Chronic insomnia is a common and significant health problem. Fortunately, the poor sleep of insomniacs is improved by both pharmacological and behavioral treatments as recent literature reviews have clearly documented. Each of these treatments, though, has accompanying strengths and weaknesses. Pharmacotherapy works rapidly, but tolerance, dependence and loss of efficacy occur with extended use. Behavioral therapy is not fast-acting and requires a few weeks of practice before benefits are realized. In addition, during this initial phase of behavioral therapy patients often complain of feeling worse than they did prior to treatment. However the improvements in sleep due to behavioral treatment are quite durable. A combination treatment which takes advantage of the strengths of both pharmacological and behavioral therapy is a reasonable approach to treating insomnia. In spite of the compelling rationale for combining the treatments, empirical results have been disappointing. Recent findings have suggested that by adding pharmacotherapy to behavioral therapy long-term treatment gains are diminished. Our preliminary data suggest that extended, intermittent use of hypnotics is not necessary when patients are concurrently treated with behavioral therapy. Our analysis of week-to-week sleep changes among patients treated with behavioral therapy alone demonstrates that significant improvement occurs by the second week of treatment. A sensible approach to combining pharmacotherapy and behavioral treatment would be to discontinue hypnotics when sleep begins to improve due to the behavioral therapy. If hypnotic use is restricted to the first three weeks of treatment, we expect that poor sleep will rapidly improve and initial discomfort will be minimized. In addition, because the initial discomfort will be minimized we expect that subjects who receive the combination treatment will be better able to comply with, and remain more engaged in, the cognitive-behavioral treatment regimen. Also, because of the reduced discomfort in the early stages of treatment, more subjects in the combined treatment will stay in treatment. Furthermore, we do not expect this limited use of pharmacotherapy to diminish the long-term treatment gains. Our preliminary data have provided initial support for our expectations and the proposed study is designed to confirm these hypotheses.
