Evidence for Efficacy

Acute-Phase Treatment
Because the recipient of light therapy is not “blind” to the treatment being administered, the development of a suitable control treatment for light therapy has been a problem since the earliest treatment studies were conducted. Despite this difficulty, dozens of controlled studies have been performed in which a treatment predicted by the investigators to be active was compared with a treatment predicted to be inactive. In many of these studies, differences between active and control treatments were found. Thus, more intense light sources have been shown to be superior to less intense ones; eye exposure has been shown to be superior to skin exposure (Wehr et al. 1987); and green light has been shown to be superior to red light (Oren et al. 1991). Two recent large-scale placebo-controlled trials have largely laid the efficacy question to rest. In one study of 158 patients with SAD (M. Terman et al. 1998), 30 minutes a day of morning 10,000-lux treatment beginning within 10 minutes of awakening proved robustly more effective than light at other times of day as well as placebo-controlled treatments. Another study of 96 subjects (Eastman et al. 1998) showed similar robust effects of morning light treatment compared with evening light treatment or placebo. These studies, together with the remaining literature, now provide strong evidence in favor of a specific antidepressant effect of light therapy in SAD. Morning light seems to provide the strongest treatment effects for SAD (Lewy et al. 1998), although evening light does provide a stronger treatment effect than placebo.

Although the conclusions described above are valid for light therapy administered with a light box, results of studies with a portable, head-mounted light visor are more difficult to interpret. In three multicenter studies of a light visor (Light Visor, Bio-Brite, Inc., Bethesda, MD) involving more than 200 patients, no difference in efficacy was observed at widely differing intensities (Joffe et al. 1993; Rosenthal 1993; Teicher et al. 1995). Response rates in these studies were comparable to those in some light box studies, but the absence of any differences between treatment conditions raises a problem in interpreting these findings. Although controlled studies are lacking, for those for whom portability is important for travel or mobility, a trial of a visor might be warranted.

Maintenance/Continuation Therapy
To date, there have been no studies of the long-term efficacy of light therapy for SAD. Indeed, the concept of long-term studies for this condition is problematic because the condition is by definition self-limiting. Nevertheless, it would be useful to know whether light therapy works for the duration of the winter, year after year. Properly controlled studies have not been undertaken to address this question. We thus have only clinical experience and systematic retrospective evaluation of SAD patients to guide us in our appraisal of the long-term efficacy of light therapy.

In a retrospective follow-up of 59 SAD patients, 24 of 57 who had originally been treated with light therapy (42%) continued to use light therapy regularly over the duration of follow-up (an average of 4 years), and all of these patients regarded light therapy as highly effective, with most reporting undiminished efficacy over time (Schwartz et al. 1996). Of the 33 patients who stopped using light therapy at some point during the follow-up period, 16 reported that light therapy had decreased in its effectiveness over time, 9 reported that light therapy was too inconvenient to continue, and 8 reported that they felt they had not become sufficiently depressed during subsequent winters to warrant renewed treatment. Although these data are of some interest, it is difficult to draw conclusions from them that bear on the likelihood that any particular individual will benefit from long-term treatment.

Clinical experience suggests that patients who respond once to light therapy will do so on subsequent occasions as well and that light therapy has a sustained benefit. In cases in which light therapy does appear to become less effective over time (a problem not unfamiliar in psychopharmacology), it is generally difficult to determine whether the decrease in effectiveness is a manifestation of tolerance or, rather, an indication that the depression has deepened. A tendency toward exacerbation of symptoms after withdrawal of light therapy would suggest that light therapy does provide ongoing antidepressant benefits. Some investigators have suggested that light therapy administered early in the winter is prophylactic against the development of SAD symptoms later in the winter, even if treatment is discontinued (Meesters et al. 1991, 1993). Others have disputed this claim and have found that patients develop symptoms as usual if treatment is discontinued (J. S. Terman et al. 1994).