Saturday, March 03, 2012

Do Sleeping Pills Kill?

A recent study published in BMJ Open found that the prescription of sleeping pills (hypnotics) was associated with a greater than threefold increase in the risk of death. The abstract and full article are available (free) here:

There are a number of problems with this study, as mentioned by Dr. Nancy Collop, President of the American Academy of Sleep Medicine:

"Although the study found that the use of hypnotic medication, or sleeping pills, was associated with an increased risk of mortality, a cause-and-effect relationship could not be established because the study only analyzed an insurance database... it was impossible for them to control for psychiatric conditions and anxiety, which is an area of significant concern to this study population..."

The major problem that I have with the study is that, as I far as I can see looking at the full article, the presence of insomnia was not controlled for (most of the comparison group did not have insomnia). Insomnia itself is associated with increased mortality (in some but not all studies) and morbidity (illness). To further illustrate why this is a problem, consider a hypothetical study in which the use of insulin was associated with an increased death rate. It is well known that dibetes itself increases the risk of death. If the comparison group in this hypothetical study did not have diabetes, I don't think much if anything could be learned from the study.

This new sleeping pill study doesn't add much new information to the field of sleep medicine, although it does add to the literature suggesting that insomnia is a serious problem. I did find this earlier editorial by Dr. Kripke (primary author of the current study) which sums up the problem with this current study:
"Numerous previous studies have shown an association of hypnotic use to mortality, which can become confounded with insomnia. Was mortality controlled for hypnotic usage in examining the association with insomnia?"
To fully apply to the current study, that last statement can be changed to: Was mortality controlled for insomnia in examining the association with hypnotic usage?

Edit: The association between insomnia and death is a complex issue. It is clearly associated with illness/morbidity. This abstract sums up the issues:

Some studies have found that while a subjective complaint of insomnia is not associated with an increased risk of death, short sleep time is.