This site is intended for health professionals only

Seasonal affective disorder: treating winter depression

John M Eagles
MB ChB MPhil FRCPsych
Consultant Psychiatrist
Royal Cornhill Hospital

The concept of seasonal affective disorder (SAD), in which winter depression is the most common presentation, was established as a respectable diagnostic entity in 1984.(1) Clinical researchers described 29 patients who experienced recurrent seasonal depressions and reported the effects on 11 patients of bright artificial light. While some doctors remain sceptical of the diagnosis, evidence accrues on the epidemiology, biology and response to treatment in winter depression.

The hallmark of winter depression is recurrent episodes that arise in autumn and winter and remit in the spring. Some clinicians regard a complete summer remission as an important diagnostic criterion; certainly, in my experience, people who report normality in the summer (even to the extent of slight overactivity, exuberance and extraversion) tend to fare better with winter treatment. As in nonseasonal depression, sufferers experience low mood (often worse in the morning), anergia, low libido, anxiety, irritability and social withdrawal. The following symptoms, however, do not tend to be found in nonseasonal depression but characterise the presentation of SAD in winter: hypersomnia; daytime somnolence, often peaking in the late afternoon; increased appetite, often for carbohydrates and chocolate; and weight gain.

Two recent UK community studies in Aberdeen and north Wales,(2,3) both using similarly tight diagnostic criteria, found prevalence rates in the adult population of 3.5% and 2.4% respectively. These figures are likely to reflect the clinically significant end of a spectrum of symptom severity. The majority of people in the UK will experience some of the winter symptoms - notably lower mood and energy, increased sleep requirement and dietary changes - to some degree. For most of us, no treatment is merited, although nearer the clinical severity end of the spectrum, whether or not the "disorder" should be recognised and "medicalised" is more contentious. Kasper et al have coined the term "subsyndromal SAD" to describe a milder version of the full syndrome on the basis of a putative response to light therapy, but such findings await confirmation.(4)

While SAD exists (and can be treated successfully) in children, its prevalence increases significantly at puberty, especially among girls, so that there is quite a marked preponderance of female adult sufferers. This female preponderance does not endure into old age, probably indicating that aetiology is partly linked to female reproductive hormones.
After screening a large number of patients in primary care, our research group compared SAD sufferers with controls who had little or no change in seasonal wellbeing. The SAD patients were heavy users of healthcare services, with more consultations, more tests and investigations, more prescriptions and more specialist referrals.(5) Symptoms were often diverse; as well as psychiatric presentations, SAD sufferers presented more often with  gastrointestinal, urinary and musculoskeletal symptoms. They presented more often with pain and with infections and received more analgesics and more antibiotics. They had more full blood counts and more thyroid function tests. As well as an excess of psychiatric referrals, they saw more physiotherapists and more gynaecologists. In short, they presented with, were investigated for and were treated for a wider variety of somatic complaints.
The diagnosis of SAD had rarely been considered in these patients, probably due to the frequency of somatised presentations that are known to obscure diagnoses of affective disorders. Along with this "somatising camouflage", however, the typical recurring symptoms of SAD can, it is to be hoped, be elicited.
It is an interesting paradigm to think of winter depression as a form of attenuated hibernation that may have conferred evolutionary advantages upon our cave-dwelling ancestors. The hormone melatonin almost certainly mediates seasonal changes in our sleep-wake cycle, and SAD patients experience circadian phase delay:  they are hypersomnolent in the morning and livelier in the evening, as in those experiencing jetlag after flying west to east. This phase-shift hypothesis is thought to underlie the effectiveness of morning-light therapy.

For mildly affected people, self-help should be advocated. The most important elements of self-help comprise regular exposure to natural daylight and remaining physically active during the winter. Walking regularly outside combines these therapeutic elements. A positive approach to the time-limited nature of the symptoms is usually helpful. For more severely affected people, self-help measures should be coupled with antidepressants and/or light therapy. Patient preference, including availability and convenience, will often influence the choice between antidepressants and light therapy. Otherwise, it is logical to start with an antidepressant when "nonseasonal" symptoms (such as weight loss and early-morning wakening) predominate, and with light therapy when symptoms of hypersomnia and carbohydrate craving are present. The following is a treatment summary:


  • Nonsedative drugs are usually required.
  • Specific serotonin reuptake inhibitors (SSRIs) are a good first choice.
  • Of the SSRIs, paroxetine may induce more weight gain.
  • Venlafaxine is a sensible second choice, particularly since its noradrenergic reuptake inhibiting (NARI) properties can induce alertness and enhance energy.
  • Reboxetine is a specific NARI with some preliminary evidence of effectiveness in SAD.

Light therapy

  • Usually administered by a lightbox that emits bright light without a UV component.
  • Most effective when used early in the morning.
  • Dawn-simulating alarm clocks may also be useful and can be more convenient.(6)


  1. Rosenthal NE, et al. Arch Gen Psychiatry 1984;41:72-80.
  2. Eagles JM, et al. Br J Psychiatry 1999;175:472-5.
  3. Michalak EE, et al. Br J Psychiatry 2001;179:31-4.
  4. Kasper S, et al. Arch Gen Psychiatry 1989;46:837-44.
  5. Eagles JM, et al. Br J Psychiatry 2002;180:449-54.
  6. Avery DH, et al. Biol Psychiatry 2001;50:205-16.

SAD Association
PO Box 989
BN44 3HG