The Big Sleep

This study by Daniel Kripke Robert Langer and Lawrence Kline in BMJ Open has gotten some publicity. But in case you haven't heard about it, it's time you did. They used data from electronic health records of Geisinger Health System, which serves a quarter of a million people in Pennsylvania.

They found that people who were prescribed sleeping pills had an excess mortality risk, over an average 2.5 years of observation, of more than 5 times for people prescribed the top third number of doses; but there was still a substantial risk for people prescribed fewer than 18 doses in a year. This pretty much applies to all kinds of sleeping pills, although the inference is strongest for the most commonly prescribed drugs, which are Ambien and similar chemicals. And it applies to all ages -- young and old. Oh yeah -- people who took sleeping pills also had an increased of being diagnosed with cancer after they got the prescriptions.

So, this just shows association, not causation, right? Technically yes, but they beat this data to death to try to rule out confounding. It is true that people who got prescriptions were more likely to have all sorts of pre-existing conditions, from asthma to heart disease to obesity. But they controlled for that. These risk ratios are for people who have the same comorbidity profile, same smoking status, and similar age. If taking sleeping pills isn't killing people, it's very hard to see what else could be.

They didn't just do this for the heck of it. There have been a lot of observations that seem to link sleeping pills with mortality but the evidence hasn't been of high quality. This is. And it isn't really all that hard to think of reasons. Sleeping pills have been associated with the onset of depression, with suicide, with accidental drug overdoses (e.g. mixing them with alcohol). Controlled trials show they cause impaired driving, more falls, bizarre behavior such as somnambulist binge eating, gastric reflux, and more infections. They may be directly carcinogenic. As these authors conclude:

The meager benefits of hypnotics, as critically reviewed by groups without financial interest, would not justify substantial risks. A consensus is developing that cognitive-behavioral therapy of chronic insomnia may be more successful than hypnotics. Against meager benefits, it is prudent to weigh the evidence from the current study and 24 previous reports, in order to reconsider whether even short-term use of hypnotics . . . is sufficiently safe.

Indeed. You will see these drugs heavily advertised on TV. I'm not a real doctor so I'm mustn't advice, but I sure as hell wouldn't take these and I really won't like it if you do. It's probably time to think about revoking approval.