I don't recall - I must have slept well

The New York Times Magazine ran a story on the sleep industrial complex. Of course, a big segment of that are medications:

...In one six-week trial, for example, people taking Ambien every night fell asleep, on average, only 23 minutes faster than those taking the placebo. They spent 88 percent of their time in bed asleep, as opposed to 82 percent. Given that their objectively measured improvements are frequently this meager, why do sleeping pills create incommensurate feelings of having slept so well?

A popular theory is that one of the pill’s side-effects is actually contributing to their success. Most sleeping pills are known to block the formation of memories during their use, creating amnesia. This is why people who endure freaky side-effects — so-called “complex sleep-related behaviors” like getting into a car and driving or ravenously eating, all while asleep — don’t remember those events. Yet this amnesia could be quite beneficial, suggests Michael Bonnet, a professor of neurology at Wright State University Boonshoft School of Medicine in Dayton, Ohio. “How do you know you slept last night?” Bonnet asked me. A night of lousy, interrupted sleep, he points out, is easy to remember. “It’s full of memories, noise and pain, and heat and rolling around and obtrusive thoughts and worries — all of these various stimuli.” And we may continue to register such things even while asleep, making sleep vaguely unrefreshing. But a good night of sleep, Bonnet went on to say, “is always the antithesis to all those things, which is oblivion.” A sleeping pill, Bonnet speculates, in addition to encouraging sleep chemically in the brain, also “erases all of these thoughts that we use to define ourselves as being awake. The pill knocks them all out, and the patient says, ‘Hey, I must have been asleep because I don’t remember anything.’ ”

Drug-company representatives and consultants I spoke to confirm that their pills can create this mild form of amnesia but disagree that it contributes any significant benefit. “That is not my understanding of how Ambien works,” Dario Mirski, a psychiatrist and spokesman for Ambien’s manufacturer, Sanofi-Aventis, told me. It is difficult to find a clinical trial in which Z-drug takers drastically overestimated how long they slept.

Andrew Krystal, a Duke University psychiatrist and consultant to pharmaceutical companies like Sepracor, Lunesta’s manufacturer, acknowledges an apparent discrepancy in studies between small, objectively recorded improvements and the large percentage of subjects who end up feeling that a pill alleviated their insomnia. But because insomnia is complaint-based, he explained to me, an insomniac is cured when he stops complaining.

Interestingly, studies looking at sleep disturbance in alcoholics and relapse have found that sleep disturbance is not a great predictor of relapse, but distress over sleep disturbance is a predictor. What does that suggest?