Light therapy, in one form or another, has been used as a treatment for a number of conditions since ancient times. Nearly 2,000 years ago, Greco-Roman physicians were treating depression and lethargy with sunlight directed toward the eyes. During his Arctic expeditions in the 1890s, Frederick Cook, M.D., noticing the profound influences of light on the voyagers and Alaskan natives, described a syndrome characterized by depressed mood, fatigue, and loss of energy and sexual desire. In 1946, H. Marx reported the use of bright artificial light to treat four men who had become depressed during an Arctic winter.
Contemporary light therapy involves daily scheduled exposure to bright artificial light (Lam et al., 1999b; Partonen, 2001; Rosenthal and Matthews, 1999). The term light therapy is used to differentiate light therapy for psychiatric disorders from phototherapy for other conditions, such as hyperbilirubinemia or psoriasis.
Seasonal DepressionSince the first study of light therapy in winter seasonal affective disorder (SAD) (Rosenthal et al., 1984), a syndrome in which depression developed during fall or winter and remitted the following spring or summer for at least two successive years, numerous studies have concluded that bright light therapy is an effective treatment for SAD (Lam et al., 1999b; Magnusson and Boivin, 2003; Oren and Rosenthal, 1992; Partonen, 2001).
Light therapy is commonly administered by means of a light box--a metal structure containing fluorescent tubes behind a plastic diffusing screen (Lam et al., 1999b; Partonen, 2001; Rosenthal and Matthews, 1999). The dose of light exposure can be measured with the intensity and duration of the exposure. Efficacy is dose dependent to some extent, with both duration and intensity being important. Many studies used 2,500 lux light that was usually administered for two to six hours per day (Lam et al., 1999b; Partonen, 2001). Studies of 10,000 lux fluorescent light yielded similar results to studies using 2,500 lux for two hours (Magnusson and Kristbjarnarson, 1991; Partonen, 2001; Terman et al., 1990). To some degree, there is an inverse relationship between the intensity used and the duration required. Thus, 30 minutes of exposure to 10,000 lux may be as good as two hours of exposure to 2,500 lux, although such a linear relationship does not necessarily apply (Rosenthal and Matthews, 1999). The 10,000 lux fluorescent light box is usually used in clinical practice (Lam et al., 1999b).
A number of studies have found that morning light exposure was superior to evening light exposure (Eastman et al., 1998; Lam et al., 1999b; Lewy et al., 1998; Terman et al., 1998). It is important to note that evening light exposure was still significantly superior to placebo (Eastman et al., 1998; Terman et al., 1998). It is reasonable to start light therapy with morning exposure. Light treatment should not be administered late in the evening because it may cause insomnia.
Studies suggest that younger age and atypical symptoms such as hypersomnia, increased appetite, weight gain and increased carbohydrate consumption are associated with good response to light therapy (Lam et al., 1999b). The early response to light therapy partially predicts long-term response (Sher et al., 2001). If replicated, this observation may provide a simple test that will allow clinicians to predict which patient will respond best to light therapy.
Before starting light therapy, an ophthalmologic consultation is recommended for patients with a pre-existing retinal or eye disease (e.g., retinal detachment, retinitis pigmentosa, glaucoma, previous cataract surgery and lens removal) or a systemic illness that affects the retina (e.g., diabetes mellitus) (Lam et al., 1999b; Partonen, 2001). Side effects of light therapy include headaches, eyestrain, fatigue and insomnia (Lam et al., 1999b; Partonen, 2001; Rosenthal and Matthews, 1999). Most of these side effects respond to dose reduction. Hypomania and mania are uncommon, but serious, side effects. If a patient has prior history of mania, the risk of switch into mania might be minimized with a mood stabilizer (Kripke, 1991; Lam et al., 1999b).
