Tossing and turning for one night probably won't cause much harm besides the price of the extra shot in your latte. But for the 60 million Americans who lie awake night after night, lack of sleep can cause fatigue, irritability, and inability to concentrate. Insomnia is usually treated with drugs, though psychotherapy works as well, and it's unclear which treatment is more effective. A group of researchers from Harvard Medical School pitted the two against each other to find out.

What the researchers wanted to know: What works best to treat insomnia: psychotherapy or prescription drugs?

What they did: The researchers placed newspaper ads to recruit patients who had trouble sleeping. After screening out ineligible patients, they were left with 63 participants, whom they divided into four groups. The first group received Ambien, the most popular insomnia pill, for six weeks. The second group received four individual counseling sessions and one telephone counseling session to learn strategies for sleeping; this was also over a six-week period. A third group received both Ambien and the counseling sessions, and a fourth group was given a placebo pill with no counseling. All of the participants kept two-week sleep diaries before, in the middle of, and after their treatment, as well as one, three, six, and 12 months after the end of their treatment. Lastly, the participants had three nights of electronic sleep monitoring, using a device that measures when people are asleep or awake, and what phase of sleep they are in.

What they found: Psychotherapy alone was the best treatment overall both for participants who had a hard time falling asleep and for those who woke up during the night. All the participants took an average of about 70 minutes to fall asleep before the study started, and the psychotherapy group got that down to 34 minutes after treatment. The Ambien group took only 45 minutes to fall asleep during treatment, but after treatment ended, they went back up to almost an hour. The combination group took only 38 minutes to fall asleep after treatment, but their averages crept up again over the next few months and years, while the psychotherapy group's stayed low. Similarly, the group that received psychotherapy showed the most improvement in the amount of time they spent asleep after they turned out their lights.

What it means to you: While prescription drugs are the most commonly used treatment for insomnia, this study shows they are not necessarily the most effective. In addition, they are more expensive than psychotherapy, especially because the effect of a prescription wears off after you stop taking the drugs, while the effects of therapy appear to last longer. The people who were given behavior therapy were taught to help themselves go to sleep faster using a routine. First, they were not supposed to use their bed for anything other than sleeping or sex; they were to go to bed only when drowsy, and, if they couldn't go to sleep within 20 or 30 minutes, they were supposed to go to another room and do something quietly until they felt sleepy.

Caveats: Each treatment group in the study was pretty small, about 15 people each. For a statistician, that means that the results of each different treatment have to be pretty far apart to know that it is a real difference and not due to a coincidence. (If there are more people, coincidences in the numbers happen less frequently.) So just the fact that they found differences between the groups is impressive—and it is conceivable that the effects the researchers found in this study would be stronger, given more people. In addition, this study only evaluated the effects of treatment on people whose primary problem was insomnia; it may not apply to people who have insomnia caused by something else such as depression or drug use.

Find out more: The National Women's Health Information Center has a Web page—not just for women—all about insomnia.

The National Center on Sleep Disorders Research has tons of information about sleep, including this quiz, where you can test your own knowledge.

Read the article: Jacobs, G.D. et al. "Cognitive Behavior Therapy and Pharmacotherapy for Insomnia." Archives of Internal Medicine. Sept. 27, 2004, Vol. 164, No. 17, pp. 1888–1896.

Abstract online: http://archinte.ama-assn.org