Almost every credible doctor agrees you should attempt to avoid sleeping pills if possible, and indeed drugs should not be taken unless there is a good reason. However, most doctors agree modern pharmaceuticals are a technology we can use wisely to make our lives better, and that under certain circumstances, the patient can and should use over-the-counter or prescription drugs.
But there are skeptics – people who believe such pills are vastly overused, and some who don’t think they should be used in any but extreme situations. These public-health-minded doctors feel that the risks of hypnotics are underappreciated and include a shortening of life by a mechanism of unknown etiology.
These concerns are beyond the normal worries about publicized side effects. Some of the more known side effects can be dangerous, such as morning grogginess, drowsy or even sleep driving, and the zombie-like behavior caused by z-drugs. Doctors know about these and any insomniac who takes his healthcare responsibly should, too.
The further concerns – the ones most patients, doctors, and pharmacists never think about – revolve around the long-term use of hypnotics and the possibility that they raise the risk for cancer. The disingenuous response to this concern is that such drugs are not designed to be used over extended periods of time. They are for short-term alleviation of an acute problem, and the FDA doesn’t approve hypnotics for use for more than a month (this was pretty much true before the approval of Lunesta.) However, it is an open secret that long-term use occurs, often with the consent of the physician. Some patients either take the same drug for months or years continuously, or they engage in a stop-and-start intermittent pattern. Or they may switch among different hypnotics. Any way it happens, many people are taking these drugs over an extended period.
There have been concerns about increased cancer risk associated with hypnotic medicines going back to at least the 1970s when an American Cancer Society study found a correlation.
A recent longitudinal study found that people who took sleeping pills had "four times the mortality as the matched hypnotic-free control patients" and that even moderate use raised the risk of death: "threefold increased hazards of death even when prescribed <18 pills/year". The types of hypnotic drugs studied included antihistamines, benzodiazepines, and z-drugs, and all showed an increased mortality even though they work by different mechanisms. Of course, this does not prove that hypnotic drugs cause death. The familiar refrain from epidemiologists is always "correlation does not imply causation". Still, such findings raise eyebrows. What might the extra deaths be due to? Cancer, heart disease, etc – the regular things people die from. The study authors claim it shows general, if unspecific, dangers of long-term use of sleeping pills.
Criticism of this study highlights the fact that the cohorts were not matched for underlying sleep disorder – and it is known that sleep disorders correlate with shortened lifespans. Further, the patients on the drugs had more co-morbidities going into the study, and critics say this was not fully accounted for in the analysis.
Hypnotic drugs generally do not significantly extend sleep time, so skeptics wonder if even a small increase in risk is worth the little gain in sleep. A counterargument is that the drugs do improve sleep continuity, which can be beneficial even if total sleep time does not increase much.
Previous studies - see http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3293137/bin/bmjopen-2012-000850-s2.pdf - have considered the effect of long-term sleep aid use on mortality, and most found some sort of statistically significant increase in mortality, although the studies were not always designed to measure this question directly, and there may be confounding factors.
The more dramatic findings of the 2012 longitudinal study suggest "hypnotics may have been associated with" several hundreds of thousands of excess deaths in the US in any recent year.
A formal randomized trial could perhaps get to the bottom of the question of whether sleeping pills increase risk of death, but such a trial would be impractical to mount, and perhaps unethical.
The regulatory authorities and medical establishment are conservative in their recommendations of drug use, which is why the labels on sleeping pills limit use to a few weeks. Some feel the efficacy of these medications goes on for an extended period, and patients do not build up a tolerance. These is some evidence that these medications are still valuable to patients after several months although trials going on for longer than a year need to be done. Some of the science that was used to discourage long-term use was based on benzodiazepines, which are rarely employed primarily for insomnia any more.
Chinese scientists examined thousands of insomniacs and found that hypnotic drugs did indeed increase the risk of dementia. People in the age range 50 to 65 had about twice as high a risk for dementia when they were long-term users of these drugs. The risk increased with the dosage and with the half-life of the drug.