Related Papers Menu

Seasonal Affective Disorder
Permission graciously given by the author to reproduce this paper:   

Seasonal Changes of Light and Their Possible Contribution to Seasonal Affective Disorder

Sabine Scoggins


This paper presents an overview of recent findings, which identify light as a key contributing factor in the etiology of winter seasonal affective disorder.  Current research proposes the central nervous system’s ability to respond to changes in light is under the direction of circadian rhythm clock genes.  Their input, in turn, is responsible for processes involved in physiological rhythmicity, including the stimulation of the uprachiasmatic nucleus through light-triggered electronic potentials transduced by the retinal photoreceptors.  The writer discusses how circadian clock gene aberrations adversely impact the strength of stimulation the suprachiasmatic nucleus receives from retinal photoreceptors in times of diminished light availability, such as is typical in the northern hemisphere during the fall and winter seasons, and its domino effect on mood.  This paper also describes findings regarding efficient assessment techniques and treatment options available to treat winter seasonal affective disorder. 


    Over the years observations from clinicians have initiated research into seasonal patterns that cause variations in light as possible contributing factors in the occurrence of a number of disorders (e.g. Bunney and Bunney, 2000; Ghadirian, Marini, Jabalpurwala, and Steiger, 1999; Parry and Newton, 2001).  Among these are disorders that entail depressive episodes, including Bipolar I, Bipolar II, and Major Depression (Diagnostic and Statistical Manual, 1994, fourth edition).   In a recent  study conducted in the Netherlands, seasonality was found to play a role in suicidal behavior, making suicide attempts more likely during certain hours in the day (van Houwelingen and Beersma, 2001). Other mood-related disorders, which seem to be exasperated by seasonal patterns, include post-partum depression and  premenstrual dysphoric disorder or PMDD (Parry and Newton, 2001).  Even eating disorders, specifically bulimia nervosa, can also be adversely affected by this factor (Ghadirian, Marini, Jabalpurwala, and Steiger, 1999). 

    This paper, therefore, presents an investigation into some of the current findings on the  neuropsychological mechanisms involved in the way seasonal light fluctuations appear to impact human behavior.  Because symptoms related to Seasonal Affective Disorder (SAD) are the identifier for seasonality (e.g. Bunney and Bunney, 2000), this writer has used research that focused on this disorder.  In addition, SAD occurs mainly during the darker months of the year, which in the northern hemisphere are the months of fall and winter (Willerman and Cohen, 1990).  Therefore, the research supporting this paper was primarily taken from studies which investigated winter seasonal affective disorder (W-SAD). 

Neurophysiological Mechanisms Involved in Seasonality

Light as a possible factor.   Researchers have investigated a number of possible neurophysiological mechanisms that might contribute to the occurrence of seasonally-triggered depressive episodes, including retinal sensitivity to light (Terman and Terman, 1999), molecular clock genes (Bunney and Bunney, 2000), social zeitgebers (Reid, Towell, and Golding, 2000) monoamine oxidase acitvity (Reichborn-Kjennerud, Lindaerde, and Oreland, 1996), and serotonin levels (Partonen and Loennqvist, 1998). 

    In many of these findings, exposure (e.g. Rowan and Sigmon, 2000) and sensitivity to light (e.g.  Graw, Recker, Sand, Kraechi, and Wirz-Justice,  1999; Lam, Tam, Yatham, Shiah, and Zis, 2001) have been implicated as possible factors in the development of W-SAD.   For instance, Terman and Terman (1999) compared retinal sensitivity to light in W-SAD patients with healthy control subjects.  Overall results indicated significant differences between the two groups during the reduced-light months (fall and winter), with subjects in the W-SAD group displaying retinal hypersensitivity when compared to subjects in the control group.  Their findings lead them to the conclusion that the degree of reactivity to light by the photoreceptors, specifically rods and cones, exerts a key role in the occurrence of W-SAD, as well as in the improvement through light treatment. 

   The importance of light as a factor in the etiology of W-SAD is again stressed in a study by Graw, Recker, Sand, Kraechi, and Wirz-Justice (1999).  The researcher compared 22 women diagnosed with W-SAD and 47 healthy women in regard to time spent outside during daylight hours, mood, alertness, and sleep.  Result indicated that women in the SAD group had spent significantly more time outside during the summer and, therefore, had received more exposure to light.  However, during the winter season there were no differences in light exposure between the groups, yet the women in the W-SAD group had developed a classical symptoms of the disorder.  Therefore, the authors concluded that rather than duration to light exposure, the crucial factor for the development of winter depression might instead be vulnerability under lower light conditions.  In other words, when compared to non-SAD women, SAD women may be more vulnerable to winter depression in lower light conditions and appear to need greater exposure to light to retain normal mood. 

    Moreover, results implicating sensitivity to light are supported by findings that have used light as a treatment (e.g. Parry and Newton, 2001).  In this study, early morning exposure to as little as 200 lux was found sufficient to suppress melatonin release in W-SAD subjects but not in the control group.  These findings suggest a cause-and-effect relationship between retinal sensitivity, overall exposure to light, and the development of W-SAD, in this cause through the prevention of melatonin suppression that occurred following the supplemental exposure to light in the early morning.

Light and ciracdian clock genes.  The ability of the retinal photoreceptors to appropriately react to the presence of light appears to represent a vital link in a chain of mechanisms that, according to Bunney and Bunney (2000), regulate physiological rhythmicity and, if disrupted, may adversely affect a number of functions, including mood, body temperature, sleep-wake cycles, hunger, thirst, alertness, and motivation (e.g. Durat, Foret, Touitou, and Benoit, 1996; Partonen and Loennqvist, 2000).   In their review of literature, Bunney and Bunney discuss the presence of clock genes in all organisms, which responds to environmental changes (e.g. light, noise level, temperature) by adjusting physiological mechanisms and behaviors accordingly – a process called entrainment, which can be defined as the advancing or delaying of rhythmicity in response to circadian rhythm cues (Albrecht et al., 1997, in Bunney and Bunney, 2000). 

    In humans, the most important circadian regulator is the suprochiasmatic nucleus (SCN) located above the optic chiasm in the anterior hypothalamus (Bunney and Bunney, 2000).  The cells of this nucleus are interconnected and create a rhythmic, autonomic pattern of neural stimulation resulting from the shutting down of fos protein production once certain levels have been exceeded and turning it back on once levels have fallen sufficiently.   Based on this information the researchers suggested that mutations of the clock gene interfere with protein production and thereby alter the speed of the circadian rhythm and period lengths.  Consequently, CNS structures, such as the SCN, which act under the direction of clock genes, could create the symptoms typically associated with SAD, such as sleep disturbances, increases in night-time temperature, as well as disturbances in the release of melatonin and cortisol.  Interestingly, symptoms in certain clinically depressed patients that indicate circadian rhythm aberrations, include changes in these areas –  temperature and melatonin secretion during sleep, shortened REM latency, and mood swings marked by depression in the morning and near euthymia by evening. 

    Moreover, current findings indicate that the SCN receives retinal impulses either directly via the retinohypothalmic tract, or indirectly via the geniculohypothalmic pathway, which it uses  for the perpetuation of circadian rhythmicity (Bunney and Bunney, 2000).  Therefore, an anomaly in retinal sensitivity to light  (Terman and Terman; 1999) could disrupt the type of input the SCN needs to function normally.   Light, on the other hand,  is translated into electrical potentials by photoreceptors as it falls onto the retina (Bernstein, Roy, Srull, Wickens, 1991; Carlson, 2001).  These potentials, in turn, stimulate the SCN.  Neural interactions between the SCN and pineal gland and hypothalamus, then, result in behavioral changes relative to hunger, sleep, night-time body temperature, cortisol, as well as the synthesis of melatonin and neuropeptides.  Consequently, abnormal entrainment of seasonal changes in light by the  SCN contribute to the symptoms typically associated with W-SAD, such as hypersomnia, hyperphagia, and cravings for carbohydrate-containing foods  (Bunney and Bunney, 2000; Willerman and Cohen, 1990). 

   Further evidence for the existence of a disruption in the pathways connecting the retina with the SCN was found in research that measured the release of melatonin during dim and bright light conditions in individuals with either bipolar, unipolar, or winter depression (Nathan, Burrows, and Norman, 1999).  Comparative analysis indicated that patients in the dim light bipolar and winter depression groups were very reactive as measured by their melatonin suppression and  favorable remission rate.   Participants in the unipolar depression group, however, showed no improvement of symptoms.  In regard to W-SAD, which typically shows a full-remission during the summer (DSM-IV, 1994, fourth edition), these results suggest the presence of an interactive mechanism that includes such structures as the hypothalamus, pineal gland, SCN, and retinal pathways, and initiates a phase-delay of the circadian rhythm as a result of a lack of electrical potentials traveling from the retinal photoreceptors to the SCN.  This mechanism did not exist in patients with major depressive disorders, whose melatonin suppression rate had remained stable following the exposure to either light.  In other words, the stimulation of the photoreceptors or lack thereof might adversely affect melatonin suppression in vulnerable individuals, who then respond with a seasonal pattern in their mood variability. 

    Additional support for the hypothesis that mood can be adversely affected by the circadian pacemaker comes from a case study by Koorengevel, Beersma, Gordijn, den Boer, and van den Hoofdakker (2001).   The researchers used a 120-hour forced desynchrony protocol that strictly controlled light exposure of a male W-SAD subject and measured the circadian pacemaker functioning by assessing core body temperature and mood.   Although results were considered tentative relative to W-SAD, in this particular subject they indicated a phase-delay shift of up to one hour for body temperature and two hours for mood. 

    Moreover, abnormal brain temperatures during sleep appear to yet further implicate the circadian clock functions in the context of W-SAD (Schwartz, et al., 2000).  In other words, when compared to healthy subjects, those with W-SAD show less of a decline in brain temperatures during the acute stage of the disorder, than do their healthy counterparts.  Not surprisingly, this trend returns to normal  during remission.  Interestingly, W-SAD patients also showed a significant increase in mean slow-wave activity and NREM episode, whereas the intraepisodic dynamics of slow-wave activity remained normal. 

   Further evidence of an association between light availability and mood can also be established in healthy subjects.  For instance, exposure to bright light has been associated with elevation in mood, body temperature, alertness, performance, and motivation as well as suppression of melatonin release (Daurat, Foret, Touitou, and Benoit, 1996).  These were some of the findings in an experimental study that exposed subjects to bright light during the night.  Interestingly, in the early morning hours bright light became ineffective and a reversal of functioning occurred, apparently associated with an increase in melatonin release.  In other words, the increase in melatonin release was associated with decreases in alertness, performance, motivation, and mood.  Again, these results seem congruent with other research on circadian rhythm related functions (e.g. (Nathan, Burrows, and Norman, 1999; Willerman and Cohen, 1990).  In this writer’s view, the eventual decline in functioning, temperature, and mood were a normal reaction of the brain’s need for sleep and did not contradict earlier findings that showed overall improvement and a suppression of melatonin release with exposure to bright light.   These findings also point out again that bright light plays a crucial role in these circadian rhythm-related physiological reactions. 

Summary.  Current findings suggest that retinal sensitivity to light presents a key factor in the development of W-SAD in vulnerable individuals.  Consequently, diminished availability of light, a typical scenario in northern regions, appears to fail providing enough stimulation for the retinal photoreceptors in vulnerable individuals, which in turn transmit a lower number of electrical potentials to the SCN.  Understimulation of the SCN appears to adversely affect the physiological functions directed by this nucleus, thereby adversely affecting mood. 


    The process of assessing clients for W-SAD should include an informal interview and the use of standardized assessment tools (Othmer and Othmer, 1994).  During the informal interview clients may present with symptoms that reflect a depressive episode, including feelings of sadness, irritability, generally reduced interest and ability to experience pleasure, lethargy, decline in the ability to concentrate, hopelessness, and low self-worth    (DSM-IV, 1994, fourth edition).  Although clients may already be diagnosed or present with bipolar or unipolar depression, they may additionally complain of symptoms that warrant a seasonal pattern specifier.  This particular specifier is warranted under following conditions:  for a minimum of the previous two years there have to have been two occurrences; onset and full remission of a major depressive episode have occurred twice around the same time during the year.  For winter depression onset is during the fall or winter and remission during the spring and summer.  The occurrence of symptoms have to be independent of seasonally-related psychosocial stressors, such as the threat of lay-offs for construction workers during the cold months.  Moreover, there has to be a major discrepancy between the number of nonseasonal and seasonal major depressive episodes during the person’s lifetime, with the latter outnumbering the former. 

    In addition to the above describe seasonal pattern, there are several dynamics which set this disorder apart from major depression.  For instance, W-SAD tends to be more common in northern regions and among younger to middle aged adults (Carlson, Butcher, and Mineka, 2000; Rowan and Sigmon, 2000).  Furthermore, it entails atypical vegetative symptoms, including excessive fatigue, hypersomnia, hyperphagia, and cravings for high-carbohydrate foods, such as pasta, chocolate, and donuts (Willerman and Cohen, 1990). 

    Once an informal assessment suggests symptoms to warrant a diagnosis of SAD, clinicians may want to use a structured interview or psychological test to verify their conclusion.  The Beck Depression Inventory is an example of an instrument that could be used to initially confirm the presence of depressive episodes (Beck, Rial, and Rickels, 1974a in Groth-Marnat, 1990).  In an effort to verify seasonality, clinicians might want to use the Seasonal Pattern Assessment Questionnaire (SPAQ) by Rosenthal (as cited in Magnusson, Friies, and Objordsmoen, 1997), which provides high internal consistency in those scales measuring seasonality.  Although less helpful in predicting future course of the disorder, the SPAQ has been found to quickly collect information concerning recent seasonal variations (Raheja, King, and Thompson, 1996).  At the same time, the test’s ability to successfully identify SAD was only at 57% and showed low test-retest reliability.  Consequently, clinicians might get more reliable results by using more than one instrument.  In a recent study by Thompson and Cowan (2001) during which the SPAQ was compared to a new instrument, the Seasonal Health Questionnaire (SHQ), the latter was found to provide more detailed information, whereas the former tended to produce false positives.  Because the SHQ has not been sufficiently field-tested, the use of both tests might render the most reliable results. 

Treatment Options

    Clients diagnosed with SAD have several treatment options at their disposal, one of which is the deliberate exposure to light (e.g. Perry and Newton, 2001).   Although some findings have suggested light exposure of as little as 200 lux to decrease melatonin release in patients with W-SAD, the standard amount of treatment for its efficacy in success ranges between 2,500 lux and 6,000 lux for two hours in the early morning hours for over a period of at least one to three weeks (Jacobson et al.,1987 and  Lewy, et al., 1998 in Bunney and Bunney, 2000; Pinchasova, Shurgajaa, Grischina, and Putilov, 2000; Rosenthal, 1993 in Bunney and Bunney, 2000). 

    However, not all findings supported the use of light therapy to be effective in the treatment of SAD.  For instance, Koorengevel et al. (2001) found no differences in effect on SAD patients after light treatment when compared with the placebo control group.  At the same time, treatment duration and timing should be considered when interpreting these results. For instance, although exposure to light was stronger (13,000 lux) it occurred later in the day – between 8:00 a.m. and 11:00 a.m. – and only lasted five days.  Moreover, exposure time might exert greater impact on remission than does brightness of light.   In fact, findings suggest early morning exposure (4:30 a.m. to 6:00 a.m.) of a dim light releasing as little as 250 lux for six weeks shows greater improvement of symptoms than does bright light and placebo – both of which had equal rates of improvement (Avery et al., 2001).   Interestingly, pretreatment exposure to natural sunlight had a positive impact on remission, as well as response rate. 

    In addition, clinicians should consider confounding factors that potentially contribute to the etiology of winter depression and adversely impact treatment efforts.  An example of a contributing factor was found in a recent study by Lam, Tam, Yatham, Shiah, and Zis (2001), who – contrary to the DSM-V (1994, fourth edition) of symptoms warranting the use of a seasonal specifier – differentiated between three types of seasonal winter depressive episodes:   traditional SAD symptoms; depressive episodes with a clear seasonal trend yet incomplete remission of symptoms during the summer; or symptoms that show a trend toward winter depression but do not fully meet criteria for SAD (sub-syndromal SAD).  Comparison of treatment efficacy with light showed differences in success rates between those groups, with patients in the sub-syndromal SAD group being most responsive, followed by the SAD subjects with a success rate of 66%, and those with incomplete summer remission producing the lowest remission rates.  In other words, treatment options have to be carefully matched with symptom presentation.  In addition, Reichborn-Kjennerud and Lindaerde in their 1996 study found differences in treatment success rate using light between patients with personality disorders and those without.  However, light exposure was only at 1,500 lux and at a later time during the morning.  The question this writer raises is whether light treatment application with a better track record, such as dim light at dawn, would have diminished those differences. 

    As previously mentioned, these differences in success rate point out that clients need to receive individually-tailored treatment.  For instance, clients who show resistance to bright light treatment should try the dawn dim light version (Avery et al., 2001).  This latter treatment option has shown greater success rate and therefore should be the one to start with, given clients are comfortable with the approach.  Furthermore, research findings have indicated SAD to be associated with low serotonin levels (Partonen and Loennqvist, 1998).  Medicinal approaches therefore include antidepressants that increase serotonin levels, such as reboxetine, a selective noradrenaline inhibitor (Hilger,  Willeit, Praschak-Rieder, Stastny, Neumeister, and Kasper, 2001).  Drug intervention is an efficient treatment of choice for patients who are uncomfortable with or unresponsive to light therapy. 

    Finally, activities that raise oxygen levels in the body have been found as effective as light treatment (Pinchasova, Shurgajaa, Grischina, and Putilov, 2000).  Therefore, clinicians have the option of suggesting cardiovascular exercise, light, drug treatment, or a combination of treatment choices to their clients.  In an effort to ensure treatment success, clinicians should employ the use of an instrument that measures therapy efficacy, such as the visual analogue scale (VAS) developed by in 1969 by Zealley and Aitiken (cited in Lingjaerde and Foreland, 1998).  This tool has shown good test-retest (r=0.96) and validity (r=0.85) scores in the assessment of W-SAD symptoms in patients, who had been in treatment for one to four weeks.  Assessment after initiation of treatment presents a vital means of ensuring therapy efficacy. 

    Finally, clinicians should encourage patients who develop W-SAD on an annual basis to consider prophylactic light treatment.   Exposure to either bright or infrared light has been found efficient in the prevention of the disorder in a 1999 experimental study conducted in the Netherlands by Meester, Beersma, Bouhuys, and van den Hoofdakker.  Research results suggested that exposing these patients to light treatment before typical symptom onset, such as in the month of October, can at least diminish symptoms of the disorder in a significant number of patients with recurring W-SAD.  However, these results also indicate that the use of a light visor might not be as efficient as more traditional approaches to light therapy.  In fact, this writer would hopes that additional research will soon determine whether the use of early dawn dim light treatment in the months of fall diminish or perhaps even prevent the onset of W-SAD. 


American Psychiatric Association.  (1994).  Diagnostic and statistical manual of mental disorders  (4th ed.).  Washington, DC:  Author.

Avery, D. H., Eder, D. N., Bolte, M. A., Hellekson, C. J., Dunner, D. L, Vitiello, M. V., & Prinz, P. N. (2001).  Dawn stimulation and bright light in the treatment of SAD:  a controlled study.  Biological Psychiatry, 50 (3), 205-216. 

Bernstein, D. A., Roy, E. J., Srull, T. K.,  & Wickens, C. D. (1991).  Psychology (2nd ed.).  Princeton, N. J.:  Houghton Mifflin Company.

Bunney, W. E. & Bunney, B. G. (2000).  Molecular clock genes in man and lower animals:  possible implications for circadian abnormalities in depression.  Neuropsychopharmacology, 22 (4), 335-345. 

Carlson, N. R. (2001).  Physiology of Behavior (7th ed.).  Needham Heights, MA:  A Pearson Education Company. 

Carson, R. C., Butcher, J. N., & Mineka, S.  (2000).  Abnormal psychology and modern life 11th ed.).  Boston, MA:  Pearson Education Company. 
Daurat, A., Foret, J., Touitou, Y., & Benoit, O. (1996).  Detrimental influence of bright light exposure on alertness, performance, and mood in the early morning.  Neurophysiol Clin, 26, 8-14. 

Ghadirian, A.-M., Marini, N., Jabalpurwala, S., & Steiger, H.  (1999).  Seasonal mood patterns in eating disorders.  General Hospital Psychiatry, 21 (5), 352-359. 

Graw, P., Recker, S., Sand, L., Kraechi, K., & Wirz-Justice, A. (1999)  Winter and summer outdoor light exposure in women with and without seasonal affective disorder.  Journal of Affective Disorders, 56 (2-3), 163-169. 

Groth-Marnat, G. (1990).  Handbook of psychological assessment (2nd ed.).  New York:  A Wiley-Interscience Publication. 

Hilger, E., Willeit, M., Praschak-Rieder, N., Stastny, J., Neumeister, A., & Kasper, S. (2001).  Reboxetine in seasonal affective disorder:  an open trial.  European Neuropsychopharmacology, 11 (1), 1-5. 

Koorengevel, K. M., Beersma, D. G.,  Gordijn, M. C., den Boer, J. A., & van den Hoofdakker, R. H. (2000).  Body temperature and mood variations during forced desynchronization in winter depression: a preliminary report.  Biological Psychiatry, 47 (4), 355-358. 

Koorengevel, K. M., Gordijn, M. C. Beersma, D. G., Meesters, Y, den Boer, J. A., van den Beersma, & Daan, S. (2001).  Extraocular light therapy in winter depression:  a double-blind placebo-controlled study.   Journal of Affective Disorders, 50 (9), 691-698.

van Houwelingen, C. A. & Beersma, D. G. (2001)  Seasonal changes in 24-h patterns of suicide rates:  a study on train suicides in The Netherlands.  Journal of Affective Disorders, 66 (2-3), 215-223

Lam, R.. W., Tam, E. M, Yatham, L. N., Shiah, I.-S., & Zis, A. P.  (2001).  Seasonal depression:  the dual vulnerability hypothesis revisited.  Journal of Affective Disorders, 63 (1-3), 123-132. 

Lingjaerde, O. & Foreland, A. R. (1998).  Direct assessment of improvement in winter depression with a visual analogue scale: high reliability and validity.  Psychiatric Research, 89 (3), 387-392. 

Magnusson, A., Friies, S., & Opjordsmoen, S. (1997).  Internal consistency of the Seasonal Pattern Assessment Questionnaire.  Journal of Affective Disorder, 42, 113-116. 

Meester, Y., Beersm, D. G., Bouhuys, A. L., & van den Hoofdakker, R. H. (1999).  Prophylactic treatment of seasonal affective disorder (SAD) by using light visors: bright white or infrared light?  Biological Psychiatry, 46 (2), 239-246.

Nathan, P. J., Burrows, G. D., & Norman, T. R. (1999).  Melatonin sensitivity to dim white light in affective disorders.  Neuropsychopharamcology, 21 (3), 408-413.

Othmer, E. & Othmer, S. C. (1994).  The clinical interview using DSM-IV Volume !:  Fundamentals.  Washington, D.C.:  American Psychiatric Press, Inc.

Parry, B. L. & Newton, R. P. (2001). Chronobiological basis of female-specific mood disorders.  Neuropsychopharmacology, 25 (5), 102-108. 

Partonen, T. & Loennqvist, J.  (1998).  Seasonal affective disorder.  The Lancet, 352 (9137), 1369-1474. 

Partonen, T. & Loennqvist, J. (2000).  Bright light improves vitality and alleviates distress in healthy people.  Journal of Affective Disorders, 53 (1-3), 55-61. 

Pinchasova, B. B., Shurgajaa, A. M., Grischina, 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. 

Raheja, S. K., King, E. A., & Thompson, C. (1996).  The seasonal pattern assessment questionnaire for assessing seasonal affective disorders.  Journal of Affective Disorders, 41, 193-199. 

Reichborn-Kjennerud, T. & Lindaerde, O.  (1996).  Response to light therapy in seasonal affective disorder:  personality disorders and temperament as predictors of outcome.  Journal of Affective Disorders, 41, 101-110.

Reichborn-Kjennerud, T., Lindaerde, O., & Oreland, L. (1996).  Platelet monamine oxides activity in patients with winter seasonal affective disorder.  Psychiatric Research, 62, 273-280. 

Reid, S., Towell, A. D., & Golding, J. F. (2000).  Seasonality, social zeitgebers and mood  variability in entrainment of  mood:  Implications for seasonal affective disorder.   .     Journal of Affective Disorders, 59 (1), 47-54.

Rowan, K. J. & Sigmon, S. T. (2000).  Seasonal mood patterns in a northeastern college sample.     Journal of Affective Disorders, 59 (2), 85-96.

Schwartz, P. J., Rosenthal, N. E., Kajimura, N., Han, L, Turner, E. H., Bender, C., & Wehr, T. A.  (2000).  Ultradian oscillations in cranial thermoregulation and electroencephalographic slow-wave activity during sleep are abnormal in humans with annual winter depression.  Brain Research, 866 (1-2), 152-167. 

Terman, J. S. & Terman, M. (2000).  Photopic and scotopic light detection in patients with seasonal affective disorder and control subjects.  Journal of Affective Disorders, 59 (1), 47-54. 

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. 

Willerman, L. & Cohen, D. B. (1990).  Psychopathology.  New-York:  McGraw-Hill.