Phase Typing of Patients with Seasonal Affective Disorder: A Test for the Phase Shift Hypothesis

Abstract

The study evaluates the phase-shift hypothesis for seasonal affective disorder (Lewy et al., 1987, 1988) in parallel-design comparison of effects of morning (800-1000) or afternoon (1600-1800) light treatment on mood and circadian phase. Subjective arousal, body temperature, melatonin and cortisol were measured at 800, 1200, 1600, 2000 and 2400 in 23 women with seasonal depression and 20 controls before and after a week of bright light (2 hours per day). The rates of clinical response to both treatments were similar. Comparison of circadian variations did not provide evidence for significant phase-delay in patients compared to controls. However, morning light produced significant phase advance in patients, but not in controls. Also we found that advance phase shifts in well-responded patients were more often than in patients with worse response and controls. Before light treatment phase concordance between different variables in patients was lower compared to either themselves after light treatment or controls before and after light treatment. Dependence of antidepressant response to light from pretreatment circadian phases was also observed. Those patients who responded worse to morning light tended to have advance circadian phases, while those who responded worse to afternoon light tended to have delay phases. Although some results are lending support for the phase-shift hypothesis, other explanations for mechanisms by which biological rhythms are implicated in winter depression and light treatment might be suggested.