotheses were tested in this study. (1) Trends for patients with MDD and healthy individuals would be different when facing social feedback of divergent valence (success or failure), supporting different theoretical models. First, findings in depressed individuals would not change with the valence of social feedback, in support of the ECI theory. Second, healthy individuals’ depressed emotions would support the mood-congruent theory, in that their depression level would decrease after receiving success feedback and increase after receiving failure feedback. (2) Healthy controls receiving success feedback would report that the test measured their real intelligence level, whereas those receiving failure feedback would not, indicating a self-serving bias in causal attribution. In contrast, the results in the depression group might support the learned helplessness theory and indicate no self-serving attribution bias after receiving any type of feedback.

Materials and Methods

Participants.

A total of 130 Chinese participants took part in this experiment, including a depression group of 71 outpatients recruited from the Psychiatric Clinic of Xiangya Second Hospital at Central South University in China (35 male, 36 female; mean age 25.39 years, SD = 9.152). For the control group, 59 healthy participants were recruited by media advertisements (32 female, 27 male; mean age 24.85 years, SD = 8.672); these participants reported no prior or current history of depression.

The inclusion criteria for the depression group were as follows: (1) acute phase of the first episode of MDD; (2) screened with the Mini-International Neuropsychiatric Interview (MINI) and diagnosed as meeting Diagnostic and Statistical Manual of Mental Disorders (DSM)-V diagnostic criteria for MDD by a qualified clinical psychologist; and (3) total depression scores on the Zung Self-Rating Depression Scale (SDS) ≥42 ( Duan, 2012 ). The mean SDS scores for the depression and control groups were 52.141 (SD = 6.545) and 35.085 (SD = 4.403), respectively. The exclusion criteria for both groups were: (1) psychiatric medicine use; (2) suffering from Persistent Depressive Disorder (Dysthymia) or other psychiatric disorders; (3) cognitive impairment caused by neurological disorder or other physical diseases; or (4) diagnosed with MDD in remission.

All participants signed informed consent forms, and the study was approved by the local ethics committee. General participant information is summarized in Table 1 .

Table 1. Participant characteristics by group.

Measurements

Mini-international neuropsychiatric interview.

The MINI is a short, structured, diagnostic interview designed to assess diagnostic criteria according to the DSM-IV. It takes 15 min to administer, meeting the time limitations of clinical trials and epidemiological studies ( Sheehan et al., 2010 ).

Zung’s Self-Rating Depression Scale

We used the 20 items of the Zung SDS as a measurement to recruit and classify participants, and the eight items of the Core Depressive Factor (items 1, 3, 6, 14, 17, 18, 19, and 20) to examine changes in depressed emotion. The Core Depressive Factor (CDF) has the greatest weight, accounting for 23.8% of the SDS variance, which mainly reflects emotional or affective symptoms of depression ( Romera et al., 2008 ). Besides, Romera et al. (2008) reported that the Congruence Coefficient of the CDF was 0.98, which represents very high agreement according to Sakamoto et al. (1998) . Additionally, Cronbach’s alpha was 0.837 for the CDF among our participants ( n = 130), indicating a satisfactory internal consistency.

Raven’s Standard Progressive Matrices

Raven’s SPM is a non-verbal test that overcomes limitations of language and educational background and has been used to assess reasoning ability. Twenty problems from the SPM were selected so that the difficulty level of the test could be varied to ensure that it reflected either a success or failure condition. Half of the problems were easy (success condition), and the rest were difficult (failure condition). The difficulty level was determined on the basis of prior testing with an independent sample, and the difficulty coefficients of the easy and difficult questions in the current study were approximately 0.8 and 0.2, respectively. Each participant was asked to complete 10 questions within 5 min ( Raven, 1983 ).

The study adopted a 2 × 2 mixed experimental design [(groups: MDD, healthy) × (feedback conditions: success, failure)]. At the start of the experiment, participants completed the eight items from the Core Depressive Factor of the Zung SDS, answering the questions according to their feelings at that moment. A 4-point scale was adopted, ranging from 1 (“strongly disagree”) to 4 (“strongly agree”). They were then asked to estimate how many problems (out of 10) they expected to solve correctly in the subsequent SPM test.

The reasoning test of the SPM was designed as the basic task for the entire experiment. Each participant randomly conducted Task A or Task B before receiving performance feedback, and their actual score was recorded. Task A was designed to be a successful experience, involving 10 easy problems, and positive feedback was provided irrespective of outcome. Conversely, Task B required participants to solve 10 difficult problems within 5 min and was followed by negative feedback irrespective of outcome to elicit feelings of failure. The success feedback was: “Congratulations! The results of our most authoritative intelligence test show that you have brilliant reasoning capacity, significantly exceeding 90% of your peers.” The failure feedback was: “I’m so sorry that you have failed the test. The results of our most authoritative intelligence test shows that you have poor reasoning capacity, lagging behind 90% of your peers.” Both groups of subjects were randomly assigned to the two feedback conditions. In the depression group, 39 and 32 patients were assigned to the success and failure conditions, respectively; the corresponding numbers for the healthy group were 31 and 28.

After completing the SDS and SPM, two additional questionnaires were administered to all participants. First, the eight items of the Core Depressive Factor were readministered. Second, the participants were asked to rate the accuracy of the SPM (“How accurately do you think the test assessed your actual reasoning capacity?”) on a 5-point scale ranging from 1 (“cannot detect”) to 5 (“can detect”). When they had completed these items, participants informed the experimenter that they had finished. They were then debriefed, thanked, and excused.

Data Analysis

The collected data were analyzed using SPSS Version 21.0. Independent-samples t -tests were conducted to compare differences between the depression and control groups for age and the expected and actual numbers of SPM problems that participants solved. Group differences in gender and educational background were assessed with chi-square tests. Paired-samples t -tests were performed to compare pre- and post-test Core Depressive Factor scores and to explore the trends in depressed emotion. Analysis of variance (ANOVA) was performed to explore the effect of participant type (depressive or healthy control) and feedback condition (success or failure) on emotional reactivity. Finally, simple effects tests were performed when the interaction terms were significant.

Descriptive Statistics

Independent-samples t -tests revealed no significant difference between the two groups in age or the actual number of problems solved. Regarding the expected number, the healthy group predicted that they would solve more problems than the patients with depression ( M = 8.627, M = 7.648; t = −3.352, p < 0.05, Cohen’s d = 0.597) ( Table 2 ). Regarding the Core Depressive Factor scores at baseline, the depression group scored significantly higher than the controls ( M = 21.521, M = 13.814; t = 13.331, p < 0.01, Cohen’s d = 2.39). There were no group differences in gender or educational background.

Table 2. Core Depressive Factor scores and the actual and expected numbers of SPM problems solved at baseline.

Depressed Emotion Results

Differences between pre- and post-test in depressed emotions.

For the depression group, the pre- and post-test depressed emotion scores (i.e., Core Depressive Factor scores) did not differ, regardless of the feedback condition ( t = −0.875, p = 0.387; t = 1.408, p = 0.169), indicating that the performance feedback valence had little effect in subjects with MDD. For the healthy control group, pre- and post-test depressed emotion scores did not differ in the success feedback condition ( t = 1.233, p = 0.231), but there was a higher post-test score in the failure condition ( M = 13.321, M = 14.571; t = −3.35, p < 0.011, Cohen’s d = 0.44). This suggests that increased depressed emotions in the control group were only provoked by failure feedback.

ANOVA of Depressed Emotion

Table 3 shows that depressed emotion was potentially influenced by participant type and feedback valence. Therefore, we performed a 2 (groups: depression, healthy) × 2 (conditions: success, failure) ANOVA on depressed emotion, and the difference between pre- and post-test scores was used to compute the effect on depressed emotion. The analysis revealed a non-significant main effect for group [ F (1,126) = 1.599, p = 0.208] and a non-significant main effect for condition [ F (1,126) = 1.945, p = 0.166]. However, there was a significant group × condition interaction effect [ F (1,126) = 11.952, p < 0.01, Cohen’s f = 0.31]. Taken together, these results indicate that the differential effect of feedback condition on depressed emotion also depended on the levels of a second variable, namely, the group.

Table 3. Pre- and post-test depressed emotion scores and SPM accuracy ratings.

Simple Effects Tests for Depressed Emotion

Given the significant group × condition interaction, we examined two simple effects of condition: the effect of condition for the depressed group and the effect of condition for the control group. As the results of depressed emotion scores across feedback conditions within each group showed, no difference was observed between the failure and success feedback conditions for depressed individuals [ F (1,126) = 2.591, p = 0.112]. In contrast, for the controls, the amplitude of emotional fluctuation was larger in the failure feedback condition ( M = −1.250) compared to the success condition [ M = 0.516; F (1,126) = 9.652, p < 0.01, Cohen’s f = 0.38]. Similarly, two simple effects of group were examined: the effect of group for the success condition and the effect of group for the failure condition. The results of depressed emotion scores between the two groups within each feedback condition revealed no significant difference between the depression and control groups in the success feedback condition [ F (1,126) = 2.341, p = 0.131]. However, in the failure condition, the amplitude of the emotional fluctuation of depressed participants ( M = 0.469) was smaller than that of the controls [ M = −1.25; F (1,126) = 11.889, p < 0.01, Cohen’s f = 0.37] ( Table 4 ). Overall, these findings suggest that participant type may have affected the feedback-related differences in depressed emotion.

Table 4. ANOVA of Core Depressive Factor scores.

Results of Raven’s Standard Progressive Matrices Accuracy Ratings

Anova of raven’s standard progressive matrices accuracy ratings.

To investigate the impacts of participant type and feedback condition on causal attribution, we performed ANOVA on the SPM accuracy ratings and observed a non-significant main effect for participant type. However, there was a substantial main effect for feedback condition [ F (1,126) = 6.051, p < 0.05, Cohen’s f = 0.22] and a significant group × condition interaction effect [F (1,126) = 5.208, p < 0.05, Cohen’s f = 0.20] ( Table 5 ). This suggests that the differential effect of feedback condition on the perceived SPM accuracy was also affected by participant type.

Table 5. ANOVA of the SPM accuracy ratings.

Simple Effects Tests of Raven’s Standard Progressive Matrices Accuracy Ratings

Given the significant interaction effect, further simple effects tests were performed. Firstly, the changes of perceived SPM accuracy rating across feedback conditions within each participant group was used to examine the simple effect of feedback condition. The results showed that there was no significant difference in perceived SPM accuracy ratings among subjects with depression across feedback conditions [ F (1,126) = 0.014, p = 0.017]. Conversely, perceived accuracy ratings were higher in the control group for the success condition ( M = 3.355) than for the failure condition [ M = 2.607; F (1,126) = 10.709, p < 0.01, Cohen’s f = 0.40]. The differences in perceived SPM accuracy rating between the two groups within each feedback condition was then used to examine the simple effect of group. The results showed that there was no significant difference between the depression and control groups in the failure condition. In the success condition, however, healthy controls reported higher SPM accuracy ratings ( M = 3.355) than MDD subjects [ M = 2.872; F (1,126) = 4.459, p < 0.01, Cohen’s f = 0.25].

Comparing individuals with MDD and healthy individuals at baseline, we found that the depressed emotion scores and number of SPM problems that participants expected to solve were significantly different between the two groups. Before the SPM was administered, the number of problems participants expected to solve was lower for subjects with MDD than for controls, but the actual performance was similar between groups. To rule out the possibility that education level affected the correct number of SPM problems, chi-square tests were performed to examine the between-group difference in educational background, and the tests revealed no significant difference among the three levels (junior high school and below; senior high school; undergraduate and above). In general, depressed individuals exhibited increased negative expectations regarding their personal futures compared to controls, which could be considered as hopelessness. Macleod and Salaminiou (2001) administered the Future-Thinking Task and found that depressed patients reported decreased positive expectations and fewer enjoyable experiences than healthy participants. Further, in line with the clinical manifestations of MDD (prominent and persistent low moods), the results confirmed a higher level of depressed emotion in the depression group compared to the control group.

ANOVA of depressed emotion scores demonstrated a significant group × condition interaction. After performing additional simple effects tests, we found in the healthy control group that the amplitude of emotional fluctuation following failure feedback increased drastically compared to that in response to success feedback. This could be explained by the cognitive dissonance hypothesis, as formulated by Leon Festinger, who proposed that cognitive dissonance causes psychological stress: cognitive dissonance refers to the existential inconsistency between two contradictory beliefs about social reality, or the contradiction between a person’s belief and an action s/he has taken ( Brehm, 2010 ). Thus, failure feedback given by others was largely inconsistent with the self-view held by the controls, which triggered a stronger mood swing. Paired-samples t -tests showed that depressed emotion increased in the control group in response to failure feedback, though not to the level seen in subjects with MDD ( Figure 1 ). This result supports the cognitive schema aspect of mood-congruent theory ( Beck, 1967 ; Bower, 1981 ) and is also in line with the results reported by Zhang and Tian (2005) , who performed a failure feedback experiment with 129 college students in China and found that failure feedback was effective in eliciting negative emotions (i.e., anxiety or depression). However, the results for the success feedback condition did not support our hypothesis. The reason for this discrepancy could be our measure of depressed emotion, which mainly reflected changes in negative emotions but not positive ones. Affective science, proposed by the psychologists Barrett and Russell (1998) , holds that the positive and negative valences of emotion are independent of each other and are characterized by drastic bipolarity in experience and expression. Empirical research conducted by Warr et al. (1983) showed that the frequency of engaging in desirable life activities is related to increased positive emotion but not negative emotion. More recently, Ellis et al. (2009) suggested that healthy individuals reported increased positive affect after success feedback, all of which leads us to speculate that success feedback might have a greater influence on the regulation of positive emotion than on negative emotion regulation. This greater influence might explain why the depressed emotion levels of the controls were similar before and after the test in the positive feedback condition. Further investigations are required to gather more evidence about positive emotion changes in response to different valences in social feedback.

Figure 1. Mean depressed emotion scores by participant type and feedback condition.

Among subjects with MDD, the amplitude of emotional fluctuation did not differ between the failure and success feedback conditions ( Figure 1 ), favoring our prior assumption that depressed emotion in MDD subjects would not change after any type of feedback, and also favoring the ECI theory, which suggests that both positive and negative stimuli elicit reduced emotional responses in MDD patients. The finding is largely consistent with the common phenomenon experienced by MDD patients where perceptions of the world are flat and dull, and patients maintain that “everything is the same” ( Healy and Williams, 1988 ). Empirical research conducted by Dichter et al. (2010) reported that, compared with controls, depressed individuals did not have a significant change in attitude when they viewed affective images of different valences. Additionally, a similar research in Singapore also found no significant difference between success and failure conditions among depressed patients, even with an alternative instrument of measurement ( Yeo et al., 2017 ). Specifically, participants were asked to complete a digit-span memory task, and the Positive and Negative Affect Schedule (PANAS) was used to assess changes in negative emotion. Two consistent results predicted that the insensitivity of MDD to evaluative information from external social networks might not be associated with the instrument of measurement.

In light of these considerations, we propose that patients with MDD show blunted emotional reactivity to new stimuli with positive or negative valence. This could further explain why symptom alleviation is difficult in MDD patients experiencing an acute episode, even when they are encouraged by friends and family members around them. To optimize the therapeutic effects for patients with depressive disorders, medication is generally the first-line treatment for MDD, especially for severe cases involving suicidal thoughts. Once a patient’s condition is improved, psychological counseling or psychotherapy can be introduced ( Morin et al., 2009 ). Stressful life events have been well recognized in previous research as precipitants of major depressive episodes. However, the presence or absence of adverse events could not provide a useful guide to the prognosis or treatment of depression. Therefore, we speculated that social feedback based on success or failure performance might have a more substantial effect on depressed patients in remission, which was confirmed by the control group in the current study. Furthermore, some psychotherapy modalities, such as supportive psychotherapy and cognitive behavioral therapy (CBT), could be more useful for patients in partial or full remission.

According to cognitive theories of emotion, emotional reactivity to success or failure outcomes is influenced by a person’s views about the causes of their performance ( Weiner, 1980 ). We required participants to rate the accuracy of the SPM in assessing reasoning ability on a five-point scale after receiving feedback (success or failure). If the rating was high following success feedback, it meant that participants thought that the SPM test was an accurate measure of their reasoning capacity, indicating that they made internal attributions for a successful outcome. If the rating was low after failure feedback, it meant that they did not think that the SPM test was valid when assessing their reasoning ability and thus made an external attribution for the failure. Miller and Ross (1976) defined the self-serving attribution bias as the tendency to give credit to ourselves for success but attribute failure to external sources ( Zuckerman, 2010 ).

The differences in self-serving attribution bias between the depression and control groups across feedback conditions were analyzed with ANOVA, and the result revealed a significant group × condition interaction effect. Using a simple effects test, we noted that SPM accuracy rating was higher following success feedback than that following failure feedback in the control group ( Figure 2 ), indicating that healthy individuals thought that the test provided a more accurate assessment of their ability when they performed well. This result is in line with our hypothesis that self-serving attribution bias is commonly held by the general public; the result is also consistent with much of the existing literature ( Streufert and Streufert, 1969 ; Miller, 1976 ). However, a meta-analysis of cross-cultural studies revealed that Westerners showed a clear self-serving bias, whereas East Asians of a relatively more collectivistic nature did not ( Heine and Hamamura, 2007 ). The current study found that the self-serving attribution bias, which is said to be independent of collectivist culture, also occurs among the Chinese.

Figure 2. Mean Raven’s Standard Progressive Matrices (SPM) accuracy ratings by participant type and feedback condition.

In the depression group, the SPM accuracy ratings were not significantly different between the two feedback conditions, suggesting that patients with MDD regarded success and failure as equally diagnostic of their ability level. Moreover, their causal attribution lacked a self-serving bias, unlike controls, which is in line with Kuiper’s findings ( Kuiper, 1978 ). The learned helplessness model can be used to explain the attribution style of the MDD patients, characterized by a lack of difference in internal–external causal attributions in response to changes in environmental contingencies ( Seligman, 1975 ). Thus, insensitivity to various environmental events was evident both in attribution style and emotional reactivity. Research has reported a strong correlation between causal attribution and depressive disorder ( Lau and Eley, 2008 ; Rueger and Malecki, 2011 ), but how depressed patients attribute the success and failure to specific kinds of control locus (internal and external) remains unclear in the context of collectivist culture, and further research is required. A study by Fu and Chen (2010) in China suggested that individuals with depression also exhibited a self-serving bias in causal attribution, tending to take personal credit for success and blaming the external environment for failure. However, Yeo et al. (2017) proposed that patients with MDD commonly exhibit a reverse self-serving bias in causal attribution, as negative events were usually ascribed to internal, stable, and global causes. In other words, subjects with depression blamed themselves for failure but did not praise themselves for success.

Future research should also aim to rectify two major shortcomings in the current study. The first is the use of the Core Depressive Factor, which might be related to the emotional insensitivity of depressed patients to the performance-based feedback. The Core Depressive Factor has been labeled as “the affective symptoms” ( Kitamura et al., 2004 ) and the “general depression”( American Psychiatric Association, 2013 ). It was used to assess the depressed emotions of patients by most researches and was hardly distinguished from emotional symptoms of depressive disorders. However, Passik et al. (2000) suggested that the emotional symptoms reflected in the core factor might be more variable and transient, and therefore less indicative of a depressive illness. Even in the case of suicidal ideation, many people express such thoughts as an indication of frustration or to “blow off steam” rather than as a genuine desire to die. Thus, to some extent, the Core Depressive Factor could be used to assess negative emotion with enriched content in healthy individuals, rather than levels of depression only. Therefore, future studies could combine the different scales. Additionally, objective physiological indicators, such as heart rate and skin electricity, could be added in order to confirm these findings.

The second shortcoming is that we did not assess the change in positive emotion after the performance appraisal. The control participants showed no difference in depressed emotions after receiving success feedback, which did not support our hypothesis. However, a previous study by Ellis et al. (2009) reported increased positive reactions in a healthy group following success feedback. We speculate therefore that positive feedback might be more effective in influencing positive emotion in healthy individuals, rather than negative emotions. Further investigations are required to gather more evidence about positive emotion changes in response to different valences in social feedback.

In summary, this study has investigated the effects of performance feedback on MDD patients and healthy participants. First, the MDD patients showed little change in depressed emotions in response to both types of feedback, supporting Rottenberg et al.’s (2005) theory of ECI. Second, healthy individuals described enhanced depressed emotion in response to failure feedback, supporting the mood-congruent theory. There was no change in response to success feedback possibly because success feedback may be more effective at influencing positive rather than negative emotions. Third, self-serving attribution bias was not evident in the MDD group, supporting Seligman’s theory of learned helplessness. In contrast, self-serving attribution bias was evident in the healthy control group.

Data Availability Statement

The datasets generated for this study are available on request to the corresponding author.

Ethics Statement

The studies involving human participants were reviewed and approved by the Ethics Committee of Second Xiangya Hospital, Central South University. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

Author Contributions

TC and XL designed the whole study. HG and QC conceived the study. DZ, QC, and YC collected the data. HG analyzed the data and wrote the first draft of the manuscript. All authors commented on the successive drafts, contributed to the interpretation of the results, and approved the final version of the manuscript.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

The authors gratefully acknowledge each participant. In addition, HG would like to thank the inimitable care and spiritual support of Bangpei Zhang over the years, she said she loved him very much.

American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental Disorders (DSM-5). Virginia: American Psychiatric Association.

Google Scholar

Barrett, L. F., and Russell, J. A. (1998). Independence and bipolarity in the structure of current affect. J. Pers. Soc. Psychol. 74, 967–984. doi: 10.1037/0022-3514.74.4.967

CrossRef Full Text | Google Scholar

Beck, A. T. (1967). Depression. Clinical, experimental and theoretical aspects. JAMA J. Am. Med. Assoc. 203:1144. doi: 10.1001/jama.1968.03140130056023

Bora, E., and Berk, M. (2016). Theory of mind in major depressive disorder: a meta-analysis. J. Affect. Disord. 191, 49–55. doi: 10.1016/j.jad.2015.11.023

PubMed Abstract | CrossRef Full Text | Google Scholar

Bower, G. H. (1981). Mood and memory. Am. Psychol. 36, 129–148.

Brehm, J. W. (2010). A brief history of dissonance theory. Soc. Pers. Psychol. Compass 1, 381–391. doi: 10.1111/j.1751-9004.2007.00035.x

Brinkmann, C., and Brixius, K. (2017). Erratum to: peroxiredoxins and sports: new insights on the antioxidative defense. J. Physiol. Sci. 67:637. doi: 10.1007/s12576-017-0551-y

Cai, H. J., and Yang, Z. L. (2003). The stability of implicit self-esteem-the effect of the manipulation of success and failure. Psychol. Sci. 26, 461–464.

Clark, L. A., Watson, D., and Mineka, S. (1994). Temperament, personality, and the mood and anxiety disorders. J. Abnorm. Psychol. 103, 103–116. doi: 10.1037/0021-843x.103.1.103

Dichter, G. S., Tomarken, A. J., Shelton, R. C., and Sutton, S. K. (2010). Early- and late-onset startle modulation in unipolar depression. Psychophysiology 41, 433–440. doi: 10.1111/j.1469-8986.00162.x

Duan, Q. Q. (2012). Differential validity of SAS and SDS among psychiatric non-psychotic outpatients and their partners. Chin. Mental Health J. 26: 676–679.

Dutton, K. A., and Brown, J. D. (1997). Global self-esteem and specific self-views as determinants of people’s reactions to success and failure. J. Pers. Soc. Psychol. 73, 139–148. doi: 10.1037/0022-3514.73.1.139

Ellis, A. J., Beevers, C. G., and Wells, T. T. (2009). Emotional dysregulation in dysphoria: support for Emotion Context Insensitivity in response to performance-based feedback. J. Behav. Ther. Exp. Psychiatry 40, 443–454. doi: 10.1016/j.jbtep.2009.05.002

Fu, C. J., and Chen, L. (2010). The effects of success or failure manipulation on objective setting in dysphoria. Chin. J. Behav. Med. Brain Sci. 19, 497–498.

Guo, J., Hou-Chao, L. V., Huang, X. T., and Chen, X. J. (2011). Review and prospect on the research of self-serving Bias. Adv. Psychol. Sci. 12:7.

Healy, D., and Williams, J. M. (1988). Dysrhythmia, dysphoria, and depression: the interaction of learned helplessness and circadian dysrhythmia in the pathogenesis of depression. Psychol. Bull. 103, 163–178. doi: 10.1037/0033-2909.103.2.163

Heine, S. J., and Hamamura, T. (2007). In search of east asian self-enhancement. Pers. Soc. Psychol. Rev. 11, 4–27. doi: 10.1177/1088868306294587

Jonas, E., Wouter, D., and Koster, E. H. W. (2014). Attention, interpretation, and memory biases in subclinical depression: a proof-of-principle test of the combined cognitive biases hypothesis. Emotion 14, 331–340. doi: 10.1037/a0035250

Kayikcioglu, O., Bilgin, S., and Uyar, M. (2016). Review on the eyedrop self-instillation techniques and factors affecting these techniques in Glaucoma patients. Scientifica 2016, 1–5. doi: 10.1155/2016/9183272

Kitamura, T., Hirano, H., Chen, Z., and Hirata, M. (2004). Factor structure of the Zung Self-rating Depression Scale in first-year university students in Japan. Psychiatry Res. 128, 0–287.

PubMed Abstract | Google Scholar

Kuiper, N. A. (1978). Depression and causal attributions for success and failure. J. Pers. Soc. Psychol. 36, 236–246. doi: 10.1037/0022-3514.36.3.236

Kupferberg, A., Bicks, L., and Hasler, G. (2016). Social functioning in major depressive disorder. Neurosci. Biobehav. Rev. 69, 313–332. doi: 10.1016/j.neubiorev.2016.07.002

Lau, J. Y., and Eley, T. C. (2008). Attributional style as a risk marker of genetic effects for adolescent depressive symptoms. J. Abnorm. Psychol. 117, 849–859. doi: 10.1037/a0013943

Linville, P. W. (1985). Self-complexity and affective extremity: don’t put all of your eggs in one cognitive basket. Soc. Cogn. 3, 94–120. doi: 10.1521/soco.1985.3.1.94

Macleod, A. K., and Salaminiou, E. (2001). Reduced positive future-thinking in depression: cognitive and affective factors. Cogn. Emot. 15, 99–107. doi: 10.1080/0269993004200006

Miller, D. T. (1976). Ego Involvement and Attributions for Success and Failure. J. Pers. Soc. Psychol. 34, 901–906. doi: 10.1037/0022-3514.34.5.901

Miller, D. T., and Ross, M. (1976). Self-serving biases in the attribution of causality: fact or fiction? Psychol. Bull. 82, 213–225. doi: 10.1037/h0076486

Morin, C. M., Valli, A., Guay, B., Ivers, H., Savard, J., Rette, C. M., et al. (2009). Cognitive behavioral therapy, singly and combined with medication, for persistent insomnia: a randomized controlled trial. JAMA 301, 2005–2015.

Morris, S. J. (2010). Attributional biases in subclinical depression: a schema-based account. Clin. Psychol. Psychother. 14, 32–47. doi: 10.1002/cpp.512

Natoli, A. P., Nelson, S. M., Lengu, K. J., and Huprich, S. K. (2016). Sensitivity to criticism differentially mediates the relationship between interpersonal problems and state and trait depression. Pers. Mental Health 10, 293–304. doi: 10.1002/pmh.1338

Passik, S. D., Lundberg, J. C., Rosenfeld, B., Kirsh, K. L., Donaghy, K., Theobald, D., et al. (2000). Factor analysis of the zung self-rating depression scale in a large ambulatory oncology sample. Psychosomatics 41, 121–127. doi: 10.1176/appi.psy.41.2.121

Raven, J. (1983). “The progressive matrices and mill hill vocabulary scale in western societies,” in Human Assessment and Cultural Factors. NATO Conference Series (III Human Factors) eds S. H. Irvine and J. W. Berry (Boston, MA: Springer).

Romera, I., Delgado-Cohen, H., Perez, T., Caballero, L., and Gilaberte, I. (2008). Factor analysis of the Zung self-rating depression scale in a large sample of patients with major depressive disorder in primary care. BMC Psychiatry 8:4–4. doi: 10.1186/1471-244X-8-4

Rottenberg, J. (2007). “Major depressive disorder: emerging evidence for emotion context insensitivity,” in Emotion and Psychopathology: Bridging Affective and Clinical Science , eds J. Rottenberg and S. L. Johnson (Washington, DC: American Psychological Association).

Rottenberg, J., Gross, J. J., and Gotlib, I. H. (2005). Emotion context insensitivity in major depressive disorder. J. Abnorm. Psychol. 114, 627–639. doi: 10.1037/0021-843x.114.4.627

Rueger, S. Y., and Malecki, C. K. (2011). Effects of stress, attributional style and perceived parental support on depressive symptoms in early adolescence: a prospective analysis. J. Clin. Child Adolesc. Psychol. 40, 347–359. doi: 10.1080/15374416.2011.563461

Ruff, C. C., and Ernst, F. (2014). The neurobiology of rewards and values in social decision making. Nat. Rev. Neurosci. 15, 549–562. doi: 10.1038/nrn3776

Rygula, R., and Popik, P. (2016). Trait “pessimism” is associated with increased sensitivity to negative feedback in rats. Cogn. Affect. Behav. Neurosci. 16, 516–526. doi: 10.3758/s13415-016-0410-y

Sakamoto, S., Kijima, N., Tomoda, A., and Kambara, M. (1998). Factor structures of the zung self- rating depression scale (SDS) for undergraduates. J. Clin. Psychol. 54, 477–487. doi: 10.1002/(sici)1097-4679(199806)54:4<477::aid-jclp9>3.0.co;2-k

Seligman, M. E. P. (1975). Helplessness: On Depression, Development, and Death. New York, NY: W. H. Freeman.

Sheehan, D. V., Sheehan, K. H., Douglas, S., Juris, J., Yvonne, B., Rogers, J. E., et al. (2010). Reliability and validity of the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID). Journal of Clinical Psychiatry 71, 313–326. doi: 10.4088/jcp.09m05305whi

Steele, J. D., Kumar, P., and Ebmeier, A. K. P. (2007). Blunted response to feedback information in depressive illness. Brain 130, 2367–2374. doi: 10.1093/brain/awm150

Streufert, S., and Streufert, S. C. (1969). Effects of conceptual structure, failure, and success on attribution of causality and interpersonal attitudes. J. Pers. Soc. Psychol. 9, 138–147. doi: 10.1037/h0027047

Warr, P. B., Barter, J., and Brownbridge, G. (1983). On the independence of positive and negative affect. J. Pers. Soc. Psychol. 44, 644–651.

Weiner, B. (1980). A cognitive (attribution)-emotion-action model of motivated behavior: an analysis of judgments of help-giving. J. Pers. Soc. Psychol. 39, 186–200. doi: 10.1037/0022-3514.39.2.186

Xi, M. J., Zhang, Y. J., Li, Y. X., and Yan, K. L. (2007). Implicit self-esteem level and stability of depressive patients. Chin. Mental Health J. 21, 756–767.

Yeo, S. N., Zainal, H., Tang, C. S., Tong, E. M., Ho, C. S., and Ho, R. C. (2017). Success/failure condition influences attribution of control, negative affect, and shame among patients with depression in Singapore. BMC Psychiatry 17:285. doi: 10.1186/s12888-017-1451-7

Zhang, X., and Tian, L. (2005). The buffering effect of self-esteem on the depressive and anxious reactions to induced failure. Acta Psychol. Sin. 37, 241–245.

Zuckerman, M. (2010). Attribution of success and failure revisited: or the motivational bias is alive and well in attribution theory. J. Pers. 47, 245–287. doi: 10.1111/j.1467-6494.1979.tb00202.x

Keywords : emotional reaction, attribution, depression, success feedback, failure feedback

Citation: Gao H, Che Q, Zhang D, Chai Y, Luo X and Cai T (2020) Insensitivity to Success and Failure: An Experimental Study of Performance-Based Feedback in Depression. Front. Psychol. 11:670. doi: 10.3389/fpsyg.2020.00670

Received: 05 July 2019; Accepted: 19 March 2020; Published: 23 April 2020.

Reviewed by:

Copyright © 2020 Gao, Che, Zhang, Chai, Luo and Cai. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Xingwei Luo, [email protected] ; Taisheng Cai, [email protected]

† These authors have contributed equally to this work and share first authorship

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• Current issue
• BMJ Journals

You are here

• Volume 72, Issue 3
• Rumination and executive function in depression: an experimental study
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• E Watkins ,
• Department of Psychology, Institute of Psychiatry, Kings College, London, UK
• Correspondence to:
 Dr E Watkins, Department of Psychology, Institute of Psychiatry, De Crespingny Park, Denmark Hill, London SE5 8AF, UK

Background: Major depression is associated with cognitive deficits, particularly those requiring central executive functioning. Depressed patients also tend to focus on and think about their symptoms and problems (“ruminate”) more than non-depressed controls. Although an association has been found between rumination and impaired performance on a central executive processing task, the causal relation between impaired executive functioning and rumination has not been determined. This study sought to directly manipulate rumination and assess the impact on executive functioning in depression as measured by random number generation.

Methods: Depressed patients (n=14) and non-depressed controls (n=14) were compared on a random number generation task, performed after both a rumination induction and after a distraction induction, with order of inductions counter balanced within each group.

Results: Compared with the distraction induction, the rumination induction produced a significant increase in both ruminations and the tendency towards stereotyped counting responses (thought to reflect a failure of inhibitory executive control) in depressed patients but not in controls. However, after distraction, no difference in random number generation or rumination was found between the two groups.

Conclusions: The aspects of executive function involved in random number generation are not fundamentally impaired in depressed patients. In depressed patients, the rumination induction seems to trigger the continued generation of ruminative stimulus independent thoughts, which interferes with concurrent executive processing.

• executive function
• WCST, Wisconsin card sorting test
• SITs, stimulus independent thoughts
• BDI, Beck depression inventory
• RRS, ruminative responses scale
• RTF, ruminative thought form

https://doi.org/10.1136/jnnp.72.3.400

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Read the full text or download the PDF:

Experimental depression treatment is nearly 80% effective in controlled study

By Mandy Erickson

A new type of magnetic brain stimulation brought rapid remission to almost 80% of participants with severe depression in a study conducted at the  Stanford University School of Medicine .

The treatment, known as Stanford accelerated intelligent neuromodulation therapy (SAINT) or simply Stanford neuromodulation therapy, is an intensive, individualized form of transcranial magnetic stimulation. In the study, remission typically occurred within days and lasted months. The only side effects were temporary fatigue and headaches.

“It works well, it works quickly and it’s noninvasive,” said  Nolan Williams , MD, an assistant professor of psychiatry and behavioral sciences, and co-director of the Koret Human Neurosciences Community Laboratory at the Wu Tsai Neurosciences Institute . “It could be a game changer.” Williams is the senior author of the study, which was  published  Oct. 29 in the  American Journal of Psychiatry .

Twenty-nine people with treatment-resistant depression participated in the study: About half received SAINT, and the rest underwent a placebo procedure that mimicked the real treatment. After five days of treatment, 78.6% of the participants in the treatment group were no longer depressed, according to several standard methods of evaluation. “It’s quite a dramatic effect, and it’s quite sustained,” said  Alan Schatzberg , MD, the Kenneth T. Norris, Jr. Professor in Psychiatry and Behavioral Sciences, who was a co-author of the study.

A lifetime of depression

Tommy Van Brocklin, 60, has suffered from depression since he was 15. “In 1975, they didn’t have the medication and understanding they do now,” he said. “I was told I wasn’t trying hard enough.”

“I’ve functioned all these years, but it’s been very difficult at times,” the civil engineer added. Talk therapy helped “for about half a day after an appointment.” When selective serotonin reuptake inhibitors became available in the 1990s, he started on paroxetine, commonly sold under the brand name Paxil.

“It worked like a miracle drug,” he said, but after 10 or 15 years it started to lose its effect. After 25 years, it stopped working entirely. He tried other medications, but none helped; one even made him suicidal. 

His sister, who lives near Stanford, connected him with the researchers studying SAINT. He flew from his home in Memphis, Tennessee, and underwent the treatment in September. He felt nothing the first day; on day two, he began feeling emotional — “I felt the struggle of what I’d been through all these years.”

“The next day, all of a sudden, it broke through,” he said. “I felt so much better, and it’s stuck with me.”

Specialized magnetic stimulation

The transcranial magnetic stimulation treatment currently approved by the Food and Drug Administration requires six weeks of once-daily sessions. Only about half of patients who undergo the treatment improve, and only about a third experience remission from depression.

SAINT advances that treatment by targeting the magnetic pulses according to each patient’s neurocircuitry and providing a greater number of pulses at a faster pace.

In the study, the researchers first used MRI to locate the best location to target within each participant’s dorsolateral prefrontal cortex, which regulates executive functions, such as problem solving and inhibiting unwanted responses. They applied the stimulation in a subregion that has the strongest relationship with the subgenual cingulate, a part of the brain that is overactive in people experiencing depression. The transcranial magnetic stimulation strengthens the connection between the two regions, facilitating dorsolateral prefrontal cortex control of the activity in the subgenual cingulate.

The researchers also used 1,800 pulses per session instead of 600. (The larger amount has been used safely in other forms of brain stimulation for neurological disorders such as Parkinson’s disease.) And instead of providing one treatment a day, they gave participants 10 10-minute treatments, with 50-minute breaks in between.

For the control group, the researchers disguised the treatment with a magnetic coil that mimicked the experience of the magnetic pulse; both the control and active treatment groups wore noise-canceling earphones and received a topical ointment to dull sensation. Neither the researcher administering the procedure nor the participant knew whether the participant was receiving real treatment.

A hard-to-treat group

The trial participants ranged in age from 22 to 80; on average, they had suffered depression for nine years. They had tried medications, but either they had had no effect or they had stopped working. During the trial, participants who were on medication maintained their regular dosage; participants who weren’t taking medications did not start any.

Within four weeks after treatment, 12 of the 14 participants who had received the treatment improved, and 11 of them met FDA criteria for remission. In contrast, only two of the 15 participants who had received the placebo met the criteria for remission.

Because the study participants typically felt better within days of starting SAINT, the researchers are hoping it can be used to quickly treat patients who are at a crisis point. Patients who start taking medication for depression typically don’t experience any reduction of symptoms for a month.

“We want to get this into emergency departments and psychiatric wards where we can treat people who are in a psychiatric emergency,” Williams said. “The period right after hospitalization is when there’s the highest risk of suicide.”

Van Brocklin said that since he returned home following treatment, he’s made some radical changes. “I have a really strong desire to get my life together,” he said.

“I don’t procrastinate anymore,” he added. “I’m sleeping better. I completely quit alcohol. I’m walking my dog and playing the guitar again, for nothing more than the sheer joy of it.”

Most importantly, he said, “I’m remaining positive and being respectful of others. These are big changes in my life.”

Other Stanford scientists who contributed to the study are former postdoctoral scholars Eleanor Cole, PhD, and Angela Phillips, PhD; Brandon Bentzley, MD, PhD, David Carreon, MD, Jennifer Keller, PhD, Kristin Raj, MD, and Flint Espil, PhD, all clinical assistant professors of psychiatry and behavioral sciences; clinical research coordinators Katy Stimpson, Romina Nejad, Clive Veerapal, Nicole Odenwald and Maureen Chang; former clinical research coordinators Fahim Barmak, MD, Naushaba Khan and Rachel Rapier; postdoctoral scholars Kirsten Cherian, PhD, James Bishop, PhD, Azeezat Azeez, PhD, and John Coetzee, PhD; life science research professional Heather Pankow; clinical research manager Jessica Hawkins; Charles DeBattista, MD, professor of psychiatry and behavioral sciences; and Booil Jo, PhD, associate professor of psychiatry and behavioral sciences.

Scientists from the U.S. Department of Veterans Affairs; Palo Alto University; the Centre for Neuroimaging and Cognitive Genomics at the National University of Ireland; and the School of Medicine at Southern Illinois University, Carbondale, contributed to the research.

The research was funded by a Brain and Behavior Research Foundation Young Investigator Award, Charles R. Schwab, the David and Amanda Chao Fund II, the Amy Roth PhD Fund, the Neuromodulation Research Fund, the Lehman Family, the Still Charitable Trust, the Marshall and Dee Ann Payne Fund, and the Gordie Brookstone Fund.

Stanford’s  Department of Psychiatry and Behavioral Sciences  also contributed to the work.

• See us on facebook
• See us on twitter
• See us on youtube
• See us on linkedin
• See us on instagram

Stanford researchers devise treatment that relieved depression in 90% of participants in small study

Stanford Medicine researchers used high doses of magnetic stimulation, delivered on an accelerated timeline and targeted to individual neurocircuitry, to treat patients with severe depression.

April 6, 2020 - By Mandy Erickson

Deirdre Lehman, who suffered from depression, with Nolan Williams, who oversaw a clinical trial of a potential treatment that uses transcranial magnetic stimulation. In this photo, Williams and Lehman demonstrate how a patient is positioned and the equipment is used during the treatment.   Steve Fisch

A new form of magnetic brain stimulation rapidly relieved symptoms of severe depression in 90% of participants in a small study conducted by researchers at the Stanford University School of Medicine .

The researchers are conducting a larger, double-blinded trial in which half the participants are receiving fake treatment. The researchers are optimistic the second trial will prove to be similarly effective in treating people whose condition hasn’t improved with medication, talk therapy or other forms of electromagnetic stimulation.

The treatment is called Stanford Accelerated Intelligent Neuromodulation Therapy, or SAINT. It is a form of transcranial magnetic stimulation, which is approved by the Food and Drug Administration for treatment of depression. The researchers reported that the therapy improves on current FDA-approved protocols by increasing the number of magnetic pulses, speeding up the pace of the treatment and targeting the pulses according to each individual’s neurocircuitry.

Before undergoing the therapy, all 21 study participants were severely depressed, according to several diagnostic tests for depression. Afterward, 19 of them scored within the nondepressed range. Although all of the participants had suicidal thoughts before the therapy, none of them reported having suicidal thoughts after treatment. All 21 participants had previously not experienced improvements with medications, FDA-approved transcranial magnetic stimulation or electroconvulsive therapy.

The only side effects of the new therapy were fatigue and some discomfort during treatment, the study reported. The results were published online April 6 in the  American Journal of Psychiatry.

“There’s never been a therapy for treatment-resistant depression that’s broken 55% remission rates in open-label testing,” said Nolan Williams , MD, assistant professor of psychiatry and behavioral sciences and a senior author of the study. “Electroconvulsive therapy is thought to be the gold standard, but it has only an average 48% remission rate in treatment-resistant depression. No one expected these kinds of results.”

Calming the brain chatter

When Deirdre Lehman, 60, woke up the morning of June 30, 2018, she said she was hit by “a tsunami of darkness.” Lehman had struggled with bipolar disorder all her adult life, but with medications and psychotherapy her mood had been stable for 15 years.

After undergoing the experimental treatment, Lehman's depression dissipated. She has since completed a bachelor’s degree at the University of California-Santa Barbara, where she had dropped out as a young woman when her bipolar symptoms overwhelmed her studies. Steve Fisch

“There was a constant chattering in my brain: It was my own voice talking about depression, agony, hopelessness,” she said. “I told my husband, ‘I’m going down and I’m heading toward suicide.’ There seemed to be no other option.”

Lehman’s psychiatrist had heard of the SAINT study and referred her to Stanford. After researchers pinpointed the spot in her brain that would benefit from stimulation, Lehman underwent the therapy. 

“By the third round, the chatter started to ease,” she said. “By lunch, I could look my husband in the eye. With each session, the chatter got less and less until it was completely quiet.

“That was the most peace there’s been in my brain since I was 16 and started down the path to bipolar disorder.”

In transcranial magnetic stimulation, electric currents from a magnetic coil placed on the scalp excite a region of the brain implicated in depression. The treatment, as approved by the FDA, requires six weeks of once-daily sessions. Only about half of patients who undergo this treatment improve, and only about a third experience remission from depression.

Stanford researchers hypothesized that some modifications to transcranial magnetic stimulation could improve its effectiveness. Studies had suggested that a stronger dose, of 1,800 pulses per session instead of 600, would be more effective. The researchers were cautiously optimistic of the safety of the treatment, as that dose of stimulation had been used without harm in other forms of brain stimulation for neurological disorders, such as Parkinson’s disease.

Other studies suggested that accelerating the treatment would help relieve patients’ depression more rapidly. With SAINT, study participants underwent 10 sessions per day of 10-minute treatments, with 50-minute breaks in between. After a day of therapy, Lehman’s mood score indicated she was no longer depressed; it took up to five days for other participants. On average, three days of the therapy were enough for participants to have relief from depression.

“The less treatment-resistant participants are, the longer the treatment lasts,” said postdoctoral scholar  Eleanor Cole , PhD, a lead author of the study.

Strengthening a weak connection

The researchers also conjectured that targeting the stimulation more precisely would improve the treatment’s effectiveness. In transcranial magnetic stimulation, the treatment is aimed at the location where most people’s dorsolateral prefrontal cortex lies. This region regulates executive functions, such as selecting appropriate memories and inhibiting inappropriate responses.

For SAINT, the researchers used magnetic-resonance imaging of brain activity to locate not only the dorsolateral prefrontal cortex, but a particular subregion within it. They pinpointed the subregion in each participant that has a relationship with the subgenual cingulate, a part of brain that is overactive in people experiencing depression.

In people who are depressed, the connection between the two regions is weak, and the subgenual cingulate becomes overactive, said Keith Sudheimer , PhD, clinical assistant professor of psychiatry and a senior author of the study. Stimulating the subregion of the dorsolateral prefrontal cortex reduces activity in the subgenual cingulate, he said.

To test safety, the researchers evaluated the participants’ cognitive function before and after treatment. They found no negative side effects; in fact, they discovered that the participants’ ability to switch between mental tasks and to solve problems had improved — a typical outcome for people who are no longer depressed.

One month after the therapy, 60% of participants were still in remission from depression. Follow-up studies are underway to determine the duration of the antidepressant effects.

The researchers plan to study the effectiveness of SAINT on other conditions, such as obsessive-compulsive disorder, addiction and autism spectrum disorders.

‘Resilient and stable’

The depression Lehman woke up to almost two years ago was the worst episode she had ever experienced. Today, she said, she is happy and calm.

Since undergoing SAINT treatment, she has completed a bachelor’s degree at the University of California-Santa Barbara; she had dropped out as a young woman when her bipolar symptoms overwhelmed her studies.

“I used to cry over the slightest thing,” she said. “But when bad things happen now, I’m just resilient and stable. I’m in a much more peaceful state of mind, able to enjoy the positive things in life with the energy to get things done.”

Graduate student Katy Stimpson and Brandon Bentzley , MD, PhD, a medical fellow in psychiatry and behavioral sciences, are also lead authors.

Other Stanford co-authors are former lab manager Merve Gulser; graduate students Kirsten Cherian, Elizabeth Choi, Haley Aaron and Austin Guerra; Flint Espil , PhD, clinical assistant professor of psychiatry and behavioral sciences; research coordinators Claudia Tischler, Romina Nejad and Heather Pankow; medical student Jaspreet Pannu; postdoctoral scholars Xiaoqian Xiao, PhD, James Bishop , PhD, John Coetzee , PhD, and Angela Phillips , PhD; Hugh Solvason , MD, PhD, clinical professor of psychiatry and behavioral sciences; research manager Jessica Hawkins; Booil Jo , PhD, associate professor of psychiatry and behavioral sciences; Kristin Raj , MD, clinical assistant professor of psychiatry and behavioral sciences; Charles DeBattista , MD, professor of psychiatry and behavioral sciences; Jennifer Keller , PhD, clinical associate professor of psychiatry and behavioral sciences; and Alan Schatzberg , MD, professor of psychiatry and behavioral sciences.

The research was supported by Charles R. Schwab, the Marshall and Dee Ann Payne Fund, the Lehman Family Neuromodulation Research Fund, the Still Charitable Fund, the Avy L. and Robert L. Miller Foundation, a Stanford Psychiatry Chairman’s Small Grant, the Stanford CNI Innovation Award, the National Institutes of Health (grants T32035165 and UL1TR001085), the Stanford Medical Scholars Research Scholarship, the NARSAD Young Investigator Award and the Gordie Brookstone Fund. 

If you're interested in participating in a study, please email [email protected] .

A 1:2:1  podcast  on the study's findings features Williams in conversation with Paul Costello, senior communications strategist and adviser for Stanford Health Care and the School of Medicine.

• Mandy Erickson Mandy Erickson is a science writer in the Office of Communications. Email her at [email protected].

About Stanford Medicine

Stanford Medicine is an integrated academic health system comprising the Stanford School of Medicine and adult and pediatric health care delivery systems. Together, they harness the full potential of biomedicine through collaborative research, education and clinical care for patients. For more information, please visit med.stanford.edu .

Hope amid crisis

Psychiatry’s new frontiers

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• My Bibliography
• Collections
• Citation manager

Save citation to file

Email citation, add to collections.

• Create a new collection
• Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

• Search in PubMed
• Search in NLM Catalog
• Add to Search

Working memory in clinical depression: an experimental study

Affiliation.

• 1 Department of Academic Psychiatry, University College and Middlesex School of Medicine, London.
• PMID: 8475219
• DOI: 10.1017/s0033291700038873

This study compared clinically depressed subjects with normal controls on a range of working memory tasks. The findings suggested the articulatory loop and visuospatial sketch pad components of working memory to be unimpaired in depression. On a range of clinical tasks likely to involve central executive function, depressed subjects showed impairment only on some tasks.

PubMed Disclaimer

Similar articles

• [Gender differences in the neurocognitive components of depression]. Sárosi A, Gonda X, Balogh G, Székely A, Sasvári M, Faludi G. Sárosi A, et al. Neuropsychopharmacol Hung. 2008 Oct;10(4):191-9. Neuropsychopharmacol Hung. 2008. PMID: 19213198 Hungarian.
• Cognitive inhibition and working memory in unipolar depression. Gohier B, Ferracci L, Surguladze SA, Lawrence E, El Hage W, Kefi MZ, Allain P, Garre JB, Le Gall D. Gohier B, et al. J Affect Disord. 2009 Jul;116(1-2):100-5. doi: 10.1016/j.jad.2008.10.028. Epub 2008 Nov 29. J Affect Disord. 2009. PMID: 19042027
• Exploration of a 'double-jeopardy' hypothesis within working memory profiles for children with specific language impairment. Briscoe J, Rankin PM. Briscoe J, et al. Int J Lang Commun Disord. 2009 Mar-Apr;44(2):236-50. doi: 10.1080/13682820802028760. Int J Lang Commun Disord. 2009. PMID: 18821159
• [Deficits in memory retrieval: an argument in favor of frontal subcortical dysfunction in depression]. Fossati P, Deweer B, Raoux N, Allilaire JF. Fossati P, et al. Encephale. 1995 Jul-Aug;21(4):295-305. Encephale. 1995. PMID: 7588169 French.
• Clinical depression and implicit memory. Elliott CL, Greene RL. Elliott CL, et al. J Abnorm Psychol. 1992 Aug;101(3):572-4. doi: 10.1037//0021-843x.101.3.572. J Abnorm Psychol. 1992. PMID: 1500615
• Executive Function Deficits in Daily Life Prospectively Predict Increases in Depressive Symptoms. Letkiewicz AM, Miller GA, Crocker LD, Warren SL, Infantolino ZP, Mimnaugh KJ, Heller W. Letkiewicz AM, et al. Cognit Ther Res. 2014 Dec;38(6):612-620. doi: 10.1007/s10608-014-9629-5. Epub 2014 Aug 6. Cognit Ther Res. 2014. PMID: 37786427 Free PMC article.
• Shared and distinct brain fMRI response during performance of working memory tasks in adult patients with schizophrenia and major depressive disorder. Wang X, Cheng B, Roberts N, Wang S, Luo Y, Tian F, Yue S. Wang X, et al. Hum Brain Mapp. 2021 Nov;42(16):5458-5476. doi: 10.1002/hbm.25618. Epub 2021 Aug 25. Hum Brain Mapp. 2021. PMID: 34431584 Free PMC article. Review.
• Examining Specificity of Neural Correlates of Childhood Psychotic-like Experiences During an Emotional n-Back Task. O'Brien KJ, Barch DM, Kandala S, Karcher NR. O'Brien KJ, et al. Biol Psychiatry Cogn Neurosci Neuroimaging. 2020 Jun;5(6):580-590. doi: 10.1016/j.bpsc.2020.02.012. Epub 2020 Mar 19. Biol Psychiatry Cogn Neurosci Neuroimaging. 2020. PMID: 32354687 Free PMC article.
• Alterations in visual cortical activation and connectivity with prefrontal cortex during working memory updating in major depressive disorder. Le TM, Borghi JA, Kujawa AJ, Klein DN, Leung HC. Le TM, et al. Neuroimage Clin. 2017 Jan 7;14:43-53. doi: 10.1016/j.nicl.2017.01.004. eCollection 2017. Neuroimage Clin. 2017. PMID: 28138426 Free PMC article.
• WORKING MEMORY IMPAIRMENT AS AN ENDOPHENOTYPIC MARKER OF A SCHIZOPHRENIA DIATHESIS. Park S, Gooding DC. Park S, et al. Schizophr Res Cogn. 2014 Sep 1;1(3):127-136. doi: 10.1016/j.scog.2014.09.005. Schizophr Res Cogn. 2014. PMID: 25414816 Free PMC article.

Publication types

• Search in MeSH

LinkOut - more resources

• Genetic Alliance
• MedlinePlus Health Information
• Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

Change Password

Your password must have 6 characters or more:.

• a lower case character, 
• an upper case character, 
• a special character 

Password Changed Successfully

Your password has been changed

Create your account

Forget yout password.

Enter your email address below and we will send you the reset instructions

If the address matches an existing account you will receive an email with instructions to reset your password

Forgot your Username?

Enter your email address below and we will send you your username

If the address matches an existing account you will receive an email with instructions to retrieve your username

• June 01, 2024 | VOL. 181, NO. 6 CURRENT ISSUE pp.461-564
• May 01, 2024 | VOL. 181, NO. 5 pp.347-460
• April 01, 2024 | VOL. 181, NO. 4 pp.255-346
• March 01, 2024 | VOL. 181, NO. 3 pp.171-254
• February 01, 2024 | VOL. 181, NO. 2 pp.83-170
• January 01, 2024 | VOL. 181, NO. 1 pp.1-82

The American Psychiatric Association (APA) has updated its Privacy Policy and Terms of Use , including with new information specifically addressed to individuals in the European Economic Area. As described in the Privacy Policy and Terms of Use, this website utilizes cookies, including for the purpose of offering an optimal online experience and services tailored to your preferences.

Please read the entire Privacy Policy and Terms of Use. By closing this message, browsing this website, continuing the navigation, or otherwise continuing to use the APA's websites, you confirm that you understand and accept the terms of the Privacy Policy and Terms of Use, including the utilization of cookies.

Exercise and the Prevention of Depression: Results of the HUNT Cohort Study

• Samuel B. Harvey , F.R.A.N.Z.C.P., Ph.D. ,
• Simon Øverland , Ph.D. ,
• Stephani L. Hatch , Ph.D. ,
• Simon Wessely , F.R.C.Psych., M.D. ,
• Arnstein Mykletun , Ph.D. ,
• Matthew Hotopf , F.R.C.Psych., Ph.D.

Search for more papers by this author

The purpose of the present study was to address 1) whether exercise provides protection against new-onset depression and anxiety and 2) if so, the intensity and amount of exercise required to gain protection and, lastly, 3) the mechanisms that underlie any association.

A “healthy” cohort of 33,908 adults, selected on the basis of having no symptoms of common mental disorder or limiting physical health conditions, was prospectively followed for 11 years. Validated measures of exercise, depression, anxiety, and a range of potential confounding and mediating factors were collected.

Undertaking regular leisure-time exercise was associated with reduced incidence of future depression but not anxiety. The majority of this protective effect occurred at low levels of exercise and was observed regardless of intensity. After adjustment for confounders, the population attributable fraction suggests that, assuming the relationship is causal, 12% of future cases of depression could have been prevented if all participants had engaged in at least 1 hour of physical activity each week. The social and physical health benefits of exercise explained a small proportion of the protective effect. Previously proposed biological mechanisms, such as alterations in parasympathetic vagal tone, did not appear to have a role in explaining the protection against depression.

Conclusions:

Regular leisure-time exercise of any intensity provides protection against future depression but not anxiety. Relatively modest changes in population levels of exercise may have important public mental health benefits and prevent a substantial number of new cases of depression.

The rising costs associated with depression and anxiety constitute a major public health problem across the developed and developing world ( 1 ). While the need to address the growing burden of these common mental disorders is not in doubt, there has been little consensus on how this should be executed. Although effective treatments are available, cost-effectiveness models suggest that even in the unlikely event of optimal treatment being delivered in all cases, only 35%–50% of the overall burden of depression and anxiety would be alleviated ( 2 ). As a result, a number of agencies have begun to consider strategies aimed at primary prevention of both depression and anxiety.

Rose ( 3 ) argued that the most appropriate method for preventing common, multifactorial diseases is to shift the entire population distribution of known risk factors. Such strategies are well established for the prevention of other conditions, such as cardiovascular disease. However, most of the known risk factors for depression and anxiety, such as familial risk, socioeconomic position, and life events, are difficult or impossible to modify ( 4 ).

There is, however, some emerging evidence that lifestyle factors, such as physical inactivity, may be potential targets for strategies aimed at preventing depression and anxiety ( 5 , 6 ). A variety of health surveys have demonstrated a cross-sectional association between exercise and lower rates of both depression and anxiety ( 7 , 8 ). However, the possibility of reverse causation (low mood or anxiety leading to reduced levels of exercise) has limited the interpretation of such studies. To date, the results of prospective studies have been more mixed. Some studies have found no prospective association between levels of exercise and depression and anxiety ( 9 – 11 ), while others have suggested that any beneficial effects of exercise may be limited to certain subgroups or age groups, or only associated with intensive exercise ( 12 – 14 ). The evidence base has been further confused by many, but not all, of the published reports conflating depression and anxiety disorders, despite each having unique risk factors and distinct biological processes ( 15 ). The evidence base for exercise as a treatment for current depression is more established, with numerous reviews concluding that exercise is moderately effective for reducing the symptoms of depression ( 16 , 17 ). Recent analyses of the data used in these reviews have added further evidence for the antidepressant effect of exercise, with findings that both publication bias and enhanced control group responses may have led to an underestimate of the true effect size of exercise as an intervention in depression ( 18 , 19 ). However, systematic reviews of the evidence for exercise in preventing new-onset depression and/or anxiety have needed to be more tempered in their conclusions, particularly regarding the relative importance of the intensity and amount of exercise required to convey any protective effect ( 20 ).

A number of theories have been proposed as to how exercise may prevent mental illness, yet to date none of these have been formally evaluated in prospective epidemiological studies ( 5 ). Exercise is associated with a number of biological changes that could have an impact on mental health. One purported biological mechanism is alteration in the activity of the autonomic nervous system ( 21 , 22 ). Regular exercise increases parasympathetic vagal tone, leading to physiological changes such as resting bradycardia ( 23 ). Alterations in autonomic nervous system activity have been observed in those suffering from depression, and vagal nerve stimulation has been used to treat depression ( 21 ). Other explanations for any association between exercise and depression and anxiety focus on the physical health, self-esteem, or social benefits of exercise.

Addressing the uncertainty surrounding the relationship between exercise and depression and anxiety is important. While many agencies are keen to promote the potential mental health benefits of exercise, at present the literature is unable to provide the most basic information needed for effective, targeted, evidence-based public health campaigns concerning depression and anxiety. The aim of the present study was to utilize a large (N=33,908) prospective cohort to address three questions. First, does exercise provide protection against new-onset depression and anxiety? Second, if so, what intensity and total amount of exercise is required to gain protection? Third, what causal mechanisms underlie any association between exercise and later depression and anxiety?

Study Design

The HUNT study [Health Study of Nord-Trøndelag County] represents one of the largest and most comprehensive population-based health surveys ever undertaken. The Nord-Trøndelag County of Norway covers a mainly rural area, with a total population of 127,000 at the time of the study commencement. In phase 1 (HUNT 1) of the study, conducted between January 1984 and February 1986, all inhabitants of the county aged 20 years or older (N=85,100) were invited to complete questionnaires on their lifestyle and medical history and to attend a physical examination. A total of 74,599 individuals participated (87.7%). All participants were then followed up 9 to 13 years later (between August 1995 and June 1997) in phase 2 of the study (HUNT 2). Detailed information on the HUNT cohort study has been published elsewhere ( 24 , 25 ). The STROBE [Strengthening the Reporting of Observational studies in Epidemiology] checklist (see appendix SA1 in the data supplement accompanying the online version of this article) was followed throughout the study.

Selection of a “Healthy” Sample

In order to be more confident regarding the direction of any associations found, data from HUNT 1 were used to select a “healthy” cohort, without any evidence of current physical illness or depressive or anxiety disorders at baseline.

Symptoms of depression and anxiety at baseline.

The presence of depression and anxiety at baseline (HUNT 1) was detected in two ways. Firstly, all participants completed the 12-item Anxiety and Depression Symptom Index. This measure, designed to capture a range of symptoms suggestive of depression and anxiety, has been validated and shown to have a good test-retest correlation ( 26 , 27 ). A total of 60,980 respondents (81.7%) returned an adequately completed questionnaire. Previously, a cut-off at the 80th percentile of the total score on the questionnaire has been used to define caseness ( 28 ). In order to be more conservative in creating a “healthy” cohort, only those falling below the 70th percentile of the total score were selected (N=42,686). Secondly, participants were also asked whether they suffered from any impairments due to psychological complaints. An additional 726 individuals who indicated baseline psychological impairments were excluded.

Assessment of physical health at baseline.

Physical ill-health may prevent individuals from participating in exercise and is an independent predictor of common mental disorders ( 29 ). Participants were asked whether they suffered from any impairments with regard to motor abilities or impairments due to physical illness, or suffered from, or had ever been diagnosed with, diabetes, angina, myocardial infarction, stroke, or cerebral hemorrhage. Consequently, an additional 8,444 participants were excluded, yielding a final “healthy” cohort of 33,908 individuals.

Measurement of Exercise at Baseline

At the time of their baseline assessment (HUNT 1), all participants were asked how often they engaged in exercise (such as walking or swimming). They were provided with five options: 1) never, 2) less than once a week, 3) once a week, 4) two to three times a week, and 5) nearly every day. Participants were asked, on average, how long they exercised on each occasion. By combining the answers for both of these questions, it was possible to produce an estimate of the total number of minutes per week each individual spent exercising. Participants were also asked about the intensity of their exercise, with three possible options: 1) exercise without becoming breathless or sweating, 2) exercise resulting in breathlessness and sweating, or 3) exercise resulting in near exhaustion. The last two options were combined into one category.

The reliability and validity of these questions and a composite total time spent engaging in exercise per week have been demonstrated against three additional objective measures of physical activity: maximal oxygen uptake, measures of body position and motion over 7 days using an ActiReg recording instrument, and the International Physical Activity Questionnaire ( 30 ).

Assessment of Depression and Anxiety in HUNT 2

At follow-up (HUNT 2), all participants were asked to complete the Hospital Anxiety and Depression Scale ( 31 ). The Hospital Anxiety and Depression Scale is a self-report questionnaire comprising 14 4-point Likert-scale items covering anxiety and depression symptoms over the previous 2 weeks. A cut-off score of 8 in each subscale (anxiety subscale, depression subscale) has been found to be optimal for case finding, with sensitivity and specificity estimates of around 0.80 ( 32 ).

Potential Confounding and Mediating Variables

In order to facilitate the differentiating of confounders and mediators, a conceptual hierarchical framework (see Figure 1 ) was constructed a priori to outline how various factors may relate to both exercise and depression and anxiety.

FIGURE 1. Theoretical Hierarchical Model Demonstrating How Potential Confounding and Mediating Factors May Interact With Both Levels of Exercise and Depression and Anxiety a

a Abbreviations: ADI-12=Anxiety and Depression Symptom Index-12 item; ANS=autonomic nervous system; HADS=Hospital Anxiety and Depression Scale.

Demographic and socioeconomic factors.

Information on participants’ age, gender, and marital status was obtained from the Norwegian National Population Registry. Participants were asked to record their highest completed education level. Occupational social class was calculated according to the International Erikson-Goldthorpe-Portocareros classification ( 33 ).

Substance use.

Participants were asked to report the total number of cigarettes consumed per day and how frequently they drank alcohol, with five options ranging from total abstainer to 10 or more times in the past week.

Body-mass index (BMI).

A specially trained nurse took measurements calculating the BMI of each participant.

New-onset physical illness.

The same somatic conditions and limitations that were considered at baseline as exclusion measures were also assessed at follow-up, allowing the identification of any new-onset illnesses or impairments throughout the course of the study.

Autonomic nervous system activity.

Lowered resting pulse is a biological adaptation to regular exercise ( 34 ), due to increased parasympathetic vagal tone ( 23 ). At the HUNT 1 physical examination, each participant’s resting pulse was measured after at least 4 minutes of rest in the sitting position by palpation over the radial artery for 15 seconds.

Perceived social support.

Each participant’s perceived social support (both instrumental and emotional) was assessed via a single question: “If you fell ill and had to stay in bed for a significant period, how likely do you think it is that you would get the necessary help and support from family, friends, or neighbors?” Five options were provided: 1) very likely, 2) quite likely, 3) doubtful, 4) unlikely, 5) not likely at all.

Statistical Analysis

All analyses were conducted using STATA statistical software, Version 10.1 (StataCorp, College Station, Tex.) ( 35 ). A multiple imputation model was constructed to replace missing values using the imputation by chained approach method. Thirty imputed data sets were created. All variables used in the analyses were included in the imputation model. Sensitivity analyses utilizing complete case analysis were undertaken to ensure results were not significantly altered by the multiple imputation process.

The associations between level of physical activity and both depression and anxiety were assessed using univariate and then multivariate logistic regression. The total amount of exercise undertaken each week was divided into six categories ranging from none up to more than 4 hours. The relative confounding effect of each of the variables outlined above was examined in turn, before a final multivariate model containing all potential confounders was constructed. Interactions by gender, age subgroup, and exercise intensity were tested using postestimation Wald tests. The correlation between levels of exercise at baseline and follow-up was examined using Spearman’s rank-order correlation tests. In addition to reporting odds ratios, the relative importance of exercise in predicting future depression and anxiety was examined using population attributable fractions.

Finally, the importance of the following three potential mediating factors was considered: 1) new-onset physical illness, 2) autonomic nervous system activity (resting pulse), and 3) the level of perceived social support. Although a “healthy” cohort was selected for this study, the possibility of reverse causation remained, with subthreshold symptoms of depression and anxiety leading to reduced levels of exercise. Therefore, a fourth potential pathway, reverse causation, was also considered. The associations between each of the four potential mediating factors and both exercise level and levels of depression and anxiety were examined using linear regression, before the effect of adding each potential mediating factor to a final model was assessed.

Both the HUNT 1 and HUNT 2 phases of the study were approved by the National Data Inspectorate and the Board of Research Ethics in Health Region IV of Norway.

Characteristics of the Study Sample

The characteristics of the “healthy” cohort of 33,908 individuals are summarized in Table 1 . Of these, 22,564 (66.5%) were successfully followed up at the HUNT 2 phase. Females (p<0.001) and younger participants (p<0.0001) were more likely to be followed up. The frequency of exercise undertaken at baseline did not predict loss to follow-up once the effect of gender and age was considered (p=0.19). Of the 22,564 individuals followed up, 1,578 (7.0%) developed case-level symptoms of depression, and 1,972 (8.7%) developed case-level symptoms of anxiety. All participants gave their informed consent to participate in this study.

TABLE 1. Characteristics of the “Healthy” Cohort of Individuals Followed Prospectively From Phase 1 to Phase 2 of the Health Study of Nord-Trøndelag County

a Based on current occupation.

b Data indicate frequency within the previous 7 days.

TABLE 1. Characteristics of the “Healthy” Cohort of Individuals Followed Prospectively From Phase 1 to Phase 2 of the Health Study of Nord-Trøndelag County

Exercise at Baseline and New-Onset Depression and Anxiety

There was a negative relationship between the total amount of exercise undertaken at baseline and risk of future depression (p=0.001). In contrast, the prevalence of case-level anxiety was similar regardless of the levels of baseline exercise (p=0.21). Logistic regression models of the associations between the total amount of exercise at baseline and later depression and anxiety are displayed in Table 2 . After adjustment for a range of confounders, those who reported undertaking no exercise at baseline had a 44% (95% confidence interval = 17%–78%) increased odds of developing case-level depression compared with those who were exercising 1–2 hours a week. The models presented in Table 2 confirm the lack of any association between baseline exercise levels and later case-level anxiety (p=0.27).

TABLE 2. Prospective Associations Between Total Amount of Exercise at Baseline and Later Depression or Anxiety

a The correlation between regular exercise and reduced incidence of future depression reached statistical significance (p=0.003).

b The correlation between regular exercise and reduced incidence of future anxiety was not statistically significant (p=0.27).

c Adjusted for age and gender.

d Adjusted for age, gender, marital status, education, and social class.

e Adjusted for sociodemographic characteristics plus cigarette and alcohol use.

f Adjusted for sociodemographic characteristics plus body mass index.

g Adjusted for age, gender, marital status, education, social class, number of cigarettes consumed, alcohol use, and body mass index.

TABLE 2. Prospective Associations Between Total Amount of Exercise at Baseline and Later Depression or Anxiety

There was no evidence of interaction by gender (all p values >0.2) or when stratified by age group (greater than or less than 50 years old, p=0.96) in the association between the total amount of exercise at baseline and later case-level depression or anxiety. A similar significant association was seen between baseline levels of exercise and later depression in those aged less than 50 years (p=0.04) and those aged 50 years and older (p=0.03). As expected, there was a significant correlation between the amount of exercise undertaken at baseline and follow-up (p<0.001).

Dose-Response Relationship

Visual representations of the dose-response relationship between total exercise at baseline and the odds of later case-level depression are provided in Figure 2 . Most of the protective effect of exercise is realized with relatively low levels of exercise, with no indication of any additional benefit beyond 1 hour of exercise each week. Maximum likelihood ratio tests suggest that an exponential decay model (with decreasing benefit as the total time of exercise increases) was a better fit for the data than a linear model (test for difference between models, p=0.004). The combined population attributable fraction for less than 1 hour of exercise per week was 11.9%. There was no evidence of an interaction by intensity of exercise (p=0.96).

FIGURE 2. Adjusted Odds Ratios (With 95% CIs) for Case-Level Depression at Follow-Up According to the Overall Amount of Exercise Reported at Baseline a

a All odds ratios are adjusted for age, gender, marital status, education, social class, number of cigarettes consumed, alcohol use, and body mass index.

Possible Mediating Pathways

In line with a priori predictions, those who engaged in less exercise at baseline tended to have higher resting pulse, lower levels of perceived social support, and more subthreshold symptoms of depression and anxiety, and they were more likely to develop new-onset physical illnesses over the course of the study (p<0.001). Table 3 demonstrates that three of the four potential mediating pathways considered accounted for some of the observed association: reverse causation, lower levels of perceived social support, and new-onset physical illness. However, each of these modeled pathways explained only a very small proportion of the observed effect, with the majority of the protective effect of exercise remaining unaccounted for by measured factors.

TABLE 3. Additional Multivariable Models to Investigate Possible Mediating Pathways Between Total Amount of Exercise at Baseline and Later New Case-Level Depression

a Postestimation was conducted using the Wald test.

b Adjusted for age, gender, marital status, education, social class, number of cigarettes consumed, alcohol use, and body mass index.

c Adjusted for sociodemographic characteristics (as in Model 1) plus baseline symptoms using the 12-item Anxiety and Depression Symptom Index.

d Adjusted for sociodemographic characteristics (as in Model 1) plus resting pulse.

e Adjusted for sociodemographic characteristics (as in Model 1) plus likelihood of support and perceived support from family, friends, or neighbors.

f Adjusted for sociodemographic characteristics (as in Model 1) plus new-onset physical illness.

g The sum of new diagnoses and the level of new impairment were added as separate variables.

TABLE 3. Additional Multivariable Models to Investigate Possible Mediating Pathways Between Total Amount of Exercise at Baseline and Later New Case-Level Depression

Using a large population cohort study, we have observed that relatively small amounts of exercise can provide significant protection against future depression but not anxiety. This protective effect was seen equally across all groups, regardless of the intensity of exercise that was undertaken or the gender or age of the participants. Assuming there is no residual confounding in our final model and the observed relationship is causal, our results suggest that if all participants had exercised for at least 1 hour each week, 12% of the cases of depression at follow-up could have been prevented.

The key strengths of this study are its large sample size, prospective data collection, use of validated measures of physical activity and mental disorder, and the detailed information available on a wide range of potential confounding and mediating factors. Despite these strengths, the analyses presented have some important limitations. Regarding the study design, while individuals reporting current symptoms and/or impairment of depression or anxiety at baseline were excluded using a two-step process, we were not able to exclude individuals with a history of prior episodes of depression and anxiety. Thus, it is possible that some individuals with a lifetime history of depression or anxiety may have been included in the “healthy” cohort, and thus a proportion of the future cases may be recurrent episodes of depression or anxiety. While the long follow-up time is a strength, the use of a single measure of a relapsing and remitting condition such as depression means that some misclassification will have occurred. Such misclassification is likely to be random and therefore results in regression dilution bias and an attenuation of effect sizes. This has important consequences for the interpretation of the results and suggests that the actual protective effect of exercise may be even greater than that reported in this study. The measurement of exercise at a single time point will also have created some misclassification, although there was a significant correlation between levels of exercise at baseline and follow-up. The majority of other limitations relate to the measures used. While the Hospital Anxiety and Depression Scale is one of the most widely used and validated measures of depression and anxiety, the operationalization of any mental disorder via a self-report screening tool cannot be considered equivalent to a clinical diagnosis, and a risk of misclassification remains. Similarly, while the measures of exercise used have been extensively validated ( 30 ), they remain reliant on self-report, and each factor considered in the analysis of mediation was measured with a single item that may not have fully captured the constructs being considered. In addition to standard regression, population attributable fractions were used to describe the relative importance of exercise as a possible preventative strategy. Population attributable fractions can be a useful way to help guide public health interventions, but any estimates of population attributable fractions assume a causal relationship with no residual confounding. While attempts were made to account for many confounders, a number of important potential confounding variables, such as personality, attitude toward health, diet, seasonal weather variations, and the degree to which each participant’s local environment is conducive to regular exercise, remained unmeasured. Nord-Trøndelag County is situated between northern latitudes 63° and 65°. As a result, there is considerable seasonal variation in the number of daylight hours. Previous studies have shown an associated seasonal variation in depression prevalence within Nord-Trøndelag County, with higher rates of depression between December and April ( 36 ). If rates of physical activity were also lower in the winter months, then the confounding effect of the season at the time of assessment could affect any observed cross-sectional association between exercise and depression. However, this longitudinal study mitigates against this by the fact that there was no link between the seasonal timing of the assessments in HUNT 1 (when exercise levels were measured) and HUNT 2 (when levels of depression were assessed). The invitation for participation in the HUNT 2 phase was sent to all residents of the county at a time that was unrelated to when each individual had been assessed during the HUNT 1 phase. The equal popularity of both winter and summer sports in Norway may also have reduced the possibility of seasonal confounding.

This study represents, to our knowledge, the largest and most detailed modeling of the prospective dose-response relationship between exercise and depression and the first published epidemiological exploration of the causal pathways involved. In addition to confirming that more active individuals are less likely to develop depression, we were able to demonstrate that this was most accurately modeled as an exponential decay model, with decreasing benefit as the total time spent exercising increases. This supports and expands on the tentative conclusions from a review published in 2014, which highlighted that substantial mental health benefits may be gained from relatively moderate levels of exercise ( 37 ). Importantly, the majority of the protective effects of exercise against depression are realized within the first hour of exercise undertaken each week, which provides some clues regarding causation and has major implications for possible future public mental health campaigns. The majority of studies examining the role of physical activity in preventing cardiovascular disease have found that the beneficial cardiovascular effects continue to increase up to around 2–3 hours of exercise per week ( 38 , 39 ). Thus, while there are similarities in the overall shape of the dose-response relationship between exercise and depression and exercise and somatic illness, the level of activity needed to realize the bulk of the possible protective effects are very different. Our finding that more vigorous-intensity exercise had no additional protective effects against future case-level depression is also in contrast to previous findings regarding protective factors against cardiovascular disease ( 39 ).

Taken together, these results suggest that processes such as alterations in autonomic nervous system activity and modification of metabolic factors, which require more regular or strenuous exercise, may be less important when considering the protective effects of exercise against future depressive illness. In keeping with this hypothesis, our results suggest that the perceived social benefits of exercise may mediate some of the protective effects against depression. However, within our analysis, the increased levels of perceived social support accounted for only a small proportion of the effect observed, meaning the bulk of the observed protective effect remains unexplained. People’s perception of their social support may be subject to bias relating to their current mental state. This type of reporting bias could lead to an overestimate of the mediating effect of perceived social support, meaning an even greater proportion of the observed protective effect of exercise may be unexplained. We propose two possible explanations to account for the unexplained protective effect of exercise. Firstly, the remaining prospective associations may be due to confounding from factors not measured, such as shared genetic factors, personality, or individual attitudes toward health ( 40 ). Secondly, or alternatively, there may be other causal factors not measured in this study, such as changes in self-esteem, serotonin release ( 41 ), increased expression of neuroprotective proteins such as brain-derived neurotropic factor, altered hippocampal neurogenesis, or modifications to the activity levels of the hypothalamic-pituitary-adrenal axis ( 42 ). The lack of any association between exercise and future anxiety disorders suggests that the link between exercise and depression is not merely related to a general increase in mental well-being and is unlikely to involve risk factors shared between depression and anxiety.

Despite the remaining uncertainty regarding causal pathways, the findings presented in this study have important public health implications. There is evidence that the levels of exercise in the general population in developed countries have decreased considerably over the recent decades ( 43 ), with similar trends now also being observed in developing countries. The results of this study indicate that relatively modest increases in the overall amount of time spent exercising per week may be able to prevent a substantial number of new cases of depression. If causality is assumed and there are no other major cofounders, our results suggest that at least 12% of new cases of depression could be prevented if all adults participated in at least 1 hour of exercise each week. While education regarding the higher levels of exercise required to achieve maximum cardiovascular and metabolic benefits remains important, informing individuals that significant mental health benefits may be achieved with small changes in their behavior may be valuable in facilitating behavioral change. Given that the intensity of exercise does not appear to be important, it may be that the most effective public health measures are those that encourage and facilitate increased levels of everyday activities, such as walking or cycling. The results presented in this study provide a strong argument in favor of further exploration of exercise as a strategy for the prevention of depression.

Prof. Mykletun and Dr. Hotopf are joint last authors.

The HUNT study [Health Study of Nord-Trøndelag County] is a collaboration between the HUNT Research Centre, Faculty of Medicine, Norwegian University of Science and Technology (NTNU, Verdal), Norwegian Institute of Public Health and Nord-Trøndelag County Council.

This study represents independent research funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health. None of the funders had any role in the study design, collection, analysis, interpretation of data, writing of the manuscript, or the decision to submit this study for publication.

A/Prof. Harvey was funded by NSW Health and a grant from the Institute of Social Psychiatry. Prof. Mykletun and Prof. Øverland were funded by the Norwegian Research Council. All other authors report no financial relationships with commercial interests.

The authors thank Erlend Bergesen, who was initially involved in this study but, sadly, died in August 2006.

1 Murray CJ, Vos T, Lozano R, et al. : Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010 . Lancet 2012 ; 380:2197–2223 Crossref , Medline ,  Google Scholar

2 Andrews G, Issakidis C, Sanderson K, et al. : Utilising survey data to inform public policy: comparison of the cost-effectiveness of treatment of ten mental disorders . Br J Psychiatry 2004 ; 184:526–533 Crossref , Medline ,  Google Scholar

3 Rose G : The Strategy of Preventive Medicine . Oxford, United Kingdom, Oxford University Press, 1993 Google Scholar

4 Weich S : Prevention of the common mental disorders: a public health perspective . Psychol Med 1997 ; 27:757–764 Crossref , Medline ,  Google Scholar

5 Teychenne M, Ball K, Salmon J : Physical activity and likelihood of depression in adults: a review . Prev Med 2008 ; 46:397–411 Crossref , Medline ,  Google Scholar

6 Schuch FB, Vancampfort D, Sui X, et al. : Are lower levels of cardiorespiratory fitness associated with incident depression? A systematic review of prospective cohort studies . Prev Med 2016 ; 93:159–165 Crossref , Medline ,  Google Scholar

7 Goodwin RD : Association between physical activity and mental disorders among adults in the United States . Prev Med 2003 ; 36:698–703 Crossref , Medline ,  Google Scholar

8 Harvey SB, Hotopf M, Overland S, et al. : Physical activity and common mental disorders . Br J Psychiatry 2010 ; 197:357–364 Crossref , Medline ,  Google Scholar

9 Birkeland MS, Torsheim T, Wold B : A longitudinal study of the relationship between leisure-time physical activity and depressed mood among adolescents . Psychol Sport Exerc 2009 ; 10:25–34 Crossref ,  Google Scholar

10 Cooper-Patrick L, Ford DE, Mead LA, et al. : Exercise and depression in midlife: a prospective study . Am J Public Health 1997 ; 87:670–673 Crossref , Medline ,  Google Scholar

11 Kritz-Silverstein D, Barrett-Connor E, Corbeau C : Cross-sectional and prospective study of exercise and depressed mood in the elderly: the Rancho Bernardo Study . Am J Epidemiol 2001 ; 153:596–603 Crossref , Medline ,  Google Scholar

12 Bhui K, Fletcher A : Common mood and anxiety states: gender differences in the protective effect of physical activity . Soc Psychiatry Psychiatr Epidemiol 2000 ; 35:28–35 Crossref , Medline ,  Google Scholar

13 Farmer ME, Locke BZ, Mościcki EK, et al. : Physical activity and depressive symptoms: the NHANES I Epidemiologic Follow-Up Study . Am J Epidemiol 1988 ; 128:1340–1351 Crossref , Medline ,  Google Scholar

14 Wiles NJ, Haase AM, Gallacher J, et al. : Physical activity and common mental disorder: results from the Caerphilly Study . Am J Epidemiol 2007 ; 165:946–954 Crossref , Medline ,  Google Scholar

15 Mathew AR, Pettit JW, Lewinsohn PM, et al. : Co-morbidity between major depressive disorder and anxiety disorders: shared etiology or direct causation? Psychol Med 2011 ; 41:2023–2034 Crossref , Medline ,  Google Scholar

16 Cooney GM, Dwan K, Greig CA, et al. : Exercise for depression . Cochrane Database Syst Rev 2013 ; (9):CD004366 Medline ,  Google Scholar

17 Rosenbaum S, Tiedemann A, Sherrington C, et al. : Physical activity interventions for people with mental illness: a systematic review and meta-analysis . J Clin Psychiatry 2014 ; 75:964–974 Crossref , Medline ,  Google Scholar

18 Schuch FB, Vancampfort D, Richards J, et al. : Exercise as a treatment for depression: a meta-analysis adjusting for publication bias . J Psychiatr Res 2016 ; 77:42–51 Crossref , Medline ,  Google Scholar

19 Stubbs B, Vancampfort D, Rosenbaum S, et al. : Challenges establishing the efficacy of exercise as an antidepressant treatment: a systematic review and meta-analysis of control group responses in exercise randomised controlled trials . Sports Med 2016 ; 46:699–713 Crossref , Medline ,  Google Scholar

20 Mammen G, Faulkner G : Physical activity and the prevention of depression: a systematic review of prospective studies . Am J Prev Med 2013 ; 45:649–657 Crossref , Medline ,  Google Scholar

21 Rottenberg J : Cardiac vagal control in depression: a critical analysis . Biol Psychol 2007 ; 74:200–211 Crossref , Medline ,  Google Scholar

22 Zoeller RF : Depression, anxiety, physical activity, and cardiovascular disease: What is the connection? Am J Lifestyle Med 2007 ; 1:175–180 Crossref ,  Google Scholar

23 Seals DR, Chase PB : Influence of physical training on heart rate variability and baroreflex circulatory control . J Appl Physiol (1985) 1989 ; 66:1886–1895 Crossref , Medline ,  Google Scholar

24 Krokstad S, Langhammer A, Hveem K, et al. : Cohort Profile: the HUNT Study, Norway . Int J Epidemiol 2013 ; 42:968–977 Crossref , Medline ,  Google Scholar

25 Holmen J, Bjartveit K, Hjort PF, et al: The NordTrøndelag Health Survey 1984–86: Purpose, Background and Methods: Participation, Non-Participation and Frequency Distributions. Verdal, Norway, Statens institutt for folkehelse, 1990 Google Scholar

26 Derogatis LR, Lipman RS, Rickels K, et al. : The Hopkins Symptom Checklist (HSCL): a self-report symptom inventory . Behav Sci 1974 ; 19:1–15 Crossref , Medline ,  Google Scholar

27 Tambs K, Moum T : Low genetic effect and age-specific family effect for symptoms of anxiety and depression in nuclear families, halfsibs and twins . J Affect Disord 1993 ; 27:183–195 Crossref , Medline ,  Google Scholar

28 Hildrum B, Mykletun A, Holmen J, et al. : Effect of anxiety and depression on blood pressure: 11-year longitudinal population study . Br J Psychiatry 2008 ; 193:108–113 Crossref , Medline ,  Google Scholar

29 Harvey SB, Ismail K : Psychiatric aspects of chronic physical disease . Medicine 2008 ; 36:471–474 Crossref ,  Google Scholar

30 Kurtze N, Rangul V, Hustvedt BE, et al. : Reliability and validity of self-reported physical activity in the Nord-Trøndelag Health Study: HUNT 1 . Scand J Public Health 2008 ; 36:52–61 Crossref , Medline ,  Google Scholar

31 Zigmond AS, Snaith RP : The Hospital Anxiety and Depression Scale . Acta Psychiatr Scand 1983 ; 67:361–370 Crossref , Medline ,  Google Scholar

32 Bjelland I, Dahl AA, Haug TT, et al. : The validity of the Hospital Anxiety and Depression Scale: an updated literature review . J Psychosom Res 2002 ; 52:69–77 Crossref , Medline ,  Google Scholar

33 Krokstad S, Westin S : Health inequalities by socioeconomic status among men in the Nord-Trøndelag Health Study, Norway . Scand J Public Health 2002 ; 30:113–124 Medline ,  Google Scholar

34 Scheuer J, Tipton CM : Cardiovascular adaptations to physical training . Annu Rev Physiol 1977 ; 39:221–251 Crossref , Medline ,  Google Scholar

35 StataCorp: Stata Statistical Software, Release 12.1. College Station, Tex, Stata Corporation, 2012 Google Scholar

36 Stordal E, Morken G, Mykletun A, et al. : Monthly variation in prevalence rates of comorbid depression and anxiety in the general population at 63–65 degrees North: the HUNT study . J Affect Disord 2008 ; 106:273–278 Crossref , Medline ,  Google Scholar

37 Rosenbaum S, Tiedemann A, Sherrington C, et al. : Physical activity interventions for people with mental illness: a systematic review and meta-analysis . J Clin Psychiatry 2014 ; 75:964–974 Crossref , Medline ,  Google Scholar

38 Bauman AE : Updating the evidence that physical activity is good for health: an epidemiological review 2000–2003 . J Sci Med Sport 2004 ; 7(Suppl):6–19 Crossref , Medline ,  Google Scholar

39 Haskell WL, Lee IM, Pate RR, et al. : Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association . Circulation 2007 ; 116:1081–1093 Crossref , Medline ,  Google Scholar

40 De Moor MH, Boomsma DI, Stubbe JH, et al. : Testing causality in the association between regular exercise and symptoms of anxiety and depression . Arch Gen Psychiatry 2008 ; 65:897–905 Crossref , Medline ,  Google Scholar

41 Meeusen R, Thorré K, Chaouloff F, et al. : Effects of tryptophan and/or acute running on extracellular 5-HT and 5-HIAA levels in the hippocampus of food-deprived rats . Brain Res 1996 ; 740:245–252 Crossref , Medline ,  Google Scholar

42 Ernst C, Olson AK, Pinel JP, et al. : Antidepressant effects of exercise: evidence for an adult-neurogenesis hypothesis? J Psychiatry Neurosci 2006 ; 31:84–92 Medline ,  Google Scholar

43 Carlson SA, Densmore D, Fulton JE, et al. : Differences in physical activity prevalence and trends from 3 U.S. surveillance systems: NHIS, NHANES, and BRFSS . J Phys Act Health 2009 ; 6(Suppl 1):S18–S27 Crossref , Medline ,  Google Scholar

• The relationship between depression and relative fat mass (RFM): A population-based study Journal of Affective Disorders, Vol. 356
• The impact of educational attainment on mental health: A Causal Assessment from the UKB and FinnGen Cohorts Medicine, Vol. 103, No. 26
• Association between weight-adjusted-waist index and depression in US adults: A cross-sectional study Journal of Affective Disorders, Vol. 355
• Ratings and experiences in using a mobile application to increase physical activity among university students: implications for future design 5 January 2023 | Universal Access in the Information Society, Vol. 23, No. 2
• Social dance and movement for mental health: A narrative review 1 April 2024 | Mental Health Science, Vol. 2, No. 2
• Neuroimmune modulation by tryptophan derivatives in neurological and inflammatory disorders European Journal of Cell Biology, Vol. 103, No. 2
• The effect of self‐designed metabolic equivalent exercises on cancer‐related fatigue in patients with gastric cancer: A randomized controlled trial 8 May 2024 | Cancer Medicine, Vol. 13, No. 9
• Yurtlarda Kalan Üniversite Öğrencilerinin Fiziksel Aktivite İle Psikolojik İyi Oluş Düzeyleri: Adana Örneği 30 April 2024 | Gençlik Araştırmaları Dergisi, Vol. 12, No. 32
• Activity and Positive Engagement within a PICU
• Leisure-Time Physical Activity Impact on Quality of Life, Anxiety, Depression, and Clinical Intensity of Pemphigus Vulgaris 26 March 2024 | Health Scope, Vol. 13, No. 1
• Editorial: Advances in the understanding of the affective and cognitive effects of physical activity, exercise, and sports 13 March 2024 | Frontiers in Psychology, Vol. 15
• Mental workload, musculoskeletal discomfort and physical activity level in master’s degree students in the COVID-19 context: A pilot study Work, Vol. 77, No. 3
• Physical activity and specific symptoms of depression: A pooled analysis of six cohort studies Journal of Affective Disorders, Vol. 348
• Modulatory effects of aerobic training on the degree centrality of brain functional activity in subthreshold depression Brain Research, Vol. 1827
• Self-efficacy for exercise in adults with lifetime depression and low physical activity Psychiatry Research Communications, Vol. 4, No. 1
• Microbiota–gut–brain axis: the mediator of exercise and brain health 19 April 2024 | Psychoradiology, Vol. 4
• The influence of lifestyle habits on levels of depression among rural middle school students in Northeastern China 29 January 2024 | Frontiers in Public Health, Vol. 12
• What digital health technology types are used in mental health prevention and intervention? Review of systematic reviews for systematization of technologies 9 November 2023 | Journal of Occupational Health, Vol. 66, No. 1
• Nature-Based Education for Facilitating Resilience and Well-Being among Youth—A Nordic Perspective 29 December 2023 | Education Sciences, Vol. 14, No. 1
• Sporting Mind: The Interplay of Physical Activity and Psychological Health 22 January 2024 | Sports, Vol. 12, No. 1
• Lessons and Untapped Potential of Smartphone-Based Physical Activity Interventions for Mental Health: Narrative Review 15 March 2024 | JMIR mHealth and uHealth, Vol. 12
• Longitudinal Assessment of Seasonal Impacts and Depression Associations on Circadian Rhythm Using Multimodal Wearable Sensing: Retrospective Analysis 28 June 2024 | Journal of Medical Internet Research, Vol. 26
• Comparing resting state and task-based EEG using machine learning to predict vulnerability to depression in a non-clinical population 8 May 2023 | Scientific Reports, Vol. 13, No. 1
• Dissecting the causal association between social or physical inactivity and depression: a bidirectional two-sample Mendelian Randomization study 8 June 2023 | Translational Psychiatry, Vol. 13, No. 1
• Exploring the role of esketamine in alleviating depressive symptoms in mice via the PGC-1α/irisin/ERK1/2 signaling pathway 3 October 2023 | Scientific Reports, Vol. 13, No. 1
• The effectiveness of mindfulness yoga on patients with major depressive disorder: a systematic review and meta-analysis of randomized controlled trials 8 September 2023 | BMC Complementary Medicine and Therapies, Vol. 23, No. 1
• Embedding action on social and structural determinants of mental health into a national framework: An “immunisation schedule” for the prevention of common mental disorders Mental Health & Prevention, Vol. 32
• Active commuting associations with BMI and self-rated health: a cross-sectional analysis of the Healthy Ireland survey 8 September 2022 | Journal of Public Health, Vol. 31, No. 11
• Association between physical activity and depressive symptoms in adolescents: A prospective cohort study Psychiatry Research, Vol. 329
• Mitochondrial dysfunction: A fatal blow in depression Biomedicine & Pharmacotherapy, Vol. 167
• INVESTIGATION OF THE RELATIONSHIP BETWEEN PSYCHOLOGICAL RESILIENCE LEVELS OF TAEKWONDO PLAYERS AND THEIR LEVELS OF STATE AND TRAIT ANXIETYINVESTIGATION OF THE RELATIONSHIP BETWEEN PSYCHOLOGICAL RESILIENCE LEVELS OF TAEKWONDO PLAYERS AND THEIR LEVELS OF STATE AND TRAIT ANXIETY 26 October 2023 | Akdeniz Spor Bilimleri Dergisi
• Association between Perceived Psychological Stress and Exercise Behaviors: A Cross-Sectional Study Using the Survey of National Physical Fitness 14 October 2023 | Life, Vol. 13, No. 10
• Beden Eğitimi Dersinin Öğrencilerin Mutluluk Düzeyine Ve Beden Takdirine İlişkin Etkilerinin Çeşitli Değişkenlere Göre İncelenmesi 25 September 2023 | Akdeniz Spor Bilimleri Dergisi
• Effects of Exercise on Functional Recovery in Patients with Bipolar Depression: A Study Protocol for a Randomized Controlled Trial 30 August 2023 | Metabolites, Vol. 13, No. 9
• Changes in Body Mass Index and Their Associations with Psychological Distress, Worries, and Emotional Eating during the COVID-19 Pandemic: A Norwegian Cohort Study 2 September 2023 | Nutrients, Vol. 15, No. 17
• The moderating role of physical activity on the relationship between work intensity and depressive symptoms among the employees SSM - Population Health, Vol. 23
• Physical activity, sleep, and mental health during the COVID-19 pandemic: A one-year longitudinal study of Spanish university students Heliyon, Vol. 9, No. 9
• Depression and lifestyle: Focusing on nutrition, exercise, and their possible relevance to molecular mechanisms 25 April 2023 | Psychiatry and Clinical Neurosciences, Vol. 77, No. 8
• Resilience and self-compassion among persons with depressive disorders: Prerequisite for a positive mental health approach 29 August 2023 | Journal of Family Medicine and Primary Care, Vol. 12, No. 8
• Exercise for Mental Well-Being: Exploring Neurobiological Advances and Intervention Effects in Depression 4 July 2023 | Life, Vol. 13, No. 7
• How school closures affected learning and the physical and mental health of Chinese university students during the COVID-19 pandemic? Building and Environment, Vol. 229
• Sport and physical exercise in sustainable mental health care of common mental disorders: Lessons from the COVID-19 pandemic Sports Medicine and Health Science, Vol. 5, No. 2
• Düzenli Fiziksel Aktivitenin Stres, Kaygı, Depresyon, Yaşam Memnuniyeti, Psikolojik İyi Oluş ve Pozitif-Negatif Duygu Üzerine Etkisi 24 May 2023 | Akdeniz Spor Bilimleri Dergisi
• Symptoms of depression and anxiety among elite high school student-athletes in Sweden during the COVID-19 pandemic: A repeated cross-sectional study 1 August 2023 | Journal of Sports Sciences, Vol. 41, No. 9
• Physical activity domains and incident clinical depression: A 4-year follow-up analysis from the ELSA-Brasil cohort Journal of Affective Disorders, Vol. 329
• The Relationship Between Weight Stigmatization, Avoidance, Enjoyment and Participation in Physical Activity and Sport, and Psychological Distress American Journal of Health Behavior, Vol. 47, No. 2
• Psychological Ill Health 28 April 2023
• A proposal for emotional intelligence development through dance movement therapy 1 June 2022 | Body, Movement and Dance in Psychotherapy, Vol. 18, No. 2
• The associations between body dissatisfaction, exercise intensity, sleep quality, and depression in university students in southern China 20 March 2023 | Frontiers in Psychiatry, Vol. 14
• An insight into physical activity across domains: implications for depression and perceived health 16 March 2021 | European Sport Management Quarterly, Vol. 23, No. 2
• Impact of prestroke physical activity and citalopram treatment on poststroke depressive symptoms: a secondary analysis of data from the TALOS randomised controlled trial in Denmark 30 March 2023 | BMJ Open, Vol. 13, No. 3
• Associations of parent-child exercise with family relations and parental mental health during the COVID-19 pandemic Journal of Affective Disorders, Vol. 324
• Sugar-sweetened beverage consumption retarded weight gain but not induced depression and anxiety-like behaviors in mice Life Sciences, Vol. 317
• Physical Activity Level amongst University Students and Lecturers across Majors and Programs in Indonesia 28 February 2023 | Physical Education Theory and Methodology, Vol. 23, No. 1
• COVID-19 pandemic experiences of students from BAME and White ethnic groups pursuing higher education in the UK: A qualitative comparative exploration 23 February 2023 | Frontiers in Psychiatry, Vol. 13
• Modern Methods of Diagnostics and Treatment of Neurodegenerative Diseases and Depression 3 February 2023 | Diagnostics, Vol. 13, No. 3
• The Principal's Strategy for Strengthening National Identity in Globalization Era 31 January 2023 | IJORER : International Journal of Recent Educational Research, Vol. 4, No. 1
• Female rats are more responsive than are males to the protective effects of voluntary physical activity against the behavioral consequences of inescapable stress 23 August 2023 | Stress, Vol. 26, No. 1
• NLRP3 Inflammasome: From Pathophysiology to Therapeutic Target in Major Depressive Disorder 21 December 2022 | International Journal of Molecular Sciences, Vol. 24, No. 1
• A Cross-Sectional Study Exploring the Physical Activity Levels of Afghans and Other South Asian Youth in the UK 7 January 2023 | International Journal of Environmental Research and Public Health, Vol. 20, No. 2
• Does the COVID-pandemic affect the educational and financial inequality in weekly sport participation in the Netherlands? 26 December 2022 | Contemporary Social Science, Vol. 18, No. 1
• Medication for Depression: Monoamine Enhancers and Esketamine (Antidepressants) 6 May 2023
• Prävention 24 March 2023
• Psychiatric Epidemiology 2 June 2023
• Exercise, Depression–Anxiety Disorders, and Sex Hormones 6 June 2023
• Sport: Wie Sie von den Vorteilen eines aktiven Lebensstils profitieren 20 July 2023
• The inter‐relatedness and demographic predictors of physical activity, self‐rated health, and mental well‐being: A three‐wave study in secondary school children 15 November 2022 | Journal of Adolescence, Vol. 95, No. 1
• VCOACH: A Virtual Coaching System Based on Visual Streaming Procedia Computer Science, Vol. 222
• Associação entre sintomas depressivos e inatividade física em trabalhadores técnico-administrativos de uma universidade pública do Nordeste do Brasil: estudo transversal 1 January 2023 | Revista Brasileira de Saúde Ocupacional, Vol. 48
• Association between depressive symptoms and physical inactivity in technical-administrative workers from a public university in Northeast Brazil: a cross-sectional study 1 January 2023 | Revista Brasileira de Saúde Ocupacional, Vol. 48
• Effects of linear periodization of combined training on quality of life of adults with obesity: a blind randomized controlled trial 16 February 2022 | Scientific Reports, Vol. 12, No. 1
• Association between changes in handgrip strength and depression in Korean adults: a longitudinal panel study 11 August 2022 | Scientific Reports, Vol. 12, No. 1
• Restricted cafeteria feeding and treadmill exercise improved body composition, metabolic profile and exploratory behavior in obese male rats 15 November 2022 | Scientific Reports, Vol. 12, No. 1
• Identifying long-term and imminent suicide predictors in a general population and a clinical sample with machine learning 15 February 2022 | BMC Psychiatry, Vol. 22, No. 1
• The impact of the COVID-19 lockdown on depression sufferers: a qualitative study from the province of Zaragoza, Spain 18 April 2022 | BMC Public Health, Vol. 22, No. 1
• Physical activity and healthcare utilization in France: evidence from the European Health Interview Survey (EHIS) 2014 15 July 2022 | BMC Public Health, Vol. 22, No. 1
• The knowledge, barriers and opportunities to improve nutrition and physical activity amongst young people attending an Australian youth mental health service: a mixed-methods study 17 June 2022 | BMC Health Services Research, Vol. 22, No. 1
• Swimming exercise ameliorates depressive-like behavior by anti-inflammation activity, rebalancing gut Escherichia coli and Lactobacilli Brain Research, Vol. 1797
• Understanding the Complex Interactions between Coffee, Tea Intake and Neurologically Relevant Tissues Proteins in the Development of Anxiety and Depression 3 December 2022 | The journal of nutrition, health & aging, Vol. 26, No. 12
• Psychometric Evaluation of the Revised Exercise Addiction Inventory (EAI-R) Among Chinese College Students 22 November 2022 | International Journal of Mental Health and Addiction, Vol. 7
• Depression symptoms and quality of life in empty-nest elderly among Chengdu: A cross-sectional study 8 November 2022 | Frontiers in Psychiatry, Vol. 13
• Resilience resources and coping strategies of COVID-19 female long haulers: A qualitative study 4 November 2022 | Frontiers in Public Health, Vol. 10
• Intervention of Physical Activity for University Students with Anxiety and Depression during the COVID-19 Pandemic Prevention and Control Period: A Systematic Review and Meta-Analysis 20 November 2022 | International Journal of Environmental Research and Public Health, Vol. 19, No. 22
• Household Physical Activity for Adults in the Context of the Pandemic: A Systematic Review 17 November 2022 | Sustainability, Vol. 14, No. 22
• Functional significance of the dorsolateral prefrontal cortex during exhaustive exercise Biological Psychology, Vol. 175
• Associated Determinants and Music Genres in A Few Fitness Facilities 20 November 2022 | Journal of Humanities,Music and Dance, No. 26
• Exploring the self-reported physical fitness and self-rated health, mental health disorders, and body satisfaction among Chinese adolescents: A cross-sectional study 15 September 2022 | Frontiers in Psychology, Vol. 13
• Depressive Symptoms Are Associated With Accelerometer-Measured Physical Activity and Time in Bed Among Working-Aged Men and Women Journal for the Measurement of Physical Behaviour, Vol. 5, No. 3
• Assessment of the relationship between living alone and the risk of depression based on longitudinal studies: A systematic review and meta-analysis 30 August 2022 | Frontiers in Psychiatry, Vol. 13
• Algorithm for Evaluating the Intervention Effect of Physical Exercise on Stress Groups Mobile Information Systems, Vol. 2022
• Effects of tai chi cardiac rehabilitation program on anxiety and depression in patients with coronary heart disease: A randomized controlled clinical trial European Journal of Integrative Medicine, Vol. 53
• Exercise more efficiently regulates the maturation of newborn neurons and synaptic plasticity than fluoxetine in a CUS-induced depression mouse model Experimental Neurology, Vol. 354
• The impact of physical activity rate on subjective well-being among North-Western Italian population during COVID-19 nationwide lockdown The Journal of Sports Medicine and Physical Fitness, Vol. 62, No. 8
• Are You Happy? A Validation Study of a Tool Measuring Happiness 19 August 2022 | Behavioral Sciences, Vol. 12, No. 8
• Effectiveness of a lifestyle modification programme in the treatment of depression symptoms in primary care 26 July 2022 | Frontiers in Medicine, Vol. 9
• Effects of Mood Regulation on Sociodemographic Status, Exercise Pattern, and Physical Conditions of Chinese Adults and the Elderly Computational and Mathematical Methods in Medicine, Vol. 2022
• Influence of motivational placebo-related factors on the effects of exercise treatment in depressive adolescents 12 March 2021 | European Child & Adolescent Psychiatry, Vol. 31, No. 7
• Physical Activity, Life Satisfaction, Stress Perception and Coping Strategies of University Students in Belarus during the COVID-19 Pandemic 15 July 2022 | International Journal of Environmental Research and Public Health, Vol. 19, No. 14
• Getting to the Bottom of Saddle Sores: A Scoping Review of the Definition, Prevalence, Management and Prevention of Saddle Sores in Cycling 30 June 2022 | International Journal of Environmental Research and Public Health, Vol. 19, No. 13
• Association Between Physical Activity and Risk of Depression JAMA Psychiatry, Vol. 79, No. 6
• Effects of Taijiquan and Qigong exercises on depression and anxiety levels in patients with substance use disorders: A systematic review and meta-analysis Sports Medicine and Health Science, Vol. 4, No. 2
• Prior short-term exercise prevents behavioral and biochemical abnormalities induced by single prolonged stress in a rat model of posttraumatic stress disorder Behavioural Brain Research, Vol. 428
• The impact of physical activity on well-being, lifestyle and health promotion in an era of COVID-19 and SARS-CoV-2 variant 30 March 2022 | Postgraduate Medicine, Vol. 134, No. 4
• Tratamento para o transtorno da ansiedade generalizada por meio da prática do exercício físico: uma revisão sistemática 4 May 2022 | Revista Científica Multidisciplinar Núcleo do Conhecimento
• Treatment for generalized anxiety disorder through physical exercise: a systematic review 4 May 2022 | Revista Científica Multidisciplinar Núcleo do Conhecimento
• Tratamiento del trastorno de ansiedad generalizada a través del ejercicio físico: una revisión sistemática 4 May 2022 | Revista Científica Multidisciplinar Núcleo do Conhecimento
• Effects of Nutritional Health Risk and Physical Activity on Depressive Symptoms in Older Adults 31 May 2022 | Exercise Science, Vol. 31, No. 2
• Impact of High-Intensity Interval Training on Body Composition and Depressive Symptoms in Adults under Home Confinement 18 May 2022 | International Journal of Environmental Research and Public Health, Vol. 19, No. 10
• To drink, or to exercise: That is (not) the question! Daily effects of alcohol consumption and exercise on well‐being 18 November 2021 | Applied Psychology: Health and Well-Being, Vol. 14, No. 2
• Enhancing Mental Health, Well-Being and Active Lifestyles of University Students by Means of Physical Activity and Exercise Research Programs 25 April 2022 | Frontiers in Public Health, Vol. 10
• Changes in anxiety and depression levels after two exercises programs with women attending the Health Academy Program 6 June 2022 | Revista Brasileira de Atividade Física & Saúde, Vol. 27
• Emotional distress in young adults during the COVID-19 pandemic: evidence of risk and resilience from a longitudinal cohort study 23 June 2020 | Psychological Medicine, Vol. 52, No. 5
• State of Knowledge on Molecular Adaptations to Exercise in Humans: Historical Perspectives and Future Directions 9 March 2022
• Körperliche Aktivität und psychische Gesundheit: Fokus Alter Praxis, Vol. 111, No. 4
• Neural Correlates of Physical Activity Moderate the Association Between Problematic Mobile Phone Use and Psychological Symptoms 1 February 2022 | Frontiers in Behavioral Neuroscience, Vol. 15
• Movement behaviors and mental health of caregivers of preschoolers in China during the COVID-19 pandemic Preventive Medicine, Vol. 155
• Reflections of parents of children with 22q11.2 Deletion Syndrome on the experience of receiving psychiatric genetic counseling: ‘Awareness to Act’ 5 July 2021 | Journal of Genetic Counseling, Vol. 31, No. 1
• Psychological Health 31 August 2022
• Ausdauersport und das Gehirn 24 April 2022
• Health and Well-Being: What Can Everyone Do? 1 January 2022
• Endurance Sports and The Brain 30 August 2022
• Exercise and menopause
• Physical activity in women with depression in the Spanish national health survey: a mediation analysis 20 October 2022 | MOJ Sports Medicine, Vol. 5, No. 3
• Association Between Emotional Eating, Sociodemographic Characteristics, Physical Activity, Sleep Duration, and Mental and Physical Health in Young Adults 1 December 2022 | Journal of Multidisciplinary Healthcare, Vol. Volume 15
• Effectiveness and Benefits of Exercise on Older People Living With Mental Illness’ Physical and Psychological Outcomes in Regional Australia: A Mixed-Methods Study Journal of Aging and Physical Activity, Vol. 25
• The Effect and Mechanism of Exercise Training on Rats with Poststroke Depression Based on the Intestinal Flora Computational and Mathematical Methods in Medicine, Vol. 2021
• Effects of Personality Characteristics of Different Genders and Quality of Life Analysis on Risk Factors for Suicide in Depressive Patients Computational and Mathematical Methods in Medicine, Vol. 2021
• Association between circadian activity rhythms and mood episode relapse in bipolar disorder: a 12-month prospective cohort study 13 October 2021 | Translational Psychiatry, Vol. 11, No. 1
• Life satisfaction and depressive symptoms of mentally active older adults in Poland: a cross-sectional study 18 August 2021 | BMC Geriatrics, Vol. 21, No. 1
• Risk and protective factors of depression in the general population during the COVID-19 epidemic in Korea 8 September 2021 | BMC Psychiatry, Vol. 21, No. 1
• Mental health interventions following COVID-19 and other coronavirus infections: a systematic review of current recommendations and meta-analysis of randomized controlled trials 1 December 2021 | Brazilian Journal of Psychiatry, Vol. 43, No. 6
• A randomised controlled trial of energetic activity for depression in young people (READY): a multi-site feasibility trial protocol 4 January 2021 | Pilot and Feasibility Studies, Vol. 7, No. 1
• Is Physical Activity Protective against Emotional Eating Associated Factors during the COVID-19 Pandemic? A Cross-Sectional Study among Physically Active and Inactive Adults 28 October 2021 | Nutrients, Vol. 13, No. 11
• Increased metabolic variability is associated with newly diagnosed depression: A nationwide cohort study Journal of Affective Disorders, Vol. 294
• Roger S. McIntyre , M.D.
• Exploring the meaning of life from the perspective of patients with depression: A phenomenological study Archives of Psychiatric Nursing, Vol. 35, No. 5
• Online Education in the Time of COVID-19 Crisis 7 October 2021 | International Journal of Engineering Technologies and Management Research, Vol. 8, No. 9
• Learning oriented physiotherapy (LOP) in anxiety and depression: an 18 months multicentre randomised controlled trial (RCT) 19 March 2020 | European Journal of Physiotherapy, Vol. 23, No. 5
• Integrative health services use for depression in middle-aged and older Australian women European Journal of Integrative Medicine, Vol. 46
• How Exercise Protects Against Mild Cognitive Impairment in Nursing Home–Dwelling Older Adults Journal of Nervous & Mental Disease, Vol. 209, No. 9
• The Views of People Who Experienced the Reduction of the Concerns in the COVID-19 Quarantine Process by Making Physical Activity at Home 5 July 2021 | Social Work in Public Health, Vol. 36, No. 6
• Physical Exercise: An Overview of Benefits From Psychological Level to Genetics and Beyond 12 August 2021 | Frontiers in Physiology, Vol. 12
• Biopsychosocial Functions of Human Walking and Adherence to Behaviourally Demanding Belief Systems: A Narrative Review 4 August 2021 | Frontiers in Psychology, Vol. 12
• Changes in mood and health‐related quality of life in Look AHEAD 6 years after termination of the lifestyle intervention 14 July 2021 | Obesity, Vol. 29, No. 8
• Effects of Physical Activity and Training Routine on Mental Health During the COVID-19 Pandemic and Curfew 3 June 2021 | Frontiers in Psychology, Vol. 12
• Differences in objectively measured daily physical activity patterns related to depressive symptoms in community dwelling women – mPED trial Preventive Medicine Reports, Vol. 22
• Hippocampal PGC-1α-mediated positive effects on parvalbumin interneurons are required for the antidepressant effects of running exercise 15 April 2021 | Translational Psychiatry, Vol. 11, No. 1
• Mental health during the COVID-19 pandemic and beyond: The importance of the vagus nerve for biopsychosocial resilience Neuroscience & Biobehavioral Reviews, Vol. 125
• Exercise is a Life-Changer for Those with Autism ACSM'S Health & Fitness Journal, Vol. 25, No. 3
• Trajectories and associations between depression and physical activity in patients with cardiovascular disease during participation in an internet-based cognitive behavioural therapy programme 5 August 2020 | European Journal of Cardiovascular Nursing, Vol. 20, No. 2
• Psychological distress in the academic population and its association with socio-demographic and lifestyle characteristics during COVID-19 pandemic lockdown: Results from a large multicenter Italian study 10 March 2021 | PLOS ONE, Vol. 16, No. 3
• Does context moderate physical activity's relations with depression?: A cross-sectional study Mental Health and Physical Activity, Vol. 20
• Sports participation and quality of life in individuals with visual impairment 25 June 2020 | Irish Journal of Medical Science (1971 -), Vol. 190, No. 1
• Physical Activity, Mental Health and Consumption of Medications in Pre-Elderly People: The National Health Survey 2017 26 January 2021 | International Journal of Environmental Research and Public Health, Vol. 18, No. 3
• The Development Strategy of Home-Based Exercise in China Based on the SWOT-AHP Model 29 January 2021 | International Journal of Environmental Research and Public Health, Vol. 18, No. 3
• Real-time fMRI neurofeedback amygdala training may influence kynurenine pathway metabolism in major depressive disorder NeuroImage: Clinical, Vol. 29
• The 2020 Royal Australian and New Zealand College of Psychiatrists clinical practice guidelines for mood disorders 22 December 2020 | Australian & New Zealand Journal of Psychiatry, Vol. 55, No. 1
• Exercise, Flourishing, and the Positivity Ratio in Seventh-Day Adventists: A Prospective Study 17 June 2020 | American Journal of Health Promotion, Vol. 35, No. 1
• Aerobic Exercise Prevents Depression via Alleviating Hippocampus Injury in Chronic Stressed Depression Rats 23 December 2020 | Brain Sciences, Vol. 11, No. 1
• The association between prior physical fitness and depression in young adults during the COVID-19 pandemic—a cross-sectional, retrospective study 11 March 2021 | PeerJ, Vol. 9
• Effects of Physical Exercise on Neuroplasticity and Brain Function: A Systematic Review in Human and Animal Studies Neural Plasticity, Vol. 2020
• Acute exercise enhances fear extinction through a mechanism involving central mTOR signaling Neurobiology of Learning and Memory, Vol. 176
• Sex differences in type of exercise associated with depression in South Korean adults 26 October 2020 | Scientific Reports, Vol. 10, No. 1
• Factors Associated with Psychological Distress and Physical Activity During the COVID-19 Pandemic Health Security, Vol. 18, No. 6
• Effectiveness and cost-effectiveness of a lifestyle modification programme in the prevention and treatment of subclinical, mild and moderate depression in primary care: a randomised clinical trial protocol 28 December 2020 | BMJ Open, Vol. 10, No. 12
• Prevalence, clustering and combined effects of lifestyle behaviours and their association with health after retirement age in a prospective cohort study, the Nord-Trøndelag Health Study, Norway 10 June 2020 | BMC Public Health, Vol. 20, No. 1
• Veterinary-prescribed physical activity promotes walking in healthy dogs and people 1 December 2020 | BMC Veterinary Research, Vol. 16, No. 1
• Long-term effects of group exercise intervention on maximal step-up height in middle-aged female primary care patients with obesity and other cardio-metabolic risk factors 16 March 2020 | BMC Sports Science, Medicine and Rehabilitation, Vol. 12, No. 1
• Environmental exposures and health behavior in association with mental health: a study design 21 October 2020 | Archives of Public Health, Vol. 78, No. 1
• Associations Between Doing Planned Exercise and Probable Major Depressive Disorder in Individuals Following Spinal Cord Injury Topics in Spinal Cord Injury Rehabilitation, Vol. 26, No. 1
• The experience of coaching whilst walking: A pilot study 1 December 2020 | The Coaching Psychologist, Vol. 16, No. 2
• Nutrition and depression: Summary of findings from the EU‐funded MooDFOOD depression prevention randomised controlled trial and a critical review of the literature 15 July 2020 | Nutrition Bulletin, Vol. 45, No. 4
• Physical Inactivity Is Associated With Increased Levels of Anxiety, Depression, and Stress in Brazilians During the COVID-19 Pandemic: A Cross-Sectional Study 17 November 2020 | Frontiers in Psychiatry, Vol. 11
• Beneficial Effects of Exercise on Depression and Anxiety During the Covid-19 Pandemic: A Narrative Review 4 November 2020 | Frontiers in Psychiatry, Vol. 11
• Psychological, addictive, and health behavior implications of the COVID-19 pandemic Behaviour Research and Therapy, Vol. 134
• Major depressive disorder due to social isolation during the coronavirus disease 2019 pandemic: an elderly woman's first episode of major depressive disorder 14 September 2020 | Psychogeriatrics, Vol. 20, No. 6
• Effective strategies for prevention of cardiovascular complications in elderly patients with arterial hypertension and obesity (clinical example) 15 October 2020 | Medical alphabet, No. 21
• The effects of physical activity and sedentary behavior in the associations between cardiovascular diseases and depression: A four-way decomposition Journal of Affective Disorders, Vol. 275
• Associations of exercise frequency and cardiorespiratory fitness with symptoms of depression and anxiety - a cross-sectional study of 36,595 adults Mental Health and Physical Activity, Vol. 19
• Occupational Physical Activity and Health-Related Quality of Life among Nigerian Vocational-Skilled Workers 1 October 2020 | Journal of Occupational Health and Epidemiology, Vol. 9, No. 4
• Mental health and movement behaviour during the COVID-19 pandemic in UK university students: Prospective cohort study Mental Health and Physical Activity
• Associations between 24 h Movement Behavior and Mental Health in Office Workers 27 August 2020 | International Journal of Environmental Research and Public Health, Vol. 17, No. 17
• Exercise, brain plasticity, and depression 3 June 2020 | CNS Neuroscience & Therapeutics, Vol. 26, No. 9
• Dose‐response relationship of active commuting to work: Results of the GISMO study 27 August 2020 | Scandinavian Journal of Medicine & Science in Sports, Vol. 30, No. S1
• More habitual physical activity is linked to the use of specific, more adaptive cognitive reappraisal strategies in dealing with stressful events 9 February 2020 | Stress and Health, Vol. 36, No. 3
• Effects of Exercise Training on Anxious–Depressive-like Behavior in Alzheimer Rat 6 February 2020 | Medicine & Science in Sports & Exercise, Vol. 52, No. 7
• Sex differences in resilience: Experiential factors and their mechanisms 20 December 2019 | European Journal of Neuroscience, Vol. 52, No. 1
• Reducing Allostatic Load in Depression and Anxiety Disorders: Physical Activity and Yoga Practice as Add-On Therapies 4 June 2020 | Frontiers in Psychiatry, Vol. 11
• Physical Activity of the Population of the Most Obese Country in Europe, Hungary 2 June 2020 | Frontiers in Public Health, Vol. 8
• Psychological impact of the COVID-19 pandemic on healthcare workers: a cross-sectional study in China 14 June 2020 | General Psychiatry, Vol. 33, No. 3
• Regular Physical Activity, Short-Term Exercise, Mental Health, and Well-Being Among University Students: The Results of an Online and a Laboratory Study 26 May 2020 | Frontiers in Psychology, Vol. 11
• Antidepressant Drugs and Physical Activity: A Possible Synergism in the Treatment of Major Depression? 6 May 2020 | Frontiers in Psychology, Vol. 11
• The PROSPER-HIV Study: A Research Protocol to Examine Relationships Among Physical Activity, Diet Intake, and Symptoms in Adults Living With HIV Journal of the Association of Nurses in AIDS Care, Vol. 31, No. 3
• Prospective relationships of mentally passive sedentary behaviors with depression: Mediation by sleep problems Journal of Affective Disorders, Vol. 265
• The association of physical activity and sleep duration with incident anxiety symptoms: A cohort study of 134,957 Korean adults Journal of Affective Disorders, Vol. 265
• A Network Approach to Understanding the Emotion Regulation Benefits of Aerobic Exercise 31 August 2019 | Cognitive Therapy and Research, Vol. 44, No. 1
• Developing algorithms to predict adult onset internalizing disorders: An ensemble learning approach Journal of Psychiatric Research, Vol. 121
• Who does what the cardiologist recommends? Psychosocial markers of unhealthy behavior in coronary disease patients 31 January 2020 | PLOS ONE, Vol. 15, No. 1
• Depression in Athletes: Incidence, Prevalence, and Comparisons with the Nonathletic Population 31 May 2020
• Gesundheit und Wohlbefinden – was kann jeder tun? 6 October 2020
• Exercise and Depressive Disorder 28 April 2020
• Exercise and Schizophrenia 28 April 2020
• The kynurenine pathway: a finger in every pie 12 April 2019 | Molecular Psychiatry, Vol. 25, No. 1
• Nature sports, health and ageing: the value of euphoria 3 June 2018 | Annals of Leisure Research, Vol. 23, No. 1
• Passive Versus Mentally Active Sedentary Behaviors and Depression 29 October 2019 | Exercise and Sport Sciences Reviews, Vol. 48, No. 1
• School-based physical activity interventions, physical fitness and mental health among adolescents; Effects, associations and lessons learned from the School in Motion study 1 January 2021 | , Vol. 1
• Management of nonarthritic meniscal tears: a qualitative, patient-centered assessment of decision making 1 July 2020 | Current Orthopaedic Practice, Vol. 31, No. 4
• Regular Moderate- to Vigorous-Intensity Physical Activity Rather Than Walking Is Associated with Enhanced Cognitive Functions and Mental Health in Young Adults 18 January 2020 | International Journal of Environmental Research and Public Health, Vol. 17, No. 2
• Association of living arrangements with depressive symptoms among older adults in China: a cross-sectional study 29 July 2019 | BMC Public Health, Vol. 19, No. 1
• Running exercise protects oligodendrocytes in the medial prefrontal cortex in chronic unpredictable stress rat model 28 November 2019 | Translational Psychiatry, Vol. 9, No. 1
• Association between use of oral hypoglycemic agents in Japanese patients with type 2 diabetes mellitus and risk of depression: A retrospective cohort study 21 November 2019 | Pharmacology Research & Perspectives, Vol. 7, No. 6
• The Impact of Exercise, Lifestyle, and Clinical Factors on Perceived Cognitive Function in Patients with Rheumatoid Arthritis: Results from a Prospective Cohort Study 24 October 2019 | ACR Open Rheumatology, Vol. 1, No. 10
• Health Benefits of Physical Activity: A Strengths-Based Approach 21 November 2019 | Journal of Clinical Medicine, Vol. 8, No. 12
• The Role of Leisure-Time Physical Activity in the Change of Work-Related Stress (ERI) over Time 2 December 2019 | International Journal of Environmental Research and Public Health, Vol. 16, No. 23
• Metabolic profiling identifies TC and LDL as potential serum biomarkers for depressive symptoms in schizophrenia Psychiatry Research, Vol. 281
• Long distance ski racing is associated with lower long-term incidence of depression in a population based, large-scale study Psychiatry Research, Vol. 281
• Physical Activity and Fitness in White- and Blue-Collar Retired Men 18 December 2019 | American Journal of Men's Health, Vol. 13, No. 6
• Voluntary exercise enables stress resistance in females Behavioural Brain Research, Vol. 369
• The evidence for physical activity in the management of major mental illnesses Current Opinion in Psychiatry, Vol. 32, No. 5
• Investigating Exercise for Anxiety and Depression Treatment: A Case Report Journal of Clinical Exercise Physiology, Vol. 8, No. 3
• Physical Activity and Depression ACSM'S Health & Fitness Journal, Vol. 23, No. 5
• Retracted : Efficacy of vortioxetine combined cognitive behaviour intervention therapy on brain‐derived neurotrophic factor level on depressive patients 19 February 2019 | Psychogeriatrics, Vol. 19, No. 5
• Mental health informed physical activity for first responders and their support partner: a protocol for a stepped-wedge evaluation of an online, codesigned intervention 11 September 2019 | BMJ Open, Vol. 9, No. 9
• Cross-sectional and longitudinal association of non-exercise estimated cardiorespiratory fitness with depression and anxiety in the general population: The HUNT study Journal of Affective Disorders, Vol. 252
• LDL cholesterol relates to depression, its severity, and the prospective course Progress in Neuro-Psychopharmacology and Biological Psychiatry, Vol. 92
• Is sport an untapped resource for recovery from first episode psychosis? A narrative review and call to action 24 July 2018 | Early Intervention in Psychiatry, Vol. 13, No. 3
• Exploring how medical students learn during clinical rotations: a pilot study with a mobile application 27 February 2019 | Health and Technology, Vol. 9, No. 3
• Variation in the prevalence of depression and patterns of association, sociodemographic and lifestyle factors in community-dwelling older adults in six low- and middle-income countries Journal of Affective Disorders, Vol. 251
• Lactate is an antidepressant that mediates resilience to stress by modulating the hippocampal levels and activity of histone deacetylases 8 January 2019 | Neuropsychopharmacology, Vol. 44, No. 6
• The Transcriptional Signature of a Runner’s High Medicine & Science in Sports & Exercise, Vol. 51, No. 5
• Examining the Effects of Exercise on Pattern Separation and the Moderating Effects of Mood Symptoms Behavior Therapy, Vol. 50, No. 3
• Physical Activity Intensity is Associated with Symptom Distress in the CNICS Cohort 27 October 2018 | AIDS and Behavior, Vol. 23, No. 3
• JAMA Psychiatry, Vol. 76, No. 4
• Depression and Anxiety, Vol. 36, No. 9
• American Journal of Preventive Medicine, Vol. 56, No. 5
• Behavioural Brain Research, Vol. 374
• Brain Research, Vol. 1722
• Current Opinion in Psychology, Vol. 28
• Experimental Neurology, Vol. 315
• General Hospital Psychiatry, Vol. 60
• Journal of Affective Disorders, Vol. 251
• SSM - Population Health, Vol. 8
• Prospective associations between physical activity and clinician diagnosed major depressive disorder in adults: A 13-year cohort study Preventive Medicine, Vol. 118
• Orthopaedic Journal of Sports Medicine, Vol. 7, No. 3
• PLOS ONE, Vol. 14, No. 11
• Journal of Korean Medical Science, Vol. 34, No. 7
• Frontiers in Behavioral Neuroscience, Vol. 13
• Passive and mentally-active sedentary behaviors and incident major depressive disorder: A 13-year cohort study Journal of Affective Disorders, Vol. 241
• Frequency and pattern of exercise and depression after two years in older Japanese adults: the JAGES longitudinal study 25 July 2018 | Scientific Reports, Vol. 8, No. 1
• The impact of chronic fluoxetine on conditioned fear expression and hippocampal FGF2 in rats: Short- and long-term effects Neurobiology of Learning and Memory, Vol. 155
• Associations of physical activity and depression: Results from the Irish Longitudinal Study on Ageing Experimental Gerontology, Vol. 112
• Association between physical exercise and mental health in 1·2 million individuals in the USA between 2011 and 2015: a cross-sectional study The Lancet Psychiatry, Vol. 5, No. 9
• Changes in the mode of travel to work and the severity of depressive symptoms: a longitudinal analysis of UK Biobank Preventive Medicine, Vol. 112
• Working holistically: Body, mind, spirit, social world and political context 1 May 2018 | Clinical Psychology Forum, Vol. 1, No. 305
• Analysis of heart rate variability and its application in sports medicine: A review
• Gregory Simon , M.D., M.P.H.
• JAMA Psychiatry, Vol. 75, No. 9
• Current Psychiatry Reports, Vol. 20, No. 8
• International Journal of Epidemiology, Vol. 47, No. 5
• Health and Quality of Life Outcomes, Vol. 16, No. 1
• Trials, Vol. 19, No. 1
• Journal of Medical Internet Research, Vol. 20, No. 7
• Frontiers in Psychology, Vol. 9
• Frontiers in Public Health, Vol. 6
• Journal of Korean Neuropsychiatric Association, Vol. 57, No. 2
• MMW - Fortschritte der Medizin, Vol. 159, No. 20
• Mindset kinetics and some depression status: A new quantitative model under biochemical - toxicology approach? 1 January 2017 | Insights on the Depression and Anxiety, Vol. 2, No. 1

Editor Spotlight

• Mental disorders

• Search Menu
• Sign in through your institution
• Advance Articles
• Supplements
• Editor's Choice
• Virtual Issues
• Virtual Roundtables
• Abstract Supplements
• Basic Science
• Circadian Disorders
• Cognitive, Affective and Behavioral Neuroscience of Sleep
• Neurological Disorders
• Sleep Across the Lifespan
• Sleep and Metabolism
• Sleep Disordered Breathing
• Sleep Health and Safety
• Author Guidelines
• Instructions for Reviewers
• Submission Site
• Open Access Options
• Additional Resources
• Self-Archiving Policy
• Why publish with this journal?
• About SLEEP
• Editorial Board
• Dispatch Dates
• Permissions
• Advertising & Corporate Services
• Journals Career Network
• Reprints and ePrints
• Sponsored Supplements
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

Introduction, disclosure statement.

• < Previous

Sleep duration and mood in adolescents: an experimental study

• Article contents
• Figures & tables
• Supplementary Data

Stephen A Booth, Mary A Carskadon, Robyn Young, Michelle A Short, Sleep duration and mood in adolescents: an experimental study, Sleep , Volume 44, Issue 5, May 2021, zsaa253, https://doi.org/10.1093/sleep/zsaa253

• Permissions Icon Permissions

This study examines the relationship between experimentally manipulated sleep duration and mood in adolescents.

Thirty-four adolescents (20 male), aged 15–17 years, lived in a sleep laboratory for 10 days and 9 nights. They were allocated to one of three sleep “doses” for five consecutive nights for 5, 7.5, or 10 h sleep opportunity per night. Two baseline nights and two recovery nights entailed 10 h sleep opportunity per night. Mood was measured every 3 h during wake using unipolar visual analogue scales measuring the mood states “depressed,” “afraid,” “angry,” “confused,” “anxious,” “happy,” and “energetic.”

Mixed models analyses with post hoc comparisons revealed that participants in the 5-h group, but not the 7.5- or 10-h groups, reported being significantly more depressed, angry, and confused during sleep restriction than at baseline. Adolescents were significantly less happy and energetic during sleep restricted to 5 h and significantly less energetic during sleep restricted to 7.5 h. When adolescents had 10 h sleep opportunities their happiness significantly increased. No statistically significant effects of sleep restriction were found for fear or anxiety, although small-to-moderate effects of sleep restricted to 5 or 7.5 h were found. Two nights of recovery sleep was not sufficient to recover from increased negative mood states for the 5-h group, although recovery occurred for positive mood states.

Given the prevalence of insufficient sleep and the rising incidence of mood disorders and dysregulation in adolescents, these findings highlight the importance of sufficient sleep to mitigate these risks.

Adolescence is a critical maturational stage in terms of heightened risk of the onset of mood disorders. Insufficient experimental evidence exists that elucidates the effect of sleep duration on a range of positive and negative mood states in adolescents. The present study uses sleep restriction and sleep extension protocols to experimentally manipulate sleep duration in 34 adolescents. Results indicated that adolescents reported deteriorated in terms of depression, happiness, anger, confusion and energy. Two nights of recovery sleep did not eliminate mood deficits for negative mood states for the 5-h group, although recovery occurred for positive mood states. Sufficient sleep is crucial to guard against mood deficits in otherwise healthy adolescents.

Adolescence is a time of significant psychological, social, and physiological change [ 1 ] and a vulnerable developmental period during which individuals are at heightened risk of developing a mental illness [ 2 ]. Despite the importance of mood and the commonly held belief that sleep loss perturbs adolescent mood [ 3 , 4 ], rigorous experimental evidence supporting a causal relationship between sleep loss and mood deficits in adolescents is scant. The current experiment tests the causal association between sleep duration and adolescent mood.

Empirical literature on sleep duration and mood has overwhelmingly focused on adults [ 5–7 ]; however, adult findings may not generalize to adolescents. Adolescents differ in terms of their greater sleep need [ 8 , 9 ], types of affective challenges they face [ 10 ], and less mature prefrontal brain regions that are crucial to affective regulation [ 11 ]. As such, research focusing specifically on adolescents is needed. Most of the extant literature focusing on sleep duration and mood in adolescents is cross-sectional. These studies report a correlation of sleep duration and mood, with shorter sleep associated with worse mood [ 12 , 13 ]. Due to the cross-sectional nature of such studies, however, a casual relationship cannot be concluded.

A recent study investigated the effect of sleep restriction and sleep extension on a group of 48 adolescents aged 14–17 years with Attention Deficit Hyperactivity Disorder using a 3-week sleep protocol with an experimental crossover design in participants’ homes [ 14 ]. Sleep restriction involved a week with a 6.5 h sleep opportunity per night while sleep extension involved a week of 9.5 h sleep opportunities per night. Parent- and self-reported depressive symptoms were greater during sleep restriction compared to sleep extension, while positive affect was lower. Parents also reported increased negative affect and emotion dysregulation among adolescents during sleep restriction when compared to extension.

Among the limited experimental studies including adolescents, two were performed under continuous monitoring. The sleep of 113 adolescents, aged 15–19 years, was restricted to a 5-h sleep opportunity for seven consecutive nights, with or without a 1-h daytime nap opportunity [ 15 , 16 ]. Mood was assessed three times per day using Positive and Negative Affect Scales (PANAS). Compared to controls, who had 9-h sleep opportunity per night, sleep-restricted adolescents reported significantly lower positive mood scores. The detriment to positive mood was only partially ameliorated by an afternoon nap. Unexpectedly, no change was observed to negative mood. Lo et al. note that adolescents reported that many of the mood states assessed by the PANAS negative mood subscale, such as guilty, afraid, and scared, were not relevant to them [ 15 ]. Differential sensitivity to sleep loss among mood states has been reported in a recent meta-analysis of 361,505 adolescents [ 17 ], with positive mood showing the largest effect in response to shorter sleep (OR = 1.02), followed by anger (OR = .83), depression (OR = .62), and anxiety (OR = .41). The present literature is limited by the paucity of studies that have examined specific mood states [ 18 ]. Other challenges arise in those studies using home-based sleep restriction schedules where breaches of adherence to study protocols regarding sleep and abstinence from napping and caffeine may muddle interpretation [ 19 , 20 ].

The present study addresses the gaps and limitations in current literature by employing a laboratory-based experimental design to measure several discrete mood states over varying “doses” of sleep. The laboratory environment ensures adherence to study protocols and control of environmental variables, such as diet, caffeine consumption, and exercise that are known to affect sleep and mood [ 21–23 ]. The repeated-measures design whereby participants in each condition have their mood compared between baseline, the sleep dose condition, and following recovery sleep provides the opportunity for robust conclusions regarding the causal link between sleep duration and mood.

We hypothesized that self-reported mood will be significantly worse during sleep restriction when compared to extended 10-h baseline and recovery sleep opportunity, with positive moods decreasing and negative moods increasing when sleep is restricted to 5- or 7.5-h sleep opportunity per night, as neither duration allows for sleep that is within the recommended range for adolescents [ 24 ] (i.e. 8–10 h sleep per night).

Participants

Participants were 34 adolescents aged 15–17 years (20 male, M age = 15.91 years ± 0.86) from South Australian high schools. All participants were late- or postpubertal adolescents (Tanner Stage 4 or 5 on the Pubertal Development Scale) [ 25 ]. Prescreening by parent and adolescent self-report, showed that participants were physically and psychologically healthy and were medication-free, with the exception of birth control. Participants were good sleepers, with average sleep durations ≥8 h per night [ 26 ], average sleep onset latencies of ≤30 min per night [ 27 ], and weeknight/weekend bedtime discrepancy less than 2 h [ 28 ], as determined by a survey and a 7-day sleep diary during screening, to reduce confounding effects of preexisting poor sleep and/or sleep disorders. Extreme morning or evening chronotypes were not included (≥44 or ≤22 on the Composite Morningness/Eveningness Scale) [ 29 ] due to effects of chronotype on adolescent sleep and mood [ 30 ], nor were participants indicating less than 8 h sleep per night on average, weekend bedtime delay of 2 h or more, or sleep onset latencies greater than 30 min per night.

Materials and measures

Mood was measured using a series of 100 mm unipolar visual analog scale (VAS), similar to those used in a study by Stern et al. [ 31 ]. These consisted of scales with labels to demonstrate the spectrum of discrete mood experience, with “Not at all (mood)” at the left endpoint and “Extremely (mood)” on the right endpoint. Mood states included “Depressed,” “Afraid,” “Angry,” “Confused,” “Anxious,” “Happy,” and “Energetic.”

At each test administration, participants were asked to describe how they felt “RIGHT NOW” by marking a short vertical line intersecting the scale at some point, placing the mark further to the right to reflect the greater intensity of that mood. Items were scored by measuring the distance in millimeters from the left anchor of the scale to the point at which the participant intersects the scale with a line. Unipolar scales (i.e. spanning the range of not at all sad to extremely sad, as opposed to bipolar scales which may span happy to sad) were selected as they reduce participant confusion as compared to a bipolar scale [ 31 ].

Sleep was recorded prior to the laboratory experiment with sleep diaries, actigraphy, and by getting participants to call and leave a message on the sleep laboratory answering machine confirming their and waketime. During the laboratory experiment, sleep was recording using nightly polysomnography. Further information regarding sleep measurement is provided elsewhere [ 8 ].

This study used a mixed experimental design. The independent variables were sleep duration dose (5, 7.5, and 10 h groups), and sleep condition (Baseline, Experimental Sleep, and Recovery); dependent variables were the mood terms: “Depressed,” “Afraid,” “Angry,” “Confused,” “Anxious,” “Happy,” and “Energetic.”

An active recruitment process was used to recruit adolescent participants through announcements in South Australian high-school newsletters. Parents of potential participants provided an initial telephone screen using the Sleep, Medical, Educational, and Family History Survey [ 12 ]. Adolescent participants meeting preliminary inclusion criteria were sent a questionnaire package that included a modified Sleep Habits Survey, the Smith Composite Morningness/Eveningness questionnaire, the Pubertal Development Scale and the Sleep, Medical, Education and Family History Survey and a 7-day sleep diary and were invited to attend an interview. Upon confirming eligibility and obtaining parental consent and adolescent assent, adolescents were provided a sleep diary and a wrist activity monitor for the week prior to the in-lab study. Participants were allocated in blocks of four participants to one of the three sleep-dose conditions. Adolescents in each study run were in the same condition and were not informed of their sleep opportunity each night until the end of the study. Adolescents were required to maintain a 9.5 h sleep opportunity between 9:30 pm and 07:00 am for five nights prior to the study to eliminate any existing sleep debt before the start of the study. Thirty-seven participants (21 males) were recruited; however, one did not follow the required pre-study sleep protocol and was excluded from the study and two others (both females assigned to the 7.5 h condition) discontinued their participation before the study’s conclusion.

On the first two of nine consecutive laboratory nights, a 10 h sleep opportunity was provided to extinguish any residual sleep debt and provide baseline sleep and mood data. Adolescents’ polysomnographically estimated sleep durations on the adaptation and baseline nights were not significantly different between conditions (all p < .05). Five experimental nights of 5, 7.5, or 10 h sleep opportunity followed, with wake time at 7:30 am regardless of sleep duration. This wake time was chosen to simulate typical rising early for school, as adolescents generally restrict sleep by staying up late rather than getting up early [ 32 ]. Two recovery nights of 10 h sleep opportunity concluded the experiment. The VAS was administered three-hourly across all wake periods, as shown in Figure 1 .

Schematic of the study protocol, displaying days by hours over three sleep dose conditions.

To control environmental variables, participants completed the study in a laboratory free of time-cues, without access to caffeine or to live television or Internet. Participants had access to mobile phones for one 15-min period each day, although Internet access was disabled and phone clocks were altered. Participants spent their time playing board games, doing craft activities, watching prerecorded movies and television series, and interacting with one another and with research staff. The laboratory was light- (<50 lx during wake periods and <1 lx during sleep opportunities) and temperature- (21 ○ C) controlled and sound-attenuated. Participants’ scheduled sleep episodes were recorded using polysomnography to confirm participant sleep times. Findings regarding changes to sleep and attention variables are reported elsewhere [ 8 ].

Statistical analyses

Linear mixed-effects models were used to test differences in mood across sleep conditions (baseline, sleep restriction, and recovery) for both males and females. This analytic approach accounted for both within and between-participant variance [ 33 ]. All models specified a random effect of subject ID. Models for mood specified “Depressed,” “Afraid,” “Angry,” “Confused,” “Anxious,” “Happy,” and “Energetic” as dependent variables. Each model was fully saturated, including all main and interaction effects, for sleep dose (5, 7.5, and 10 h), sleep condition (baseline, sleep restriction, and recovery), and sex. Baseline data collected on day 2 were excluded to reduce acclimatization effects as participants adjusted to a novel environment. Cohen’s d was calculated to indicate effect size.

Results are presented for the interactions between sleep dose group (i.e. 5, 7.5, or 10 h) and study phase (baseline, sleep restriction, or recovery) in Tables 1–2 and Figure 2 . Noted differences refer to statistical significance at a p < .05 level. Effect sizes are presented using Cohen’s d , where .2, .5, and .8 indicate a small, medium, and large effect size, respectively.

Inferential statistics for main effects and interactions of sleep dose and sleep condition on negative mood variables

Final column displays significant post hoc comparisons ( p < .05). Post hoc comparisons for the main effect of “dose” are between subjects’ comparisons while all remaining comparisons are within-subjects.

Note. BL, baseline sleep condition; ES, experimental sleep dose; RC, recovery sleep condition; 5 h, 5 h experimental sleep dose; 7.5 h, 7.5 h experimental sleep dose; 10 h, control group with 10 h sleep dose.

Inferential statistics for the main effects and interactions of experimental sleep dose and sleep condition on positive mood variables

Effect sizes (Cohen’s d ) of changes to mood states between baseline and experimental sleep dose for the 5-, 7.5-, and 10-h sleep dose conditions.

Sleep duration and negative mood

Descriptive statistics for all mood states are provided in the Supplementary Material , and inferential statistics with post hoc test results are provided in Table 1 . Figure 2 illustrates the effect size of the changes in mood states between baseline and experimental sleep dose across the three groups. Overall, the change to mood states across the phases of the study (baseline, experimental sleep dose, and recovery) varied between experimental sleep dose groups for depressed mood, anger, and confusion, but not fear or anxiety. Specifically, participants reported greater depressed mood, anger, and confusion during the experimental sleep dose when compared to baseline, but this effect was only seen when sleep was restricted to 5 h TIB and not during the 7.5 h sleep dose. Depressed mood and fear also increased from experimental sleep dose to recovery for participants in the 5 h group. Confusion increased during recovery when compared to baseline and experimental sleep dose for participants in the 10 h group. No significant changes to other mood states were found between experimental sleep dose and recovery in the 7.5 or 10 h sleep groups.

Sleep duration and positive mood

Inferential statistics and post hoc results are provided in Table 2 . The change to mood states across the phases of the study (baseline, experimental sleep dose, and recovery) varied between sleep dose groups for both happiness and energy. Specifically, participants reported significantly reduced happiness when sleep was restricted to 5 h. They also reported significantly less energy when sleep was restricted to either 5 or 7.5 h sleep opportunities per night. Participants’ happiness increased from baseline to experimental sleep dose in the 10 h sleep dose (control group), and there was a small but nonsignificant increase in energy. Significant increases in happiness and energy also occurred between the experimental sleep period and recovery for adolescents in the 5 and 7.5 h sleep dose groups, while happiness increased between the sleep dose phase and recovery for participants in the 7.5 and 10 h groups.

The aim of the current study was to explore the effect of five nights of sleep limited to either a 5, 7.5, or 10 h sleep opportunity per night on adolescent mood, when compared with baseline and recovery conditions which had 10-h sleep opportunities. Consistent with previous research [ 15–17 , 34 ], positive moods of happiness and energy significantly decreased when sleep was restricted to 5 h sleep opportunity per night, with large effect sizes for both moods. This finding is consistent with results found in adult studies that used a VAS to measure mood [ 5 ], establishing the sensitivity of happiness and energy to sleep loss. It is interesting to note that, although obtaining less than the recommended 8–10 h sleep per night [ 24 , 26 ], positive moods only decreased for energy but not happiness in the 7.5 h group, although a small-to-medium effect size was observed. Conversely, happiness, but not energy, increased from baseline to experimental sleep dose in the 10 h control group, suggesting that when adolescents consistently have the opportunity to obtain optimal sleep, happiness increases.

Results regarding negative mood states of depression, fear, anger, confusion, and anxiety were mixed. Participants in the 5 h group were significantly more depressed, angry, and confused when restricted to 5 h TIB compared to baseline, with large effect sizes for all changes. Likewise, extant research has found that depressed mood, anger, and confusion increased in response to less sleep [ 19 , 35 ]. Similar to results found for positive moods, no changes to negative mood states were observed in the 7.5-h group between baseline and experimental sleep dose, despite not obtaining the recommended duration of sleep.

Fear and anxiety did not increase during the experimental sleep phase for adolescents in the 5 or 7.5 h sleep dose groups. Findings in regard to the sensitivty of anxiety to sleep loss have been mixed [ 34 ]. There are several possible explanations for this discrepancy. First, experimental studies have struggled to replicate cross-sectional results linking less sleep to increased negative mood in adolescents [ 18 , 19 ]. A lack of significant findings in some moods may be a result of a differential sensitivity. It is possible that the “dosage” and chronicity of sleep restriction implemented in the current study, as well as previous experimental research, were sufficient to elicit an increase in depressed mood, anger, and confusion in adolescents, but not fear or anxiety. More chronic sleep restriction may be required to find observable effects. It is important to note that while statistical significance was not reached, the effect size for the increase in anxiety from baseline to sleep restriction was medium ( d = .51) for the 5 h group and small ( d = .27) for the 7.5 h group, thus part of this nonsignificant finding may also reflect a lack of statistical power in the current analyses.

Another possible explanation for the lack of a significant relationship between sleep duration and fear and anxiety has been suggested [ 36 ]. It is argued that individual differences can predict mood responses to restricted sleep, such as a phenomenon where stressful life events influence the development of a person’s genes, making them more susceptible to mood disturbances upon reduced sleep. Restricting sleep reduces cognitive resources needed to dismiss negative stimuli, and those susceptible individuals, having biased attention for negativity, become less able to regulate or reappraise them. This is supported by Gregory et al. [ 37 ], who found that the greatest variance in the relationship between sleep duration and anxiety was accounted for by genetic factors. As the participants recruited for the current study underwent an extensive screening process, ensuring physical and psychological health, this may have minimized the possibility that individuals at higher risk of mood and/or sleep disorders would be included. Fuligni et al. have similarly found that adolescents with greater depressed mood and anxiety need more sleep for optimal mood [ 38 ]. Thus, a sample of adolescents with heightened depressed mood or anxiety at baseline may be less resilient to sleep loss. Although literature reporting this effect focuses on depressed mood and anxiety [ 36 , 37 ], it is possible that the same vulnerabilities apply to other mood states, such as fear. As such, the present results may underestimate the effect that sleep loss has on mood states among the general adolescent population.

As seen in the 5 h sleep dose group, positive mood states of happiness and energy significantly increased from sleep restriction to recovery. In addition, the 7.5 h sleep dose group displayed more energy but not happiness between the sleep restriction and recovery phases. These increases in positive mood states following recovery sleep provide support for the restorative effects of optimal sleep following cumulative sleep loss, allowing adolescents mood to recover to baseline values. For the 10 h sleep dose, happiness increased from baseline to the sleep dose condition, then further increased during recovery. This demonstrates that obtaining optimal sleep increases happiness, with 10 h group participants’ self-reported happiness increasing by more than 10% over the course of the study. Larger effects of sleep loss on mood were found for positive mood states, consistent with recent meta-analytic findings in adolescents [ 17 ]. This highlights the importance of considering positive mood states in research into the impact of sleep on mood and also has important clinical ramifications, given the role of anhedonia in psychological disorders such as depression.

It was expected that negative moods would recover from sleep restriction to recovery for participants in the 5 and 7.5 h sleep restriction conditions. Conversely, in the 5 h sleep dose, depressed mood and fear significantly increased from sleep restriction to recovery. It has been suggested that this increase in negative mood may be a natural effect of living in a laboratory environment, as has observed in adult participants given 9 h sleep opportunities over nine nights [ 39 ]. If so, it would be expected that this same pattern would be observed in the 10 h sleep dose group. However, the only mood to demonstrate this effect without sleep loss was confusion. It is possible that the factors leading to increased negative moods in adults placed in a laboratory environment do not affect adolescents in the same way. This may be due to differences in study protocols, or adolescents’ reduced ability to regulate mood following a period of sleep restriction [ 19 ] when faced with a mood-evoking situation (i.e. leaving the laboratory and their new-found friends). As reported elsewhere [ 8 ], salivary dim light melatonin onset showed a significant and dose–response delay in response to sleep restriction. As such, participants in the 5 h condition completed the study with a circadian rhythm that ran nearly 3 h later than it did at baseline. As a result, adolescents were likely to be waking closer to their circadian nadir, which may result in increased sleep inertia and worse mood, even following recovery sleep.

The current study was able to control for many of the confounding factors, which may have influenced the outcomes of prior adolescent sleep research. No other identified adolescent study examining the effect of sleep on mood was completed entirely under laboratory conditions. As such, previous studies were not able to control for exposure to environmental variables and diet, such as caffeine or excessive sugar to the same degree. The laboratory conditions allowed enforced bed/wake times, permitting stronger causal conclusions to be made about the effect of sleep duration on mood without having to consider response biases inflating relationships between subjective sleep and mood measures when both sleep and mood are measured subjectively. However, laboratory conditions present additional challenges, with reduced ecological validity.

Some of the challenges of the laboratory environment in measuring mood outcomes include the effect of an unfamiliar environment, socializing with peers who are not part of their normal friendship groups, and lack of contact with friends and family, which could have been confounding factors in the effect of sleep restriction on mood. It is a possibility that in a more familiar environment, such as participants’ homes, we may expect to see a more ecologically valid indication of the effect of sleep on mood; however, this comes at a cost of greater exposure to extraneous variables. Nonetheless, the inclusion of 7.5 and 10 h sleep dose groups provides a direct comparison between conditions to test the independent effects of sleep “dose” on mood and helps to distinguish between the effects of sleep loss and what may result simply from being in a laboratory environment for an extended period.

An important consideration regarding the current study is that the screening process ensured that participants were both physically and psychologically healthy. Although this is important to minimize exposure to at-risk individuals and to control for confounding variables, it is possible that the sample of the current study was more impervious or resilient to many of the mood disturbances often associated with inadequate sleep. As such, mood effects witnessed in the present study may have been felt more acutely in at-risk individuals, as indicated by prior research [ 36 , 38 ].

Patterns of cumulative sleep loss are increasingly prevalent among adolescents [ 40 ]. The current study found that, when restricted to 5 h sleep for five nights, adolescents’ happiness and energy decreased, depressed mood, anger, and confusion increased, while fear and anxiety did not change. For participants in the 7.5 sleep dose, no significant changes to positive or negative moods were observed between baseline and sleep restriction conditions, and this degree of sleep restriction may require a longer period of time to observe detrimental effects to mood. It is important to note that, while statistical significance was not reached, small-to-medium effect sizes in changes to these mood states were observed between baseline and sleep restriction in the 7.5-h condition. As such, we cannot be sure that, over an extended period of time, that sleep restricted to 7.5 h sleep opportunity per night may not be damaging to mental health.

The implications of the effect of sleep duration on mood relate to the increasing incidence of both sleep loss and mood disorders in adolescents [ 41–43 ], suggesting a greater need for awareness, support, and intervention in promoting healthy sleep for adolescents [ 44 ]. In addition, findings of the current study demonstrate the rapidity of mood decline when adolescent sleep is restricted to 5 h per night, while a more modest amount of sleep loss may require an extended period to see similar effects. Given the prevalence of insufficient sleep and the rising incidence of mood disorders and dysregulation in adolescents, these findings highlight the importance of sufficient sleep to mitigate these risks.

Financial disclosure: The authors have no financial conflicts of interest to declare.

Non-financial disclosure: The authors have no non-financial conflicts of interest to declare.

Brand S , et al.  “Sleep well, our tough heroes!”—in adolescence, greater mental toughness is related to better sleep schedules . Behav Sleep Med. 2014 ; 12 ( 6 ): 444 – 454 .

Google Scholar

Paus T , et al.  Why do many psychiatric disorders emerge during adolescence? Nat Rev Neurosci. 2008 ; 9 ( 12 ): 947 – 957 .

Fredriksen K , et al.  Sleepless in Chicago: tracking the effects of adolescent sleep loss during the middle school years . Child Dev. 2004 ; 75 ( 1 ): 84 – 95 .

Wolfson AR , et al.  Sleep schedules and daytime functioning in adolescents . Child Dev. 1998 ; 69 ( 4 ): 875 – 887 .

Franzen PL , et al.  Relationships between affect, vigilance, and sleepiness following sleep deprivation . J Sleep Res. 2008 ; 17 ( 1 ): 34 – 41 .

Dinges DF , et al.  Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4–5 hours per night . Sleep. 1997 ; 20 ( 4 ): 267 – 277 .

Kahn M , et al.  Effects of one night of induced night-wakings versus sleep restriction on sustained attention and mood: a pilot study . Sleep Med. 2014 ; 15 ( 7 ): 825 – 832 .

Short MA , et al.  Estimating adolescent sleep need using dose-response modelling . Sleep. 2018;41(4). doi: 10.1093/sleep/zsy011 .

Carskadon MA , et al.  Pubertal changes in daytime sleepiness . Sleep. 1980 ; 2 ( 4 ): 453 – 460 .

Van Petegem S , et al.  On the association between adolescent autonomy and psychosocial functioning: examining decisional independence from a self-determination theory perspective . Dev Psychol. 2012 ; 48 ( 1 ): 76 – 88 .

Strang NM , et al.  Developmental changes in adolescents’ neural response to challenge . Dev Cogn Neurosci. 2011 ; 1 ( 4 ): 560 – 569 .

Short MA , et al.  The sleep patterns and well-being of Australian adolescents . J Adolesc. 2013 ; 36 ( 1 ): 103 – 110 .

Lovato N , et al.  A meta-analysis and model of the relationship between sleep and depression in adolescents: recommendations for future research and clinical practice . Sleep Med Rev. 2014 ; 18 ( 6 ): 521 – 529 .

Becker SP , et al.  Impact of sleep restriction on affective functioning in adolescents with attention-deficit/hyperactivity disorder . J Child Psychol Psychiatry. 2020 ; 61 ( 10 ): 1160 – 1168 .

Lo JC , et al.  Cognitive performance, sleepiness, and mood in partially sleep deprived adolescents: the need for sleep study . Sleep. 2016 ; 39 ( 3 ): 687 – 698 .

Lo JC , et al.  Neurobehavioral impact of successive cycles of sleep restriction with and without naps in adolescents . Sleep. 2016 ; 40 ( 2 ). doi: 10.1093/sleep/zsw042 .

Short MA , et al.  The relationship between sleep duration and mood in adolescents: a systematic review and meta-analysis . Sleep Med Rev. 2020 ; 52 : 101311 .

Talbot LS , et al.  Sleep deprivation in adolescents and adults: changes in affect . Emotion. 2010 ; 10 ( 6 ): 831 – 841 .

Baum KT , et al.  Sleep restriction worsens mood and emotion regulation in adolescents . J Child Psychol Psychiatry. 2014 ; 55 ( 2 ): 180 – 190 .

Jiang F , et al.  Effect of chronic sleep restriction on sleepiness and working memory in adolescents and young adults . J Clin Exp Neuropsychol. 2011 ; 33 ( 8 ): 892 – 900 .

Fox KR . The influence of physical activity on mental well-being . Public Health Nutr. 1999 ; 2 ( 3A ): 411 – 418 .

Lieberman HR , et al.  Effects of caffeine, sleep loss, and stress on cognitive performance and mood during U.S. Navy SEAL training. Sea-Air-Land . Psychopharmacology. 2002 ; 164 ( 3 ): 250 – 261 .

Lopresti AL , et al.  A review of lifestyle factors that contribute to important pathways associated with major depression: diet, sleep and exercise . J Affect Disord. 2013 ; 148 ( 1 ): 12 – 27 .

Hirshkowitz M , et al.  National Sleep Foundation’s sleep time duration recommendations: methodology and results summary . Sleep Health. 2015 ; 1 ( 1 ): 40 – 43 .

Petersen AC , et al.  A self-report measure of pubertal status: reliability, validity, and initial norms . J Youth Adolesc. 1988 ; 17 ( 2 ): 117 – 133 .

Paruthi S , et al.  Recommended amount of sleep for pediatric populations: a consensus statement of the American Academy of Sleep Medicine . J Clin Sleep Med. 2016 ; 12 ( 6 ): 785 – 786 .

Buysse DJ , et al.  Recommendations for a standard research assessment of insomnia . Sleep. 2006 ; 29 ( 9 ): 1155 – 1173 .

Smith CS , et al.  Evaluation of three circadian rhythm questionnaires with suggestions for an improved measure of morningness . J Appl Psychol. 1989 ; 74 ( 5 ): 728 – 738 .

Tonetti L , et al.  Circadian type and mood seasonality in adolescents . Psychiatry Clin Neurosci. 2012 ; 66 ( 2 ): 157 – 159 .

Stern RA , et al.  Visual analogue mood scales to measure internal mood state in neurologically impaired patients: description and initial validity evidence . Aphasiology. 1997 ; 11 ( 1 ): 59 – 71 .

Boergers J , et al.  Later school start time is associated with improved sleep and daytime functioning in adolescents . J Dev Behav Pediatr. 2014 ; 35 ( 1 ): 11 – 17 .

Van Dongen HP , et al.  Mixed-model regression analysis and dealing with interindividual differences . Methods Enzymol. 2004 ; 384 : 139 – 171 .

Fuligni AJ , et al.  Daily variation in adolescents’ sleep, activities, and psychological well-being . J Res Adolesc. 2006 ; 16 ( 3 ): 353 – 378 .

Dewald-Kaufmann JF , et al.  Why sleep matters: differences in daytime functioning between adolescents with low and high chronic sleep reduction and short and long sleep durations . J Cog and Behav Psychotherapies . 2013 ; 13 ( 1A ): 171 .

Carskadon MA , et al.  Short sleep as an environmental exposure: a preliminary study associating 5-HTTLPR genotype to self-reported sleep duration and depressed mood in first-year university students . Sleep. 2012 ; 35 ( 6 ): 791 – 796 .

Gregory AM , et al.  Associations between sleep quality and anxiety and depression symptoms in a sample of young adult twins and siblings . J Psychosom Res. 2011 ; 71 ( 4 ): 250 – 255 .

Fuligni AJ , et al.  Individual differences in optimum sleep for daily mood during adolescence . J Clin Child Adolesc. 2017 ; 53 : 1 – 11 .

Paterson JL , et al.  What happens to mood, performance and sleep in a laboratory study with no sleep deprivation? Sleep Biol Rhythms. 2013 ; 11 ( 3 ): 200 – 209 .

Galland BC , et al.  Establishing normal values for pediatric nighttime sleep measured by actigraphy: a systematic review and meta-analysis . Sleep. 2018 . doi: 10.1093/sleep/zsy017 .

Dahl RE , et al.  Pathways to adolescent health sleep regulation and behavior . J Adolesc Health. 2002 ; 31 ( 6 Suppl ): 175 – 184 .

Ireland JL , et al.  The relationship between sleeping problems and aggression, anger, and impulsivity in a population of juvenile and young offenders . J Adolesc Health. 2006 ; 38 ( 6 ): 649 – 655 .

Short MA , et al.  Adolescent sleep restriction effects on cognition and mood . Prog Brain Res. 2019 ; 246 : 55 – 71 .

Short MA , et al.  Sleep and mental health in children and adolescents . In: Grandner MA Sleep and Health . Academic Press ; 2019 : 435 – 445 .

Google Preview

• depressive disorders
• sleep restriction therapy
• sleep duration

Supplementary data

Email alerts

Citing articles via, looking for your next opportunity.

• Recommend to Your Librarian
• Advertising and Corporate Services

Affiliations

• Online ISSN 1550-9109
• Print ISSN 0161-8105
• Copyright © 2024 Sleep Research Society
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Systematic Review
• Open access
• Published: 24 June 2024

Placebo effects in randomized trials of pharmacological and neurostimulation interventions for mental disorders: An umbrella review

• Nathan T. M. Huneke   ORCID: orcid.org/0000-0001-5981-6707 1 , 2 ,
• Jay Amin   ORCID: orcid.org/0000-0003-3792-0428 1 , 2 ,
• David S. Baldwin 1 , 2 , 3 ,
• Alessio Bellato 4 , 5 ,
• Valerie Brandt   ORCID: orcid.org/0000-0002-3208-2659 5 , 6 ,
• Samuel R. Chamberlain 1 , 2 ,
• Christoph U. Correll   ORCID: orcid.org/0000-0002-7254-5646 7 , 8 , 9 , 10 ,
• Luis Eudave 11 ,
• Matthew Garner 1 , 5 , 12 ,
• Corentin J. Gosling 5 , 13 , 14 ,
• Catherine M. Hill 1 , 15 ,
• Ruihua Hou 1 ,
• Oliver D. Howes   ORCID: orcid.org/0000-0002-2928-1972 16 , 17 , 18 ,
• Konstantinos Ioannidis 1 , 2 ,
• Ole Köhler-Forsberg 19 , 20 ,
• Lucia Marzulli 21 ,
• Claire Reed   ORCID: orcid.org/0000-0003-1385-4729 5 ,
• Julia M. A. Sinclair 1 ,
• Satneet Singh 2 ,
• Marco Solmi   ORCID: orcid.org/0000-0003-4877-7233 5 , 22 , 23 , 24 , 25   na1 &
• Samuele Cortese   ORCID: orcid.org/0000-0001-5877-8075 1 , 5 , 26 , 27 , 28   na1  

Molecular Psychiatry ( 2024 ) Cite this article

582 Accesses

5 Altmetric

Metrics details

• Drug discovery
• Neuroscience
• Psychiatric disorders

There is a growing literature exploring the placebo response within specific mental disorders, but no overarching quantitative synthesis of this research has analyzed evidence across mental disorders. We carried out an umbrella review of meta-analyses of randomized controlled trials (RCTs) of biological treatments (pharmacotherapy or neurostimulation) for mental disorders. We explored whether placebo effect size differs across distinct disorders, and the correlates of increased placebo effects. Based on a pre-registered protocol, we searched Medline, PsycInfo, EMBASE, and Web of Knowledge up to 23.10.2022 for systematic reviews and/or meta-analyses reporting placebo effect sizes in psychopharmacological or neurostimulation RCTs. Twenty meta-analyses, summarising 1,691 RCTs involving 261,730 patients, were included. Placebo effect size varied, and was large in alcohol use disorder ( g  = 0.90, 95% CI [0.70, 1.09]), depression ( g  = 1.10, 95% CI [1.06, 1.15]), restless legs syndrome ( g  = 1.41, 95% CI [1.25, 1.56]), and generalized anxiety disorder ( d  = 1.85, 95% CI [1.61, 2.09]). Placebo effect size was small-to-medium in obsessive-compulsive disorder ( d  = 0.32, 95% CI [0.22, 0.41]), primary insomnia ( g  = 0.35, 95% CI [0.28, 0.42]), and schizophrenia spectrum disorders (standardized mean change = 0.33, 95% CI [0.22, 0.44]). Correlates of larger placebo response in multiple mental disorders included later publication year (opposite finding for ADHD), younger age, more trial sites, larger sample size, increased baseline severity, and larger active treatment effect size. Most (18 of 20) meta-analyses were judged ‘low’ quality as per AMSTAR-2. Placebo effect sizes varied substantially across mental disorders. Future research should explore the sources of this variation. We identified important gaps in the literature, with no eligible systematic reviews/meta-analyses of placebo response in stress-related disorders, eating disorders, behavioural addictions, or bipolar mania.

Similar content being viewed by others

Effects of open-label placebos in clinical trials: a systematic review and meta-analysis

A transdiagnostic meta-analysis of acute augmentations to psychological therapy

Treatment resistance in psychiatry: state of the art and new directions

Introduction.

A placebo is an ‘inactive’ substance or ‘sham’ technique that is used as a control for assessing the efficacy of an active treatment [ 1 ]. However, study participants in a placebo control group may experience considerable symptom improvements - a ‘placebo response’ [ 1 , 2 , 3 ]. Statistical artifacts or non-specific effects account for some of the placebo response. For example, many individuals seek treatment and are enrolled in clinical trials while their symptoms are at their worst. Their symptoms will gradually return to their usual severity (‘regression to the mean’), giving the appearance of a placebo response [ 4 ]. Further, it has been suggested that the placebo response is exacerbated due to unreliable ratings as well as baseline symptom severity inflation if raters are aware of severity criteria for entry to a trial [ 5 , 6 ]. Other potential sources of apparent placebo responses include sampling biases caused by the withdrawal of the least improved patients in the placebo arm, non-specific beneficial effects resulting from interactions with staff delivering the trial, environmental effects due to inpatient care during placebo-controlled trials, or other unaccounted for factors, such as dietary or exercise changes during the trial [ 7 , 8 , 9 ]. Nonetheless, there is evidence that placebo administration results in ‘true’ - or non-artefactual - placebo effects, that is, identifiable changes in biological systems [ 1 , 10 , 11 ]. For example, placebo administration is capable of causing immunosuppression [ 12 , 13 ], placebo effects in Parkinson’s disease are driven by striatal dopamine release [ 10 , 14 ], and placebo analgesia is mediated by endogenous opioid release [ 15 , 16 ]. Furthermore, there is evidence that placebo effects in depressive and anxiety disorders are correlated with altered activity in the ventral striatum, orbitofrontal cortex, rostral anterior cingulate cortex, and the default mode network [ 17 ]. The placebo effect size can be increased through the use of verbal suggestions and conditioning procedures, thus suggesting the underlying role of psychological mechanisms including learning and expectations [ 11 , 18 ].

Across age groups, treatment modalities, and diverse mental disorders, biological treatments (pharmacotherapy or neurostimulation) do reduce symptoms [ 19 , 20 , 21 , 22 ], but only a subgroup of patients experience a clinically significant symptom response or enter remission [ 23 , 24 , 25 ]. Furthermore, current medications may also have unfavourable side effects [ 23 , 26 , 27 , 28 , 29 , 30 , 31 ]. Given the high prevalence of mental disorders and their significant socioeconomic burden [ 32 , 33 , 34 ], there is a need to develop more effective and safer psychopharmacologic and neurostimulation treatments. However, in randomized-controlled trials (RCTs), the magnitude of the placebo response may be considerable, which can affect the interpretation of their results [ 35 , 36 , 37 ]. For example, in antipsychotic trials over the past 40 years, placebo response has increased while medication response has remained consistent [ 38 , 39 ]. Consequently, the trial’s ability to statistically differentiate between an active medication and a placebo is diminished [ 40 ]. Indeed, large placebo response rates have been implicated in hindering psychotropic drug development [ 41 , 42 ]. The increased placebo response can also affect larger data synthesis approaches, such as network meta-analysis, in which assumptions about placebo responses (e.g. stability over time) might affect the validity of results [ 43 ].

Improved understanding of participant, trial, and mental disorder-related factors that contribute to placebo response might allow better clinical trial design to separate active treatment from placebo effects. There is a growing body of research, including individual studies and systematic reviews/meta-analyses, examining the placebo response within specific mental disorders [ 35 ]. However, to date, no overarching synthesis of this literature, to detect any similarities or differences across mental disorders, has been published. We therefore carried out an umbrella review of meta-analyses to address this need. We aimed to assess the placebo effect size in RCTs for a range of mental disorders, whether the effect size differs across distinct mental disorders, and identify any correlates of increased placebo effect size or response rate.

The protocol for this systematic umbrella review was pre-registered on the open science framework ( https://osf.io/fxvn4/ ) and published [ 44 ]. Deviations from this protocol, and additions to it, were: eight authors were involved in record screening rather than two; we reported effect sizes pooled across age groups and analyses comparing placebo effect sizes between age groups; and we included a meta-analysis that incorporated trials of dietary supplements as well as medications in autism. For the rationale behind these decisions, see eMethods.

Eight authors (NH, AB, VB, LE, OKF, LM, CR, SS) carried out the systematic review and data extraction independently in pairs. Discrepancies were resolved through consensus or through arbitration by a third reviewer (NH or SCo). We searched, without date or language restrictions, up to 23.10.2022, Medline, PsycInfo, EMBASE + EMBASE Classic, and Web of Knowledge for systematic reviews with or without meta-analyses of RCTs of biological treatments (psychopharmacotherapy or neurostimulation) compared with a placebo or sham treatment in individuals with mental disorders diagnosed according to standardized criteria. The full search strategy is included in eMethods. We also sought systematic reviews of RCTs conducted in patients with sleep-wake disorders, since these disorders are included in the DSM-5 and their core symptoms overlap with those of mental disorders [ 45 ]. We retained systematic reviews with or without meta-analyses that reported within-group changes in symptoms in the placebo arm.

Next, to prevent duplication of data, a matrix containing all eligible systematic reviews/meta-analyses for each category of mental disorder was created. Where there were multiple eligible systematic reviews/meta-analyses for the same disorder and treatment, we preferentially included meta-analyses, and if multiple eligible meta-analyses remained, then we included the one containing the largest number of studies for the same disorder and treatment, in line with recent umbrella reviews [ 46 , 47 ].

Data were extracted by at least two among six reviewers (AB, VB, LE, OKF, CR, SS) independently in pairs via a piloted form. All extracted data were further checked by a third reviewer (NH). See eMethods for a list of extracted data.

Our primary outcome was the pre-post effect size of the placebo/sham related to the condition-specific primary symptom change for each mental disorder. Secondary outcomes included any other reported clinical outcomes in eligible reviews. We report effect sizes calculated within-group from baseline and post-treatment means by meta-analysis authors, including Cohen’s d and Hedges’ g for repeated measures, which account for both mean difference and correlation between paired observations; and standardized mean change, where the average change score is divided by standard deviation of the change scores. We interpreted the effect size in line with the suggestion by Cohen [ 48 ], i.e. small (~0.2), medium (~0.5), or large (~0.8).

In addition, we extracted data regarding potential correlates of increased placebo effect size or response rate (as defined and assessed by the authors of each meta-analysis) in each mental disorder identified through correlation analyses or meta-regression. Where available, results from multivariate analyses were preferred.

The methodological quality of included reviews was assessed by at least two among six reviewers (AB, VB, LE, OKF, NH, CR) independently and in pairs using the AMSTAR-2 tool, a critical appraisal tool that enables reproducible assessments of the conduct of systematic reviews [ 49 ]. The methodological quality of each included review was rated as high, moderate, low, or critically low.

Our initial search identified 6,108 records. After screening titles and abstracts, we obtained and assessed 115 full-text reports (see eResults for a list of articles excluded following full-text assessment, with reasons). Of these, 20 were deemed eligible, and all were systematic reviews with meta-analysis (Fig.  1 ). In total, the 20 included meta-analyses synthesized data from 1,691 RCTs (median 55) involving 261,730 patients (median 5,365). These meta-analyses were published between 2007 and 2022 and involved individuals with the following mental disorders: major depressive disorder (MDD; n  = 6) [ 50 , 51 , 52 , 53 , 54 , 55 ], anxiety disorders ( n  = 4) [ 55 , 56 , 57 , 58 ], schizophrenia spectrum disorders ( n  = 3) [ 38 , 59 , 60 ], alcohol use disorder (AUD; n  = 1) [ 61 ], attention-deficit/hyperactivity disorder (ADHD; n  = 1) [ 62 ], autism spectrum disorders ( n  = 1) [ 63 ], bipolar depression ( n  = 1) [ 64 ], intellectual disability ( n  = 1) [ 65 ], obsessive-compulsive disorder (OCD; n  = 1) [ 66 ], primary insomnia ( n  = 1) [ 67 ], and restless legs syndrome (RLS; n  = 1) [ 68 ].

Twenty meta-analyses were included.

The methodological quality of the included meta-analyses according to AMSTAR-2 ratings was high in two meta-analyses (ADHD and autism), low in four meta-analyses, and critically low in the remaining 14 meta-analyses (Table  1 ). The most common sources of bias that led to downgrading on the AMSTAR-2 were: no list of excluded full-text articles with reasons ( k  = 14), no explicit statement that the protocol was pre-registered ( k  = 14), and no assessment of the potential impact of risk of bias in individual studies on the results ( k  = 13). The full reasoning behind our AMSTAR-2 ratings is included in eResults.

Our first objective was to determine placebo effect sizes across mental conditions. Data regarding within-group placebo efficacy were reported in sixteen of the included meta-analyses [ 38 , 50 , 52 , 53 , 55 , 56 , 57 , 58 , 60 , 61 , 62 , 63 , 65 , 66 , 67 , 68 ]. Placebo effect sizes for the primary outcomes ranged from 0.23 to 1.85, with a median of 0.64 (Fig.  2 ). Median heterogeneity across meta-analyses was I 2  = 72%, suggesting a generally high percentage of heterogeneity due to true variation across studies.

Dots represent placebo group effect size while triangles represent active effect size. CI confidence interval, MDD major depressive disorder, GAD generalized anxiety disorder, SAD social anxiety disorder, OCD obsessive-compulsive disorder, g Hedges’ g, d Cohen’s d, SMC standardized mean change, NR not reported.

A detailed description of each meta-analysis included for this objective is included in eResults. Here, we report a summary of these results in order of the greatest number of RCT’s and meta-analyses included per disorder. In MDD, a large within-group placebo effect was observed ( g  = 1.10, 95% CI [1.06, 1.15]), although active medication had an even larger effect size ( g  = 1.49, 95% CI [1.44, 1.53]) [ 50 ]. Similarly, in children and adolescents with MDD, placebo effect size was large ( g  = 1.57, 95% CI [1.36, 1.78]), as was serotonergic medication effect size ( g  = 1.85, 95% CI [1.70, 2.00]) [ 55 ]. In treatment-resistant MDD, the within-group placebo effect size was smaller than in non-treatment-resistant MDD ( g  = 0.89, 95% CI [0.81, 0.98]) [ 52 ]. In neuromodulation trials for MDD, the effect size of sham was g  = 0.80 (95% CI [0.65, 0.95]) [ 53 ]. In this meta-analysis, the effect size was larger for non-treatment-resistant ( g  = 1.28, 95% CI [0.47, 2.97]) compared to treatment-resistant participants (g = 0.50 95% CI [0.03, 0.99]) [ 53 ]. In adults with anxiety disorders, placebo effect sizes varied across disorders, with a medium effect size in panic disorder ( d  = 0.57, 95% CI [0.50, 0.64]) [ 56 ] and large effect sizes in generalized anxiety disorder (GAD) ( d  = 1.85, 95% CI [1.61, 2.09]) and social anxiety disorder (SAD) ( d  = 0.94, 95% CI [0.77, 1.12]) [ 57 ]. Other meta-analyses in children and adolescents and older adults pooled RCTs across anxiety disorders, and found large placebo effect sizes ( g  = 1.03, 95% CI [0.84, 1.21] and d  = 1.06, 95% CI [0.71, 1.42], respectively) [ 55 , 58 ]. In ADHD, placebo effect size was medium-to-large for clinician-rated outcomes (SMC = 0.75, 95% CI [0.67, 0.83]) [ 62 ]. There was additionally a significant negative relationship between placebo effect size and drug-placebo difference (−0.56, p  < 0.01) for self-rated outcomes [ 62 ]. In schizophrenia spectrum disorders, placebo effect size was small-to-medium in antipsychotic RCTs (SMC = 0.33, 95% CI [0.22, 0.44]) [ 38 ] and medium in RCTs focusing specifically on negative symptoms ( d  = 0.64, 95% CI [0.46, 0.83]) [ 60 ]. Placebo effect size in RLS was large when measured via rating scales ( g  = 1.41, 95% CI [1.25, 1.56]), but small ( g  = 0.02 to 0.24) in RCTs using objective outcomes [ 68 ]. In autism, placebo effect sizes were small (SMC ranged 0.23 to 0.36) [ 63 ]. Similarly, placebo effect size was small in OCD ( d  = 0.32, 95% CI [0.22, 0.41]), although larger in children and adolescents ( d  = 0.45, 95% CI [0.35, 0.56]) compared with adults ( d  = 0.27, 95% CI [0.15, 0.38]) [ 66 ]. Placebo effect size was large in AUD ( g  = 0.90, 95% CI [0.70, 1.09]) [ 61 ], small in primary insomnia ( g ranged 0.25 to 0.43) [ 67 ], and medium in intellectual disability related to genetic causes ( g  = 0.47, 95% CI [0.18, 0.76]) [ 65 ].

Our second objective was to examine the correlates of increased placebo response. We included 14 meta-analyses that reported correlates of placebo effect size or response rate through correlation analysis or meta-regression [ 38 , 51 , 53 , 54 , 56 , 57 , 59 , 60 , 61 , 62 , 63 , 64 , 66 , 68 ]. The key correlates extracted from these studies are summarized in Table  2 .

Several variables were consistently identified across meta-analyses. Increased number of trial sites was a positive correlate of increased placebo response in MDD [ 51 , 54 ], schizophrenia spectrum disorders [ 59 ], and autism spectrum disorders [ 63 ]. Similarly, increased sample size was positively associated with placebo effect size in schizophrenia spectrum disorders [ 59 ], OCD [ 66 ], and panic disorder [ 56 ]. Later publication or study year was associated with greater placebo response in anxiety disorders [ 56 , 57 ], schizophrenia spectrum disorders [ 38 ], AUD [ 61 ], and OCD [ 66 ] but not in MDD [ 51 ], and with reduced placebo response in ADHD [ 62 ]. Younger age was associated with increased placebo responses in schizophrenia spectrum disorders [ 38 , 59 ] and OCD [ 66 ]. Increased baseline illness severity was associated with increased placebo response in schizophrenia spectrum disorders [ 38 ], ADHD [ 62 ], and AUD [ 61 ]. Increased trial or follow-up duration was positively associated with increased placebo response in MDD [ 51 ], but negatively associated with placebo response in schizophrenia spectrum disorders [ 38 , 60 ] and OCD [ 66 ]. Finally, the effect size of active treatment was positively associated with increased placebo response in neurostimulation trials for MDD [ 53 ], bipolar depression [ 64 ], autistic spectrum disorders [ 63 ], and ADHD [ 62 ].

There were also some variables associated with increased placebo response in single disorders only. Flexible dosing, rather than fixed dosing, was associated with increased placebo response in MDD [ 51 ]. Increased illness duration was associated with reduced placebo response in schizophrenia spectrum disorders [ 38 ]. In RCTs for negative symptoms of schizophrenia, a higher number of active treatment arms was associated with increased placebo response [ 60 ]. A number of treatment administrations was a positive correlate of increased placebo response in patients with AUD [ 61 ]. A low risk of bias in selective reporting was associated with increased placebo response in ADHD [ 62 ]. Finally, a low risk of bias in allocation concealment was associated with increased placebo response in autism [ 63 ].

To our knowledge, this is the first overarching synthesis of the literature exploring the placebo response in RCTs of biological treatments across a broad range of mental disorders. We found that placebo responses were present and detectable across mental disorders. Further, the placebo effect size across these disorders varied between small and large (see Fig.  3 ). Additionally, several variables appeared to be associated with increased placebo effect size or response rate across a number of disorders, while others were reported for individual disorders only.

CI confidence interval, MDD major depressive disorder, GAD generalized anxiety disorder, SAD social anxiety disorder, OCD obsessive-compulsive disorder, g Hedges’ g, d Cohen’s d, SMC standardized mean change.

Our umbrella review distinguishes itself from a recent publication on placebo mechanisms across medical conditions [ 69 ]. Only four systematic reviews of research in mental disorders were included in that recent review [ 69 ], none of which were eligible for inclusion in our umbrella review, as we focus specifically on RCTs in mental disorders. Thus, our current umbrella review synthesizes different literature and is complementary [ 69 ].

We found substantial variation in placebo effect sizes across mental disorders. In GAD, SAD, MDD, AUD, and RLS (for subjective outcomes), placebo effects were large (>0.9), while they were small (approximately 0.3) in OCD, primary insomnia, autism, RLS (for objective outcomes), and schizophrenia spectrum disorders. It is noteworthy that placebo effect size/response rate correlated with active treatment effect size/response rate in many disorders (MDD, bipolar depression, ADHD, and autism). Nonetheless, where reported, active treatment was always superior. This possibly suggests an underlying ‘treatment responsiveness’ of these disorders that can vary in size. Perhaps, the natural history of a disorder is an important factor in ‘responsiveness’, i.e., disorders in which there is greater natural fluctuation in severity will show larger placebo (and active treatment) effect sizes. Supporting this hypothesis, increased trial duration predicted a larger placebo effect size in MDD, a disorder in which the natural course includes improvement [ 31 , 51 , 70 ]. Conversely, in schizophrenia spectrum disorders where improvement (particularly of negative symptoms) is less likely [ 71 ], increased trial and illness duration predicted a smaller placebo effect size [ 38 , 60 ]. However, previous meta-analyses suggest that natural improvement, for example, measured via waiting list control, does not fully account for the placebo effect in depression and anxiety disorders [ 72 , 73 ]. Statistical artifact, therefore, does not seem to fully explain the variation in effect size.

Non-specific treatment mechanisms are likely an additional source of the observed placebo effect. For example, those with treatment-resistant illness might have reduced expectations regarding treatment. This assumption is supported by the subgroup analysis reported by Razza and colleagues showing sham neuromodulation efficacy reduced as the number of previous failed antidepressant trials increased [ 53 ]. Another factor to consider is the outcome measure chosen. For example, the placebo effect size in panic disorder was smaller when calculated with objective or self-report measures compared with clinician-rated measures [ 56 ]. A similar finding was reported in ADHD trials [ 62 ]. Why placebo effect sizes would differ with clinician-rated versus self-rated scales is unclear. This might result from ‘demand characteristics’ (i.e., cues that suggest to a patient how they ‘should’ respond), or unblinding of the rater, or a combination of the two [ 74 , 75 ].

Several correlates of increased placebo response were reported in included meta-analyses. These included a larger sample size, more study sites, a later publication year (but with an opposite finding for ADHD), younger age, and increased baseline illness severity. This might reflect changes in clinical trial methods over time, the potential for increased ‘noise’ in the data with larger samples or more study sites, and, more speculatively, variables associated with increased volatility in symptoms [ 39 , 51 , 76 ]. A more extensive discussion regarding the potential reasons these variables might correlate with, or predict, placebo response is included in the eDiscussion. Although some correlates of increased placebo response were identified, perhaps more pertinently, it is unknown whether these also predict the separation between active treatment and placebo in most mental disorders. Three included meta-analyses did show that as placebo response increases, the likelihood of drug-placebo separation decreases [ 38 , 62 , 64 ]. This suggests correlates of placebo effect size are also correlates of trial success or failure, but this hypothesis needs explicit testing. In addition, few of the meta-analyses we included explored whether correlates of placebo response differed from correlates of active treatment response. For example, in clinical trials for gambling disorder, response to active treatment was predicted by weeks spent in the trial and by baseline severity, while response to placebo was predicted by baseline depressive and anxiety symptoms [ 77 ]. Furthermore, there is evidence that industry sponsorship is a specific correlate of reduced drug-placebo separation in schizophrenia spectrum disorders [ 78 ]. The largest meta-analysis that we included (conducted by Scott et al. [ 50 ]) did not explore correlates of increased placebo response through meta-regression analysis; rather, it was designed specifically to assess the impact of the use of placebo run-in periods in antidepressant trials. The authors found that use of a placebo run-in was associated with reduced placebo response. However, this effect did not enhance sensitivity to detect medication efficacy versus control groups, as trials with placebo run-in periods were also associated with a reduced medication response. Similar effects of placebo run-in were seen in univariate (but not multivariable) models in ADHD, where placebo run-in reduced placebo effect size in youth, but did not affect drug vs placebo difference [ 62 ]. Further work should be undertaken to ascertain whether trial-level correlates (including the use of placebo run-in) differentially explain active treatment or placebo response and whether controlling for these can improve drug-placebo separation.

Our results should be considered in the light of several possible limitations. First, as in any umbrella review, we were limited by the quality of the meta-analyses we included. Our AMSTAR-2 ratings suggest that confidence in the conclusions of most included meta-analyses should be critically low or low. Indeed, several meta-analyses did not assess for publication bias or for bias in included RCTs. This is relevant, as the risk of bias in selective reporting was highlighted as potentially being associated with placebo effect size in ADHD [ 62 ], and might therefore be relevant in other mental disorders. Second, our results are potentially vulnerable to biases or unmeasured confounders present in the included meta-analyses. Third, we attempted to prevent overlap and duplication of information by including only the meta-analyses with the most information. This might, however, have resulted in some data not being included in our synthesis. Fourth, an exploration of the potential clinical relevance of the placebo effect sizes reported here was outside the scope of the current review but should be considered an important question for future research. Finally, the meta-analyses we included encompassed RCTs with different levels of blinding (double-blind, single-blind). Although the majority of trials were likely double-blind, it is possible that different levels of blinding could have influenced placebo effect sizes through effects on expectations. Future analyses of placebo effects and their correlates should either focus on double-blind trials or compare results across levels of blinding. Related to this, the included meta-analyses pooled phase 2 and phase 3 trials (the latter of which will usually follow positive phase 2 trials), which might result in different expectation biases. Therefore, placebo effects should be compared between phase 2 and phase 3 trials in the future.

In this umbrella review, we found placebo effect sizes varied substantially across mental disorders. The sources of this variation remain unknown and require further study. Some variables were correlates of increased placebo response across mental disorders, including larger sample size, higher number of study sites, later publication year (opposite for ADHD), younger age, and increased baseline illness severity. There was also evidence that clinician-rated outcomes were associated with larger placebo effect sizes than self-rated or objective outcomes. We additionally identified important gaps in the literature, with no eligible systematic reviews identified in stress-related disorders, eating disorders, behavioural addictions, or bipolar mania. In relation to these disorders, some analyses have been published but they have not been included in systematic reviews/meta-analyses (e.g. analyses of individual patient data pooled across RCTs in acute mania [ 79 ] or gambling disorder [ 77 , 80 ]) and therefore were not eligible for inclusion here. We also focused on placebo response in RCTs of pharmacotherapies and neurostimulation interventions for mental disorders. We did not include placebo effects in psychosocial interventions, but such an analysis would also be valuable. Future studies should address these gaps in the literature and furthermore should compare findings in placebo arms with active treatment arms, both regarding treatment effect size and its correlates. Gaining additional insights into the placebo response may improve our ability to separate active treatment effects from placebo effects, thus paving the way for potentially effective new treatments for mental disorders.

Data availability

The datasets generated during and/or analysed during the current study are available in the Open Science Framework repository, https://osf.io/fxvn4/ .

Evers AWM, Colloca L, Blease C, Annoni M, Atlas LY, Benedetti F, et al. Implications of placebo and nocebo effects for clinical practice: Expert Consensus. Psychother Psychosom. 2018;87:204–10.

Article   PubMed   Google Scholar  

McQueen D, Cohen S, John-Smith PS, Rampes H. Rethinking placebo in psychiatry: the range of placebo effects. Adv Psychiatr Treat. 2013;19:162–70.

Article   Google Scholar  

Beecher HK. The powerful placebo. J Am Med Assoc. 1955;159:1602–6.

Article   CAS   PubMed   Google Scholar  

Harris I. When the placebo effect is not an effect. Acta Orthop. 2021;92:501–2.

Article   PubMed   PubMed Central   Google Scholar  

Landin R, DeBrota DJ, DeVries TA, Potter WZ, Demitrack MA. The impact of Restrictive Entry Criterion during the placebo lead-in period. Biometrics. 2000;56:271–8.

Jones BDM, Razza LB, Weissman CR, Karbi J, Vine T, Mulsant LS, et al. Magnitude of the Placebo response across treatment modalities used for treatment-resistant depression in adults: a systematic review and meta-analysis. JAMA Netw Open. 2021;4:e2125531.

Miller FG, Rosenstein DL. The nature and power of the placebo effect. J Clin Epidemiol. 2006;59:331–5.

Ashar YK, Chang LJ, Wager TD. Brain mechanisms of the Placebo effect: an affective appraisal account. Annu Rev Clin Psychol. 2017;13:73–98.

Ernst E, Resch KL. Concept of true and perceived placebo effects. BMJ. 1995;311:551–3.

Article   CAS   PubMed   PubMed Central   Google Scholar  

De La Fuente-Fernandez R. Expectation and Dopamine release: mechanism of the Placebo effect in Parkinson’s disease. Science. 2001;293:1164–6.

Benedetti F, Carlino E, Pollo A. How Placebos change the patient’s brain. Neuropsychopharmacology. 2011;36:339–54.

Goebel MU, Trebst AE, Steiner J, Xie YF, Exton MS, Frede S, et al. Behavioral conditioning of immunosuppression is possible in humans. FASEB J. 2002;16:1869–73.

Albring A, Wendt L, Benson S, Witzke O, Kribben A, Engler H, et al. Placebo effects on the immune response in humans: the role of learning and expectation. PloS One. 2012;7:e49477.

Lidstone SC, Schulzer M, Dinelle K, Mak E, Sossi V, Ruth TJ, et al. Effects of expectation on placebo-induced Dopamine release in Parkinson disease. Arch Gen Psychiatry. 2010;67:857–65.

Amanzio M, Benedetti F. Neuropharmacological dissection of placebo analgesia: expectation-activated opioid systems versus conditioning-activated specific subsystems. J Neurosci. 1999;19:484–94.

Amanzio M, Pollo A, Maggi G, Benedetti F. Response variability to analgesics: a role for non-specific activation of endogenous opioids. Pain. 2001;90:205–15.

Huneke NTM, Aslan IH, Fagan H, Phillips N, Tanna R, Cortese S, et al. Functional neuroimaging correlates of placebo response in patients with depressive or anxiety disorders: A systematic review. Int J Neuropsychopharmacol. 2022;25:433–47.

Vase L, Riley JL, Price DD. A comparison of placebo effects in clinical analgesic trials versus studies of placebo analgesia. Pain. 2002;99:443–52.

Solmi M, Croatto G, Piva G, Rosson S, Fusar-Poli P, Rubio JM, et al. Efficacy and acceptability of psychosocial interventions in schizophrenia: systematic overview and quality appraisal of the meta-analytic evidence. Mol Psychiatry. 2023;28:354–68.

Monteleone AM, Pellegrino F, Croatto G, Carfagno M, Hilbert A, Treasure J, et al. Treatment of eating disorders: A systematic meta-review of meta-analyses and network meta-analyses. Neurosci Biobehav Rev. 2022;142:104857.

Rosson S, de Filippis R, Croatto G, Collantoni E, Pallottino S, Guinart D, et al. Brain stimulation and other biological non-pharmacological interventions in mental disorders: An umbrella review. Neurosci Biobehav Rev. 2022;139:104743.

Correll CU, Cortese S, Croatto G, Monaco F, Krinitski D, Arrondo G, et al. Efficacy and acceptability of pharmacological, psychosocial, and brain stimulation interventions in children and adolescents with mental disorders: an umbrella review. World Psychiatry. 2021;20:244–75.

Gaynes BN, Warden D, Trivedi MH, Wisniewski SR, Fava M, Rush AJ. What did STAR*D teach us? Results from a large-scale, practical, clinical trial for patients with depression. Psychiatr Serv. 2009;60:1439–45.

Stone MB, Yaseen ZS, Miller BJ, Richardville K, Kalaria SN, Kirsch I. Response to acute monotherapy for major depressive disorder in randomized, placebo controlled trials submitted to the US Food and Drug Administration: individual participant data analysis. BMJ. 2022;378:e067606.

Hendriks SM, Spijker J, Licht CMM, Hardeveld F, de Graaf R, Batelaan NM, et al. Long-term disability in anxiety disorders. BMC Psychiatry. 2016;16:248.

Dragioti E, Solmi M, Favaro A, Fusar-Poli P, Dazzan P, Thompson T, et al. Association of antidepressant use with adverse health outcomes: a systematic umbrella review. JAMA Psychiatry. 2019;76:1241–55.

Croatto G, Vancampfort D, Miola A, Olivola M, Fiedorowicz JG, Firth J, et al. The impact of pharmacological and non-pharmacological interventions on physical health outcomes in people with mood disorders across the lifespan: An umbrella review of the evidence from randomised controlled trials. Mol Psychiatry. 2023;28:369–90.

Papola D, Ostuzzi G, Gastaldon C, Morgano GP, Dragioti E, Carvalho AF, et al. Antipsychotic use and risk of life-threatening medical events: umbrella review of observational studies. Acta Psychiatr Scand. 2019;140:227–43.

Linden M. How to define, find and classify side effects in psychotherapy: from unwanted events to adverse treatment reactions. Clin Psychol Psychother. 2013;20:286–96.

Reynolds GP, Kirk SL. Metabolic side effects of antipsychotic drug treatment – pharmacological mechanisms. Pharmacol Ther. 2010;125:169–79.

Cuijpers P, Karyotaki E, Weitz E, Andersson G, Hollon SD, van Straten A. The effects of psychotherapies for major depression in adults on remission, recovery and improvement: A meta-analysis. J Affect Disord. 2014;159:118–26.

Bloom DE, Cafiero E, Jané-Llopis E, Abrahams-Gessel S, Bloom LR, Fathima S, et al. The global economic burden of noncommunicable diseases. PGDA Work Pap. (2012).

Vos T, Lim SS, Abbafati C, Abbas KM, Abbasi M, Abbasifard M, et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396:1204–22.

Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ, Erskine HE, et al. Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010. The Lancet. 2013;382:1575–86.

Huneke NTM, van der Wee N, Garner M, Baldwin DS. Why we need more research into the placebo response in psychiatry. Psychol Med. 2020;50:2317–23.

Huneke NTM. Is superiority to placebo the most appropriate measure of efficacy in trials of novel psychotropic medications? Eur Neuropsychopharmacol. 2022;62:7–9.

Khan A, Brown WA. Antidepressants versus placebo in major depression: An overview. World Psychiatry. 2015;14:294–300.

Agid O, Siu CO, Potkin SG, Kapur S, Watsky E, Vanderburg D, et al. Meta-regression analysis of placebo response in antipsychotic trials, 1970–2010. Am J Psychiatry. 2013;170:1335–44.

Leucht S, Leucht C, Huhn M, Chaimani A, Mavridis D, Helfer B, et al. Sixty years of placebo-controlled antipsychotic drug trials in acute schizophrenia: systematic review, bayesian meta-analysis, and meta-regression of efficacy predictors. Am J Psychiatry. 2017;174:927–42.

Enck P, Bingel U, Schedlowski M, Rief W. The placebo response in medicine: minimize, maximize or personalize? Nat Rev Drug Discov. 2013;12:191–204.

Correll CU, Solmi M, Cortese S, Fava M, Højlund M, Kraemer HC, et al. The future of psychopharmacology: a critical appraisal of ongoing phase 2/3 trials, and of some current trends aiming to de-risk trial programmes of novel agents. World Psychiatry. 2023;22:48–74.

Stahl SM, Greenberg GD. Placebo response rate is ruining drug development in psychiatry: why is this happening and what can we do about it? Acta Psychiatr Scand. 2019;139:105–7.

Nikolakopoulou A, Chaimani A, Furukawa TA, Papakonstantinou T, Rücker G, Schwarzer G. When does the placebo effect have an impact on network meta-analysis results? BMJ Evid-Based Med. 2023. https://doi.org/10.1136/bmjebm-2022-112197 .

Huneke NTM, Amin J, Baldwin DS, Chamberlain SR, Correll CU, Garner M, et al. Placebo effects in mental health disorders: protocol for an umbrella review. BMJ Open. 2023;13:e073946.

Gauld C, Lopez R, Morin CM, Maquet J, Mcgonigal A, Geoffroy P-A, et al. Why do sleep disorders belong to mental disorder classifications? A network analysis of the “Sleep-Wake Disorders” section of the DSM-5. J Psychiatr Res. 2021;142:153–9.

Köhler-Forsberg O, Stiglbauer V, Brasanac J, Chae WR, Wagener F, Zimbalski K, et al. Efficacy and safety of antidepressants in patients with comorbid depression and medical diseases: an umbrella systematic review and meta-Analysis. JAMA Psychiatry. 2023. https://doi.org/10.1001/jamapsychiatry.2023.2983 .

Belbasis L, Bellou V, Ioannidis JPA. Conducting umbrella reviews. BMJ Med. 2022;1:e000071.

Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; (1988).

Shea BJ, Reeves BC, Wells G, Thuku M, Hamel C, Moran J, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008.

Scott AJ, Sharpe L, Quinn V, Colagiuri B. Association of Single-blind Placebo Run-in Periods With the Placebo Response in Randomized Clinical Trials of Antidepressants: A Systematic Review and Meta-analysis. JAMA Psychiatry. 2022;79:42.

Furukawa TA, Cipriani A, Atkinson LZ, Leucht S, Ogawa Y, Takeshima N, et al. Placebo response rates in antidepressant trials: a systematic review of published and unpublished double-blind randomised controlled studies. Lancet Psychiatry. 2016;3:1059–66.

Scott F, Hampsey E, Gnanapragasam S, Carter B, Marwood L, Taylor RW, et al. Systematic review and meta-analysis of augmentation and combination treatments for early-stage treatment-resistant depression. J Psychopharmacol. 2023;37:268–78.

Razza LB, Moffa AH, Moreno ML, Carvalho AF, Padberg F, Fregni F, et al. A systematic review and meta-analysis on placebo response to repetitive transcranial magnetic stimulation for depression trials. Prog Neuropsychopharmacol Biol Psychiatry. 2018;81:105–13.

Meister R, Abbas M, Antel J, Peters T, Pan Y, Bingel U, et al. Placebo response rates and potential modifiers in double-blind randomized controlled trials of second and newer generation antidepressants for major depressive disorder in children and adolescents: a systematic review and meta-regression analysis. Eur Child Adolesc Psychiatry. 2020;29:253–73.

Locher C, Koechlin H, Zion SR, Werner C, Pine DS, Kirsch I, et al. Efficacy and safety of selective Serotonin reuptake inhibitors, Serotonin-Norepinephrine Reuptake inhibitors, and placebo for common psychiatric disorders among children and adolescents: a systematic review and meta-analysis. Jama Psychiatry. 2017;74:1011–20.

Ahmadzad-Asl M, Davoudi F, Mohamadi S, Hadi F, Nejadghaderi SA, Mirbehbahani SH, et al. Systematic review and meta-analysis of the placebo effect in panic disorder: Implications for research and clinical practice. Aust N Z J Psychiatry. 2022;56:1130–41.

Bandelow B, Reitt M, Röver C, Michaelis S, Görlich Y, Wedekind D. Efficacy of treatments for anxiety disorders: a meta-analysis. Int Clin Psychopharmacol. 2015;30:183–92.

Pinquart M, Duberstein PR. Treatment of anxiety disorders in older adults: a meta-analytic comparison of behavioral and pharmacological interventions. Am J Geriatr Psychiatry. 2007;15:639–51.

Leucht S, Chaimani A, Leucht C, Huhn M, Mavridis D, Helfer B, et al. 60 years of placebo-controlled antipsychotic drug trials in acute schizophrenia: Meta-regression of predictors of placebo response. Schizophr Res. 2018;201:315–23.

Czobor P, Kakuszi B, Bitter I. Placebo response in trials of negative symptoms in Schizophrenia: A critical reassessment of the evidence. Schizophr Bull. 2022;48:1228–40.

Del Re AC, Maisel N, Blodgett J, Wilbourne P, Finney J. Placebo group improvement in trials of pharmacotherapies for alcohol use disorders: a multivariate meta-analysis examining change over time. J Clin Psychopharmacol. 2013;33:649.

Faraone SV, Newcorn JH, Cipriani A, Brandeis D, Kaiser A, Hohmann S, et al. Placebo and nocebo responses in randomised, controlled trials of medications for ADHD: a systematic review and meta-analysis. Mol Psychiatry. 2022;27:212–9.

Siafis S, Çıray O, Schneider-Thoma J, Bighelli I, Krause M, Rodolico A, et al. Placebo response in pharmacological and dietary supplement trials of autism spectrum disorder (ASD): systematic review and meta-regression analysis. Mol Autism. 2020;11:66.

Iovieno N, Nierenberg AA, Parkin SR, Hyung Kim DJ, Walker RS, Fava M, et al. Relationship between placebo response rate and clinical trial outcome in bipolar depression. J Psychiatr Res. 2016;74:38–44.

Curie A, Yang K, Kirsch I, Gollub RL, des Portes V, Kaptchuk TJ, et al. Placebo responses in genetically determined intellectual disability: a meta-analysis. PloS One. 2015;10:e0133316.

Mohamadi S, Ahmadzad-Asl M, Nejadghaderi SA, Jabbarinejad R, Mirbehbahani SH, Sinyor M, et al. Systematic review and meta-analysis of the placebo effect and its correlates in obsessive compulsive disorder. Can J Psychiatry. 2023;68:479–94.

Winkler A, Rief W. Effect of placebo conditions on polysomnographic parameters in primary insomnia: a meta-analysis. Sleep. 2015;38:925–31.

PubMed   PubMed Central   Google Scholar  

Silva MA, Duarte GS, Camara R, Rodrigues FB, Fernandes RM, Abreu D, et al. Placebo and nocebo responses in restless legs syndrome: A systematic review and meta-analysis. Neurology. 2017;88:2216–24.

Frisaldi E, Shaibani A, Benedetti F, Pagnini F. Placebo and nocebo effects and mechanisms associated with pharmacological interventions: an umbrella review. BMJ Open. 2023;13:e077243.

Cuijpers P, Stringaris A, Wolpert M. Treatment outcomes for depression: challenges and opportunities. Lancet Psychiatry. 2020;7:925–7.

Bromet EJ, Fennig S. Epidemiology and natural history of schizophrenia. Biol Psychiatry. 1999;46:871–81.

Rutherford BR, Mori S, Sneed JR, Pimontel MA, Roose SP. Contribution of spontaneous improvement to placebo response in depression: A meta-analytic review. J Psychiatr Res. 2012;46:697–702.

Fernández-López R, Riquelme-Gallego B, Bueno-Cavanillas A, Khan KS. Influence of placebo effect in mental disorders research: A systematic review and meta-analysis. Eur J Clin Invest. 2022;52:e13762.

Goodwin GM, Croal M, Marwood L, Malievskaia E. Unblinding and demand characteristics in the treatment of depression. J Affect Disord. 2023;328:1–5.

Coles NA, Gaertner L, Frohlich B, Larsen JT, Basnight-Brown DM. Fact or artifact? Demand characteristics and participants’ beliefs can moderate, but do not fully account for, the effects of facial feedback on emotional experience. J Pers Soc Psychol. 2023;124:287–310.

Weimer K, Colloca L, Enck P. Placebo eff ects in psychiatry: mediators and moderators. Lancet Psychiatry. 2015;2:246–57.

Huneke NTM, Chamberlain SR, Baldwin DS, Grant JE. Diverse predictors of treatment response to active medication and placebo in gambling disorder. J Psychiatr Res. 2021;144:96–101.

Leucht S, Chaimani A, Mavridis D, Leucht C, Huhn M, Helfer B, et al. Disconnection of drug-response and placebo-response in acute-phase antipsychotic drug trials on schizophrenia? Meta-regression analysis. Neuropsychopharmacology. 2019;44:1955–66.

Welten CCM, Koeter MWJ, Wohlfarth T, Storosum JG, van den Brink W, Gispen-de Wied CC, et al. Placebo response in antipsychotic trials of patients with acute mania. Eur Neuropsychopharmacol. 2015;25:1018–26.

Grant JE, Chamberlain SR. The placebo effect and its clinical associations in gambling disorder. Ann Clin Psychiatry. 2017;29:167.

Download references

Acknowledgements

Dr Nathan TM Huneke is an NIHR Academic Clinical Lecturer. The views expressed in this publication are those of the author(s) and not necessarily those of the NIHR, NHS, or the UK Department of Health and Social Care. For the purpose of open access, the author has applied a Creative Commons Attribution License (CC BY) to any Author Accepted Manuscript version arising from this submission.

Author contributors

NTMH, JA, DSB, SRC, CUC, MG, CMH, RH, ODH, JMAS, MS, and SCo conceptualized the study. NTMH, AB, VB, LE, CJG, OKF, LM, CR, SS, and SCo contributed to data collection, data curation, or data analysis. NTMH, MS, and SCo wrote the first draft of the manuscript. All authors had access to the raw data. All authors reviewed and edited the manuscript and had final responsibility for the decision to submit it for publication.

Author information

These authors contributed equally: Marco Solmi, Samuele Cortese.

Authors and Affiliations

Clinical and Experimental Sciences, Faculty of Medicine, University of Southampton, Southampton, UK

Nathan T. M. Huneke, Jay Amin, David S. Baldwin, Samuel R. Chamberlain, Matthew Garner, Catherine M. Hill, Ruihua Hou, Konstantinos Ioannidis, Julia M. A. Sinclair & Samuele Cortese

Southern Health NHS Foundation Trust, Southampton, UK

Nathan T. M. Huneke, Jay Amin, David S. Baldwin, Samuel R. Chamberlain, Konstantinos Ioannidis & Satneet Singh

University Department of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa

David S. Baldwin

School of Psychology, University of Nottingham Malaysia, Semenyih, Malaysia

Alessio Bellato

Centre for Innovation in Mental Health, School of Psychology, Faculty of Environmental and Life Sciences, University of Southampton, Southampton, UK

Alessio Bellato, Valerie Brandt, Matthew Garner, Corentin J. Gosling, Claire Reed, Marco Solmi & Samuele Cortese

Clinic of Psychiatry, Social Psychiatry and Psychotherapy, Hannover Medical School, Hanover, Germany

Valerie Brandt

Department of Child and Adolescent Psychiatry, Charité Universitätsmedizin Berlin, Berlin, Germany

Christoph U. Correll

Department of Psychiatry, Zucker Hillside Hospital, Northwell Health, Glen Oaks, NY, USA

Department of Psychiatry and Molecular Medicine, Zucker School of Medicine at Hofstra/Northwell, Hempstead, NY, USA

Center for Psychiatric Neuroscience, Feinstein Institute for Medical Research, Manhasset, NY, USA

Faculty of Education and Psychology, University of Navarra, Pamplona, Spain

Luis Eudave

School of Psychology, Faculty of Environmental and Life Sciences, University of Southampton, Southampton, UK

Matthew Garner

Université Paris Nanterre, DysCo Lab, F-92000, Nanterre, France

Corentin J. Gosling

Université de Paris, Laboratoire de Psychopathologie et Processus de Santé, F-92100, Boulogne-Billancourt, France

Department of Sleep Medicine, Southampton Children’s Hospital, Southampton, UK

Catherine M. Hill

Department of Psychosis Studies, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, UK

Oliver D. Howes

H Lundbeck A/s, Iveco House, Watford, UK

Institute of Clinical Sciences (ICS), Faculty of Medicine, Imperial College London, London, UK

Department of Clinical Medicine, Aarhus University, Aarhus, Denmark

Ole Köhler-Forsberg

Psychosis Research Unit, Aarhus University Hospital–Psychiatry, Aarhus, Denmark

Department of Translational Biomedicine and Neuroscience (DIBRAIN), University of Studies of Bari “Aldo Moro”, Bari, Italy

Lucia Marzulli

Department of Psychiatry, University of Ottawa, Ottawa, ON, Canada

Marco Solmi

Department of Mental Health, Ottawa Hospital, Ottawa, ON, Canada

Ottawa Hospital Research Institute (OHRI) Clinical Epidemiology Program, University of Ottawa, Ottawa, ON, Canada

School of Epidemiology and Public Health, Faculty of Medicine, University of Ottawa, Ottawa, ON, Canada

Solent NHS Trust, Southampton, UK

Samuele Cortese

DiMePRe-J-Department of Precision and Regenerative Medicine-Jonic Area, University “Aldo Moro”, Bari, Italy

Hassenfeld Children’s Hospital at NYU Langone, New York University Child Study Center, New York, NY, USA

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Nathan T. M. Huneke .

Ethics declarations

Competing interests.

DSB is President of the British Association for Psychopharmacology, Editor of the Human Psychopharmacology journal (for which he receives an editor’s honorarium), and has received royalties from UpToDate. CMH has acted on an expert advisory board for Neurim Pharmaceuticals. ODH is a part-time employee and stockholder of Lundbeck A/s. He has received investigator-initiated research funding from and/or participated in advisory/speaker meetings organized by Angellini, Autifony, Biogen, Boehringer-Ingelheim, Eli Lilly, Heptares, Global Medical Education, Invicro, Jansenn, Lundbeck, Neurocrine, Otsuka, Sunovion, Recordati, Roche and Viatris/Mylan. ODH has a patent for the use of dopaminergic imaging. All other authors declare no competing interests. MS has received honoraria/has been a consultant for Angelini, Lundbeck, and Otsuka. SCo has received honoraria from non-profit associations (BAP, ACAMH, CADDRA) for educational activities and an honorarium from Medice. KI has received honoraria from Elsevier for editorial work. SRC receives honoraria from Elsevier for associate editor roles at comprehensive psychiatry and NBR journals. CUC has been a consultant and/or advisor to or has received honoraria from: AbbVie, Acadia, Adock Ingram, Alkermes, Allergan, Angelini, Aristo, Biogen, Boehringer-Ingelheim, Bristol-Meyers Squibb, Cardio Diagnostics, Cerevel, CNX Therapeutics, Compass Pathways, Darnitsa, Denovo, Gedeon Richter, Hikma, Holmusk, IntraCellular Therapies, Jamjoom Pharma, Janssen/J&J, Karuna, LB Pharma, Lundbeck, MedAvante-ProPhase, MedInCell, Merck, Mindpax, Mitsubishi Tanabe Pharma, Mylan, Neurocrine, Neurelis, Newron, Noven, Novo Nordisk, Otsuka, Pharmabrain, PPD Biotech, Recordati, Relmada, Reviva, Rovi, Sage, Seqirus, SK Life Science, Sumitomo Pharma America, Sunovion, Sun Pharma, Supernus, Takeda, Teva, Tolmar, Vertex, and Viatris. He provided expert testimony for Janssen and Otsuka. He served on a Data Safety Monitoring Board for Compass Pathways, Denovo, Lundbeck, Relmada, Reviva, Rovi, Supernus, and Teva. He has received grant support from Janssen and Takeda. He received royalties from UpToDate and is also a stock option holder of Cardio Diagnostics, Kuleon Biosciences, LB Pharma, Mindpax, and Quantic.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

41380_2024_2638_moesm1_esm.docx.

PLACEBO EFFECTS IN RANDOMIZED TRIALS OF PHARMACOLOGICAL AND NEUROSTIMULATION INTERVENTIONS FOR MENTAL DISORDERS: AN UMBRELLA REVIEW SUPPLEMENTARY APPENDIX

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Huneke, N.T.M., Amin, J., Baldwin, D.S. et al. Placebo effects in randomized trials of pharmacological and neurostimulation interventions for mental disorders: An umbrella review. Mol Psychiatry (2024). https://doi.org/10.1038/s41380-024-02638-x

Download citation

Received : 01 February 2024

Revised : 17 June 2024

Accepted : 19 June 2024

Published : 24 June 2024

DOI : https://doi.org/10.1038/s41380-024-02638-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List

Romantic relationship breakup: An experimental model to study effects of stress on depression (-like) symptoms

Anne m. verhallen.

University of Groningen, University Medical Center Groningen, Department of Biomedical Sciences of Cells & Systems, Cognitive Neuroscience Center, Groningen, the Netherlands

Remco J. Renken

Jan-bernard c. marsman, gert j. ter horst, associated data.

The data will not be stored at a public repository due to restrictions from the informed consent (subjects have not given consent to have their data publicly stored) and European data privacy regulations (GDPR). The data are available on request. A data access committee has been put in place, consisting of Prof G.J. ter Horst (Principal Investigator of the study), J.B.C. Marsman and Prof A. Aleman (head of the Cognitive Neuroscience Center), who will review requests and assure accessibility of the data. This access committee can be reached via [email protected] .

The occurrence of a stressful event is considered to increase the risk of developing depression. In the present study we explore whether the breakup of a romantic relationship can be used as an experimental model to study a depression-like state during a period of stress in individuals without a psychiatric disorder. The primary aim of our study was to investigate: 1) whether individuals with a recent romantic relationship breakup (‘‘heartbreak”) demonstrate symptoms of depression, 2) how to describe heartbreak characteristics based on data from a comprehensive questionnaire battery, and 3) whether this description can capture severity of depression symptoms. Secondary, we were interested in gender differences with regard to the above study objectives. Subjects who have experienced a relationship breakup in the preceding six months ( N = 71) or are in a romantic relationship ( N = 46) participated in our study. A questionnaire battery was administered to acquire information related to depression, mood, the breakup and (former) relationship. Principal Component Analysis with Procrustes bootstrapping was performed to extract components from the questionnaire data. Even though our sample of individuals who recently have experienced a relationship breakup can be on average considered non-depressed, group-level depression scores were elevated compared to individuals in a relationship ( p = .001) and 26.8% reported symptoms corresponding to mild, moderate or severe depression. We described heartbreak by two principal components interpreted as ‘‘sudden loss” and ‘‘lack of positive affect”, respectively. Highly significant correlations between the component scores and depression scores were found ( p < .001 and p < .001, respectively), although these correlations differed between the genders. Based on these findings, we propose that the experience of a romantic relationship breakup is a viable experimental model to examine symptoms of depression in individuals without a psychiatric disorder. This way, stress-related coping and depression vulnerability can be studied in further research.

Introduction

Stressful life-events are considered to be risk factors for the development of depression[ 1 ]. Kendler et al.[ 2 ] investigated the interplay between stressful events, genetic predisposition and depression among female twins and found that both heredity and occurrence of stressful events contributed to the onset of depressive episodes independently. Especially events with a high impact, such as death of a close family member and divorce, elevated the probability of developing a depressive episode[ 2 ], although the majority of people do not develop a depressive episode following the experience of an upsetting event. Hence, research focusing on stressful and emotionally upsetting events can give valuable insights into individual differences regarding stress-related coping and the link between stress and depression.

In this study we set out to investigate mood and depression symptoms during a period of stress in a population without a psychiatric disorder. More precisely, we explore whether the breakup of a romantic relationship can be used as an experimental model to study a depression-like state. Previous research already showed that the breakup of a romantic relationship can be seen as an emotionally upsetting event that can lead to multiple symptoms related to sadness, grief and depression and even can result in an increased risk of developing a depressive episode[ 3 – 6 ]. In a university student sample, severe breakup distress, measured with a questionnaire concerning symptoms of grief, was accompanied by feelings of betray and rejection, depression symptoms, anxiety symptoms, intrusive thoughts about the ex-partner and sleep disturbances[ 3 ]. The elapsed time since the breakup, self-reported quality of the former relationship, feelings of betray and depression scores especially predicted the severity of breakup distress[ 3 ]. Additionally, women reported higher breakup distress scores compared to men in that study[ 3 ]. In a study of Stoessel et al.[ 4 ], all of the subjects with a relationship breakup in the preceding six months and experiencing feelings of sadness about the breakup reported symptoms corresponding to clinical depression. In women with a breakup in the preceding four months, high levels of complicated grief (extreme symptoms of grief interfering with daily life functioning) were present in four of the eight subjects. In addition, a different brain pattern (increased activity in posterior regions such as the cerebellum and decreased activity in anterior regions such as the insula and temporal cortex) was found in these women when ruminating about their ex-partner in comparison with thinking about an acquaintance in a neutral manner[ 5 ]. Moreover, epidemiological data indicated an association between the occurrence of a romantic relationship breakup and first onset of major depression in a young population[ 6 ].

As it is known that the prevalence of depression is higher in women, we were also interested in differences in depression (-like) symptoms between the genders in our study. For example, data from a United States survey revealed a 1.7 times higher lifetime prevalence of depressive episodes among women[ 7 ]. Differences in stress sensitivity between the genders could play a role, as stress paradigms in rodents elucidated different stress responses between males and females[ 8 , 9 ]. Moreover, gender differences with regard to rumination might be involved. It is known that women tend to ruminate more during periods of stress[ 10 ]. A ruminative coping strategy was associated with both anxiety and depression symptoms and correlated significantly with occurrence of new depressive episodes in patients with major depressive disorder[ 11 ]. In addition, experiencing ruminating thoughts about the loss during grief was found to be related to maladaptive grieving and the development of symptoms of depression[ 12 ].

In the present study, we primarily aimed to investigate: 1) whether individuals with a recent romantic relationship breakup (‘‘heartbreak”) demonstrate symptoms of depression, 2) how to describe heartbreak characteristics based on data from a comprehensive questionnaire battery, and 3) whether this description can capture severity of depression symptoms. Secondary, we were interested in gender differences with regard to the above study objectives. To this end, young men and women, either with a recent romantic relationship breakup (the ‘‘heartbreak group”) or in a romantic relationship (the ‘‘relationship group”) participated. The relationship group was included in the study as a reference group with absence of stress resulting from a romantic relationship breakup. We expected a higher severity of depression symptoms in the heartbreak group compared to the relationship group. Given that women are more at risk for developing depression in the general population, we expected a higher severity of depression symptoms among the women in the heartbreak group than the men in the heartbreak group.

Materials and methods

Experimental design.

Subjects were invited to our laboratory to participate in the study between 2011 and 2013. The experiment comprised a self-report questionnaire battery and fMRI paradigm with a cross-sectional design. fMRI results will be reported elsewhere. Before the start of the study, written informed consent was obtained from every subject. Study procedures were approved by the Medical Ethical Committee of the University Medical Center Groningen and conducted in accordance with the principles of the Declaration of Helsinki. Subjects received a financial compensation for their participation.

Recruitment strategy

Subjects were recruited by distributing posters around faculty buildings of the University of Groningen and promoting the study using (social) media. Women of the heartbreak group (‘‘heartbroken females”) were recruited using recruitment material with terminology implying that subjects have to suffer from breakup distress to participate. With this recruitment strategy it was not possible to include a sufficient number of male subjects. Therefore, a subsample of the men of the heartbreak group (‘‘heartbroken males”) was recruited using recruitment material referring to the experience of a relationship breakup instead of suffering from breakup distress. Potential subjects could send an email to show their interest in the study and exchange contact information. A telephone intake interview was planned to explain study procedures and check inclusion and exclusion criteria. Additionally, subjects received written information. During the first stage of the study, heartbroken females were pre-selected at the intake interview based on their self-report level of sadness about the breakup on a scale from 1 to 10, because at that time we intended to compare women with contrasting levels of breakup distress. For the results presented in this paper we did not divide the heartbreak group in subgroups based on information obtained at the intake interview.

Inclusion and exclusion criteria

For both the heartbreak group and the relationship group, subjects had to be between the age of 18 and 26 years, right-handed, Western, heterosexual and Dutch speaking. Female subjects could only participate if they used hormonal contraception and were in the continuation phase on the day of the experiment to minimize possible effects of fluctuating sex hormone levels on our outcome measures. To participate in the heartbreak group, subjects had to have a relationship breakup within the preceding six months and a relationship duration of at least six months. To participate in the relationship group, subjects had to have a relationship duration between 6 and 24 months because we intended to include subjects whose relationship has not yet evolved into a companionate stage[ 13 ]. Subjects with a relationship duration shorter than 6 months were excluded due to previous research on increased stress hormone level during these first periods[ 14 ]. For both the heartbreak group and the relationship group, subjects with neurological abnormalities, MRI contraindications such as ferromagnetic metal parts in the body, (suspected) pregnancy and claustrophobia, use of psychotropic medication in the last five years, alcohol and/or drug abuse and physical and/or sexual abuse during the relationship (all self-reported) were not allowed to participate in the study. 71 and 46 subjects were included in the heartbreak group and the relationship group, respectively. After the fMRI scanning session, one subject from the heartbreak group was excluded because of substantial brain ventricle abnormalities.

Questionnaire battery

A self-report questionnaire battery in Dutch was administered to assess psychological and behavioral information of the subjects. Before filling in the questionnaire battery, background information, such as highest completed educational level according to the Dutch educational system and current occupation status, was acquired from the subjects. The heartbreak questionnaire battery consisted of the Major Depression Inventory (MDI) and adjusted versions of the Inventory of Complicated Grief (ICG), Positive and Negative Affect Schedule (PANAS), Perceived Relationship Quality Components Inventory (PRQC), the Hurt-Proneness Scale and the Passionate Love Scale (PLS)[ 13 , 15 – 19 ]. Additionally, in-house designed questions about the breakup were added to the questionnaire battery, covering aspects such as unexpectedness of the breakup and ruminating thoughts about the ex-partner. For each questionnaire, except the in-house designed questionnaire about the breakup, total scores were calculated and used in further analyses. Cronbach’s alpha scores were calculated for each questionnaire (can be found in S1 Table ). The MDI is a 10-item questionnaire to assess symptoms of depression (both core symptoms such as anhedonia and accompanying symptoms such as sleeping difficulties), based on the DSM-IV and ICD-10 diagnostic criteria[ 15 , 20 ]. MDI scores were calculated according to the scoring guidelines for the use of the MDI as a rating scale to measure severity of depression symptoms[ 20 ]. MDI scores theoretically range between 0 and 50. Scores between 0 and 20 indicate absence of clinical depression, scores between 21 and 25 correspond to mild depression, scores between 26 and 30 and scores above 31 indicate respectively moderate and severe depression[ 20 ]. The ICG is used to assess maladaptive grieving after the loss of a loved one[ 16 ]. Similar to the study of Najib et al.[ 5 ], the ICG was adjusted so that it was suitable for heartbreak. Thirteen items were extracted from the original 19-item version, by removing items only applicable to death. ICG scores were calculated by summing the scores of the 13 questions and theoretically range between 13 and 130. The PANAS comprises questions about positive and negative affect, representing current mood[ 17 ]. PANAS scores were calculated for both the positive affect and negative affect part by summing the scores of the 10 questions and theoretically range between 10 and 100[ 17 ]. The PRQC was used to assess self-reported former relationship quality[ 18 ]. In this study a 9-item version of the PRQC was extracted from the original 18-item version. PRQC scores were calculated by summing the scores of the 9 questions and theoretically range between 9 and 90. To what extent the subjects were prone to experience hurt feelings was measured with the Hurt-Proneness Scale[ 19 ]. Hurt proneness scores were calculated by summing the scores of the 6 questions and theoretically range between 6 and 60. Questions 3, 4 and 6 were reversed scored because high scores characterize low hurt proneness. The PLS can be used to assess intensity of romantic love[ 13 ]. A 28-item version of the PLS was extracted from the original 30-item version, by removing two items that are not appropriate for heartbreak. The PLS was filled in exclusively by the subjects who reported to be still in love with their ex-partner at the time of the testing day. PLS scores were not analyzed further. As the PLS was only filled in by the heartbroken subjects who reported to be still in love with their ex-partner, the sample size turned out to be insufficient. All questionnaires, except the MDI, were scored on a 10-point Likert scale, ranging from 1 (‘‘not at all”) to 10 (‘‘extremely”). The MDI was rated on a 6-point Likert scale, ranging from 1 (‘‘not at all”) to 6 (‘‘all the time”). Questions about the relationship breakup were categorical or measured on a 10-point Likert scale. The questionnaire battery of the relationship group consisted of the MDI and adjusted versions of the PANAS, PRQC, Hurt-proneness scale and PLS, similar to the heartbreak group.

Data analysis

Statistical analyses were conducted with IBM SPSS Statistics version 24 for Windows. A Shapiro-Wilk test for normality was used to check if our data were normally distributed. When data distribution was found to be skewed, non-parametric statistical tests were conducted in further analysis steps.

Group-level comparisons

Background information and questionnaire data were compared between the heartbreak group and the relationship group using a Mann-Whitney U test. Concerning the questionnaire battery of the relationship group, only MDI scores are considered in this manuscript, since we aimed to compare severity of depression symptoms between the heartbreak group and the relationship group.

Principal component Analysis-varimax

An exploratory Principal Component Analysis (PCA) followed by varimax rotation was performed to extract components representing heartbreak in a data-driven manner. We intended to focus on subjective measures. Consequently, 19 variables were entered into the analysis; ‘‘unexpectedness breakup”, ‘‘feeling rejected”, ‘‘feeling betrayed”, ‘‘feeling angry”, ‘‘feeling sad”, ‘‘feeling disappointed”, ‘‘feeling independent”, ‘‘feeling alone”, ‘‘feeling relieved”, ‘‘feeling hopeful”, ‘‘ruminating thoughts”, ‘‘intrusive thoughts”, ‘‘affection for ex-partner”, ‘‘in love with ex-partner”, ‘‘ICG”, ‘‘PANAS positive”, ‘‘PANAS negative”, ‘‘PRQC” and ‘‘Hurt proneness”. Subjects with missing data were deleted listwise, resulting in a sample size of 69 for the PCA. Principal components were extracted using the correlation matrix, and rotated with varimax with Kaiser normalization[ 21 ]. Parallel analysis was performed to determine the optimal number of components[ 22 ]. We adapted the online available SPSS script for parallel analysis, written by Brian O’Connor, to our dataset ( https://people.ok.ubc.ca/brioconn/nfactors/nfactors.html ) [ 23 ]. Thousand sets of normally distributed data were randomly generated. For each component an eigenvalue belonging to the original data and an eigenvalue belonging to the 95% confidence interval (CI) of the generated data was computed. Components with an eigenvalue greater than the corresponding eigenvalue derived from random normal data generation were considered as ‘‘components”. Subsequently, a PCA followed by a varimax rotation was performed with a fixed number of components to extract, based on the results of the parallel analysis. The outcome of this combined PCA and varimax rotation, a component matrix, was used in the subsequent analyses.

Procrustes bootstrapping

A Procrustes bootstrapping PCA was performed to select the component loadings to be interpreted further. Thousand samples of component matrices were generated by resampling with replacement. To this end, we adjusted the online available SPSS script for component analysis with Procrustes bootstrapping from Linda Reichwein Zientek and Bruce Thompson[ 24 ]. Note that, just like the original PCA-varimax, components were not normalized row wise. Bootstrapping results were rotated towards a target matrix. The target matrix was constructed by binarizing the component matrix retaining the sign. Variables were assigned 1 or -1 for the component they loaded strongest on and 0 elsewhere. 95% CIs were calculated for each variable across the thousand bootstrap resamples.

Interpretation components

Only variables with a 95% CI that does not cross zero were interpreted further. Labels were assigned to each component based on the component loadings acquired with the original PCA-varimax.

Component scores analysis

For each subject, component scores were computed using regression. A Spearman rank test was conducted to see how well the component scores correlate with MDI scores. A Spearman rank test was used to assess the correlation between the component scores, time since breakup and relationship duration. Component scores were compared between men and women with an independent samples t-test. Additionally, Spearman rank correlations between the component scores and MDI scores were calculated for men and women separately.

For all conducted statistical tests, results were considered significant at p -value < .05 (uncorrected), two-tailed.

Study population

The relationship group consisted of 23 men and 23 women with a relationship duration between 6 and 24 months ( Mdn = 13.00, IQR = 9.00–19.00). Age ranged between 18 and 26 years ( Mdn = 21.00, IQR = 20.00–23.00). The heartbreak group consisted of 33 men and 38 women. Age ranged between 18 and 25 years ( Mdn = 22.00, IQR = 21.00–24.00). Relationship duration prior to the breakup ranged between 6 and 81 months ( Mdn = 20.00, IQR = 13.00–37.00). Time since breakup ranged between 0 and 5 months ( Mdn = 2.00, IQR = 1.00–4.00). In 42.3% of the subjects, the ex-partner decided to break up, whereas in 35.2% the breakup was initiated by the subject and in 22.5% the subject and ex-partner decided together. 70.4% of the subjects reported to still be in touch with their ex-partner. Five subjects (7.0%) reported to have found a new romantic partner. 70.4% reported to still think about their ex-partner on a daily basis and 25.4% experienced physical complaints after the breakup. The heartbreak group was significantly older than the relationship group ( U = 1241.00, Z = -2.21, p = .027, r = -.20). Additional background information of our study population can be found in S2 Table .

Severity of depression symptoms

Fig 1 shows the severity of depression symptoms for the relationship group and the heartbreak group. MDI total scores ranged between 2 and 22 ( Mdn = 7.00, IQR = 4.75–10.25) in the relationship group. 97.8% were found to have a MDI score below 21, corresponding to an absence of depression. 2.2% had depression symptoms corresponding to mild depression. MDI scores ranged between 1 and 45 ( Mdn = 9.00, IQR = 7.00–21.00) in the heartbreak group. 12.7% reported depression symptoms corresponding to mild depression. 1.4% and 12.7% reported symptoms corresponding to respectively moderate and severe depression. In total, 26.8% reported symptoms corresponding to mild, moderate or severe depression. MDI total scores were higher in the heartbreak group compared to the relationship group ( U = 1042.00, Z = -3.31, p = .001, r = -.31). No gender differences were found ( U = 213.50, Z = -1.13, p = .260, r = -.17) between the males ( Mdn = 6.00, IQR = 4.00–8.00) and females ( Mdn = 7.00, IQR = 5.00–14.00) in the relationship group. MDI scores differed ( U = 380.00, Z = -2.85, p = .004, r = -.34) between heartbroken males ( Mdn = 7.00, IQR = 4.50–14.00) and heartbroken females ( Mdn = 15.50, IQR = 7.75–25.00). S3 Table shows the median and interquartile range for the individual items for the two groups. With regard to the core symptoms of depression, the heartbreak group scored higher on the item ‘‘feeling sad or low in spirits” and the item ‘‘loss of interest in daily activities” ( p = .001 and p = .013), while the item ‘‘lack of energy and strength” did not differ between the two groups ( p = .218). Concerning the accompanying symptoms of depression, the items ‘‘feeling less self-confident”, ‘‘the feeling that life was not worth living”, ‘‘concentration difficulties”, ‘‘feeling restless/listless” and ‘‘sleeping difficulties” differed significantly between the two groups (all higher in the heartbreak group, p = .019, p = .002, p = .019, p < .001, and p = .004, respectively). No differences were found regarding the items ‘‘feelings of guilt” and ‘‘decreased/increased appetite” ( p = .112 and p = .151).

Outliers (values that are between Q1-1.5*IQR or Q3+1.5*IQR and Q1-3*IQR or Q3+3*IQR) are indicated with a circle. Extreme outliers (values that are beyond Q1-3*IQR or Q3+3*IQR) are indicated with a star.

Characterization of heartbreak

To characterize the heartbreak group, a PCA-varimax was performed on the questionnaire battery. Subsequently, the relation with the depression scores was investigated.

Components extraction

Parallel analysis revealed two components with corresponding eigenvalues greater than the concomitant eigenvalue calculated for a random dataset, explaining respectively 45.8% and 13.8% of the variance.

Component loadings and interpretation

Table 1 shows the component loadings for the included variables for the two components. Additionally, the 95% CIs of the component loadings are shown. Given that the variables ‘‘feeling betrayed”, ‘‘feeling rejected”, ‘‘feeling angry”, ‘‘unexpectedness breakup” and ‘‘ICG” load highly on component 1 (95% CI does not straddle zero), this component was interpreted as ‘‘sudden loss”. The variables ‘‘feeling hopeful” and ‘‘PANAS positive” load strongest (inversely) on component 2 and have 95% CIs that do not contain zero. Therefore, this component was interpreted as ‘‘lack of positive affect”.

Association between the components and time since breakup and relationship duration

No significant correlation between time since breakup and each of the two components was found (‘‘sudden loss”: r s = .06, p = .600, “lack of positive affect”: r s = -.22, p = .071). Relationship duration correlated significantly with the ‘‘lack of positive affect” component ( r s = .25, p = .039) and did not correlate significantly with the ‘‘sudden loss” component ( r s = -.07, p = .559).

Association between the components and depression scores

Positive correlations between the component scores belonging to the two extracted components and depression scores were prevalent ( r s = .57, p < .001 and r s = .49, p < .001 for the ‘‘sudden loss” component and the ‘‘lack of positive affect” component, respectively). The scatterplot between the ‘‘sudden loss” component and MDI and between the ‘‘lack of positive affect” component and MDI are shown in Fig 2A and Fig 2B , respectively.

(A) Relationship between the ‘‘sudden loss” component and MDI. (B) Relationship between the ‘‘lack of positive affect” component and MDI.

Gender effects components

Gender differences with regard to the ‘‘lack of positive affect” component were found: heartbroken females had higher component scores ( t (67) = 2.95, p = .004, r = .34). Component scores belonging to the ‘‘sudden loss” component did not differ between the genders ( t (67) = .88, p = .385, r = .11). MDI scores correlated positively with both components (see above). However, the MDI score showed a gender effect as well (see above). Therefore, correlations between MDI scores and each of the two components were examined for men and women separately. For heartbroken females, highly significant correlations were found for both ‘‘sudden loss” and ‘‘lack of positive affect” ( r s = .57, p < .001 and r s = .70, p < .001 respectively). Heartbroken males showed a partially different result: a significant correlation between MDI scores and the ‘‘sudden loss” component was prevalent ( r s = .55, p = .001). In contrast, the ‘‘lack of positive affect” component did not correlate significantly with MDI scores ( r s = -.01, p = .951).

In the present study, we primarily aimed to investigate: 1) whether individuals with a recent romantic relationship breakup demonstrate symptoms of depression, 2) how to describe heartbreak characteristics based on data from a comprehensive questionnaire battery, and 3) whether this description can capture severity of depression symptoms. Secondary, we were interested in gender differences with regard to the above study objectives.

In accordance with our expectations, severity of depression symptoms was found to be higher in the heartbreak group compared to the reference group, i.e. subjects in a romantic relationship. MDI total scores as well as individual items, including core symptoms of depression, were elevated. However, median MDI scores of the heartbreak group fell within the range of absence of depression as defined by Bech et al.[ 20 ]. Nonetheless, 26.8% and 14.1% of the heartbroken subjects reported severity of depression symptoms corresponding to respectively mild to severe depression and moderate to severe depression. In contrast, only one subject reported symptoms corresponding to (mild) depression in the relationship group. In a study by Forsell et al.[ 25 ], a mean MDI score of 8.8 (95% CI 8.6–9.0) was found in a large sample of men and women drawn from the general population. Note that even in this general population, 8.0% reported symptoms corresponding to moderate or severe depression[ 25 ] (compared to the 14.1% found in this study). Thus, we consider the heartbreak group as a good population to study a depression-like state in otherwise healthy individuals.

We described heartbreak by two principal components. Feelings of betray, rejection and anger, unexpectedness of the breakup and symptoms of complicated grief contributed substantially to the first component that was therefore interpreted as ‘‘sudden loss”. Feeling hopeful after the breakup and current positive affect (i.e. the ability to experience positive emotions) contributed largely (inversely) to the second component that was consequently interpreted as ‘‘lack of positive affect”. The finding that the feeling of being betrayed is an important parameter of heartbreak is consistent with the study of Field et al.[ 3 ]. Moreover, our findings show similarities with a retrospective study concerning emotions following a relationship dissolution by Barbara and Dion[ 26 ]. In that study, a component labeled as ‘‘negative emotions” was extracted and rejection and anger were found to be important variables for that specific component[ 26 ]. This is in accordance with the high loadings of the variables ‘‘feeling rejected” and ‘‘feeling angry” on the ‘‘sudden loss” component as found in our study.

Within the heartbreak group, both components correlated highly with depression scores. The ‘‘lack of positive affect” component is primarily defined by positive affect scores, as measured with the PANAS. This is in accordance with a study by Crawford and Henry[ 27 ] in which positive affect was found to be specifically related to depression scores in a large sample of men and women drawn from the general population. The ‘‘sudden loss” component also correlated highly with depression scores. This is consistent with literature regarding grief. For example, Keyes et al.[ 28 ] found associations between the experience of an unexpected death of a loved one and prevalence of psychiatric problems including clinical depression.

As expected, heartbroken females reported higher depression scores than heartbroken males in our study. This cannot be explained by general gender differences, given that the depression scores of the men and women of the relationship group did not differ, and therefore seems to be breakup-related. Among the heartbroken males examined separately, no association between the ‘‘lack of positive affect” component and severity of depression symptoms was found. Additionally, women had higher scores on the ‘‘lack of positive affect” component than men. Tentatively, these findings suggest that men are less likely to demonstrate and/or report reduced abilities experiencing positive emotions during a period of stress than women and this possibly relates to the well-known differential depression rates among the genders.

Limitations

By conducting the present study, detailed knowledge about behavioral and psychological consequences of a recent romantic relationship breakup and its association with symptoms of depression was acquired. A potential weakness of our study is that differences in recruitment strategy and pre-selection prior to inclusion between the genders could have influenced our findings. This makes it difficult to draw strong conclusions about effects of gender. Nevertheless, gender-specific application rates can be considered a finding as well in our opinion. Another possible weakness is that already having a new romantic partner was not an exclusion criterion in our study and in our sample five of the 71 subjects reported to have found a new partner on the day of the experiment. One could argue that this will reduce sadness and mood problems associated with a breakup. However, excluding those subjects from our dataset did not change either group-level differences regarding depression scores or the strength of the correlation between the components and depression scores noticeably (data can be found in S1 Appendix ). Therefore, possible effects of having a new romantic partner on our results were considered minimal. Perhaps, finding a new partner cannot diminish breakup-related effects within this limited period of time.

Conclusions

In the present study, we investigated whether the breakup of a romantic relationship can be used as an experimental model to study a depression-like state. We demonstrated an increased range of depression scores among our sample of individuals who recently have experienced a relationship breakup. Furthermore, our results show that the effects of experiencing a relationship breakup can be captured with two descriptors: “sudden loss” and “lack of positive affect”. Both were associated with (severity of) depression (-like) symptoms. Nota bene, this association was gender-dependent. Therefore, we propose that this life-event is a viable experimental model to investigate symptoms of depression in individuals without a psychiatric disorder. This paves the way to investigate the involvement of stress in the transition from healthy-to depressive behavior. Consequently, further longitudinal research using this model could provide new insights into individual-specific coping and vulnerability factors contributing to the development of depression symptoms during a period of stress.

Supporting information

Values are shown as percentage or median (Q1-Q3) for respectively categorical variables and numerical variables.

Values are shown as median (Q1-Q3).

S1 Appendix

Acknowledgments.

The authors would like to thank the undergraduate students who contributed to the design of the study, recruitment of subjects and data collection (Els van der Meijden, Antina de Boer, Femke van der Velde, Lisa Brouwer, Dafne Piersma, Floor Rodijk, Renske Lok, Kenney Roodakker and Thom Steenhuis). In addition, the authors would like to thank Dr. Marie-José van Tol and Sonsoles Alonso Martinez, MSc for contributing to the interpretation of the results.

Funding Statement

The study was funded by a donation of Mr Hazewinkel to the Research School of Behavioural and Cognitive Neurosciences and Prof G.J. ter Horst. Mr Hazewinkel had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability

Help | Advanced Search

Computer Science > Human-Computer Interaction

Title: using large language models to assist video content analysis: an exploratory study of short videos on depression.

Abstract: Despite the growing interest in leveraging Large Language Models (LLMs) for content analysis, current studies have primarily focused on text-based content. In the present work, we explored the potential of LLMs in assisting video content analysis by conducting a case study that followed a new workflow of LLM-assisted multimodal content analysis. The workflow encompasses codebook design, prompt engineering, LLM processing, and human evaluation. We strategically crafted annotation prompts to get LLM Annotations in structured form and explanation prompts to generate LLM Explanations for a better understanding of LLM reasoning and transparency. To test LLM's video annotation capabilities, we analyzed 203 keyframes extracted from 25 YouTube short videos about depression. We compared the LLM Annotations with those of two human coders and found that LLM has higher accuracy in object and activity Annotations than emotion and genre Annotations. Moreover, we identified the potential and limitations of LLM's capabilities in annotating videos. Based on the findings, we explore opportunities and challenges for future research and improvements to the workflow. We also discuss ethical concerns surrounding future studies based on LLM-assisted video analysis.

Submission history

Access paper:.

• HTML (experimental)
• Other Formats

References & Citations

• Google Scholar
• Semantic Scholar

BibTeX formatted citation

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

• Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .