Abstract
Seasonal affective disorder is a pattern of depressive episodes that recur at the same time each year, usually beginning in autumn or winter and remitting in spring or summer. DSM-5-TR does not list it as a separate disorder but as the specifier with seasonal pattern, attached to recurrent major depressive episodes or to bipolar disorder, and ICD-11 has a comparable seasonal qualifier. Winter episodes lean towards hypersomnia, increased appetite and carbohydrate craving; photoperiod and melatonin, circadian phase-shift, serotonin, retinal and cognitive-behavioural accounts each have some support but none is established, and the disorder's validity and its link to latitude remain disputed. Bright light, dawn simulation, cognitive-behavioural therapy and antidepressants are the tested treatments, and bupropion has modest placebo-controlled support for prevention. The page also covers lived experience, cognition, imaging, culture, comorbidity, suicide and light-therapy safety.
Keywords: seasonal affective disorder, winter depression, seasonal pattern, light therapy, circadian rhythm
Seasonal affective disorder is a recurrent depressive condition defined by its timing rather than by a distinctive set of symptoms: episodes return at about the same point in the year, most often arriving in autumn or winter and lifting in spring or summer (Magnusson & Boivin, 2003). The current American manual does not treat it as a diagnosis of its own; DSM-5-TR records the seasonal timing as a specifier added to recurrent major depressive disorder or to bipolar disorder (American Psychiatric Association, 2022). The literature also refers to the condition as winter depression, the common autumn-winter form, and as seasonal mood disorder, and MeSH indexes these names, together with the plural seasonal affective disorders, as entry terms for the same descriptor. The condition matters to cognitive psychology for two reasons. People with the disorder have shown slower processing speed and weaker working memory in summer as well as in winter in a small longitudinal study (Hjordt et al., 2017), and the main psychological account gives beliefs about the seasons a role, supported by a trial in which change in those beliefs tracked recovery during cognitive-behavioural therapy (Rohan et al., 2020). This page covers depression that follows the seasons; depressive mood in general is covered on the depression page, and the diagnosis to which the seasonal-pattern specifier most often attaches is covered on the major depressive disorder page.
The article begins with what seasonal affective disorder is, including the dispute over whether it is a distinct disorder, and with the history of the concept from the first studies of light and melatonin to its place in the diagnostic manuals. It then turns to symptoms, diagnosis and measurement, where the limits of retrospective questionnaires are set out, to the experience of the disorder as patients describe it, and to the types of seasonal affective disorder, from the winter and summer patterns to subsyndromal and childhood forms. Prevalence, course and the latitude debate follow, and then the biological accounts, sleep and circadian rhythms, and the integrative cognitive-behavioural model. Three sections examine cognition directly: attention, memory and executive function; tests of whether cognition causes seasonal episodes; and brain networks and imaging. The article then considers culture, society and who is affected, comorbidity, and suicide and self-harm. The clinical sections cover treatment evidence, the prevention of seasonal episodes, how treatments work, the side effects and safety of light therapy, and emerging treatments, before a section on related conditions and seasonal mood in the general population. A worked example shows what a positive screening result means, and the article closes with current directions in research.
Key Takeaways
- DSM-5-TR treats seasonal affective disorder as the specifier with seasonal pattern, attached to recurrent major depression or bipolar disorder rather than defined as a disorder of its own, and whether the pattern marks a distinct condition remains disputed.
- Most cases follow a winter pattern; a summer pattern, with reduced appetite and insomnia, is uncommon, and milder subsyndromal seasonality shades into the general population as a continuum.
- Retrospective questionnaires such as the Seasonal Pattern Assessment Questionnaire work as screens but tend to overestimate prevalence, prospectively tracked symptoms show only modest seasonal effects, and latitude has a modest association with questionnaire-defined prevalence while explaining little of its variation.
- Photoperiod and melatonin, phase-shift, serotonin and retinal accounts each have some evidence, but the phase-shift account is contested, the retinal account is preliminary, and no single biological mechanism is established.
- The dual vulnerability and cognitive-behavioural models link a light-triggered vegetative change to rumination and negative beliefs about the seasons; the model is supported in part, mostly by trials from its own developers.
- In one small study, people with the disorder showed working-memory, processing-speed and motor-speed deficits in summer as well as winter, and across studies winter hypersomnia is reported more readily than it is measured.
- Bright light therapy is effective with moderate certainty, because the trials are small and patients cannot be fully blinded; a seasonal form of cognitive-behavioural therapy matched light acutely and was followed by lower recurrence in the second winter in one research group's trials.
- Bupropion XL is the only preventive treatment with placebo-controlled support, and its benefit is modest; vitamin D outperformed phototherapy in one small trial but was no better than placebo in the later, underpowered placebo-controlled trial, so its value is unresolved.
- Population suicide peaks in spring rather than winter, so the peak is not evidence about risk in winter seasonal depression; the imaging findings come from small samples, and most research comes from Europe and North America.
- The disorder was strongly associated with alcohol use disorder in a Finnish population survey, and eye examinations after light therapy found no damage, while emergent hypomania is a named risk of treatment.
What Seasonal Affective Disorder Is
MeSH defines seasonal affective disorder as a syndrome of depressions that recur annually at the same time each year, usually during the winter months, with other symptoms that include anxiety, irritability, decreased energy, increased appetite with carbohydrate cravings, longer sleep and weight gain, and it notes that the condition can be treated by daily exposure to bright artificial light during the season of recurrence (MeSH, D016574). MeSH files the descriptor under depressive disorder and lists winter depression, seasonal mood disorder and seasonal affective disorders as entry terms.
A Specifier, Not a Separate Diagnosis
The current American manual does not give seasonal affective disorder a diagnostic category of its own. In DSM-5-TR it appears as the specifier with seasonal pattern, which a clinician attaches to recurrent major depressive episodes or to bipolar disorder rather than diagnosing a standalone disorder (American Psychiatric Association, 2022). Research protocols use the corresponding diagnostic name, major depression, recurrent with seasonal pattern, to define who has the condition (Rohan et al., 2024). As one review summarises the DSM-5 requirements, episodes must begin and lift at a characteristic time of year, and this timing must have held for at least two years (Melrose, 2015). In the same review's paraphrase, the person must also have had more seasons with depression than seasons without it over a lifetime (Melrose, 2015).
It entered the literature with a 1984 case series that described the syndrome and reported a preliminary benefit from bright light (Rosenthal et al., 1984); that account is traced under History of the Concept.
Reviews frame the condition in two complementary ways. One treats it as a recurrent mood disorder defined by its seasonal course, while cautioning that the disorder is less common than the prevalence of seasonal patterns in mood might lead a reader to expect (Partonen & Lönnqvist, 1998). Another emphasises the typical shape of the cycle: depression that arrives in winter and lifts in spring or summer (Magnusson & Boivin, 2003). A clinical review built on Medline and Cochrane searches added that comorbidity is common and that its pathophysiology seems to be heterogeneous, since circadian, neurotransmitter and genetic hypotheses have produced discrepant results (Westrin & Lam, 2007). Clinicians also have an overview organised as common questions and answers about the disorder (Galima et al., 2020).
How ICD-11 Handles It
The World Health Organization's counterpart is the ICD-11 Clinical Descriptions and Diagnostic Requirements for mental, behavioural and neurodevelopmental disorders, known as the CDDR (World Health Organization, 2024). Its clinical descriptions provide a seasonal-pattern qualifier that applies to recurrent depressive disorder and to bipolar disorders rather than a separate category (World Health Organization, 2024). A systematic review observes that neither classification recognises seasonal affective disorder as a diagnosis in its own right, and that ICD-11 places seasonal pattern among the specifiers for mood disorders, so the two manuals treat the condition in the same way (Øverland et al., 2019).
Clinical guidance treats the seasonal label as more than a descriptive tag. A practice guideline for major depressive disorder lists seasonal pattern among the course specifiers, with episodes concentrated in winter, and singles it out, together with psychotic features, as one of the few specifiers that carry value for choosing a treatment (Lam et al., 2024). The same guideline judges that specifiers in general add only limited value to treatment selection, and it presents its recommendations as guidance rather than fixed standards (Lam et al., 2024).
Is It a Distinct Disorder?
Whether seasonal depression is a real, separable condition has been disputed for more than a decade. The sharpest challenge came from a cross-sectional, population-based survey of U.S. adults that scored depression on the Patient Health Questionnaire-8 and tested it against latitude, season and sunlight exposure in regression models (Traffanstedt et al., 2016). None of the three predicted depression significantly, and the authors concluded that the idea of seasonal depression may be rooted in folk psychology but is not supported by objective data (Traffanstedt et al., 2016). The design was cross-sectional and measured depression in general, not diagnosed seasonal episodes followed through the year.
An earlier critique had reached a related position by another route. Hansen and colleagues' narrative review concluded that SAD does not hold up as a valid medical construct and proposed calling it recurrent depression with a seasonal pattern; winter depression in northern Norway, it reported, is no commoner than in general populations elsewhere (Hansen et al., 2008). The same review rated the Seasonal Pattern Assessment Questionnaire as psychometrically rather poor and said the subtype needs further study (Hansen et al., 2008).
The 2016 survey drew an exchange. A commentary on the paper by Traffanstedt, Mehta and LoBello took up the question of whether seasonal affective disorder exists (Young, 2017). Young held that the critique had confused the seasonal pattern with the milder winter blues; in reply, LoBello contended that Young had in effect redefined the diagnosis itself, at odds both with how the seasonal modifier developed and with the DSM editions past and present (LoBello, 2017).
Evidence on the other side comes from following people with a confirmed diagnosis into the season when, by definition, their depression should lift. When the community adults with SAD from a treatment trial were followed up prospectively the next summer, 143 of the original 153 (93.5%) supplied summer data on clinician-rated and self-rated depression measures and a longitudinal interval interview (Rohan et al., 2024). Depression scores in summer were lower than at the end of treatment in both treatment arms. Complete summer remission was the commonest outcome, at 113 of 143 (79.0%), against 19 (13.3%) with partial remission and 11 (7.7%) who had a depressive episode in summer (Rohan et al., 2024). The authors took this as support for the construct validity of the diagnosis, while acknowledging a single site, a relatively homogeneous sample and incomplete response to treatment (Rohan et al., 2024).
The two lines of evidence answer different questions. The survey asked whether depression in the general population tracks light and latitude; the follow-up asked whether people already diagnosed show the seasonal course the diagnosis requires. DSM-5-TR retains the seasonal-pattern specifier (American Psychiatric Association, 2022), but no study has settled whether the pattern marks a distinct disorder or a course feature of recurrent depression, and the question remains open. The general-population evidence is taken up under Prevalence, Course and the Latitude Debate.
History of the Concept
MeSH's own indexing history records when the condition gained a heading: Seasonal Affective Disorder was introduced in 1992, and before that the literature was indexed under depressive disorder (1984 to 1991), affective disorders (1985 to 1991) and seasons (1984 to 1991) (MeSH, D016574).
What follows is the modern research history: how a laboratory finding about light became a clinical syndrome, how that syndrome was split into opposite seasonal patterns, and how it ended up in the diagnostic manuals as a pattern of recurrence rather than a disorder of its own.
Light as a Biological Signal
The modern story starts with light intensity rather than with mood. Earlier experiments had exposed people to ordinary room light at night and found no change in melatonin, which had suggested that human physiology, unlike that of other mammals, did not respond to light in this way. A small experiment in six healthy volunteers then showed that bright light did suppress night-time melatonin secretion, while room light again had no detectable effect. The authors concluded that the human response to light resembles that of other mammals once the light is bright enough (Lewy et al., 1980). Two years later the same group reported a single case: a manic-depressive patient whose mood cycled with the seasons, treated with bright artificial light (Lewy et al., 1982).
Describing the Syndrome
The syndrome was formally described in 1984, from a case series of 29 patients with recurrent depressions tied to the season. Its clinical picture included bipolar II illness, hypersomnia and carbohydrate craving. Sleep recordings in nine of the patients while they were depressed showed that they took longer to fall asleep and had less slow-wave sleep. The same report included preliminary light studies, with 11 patients, in which bright artificial light used to lengthen the photoperiod appeared to have an antidepressant effect. The authors presented that effect as preliminary, not established (Rosenthal et al., 1984).
Opposite Seasonal Patterns
Within five years the picture was no longer a single winter syndrome. A 1989 narrative review reviewed the new work. It distinguished two opposite patterns, winter depression and summer depression. It described the two patterns as having opposite vegetative symptoms and singled out winter depression as the form that responds to a specific treatment, phototherapy (Wehr & Rosenthal, 1989). This made a distinction that later classifications kept: the seasonal pattern is defined by when episodes recur.
Entry into the Diagnostic Manuals
The term seasonal affective disorder came into use in the early 1980s, and from DSM-III-R onward the DSM has let clinicians record that major depression or bipolar disorder follows a seasonal pattern (Øverland et al., 2019). That provision survives in DSM-5-TR as a specifier (American Psychiatric Association, 2022), so the manuals adopted the idea as a seasonal pattern of recurrence, not as a separate disorder. A clinic series that applied the DSM-III-R seasonal criteria retrospectively to 146 patients found the category heterogeneous, with fall-winter and spring-summer subtypes. The data came from a single research-oriented clinic (Faedda et al., 1993). The series is reported in full under Types of Seasonal Affective Disorder.
Forty Years On
A narrative review marking forty years since the 1984 description summed up the field. Winter depression is a recognised presentation that responds to bright light treatment. The review proposed that seasonal mood changes may reflect a circannual rhythm in central nervous system function driven by the yearly cycle of light, and it presented this as a possibility rather than a demonstrated mechanism. It also noted that DSM-5 treats seasonality as a specifier for both recurrent major depressive disorder and bipolar disorder. The same review raised the possibility that the condition is a legacy of Neanderthal ancestry, a suggestion it framed as a possibility rather than a finding (Rybakowski, 2024).
The current manual, DSM-5-TR, keeps that position: seasonal pattern is a specifier attached to a mood disorder diagnosis, not a diagnosis in its own right (American Psychiatric Association, 2022). Whether the seasonal pattern marks a valid, distinct condition is still disputed, and What Seasonal Affective Disorder Is sets out both sides.
Symptoms, Diagnosis and Measurement
The Symptom Profile
Seasonal affective disorder is defined by its timing rather than by a unique set of symptoms. The psychiatric manual of the American Psychiatric Association treats it as a seasonal pattern specified for major depression or bipolar disorder, so the symptoms that qualify are those of a depressive episode, and what marks the condition out is that the episodes recur with the seasons (American Psychiatric Association, 2022). Its requirements are paraphrased under What Seasonal Affective Disorder Is.
Within that frame, winter episodes lean towards the so-called atypical or vegetative profile. A patient group comparison of 53 people with seasonal affective disorder and 54 with non-seasonal major depression found more hyperphagia and more hypersomnia in the seasonal group, and less interpersonal sensitivity; the rate of formally defined atypical depression, however, did not differ significantly between the groups, and the authors read seasonal affective disorder and atypical depression as separate subtypes whose symptom pictures overlap (Tam et al., 1997). Summer-onset episodes reverse the vegetative direction, with reduced appetite and insomnia where winter patients report increased appetite, carbohydrate craving and oversleeping, a contrast covered under Types of Seasonal Affective Disorder (Wehr et al., 1991).
Eating has been measured directly as well as by self-description. Women with the disorder and female controls completed a food and drink frequency questionnaire in each of the four seasons; the patients reported eating more carbohydrate-rich food, specifically starches rather than sweets, along with more fibre and more meals a day, and their carbohydrate intake peaked in winter (Kräuchi & Wirz-Justice, 1988). Protein intake showed no group difference and no seasonal swing, and neither body weight nor body mass index changed significantly across the year, which limits how far the reported craving translated into measurable gain. The authors likened the atypical symptoms to a medial hypothalamus syndrome. A laboratory challenge points the same way: in a crossover design, 16 depressed patients and 16 matched controls ate isocaloric meals, and a carbohydrate meal left the patients feeling more activated while it sedated the controls (Rosenthal et al., 1989). The plasma amino acid response did not differ between the groups, and the order in which meals were given strongly affected the ratings, so the authors advised that meal studies control for order.
Questionnaires and Interviews
The instruments in use fall into three kinds: screening questionnaires that flag probable cases, severity rating scales that track how ill a diagnosed patient is, and diagnostic interviews that establish the diagnosis. A PRISMA systematic review covering work up to April 2024 screened 791 articles, included 28, and identified seven scales (Ambar Akkaoui & Geoffroy, 2025). The Seasonal Pattern Assessment Questionnaire (SPAQ), a screening questionnaire, remained the reference instrument, but the review judged its reliability for diagnosing seasonal depression limited and noted that it tends to overestimate prevalence. The Seasonal Health Questionnaire, another screen, outperformed it on both sensitivity and specificity. The SIGH-SAD, a clinician-rated severity scale rather than a screen, yields scores the review described as complex to interpret, and the newer scales still lacked enough validation to replace the older ones.
Earlier comparisons anticipated these judgements. In a psychometric study of 148 college students, 73 of whom were retested two months later, the SPAQ and the Inventory for Seasonal Variation, both screening questionnaires that measure seasonality as a dimension, each performed adequately, with small advantages in psychometrics and simplicity favouring the SPAQ and supporting its continued use (Young et al., 2003); a student sample says little about clinical case-finding. Contrasting patients with seasonal affective disorder against people with recurrent non-seasonal depression, Thompson and Cowan found the SPAQ less specific than other operational criteria and proposed the Seasonal Health Questionnaire as a possibly better screen once it had been developed further (Thompson & Cowan, 2001).
Confirming a seasonal pattern requires a clinical interview. A Finnish population survey of 4,554 interviewed adults used the Munich Composite International Diagnostic Interview, a structured diagnostic interview, alongside the SPAQ (Morales-Muñoz et al., 2017), and a 10-year follow-up of bipolar I and II patients assigned the seasonal pattern through structured diagnostic interviews against DSM-IV criteria (Goikolea et al., 2007).
Retrospective Reports and Their Validity
Most seasonality scores come from retrospective reports: a respondent rates, at one sitting, how mood, sleep, appetite and energy usually change across the year. That design is the field's main measurement weakness, and studies that checked questionnaire scores against interviews or against later observation have repeatedly found its limits.
Case-finding against interviews was the first test. In a validation study, 81 participants from an Icelandic community survey were interviewed clinically; for a combined 'winter problem' group of SAD and subsyndromal SAD, the SPAQ reached 94% sensitivity but only 73% specificity, and its positive predictive value was 45%, so fewer than half of those it flagged belonged to that group on interview (Magnusson, 1996). It also separated full seasonal affective disorder poorly from the subsyndromal form, and the author judged its case-finding ability poor, although clinical evaluation arrived at a prevalence of SAD similar to the questionnaire's (Magnusson, 1996). Patient samples give the mirror image. Comparing seasonal patients with depressed outpatients, non-depressed outpatients and controls, Mersch and colleagues found a specificity of 94% but a sensitivity of 44%, with 81% of participants correctly classified and false positives confined to the depressed group; most of the seasonal patients had been treated and completed the questionnaire while not depressed, which may have lowered sensitivity, and the authors concluded that the instrument suits screening but not diagnosis (Mersch et al., 2004).
Stability over time is a second problem. When patients with seasonal-pattern major depression were interviewed again at follow-up, their SPAQ responses showed low test-retest reliability and predicted only weakly whether seasonal affective disorder was confirmed, leaving the questionnaire useful as a rapid screen without the capacity to forecast the future course (Raheja et al., 1996). Prevalence figures inherit these weaknesses: estimates built on the SPAQ were misleadingly high in Thompson and Cowan's analysis (Thompson & Cowan, 2001). The season of completion is a further question; a northern Norwegian panel answered the SPAQ every quarter for a year, comparing the responses the same people gave in different seasons (Lund & Hansen, 2001).
Prospective data offer a check that does not rely on memory. Two longitudinal community cohorts in Oregon and Iowa, 762 participants aged 14 to 36 assessed between 10 and 19 times with standard depression symptom scales, showed more depressive symptoms in early winter than in early fall, but the seasonal variation was modest, recent sunlight and weather did not account for it, and the authors suggested that earlier studies relying on long-term recall may have exaggerated both how common seasonal variation in depressive symptoms is and how much it matters (Kerr et al., 2013). The cohorts were not nationally representative and did not measure vegetative symptoms, the features most closely tied to winter episodes, so the comparison is imperfect.
Taken together, retrospective seasonality scores work as screens and do not establish a diagnosis (Mersch et al., 2004), and when symptoms are tracked forward rather than recalled, the seasonal effect is modest (Kerr et al., 2013).
What a positive seasonality screen means
Choose a validation study and set how common seasonal affective disorder is among the people screened.
| Positive screens | 3,035 |
| Positive screens that are true cases | 15.5% |
| Negative screens that are correct | 99.6% |
The Experience of Seasonal Affective Disorder
People with seasonal affective disorder describe a year split into two different lives. In Stockholm, at latitude 60°N, investigators interviewed 18 outpatients with the disorder (10 women and 8 men, aged 24–65 years, each scoring at least 12 on the self-rated MADRS-S at inclusion) and analysed the transcripts by qualitative content analysis (Rastad et al., 2017). The interviewees set summer, when they had more energy and did more, against winter, when both declined. All had just completed light-room treatment, sitting in a room lit to 1500 lux for 2 h on weekdays and attending on at least 7 of 10 consecutive weekdays, and they were interviewed two weeks after it ended. They reported that their mood, sleep and energy had improved quickly, and they rated the side effects as mild and transient and smaller than the benefits. The light room had become, in their account, a central way of coping with the dark months (Rastad et al., 2017).
These are patients' own reports rather than measured outcomes, and the study has three limits that bear on them. Clinical information at inclusion was incomplete; only people who finished treatment were interviewed, so side effects that made others stop would have been missed; and the authors allow that the sample may be biased (Rastad et al., 2017).
For some patients the disorder is present even in the bright months, as anticipation. Preventing the next episode was the focus of a second interview study, in which ten patients and five physicians took part in semi-structured interviews that were then coded by theme (Nussbaumer-Streit et al., 2018). Some patients dreaded the coming winter while it was still summer, and they described lower summer well-being than patients who were not worried; the dread was stronger in those whose past episodes had been severe. Whether a patient used preventive treatment depended less on the individual alone than on circumstances: knowing more about the disorder and having had treatment in an acute phase made use more likely, and the work and social setting, healthcare access and cost also played a part. On the physicians' side, a patient's expectations and the history and stability of the illness shaped what they recommended, while a lack of evidence counted against recommending prevention. The authors concluded, using an ecological model of health behaviour, that external factors shape the uptake of prevention (Nussbaumer-Streit et al., 2018). Their diagnoses were not formally re-evaluated, the few psychiatrists all came from one institution, accounts of past winters are open to recall bias, and some interviews were held by phone and others face to face. The treatments themselves are assessed under Preventing Seasonal Episodes.
The first-person picture therefore rests on two small interview studies in Europe, one of them confined to people who had just completed light-room treatment. People interviewed in the Stockholm study said that in winter they were less able to work, pulled back from social contact, and found their relationships with family and friends under strain (Rastad et al., 2017). Outside that study, these consequences of the winter state have received little qualitative attention.
Types of Seasonal Affective Disorder
A type of seasonal affective disorder, in this section, is a recognised presentation or a classification device, not necessarily a separate disease. The two manuals treat seasonality as a feature of a mood disorder rather than as a diagnosis in its own right: DSM-5-TR carries a seasonal-pattern specifier that can be added to two diagnoses, recurrent major depressive disorder and bipolar disorder (American Psychiatric Association, 2022), and ICD-11 records seasonality as a qualifier on a mood disorder diagnosis (World Health Organization, 2024). The ICD-11 qualifier therefore appears in the status column of Table 1 rather than as a row of its own. Within the specifier, the season in which episodes begin separates a winter pattern from a summer pattern.
Around these classified forms sit three presentations described in research rather than codified in the manuals: a milder, subsyndromal form at one end of a continuum of seasonality in the general population, a seasonal worsening of depression that is not itself seasonal, and onset in childhood or adolescence. The evidence on cognition is uneven across them. Cognitive findings exist for the winter pattern and for subsyndromal seasonality; for the summer pattern, the bipolar form, seasonal exacerbation and young people, none of the studies described here reports cognitive results specific to the type.
Table 1
Types of Seasonal Affective Disorder
| Type | Status | Typical season of onset |
|---|---|---|
| Seasonal pattern in major depressive disorder | DSM-5-TR specifier; ICD-11 seasonal qualifier | Usually autumn or winter |
| Winter pattern | Onset pattern; the common form | Autumn or winter |
| Summer pattern | Onset pattern; uncommon | Spring or summer |
| Seasonal pattern in bipolar disorder | DSM-5-TR specifier; ICD-11 seasonal qualifier | Depression and mania can each recur seasonally |
| Subsyndromal seasonal affective disorder | Research construct; milder end of a continuum | Winter |
| Seasonal exacerbation of nonseasonal depression | Boundary presentation | Variable |
| Onset in children and adolescents | Age-group presentation | Winter |
Note. The specifier and the ICD-11 qualifier are classification devices attached to a diagnosis; the other rows are presentations described in the research literature. Season of onset follows the cited studies; it is not a criterion.
Seasonal Pattern in Recurrent Major Depressive Disorder
The specifier is attached to a diagnosis of recurrent major depressive episodes, not given on its own (American Psychiatric Association, 2022). Measured in the community rather than in clinics, depression with a seasonal pattern turned out to be uncommon. A national survey of 8098 adults across 48 US states, interviewed with the Composite International Diagnostic Interview and classified by DSM-III-R criteria, found that 0.4% had ever had seasonal-pattern major depression, and 1.0% seasonal-pattern depression of either major or minor severity, well below the 2–10% that earlier SAD estimates had suggested (Blazer et al., 1998). Among respondents with major depression, a seasonal pattern was more common in men than in women and in older than in younger adults; the authors noted that the male excess ran against earlier studies (Blazer et al., 1998). Estimates across studies and latitudes are compared under Prevalence, Course and the Latitude Debate.
Cognitive Implications
People with SAD have shown deficits in working memory, processing speed and motor speed that were present in summer as well as in winter, a finding reported in full under Attention, Memory and Executive Function (Hjordt et al., 2017). The cognitive studies described here recruited people with winter-pattern depression, and none reports unipolar and bipolar cases separately, so their results cannot be assigned to the major depressive form alone.
Winter Pattern (Fall and Winter Onset)
The winter pattern is the common form: episodes begin in autumn or winter and remit in spring or summer (Magnusson & Boivin, 2003). The MeSH entry term winter depression refers to this pattern. Much of the theory of the disorder was built on such patients; the dual vulnerability hypothesis, for instance, was proposed from a study of 53 people with winter depression and is described under The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder (Young et al., 1991). In a large Australian twin study, winter-type seasonality outweighed the summer type, a result taken up under Biological Accounts (Madden et al., 1996).
Cognitive Implications
Two lines of cognitive work concern this pattern. Women with a history of SAD reported more negative automatic thoughts than matched controls even outside the winter months, a result set out under The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder (Rohan et al., 2003). Winter performance on an emotional Stroop task, and its relation to how mood changed by summer, is covered under Attention, Memory and Executive Function (Spinks & Dalgleish, 2001).
Summer Pattern (Spring and Summer Onset)
The reverse pattern was described in a 1987 case series of 12 patients whose depressions recurred regularly in summer, some of whom became hypomanic in winter; the authors suggested that heat might play a part in some cases, a hedge they kept to a subset of patients (Wehr et al., 1987). Thirty people whose depressions recurred each summer were later set against thirty sex-matched people whose depressions recurred each winter, with symptoms tracked prospectively while each episode was under way (Wehr et al., 1991). The two groups mirrored each other in vegetative symptoms. Summer episodes tended toward insomnia and loss of appetite, while winter episodes leaned toward the atypical vegetative pattern (Wehr et al., 1991). A cluster analysis of symptom profiles was used to separate the groups, and the authors described the earlier evidence for the distinction as preliminary (Wehr et al., 1991). Pooled across cross-sectional surveys, summer-type SAD was far rarer than the winter form; the figures are given under Prevalence, Course and the Latitude Debate (Kim et al., 2025).
Cognitive Implications
No study described here reports cognitive findings specific to the summer pattern.
Seasonal Pattern in Bipolar I and Bipolar II Disorder
DSM-5-TR allows the seasonal-pattern specifier for bipolar disorder as well as for recurrent depression (American Psychiatric Association, 2022). A retrospective case series from one research-oriented psychiatric clinic showed how mixed the seasonal group can be (Faedda et al., 1993). Of 146 patients whose mood disorder followed a seasonal course, about 10% of all the clinic's mood disorder cases, 51% had recurrent depression and 49% had bipolar disorder, the latter split between cases with mania (30%) and with hypomania (19%). Two recurrence types appeared at equal frequency: one with depression in fall and winter, the other with depression in spring and summer. Women made up 71% of the series; illness began at a mean age of 29 years, lasted a mean of 12 years with about eight cycles, and episodes averaged 5.0 months. The authors concluded that the DSM-III-R seasonal category was heterogeneous and held at least these two subtypes, though the criteria were applied after the fact and the patients came from a single clinic (Faedda et al., 1993).
Later reviews confirmed the seasonality of bipolar illness. A systematic review of 51 papers found that seasonal peaks in bipolar mood episodes had been replicated, and that people with bipolar disorder fluctuated with the seasons more than people with unipolar depression or healthy controls (Geoffroy et al., 2014). Depressive seasonality, at about 25%, outnumbered manic seasonality, at about 15%, and went with bipolar II, eating-disorder comorbidity, rapid cycling and relapsing more often. The reviewed studies differed widely in sample size, sex balance and methodological quality (Geoffroy et al., 2014).
The long-term course comes from a cohort of bipolar I and II patients diagnosed by DSM-IV criteria with structured interviews and followed for at least 10 years (Goikolea et al., 2007). Seventy-seven of 302 patients with a classifiable course (25.5%) had a seasonal pattern. In logistic regression, the pattern was associated with bipolar II disorder, a depressive first episode and a predominantly depressive polarity, while seasonal and nonseasonal patients did not differ in demographics, number of hospitalisations or functioning (Goikolea et al., 2007). Its suicidality result appears under Suicide and Self-Harm. Light therapy for bipolar depression, and the question of switching into mania, are discussed under Side Effects and Safety of Light Therapy.
Cognitive Implications
No study described here reports cognitive findings specific to seasonal bipolar disorder.
Subsyndromal Seasonal Affective Disorder and the Seasonality Continuum
Seasonal change in mood and behaviour is a matter of degree in the general population. Lay interviewers phoned a random-digit-dial sample of Montgomery County, Maryland households with a telephone version of the Seasonal Pattern Assessment Questionnaire, and the answers placed seasonality on a continuum whose far end is SAD (Kasper et al., 1989a). Respondents named winter as their worst season far more often than summer, by 4.5 to 1, and younger women had the highest seasonality scores. Depending on the case-finding definition, 4.3% to 10% of respondents reported seasonal impairment as severe as that seen in patients with SAD (Kasper et al., 1989a). The milder form below the SAD threshold, sometimes called the winter blues, is labelled subsyndromal SAD.
Subsyndromal SAD responds to light. In a phototherapy trial with 20 people with subsyndromal SAD and 20 controls, bright light improved mood and behaviour in the subsyndromal group and produced no change in the controls, and five hours of light split between morning and evening outperformed two hours (Kasper et al., 1989b). The authors offered the result only as a possible guide to winter lighting. A randomised field trial in 48 adults with subsyndromal SAD compared 20-minute sessions of 100-lux narrow-band blue light with 10,000-lux bright white light on five consecutive days (Meesters et al., 2016). Symptoms on the SIGH-SAD fell by 50.7% with blue light and 54.8% with white light, and the groups did not differ on that measure or in side effects; white light reduced sleepiness and deactivation more. Because the trial had no placebo condition and ran in the field, a placebo response cannot be excluded (Meesters et al., 2016). In pooled cross-sectional data, subsyndromal SAD was more prevalent than full SAD, as reported under Prevalence, Course and the Latitude Debate (Kim et al., 2025).
Cognitive Implications
Cognitive vulnerability appears to differ across the continuum. In a sample of 88 students divided into low, mild and moderate seasonality groups by the Seasonal Pattern Assessment Questionnaire and assessed in winter and outside it, those with moderate seasonality reported more automatic negative thoughts and more rumination in winter than the other two groups (Rohan et al., 2011). The moderate group also held more dysfunctional attitudes in both seasons, and both the moderate and mild groups endorsed more maladaptive attitudes about the seasons than the low group. Every group scored higher on the cognitive measures in winter, and implicit light-dark associations also shifted in winter. The authors read the pattern as a possible cognitive mechanism confined to moderate seasonality, in a student rather than a clinical sample (Rohan et al., 2011); the model it bears on is set out under The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder.
Seasonal Exacerbation of Nonseasonal Depression
Seasonal worsening of depression that does not meet the specifier is a boundary presentation, not a DSM-5-TR category. An Austrian registry of 231,824 hospitalisations for depression over 14 years, coded by ICD-10, found that admissions for moderate and severe episodes departed from their expected monthly levels, falling below them in the summer months and in December (Fellinger et al., 2022). The authors suggested that seasonality in depression may reach beyond SAD, while noting that anonymised admissions are not the same thing as the occurrence of depression (Fellinger et al., 2022); the registry's sex and age findings are covered under Culture, Society and Who Is Affected.
A Dutch cohort study gives a more restrained picture. Among 5549 primary care patients, distress scores in winter did not differ from those in other seasons, and in a diagnostic sample of 1090 people, winter rises in depressive symptoms were small in healthy controls and in some patient groups and absent in people with major depressive disorder (Winthorst et al., 2011). Atypical and melancholic features were the exception, rising in winter, while anxiety symptoms did not vary by season. The authors concluded that seasonal differences in symptoms were absent or small (Winthorst et al., 2011).
Cognitive Implications
No study described here reports cognitive findings specific to the seasonal worsening of nonseasonal depression.
Seasonal Affective Disorder in Children and Adolescents
In a pilot questionnaire survey of 9- to 19-year-old students in a Washington suburb, 1,835 analysed responses to a modified Seasonal Pattern Assessment Questionnaire yielded 60 probable cases of SAD (3.3%) (Swedo et al., 1995). The rate rose with age and was highest among postpubertal girls. The authors presented these as pilot data on probable rather than confirmed cases, and called for research on the role of puberty (Swedo et al., 1995). In a later population survey, seasonal variation in depressive symptoms was strongest in youth, a result described under Prevalence, Course and the Latitude Debate (Lukmanji et al., 2019).
Cognitive Implications
No study described here reports cognitive findings in young people with SAD.
Prevalence, Course and the Latitude Debate
How Common It Is
Prevalence estimates for seasonal affective disorder depend heavily on how a case is defined and counted. The most recent pooled figure comes from a meta-analysis of 24 cross-sectional studies covering 32,866 participants, screened with the Seasonal Pattern Assessment Questionnaire: it put winter-type SAD at 5.01%, the milder subsyndromal SAD at 9.37% and summer-type SAD at only 0.57% (Kim et al., 2025). Because every pooled study was cross-sectional, these figures describe questionnaire-defined caseness at a single point, not the incidence of the disorder over time. An earlier overview of epidemiological work had already shown how unstable such numbers are: across 20 retrospective studies, estimates ranged from 0% to 9.7%, although the same review concluded that SAD is fairly common and that depressive symptoms in the population tend to peak in winter, and that all prospective population studies except one found seasonal variations in mood (Magnusson, 2000). Part of that spread is definitional, since in an early telephone survey the criteria chosen for case-finding changed the prevalence obtained (Kasper et al., 1989a); the questionnaire itself is discussed under Symptoms, Diagnosis and Measurement. Reviews put women at higher risk than men (Magnusson & Boivin, 2003), a difference taken up under Culture, Society and Who Is Affected.
Outside clinical samples, the winter signal is weaker than the clinical picture suggests. A systematic review of 41 studies, spanning population surveys, postpartum samples, prescription records and admission registries, found an excess of depression in winter in only 13 of them; heterogeneity was too great for a meta-analysis, season definitions differed from study to study, and questionnaire studies of seasonality were excluded, so the authors judged the evidence fragmented and only partly supportive of the clinical concept (Øverland et al., 2019). Canadian Community Health Survey data, scored on the Patient Health Questionnaire-9, did show higher symptom levels in winter than in summer, with the seasonal effect stronger among people aged 12 to 24 than among adults aged 25 and over; the effect appeared at the lower cut-point of 5 or more but was not significant at the threshold of 10 or more, and the cross-sectional design and the lack of data on antidepressant use limit how far it generalises (Lukmanji et al., 2019). In a Norwegian population survey, higher seasonality went with more anxiety and depression, an association taken up under Comorbidity (Øyane et al., 2008). A crowdsourced online survey of 5,282 Dutch adults found positive affect highest in spring (d = 0.25 against summer) and depressive symptoms lowest in spring, but the effects were small or very small, appeared only in participants higher on neuroticism, and no within-person seasonal change appeared in the 503 participants measured twice, leading the authors to conclude that the seasons do not greatly affect mood (Winthorst et al., 2020).
The Latitude Debate
A long-standing prediction holds that if shortened winter days drive the disorder, its prevalence should rise with distance from the equator. Early support came from a mailed survey at four sites in the United States, which found winter SAD and subsyndromal winter SAD more common at the northern sites, with the latitude gradient carried mainly by respondents over 35 and no gradient at all for summer SAD (Rosen et al., 1990). The 2025 meta-analysis reached a similar result across its pooled studies: latitude predicted the prevalence of SAD (β = 0.2, p < 0.001) and of subsyndromal SAD (β = 0.32, p < 0.001), but not of the summer-type form (Kim et al., 2025). Weather has been tested within patients as well. When 126 people with winter depression completed the 13-item Beck Depression Inventory repeatedly between September and May over 1991 to 1994, their scores tracked minutes of sunshine, global radiation, day length and temperature recorded by the Copenhagen meteorological service, though not cloud cover, rainfall or atmospheric pressure; the authors read this as consistent with lack of light being one contributing factor (Molin et al., 1996).
Day length across the year
Move the latitude to see how much the winter photoperiod shrinks. The grey line is the equator, for comparison.
Shortest day: 7.9 hours. Longest day: 16.1 hours. Annual swing: 8.3 hours.
The contrary evidence questions how much latitude matters. A review of prevalence studies found no significant correlation with latitude across all of them, a significant positive correlation within North America and only a trend within Europe, and SAD roughly twice as prevalent in North America as in Europe; it concluded that latitude plays a minor role and that other factors are more likely to explain the variation, while calling for controlled studies (Mersch et al., 1999). Iceland is the sharpest anomaly. A random postal survey of 1,000 Icelanders aged 17 to 67 found SAD in 3.8% and subsyndromal SAD in 7.5%, rates below those on the US east coast (rate ratios 0.49 and 0.63) despite the far northern location, with younger respondents and women more often affected; the authors raised the possibility of selection for tolerance of winter darkness (Magnússon & Stefánsson, 1993). Manitoba adults of Icelandic descent also reported low rates (SAD 1.2%, subsyndromal SAD 3.3%; rate ratios against the US east coast of 0.18 and 0.38), which led the investigators to describe the latitude relation as more complex than assumed and to suggest that genetic adaptation may play a part (Magnusson & Axelsson, 1993). At the population level, screening data from 14,478 Dutch adults showed no association of temperature, sunshine or rainfall with sad mood or major depression, and depression was in fact more prevalent in summer and autumn than in the other seasons, so the authors judged a link between weather and depression probably absent (Huibers et al., 2010). A large US survey that likewise found no effect of latitude is described under What Seasonal Affective Disorder Is (Traffanstedt et al., 2016).
Read together, the studies described here support a modest positive association between latitude and questionnaire-defined prevalence of winter and subsyndromal SAD, but latitude explains little of the variation once populations, ancestry and survey methods differ, and the question remains unresolved.
Course
Over years, winter-pattern depression does not always stay seasonal. A retrospective follow-up traced 59 patients with winter SAD over a mean of 8.8 years using semistructured interviews, life charts and collateral records: 25 (42%) kept a purely seasonal course, 26 (44%) also had nonseasonal depressive episodes, and 8 (14%) had remitted (Schwartz et al., 1996). Patients whose course included nonseasonal episodes tended to be more severely ill and to respond less well to light, and for winter recurrences many preferred light to medication; the authors presented these links as possible rather than established and suggested that light alone may not suffice for the more severe cases (Schwartz et al., 1996). In bipolar disorder, a ten-year cohort tied a seasonal pattern to bipolar II disorder and a predominantly depressive polarity, as set out under Types of Seasonal Affective Disorder (Goikolea et al., 2007). Recurrence across later winters after treatment is covered under Treatment Evidence, and the follow-up over a single summer under What Seasonal Affective Disorder Is.
Biological Accounts
No single biological mechanism explains seasonal affective disorder. Several candidate accounts have been proposed, each tied to the way the body registers the shortening of daylight, and each has gathered some evidence and met some contrary results. Lam and Levitan (2000) devoted a review to the pathophysiology of the condition (Lam & Levitan, 2000). A later overview traced seasonal variation in mood, hormones and gene expression and argued that adequate light exposure matters for well-being (Wirz-Justice, 2018), while a clinical review concluded that both the cause of the disorder and the working mechanism of light remain unknown (Meesters & Gordijn, 2016). The subsections below take the main accounts in turn and close each with its current standing.
Photoperiod and Melatonin Duration
The photoperiod account holds that the brain reads the length of the night through the duration of nocturnal melatonin secretion, and that a long winter night signal contributes to depression in susceptible people. Its experimental basis came from healthy volunteers. When eight healthy people moved from a summer-like schedule of 16 hours of light to a winter-like schedule of 10 hours of light and 14 of darkness, their nightly melatonin secretion lasted longer (12.5 versus 10.3 hours) and so did their sleep (11.0 versus 7.7 hours) (Wehr, 1991). Those photoperiods were artificial, however, and the author treated effects under natural light as a question for further study.
Patients turned out to differ from healthy people in exactly this respect. In a study that sampled plasma melatonin every 30 minutes for 24 hours under constant dim light in both seasons, 55 patients with SAD secreted melatonin for longer in winter than in summer (9.0 versus 8.4 hours), whereas 55 matched healthy volunteers showed no seasonal change (Wehr et al., 2001). The authors read the pattern as consistent with the seasonal signalling circuits that humans share with other mammals, while noting that it did not establish a causal role.
A separate line of evidence argued against the idea that light works by suppressing melatonin. In controlled experiments using skeleton photoperiods, in which brief light pulses mark the edges of a day, the antidepressant effect of light did not appear to depend on when the light was given or on whether it suppressed melatonin; the investigators judged both factors probably not critical (Wehr et al., 1986). The standing of the photoperiod account is therefore partial: patients do show a seasonal change in melatonin duration that healthy people lack, but its causal role is unproven.
The Phase-Shift Hypothesis
The phase-shift hypothesis proposes that winter depression arises when the internal circadian clock drifts later relative to the timing of sleep, so that morning light, which advances the clock, should help more than evening light, which delays it (see Sleep and Circadian Rhythms). It is the most tested of the biological accounts, and the tests point in more than one direction. The hypothesis remains contested; a review by its proponents argued that their melatonin trial establishes it, while concluding that a circadian component is necessary for winter depression but not sufficient to explain it (Lewy et al., 2007).
The early support came from comparisons of light timing. In a small trial of eight patients with winter depression, a week of morning light produced more improvement than a week of evening light, and morning light moved the evening rise of melatonin earlier; the authors suggested that the timing of light may be critical (Lewy et al., 1987). A larger programme combining crossover and parallel-group designs, with 51 patients and 49 matched controls, again found morning light (6 to 8 AM) more antidepressant than evening light (7 to 9 PM), with morning light advancing the dim-light melatonin onset (DLMO) and evening light delaying it (Lewy et al., 1998b). Pooled data from 332 patients treated at 14 research centres pointed the same way: remission rates were 53% with morning light, 38% with evening light, 32% with midday light and 11% with dim light, although the advantage of morning light disappeared among patients with moderate to severe depression, and the authors cautioned that methodological problems meant firmer conclusions would need larger samples (Terman et al., 1989).
Later work tied the response to circadian phase rather than to clock time. In 42 patients given 10,000-lux light for 30 minutes, depression ratings after morning and evening light were similar, but the size of the phase advance correlated with the response to morning light (r = 0.44); in an expanded sample of 80, early-morning light outperformed both late-morning and evening light (Terman et al., 2001). Melatonin offered a test that does not depend on light. In a randomised trial of 68 patients, mostly women, low doses of melatonin (0.1 mg or less) given at the time predicted to correct each patient's circadian misalignment reduced depression more than melatonin given at the opposite time or placebo; the authors read this as evidence for a substantial circadian component, while noting small subgroups and calling for more studies (Lewy et al., 2006). In that trial, the closer a patient moved towards a phase-angle difference of about 6 hours between melatonin onset and mid-sleep, the more symptoms improved. A review by the same group concluded that phase delay underlies winter depression in most cases, with about two thirds of patients classified as phase delayed, but that the circadian component is necessary rather than sufficient (Lewy et al., 2007).
The contrary evidence is substantial. The skeleton-photoperiod experiments described above found that light timing probably did not matter (Wehr et al., 1986). A meta-analysis of light-timing studies found that a single light pulse worked about equally well in the morning, at midday or in the evening, with only a combined morning-and-evening regime of medium-intensity light outperforming single pulses; its authors favoured a photon-count account and judged the photoperiod, melatonin and phase-shift models refuted (Lee et al., 1997). An independent test in 43 symptomatic patients treated with light for 8 weeks found no support for the predicted optimal phase relation (a phase-angle difference of 3 hours in that study): neither the expected curved relation with depression nor improvement with shifts towards that value reached significance, though trends ran in the predicted direction (Murray et al., 2006). A systematic review of treatment trials also judged the circadian evidence not robust, a point taken up in How Treatments Work (Menculini et al., 2018). The phase-shift hypothesis is therefore contested: a phase delay plausibly explains part of the condition in part of the patients, and whether it is a general cause remains unresolved.
Serotonin
The serotonin account starts from the observation that the serotonin system itself varies with the seasons. Sampling blood directly from the internal jugular vein of 101 healthy men, one study found brain serotonin turnover at its lowest in winter, with turnover rising with the duration of bright sunlight on the day of sampling (Lambert et al., 2002). A positron emission tomography (PET) study of 88 drug-naive healthy volunteers found serotonin transporter binding higher in autumn and winter than in spring and summer in every region examined, and binding fell as daily sunshine increased (Praschak-Rieder et al., 2008). The transporter clears serotonin from the synapse, so more binding implies less serotonin available, and the authors proposed a possible serotonergic basis for seasonal mood change.
People with SAD differ from healthy people in how this binding changes with the seasons, although the studies disagree about the direction of the change in healthy people. In a case-control PET study of 20 patients and 20 healthy adults scanned in summer and winter, the seasonal change in transporter binding was greater in patients in the prefrontal and anterior cingulate cortex; the difference was driven by the 11 patients with severe SAD, while those with moderate SAD did not differ from healthy participants (Tyrer et al., 2016). A longitudinal study from another group, using 80 PET scans of 17 patients and 23 controls, framed its result in terms of failed adjustment: binding in patients was higher than in controls in winter but not in summer, patients changed their binding significantly less between summer and winter than controls did, and the size of each patient's seasonal change tracked the change in symptom severity (Mc Mahon et al., 2016). The same group then scanned 23 Danes considered resilient to SAD and found transporter binding lower in winter than in summer, both in the raphe nuclei and across the brain, while a serotonin receptor measured in a smaller subsample did not change; the authors proposed that downregulating the transporter in winter helps to stabilise serotonin signalling (Mc Mahon et al., 2018). The imaging is described more fully in Brain Networks and Imaging.
The most direct test removes serotonin's precursor. In a double-blind, placebo-controlled crossover study, 12 drug-free patients who had recovered with light therapy drank an amino acid mixture lacking tryptophan; plasma tryptophan fell sharply, and depressive symptoms returned, peaking the day after depletion before subsiding (Neumeister et al., 1997). The relapse was short-lived and the depletion short-term, and the authors concluded only that serotonin is possibly involved in how light therapy works. The serotonin account is supported by depletion evidence and by imaging that ties transporter binding to season, although the imaging studies disagree about the direction of change and all of the samples are small.
Catecholamines and Dopamine
Serotonin is not the only transmitter implicated. A double-blind, randomised crossover study of sixteen patients in remission after 10,000-lux light therapy compared three conditions: tryptophan depletion, catecholamine depletion with alpha-methyl-paratyrosine, and a sham condition using diphenhydramine as an active placebo (Neumeister et al., 1998). Both depletions brought back depressive symptoms relative to sham, which the authors took as confirming a serotonin role and as suggesting that catecholamines may also be involved. The catecholamine account remains preliminary, resting on small samples and a single depletion study.
Retinal Sensitivity and Melanopsin
The retinal subsensitivity account locates the problem at the start of the light pathway: if the retina of a person with SAD responds weakly to light, the shorter days of winter may fall below what the brain needs. Electroretinogram recordings supported this in one case-control study of 22 patients (20 of them women) and 16 healthy controls: in winter, patients had lower cone response amplitudes and lower rod sensitivity, both measures normalised after 4 weeks of light therapy, and the groups did not differ in summer (Lavoie et al., 2009). Because the deficit came and went with the episode, the authors suggested it may be a marker of the depressed state rather than a lasting trait.
Attention then turned to melanopsin, the photopigment of the retinal ganglion cells that signal light to the circadian system. The post-illumination pupil response, a sustained constriction after blue light that reflects melanopsin signalling, was reduced in 15 patients compared with 15 controls tested in autumn and winter, which the authors read as a possibly less sensitive light pathway needing replication in larger samples (Roecklein et al., 2013c). A genetic study of 130 patients and 90 controls found the T/T genotype at the P10L site of the melanopsin gene 5.6 times more frequent in SAD (odds ratio 5.63, 95% CI 1.22 to 26.01) (Roecklein et al., 2009). The full genotype distribution did not differ significantly between groups (p = .09), however, the T/T genotype occurred in only 5% of patients, and the two groups were recruited differently. A review by the same group set out the melanopsin hypothesis as a research agenda, treating reduced retinal sensitivity as hypothesised and melanopsin's role as possible but unconfirmed (Roecklein et al., 2013a). The retinal account is preliminary.
Genetics
Seasonality runs partly in families. Seasonality proved heritable among 4,639 adult Australian twins who returned a mailed questionnaire: every measured domain of seasonal change showed a significant genetic influence, and a genetic factor common to all the domains explained a minimum of 29% of the variance in overall seasonality, winter-pattern seasonality outnumbered the summer pattern, and the authors read this as a mostly biological predisposition, with the caveat that the twins had volunteered through a registry (Madden et al., 1996).
Specific genes have proved harder to pin down. Comparing 189 patients with 189 matched controls, Partonen and colleagues tied SAD to variants in three circadian clock genes, and the risk genotype combination carried an odds ratio of 4.43 against all other genotypes and 10.67 against the most protective one (Partonen et al., 2007). The first genome-wide association study of the disorder, in 1,380 European-American cases and 2,937 controls, found its strongest signal near the gene ZBTB20 (odds ratio 1.63), but the signal did not reach genome-wide significance, and established candidate genes, including the clock genes and the serotonin transporter gene, showed little evidence of association (Ho et al., 2018). Its authors called ZBTB20 a candidate requiring replication. A study of 4,156 Australian twins and Amish participants found no genome-wide significant locus either, but polygenic scores for schizophrenia and, more weakly, bipolar disorder tracked questionnaire-measured seasonality, whereas the score for major depression did not (Byrne et al., 2015). That study measured seasonality without clinical diagnosis, and its authors called the schizophrenia overlap unexpected. A review of clock gene variants found associations with seasonality and mood disorders but described them as controversial and unreplicated, concluding that a genetic basis has not been established (Garbazza & Benedetti, 2018). Seasonality is heritable in part, but no gene has been replicated.
Photoperiodism and Animal Models
Many mammals change their physiology and behaviour with day length, a capacity called photoperiodism, and this comparative biology underpins the photoperiod account. A review of the animal literature concluded that no valid animal model of SAD exists when strict validity criteria are applied (Workman & Nelson, 2011). It noted, however, that some photoperiod-responsive rodents show more depressive-like behaviour under short days than long days, and that many seasonal species have smaller hippocampal volume under short days, so these animals may approximate parts of the disorder's neurobiology; the duration of nightly melatonin secretion was proposed as the signal that links day length to the depressive-like state. The animal evidence is suggestive, but no animal model has been validated for the human disorder.
Evolutionary and Energy-Conservation Accounts
Evolutionary accounts ask why a winter depressive response would exist at all. One theory paper proposed that winter symptoms such as hypersomnia and weight gain are remnants of an ancestral pattern of energy conservation linked to seasonal reproduction, and it used birth-rate data to suggest that the timing of conception might be involved (Davis & Levitan, 2005). The authors acknowledged that this evidence is indirect. A more speculative proposal traces the trait to a Neanderthal inheritance, as described in History of the Concept. Both remain untested hypotheses resting on indirect evidence.
The Vitamin D Hypothesis
The vitamin D hypothesis holds that the fall in vitamin D synthesis during sunless months causes or contributes to seasonal depression. An earlier randomised trial assigned 15 patients with seasonal affective disorder to either a single 100,000 I.U. dose of vitamin D or broad-spectrum phototherapy; one month later, depression scores had improved with vitamin D but showed no significant change with phototherapy, and the rise in 25-hydroxyvitamin D went with the improvement only marginally (p = 0.05), so the authors called for further studies (Gloth et al., 1999). A later placebo-controlled trial in healthcare workers with seasonal symptoms was negative. Taking 70 μg of vitamin D a day for 12 weeks did no better than placebo on seasonal symptoms (p = 0.7), well-being or physical measures in a trial of healthcare workers that randomised 43 and kept 34 to the end (Frandsen et al., 2014). The trial reached only about 20% of the sample it needed, its volunteers were mostly women, and dropouts had lower vitamin D levels, so the authors described the result as no demonstrated effect rather than evidence of none. The hypothesis therefore remains contested for seasonal affective disorder: one small comparison favoured vitamin D over phototherapy (Gloth et al., 1999), whereas the later placebo-controlled trial found vitamin D no better than placebo (Frandsen et al., 2014).
Broader evidence concerns depression in general, not SAD. A meta-analysis of 41 placebo-controlled trials in 53,235 adults found that vitamin D supplementation reduced depressive symptoms (Hedges' g = -0.317), apparently at doses of 2,000 IU a day or more (Mikola et al., 2023). The certainty of that evidence was rated very low, heterogeneity was high, and most trials raised risk-of-bias concerns. Those trials did not select people with seasonal depression, so the meta-analysis does not show that vitamin D prevents or treats SAD. The vitamin D hypothesis is contested, both for depression in general and for SAD specifically.
Sleep and Circadian Rhythms
Disturbed sleep is among the most frequent complaints in seasonal affective disorder, but what patients describe and what instruments record do not always agree. This section concerns sleep as people report it and as it is measured; the proposal that winter depression reflects a shift in circadian phase is set out in Biological Accounts.
Winter Hypersomnia in Early Surveys and Recordings
Oversleeping, rather than sleeplessness, emerged as the typical winter pattern in an early combined study that used a symptom questionnaire, daily sleep logs and overnight EEG (Anderson et al., 1994). Of 293 patients with SAD, 80% described hypersomnia against 10% who described insomnia, and among ten patients who kept daily logs, weekend sleep at its winter peak exceeded spring and summer levels by an average of 2.7 hr per day. Two features of that study limit how far the figure can be read as specific to the disorder. A survey of 1571 individuals at four latitudes found that sleep also lengthened in winter in the general population, and the link between hypersomnia and the other symptoms of SAD was weak (r = 0.29). The authors concluded that oversleeping is common well beyond the clinical group (Anderson et al., 1994).
The EEG recordings pointed to changes in sleep structure as well as length. In winter, sleep efficiency and the percentage of delta sleep were lower than in the same patients' summer recordings and lower than in controls, while REM density was higher; REM latency, by contrast, was normal. After 9 days of light therapy these abnormalities lessened, which the authors read as evidence that the winter architecture is reversible (Anderson et al., 1994). Most of the log-keeping patients were treated with light for some weeks of the season, so the winter increases they recorded may understate the untreated pattern (Anderson et al., 1994).
Beliefs About Sleep
Later work added the cognitive side of sleep complaints. A cross-sectional winter assessment of 147 participants compared people with SAD against people with primary insomnia and people with insomnia related to nonseasonal depression (Roecklein et al., 2013b). The SAD group reported more hypersomnia than every other group on the SIGH-SAD, and on the Pittsburgh Sleep Quality Index they more often reported hypersomnia and insomnia together. Scores on the 16-item Dysfunctional Beliefs and Attitudes about Sleep scale (DBAS-16) were the strongest predictor of early insomnia. All of these measures were self-reported. The authors suggested, without testing it, that interventions targeting sleep and circadian timing could improve treatment response in SAD (Roecklein et al., 2013b).
Self-Report Versus Objective Measurement
A narrative review drew the self-report and laboratory literatures together and found that they diverge (Wescott et al., 2020). Sleep timing was delayed in SAD, and hypersomnolence appeared in questionnaires but not in actigraphy; in winter, actigraphy even recorded less sleep in people with SAD than in controls. Only three polysomnographic studies existed, with a combined sample of 96, and they showed no difference from controls in sleep continuity, depth or REM sleep, whereas sleep in major depressive disorder was more disrupted than in SAD. Among people with delayed sleep phase syndrome, the odds of meeting SAD criteria were 3.3 times higher. The review judged the sleep phenotypes of SAD unclear, noted small samples and a neglect of summer remission, and called for a study spanning several seasons and several methods (Wescott et al., 2020).
A multi-season observational study of 64 people with SAD and 80 never-depressed controls then measured the same people with actigraphy, diaries, questionnaires and interviews (Wescott et al., 2023). By interview, people with SAD slept 72 minutes longer in winter than in summer; by actigraphy, the winter gain was 23 minutes. Controls showed no seasonal change, and diary and questionnaire measures did not separate the two groups. Reports of winter hypersomnia went with greater fatigue, more total sleep time, more time in bed, napping and a later sleep midpoint. Because the interview relied on retrospective recall, the authors flagged recall bias and recommended a multimodal assessment before any sleep intervention is chosen (Wescott et al., 2023).
Sleep Timing and Eveningness
Seasonal shifts in timing preference have also been examined in people without a diagnosis. In a prospective community survey that followed 244 randomly sampled adults in Melbourne for 3 years, scores on the Horne-Ostberg Morningness-Eveningness questionnaire moved toward eveningness in winter, a shift in chronotype that was tied to lower winter mood (Murray et al., 2003). The association was small (r = 0.17, P < 0.01). The authors treated it as provisional support for the phase-shift account, cautioning that the questionnaire gives only a self-report estimate of circadian phase and that the result awaits replication with a direct circadian measure (Murray et al., 2003).
Taken together, the evidence shows winter hypersomnia to be reported more readily than it is measured. The self-reported excess is real as a complaint and is linked to fatigue and later sleep timing, but objective winter gains are modest and laboratory sleep in SAD is close to normal (Wescott et al., 2020; Wescott et al., 2023).
The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder
The main psychological account of seasonal affective disorder holds that reduced light is only half the story. On this view, winter darkness sets off a physical, seasonal change, and whether that change grows into a full depressive episode depends on how the person thinks about it and responds to it. The account developed in two steps: a dual vulnerability model, which separated a seasonal component from a depressive one, and a cognitive-behavioural model, which named the thoughts and behaviours that are said to link them. Each model remains supported only in part, the first by two daily-diary studies (Young et al., 2008) and the second by a secondary analysis of one treatment trial in which change in seasonal beliefs mediated the longer-term advantage of CBT-SAD (Rohan et al., 2023).
The Dual Vulnerability Model
The dual vulnerability model proposes two separable liabilities. The first is a seasonal, or vegetative, vulnerability that is triggered by light deprivation and shows itself in bodily symptoms. The second is a depressive, or cognitive-affective, vulnerability that governs whether cognitive-affective symptoms follow. The hypothesis grew out of a study of symptom timing in 53 patients with winter depression (Young et al., 1991). Of the 15 symptoms tracked, the onset of only 3 was tied to the onset of the episode; the onset of the remaining symptoms showed no such link. The authors read this as a core syndrome that starts the episode, with the other symptoms arriving later through a second process, and they proposed the dual vulnerability hypothesis on that basis.
A later test asked what joins the two vulnerabilities. In two within-subject studies that collected daily symptom reports, one in a subclinical sample and one in a clinical sample, rumination, meaning a habit of dwelling repetitively on one's distress, measured as a trait style in the first study and as reported ruminative responses to fatigue in the second, moderated the day-to-day link between vegetative and cognitive-affective symptoms (Young et al., 2008). When vegetative symptoms rose, cognitive-affective symptoms rose more steeply in people with a stronger ruminative style. The authors took this as support for the model. The design measured rumination rather than manipulating it and modelled daily covariation, so it shows a moderating association rather than an experimentally manipulated cause.
Seasonal Beliefs, Rumination and Behavioural Disengagement
The cognitive-behavioural model fills in the psychological half. It proposes that negative automatic thoughts, beliefs about the seasons and light, rumination, and a withdrawal from rewarding activity turn a seasonal dip in energy into a depressive episode. An early longitudinal study assessed women with a history of seasonal affective disorder and matched controls in three seasons (Rohan et al., 2003). The women with a history of the disorder reported more negative automatic thoughts than controls, and the difference was present across the year, not only in winter. In response to light-intensity stimuli, the group with a history of the disorder reported more depressive affect to low-intensity stimuli than to bright or ambiguous ones, and they responded to bright light with less depressed mood than controls did. Rumination measured in the autumn predicted how severe symptoms became the following winter. The authors concluded, with a hedge, that cognitive-behavioural factors may play a part in the disorder; the sample was women only.
Coping under stress is the behavioural side of the model. A laboratory comparison set people with seasonal depression, people with nonseasonal depression and nondepressed controls an unsolvable anagram task as a stressor (Sigmon et al., 2007). Both depressed groups reported more avoidant coping than controls and rated negative life events as having a larger impact. The groups also differed in ways that set seasonal depression apart: seasonal patients reported more coping tied to light than either other group and less acceptance-based coping than controls, while the nonseasonal group showed greater psychophysiological arousal to the stressor than both the seasonal group and controls. The authors suggested that these differences may matter for treatment.
The cognitive-behavioural model was never meant to replace biological accounts. A narrative review set the biological models of the disorder, which deal with rhythm disturbances, retinal sensitivity to light, neurotransmitters and genes, beside the psychological model of maladaptive thinking and behaviour, and found the two literatures largely unconnected (Rohan et al., 2009a). The authors argued that an integrative approach is probably the most productive way forward. They also noted that research joining the two was lacking at the time and that each micro-model had looked only at its own factors. Their call is the reason the model is described as integrative.
Standing of the Model
The integrative cognitive-behavioural model is supported in part. Its strongest current evidence is that change in seasonal beliefs during cognitive-behavioural therapy tracks and mediates symptom improvement and its durability (Rohan et al., 2020; Rohan et al., 2023; see How Treatments Work). Predictor analyses are less favourable: the cognitive vulnerabilities were non-significant predictors of treatment outcome beyond baseline severity (Camuso & Rohan, 2020; see Testing Whether Cognition Causes Seasonal Episodes). Most of this support comes from trials run by the model's own developers, and a student study bearing on the dual vulnerability model is reported under Types of Seasonal Affective Disorder (Rohan et al., 2011).
Attention, Memory and Executive Function
Research on cognition in seasonal affective disorder asks two different questions. The first is whether people with the disorder perform differently on cognitive tasks, and whether any difference comes and goes with their episodes or persists through the symptom-free summer. The second, addressed by a smaller and more recent literature, is whether cognition and brain function vary with the seasons in people who do not have the disorder at all. The studies described here are few, and most are small, so the answers to both questions remain provisional.
Cognition in Seasonal Affective Disorder
The clearest evidence of a cognitive difference comes from a longitudinal case-control study that tested the same people twice, once in summer and once in winter. Twenty-nine medication-free individuals with winter-pattern SAD were compared with 30 demographically matched healthy controls (Hjordt et al., 2017). The group with SAD performed worse on three measures: working memory, assessed with the Letter-Number Sequencing task; processing speed, assessed with the Symbol Digit Modalities Test; and motor speed, assessed with a simple reaction-time task. Crucially, these deficits were present in summer as well as winter, when depressive symptoms would be expected to have remitted. That pattern bears on the distinction between a state marker, which tracks the current episode, and a trait marker, which belongs to the person regardless of mood. The authors read the persistent slowing of processing speed as a possible trait marker of SAD, and processing speed was also negatively associated with the seasonal change in symptoms (Hjordt et al., 2017). The interpretation is tentative: with roughly 30 people per group and two testing occasions, the study shows a season-independent difference but cannot establish whether it precedes the disorder or follows from repeated episodes.
Whether light therapy alters neuropsychological function, as well as mood, in SAD was the question of an early trial (Michalon et al., 1997).
Population data point in a similar direction for memory, though with a different kind of evidence. In the Finnish Health 2011 survey, 8135 adults were invited, 5903 took part and 4554 were interviewed; SAD was associated with short-term memory problems (Morales-Muñoz et al., 2017). The same survey's principal result was a strong association between SAD and alcohol use disorder, which is taken up under Comorbidity. Because the design was a single survey, it identifies an association between SAD and memory difficulties without showing which came first or whether drinking contributes to both.
Attention to emotional material has been studied with the emotional Stroop task, a variant of the classic Stroop effect paradigm in which participants name the ink colour of words, and slower naming of threat-related words than neutral words is taken as an index of attentional bias towards threat. A longitudinal study tested individuals with SAD in winter and again in summer (Spinks & Dalgleish, 2001). Stroop performance itself did not change between the two seasons. What changed was its relation to outcome: a larger threat-word interference effect in winter was associated with more improvement by summer, and a stronger winter link between Stroop performance and mood went with poorer mood in summer. Winter attentional processing, in other words, carried information about the course of symptoms even though the bias did not itself rise and fall with the season (Spinks & Dalgleish, 2001). Whether such attentional and cognitive differences play a causal part in seasonal episodes is examined under Testing Whether Cognition Causes Seasonal Episodes.
Seasonal Variation in Cognition and Brain Function
A separate line of work asks whether the brain itself changes across the year. A systematic review of studies up to September 2024 gathered MRI, magnetic resonance spectroscopy and EEG studies of seasonal effects (Majumdar et al., 2025). It reported, among other findings, that a shorter photoperiod was associated with a smaller hippocampus, that the P3 event-related potential was larger in spring and summer than in autumn and winter, and that glutamate and GABA concentrations did not differ significantly across seasons. Its overall judgement was cautious: most of the seasonal evidence was behavioural rather than neural, the imaging and EEG studies were limited, mostly cross-sectional and observational, and they offered little mechanistic insight. The authors described the evidence as preliminary and called for adequately powered longitudinal studies (Majumdar et al., 2025).
One recent imaging study illustrates both the promise and the limits of this work. Using 7-tesla functional MRI, it scanned 29 healthy young adults (18 women; mean age 24 years) once each, in the morning, while they heard emotional and neutral vocalisations under three levels of melanopic illuminance (Campbell et al., 2025). Emotional sounds drew larger amygdala responses than neutral ones. Across the ten amygdala subparts examined, activity in the basomedial nucleus and the anterior amygdaloid area was higher for participants scanned near the start of winter than for those scanned in summer. Brighter light linearly reduced activity in the medial-superior amygdala during emotional stimuli, and worse mood on the Beck Depression Inventory-II went with greater activity in the medial and cortical nuclei and in the transition areas. Reaction times were not significantly altered by illuminance (Campbell et al., 2025). The authors proposed that the amygdala may offer a route by which light influences mood, but the study had important limits: a small sample, small effects, only four scans in winter, indoor light levels inside a scanner, no positive emotional stimuli, and an inconclusive test of the role of the light-sensitive retinal cells thought to carry non-image-forming light signals (Campbell et al., 2025).
These findings need fencing. Neither the review's seasonal brain measures nor the 7-tesla study compares people with SAD and people without it, and every participant in the 7-tesla study scored low on seasonality. They show that some brain measures may vary with season, light and mood in the general population; they do not show that the same variation is exaggerated in SAD, or that it causes seasonal episodes.
How Well the Laboratory Measures Generalise
Almost every cognitive finding on this page rests on laboratory tasks: emotional Stroop interference, exposure to controlled laboratory light, and brief laboratory stressors in the experimental studies reviewed elsewhere in the article. In the studies described here, neither the test-retest reliability of these measures in seasonal samples nor their power to predict how a person copes over a real winter has been established, so a group difference on such a task should not be read as a measure of everyday functioning. The closest approach to everyday measurement is the daily-diary work on rumination discussed under The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder, in which two within-subject studies tracked symptoms day by day rather than in a single session (Young et al., 2008).
Testing Whether Cognition Causes Seasonal Episodes
The cognitive accounts described in The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder need cognition to come before mood, not merely alongside it. Prospective designs offer the first test. Women with a history of the disorder, followed across three seasons, reported more automatic negative thoughts than matched controls throughout the year, not only in winter, and rumination measured in the fall predicted how severe winter symptoms became; the authors concluded that cognitive-behavioural factors are possibly involved (Rohan et al., 2003). Daily symptom records point the same way. Across a subclinical and a clinical sample, rumination moderated the step from vegetative symptoms to cognitive-affective ones, a pattern read as support for the dual vulnerability model (Young et al., 2008). Neither design manipulates cognition, so neither can show whether these thoughts drive an episode or remain behind after earlier ones.
Treatment trials can separate the two readings, because a therapy that changes thinking should protect against the next winter only if thinking is causal. In a secondary analysis of a three-arm randomized trial of 69 adults, dysfunctional attitudes, automatic negative thoughts and rumination fell to the same degree over 6 weeks whether participants received cognitive-behavioural therapy, light therapy or both (Evans et al., 2013). What differed was the meaning of that change: among the 52 participants with next-winter data, larger drops on the Automatic Thoughts Questionnaire and the Dysfunctional Attitude Scale predicted milder depression a year later after cognitive-behavioural therapy but not after light, and changes in rumination and distraction predicted nothing. The authors read cognitive change as a mechanism of cognitive-behavioural therapy and as a by-product of recovery under light therapy, a conclusion qualified by 25% attrition, a mostly female sample, a small follow-up subsample and some interactions that were marginal, only approaching significance (Evans et al., 2013). A larger secondary analysis of 177 adults, 89 assigned to light therapy and 88 to cognitive-behavioural therapy, asked whether vulnerabilities measured before treatment forecast who would do well. Baseline depression severity did; dysfunctional attitudes, seasonal beliefs and brooding rumination did not, neither for post-treatment outcome nor for remission or recurrence, and none of them interacted with treatment type (Camuso & Rohan, 2020). Only morningness on the Morningness-Eveningness Questionnaire carried prognostic weight, with greater morningness going with lower depression after treatment. The authors judged the cognitive vulnerabilities insufficient as predictors once baseline severity was known, while noting that the questionnaire is not a direct circadian measure, that the vulnerabilities overlapped with baseline depression, that the sample was predominantly non-Hispanic White, and that a fair test of treatment-specific prediction would need roughly 300 participants per arm (Camuso & Rohan, 2020).
The biological route to the causal question has been manipulation rather than measurement: in remitted patients, tryptophan depletion brought back depressive symptoms briefly (Neumeister et al., 1997), work described in Biological Accounts. No comparable experiment exists for cognition. None of the studies described here altered thinking in people before a first seasonal episode and watched what followed, so cognitive change is supported as a route by which cognitive-behavioural therapy works, but whether negative thinking causes seasonal episodes in the first place remains an open question.
Brain Networks and Imaging
Neuroimaging of seasonal affective disorder asks two questions: how light reaches the brain systems that regulate mood, and whether the resting brain of a person with the disorder is organised differently. Both lines of work are young, and each rests on one modest sample.
The Light-to-Mood Pathway
The clearest imaging evidence for a light-sensitive route to mood comes from a functional magnetic resonance imaging (fMRI) study run in autumn and winter. Fourteen symptomatic, untreated patients and 16 healthy controls performed an auditory emotional task while exposed to monochromatic blue or green light (Vandewalle et al., 2011). In the patients, responses in the posterior hypothalamus rose under blue light and fell under green light; in the controls, light colour had no such effect (Vandewalle et al., 2011). Electroretinography found no difference between the groups in retinal sensitivity to light (Vandewalle et al., 2011). The authors proposed the posterior hypothalamus as a substrate of the disorder, a proposal that rests on a case-control design with fewer than 20 people per group (Vandewalle et al., 2011).
Resting-State Networks
A second approach scans people at rest, with no task. In a case-control resting-state fMRI study, 45 antidepressant-free patients (30 women and 15 men, mean age 39.78 ± 10.64 years) were compared with matched healthy controls using independent component analysis with dual regression (Abou Elseoud et al., 2014). The patients differed from controls in functional connectivity within 11 of the 47 resting-state networks identified, and they showed greater amplitude of low-frequency fluctuations, a measure of the strength of spontaneous signal change, in the visual cortex and the right sensorimotor cortex (Abou Elseoud et al., 2014). The authors themselves noted that the findings need to be followed up under task conditions and that the cognitive deficits linked to these networks remain to be investigated (Abou Elseoud et al., 2014).
Two further imaging lines are reported elsewhere in this article. Positron emission tomography of the serotonin transporter across summer and winter is described under Biological Accounts (Mc Mahon et al., 2016), and work on seasonal variation in amygdala responses in healthy adults, not patients, is described under Attention, Memory and Executive Function (Campbell et al., 2025).
What the Imaging Does Not Show
The imaging studies described here do not connect the brain to the cognitive constructs that dominate psychological accounts of the disorder. None of them tested whether rumination, slowed processing or attentional bias to threat has a neural correlate in seasonal samples, and the resting-state study did not measure cognition at all. The samples are small, the findings have few or no replications, and no network finding in seasonal affective disorder can yet be called established.
Culture, Society and Who Is Affected
Seasonal depression is not spread evenly across people. Sex, age, ancestry, migration and occupation all shift how often a winter pattern is reported, and the questionnaires used to detect it were built in a narrow range of settings. The studies described here show both kinds of variation: real differences between groups, and differences that may partly reflect who was asked and how.
Sex and Age
The clearest population evidence on sex comes from the UK Biobank. Lyall and colleagues applied a cosinor method to self-reported depressive symptoms in more than 150,000 participants and found a winter peak in women but no seasonal rhythm in men (Lyall et al., 2018). Among women, longer days went with less low mood and less anhedonia, yet with more tiredness, and the day-length association could not be separated from temperature (Lyall et al., 2018). The authors read this as a sex difference in seasonal mood at the level of the population, but the design was cross-sectional, the outcomes were self-reported, and only a subset of depressive symptoms was measured, so longitudinal work is still needed (Lyall et al., 2018). Reviews of clinical seasonal affective disorder point the same way, reporting that women are affected more often than men (Rybakowski, 2024; Majumdar et al., 2025).
Hospital data add a qualification by age and by severity. In a 14-year Austrian registry of 231,824 depression admissions, a sex difference in seasonality appeared only for psychotic depression (significant in women, not in men, p = 0.291), and patients older than 55 showed a weaker seasonal pattern than younger ones; admissions, however, are not the same as the occurrence of depression (Fellinger et al., 2022). The admissions pattern itself is described under Types of Seasonal Affective Disorder.
Ethnicity and Migration
Ancestry and country of origin also matter, although the measurement problem is sharper here. A mailed survey of 1047 immigrants in Oslo, drawn from five national groups, found that winter seasonal affective disorder was more common among Iranian than among Sri Lankan respondents, with country of birth predicting both the full and the subsyndromal winter pattern (Saheer et al., 2013). Younger age, smoking and mental distress were among the other factors linked to winter cases (Saheer et al., 2013). Two caveats limit how far these group differences can be read: only 39.7% of those approached responded, and the Seasonal Pattern Assessment Questionnaire used to classify them has not been validated across these cultures (Saheer et al., 2013).
A questionnaire study of 537 African American students in Washington, D.C., found a picture close to that of the general population. Their global seasonality scores averaged 8.3, with a standard deviation of 5.3, and winter cases significantly outnumbered summer cases, so seasonality resembled that reported in general-population samples (Agumadu et al., 2004). Awareness was the striking difference: about 80% of the students had not heard of the disorder, a lower level of awareness than in the general population (Agumadu et al., 2004). The unusually low rates found in Iceland are discussed under Prevalence, Course and the Latitude Debate.
Occupation and Extreme Environments
Some working environments compress the seasonal problem into a single posting. Subsyndromal seasonal affective disorder, the polar T3 syndrome and the winter-over syndrome, three overlapping syndromes tied to the season, emerged from Palinkas and Suedfeld's review of polar expedition members who stay through the winter (Palinkas & Suedfeld, 2008). The review set these alongside positive, salutogenic outcomes of polar life and recommended screening of personnel, support during the stay and training before it (Palinkas & Suedfeld, 2008). This literature concerns selected groups in an extreme setting; it shows that seasonal mood change can be anticipated and managed in such groups, not how common it is elsewhere.
How Far the Evidence Reaches
Reported rates of seasonal affective disorder differ by region: a recent review gave background prevalence ranges of 1.4% to 9.7% for North America, 1.3% to 3.0% for Europe and 0% to 0.9% for Asia, which it attributed to latitude (Majumdar et al., 2025). Because the main screening instrument has not been culturally validated, a lower rate in an under-studied region cannot yet be separated from a difference in how symptoms are recognised, reported or counted. Whether the group differences described in this section reflect biology, culture, latitude or measurement remains unresolved.
Comorbidity
Only a handful of conditions have been examined for co-occurrence with seasonal affective disorder in samples defined by seasonality, and each study described here used a different design. None of them can say which condition comes first or whether one contributes to the other, so for every pairing below the direction of effect remains unknown.
Attention-Deficit/Hyperactivity Disorder
Seasonal depression appears common among adults with ADHD. In a clinical sample of adults with lifetime ADHD and mood disorder, 27% met screening criteria for seasonal affective disorder when the Seasonal Pattern Assessment Questionnaire was administered by telephone, and women were at higher risk than men (Amons et al., 2006). The authors concluded that the two conditions frequently co-occur. Two limits narrow the reading: the questionnaire is a screening tool rather than a diagnostic interview, and the participants were drawn from a clinic, so the figure need not hold for adults with ADHD in the community (Amons et al., 2006).
Bulimia Nervosa
Evidence for an eating-disorder overlap comes from treatment rather than prevalence. Twenty-two women with both seasonal affective disorder and bulimia nervosa received 10,000-lux light for 30 to 60 minutes a day in a 4-week open trial (Lam et al., 2001). Scores on the 29-item Hamilton Rating Scale for Depression fell by 56%, and diary records showed binges down by 46% and purges by 36%, each change significant at p < .001. Mood responded more completely than eating behaviour: 10 of the 22 women (45%) reached depressive remission, yet only 2 (9%) stopped bingeing and purging altogether. The authors judged that light therapy is likely to serve as an adjunct rather than a stand-alone treatment for the eating symptoms, and the trial had no control group and lasted only four weeks (Lam et al., 2001).
Anxiety and Depressive Symptoms
At the population level, seasonality tracks broader distress. A Norwegian cross-sectional survey analysed 2980 men and 8074 women who completed the Seasonal Pattern Assessment Questionnaire and the Hospital Anxiety and Depression Scale; higher seasonality went with more anxiety and more depressive symptoms, and the season in which people were surveyed did not significantly change that link (Øyane et al., 2008). The authors proposed, tentatively, that seasonality may be a separate dimensional trait rather than a feature confined to a clinical category, and many invited adults did not respond. A cross-sectional association of this kind cannot show whether seasonality raises anxiety, anxiety heightens awareness of seasonal change, or both share a cause. In a Dutch cohort, by contrast, anxiety symptoms themselves did not differ significantly across the seasons (Winthorst et al., 2011); that study is described under Types of Seasonal Affective Disorder.
Alcohol Use Disorder
The clearest substance-use signal comes from Finland. In the population-based Health 2011 survey, 4554 adults completed a diagnostic interview (the Munich Composite International Diagnostic Interview) alongside the Seasonal Pattern Assessment Questionnaire, and seasonal affective disorder showed a strong association with alcohol use disorder (Morales-Muñoz et al., 2017). Because the data were collected at one time point, they cannot show whether seasonal episodes encourage drinking, heavy drinking disturbs seasonal mood regulation, or a third factor underlies both.
Gaps in the Evidence
Obesity, trauma-related disorders and substance use other than alcohol have not been examined in seasonal samples in the studies described here, so nothing can yet be said about their overlap with seasonal affective disorder.
Suicide and Self-Harm
Suicide Risk in Seasonal Affective Disorder
Direct evidence on suicide in seasonal affective disorder is sparse. One cross-sectional study set 87 outpatients with SAD against 65 people with nonseasonal major depression who had attempted suicide, rating both groups on the Comprehensive Psychopathological Rating Scale under DSM-III-R criteria (Pendse et al., 2004). Total scores did not differ significantly between the groups, nor did the depression subscale, and the SAD group scored higher than the attempters on 18 individual items, while lassitude, increased sleep, hostile feelings and failing memory were among the items independently associated with SAD (Pendse et al., 2004). The authors read this as evidence of abundant, self-reported symptomatology in SAD, and the comparison set outpatients against hospitalised attempters (Pendse et al., 2004). Similar symptom totals therefore do not show that people with SAD share the attempters' risk; the study measured psychopathology, not suicidal behaviour.
The longest follow-up evidence comes from bipolar disorder rather than from SAD as such. In a cohort of 302 patients with bipolar I or II disorder followed for at least 10 years, a seasonal pattern went with bipolar II, a depressive onset and a predominantly depressive polarity, yet patients with and without the pattern did not differ in suicidality, psychotic symptoms, hospitalisations or functioning (Goikolea et al., 2007). The authors added that long-term follow-up studies of this kind are scarce (Goikolea et al., 2007). None of the studies described here measured suicide rates in people with a diagnosis of SAD, and none tested a suicide-prevention approach in a seasonal sample; both questions remain open.
Population Seasonality of Suicide
The seasonal rhythm of suicide in whole populations is a separate topic from SAD, and the evidence below concerns deaths and attempts in general populations, not people with a seasonal mood disorder. A literature review of studies published from 1979 to 2009 found that most studies reported a spring peak in suicides, with a smaller secondary rise in autumn, and that studies from both hemispheres reported a seasonal pattern; the spring excess was clearest among men and older people, although some studies found no seasonality and results were inconsistent across them (Christodoulou et al., 2012). Attempts follow a similar course: a systematic review of 29 articles from 16 countries, reported attempts most often in spring and summer, its cosinor analysis suggesting a spring peak, and suggested that some environmental factor probably modulates the rhythm, with the caveat that it grouped very diverse populations (Coimbra et al., 2016).
A Danish register study of 37,987 suicide cases between 1970 and 2001 found a spring peak both among people previously hospitalised for a mood disorder and among those never hospitalised, with a larger spring excess in the first group (men: RR = 1.18, 95% CI 1.07-1.31; women: RR = 1.20, 95% CI 1.10-1.32); the authors judged spring worsening of mood worth further study and cautioned that Danish conditions limit generalisation (Postolache et al., 2010). A narrative review covering 1979 to 2011 concluded that the spring peak replicates but that its mechanisms are poorly understood, the evidence correlational and often inconclusive, and the pattern flattening in some countries (Woo et al., 2012). A qualitative review examined how suicide seasonality relates to the availability of particular means and to weather (Ajdacic-Gross et al., 2010).
Whether sunlight itself drives the spring peak is disputed. An Austrian registry analysis of 69,462 suicides from 1970 to 2010 removed seasonal variation by differencing and still found that sunshine in the 10 days before a death went with slightly more suicides (r = 0.0370), while sunshine 14 to 60 days earlier went with slightly fewer (r = -0.0383); the authors stressed that the effect sizes were low and proposed, tentatively, a short-term triggering and longer-term protective role (Vyssoki et al., 2014). An ecological analysis of 31,060 suicides in Greece, Victoria (Australia) and Norway found a small association with mean daily sunlight before adjustment (IRR = 1.018), which vanished once month of year entered the model (IRR = 0.995, p = 0.44); the authors concluded that sunlight was likely standing in for seasonality, while noting that sunlight was never measured at the eye and attempts were not examined (White et al., 2015). The same team also set out to reproduce the Austrian daily sunshine associations in their own daily data from Victoria and Greece; once seasonal effects were removed and the many tests corrected for, no correlation reached significance and the effect sizes were nominal (White et al., 2015). Neither design settles the question, and the role of sunlight in population suicide seasonality remains unresolved.
The population peak falls in spring and early summer, whereas the episodes described in Types of Seasonal Affective Disorder typically begin in autumn and winter, so the two patterns should not be equated: a spring rise in population suicides is not evidence about risk in winter SAD. What causes the spring peak is still an open question.
Anyone thinking about suicide can find help: in the United States, the 988 Suicide and Crisis Lifeline can be reached by call or text at 988, and the International Association for Suicide Prevention keeps a directory of crisis centres in other countries.
Treatment Evidence
Four treatments have been tested in controlled trials for seasonal affective disorder: bright light, dawn simulation, a seasonal adaptation of cognitive-behavioural therapy and antidepressant medication. Light has the largest evidence base. The other three have been judged mainly by how they compare with it.
Bright Light Therapy
Meta-analyses agree that light therapy works for seasonal affective disorder. They disagree about how much. An early systematic review pooled randomised controlled trials published between 1975 and 2003. Bright light outperformed control conditions with an effect size of 0.84 (95% CI 0.60 to 1.08, eight studies), and dawn simulation did so with an effect size of 0.73 (95% CI 0.37 to 1.08, five studies) (Golden et al., 2005). The same review admitted only 13% of the reports it screened. Its authors judged that efficacy had been shown, but they noted that many trials lacked rigorous designs and adequate group sizes and that standard parameters for active versus placebo light conditions still needed to be defined (Golden et al., 2005). A later meta-analysis of 19 randomised trials in adults found a smaller advantage over placebo. Its standardised mean difference in depression scores was -0.37 (95% CI -0.63 to -0.12), and its risk ratio for treatment response was 1.42 (95% CI 1.08 to 1.85). Heterogeneity across the pooled trials was moderate and their samples were small to medium, yet no publication bias was evident; the authors judged light therapy effective while calling for bigger trials (Pjrek et al., 2020).
The main difficulty is the placebo. Trials have used dim light or sham or low-density negative-ion generators as placebo controls (Pjrek et al., 2020). One placebo-controlled trial gave 6000 lux for 1.5 hours a day over 4 weeks, with a sham negative-ion generator as the placebo. Among the 96 patients who completed it, by one strict criterion (a SIGH-SAD score below half of baseline and at most 8), the response rate was 61% with morning light, 50% with evening light and 32% with placebo (Eastman et al., 1998). The three groups expected similar benefit, which suggests the placebo was credible. However, they did not differ in mean depression scores, and the advantage of light appeared only after three weeks. The authors read the result as a specific effect of light that needs at least that long to emerge (Eastman et al., 1998). A second trial randomised 158 patients to six groups. Light was given at 10,000 lux for 30 minutes a day, and negative air ions at high density (2.7 x 10^6 ions per cubic centimetre) or at low density (1.0 x 10^4), with the low-density ions serving as a placebo. The low-density group improved less than every other group, and no other groups differed from one another (Terman et al., 1998). Morning light produced more remissions than evening light when a stringent definition of remission was used. On that evidence the authors classed both light and high-density ions as specific antidepressant treatments (Terman et al., 1998).
Not every appraisal has been so positive. A systematic review of eight trials in seasonal affective disorder found that bright white light beat control conditions at weeks two (SMD -0.50) and three (SMD -0.31) and at the end of treatment (SMD -0.54), but not at other time points (Mårtensson et al., 2015). Its authors judged the evidence to be less than unequivocal. They pointed to considerable methodological problems, to results that depended on which studies were included, and to effects shown only over the short term (Mårtensson et al., 2015).
On current reading, light therapy is effective for seasonal affective disorder, with moderate certainty, because the trials are small and none can blind patients fully. A later narrative review still called light the treatment of first choice. It also noted that no valid placebo for light exists and that meta-analytic estimates shift with the inclusion criteria, from 0.84 in one pooling to 0.54 in another (Meesters & Gordijn, 2016). A review of two decades of work described bright light as a potent treatment that does not rely on drugs (Terman & Terman, 2005). A clinical review judged light efficacious and about as effective as antidepressants (Westrin & Lam, 2007). A third overview listed light, with vitamin D and counselling, among the treatments emerging as effective (Melrose, 2015).
Light Therapy in Practice
Current therapy lamps deliver 10,000 lux (Rybakowski, 2024). A common regimen gives that intensity for 30 to 45 minutes per session; in the early days, weaker light given for longer, 2,500 lux for 3 hours, was common (Meesters & Gordijn, 2016). Another review describes sessions of 20 to 60 minutes seated before a light box, taken in the morning from early autumn until spring (Melrose, 2015). The boxes it describes screen out ultraviolet light, and in Scandinavia light rooms with indirect, evenly distributed light are also available (Melrose, 2015). Dose and time of day both need to be adjusted for the individual patient. Adding light to an antidepressant can speed improvement and leave fewer residual symptoms, and the side-effect profile of light compares favourably with that of medication (Terman & Terman, 2005). Morning sessions are usually preferred, in line with the morning advantage in the trials above. The circadian reasoning behind morning timing is covered under Biological Accounts and Sleep and Circadian Rhythms, and the side effects under Side Effects and Safety of Light Therapy.
Prescribed use and actual use differ. In a pilot trial of 19 patients, elapsed-time meters were fitted to the devices, either bright light or dim red light, and the prescribed session rose from 30 minutes in week 1 to an hour by weeks 3 and 4 (Michalak et al., 2007). On average, patients used their devices for 59.3% of the prescribed time, and 31.6% dropped out. Patients' daily diaries reported more use than the meters recorded and bore no relation to the metered figures. Dropout tended to be lower and adherence higher with bright light than with dim red light, but only as trends. The authors judged adherence to be similar to adherence with antidepressants, a conclusion limited by the pilot design (Michalak et al., 2007). The finding matters because self-reported use is the usual record of how much light a patient received.
The current depression guideline gives specifiers little value overall in choosing a treatment. Two specifiers, with psychotic features and with seasonal pattern, are the exceptions that do inform the choice (Lam et al., 2024). For seasonal (winter) depression, the guideline names light therapy as a first-line treatment, with a typical protocol of 10,000 lux for 30 minutes a day (Lam et al., 2024). The specifier itself is described under What Seasonal Affective Disorder Is.
Dawn Simulation
Dawn simulation brightens a bedroom gradually before waking, in place of a session at a light box. The strongest single result in its favour came from a six-week placebo-controlled trial of 95 medication-free patients. Dawn simulation exceeded placebo on remission (p < .05) and response (p < .001), and it also exceeded bright light on remission (p < .01) and response (p < .001) (Avery et al., 2001). In the same trial, bright light did not differ from placebo. More hours of sunshine during treatment predicted both remission and response (Avery et al., 2001).
A five-group trial of 99 adults did not support the claim that dawn simulation is superior. Dawn simulation and bright light did not differ, and neither did dawn simulation and a brief dawn pulse. The dawn pulse, however, led to a pattern of residual or worsened symptoms like that of low-density ion nonresponders (Terman & Terman, 2006). Every group, including high-density ions, outperformed low-density ions, and the authors proposed dawn simulation and high-density ions as possible alternatives to bright light and medication (Terman & Terman, 2006). A crossover study gave 40 patients with winter depression a week of each treatment. Self-rated depression fell by similar median amounts, 43.8% and 42.2%, and patients split almost evenly in their preference (21 to 19) (Danilenko & Ivanova, 2015). Those who preferred bright light rated it as more effective (p = 0.0096), and more severe depression tended to favour bright light, although only as a trend. Ratings were self-reported, and the dawn device did not imitate natural overhead light (Danilenko & Ivanova, 2015). One review reports studies in which bright light outperformed dawn simulation and one study, in patients who were all hypersomnic, in which dawn simulation did better (Meesters & Gordijn, 2016). The record for dawn simulation is mixed and rests on few trials.
Cognitive-Behavioural Therapy Versus Light Therapy
A version of cognitive behavioral therapy tailored to seasonal depression, known as CBT-SAD, has been compared with light therapy over a series of trials. A pilot trial randomised 61 adults with recurrent seasonal depression to six weeks of therapy, light at 10,000 lux, the two combined, or a minimal-contact control that offered light later (Rohan et al., 2007). Each active treatment reduced depression more than the control, and the three did not differ from one another. Remission reached 73% with the combination against 20% in the control group. The authors judged that the therapy, alone or combined with light, holds promise and warrants further study (Rohan et al., 2007).
The trials' interest lies in what happened the following winter. Pooled intention-to-treat data from two pilot trials (69 patients, with 17 dropouts handled by multiple imputation) showed recurrence of 7.0% after therapy and 5.5% after the combination, against 36.7% after light alone (Rohan et al., 2009b). Symptoms were also less severe after therapy than after light, a difference that survived adjustment. The combination, however, did not differ from light on severity. The authors framed the long-term advantage conditionally: if replicated, the therapy might be the better long-term option (Rohan et al., 2009b).
A larger head-to-head trial followed. It assigned 177 adults to 12 sessions of CBT-SAD (n = 88) or to light therapy (n = 89) for six weeks (Rohan et al., 2015). Acute outcomes were equivalent. Remission on the SIGH-SAD interview was 47.6% and 47.2%, remission on the Beck Depression Inventory-Second Edition was 56.0% and 63.6%, and neither difference was significant. Patients with comorbid disorders at baseline scored higher on depression throughout. This acute report was a secondary analysis, since the trial's primary outcome was recurrence (Rohan et al., 2015). The two arms still did not differ in the first winter after treatment. By the second winter, recurrence was 27.3% after CBT-SAD against 45.6% after light, and remission was 68.3% against 44.5% (Rohan et al., 2016).
The trials therefore show equal acute effects and a durability advantage for CBT-SAD. Those authors made the practical advantage of the therapy conditional on the findings being replicated (Rohan et al., 2009b). The cognitive model behind the therapy is set out under The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder, and whether its targets explain the benefit is examined under How Treatments Work.
Recurrence in a later winter
Each grid is 100 people; filled circles are those whose depression recurred.
CBT-SAD: 27.3% recurrence. Light therapy: 45.6% recurrence.
Antidepressant Medication
Medication has also been compared directly with light. In the Can-SAD trial, a double-blind study at four Canadian centres, 96 patients received either light or fluoxetine for 8 weeks (Lam et al., 2006). Response was 67% in each group, and remission was 50% and 54%, with no significant difference. Light produced more improvement in the first week, though that comparison was made post hoc, and fluoxetine caused more agitation, sleep disturbance and palpitations (Lam et al., 2006). The trial had no condition in which both treatments were placebos. With light therapy and fluoxetine showing no overall difference, the authors held that clinical considerations, with patient preference among them, should settle which one is used (Lam et al., 2006).
A Cochrane review of second-generation antidepressants for treating the disorder found three randomised trials with 204 adults in total (Nussbaumer-Streit et al., 2021). Against placebo, fluoxetine gave a risk ratio of 1.62 whose confidence interval (0.92 to 2.83) included no effect, and the certainty of that evidence was very low. Against light, the risk ratios were 0.98 for response and 0.81 for remission. Adverse events were similar across comparisons, though the data on them were sparse. No trial compared two antidepressants or an antidepressant with psychotherapy, and the reviewers judged the evidence insufficient for overall conclusions (Nussbaumer-Streit et al., 2021).
Two drugs that act on the melatonin system have weaker support. In an open 14-week study, 37 acutely depressed patients took agomelatine at 25 mg a day. Scores on the SIGH-SAD and on clinician global ratings fell from week 2 (p < 0.001), and response and remission rates reached 75.7% and 70.3% (Pjrek et al., 2007). Because the study had no control group, the authors' conclusion that the drug may treat seasonal depression effectively and safely remains provisional. A pilot placebo-controlled trial of melatonin in 10 patients with winter depression found that ratings fell more with melatonin than with placebo (Lewy et al., 1998a). It did not document the shift in circadian timing that the treatment was meant to produce, and its authors called for a larger replication that would measure it (Lewy et al., 1998a).
An earlier narrative review placed drug treatment second to light. Among drugs, it preferred the newer selective serotonergic and noradrenergic agents to older antidepressants, while noting small trial samples and the absence of any trial in bipolar seasonal depression (Pjrek et al., 2005).
The mechanisms proposed for these drugs follow the biological accounts. Serotonergic and catecholaminergic actions fit the depletion findings reviewed under Biological Accounts, and agomelatine and melatonin act on the melatonergic and circadian system described there and under Sleep and Circadian Rhythms. None of the trials above tested a mechanism directly; the melatonin pilot, for one, did not measure the phase shift it was designed around. One large placebo-controlled study, an eight-week multicentre trial with 187 outpatients, found sertraline better than placebo for seasonal affective disorder and well tolerated (Moscovitch et al., 2004; Meesters & Gordijn, 2016). The Cochrane review left this trial out because its sample included patients with bipolar disorder, about a tenth of the total (Nussbaumer-Streit et al., 2021). Medication is a reasonable option; one large placebo-controlled trial favoured sertraline, but beyond it the trial evidence remains limited.
Light therapy for nonseasonal depression is covered under Related Conditions and Seasonal Mood in the General Population.
Preventing Seasonal Episodes
Seasonal affective disorder is one of the few depressive conditions whose next episode can be anticipated by the calendar, and that predictability invites a different clinical question: whether treatment begun before symptoms appear can stop an autumn-winter episode from starting at all. The answer differs sharply by intervention. Only one drug has been tested against placebo for this purpose, and the evidence for light and for psychological prevention rests on one small trial each.
Bupropion XL
The pharmacological case comes from three randomized, placebo-controlled prevention trials of extended-release bupropion (bupropion XL), which together enrolled 1042 patients with a history of seasonal depression in the autumn and gave 150-300 mg daily (Modell et al., 2005). Recurrence was less frequent on the drug than on placebo in every comparison reported: 19% on bupropion and 30% on placebo (p = 0.026), 13% and 21% (p = 0.049), and 16% and 31% (Modell et al., 2005). The time-to-onset analyses were less uniform. Survival to a new episode favoured bupropion XL, yet that difference was clearly significant in only one trial (p < .001) and fell short of significance in the other two (p = .081 and p = .057) (Modell et al., 2005). The authors drew the measured conclusion that starting bupropion early in the season makes prevention possible, not that it reliably prevents every episode.
A Cochrane systematic review later pooled three randomized trials of bupropion XL, reported in four publications with data from 1100 adults with a history of the disorder, and found the risk of recurrence on bupropion XL to be little more than half that on placebo (RR 0.56, 95% CI 0.44 to 0.72) (Gartlehner et al., 2019). How many people must be treated to spare one an episode depends on how likely an episode was to begin with: the number needed to treat for benefit was 8 (95% CI 6 to 12), 5 (95% CI 4 to 7) or 4 (95% CI 3 to 6) as baseline recurrence risk rose across 30%, 50% and 60%, so the drug pays off most for those at highest risk (Gartlehner et al., 2019). Against that benefit stand its harms. Headache was more common on bupropion XL than on placebo, and so, on low-quality evidence, were insomnia and nausea (Gartlehner et al., 2019). The reviewers also flagged methodological limitations in every trial and high attrition, noted that no other second-generation antidepressant and no comparison with light therapy had been studied, and observed that, even in a high-risk population, three in four people who take the drug preventively will not benefit from it; they advised clinicians to discuss the advantages and disadvantages of preventive treatment with patients and to base the decision strongly on patient preferences (Gartlehner et al., 2019).
Light Therapy
Preventive light therapy has far thinner support. A companion Cochrane review located a single eligible randomized trial, in 46 adults with a history of seasonal affective disorder, which compared visors delivering bright white light (2500 lux), visors delivering infrared light (0.18 lux) and no treatment (Nussbaumer-Streit et al., 2019). A new episode developed in 43% (6/14) of the bright-light group, 33% (5/15) of the infrared group and 67% (6/9) of the untreated group. Both light conditions therefore had fewer episodes than no treatment, but the confidence intervals were wide and crossed no effect (risk ratios of 0.64 for bright light and 0.50 for infrared light), and the two kinds of light did about equally well (risk ratio 1.29) (Nussbaumer-Streit et al., 2019). The evidence was rated very low in quality: a small sample, high risk of bias, no blinding, outcomes rated by participants themselves, no check on whether the visors were used, and no information on adverse events. The reviewers judged that methodological limitations and the small size of the only available study precluded conclusions about preventive light therapy (Nussbaumer-Streit et al., 2019).
Psychological Prevention
Psychological prevention is in the same position. The corresponding Cochrane review found one randomized trial, again with 46 adults who had a history of the disorder, comparing mindfulness-based cognitive therapy, delivered as eight weekly individual sessions of 45 to 60 minutes, with treatment as usual (Forneris et al., 2019). A new depressive episode occurred in 65% of the 23 participants given the therapy and 74% of the 23 given usual care, a difference that was not significant (RR 0.88, 95% CI 0.60 to 1.30). Among those who developed an episode, depression severity on the self-report Inventory of Depressive Symptomatology did not differ either (mean difference 1.20, with an interval spanning no difference), and dropout was similar in the two arms (17% and 13%) (Forneris et al., 2019). With very low-quality evidence, high risk of bias, imprecision from the small sample, no data on adverse events and no comparison against other preventive options, the reviewers judged the evidence insufficient to support any conclusion about whether the therapy prevents seasonal episodes (Forneris et al., 2019). A related result from acute treatment points in a similar direction without settling the question: the lower recurrence in the second winter after cognitive-behavioural therapy tailored to the disorder (CBT-SAD), set out under Treatment Evidence, bears on prevention.
The Current Position
A narrative review written for the fortieth anniversary of the diagnosis also surveyed the prevention literature (Rybakowski, 2024). The systematic review described above, restricted to randomized trials, found only one small trial of preventive light and could reach no firm conclusion.
The current position is therefore clear in outline. Among second-generation antidepressants, only bupropion XL has been tested against placebo for prevention, and its benefit is modest, larger where the risk of recurrence is higher, and accompanied by side effects (Gartlehner et al., 2019). Preventive light therapy and mindfulness-based cognitive therapy remain unproven rather than disproven, each resting on a single small trial of very low quality, and the other candidates have not been tested. The guideline described on this page makes no separate recommendation on starting treatment before symptoms begin.
How Treatments Work
Showing that a treatment works is not the same as showing why it works. Each of the two front-line treatments for seasonal affective disorder, bright light therapy and the form of cognitive-behavioural therapy tailored to it (CBT-SAD), comes with a proposed mechanism, and the studies described here have begun to test those mechanisms rather than assume them. The two bodies of evidence differ in kind. The circadian account of light rests on a literature that its own reviewers judge inconclusive, while the cognitive account of CBT-SAD rests on a small number of careful mediation analyses drawn from a narrow base.
Light Therapy
The traditional explanation of light therapy is chronobiological. The phase-shift hypothesis proposes that light relieves seasonal depression by correcting the timing of the body's circadian rhythms; the timing evidence behind it is set out under Biological Accounts and Sleep and Circadian Rhythms. Its standing today is contested.
The most systematic look at whether treatment actually changes circadian outcomes is a PRISMA-guided review of forty eligible papers, which rated each study's risk of bias with the revised Cochrane tool or ROBINS-I (Menculini et al., 2018). Bright light therapy came out as effective for both the mood symptoms of the disorder and its circadian disturbances, and the reviewers judged other treatments promising but in need of replication. On the mechanism itself they were guarded. The circadian evidence was not robust, and whether mood improvement is tied to circadian change remained a matter of controversy; heterogeneous methods, small samples and sample selection the authors considered less than optimal all limit what the included trials can show (Menculini et al., 2018). The review therefore establishes that light helps, without establishing that it helps by resetting the clock.
A more recent narrative review goes further and argues that bright light benefits depression through routes that extend beyond its circadian effects (Kosanovic Rajacic et al., 2023). Its case rests partly on breadth: light outperformed placebo across a range of depressive disorders, so the review's scope reaches past seasonal affective disorder to non-seasonal, bipolar and treatment-resistant depression, and its discussion takes in the intrinsically photosensitive retinal ganglion cells. The authors present the treatment as easy to deliver, inexpensive and well tolerated, but they call for future studies of the clinical and biological predictors of response, of which populations may benefit, and of protocols for the timing, frequency and duration of treatment (Kosanovic Rajacic et al., 2023). Being narrative rather than systematic, the review argues for a non-circadian mechanism more than it demonstrates one.
A third strand implicates serotonin: in a small crossover study of remitted patients, short-term tryptophan depletion brought back depressive symptoms for a brief period, which the authors read as a sign that serotonin may play a part in the response to light (Neumeister et al., 1997; see Biological Accounts).
Cognitive-Behavioural Therapy
CBT-SAD is thought to work by changing the negative beliefs people hold about the seasons themselves, the cognitive vulnerability at the centre of the integrative cognitive-behavioural model (see The Integrative Cognitive-Behavioural Model of Seasonal Affective Disorder). The mediation evidence described below supports that model (Rohan et al., 2023).
Seasonal beliefs appear to explain why CBT-SAD outlasts light. In a secondary analysis of a randomised trial, 177 adults with recurrent major depression with a seasonal pattern received six weeks of either light therapy (n = 89) or CBT-SAD (n = 88) and were followed across one and two later winters (Rohan et al., 2023). Scores on the Seasonal Beliefs Questionnaire improved more with CBT-SAD, and depression at follow-up was less severe in that arm. Latent growth curve mediation models attributed CBT-SAD's long-term advantage to the change in seasonal beliefs, with indirect effects ranging from β = .091 to .162. Two rival candidates did not hold up. Light moved chronotype further toward morningness, and CBT-SAD reduced brooding rumination more, yet neither change emerged as a mediator of depression at follow-up. The analysis was secondary, its sample was modest, and missing data were handled by maximum likelihood estimation (Rohan et al., 2023).
The acute phase of treatment tells a consistent story. Framed as experimental therapeutics, an analysis of 177 adults with seasonal depression (83.6% women, 92.1% non-Hispanic White, mean age 45.6) used growth models over a six-week course to set candidate cognitive and chronobiological vulnerabilities against each other (Rohan et al., 2020). CBT-SAD lowered seasonal-belief scores more than light did, and light reduced eveningness more than CBT-SAD did. Only the cognitive change kept pace with recovery: the slope of seasonal-belief change tracked the slope of symptom decline, whereas the effects did not hold for general dysfunctional attitudes, measured with the Dysfunctional Attitudes Scale, and change in eveningness on the Morningness-Eveningness Questionnaire was unrelated to change in symptoms (Rohan et al., 2020). What appears to matter, on this evidence, is thinking about winter in particular rather than depressogenic thinking in general. A sample this homogeneous in sex and ethnicity limits how far the pattern can be generalised.
An earlier secondary analysis of 69 adults across three treatment arms found that dysfunctional attitudes, negative automatic thoughts and rumination fell to a similar degree in every arm, yet only with CBT alone did larger reductions in automatic thoughts and attitudes predict milder depression the following winter; the authors read cognitive change as a working mechanism of CBT given alone, but as a by-product of lifting depression under light therapy and under the combined treatment (Evans et al., 2013; see Testing Whether Cognition Causes Seasonal Episodes). Attrition reached 25%, the sample was mostly women, and some of the interactions only approached significance.
The Limits of the Mediator Evidence
Neither mechanism is settled. For light, a systematic review found the circadian evidence weak and a narrative review proposes a non-circadian route, so how light works remains an open question. For CBT-SAD, the mediator findings are consistent across the acute phase and later winters, but they come from secondary analyses of samples that were mostly women, and all of the CBT-SAD mediation work described here comes from a single research group. Until independent groups test seasonal beliefs as a mediator in their own trials, the cognitive account of CBT-SAD is well specified and supported, but it has not been replicated independently.
Side Effects and Safety of Light Therapy
Bright light carries a lighter burden of side effects than antidepressant medication, but it is not free of them. A review of two decades of light-therapy research judged its adverse effects to be fewer than those of drugs, while naming two that clinicians should watch for: the emergence of hypomania and over-activation of the autonomic nervous system (Terman & Terman, 2005). The same review stressed that both the dose and the time of day at which light is given need adjusting to the individual, so a regimen that suits one patient may over- or under-treat another (Terman & Terman, 2005).
Eye Safety
Because treatment means sitting close to an intense light source every winter, the eyes have been the main safety question. A prospective ophthalmological study examined 50 patients with seasonal affective disorder before and after daily sessions of 30 minutes at 10,000 lux lasting two to eight weeks, and followed 17 of them over three to six years of seasonal use, a cumulative exposure of 60 to 1,250 hours (Gallin et al., 1995). Examinations included slit-lamp biomicroscopy, visual acuity, intraocular pressure, fundus photography, the Amsler grid and a macular stress test, and the eyes after treatment showed no change from their state beforehand (Gallin et al., 1995). The authors read this as evidence of ocular safety, recommending periodic eye examinations only for some patients, and their own caveats limit the result: not every patient received every test, firm contraindications could not be specified, and people taking photosensitising drugs or with existing eye abnormalities were singled out for caution (Gallin et al., 1995). Whether patients keep using the lamp, and how many stop, is a separate practical question taken up under Light Therapy in Practice.
Mood Switching in Bipolar Depression
The hypomania risk matters most for people with bipolar disorder. The evidence in this subsection concerns bipolar depression in general, not seasonal episodes specifically. In a six-week double-blind trial, 46 medicated adults with bipolar I or II depression were randomised in equal groups of 23 to midday light at 7,000 lux or a dim 50-lux placebo (Sit et al., 2018). Remission was reached by 68.2% of the bright-light group against 22.2% of the placebo group (adjusted odds ratio 12.6), end-of-trial depression scores were lower with bright light (9.2 versus 14.9), and no patient in either group switched mood polarity (Sit et al., 2018). Sleep quality improved in both groups without a significant difference, and the trial excluded people with mania, mixed symptoms or rapid cycling, the patients in whom switching would be the greatest concern (Sit et al., 2018).
A later meta-analysis of six randomised trials with 297 participants did not confirm a clear benefit. Across trials, the difference between bright light and control was non-significant for remission (risk ratio 1.81, p = .42) and for depression scores, and light did not raise the rate of manic switching (Takeshima et al., 2020). Only the studies judged least indirect favoured light (risk ratio 3.09). The authors called the evidence inconclusive and asked for larger trials, citing few and underpowered studies, high heterogeneity and the short duration of every trial (Takeshima et al., 2020). On the present evidence, then, light has not been shown to provoke mania in bipolar depression, but the trials are too few and too brief to settle the question.
Reviews of light therapy describe mostly minor complaints. Nausea, agitation, headaches and visual complaints are reported, as are dizziness and disturbed sleep, while a switch into (hypo)mania is very rare (Meesters & Gordijn, 2016). The same review treats existing retinal disease, photosensitising medication and recent eye surgery as relative contraindications, and advises anyone uncertain about the health of their eyes to consult an ophthalmologist before starting (Meesters & Gordijn, 2016). A second review likewise warns against combining light with photosensitising drugs (Melrose, 2015). Neither review states how often each of these effects occurs.
Emerging Treatments
Several alternatives to the standard white light box are under investigation. Each rests on a small number of trials, and none has the evidence base of the treatments covered in Treatment Evidence. The standing of every option described here is best read as early.
Narrow-Band Blue Light
Two trials compared short-wavelength light with a dim red control. In the first, 24 patients with major depression in a seasonal pattern, each with a SIGH-SAD score of at least 20, received 45 minutes a day for 3 weeks of either blue LED light (468 nm, 607 microW/cm2) or dim red LED light (654 nm, 34 microW/cm2). SIGH-SAD scores fell more under the blue units (F = 6.45, p = .019), and the authors judged narrow-band blue light an effective treatment for the disorder (Glickman et al., 2006). A later double-blind trial of thirty adults with seasonal affective disorder, also run over 3 weeks, set 470 nm blue light against a red placebo. Mean improvement on the HAMD-17 was 51% with blue light and 32% with red (P=.05), and 60% of the blue group met the response criterion against 13% of the red group (P=.01). Participants whose depression was purely seasonal did numerically better than those with seasonal intensification of a year-round depression, but that split was made post hoc and the difference was numerical only; in a following open-label phase with 25 participants, both arms improved over baseline (Strong et al., 2009). Low-intensity blue light has also been tested against bright white light in people with sub-syndromal symptoms, where the two performed alike on the main symptom measure; that trial is described under Types of Seasonal Affective Disorder (Meesters et al., 2016).
All-Day Room Lighting
A different approach replaces the brief daily session with prolonged exposure to bright ambient light. In a 4-week proof-of-concept randomized trial with 62 patients, BROAD light therapy, delivered for at least 6 h per day, was compared with 30 min a day at a 10,000 lux light box. Symptom improvement, rated on the 29-item seasonal version of the Hamilton Depression Rating Scale, was similar in the two groups. Within the BROAD group, an exploratory analysis linked higher illuminance with greater improvement. The authors described the approach as feasible and as seeming to work about as well as the light box, and they treated the result as a first, preliminary estimate (Sandkühler et al., 2022).
Exercise
Physical exercise has been compared directly with light. In a one-week study of female participants from three diagnostic groups, 1 h of midday pedalling on a bicycle ergometer (between 13.00 and 14.00 h) was set against 2 h of 2500 lux bright light (between 14.00 and 16.00 h). Among the women with winter depression, exercise and light produced comparable benefits, whereas in non-seasonal depression exercise did better than light; untreated winter-depressed participants showed no significant change over the week. Oxygen consumption was lower in winter depression than in both the non-depressed and the non-seasonally depressed groups, and light did not alter oxygen consumption in non-seasonal depression. The authors took the pattern as implicating energy-regulating systems in antidepressant action (Pinchasov et al., 2000). A single week of treatment in an all-female sample leaves exercise as a candidate rather than an established alternative to light.
Related Conditions and Seasonal Mood in the General Population
Seasonal affective disorder shares low mood with several neighbouring states, so telling it apart depends on the pattern of episodes over time. No single symptom settles it, and neither does a good response to treatment. Four boundaries matter most: ordinary seasonal fluctuation in healthy people, nonseasonal depression with atypical features, the use of light therapy outside seasonal depression, and other depressive presentations that are defined by timing or course.
Ordinary Seasonal Mood Change
In the general population, mood barely changes with the seasons, and that small change is not a disorder. A crowdsourced online survey of 5,282 Dutch adults found positive affect highest and depressive symptoms lowest in spring. The differences were small or very small (d = 0.25 for positive affect, spring against summer), and the 503 respondents who completed the survey twice showed no seasonal effect within the same person (Winthorst et al., 2020). Those authors concluded that the seasons do not greatly affect mood. Respondents high in neuroticism reported more seasonality in the cross-sectional comparison, although that interaction was not significant within subjects, and the design was limited: there was no prior sample-size calculation, and most questionnaires were completed in spring. Weather tells a similar story. In screening data from 14,478 Dutch adults, temperature, hours of sunshine and duration of rainfall were each unrelated to mood or depression, and the authors judged a weather–depression link probably absent. Depression was nonetheless more prevalent in summer and autumn than in the other seasons in that sample (Huibers et al., 2010). Prevalence, Course and the Latitude Debate places these studies within the wider argument about how common seasonal depression is. For classification, a slight population-wide shift in affect is a different thing from a recurrent clinical episode.
Nonseasonal Depression with Atypical Features
The winter presentation of seasonal affective disorder overlaps with nonseasonal major depression that has atypical features, and a direct comparison suggests the overlap is partial. Tam and colleagues compared 53 patients with seasonal affective disorder and 54 with nonseasonal major depression on the atypical depression diagnostic scale (ADDS). The seasonal group reported more hyperphagia and more hypersomnia. The nonseasonal group scored higher on interpersonal sensitivity and on avoidance of rejection. The rate of atypical depression did not differ significantly between the groups, and the authors read the pattern as two separate subtypes whose symptom pictures overlap (Tam et al., 1997). The comparison was a single patient-group study, so it marks a difference in emphasis rather than a clean dividing line. Symptoms, Diagnosis and Measurement covers how the seasonal pattern itself is established.
Light Therapy in Nonseasonal Depression
Bright light also helps many people whose depression has no seasonal pattern, so a response to light does not on its own mark a case as seasonal. A systematic review and meta-analysis pooled 20 randomised trials with 881 participants who had non-seasonal depression. Light therapy lowered depression scores more than control conditions did, a small-to-moderate effect (standardised mean difference −0.41, 95% CI −0.64 to −0.18), and 59.1% of light-treated participants responded, against 38.0% of controls (Perera et al., 2016). The evidence was weak in several respects. Heterogeneity was substantial (I² = 60%) and remained unexplained, many trials were small and at high risk of bias, and GRADE rated the overall quality of evidence as low. Apparent advantages for stand-alone over adjunctive light, for morning sessions and for outpatients were not statistically significant. The authors accordingly offered light as a possible adjunct while stressing the poor quality of the evidence.
A double-blind trial was more definite. Lam and colleagues randomised 122 adults aged 19 to 60 with nonseasonal major depressive disorder to eight weeks of one of four conditions: 10,000-lux light for 30 minutes a day, fluoxetine 20 mg a day, the two combined, or a sham negative ion generator with a placebo pill. On the Montgomery-Åsberg Depression Rating Scale, the combination outperformed placebo (d = 1.11), and light alone also beat placebo (d = 0.80). Fluoxetine alone did not differ significantly from placebo (d = 0.24). Response rates were 75.9% with the combination, 50.0% with light, 29.0% with fluoxetine and 33.3% with placebo, and remission followed the same order (Lam et al., 2016). The analysis was a modified intent-to-treat one. The authors concluded that light was efficacious and well tolerated in nonseasonal major depression. Treatment Evidence reviews light therapy within seasonal affective disorder itself, and set against these results it shows that light response is shared across depressive subtypes rather than being diagnostic of a seasonal one.
Bipolar Depression, Chronic Depression and Other Timing-Defined Presentations
Depressive episodes within bipolar disorder can also follow the seasons. Types of Seasonal Affective Disorder covers the seasonal pattern in bipolar disorder, which DSM-5-TR records with the same specifier. Persistent depressive disorder is the contrast in the other direction: it is defined by chronicity, a low mood that persists over long periods, whereas seasonal affective disorder is defined by episodes that come and go with the time of year. Postpartum depression resembles the seasonal pattern in one formal respect: it too is a specifier set by when an episode begins, not by a distinctive set of symptoms. In each case the questions that separate the condition from seasonal affective disorder are about timing and course. Symptom content alone does not answer them.
Worked Example
A positive result on a seasonality questionnaire is easy to misread as a diagnosis. The arithmetic below shows why it is not, using the two published accuracy profiles of the Seasonal Pattern Assessment Questionnaire (SPAQ). In an Icelandic community sample of 81 people checked against clinical interviews, the SPAQ picked up 94% of the winter problem group (sensitivity) but correctly cleared only 73% of the remainder (specificity), and the author judged its case-finding ability for SAD poor, partly because it separated full SAD from subsyndromal SAD badly (Magnusson, 1996). A later validation that compared SAD outpatients with depressed and non-depressed outpatients and controls found the reverse balance: specificity of 94% and sensitivity of only 44%, with the false positives confined to the depressed group (Mersch et al., 2004). Those authors concluded that the instrument can serve for screening but not for diagnosis, and they cautioned that most of their SAD patients had been treated and filled in the questionnaire while not depressed, which may have lowered the sensitivity estimate (Mersch et al., 2004).
The following figures are illustrative and come from no study. Suppose 10,000 adults are screened and 500 of them (5%) have SAD. The 5% is an assumption chosen to keep the numbers round; it sits close to the pooled prevalence of 5.01% that a meta-analysis of cross-sectional studies reported (Kim et al., 2025). That leaves 9,500 people without the disorder.
With the community profile (sensitivity 94%, specificity 73%), the questionnaire flags 470 of the 500 true cases and misses 30. Among the 9,500 unaffected people, 27% are wrongly flagged, which is 2,565 false positives, and 6,935 are correctly cleared. Of the 3,035 positive screens, only 470, or 15.5%, belong to people who have SAD: this proportion is the positive predictive value. A negative screen, by contrast, is correct 6,935 times out of 6,965, a negative predictive value of 99.6%.
With the patient-group profile (sensitivity 44%, specificity 94%), the questionnaire catches 220 true cases and misses 280. Six percent of the unaffected group, 570 people, screen positive, and 8,930 screen negative. The positive predictive value rises to 27.8% (220 of 790 positive screens), while the negative predictive value falls slightly to 97.0% (8,930 of 9,210). Higher specificity roughly doubles the share of positive screens that are real cases, yet even then nearly three in four positives are false, and more than half of the true cases go undetected.
Figure 1
Share of Positive Screens That Are True Cases at an Assumed 5 Percent Prevalence
Note. Illustrative arithmetic for 10,000 people screened, of whom 5 percent are assumed to have seasonal affective disorder. Sensitivity and specificity pairs are from the community validation of the Seasonal Pattern Assessment Questionnaire (Magnusson, 1996) and the patient-group validation (Mersch et al., 2004); the prevalence is an assumption, not a study result.
The lesson is that when a disorder is uncommon, most positive screens are false, whichever accuracy profile applies. In the community validation itself, the positive predictive value was 45%, already well below the instrument's sensitivity (Magnusson, 1996); spread across a general population, the same instrument would do worse. The screening demonstration earlier in the article lets the reader vary prevalence, sensitivity and specificity and watch the positive predictive value respond.
Discussion
Taken together, the studies described on this page portray a recurrent depressive condition whose seasonal timing is real in clinical samples but weaker in the population than questionnaires suggest. People with a confirmed diagnosis mostly remit in summer, as the follow-up described under What Seasonal Affective Disorder Is shows, yet prospective community data and general-population reviews find seasonal effects that are modest, absent in many studies, or very small. Retrospective screening inflates prevalence, which is the arithmetic behind the Worked Example, and latitude explains little once populations, ancestry and survey methods differ. The disorder is therefore best read as a course feature that is clearly present in some people with recurrent depression, while its standing as a distinct entity remains open.
On cause, no single account carries the evidence. Patients lengthen their melatonin secretion in winter as healthy people do not, a phase delay plausibly explains part of the condition in part of the patients, serotonin depletion and imaging studies implicate serotonin, although the imaging studies disagree about the direction of change and all of the samples are small, and retinal findings are preliminary. These accounts need not compete: each may fit some patients. The dual vulnerability and cognitive-behavioural models add a second, psychological liability that determines whether a seasonal change in energy and sleep becomes a depressive episode, and a review of both literatures argued that an integrative approach is probably the most productive way forward (Rohan et al., 2009a).
The treatments are better established than their mechanisms. Light therapy is effective, but a systematic review judged the circadian evidence for how it works not robust, and the mediation evidence that seasonal beliefs explain the durability of cognitive-behavioural therapy comes from one research group's secondary analyses. Medication and prevention rest on fewer and weaker trials, with bupropion XL the only preventive option with placebo-controlled support.
The question shared across these literatures is direction. Slowed processing speed, negative automatic thoughts outside winter and altered serotonin transporter binding could each be causes of seasonal episodes, consequences of repeated episodes, or markers of the state. Settling that would take designs not reported in the studies described here: prospective cohorts assessed before a first seasonal episode with objective measures across several seasons, experiments that change cognition in people at risk, independent replications of the mediator trials, and screening instruments validated across cultures. Until such work exists, both the causal role of cognition and the construct validity of the diagnosis remain unresolved.
Current Directions
Recent work on seasonal affective disorder turns on a handful of open questions, each still being argued.
Does latitude shape where the disorder occurs? Pooled prevalence data now link higher latitude with more of the full and subsyndromal winter forms, which the authors took to underscore a potential role for seasonal light variation, although every pooled study was cross-sectional (Kim et al., 2025). The wider dispute is set out under Prevalence, Course and the Latitude Debate.
What does the brain do across the seasons? Imaging evidence remains preliminary and mostly cross-sectional, and its reviewers call for adequately powered longitudinal studies (Majumdar et al., 2025). Whether a possible amygdala route for the effect of light on mood holds in people with the disorder has not yet been tested; Brain Networks and Imaging covers this evidence.
How should winter sleep be measured? Patients report much more extra winter sleep at interview than actigraphy records, so hypersomnolence remains hard to pin down, and combining several measures is advised before any sleep intervention (Wescott et al., 2023). Sleep and Circadian Rhythms discusses what the gap implies.
How stable is the diagnosis? Most trial participants with confirmed disorder remitted over the following summer, which its authors read as support for the seasonal construct, though the study had one site and a fairly homogeneous sample (Rohan et al., 2024). The study is reported under What Seasonal Affective Disorder Is.
Why does psychological treatment last? In one secondary analysis of a modest sample, change in seasonal beliefs mediated the more durable benefit of cognitive-behavioural therapy for the disorder (CBT-SAD) over light therapy, while changes in brooding and morningness did not emerge as mediators of later depression (Rohan et al., 2023). How Treatments Work places this among the other mediator evidence.
How else can light be delivered? A proof-of-concept trial judged BROAD light therapy, which spreads exposure across the day, feasible and apparently similar in effect to a light box, offering only a preliminary first estimate (Sandkühler et al., 2022). Emerging Treatments takes up such alternatives.
Common Misconceptions
- Seasonal affective disorder is a separate disorder in DSM-5-TR.
- The current manual records seasonal timing as a specifier, with seasonal pattern, added to a diagnosis of recurrent major depressive disorder or bipolar disorder, not as a standalone diagnosis (American Psychiatric Association, 2022). The belief persists because the condition has its own widely used name and acronym.
- Vitamin D deficiency is the proven cause of seasonal depression.
- A placebo-controlled trial of vitamin D in healthcare professionals with seasonal symptoms found it no better than placebo, although the trial reached only about a fifth of its planned sample, so its authors described the result as no demonstrated effect rather than proof of none (Frandsen et al., 2014). The idea is plausible because sunlight drives vitamin D synthesis, but plausibility is not trial evidence.
- Any low mood in winter is seasonal affective disorder.
- Seasonal change in mood and behaviour forms a continuum in the general population, with the disorder at its extreme end, and estimates of how many people have it shift with the case definition used (Kasper et al., 1989a). A mild winter dip lies on that continuum without meeting the threshold for recurrent depressive episodes.
- Suicide peaks in the dark winter months.
- Most population studies place the suicide peak in spring, mainly among men, older people and violent methods, with a secondary peak in autumn, and studies in both hemispheres report seasonality, although some found none and results were inconsistent (Christodoulou et al., 2012). The belief borrows the gloom of winter depression, but the population spring peak and the timing of seasonal episodes are separate patterns.
Glossary
- Bright light therapy.
- Daily exposure to an intense artificial light source, commonly 10,000 lux for about 30 minutes, used to treat seasonal affective disorder and tested in some nonseasonal depressions.
- CBT-SAD.
- A form of cognitive-behavioural therapy tailored to seasonal affective disorder; in one research group's trials it matched light therapy acutely and was followed by fewer recurrences in the second winter.
- Circadian rhythm.
- The roughly daily cycle of physiology and behaviour governed by an internal clock, whose timing light can move earlier or later.
- Dawn simulation.
- A treatment in which bedroom light brightens gradually before waking, used in place of a session at a light box; trials comparing it with bright light have given mixed results.
- Dim-light melatonin onset.
- The evening time at which melatonin begins to rise when measured under dim light, used as a marker of circadian phase; morning light moves it earlier and evening light later.
- Dual vulnerability model.
- The proposal that seasonal depression combines a seasonal, vegetative vulnerability triggered by reduced light with a depressive, cognitive-affective vulnerability that governs whether low mood and negative thinking follow.
- Electroretinogram.
- A recording of the electrical response of the retina to light, used to measure the function of rods and cones.
- Hypersomnia.
- Excessive sleep or sleepiness; in seasonal affective disorder it is reported at clinical interview more often than objective recordings confirm.
- Lux.
- The unit of illuminance, the amount of light reaching a surface, in which light-therapy doses are specified.
- Melanopsin.
- The photopigment of the retinal ganglion cells that signal light to the circadian system; a role in seasonal affective disorder has been proposed but not confirmed.
- Melatonin.
- A hormone secreted at night, whose secretion bright light suppresses and whose nightly duration lengthens when the dark period is longer.
- Phase-shift hypothesis.
- The proposal that winter depression arises when the circadian clock drifts later relative to sleep, so that morning light should help more than evening light; its standing is contested.
- Photoperiod.
- The length of the daily light period; a short, winter-like photoperiod lengthened nightly melatonin secretion and sleep in healthy volunteers.
- Positive predictive value.
- The proportion of positive screening results that belong to people who truly have the condition, a share that falls as the condition becomes rarer in the screened group.
- Seasonal beliefs.
- Negative beliefs about the seasons and light, measured with the Seasonal Beliefs Questionnaire, whose change during CBT-SAD tracked and mediated improvement in one research group's analyses.
- Seasonal Pattern Assessment Questionnaire.
- The reference retrospective screening questionnaire for seasonality; it can screen but not diagnose, and its estimates tend to overstate prevalence.
- Seasonal pattern specifier.
- The DSM-5-TR label, with seasonal pattern, added to a diagnosis of recurrent major depressive disorder or bipolar disorder when episodes recur with the seasons.
- Serotonin transporter.
- The protein that clears serotonin from the synapse; in PET studies its brain binding was higher in autumn and winter in one healthy sample but lower in winter in people resilient to seasonal affective disorder, and studies of patients disagree about whether their seasonal change in binding is larger or smaller than that of healthy people.
- Subsyndromal seasonal affective disorder.
- A milder seasonal pattern below the diagnostic threshold, sometimes called the winter blues, which was more prevalent than the full disorder in pooled cross-sectional surveys.
- Tryptophan depletion.
- An experimental procedure that lowers the precursor of serotonin with an amino acid drink lacking tryptophan; in patients who had recovered with light therapy it briefly brought back depressive symptoms.
Key Researchers
Raymond W. Lam
Psychiatrist at the University of British Columbia who led the Can-SAD trial comparing light therapy with fluoxetine and reviewed the pathophysiology of seasonal affective disorder. ORCID - Google Scholar - Faculty Page
Alfred J. Lewy
Psychiatrist at Oregon Health & Science University who showed that bright light suppresses melatonin in humans and developed the phase-shift hypothesis of winter depression. Wikipedia
Teodor T. Postolache
Psychiatrist at the University of Maryland School of Medicine who studied spring peaks of suicide and seasonal mood in African American students. ORCID - Faculty Page
Kathryn A. Roecklein
Psychologist at the University of Pittsburgh who studied melanopsin, the pupil response to light and sleep in seasonal affective disorder. ORCID - Faculty Page
Kelly J. Rohan
Psychologist at the University of Vermont who developed cognitive-behavioural therapy for seasonal affective disorder and compared it with light therapy in randomised trials. ORCID
Norman E. Rosenthal
Psychiatrist, formerly at the National Institute of Mental Health and now at Georgetown University School of Medicine, who led the 1984 description of seasonal affective disorder and its first light therapy trial. Wikipedia
Michael Terman
Psychologist at Columbia University who ran controlled trials of bright light, light timing, dawn simulation and negative air ionization. Wikipedia
Thomas A. Wehr
Scientist emeritus at the National Institute of Mental Health who described summer depression and showed that patients with seasonal affective disorder signal the change of season through melatonin duration. ORCID - Wikipedia
Anna Wirz-Justice
Chronobiologist and professor emerita at the Centre for Chronobiology in Basel who studied food intake across the seasons in seasonal affective disorder and reviewed seasonality in mood disorders. Faculty Page
Michael A. Young
Psychologist and emeritus professor at the Illinois Institute of Technology who proposed the dual vulnerability model of seasonal depression and tested the role of rumination. ORCID - Faculty Page
Frequently Asked Questions
Is seasonal affective disorder a separate diagnosis in DSM-5-TR?
No. DSM-5-TR records seasonal timing as a specifier, with seasonal pattern, which a clinician adds to a diagnosis of recurrent major depressive disorder or bipolar disorder rather than diagnosing a standalone disorder (American Psychiatric Association, 2022).
What is the difference between seasonal affective disorder and the winter blues?
Seasonal change in mood and behaviour forms a continuum in the general population, with seasonal affective disorder at its extreme end, and estimates of its prevalence shift with the case definition used (Kasper et al., 1989a). The milder form below the diagnostic threshold, informally called the winter blues, is labelled subsyndromal seasonal affective disorder.
Can seasonal affective disorder happen in summer?
Yes, although the summer pattern is uncommon. Following 30 summer-depressed and 30 sex-matched winter-depressed patients prospectively, Wehr and colleagues found insomnia and poor appetite more typical of summer episodes and atypical vegetative symptoms more typical of winter ones (Wehr et al., 1991).
Does living farther north make seasonal affective disorder more likely?
Only weakly. A review of prevalence studies found no significant correlation with latitude across all studies, a positive correlation within North America and only a trend within Europe, and concluded that latitude plays a minor role (Mersch et al., 1999). Later pooled survey data found a modest positive association, and the question is not settled.
Does seasonal affective disorder affect thinking or memory?
In a small longitudinal study, 29 medication-free people with winter-pattern seasonal affective disorder performed worse than 30 matched controls on working memory, processing speed and motor speed in both summer and winter, and the authors read slowed processing speed as a possible trait marker (Hjordt et al., 2017).
How well does light therapy work?
Pooling 19 randomised trials, Pjrek and colleagues found bright light superior to placebo (SMD = -0.37 on depression; RR = 1.42 for response), although heterogeneity was moderate, the trials were small to medium and the authors asked for larger ones (Pjrek et al., 2020).
Is cognitive-behavioural therapy as effective as light therapy?
In a randomised trial of 177 adults, the two treatments did not differ in the first winter after treatment, but by the second winter recurrence was 27.3% after the seasonal form of cognitive-behavioural therapy against 45.6% after light therapy (Rohan et al., 2016). These trials come from a single research group and have not yet been independently replicated.
Can seasonal affective disorder be prevented?
A Cochrane review of three trials with 1100 adults found that bupropion XL lowered the risk of recurrence to a little more than half that on placebo, but it also raised the risk of headache, every trial had methodological limitations, and even among those at high risk, three in four people taking it preventively would not benefit (Gartlehner et al., 2019).
References
Abou Elseoud, A., Nissilä, J., Liettu, A., Remes, J., Jokelainen, J., Takala, T., Aunio, A., Starck, T., Nikkinen, J., Koponen, H., Zang, Y.-F., Tervonen, O., Timonen, M., & Kiviniemi, V. (2014). Altered resting-state activity in seasonal affective disorder. Human Brain Mapping, 35(1), 161-172. https://doi.org/10.1002/hbm.22164
Agumadu, C. O., Yousufi, S. M., Malik, I. S., Nguyen, M. C., Jackson, M. A., Soleymani, K., Thrower, C. M., Peterman, M. J., Walters, G. W., Niemtzoff, M. J., Bartko, J. J., & Postolache, T. T. (2004). Seasonal variation in mood in African American college students in the Washington, D.C., metropolitan area. American Journal of Psychiatry, 161(6), 1084-1089. https://doi.org/10.1176/appi.ajp.161.6.1084
Ajdacic-Gross, V., Bopp, M., Ring, M., Gutzwiller, F., & Rossler, W. (2010). Seasonality in suicide: A review and search of new concepts for explaining the heterogeneous phenomena. Social Science & Medicine, 71(4), 657-666. https://doi.org/10.1016/j.socscimed.2010.05.030
Ambar Akkaoui, M., & Geoffroy, P. A. (2025). Screening and evaluating seasonal affective disorder: A systematic review of available assessment tools. Journal of Psychiatric Research, 187, 223-232. https://doi.org/10.1016/j.jpsychires.2025.05.008
American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.). American Psychiatric Association Publishing. https://doi.org/10.1176/appi.books.9780890425787
Amons, P. J. T., Kooij, J. J. S., Haffmans, P. M. J., Hoffman, T. O., & Hoencamp, E. (2006). Seasonality of mood disorders in adults with lifetime attention-deficit/hyperactivity disorder (ADHD). Journal of Affective Disorders, 91(2-3), 251-255. https://doi.org/10.1016/j.jad.2005.11.017
Anderson, J. L., Rosen, L. N., Mendelson, W. B., Jacobsen, F. M., Skwerer, R. G., Joseph-Vanderpool, J. R., Duncan, C. C., Wehr, T. A., & Rosenthal, N. E. (1994). Sleep in fall/winter seasonal affective disorder: Effects of light and changing seasons. Journal of Psychosomatic Research, 38(4), 323-337. https://doi.org/10.1016/0022-3999(94)90037-x
Avery, D. H., Eder, D. N., Bolte, M. A., Hellekson, C. J., Dunner, D. L., Vitiello, M. V., & Prinz, P. N. (2001). Dawn simulation and bright light in the treatment of SAD: A controlled study. Biological Psychiatry, 50(3), 205-216. https://doi.org/10.1016/s0006-3223(01)01200-8
Blazer, D. G., Kessler, R. C., & Swartz, M. S. (1998). Epidemiology of recurrent major and minor depression with a seasonal pattern: The National Comorbidity Survey. British Journal of Psychiatry, 172(2), 164-167. https://doi.org/10.1192/bjp.172.2.164
Byrne, E. M., Raheja, U. K., Stephens, S. H., Heath, A. C., Madden, P. A., Vaswani, D., Nijjar, G. V., Ryan, K. A., Youssufi, H., Gehrman, P. R., Shuldiner, A. R., Martin, N. G., Montgomery, G. W., Wray, N. R., Nelson, E. C., Mitchell, B. D., Postolache, T. T., & Psychiatric Genomics Consortium Major Depressive Disorder Working Group. (2015). Seasonality shows evidence for polygenic architecture and genetic correlation with schizophrenia and bipolar disorder. Journal of Clinical Psychiatry, 76(2), 128-134. https://doi.org/10.4088/jcp.14m08981
Campbell, I., Balda Aizpurua, J. F., Sharifpour, R., Paparella, I., Beckers, E., Mortazavi, N., Read, J., Phillips, C., Collette, F., Talwar, P., Lamalle, L., Zubkov, M., & Vandewalle, G. (2025). Regional activity within the human amygdala varies with season, mood and illuminance. Nature Communications, 16, 11559. https://doi.org/10.1038/s41467-025-67715-3
Camuso, J. A., & Rohan, K. J. (2020). Cognitive vulnerabilities as prognostic predictors of acute and follow-up outcomes in seasonal affective disorder treatment with light therapy or cognitive-behavioral therapy. Cognitive Therapy and Research, 44(3), 468-482. https://doi.org/10.1007/s10608-020-10086-4
Christodoulou, C., Douzenis, A., Papadopoulos, F. C., Papadopoulou, A., Bouras, G., Gournellis, R., & Lykouras, L. (2012). Suicide and seasonality. Acta Psychiatrica Scandinavica, 125(2), 127-146. https://doi.org/10.1111/j.1600-0447.2011.01750.x
Coimbra, D. G., Pereira e Silva, A. C., de Sousa-Rodrigues, C. F., Barbosa, F. T., de Siqueira Figueredo, D., Araújo Santos, J. L., Barbosa, M. R., de Medeiros Alves, V., Nardi, A. E., & de Andrade, T. G. (2016). Do suicide attempts occur more frequently in the spring too? A systematic review and rhythmic analysis. Journal of Affective Disorders, 196, 125-137. https://doi.org/10.1016/j.jad.2016.02.036
Danilenko, K. V., & Ivanova, I. A. (2015). Dawn simulation vs. bright light in seasonal affective disorder: Treatment effects and subjective preference. Journal of Affective Disorders, 180, 87-89. https://doi.org/10.1016/j.jad.2015.03.055
Davis, C., & Levitan, R. D. (2005). Seasonality and seasonal affective disorder (SAD): An evolutionary viewpoint tied to energy conservation and reproductive cycles. Journal of Affective Disorders, 87(1), 3-10. https://doi.org/10.1016/j.jad.2005.03.006
Eastman, C. I., Young, M. A., Fogg, L. F., Liu, L., & Meaden, P. M. (1998). Bright light treatment of winter depression: A placebo-controlled trial. Archives of General Psychiatry, 55(10), 883-889. https://doi.org/10.1001/archpsyc.55.10.883
Evans, M., Rohan, K. J., Sitnikov, L., Mahon, J. N., Nillni, Y. I., Lindsey, K. T., & Vacek, P. M. (2013). Cognitive change across cognitive-behavioral and light therapy treatments for seasonal affective disorder: What accounts for clinical status the next winter? Cognitive Therapy and Research, 37(6), 1201-1213. https://doi.org/10.1007/s10608-013-9561-0
Faedda, G. L., Tondo, L., Teicher, M. H., Baldessarini, R. J., Gelbard, H. A., & Floris, G. F. (1993). Seasonal mood disorders: Patterns of seasonal recurrence in mania and depression. Archives of General Psychiatry, 50(1), 17-23. https://doi.org/10.1001/archpsyc.1993.01820130019004
Fellinger, M., Waldhör, T., Serretti, A., Hinterbuchinger, B., Pruckner, N., König, D., Gmeiner, A., Vyssoki, S., Vyssoki, B., & Fugger, G. (2022). Seasonality in major depressive disorder: Effect of sex and age. Journal of Affective Disorders, 296, 111-116. https://doi.org/10.1016/j.jad.2021.09.051
Forneris, C. A., Nussbaumer-Streit, B., Morgan, L. C., Greenblatt, A., Van Noord, M. G., Gaynes, B. N., Wipplinger, J., Lux, L. J., Winkler, D., & Gartlehner, G. (2019). Psychological therapies for preventing seasonal affective disorder. Cochrane Database of Systematic Reviews, 5, CD011270. https://doi.org/10.1002/14651858.CD011270.pub3
Frandsen, T. B., Pareek, M., Hansen, J. P., & Nielsen, C. T. (2014). Vitamin D supplementation for treatment of seasonal affective symptoms in healthcare professionals: A double-blind randomised placebo-controlled trial. BMC Research Notes, 7, 528. https://doi.org/10.1186/1756-0500-7-528
Galima, S. V., Vogel, S. R., & Kowalski, A. W. (2020). Seasonal affective disorder: Common questions and answers. American Family Physician, 102(11), 668-672.
Gallin, P. F., Terman, M., Remé, C. E., Rafferty, B., Terman, J. S., & Burde, R. M. (1995). Ophthalmologic examination of patients with seasonal affective disorder, before and after bright light therapy. American Journal of Ophthalmology, 119(2), 202-210. https://doi.org/10.1016/s0002-9394(14)73874-7
Garbazza, C., & Benedetti, F. (2018). Genetic factors affecting seasonality, mood, and the circadian clock. Frontiers in Endocrinology, 9, 481. https://doi.org/10.3389/fendo.2018.00481
Gartlehner, G., Nussbaumer-Streit, B., Gaynes, B. N., Forneris, C. A., Morgan, L. C., Greenblatt, A., Wipplinger, J., Lux, L. J., Van Noord, M. G., & Winkler, D. (2019). Second-generation antidepressants for preventing seasonal affective disorder in adults. Cochrane Database of Systematic Reviews, 3, CD011268. https://doi.org/10.1002/14651858.CD011268.pub3
Geoffroy, P. A., Bellivier, F., Scott, J., & Etain, B. (2014). Seasonality and bipolar disorder: A systematic review, from admission rates to seasonality of symptoms. Journal of Affective Disorders, 168, 210-223. https://doi.org/10.1016/j.jad.2014.07.002
Glickman, G., Byrne, B., Pineda, C., Hauck, W. W., & Brainard, G. C. (2006). Light therapy for seasonal affective disorder with blue narrow-band light-emitting diodes (LEDs). Biological Psychiatry, 59(6), 502-507. https://doi.org/10.1016/j.biopsych.2005.07.006
Gloth, F. M., Alam, W., & Hollis, B. (1999). Vitamin D vs broad spectrum phototherapy in the treatment of seasonal affective disorder. The Journal of Nutrition, Health & Aging, 3(1), 5-7.
Goikolea, J. M., Colom, F., Martínez-Arán, A., Sánchez-Moreno, J., Giordano, A., Bulbena, A., & Vieta, E. (2007). Clinical and prognostic implications of seasonal pattern in bipolar disorder: A 10-year follow-up of 302 patients. Psychological Medicine, 37(11), 1595-1599. https://doi.org/10.1017/s0033291707000864
Golden, R. N., Gaynes, B. N., Ekstrom, R. D., Hamer, R. M., Jacobsen, F. M., Suppes, T., Wisner, K. L., & Nemeroff, C. B. (2005). The efficacy of light therapy in the treatment of mood disorders: A review and meta-analysis of the evidence. American Journal of Psychiatry, 162(4), 656-662. https://doi.org/10.1176/appi.ajp.162.4.656
Hansen, V., Skre, I., & Lund, E. (2008). What is this thing called "SAD"? A critique of the concept of seasonal affective disorder. Epidemiologia e Psichiatria Sociale, 17(2), 120-127. https://doi.org/10.1017/s1121189x00002815
Hjordt, L. V., Stenbæk, D. S., Ozenne, B., Mc Mahon, B., Hageman, I., Hasselbalch, S. G., & Knudsen, G. M. (2017). Season-independent cognitive deficits in seasonal affective disorder and their relation to depressive symptoms. Psychiatry Research, 257, 219-226. https://doi.org/10.1016/j.psychres.2017.07.056
Ho, K. W. D., Han, S., Nielsen, J. V., Jancic, D., Hing, B., Fiedorowicz, J., Weissman, M. M., Levinson, D. F., & Potash, J. B. (2018). Genome-wide association study of seasonal affective disorder. Translational Psychiatry, 8(1), 190. https://doi.org/10.1038/s41398-018-0246-z
Huibers, M. J., de Graaf, L. E., Peeters, F. P., & Arntz, A. (2010). Does the weather make us sad? Meteorological determinants of mood and depression in the general population. Psychiatry Research, 180(2-3), 143-146. https://doi.org/10.1016/j.psychres.2009.09.016
Kasper, S., Rogers, S. L., Yancey, A., Schulz, P. M., Skwerer, R. G., & Rosenthal, N. E. (1989). Phototherapy in individuals with and without subsyndromal seasonal affective disorder. Archives of General Psychiatry, 46(9), 837-844. https://doi.org/10.1001/archpsyc.1989.01810090079011
Kasper, S., Wehr, T. A., Bartko, J. J., Gaist, P. A., & Rosenthal, N. E. (1989). Epidemiological findings of seasonal changes in mood and behavior: A telephone survey of Montgomery County, Maryland. Archives of General Psychiatry, 46(9), 823-833. https://doi.org/10.1001/archpsyc.1989.01810090065010
Kerr, D. C. R., Shaman, J., Washburn, I. J., Vuchinich, S., Neppl, T. K., Capaldi, D. M., & Conger, R. D. (2013). Two longterm studies of seasonal variation in depressive symptoms among community participants. Journal of Affective Disorders, 151(3), 837-842. https://doi.org/10.1016/j.jad.2013.07.019
Kim, K., Kim, J., Jung, S., Kim, H.-W., Kim, H.-S., Son, E., Ko, D. S., Yoon, S., Kim, B. S., Kim, W. K., Lim, C., Kim, K., Lee, D., & Kim, Y. H. (2025). Global prevalence of seasonal affective disorder by latitude: A systematic review and meta-analysis. Journal of Affective Disorders, 390, 119807. https://doi.org/10.1016/j.jad.2025.119807
Kosanovic Rajacic, B., Sagud, M., Pivac, N., & Begic, D. (2023). Illuminating the way: The role of bright light therapy in the treatment of depression. Expert Review of Neurotherapeutics, 23(12), 1157-1171. https://doi.org/10.1080/14737175.2023.2273396
Kräuchi, K., & Wirz-Justice, A. (1988). The four seasons: Food intake frequency in seasonal affective disorder in the course of a year. Psychiatry Research, 25(3), 323-338. https://doi.org/10.1016/0165-1781(88)90102-3
Lam, R. W., & Levitan, R. D. (2000). Pathophysiology of seasonal affective disorder: A review. Journal of Psychiatry & Neuroscience, 25(5), 469-480.
Lam, R. W., Kennedy, S. H., Adams, C., Bahji, A., Beaulieu, S., Bhat, V., Blier, P., Blumberger, D. M., Brietzke, E., Chakrabarty, T., Do, A., Frey, B. N., Giacobbe, P., Gratzer, D., Grigoriadis, S., Habert, J., Husain, M. I., Ismail, Z., McGirr, A., ... Milev, R. V. (2024). Canadian Network for Mood and Anxiety Treatments (CANMAT) 2023 update on clinical guidelines for management of major depressive disorder in adults. The Canadian Journal of Psychiatry, 69(9), 641-687. https://doi.org/10.1177/07067437241245384
Lam, R. W., Lee, S. K., Tam, E. M., Grewal, A., & Yatham, L. N. (2001). An open trial of light therapy for women with seasonal affective disorder and comorbid bulimia nervosa. Journal of Clinical Psychiatry, 62(3), 164-168. https://doi.org/10.4088/jcp.v62n0305
Lam, R. W., Levitt, A. J., Levitan, R. D., Enns, M. W., Morehouse, R., Michalak, E. E., & Tam, E. M. (2006). The Can-SAD study: A randomized controlled trial of the effectiveness of light therapy and fluoxetine in patients with winter seasonal affective disorder. American Journal of Psychiatry, 163(5), 805-812. https://doi.org/10.1176/ajp.2006.163.5.805
Lam, R. W., Levitt, A. J., Levitan, R. D., Michalak, E. E., Cheung, A. H., Morehouse, R., Ramasubbu, R., Yatham, L. N., & Tam, E. M. (2016). Efficacy of bright light treatment, fluoxetine, and the combination in patients with nonseasonal major depressive disorder: A randomized clinical trial. JAMA Psychiatry, 73(1), 56-63. https://doi.org/10.1001/jamapsychiatry.2015.2235
Lambert, G. W., Reid, C., Kaye, D. M., Jennings, G. L., & Esler, M. D. (2002). Effect of sunlight and season on serotonin turnover in the brain. The Lancet, 360(9348), 1840-1842. https://doi.org/10.1016/s0140-6736(02)11737-5
Lavoie, M.-P., Lam, R. W., Bouchard, G., Sasseville, A., Charron, M.-C., Gagné, A.-M., Tremblay, P., Filteau, M.-J., & Hébert, M. (2009). Evidence of a biological effect of light therapy on the retina of patients with seasonal affective disorder. Biological Psychiatry, 66(3), 253-258. https://doi.org/10.1016/j.biopsych.2008.11.020
Lee, T. M. C., Blashko, C. A., Janzen, H. L., Paterson, J. G., & Chan, C. C. H. (1997). Pathophysiological mechanism of seasonal affective disorder. Journal of Affective Disorders, 46(1), 25-38. https://doi.org/10.1016/s0165-0327(97)00076-1
Lewy, A. J., Bauer, V. K., Cutler, N. L., & Sack, R. L. (1998). Melatonin treatment of winter depression: A pilot study. Psychiatry Research, 77(1), 57-61. https://doi.org/10.1016/s0165-1781(97)00128-5
Lewy, A. J., Bauer, V. K., Cutler, N. L., Sack, R. L., Ahmed, S., Thomas, K. H., Blood, M. L., & Jackson, J. M. (1998). Morning vs evening light treatment of patients with winter depression. Archives of General Psychiatry, 55(10), 890-896. https://doi.org/10.1001/archpsyc.55.10.890
Lewy, A. J., Kern, H. A., Rosenthal, N. E., & Wehr, T. A. (1982). Bright artificial light treatment of a manic-depressive patient with a seasonal mood cycle. American Journal of Psychiatry, 139(11), 1496-1498. https://doi.org/10.1176/ajp.139.11.1496
Lewy, A. J., Lefler, B. J., Emens, J. S., & Bauer, V. K. (2006). The circadian basis of winter depression. Proceedings of the National Academy of Sciences of the United States of America, 103(19), 7414-7419. https://doi.org/10.1073/pnas.0602425103
Lewy, A. J., Rough, J. N., Songer, J. B., Mishra, N., Yuhas, K., & Emens, J. S. (2007). The phase shift hypothesis for the circadian component of winter depression. Dialogues in Clinical Neuroscience, 9(3), 291-300. https://doi.org/10.31887/DCNS.2007.9.3/alewy
Lewy, A. J., Sack, R. L., Miller, L. S., & Hoban, T. M. (1987). Antidepressant and circadian phase-shifting effects of light. Science, 235(4786), 352-354. https://doi.org/10.1126/science.3798117
Lewy, A. J., Wehr, T. A., Goodwin, F. K., Newsome, D. A., & Markey, S. P. (1980). Light suppresses melatonin secretion in humans. Science, 210(4475), 1267-1269. https://doi.org/10.1126/science.7434030
LoBello, S. G. (2017). The validity of major depression with seasonal pattern: Reply to Young (2017). Clinical Psychological Science, 5(4), 755-757. https://doi.org/10.1177/2167702617702420
Lukmanji, A., Williams, J. V. A., Bulloch, A. G. M., Bhattarai, A., & Patten, S. B. (2019). Seasonal variation in symptoms of depression: A Canadian population based study. Journal of Affective Disorders, 255, 142-149. https://doi.org/10.1016/j.jad.2019.05.040
Lund, E., & Hansen, V. (2001). Responses to the Seasonal Pattern Assessment Questionnaire in different seasons. American Journal of Psychiatry, 158(2), 316-318. https://doi.org/10.1176/appi.ajp.158.2.316
Lyall, L. M., Wyse, C. A., Celis-Morales, C. A., Lyall, D. M., Cullen, B., Mackay, D., Ward, J., Graham, N., Strawbridge, R. J., Gill, J. M. R., Ferguson, A., Bailey, M. E. S., Pell, J. P., Curtis, A. M., & Smith, D. J. (2018). Seasonality of depressive symptoms in women but not in men: A cross-sectional study in the UK Biobank cohort. Journal of Affective Disorders, 229, 296-305. https://doi.org/10.1016/j.jad.2017.12.106
Madden, P. A. F., Heath, A. C., Rosenthal, N. E., & Martin, N. G. (1996). Seasonal changes in mood and behavior: The role of genetic factors. Archives of General Psychiatry, 53(1), 47-55. https://doi.org/10.1001/archpsyc.1996.01830010049008
Magnusson, A., & Axelsson, J. (1993). The prevalence of seasonal affective disorder is low among descendants of Icelandic emigrants in Canada. Archives of General Psychiatry, 50(12), 947-951. https://doi.org/10.1001/archpsyc.1993.01820240031004
Magnusson, A., & Boivin, D. (2003). Seasonal affective disorder: An overview. Chronobiology International, 20(2), 189-207. https://doi.org/10.1081/CBI-120019310
Magnússon, A., & Stefánsson, J. G. (1993). Prevalence of seasonal affective disorder in Iceland. Archives of General Psychiatry, 50(12), 941-946. https://doi.org/10.1001/archpsyc.1993.01820240025002
Magnusson, A. (1996). Validation of the Seasonal Pattern Assessment Questionnaire (SPAQ). Journal of Affective Disorders, 40(3), 121-129. https://doi.org/10.1016/0165-0327(96)00036-5
Magnusson, A. (2000). An overview of epidemiological studies on seasonal affective disorder. Acta Psychiatrica Scandinavica, 101(3), 176-184. https://doi.org/10.1034/j.1600-0447.2000.101003176.x
Majumdar, J., Rangaprakash, D., & Deshpande, G. (2025). Effects of seasonal factors on brain function: Systematic review and future perspectives. iScience, 28(12), 114066. https://doi.org/10.1016/j.isci.2025.114066
Mårtensson, B., Pettersson, A., Berglund, L., & Ekselius, L. (2015). Bright white light therapy in depression: A critical review of the evidence. Journal of Affective Disorders, 182, 1-7. https://doi.org/10.1016/j.jad.2015.04.013
Mc Mahon, B., Andersen, S. B., Madsen, M. K., Hjordt, L. V., Hageman, I., Dam, H., Svarer, C., da Cunha-Bang, S., Baaré, W., Madsen, J., Hasholt, L., Holst, K., Frokjaer, V. G., & Knudsen, G. M. (2016). Seasonal difference in brain serotonin transporter binding predicts symptom severity in patients with seasonal affective disorder. Brain, 139(5), 1605-1614. https://doi.org/10.1093/brain/aww043
Mc Mahon, B., Nørgaard, M., Svarer, C., Andersen, S. B., Madsen, M. K., Baaré, W. F. C., Madsen, J., Frokjaer, V. G., & Knudsen, G. M. (2018). Seasonality-resilient individuals downregulate their cerebral 5-HT transporter binding in winter: A longitudinal combined 11C-DASB and 11C-SB207145 PET study. European Neuropsychopharmacology, 28(10), 1151-1160. https://doi.org/10.1016/j.euroneuro.2018.06.004
Meesters, Y., & Gordijn, M. C. M. (2016). Seasonal affective disorder, winter type: Current insights and treatment options. Psychology Research and Behavior Management, 9, 317-327. https://doi.org/10.2147/PRBM.S114906
Meesters, Y., Winthorst, W. H., Duijzer, W. B., & Hommes, V. (2016). The effects of low-intensity narrow-band blue-light treatment compared to bright white-light treatment in sub-syndromal seasonal affective disorder. BMC Psychiatry, 16, 27. https://doi.org/10.1186/s12888-016-0729-5
Melrose, S. (2015). Seasonal affective disorder: An overview of assessment and treatment approaches. Depression Research and Treatment, 2015, 178564. https://doi.org/10.1155/2015/178564
Menculini, G., Verdolini, N., Murru, A., Pacchiarotti, I., Volpe, U., Cervino, A., Steardo, L., Moretti, P., Vieta, E., & Tortorella, A. (2018). Depressive mood and circadian rhythms disturbances as outcomes of seasonal affective disorder treatment: A systematic review. Journal of Affective Disorders, 241, 608-626. https://doi.org/10.1016/j.jad.2018.08.071
Mersch, P. P. A., Middendorp, H. M., Bouhuys, A. L., Beersma, D. G. M., & van den Hoofdakker, R. H. (1999). Seasonal affective disorder and latitude: A review of the literature. Journal of Affective Disorders, 53(1), 35-48. https://doi.org/10.1016/S0165-0327(98)00097-4
Mersch, P. P. A., Vastenburg, N. C., Meesters, Y., Bouhuys, A. L., Beersma, D. G. M., van den Hoofdakker, R. H., & den Boer, J. A. (2004). The reliability and validity of the Seasonal Pattern Assessment Questionnaire: A comparison between patient groups. Journal of Affective Disorders, 80(2-3), 209-219. https://doi.org/10.1016/S0165-0327(03)00114-9
Michalak, E. E., Murray, G., Wilkinson, C., Dowrick, C., & Lam, R. W. (2007). A pilot study of adherence with light treatment for seasonal affective disorder. Psychiatry Research, 149(1-3), 315-320. https://doi.org/10.1016/j.psychres.2006.05.005
Michalon, M., Eskes, G. A., & Mate-Kole, C. C. (1997). Effects of light therapy on neuropsychological function and mood in seasonal affective disorder. Journal of Psychiatry and Neuroscience, 22(1), 19-28.
Mikola, T., Marx, W., Lane, M. M., Hockey, M., Loughman, A., Rajapolvi, S., Rocks, T., O'Neil, A., Mischoulon, D., Valkonen-Korhonen, M., Lehto, S. M., & Ruusunen, A. (2023). The effect of vitamin D supplementation on depressive symptoms in adults: A systematic review and meta-analysis of randomized controlled trials. Critical Reviews in Food Science and Nutrition, 63(33), 11784-11801. https://doi.org/10.1080/10408398.2022.2096560
Modell, J. G., Rosenthal, N. E., Harriett, A. E., Krishen, A., Asgharian, A., Foster, V. J., Metz, A., Rockett, C. B., & Wightman, D. S. (2005). Seasonal affective disorder and its prevention by anticipatory treatment with bupropion XL. Biological Psychiatry, 58(8), 658-667. https://doi.org/10.1016/j.biopsych.2005.07.021
Molin, J., Mellerup, E., Bolwig, T., Scheike, T., & Dam, H. (1996). The influence of climate on development of winter depression. Journal of Affective Disorders, 37(2-3), 151-155. https://doi.org/10.1016/0165-0327(95)00090-9
Morales-Muñoz, I., Koskinen, S., & Partonen, T. (2017). Seasonal affective disorder and alcohol abuse disorder in a population-based study. Psychiatry Research, 253, 91-98. https://doi.org/10.1016/j.psychres.2017.03.029
Moscovitch, A., Blashko, C. A., Eagles, J. M., Darcourt, G., Thompson, C., Kasper, S., & Lane, R. M. (2004). A placebo-controlled study of sertraline in the treatment of outpatients with seasonal affective disorder. Psychopharmacology, 171(4), 390-397. https://doi.org/10.1007/s00213-003-1594-8
Murray, G., Allen, N. B., & Trinder, J. (2003). Seasonality and circadian phase delay: Prospective evidence that winter lowering of mood is associated with a shift towards Eveningness. Journal of Affective Disorders, 76(1-3), 15-22. https://doi.org/10.1016/S0165-0327(02)00059-9
Murray, G., Michalak, E. E., Levitt, A. J., Levitan, R. D., Enns, M. W., Morehouse, R., & Lam, R. W. (2006). O sweet spot where art thou? Light treatment of seasonal affective disorder and the circadian time of sleep. Journal of Affective Disorders, 90(2-3), 227-231. https://doi.org/10.1016/j.jad.2005.10.010
Neumeister, A., Praschak-Rieder, N., Besselmann, B., Rao, M. L., Glück, J., & Kasper, S. (1997). Effects of tryptophan depletion on drug-free patients with seasonal affective disorder during a stable response to bright light therapy. Archives of General Psychiatry, 54(2), 133-138. https://doi.org/10.1001/archpsyc.1997.01830140043008
Neumeister, A., Turner, E. H., Matthews, J. R., Postolache, T. T., Barnett, R. L., Rauh, M., Vetticad, R. G., Kasper, S., & Rosenthal, N. E. (1998). Effects of tryptophan depletion vs catecholamine depletion in patients with seasonal affective disorder in remission with light therapy. Archives of General Psychiatry, 55(6), 524-530. https://doi.org/10.1001/archpsyc.55.6.524
Nussbaumer-Streit, B., Forneris, C. A., Morgan, L. C., Van Noord, M. G., Gaynes, B. N., Greenblatt, A., Wipplinger, J., Lux, L. J., Winkler, D., & Gartlehner, G. (2019). Light therapy for preventing seasonal affective disorder. Cochrane Database of Systematic Reviews, 2019(3), CD011269. https://doi.org/10.1002/14651858.CD011269.pub3
Nussbaumer-Streit, B., Pjrek, E., Kien, C., Gartlehner, G., Bartova, L., Friedrich, M.-E., Kasper, S., & Winkler, D. (2018). Implementing prevention of seasonal affective disorder from patients' and physicians' perspectives – a qualitative study. BMC Psychiatry, 18(1), 372. https://doi.org/10.1186/s12888-018-1951-0
Nussbaumer-Streit, B., Thaler, K., Chapman, A., Probst, T., Winkler, D., Sönnichsen, A., Gaynes, B. N., & Gartlehner, G. (2021). Second-generation antidepressants for treatment of seasonal affective disorder. Cochrane Database of Systematic Reviews, 2021(3), CD008591. https://doi.org/10.1002/14651858.CD008591.pub3
Palinkas, L. A., & Suedfeld, P. (2008). Psychological effects of polar expeditions. The Lancet, 371(9607), 153-163. https://doi.org/10.1016/s0140-6736(07)61056-3
Partonen, T., & Lönnqvist, J. (1998). Seasonal affective disorder. The Lancet, 352(9137), 1369-1374. https://doi.org/10.1016/s0140-6736(98)01015-0
Partonen, T., Treutlein, J., Alpman, A., Frank, J., Johansson, C., Depner, M., Aron, L., Rietschel, M., Wellek, S., Soronen, P., et al. (2007). Three circadian clock genes Per2, Arntl, and Npas2 contribute to winter depression. Annals of Medicine, 39(3), 229-238. https://doi.org/10.1080/07853890701278795
Pendse, B. P., Engström, G., & Träskman-Bendz, L. (2004). Psychopathology of seasonal affective disorder patients in comparison with major depression patients who have attempted suicide. The Journal of Clinical Psychiatry, 65(3), 322-327. https://doi.org/10.4088/jcp.v65n0306
Perera, S., Eisen, R., Bhatt, M., Bhatnagar, N., de Souza, R., Thabane, L., & Samaan, Z. (2016). Light therapy for non-seasonal depression: Systematic review and meta-analysis. BJPsych Open, 2(2), 116-126. https://doi.org/10.1192/bjpo.bp.115.001610
Pinchasov, B. B., Shurgaja, A. M., Grischin, O. V., & Putilov, A. A. (2000). Mood and energy regulation in seasonal and non-seasonal depression before and after midday treatment with physical exercise or bright light. Psychiatry Research, 94(1), 29-42. https://doi.org/10.1016/s0165-1781(00)00138-4
Pjrek, E., Friedrich, M.-E., Cambioli, L., Dold, M., Jäger, F., Komorowski, A., Lanzenberger, R., Kasper, S., & Winkler, D. (2020). The efficacy of light therapy in the treatment of seasonal affective disorder: A meta-analysis of randomized controlled trials. Psychotherapy and Psychosomatics, 89(1), 17-24. https://doi.org/10.1159/000502891
Pjrek, E., Winkler, D., & Kasper, S. (2005). Pharmacotherapy of seasonal affective disorder. CNS Spectrums, 10(8), 664-669. https://doi.org/10.1017/s1092852900019623
Pjrek, E., Winkler, D., Konstantinidis, A., Willeit, M., Praschak-Rieder, N., & Kasper, S. (2007). Agomelatine in the treatment of seasonal affective disorder. Psychopharmacology, 190(4), 575-579. https://doi.org/10.1007/s00213-006-0645-3
Postolache, T. T., Mortensen, P. B., Tonelli, L. H., Jiao, X., Frangakis, C., Soriano, J. J., & Qin, P. (2010). Seasonal spring peaks of suicide in victims with and without prior history of hospitalization for mood disorders. Journal of Affective Disorders, 121(1-2), 88-93. https://doi.org/10.1016/j.jad.2009.05.015
Praschak-Rieder, N., Willeit, M., Wilson, A. A., Houle, S., & Meyer, J. H. (2008). Seasonal variation in human brain serotonin transporter binding. Archives of General Psychiatry, 65(9), 1072-1078. https://doi.org/10.1001/archpsyc.65.9.1072
Raheja, S. K., King, E. A., & Thompson, C. (1996). The Seasonal Pattern Assessment Questionnaire for identifying seasonal affective disorders. Journal of Affective Disorders, 41(3), 193-199. https://doi.org/10.1016/s0165-0327(96)00087-0
Rastad, C., Wetterberg, L., & Martin, C. (2017). Patients' experience of winter depression and light room treatment. Psychiatry Journal, 2017, Article 6867957. https://doi.org/10.1155/2017/6867957
Roecklein, K. A., Carney, C. E., Wong, P. M., Steiner, J. L., Hasler, B. P., & Franzen, P. L. (2013). The role of beliefs and attitudes about sleep in seasonal and nonseasonal mood disorder, and nondepressed controls. Journal of Affective Disorders, 150(2), 466-473. https://doi.org/10.1016/j.jad.2013.04.041
Roecklein, K. A., Rohan, K. J., Duncan, W. C., Rollag, M. D., Rosenthal, N. E., Lipsky, R. H., & Provencio, I. (2009). A missense variant (P10L) of the melanopsin (OPN4) gene in seasonal affective disorder. Journal of Affective Disorders, 114(1-3), 279-285. https://doi.org/10.1016/j.jad.2008.08.005
Roecklein, K. A., Wong, P. M., Miller, M. A., Donofry, S. D., Kamarck, M. L., & Brainard, G. C. (2013). Melanopsin, photosensitive ganglion cells, and seasonal affective disorder. Neuroscience and Biobehavioral Reviews, 37(3), 229-239. https://doi.org/10.1016/j.neubiorev.2012.12.009
Roecklein, K., Wong, P., Ernecoff, N., Miller, M., Donofry, S., Kamarck, M., Wood-Vasey, W. M., & Franzen, P. (2013). The post illumination pupil response is reduced in seasonal affective disorder. Psychiatry Research, 210(1), 150-158. https://doi.org/10.1016/j.psychres.2013.05.023
Rohan, K. J., Burt, K. B., Camuso, J., Perez, J., & Meyerhoff, J. (2020). Applying experimental therapeutics to examine cognitive and chronological vulnerabilities as mediators of acute outcomes in cognitive-behavioral therapy and light therapy for winter depression. Journal of Consulting and Clinical Psychology, 88(8), 786-797. https://doi.org/10.1037/ccp0000499
Rohan, K. J., Burt, K. B., Norton, R. J., Perez, J., Iyiewuare, P., & Terman, J. M. (2023). Change in seasonal beliefs mediates the durability advantage of cognitive-behavioral therapy over light therapy for winter depression. Behavior Therapy, 54(4), 682-695. https://doi.org/10.1016/j.beth.2022.11.002
Rohan, K. J., Mahon, J. N., Evans, M., Ho, S.-Y., Meyerhoff, J., Postolache, T. T., & Vacek, P. M. (2015). Randomized trial of cognitive-behavioral therapy versus light therapy for seasonal affective disorder: Acute outcomes. American Journal of Psychiatry, 172(9), 862-869. https://doi.org/10.1176/appi.ajp.2015.14101293
Rohan, K. J., Meyerhoff, J., Ho, S.-Y., Evans, M., Postolache, T. T., & Vacek, P. M. (2016). Outcomes one and two winters following cognitive-behavioral therapy or light therapy for seasonal affective disorder. American Journal of Psychiatry, 173(3), 244-251. https://doi.org/10.1176/appi.ajp.2015.15060773
Rohan, K. J., Nillni, Y. I., Mahon, J. N., Roecklein, K. A., Sitnikov, L., & Haaga, D. A. F. (2011). Cognitive vulnerability in moderate, mild, and low seasonality. Journal of Nervous and Mental Disease, 199(12), 961-970. https://doi.org/10.1097/NMD.0b013e3182392948
Rohan, K. J., Roecklein, K. A., & Haaga, D. A. F. (2009). Biological and psychological mechanisms of seasonal affective disorder: A review and integration. Current Psychiatry Reviews, 5(1), 37-47. https://doi.org/10.2174/157340009787315299
Rohan, K. J., Roecklein, K. A., Lacy, T. J., & Vacek, P. M. (2009). Winter depression recurrence one year after cognitive-behavioral therapy, light therapy, or combination treatment. Behavior Therapy, 40(3), 225-238. https://doi.org/10.1016/j.beth.2008.06.004
Rohan, K. J., Roecklein, K. A., Tierney Lindsey, K., Johnson, L. G., Lippy, R. D., Lacy, T. J., & Barton, F. B. (2007). A randomized controlled trial of cognitive-behavioral therapy, light therapy, and their combination for seasonal affective disorder. Journal of Consulting and Clinical Psychology, 75(3), 489-500. https://doi.org/10.1037/0022-006x.75.3.489
Rohan, K. J., Sigmon, S. T., & Dorhofer, D. M. (2003). Cognitive-behavioral factors in seasonal affective disorder. Journal of Consulting and Clinical Psychology, 71(1), 22-30. https://doi.org/10.1037/0022-006x.71.1.22
Rohan, K. J., Terman, J. M., Iyiewuare, P., Perez, J., Camuso, J. A., Postolache, T. T., DeSarno, M. J., & Vacek, P. M. (2024). Prospectively assessed summer mood status in major depression, recurrent with seasonal pattern: Evidence for SAD's construct validity. Journal of Affective Disorders, 349, 32-38. https://doi.org/10.1016/j.jad.2023.12.070
Rosen, L. N., Targum, S. D., Terman, M., Bryant, M. J., Hoffman, H., Kasper, S. F., Hamovit, J. R., Docherty, J. P., Welch, B., & Rosenthal, N. E. (1990). Prevalence of seasonal affective disorder at four latitudes. Psychiatry Research, 31(2), 131-144. https://doi.org/10.1016/0165-1781(90)90116-M
Rosenthal, N. E., Genhart, M. J., Caballero, B., Jacobsen, F. M., Skwerer, R. G., Coursey, R. D., Rogers, S., & Spring, B. J. (1989). Psychobiological effects of carbohydrate- and protein-rich meals in patients with seasonal affective disorder and normal controls. Biological Psychiatry, 25(8), 1029-1040. https://doi.org/10.1016/0006-3223(89)90291-6
Rosenthal, N. E., Sack, D. A., Gillin, J. C., Lewy, A. J., Goodwin, F. K., Davenport, Y., Mueller, P. S., Newsome, D. A., & Wehr, T. A. (1984). Seasonal affective disorder: A description of the syndrome and preliminary findings with light therapy. Archives of General Psychiatry, 41(1), 72-80. https://doi.org/10.1001/archpsyc.1984.01790120076010
Rybakowski, J. (2024). Forty years of seasonal affective disorder. Psychiatria Polska, 58(5), 747-759. https://doi.org/10.12740/pp/186721
Saheer, T. B., Lien, L., Hauff, E., & Kumar, B. N. (2013). Ethnic differences in seasonal affective disorder and associated factors among five immigrant groups in Norway. Journal of Affective Disorders, 151(1), 237-242. https://doi.org/10.1016/j.jad.2013.05.086
Sandkühler, J. F., Brochhagen, S., Rohde, P., Muscheidt, R. C., Grömer, T. W., Müller, H., & Brauner, J. M. (2022). 100,000 lumens to treat seasonal affective disorder: A proof of concept RCT of Bright, whole-ROom, All-Day (BROAD) light therapy. Depression and Anxiety, 39(12), 760-769. https://doi.org/10.1002/da.23281
Schwartz, P. J., Brown, C., Wehr, T. A., & Rosenthal, N. E. (1996). Winter seasonal affective disorder: A follow-up study of the first 59 patients of the National Institute of Mental Health Seasonal Studies Program. American Journal of Psychiatry, 153(8), 1028-1036. https://doi.org/10.1176/ajp.153.8.1028
Sigmon, S. T., Pells, J. J., Schartel, J. G., Hermann, B. A., Edenfield, T. M., LaMattina, S. M., Boulard, N. E., & Whitcomb-Smith, S. R. (2007). Stress reactivity and coping in seasonal and nonseasonal depression. Behaviour Research and Therapy, 45(5), 965-975. https://doi.org/10.1016/j.brat.2006.07.016
Sit, D. K., McGowan, J., Wiltrout, C., Diler, R. S., Dills, J. J., Luther, J., Yang, A., Ciolino, J. D., Seltman, H., Wisniewski, S. R., Terman, M., & Wisner, K. L. (2018). Adjunctive bright light therapy for bipolar depression: A randomized double-blind placebo-controlled trial. American Journal of Psychiatry, 175(2), 131-139. https://doi.org/10.1176/appi.ajp.2017.16101200
Spinks, H., & Dalgleish, T. (2001). Attentional processing and levels of symptomatology in Seasonal Affective Disorder (SAD): A preliminary longitudinal study. Journal of Affective Disorders, 62(3), 229-232. https://doi.org/10.1016/s0165-0327(00)00155-5
Strong, R. E., Marchant, B. K., Reimherr, F. W., Williams, E., Soni, P., & Mestas, R. (2009). Narrow-band blue-light treatment of seasonal affective disorder in adults and the influence of additional nonseasonal symptoms. Depression and Anxiety, 26(3), 273-278. https://doi.org/10.1002/da.20538
Swedo, S. E., Pleeter, J. D., Richter, D. M., Hoffman, C. L., Allen, A. J., Hamburger, S. D., Turner, E. H., Yamada, E. M., & Rosenthal, N. E. (1995). Rates of seasonal affective disorder in children and adolescents. American Journal of Psychiatry, 152(7), 1016-1019. https://doi.org/10.1176/ajp.152.7.1016
Takeshima, M., Utsumi, T., Aoki, Y., Wang, Z., Suzuki, M., Okajima, I., Watanabe, N., Watanabe, K., & Takaesu, Y. (2020). Efficacy and safety of bright light therapy for manic and depressive symptoms in patients with bipolar disorder: A systematic review and meta-analysis. Psychiatry and Clinical Neurosciences, 74(4), 247-256. https://doi.org/10.1111/pcn.12976
Tam, E. M., Lam, R. W., Robertson, H. A., Stewart, J. N., Yatham, L. N., & Zis, A. P. (1997). Atypical depressive symptoms in seasonal and non-seasonal mood disorders. Journal of Affective Disorders, 44(1), 39-44. https://doi.org/10.1016/s0165-0327(97)01447-x
Terman, J. S., Terman, M., Lo, E.-S., & Cooper, T. B. (2001). Circadian time of morning light administration and therapeutic response in winter depression. Archives of General Psychiatry, 58(1), 69-75. https://doi.org/10.1001/archpsyc.58.1.69
Terman, M., & Terman, J. S. (2005). Light therapy for seasonal and nonseasonal depression: Efficacy, protocol, safety, and side effects. CNS Spectrums, 10(8), 647-663. https://doi.org/10.1017/s1092852900019611
Terman, M., & Terman, J. S. (2006). Controlled trial of naturalistic dawn simulation and negative air ionization for seasonal affective disorder. American Journal of Psychiatry, 163(12), 2126-2133. https://doi.org/10.1176/ajp.2006.163.12.2126
Terman, M., Terman, J. S., & Ross, D. C. (1998). A controlled trial of timed bright light and negative air ionization for treatment of winter depression. Archives of General Psychiatry, 55(10), 875-882. https://doi.org/10.1001/archpsyc.55.10.875
Terman, M., Terman, J. S., Quitkin, F. M., McGrath, P. J., Stewart, J. W., & Rafferty, B. (1989). Light therapy for seasonal affective disorder: A review of efficacy. Neuropsychopharmacology, 2(1), 1-22. https://doi.org/10.1016/0893-133x(89)90002-x
Thompson, C., & Cowan, A. (2001). The Seasonal Health Questionnaire: A preliminary validation of a new instrument to screen for seasonal affective disorder. Journal of Affective Disorders, 64(1), 89-98. https://doi.org/10.1016/s0165-0327(00)00208-1
Traffanstedt, M. K., Mehta, S., & LoBello, S. G. (2016). Major depression with seasonal variation: Is it a valid construct? Clinical Psychological Science, 4(5), 825-834. https://doi.org/10.1177/2167702615615867
Tyrer, A. E., Levitan, R. D., Houle, S., Wilson, A. A., Nobrega, J. N., & Meyer, J. H. (2016). Increased seasonal variation in serotonin transporter binding in seasonal affective disorder. Neuropsychopharmacology, 41(10), 2447-2454. https://doi.org/10.1038/npp.2016.54
Vandewalle, G., Hébert, M., Beaulieu, C., Richard, L., Daneault, V., Garon, M.-L., Leblanc, J., Grandjean, D., Maquet, P., Schwartz, S., Dumont, M., Doyon, J., & Carrier, J. (2011). Abnormal hypothalamic response to light in seasonal affective disorder. Biological Psychiatry, 70(10), 954-961. https://doi.org/10.1016/j.biopsych.2011.06.022
Øverland, S., Woicik, W., Sikora, L., Whittaker, K., Heli, H., Skjelkvåle, F. S., Sivertsen, B., & Colman, I. (2019). Seasonality and symptoms of depression: A systematic review of the literature. Epidemiology and Psychiatric Sciences, 29, e31. https://doi.org/10.1017/s2045796019000209
Vyssoki, B., Kapusta, N. D., Praschak-Rieder, N., Dorffner, G., & Willeit, M. (2014). Direct effect of sunshine on suicide. JAMA Psychiatry, 71(11), 1231-1237. https://doi.org/10.1001/jamapsychiatry.2014.1198
Wehr, T. A., & Rosenthal, N. E. (1989). Seasonality and affective illness. American Journal of Psychiatry, 146(7), 829-839. https://doi.org/10.1176/ajp.146.7.829
Wehr, T. A. (1991). The durations of human melatonin secretion and sleep respond to changes in daylength (photoperiod). The Journal of Clinical Endocrinology & Metabolism, 73(6), 1276-1280. https://doi.org/10.1210/jcem-73-6-1276
Wehr, T. A., Duncan, W. C., Sher, L., Aeschbach, D., Schwartz, P. J., Turner, E. H., Postolache, T. T., & Rosenthal, N. E. (2001). A circadian signal of change of season in patients with seasonal affective disorder. Archives of General Psychiatry, 58(12), 1108-1114. https://doi.org/10.1001/archpsyc.58.12.1108
Wehr, T. A., Giesen, H. A., Schulz, P. M., Anderson, J. L., Joseph-Vanderpool, J. R., Kelly, K., Kasper, S., & Rosenthal, N. E. (1991). Contrasts between symptoms of summer depression and winter depression. Journal of Affective Disorders, 23(4), 173-183. https://doi.org/10.1016/0165-0327(91)90098-d
Wehr, T. A., Jacobsen, F. M., Sack, D. A., Arendt, J., Tamarkin, L., & Rosenthal, N. E. (1986). Phototherapy of seasonal affective disorder: Time of day and suppression of melatonin are not critical for antidepressant effects. Archives of General Psychiatry, 43(9), 870-875. https://doi.org/10.1001/archpsyc.1986.01800090060008
Wehr, T. A., Sack, D. A., & Rosenthal, N. E. (1987). Seasonal affective disorder with summer depression and winter hypomania. American Journal of Psychiatry, 144(12), 1602-1603. https://doi.org/10.1176/ajp.144.12.1602
Wescott, D. L., Franzen, P. L., Hasler, B. P., Miller, M. A., Soehner, A. M., Smagula, S. F., Wallace, M. L., Hall, M. H., & Roecklein, K. A. (2023). Elusive hypersomnolence in seasonal affective disorder: Actigraphic and self-reported sleep in and out of depressive episodes. Psychological Medicine, 53(4), 1313-1322. https://doi.org/10.1017/S003329172100283X
Wescott, D. L., Soehner, A. M., & Roecklein, K. A. (2020). Sleep in seasonal affective disorder. Current Opinion in Psychology, 34, 7-11. https://doi.org/10.1016/j.copsyc.2019.08.023
Westrin, Å., & Lam, R. W. (2007). Seasonal affective disorder: A clinical update. Annals of Clinical Psychiatry, 19(4), 239-246. https://doi.org/10.1080/10401230701653476
White, R. A., Azrael, D., Papadopoulos, F. C., Lambert, G. W., & Miller, M. (2015). Does suicide have a stronger association with seasonality than sunlight? BMJ Open, 5(6), e007403. https://doi.org/10.1136/bmjopen-2014-007403
Winthorst, W. H., Bos, E. H., Roest, A. M., & de Jonge, P. (2020). Seasonality of mood and affect in a large general population sample. PLoS ONE, 15(9), e0239033. https://doi.org/10.1371/journal.pone.0239033
Winthorst, W. H., Post, W. J., Meesters, Y., Penninx, B. W. J. H., & Nolen, W. A. (2011). Seasonality in depressive and anxiety symptoms among primary care patients and in patients with depressive and anxiety disorders; results from the Netherlands Study of Depression and Anxiety. BMC Psychiatry, 11, 198. https://doi.org/10.1186/1471-244X-11-198
Wirz-Justice, A. (2018). Seasonality in affective disorders. General and Comparative Endocrinology, 258, 244-249. https://doi.org/10.1016/j.ygcen.2017.07.010
Woo, J.-M., Okusaga, O., & Postolache, T. T. (2012). Seasonality of suicidal behavior. International Journal of Environmental Research and Public Health, 9(2), 531-547. https://doi.org/10.3390/ijerph9020531
Workman, J. L., & Nelson, R. J. (2011). Potential animal models of seasonal affective disorder. Neuroscience and Biobehavioral Reviews, 35(3), 669-679. https://doi.org/10.1016/j.neubiorev.2010.08.005
World Health Organization. (2024). Clinical descriptions and diagnostic requirements for ICD-11 mental, behavioural and neurodevelopmental disorders (CDDR). World Health Organization.
Øyane, N. M. F., Bjelland, I., Pallesen, S., Holsten, F., & Bjorvatn, B. (2008). Seasonality is associated with anxiety and depression: The Hordaland health study. Journal of Affective Disorders, 105(1-3), 147-155. https://doi.org/10.1016/j.jad.2007.05.002
Young, M. A. (2017). Does seasonal affective disorder exist? A commentary on Traffanstedt, Mehta, and LoBello (2016). Clinical Psychological Science, 5(4), 750-754. https://doi.org/10.1177/2167702616689086
Young, M. A., Blodgett, C., & Reardon, A. (2003). Measuring seasonality: Psychometric properties of the Seasonal Pattern Assessment Questionnaire and the Inventory for Seasonal Variation. Psychiatry Research, 117(1), 75-83. https://doi.org/10.1016/s0165-1781(02)00299-8
Young, M. A., Reardon, A., & Azam, O. (2008). Rumination and vegetative symptoms: A test of the dual vulnerability model of seasonal depression. Cognitive Therapy and Research, 32(4), 567-576. https://doi.org/10.1007/s10608-008-9184-z
Young, M. A., Watel, L. G., Lahmeyer, H. W., & Eastman, C. I. (1991). The temporal onset of individual symptoms in winter depression: Differentiating underlying mechanisms. Journal of Affective Disorders, 22(4), 191-197. https://doi.org/10.1016/0165-0327(91)90065-Z