Hypnotic drugs are consumed mainly by chronic users, who often take hypnotics for years continuously. Two studies of the American Cancer Society, each involving over 1 million subjects followed prospectively for 6 years, showed that hypnotic use predicted increased mortality after control for comorbidities. At least 8 other studies have also noted that hypnotic use predicts increased mortality. There is an epidemiologic consensus that hypnotic use is associated with increased risk of death. Chronic hypnotic use is also strongly associated with insomnia, poor function, and poor quality of life. There is considerable evidence that hypnotics may in part cause this increased morbidity, as well as automobile accidents and falls. In contrast, there is no persuasive evidence that long-term use of hypnotics produces any benefit. Available evidence is that the risks of chronic hypnotic use outweigh the benefits.

Introduction

Most prescriptions of hypnotic drugs are consumed by chronic consumers, who often take hypnotics nightly for years at a time. The greatest sales, the greatest consumption, the greatest health impact and the greatest economic impact are with the chronic users. From U.S. data, it has been estimated that perhaps 2/3 or 3/4 of prescriptions go to the chronic user.^1,2 In the European setting, the percentage of chronic use might be rather higher. Ohayon found that 80% of those in France and 76% of those in Quebec who were taking a sleep-promoting drug had consumed it for at least 6 months.³ In Italy, 73% taking a hypnotic had done so for a year or more.⁴ In Spain, 72% had used the hypnotic they received for over 3 months.⁵ Until recently, the dispensing of hypnotic drugs had more economic impact in American medicine than all other aspects of sleep medicine combined. That may still be so. It is a shame that we have no random clinical trials of hypnotics tested objectively over several months or years. Thus, we must make our best judgement of the risks and benefits of chronic hypnotic use from other evidence.

This discussion will focus on the risks of chronic hypnotic use, because substantial evidence is available about such risks. Little can be said about the benefits of chronic hypnotic use, for there is no persuasive scientific evidence that such benefits occur. Accordingly, this discussion will review the mortality risks and the morbidity risks of chronic hypnotic use, followed by a brief critique of the information on long-term benefits.

Mortality Associated with Prescription Hypnotics

In questionnaires collected from more than 1 million people in 1959, the American Cancer Society asked participants if they took “sleeping pills” “never,” “seldom” or “often.” Men who indicated that they took sleeping pills “often” had 57% increased mortality and women had 54% increased mortality after prospective 6-year follow-up, controlling for age and for reported sleep duration.⁶ The ratios for “seldom” using sleeping pills were 15% and 13% increased, respectively, also statistically significant. The risk for sleeping pill users was more than 100% greater among those who took sleeping pills “often”, if reported insomnia and sleep duration were not controlled. There were four important limitations to this study. First, the study did not distinguish between prescription versus over-the-counter “sleeping pills.” Second, when the questionnaires were distributed in 1959, barbiturates were the predominant hypnotics, but in the next decade they were rapidly replaced by benzodiazepines with less acute overdose toxicity. Third, there was uncertainty of what frequency of use participants signified when they indicated that they took sleeping pills “often.” Fourth, because of limitations on computer facilities at the time the data were analyzed, it was not possible to control extensively for comorbidities.

Most of the limitations of this study were overcome in the American Cancer Society's Cancer Prevention Study II (CPSII), which received questionnaires from 1.1 million participants in 1982, at a time when benzodiazepines dominated the hypnotics market.⁷ Participants were asked “how many times in the last month” they used “prescription sleeping pills”. Before controlling for comorbidities, as in the previous study, use of sleeping pills 30 or more times per month (essentially every night) was associated with more than doubled mortality hazard.² Controlling simultaneously for 32 risk factors and comorbidities, the hazard associated with sleeping pill use 30+ times per month was increased 24% for women and 25% for men.⁷ For use 1-29 times per month (median 3), the hazard was increased 10% for women and 15% for men. In all these analyses, the adjusted hazard of sleeping pill use was statistically significant. The risk was particularly elevated for suicide, but was also elevated for cancer.⁸ A significantly elevated risk for cancer was found, even excluding those sleeping pill users who said that they had cancer at the time they completed the questionnaire (unpublished).

Strengths of this study were the extensive control for concomitant risk factors and the questionnaire_s distinction of “prescription sleeping pills” from “Valium” and “Librium,” which were the most popular minor tranquilizers in 1982. Limitations were the inability to specify exactly which “prescription sleeping pills” each participant was taking, and the fact that some popular contemporary hypnotics were not marketed at the time. The most popular hypnotics in 2002 were benzodiazepine agonists such as zolpidem and zaleplon which are not actually benzodiazepines. These contemporary agents bind somewhat selectively to the various benzodiazepine receptors. Since neither Valium or Librium was associated with excess mortality hazard,⁷ one might be skeptical that the long-acting hypnotics such as flurazepam and quazepam, which are very similar pharmacologically, would be at fault, but shorter-acting benzodiazepines such as temazepam and triazolam might pose a higher risk. The kinetics of zolpidem and zaleplon at the receptor are rather similar to those of triazolam.

Several much-smaller studies have also examined the mortality hazards of hypnotics with a variety of control methods for confounding risk factors. Allgulander and Nowak found that men and women who took hypnotics and tranquilizers had about 7-fold increases in suicide.⁹ Allgulander also noticed a high rate of suicides among those admitted for sedative-hypnotic dependence.¹⁰ Thorogood and colleagues found that benzodiazepine and barbiturate use significantly predicted 4-fold and 6-fold risks of fatal myocardial infarction.¹¹ Rumble and Morgan found that in Nottingham, use of sleep medication significantly predicted mortality.¹² The effect was stronger for medications not usually classified as hypnotics, but the numbers were small. Brabbins et al found no significant relationship of hypnotic use to mortality in Liverpool, but the small sample included only 86 hypnotics users. ¹³ Sundquist et al observed that municipalities with higher sedative/hypnotic use had higher mortality.¹⁴ Merlo et al found an increased risk of cancer mortality in men using anxiolytics-hypnotics.¹⁵ Hays, Blazer, and Foley found no significant relationship of sleep medications to survival.¹⁶ Kojima and colleagues found that use of drugs for sleep was associated with total mortality with adjusted risk ratios of 1.28 for men (not significant) and 1.81 for women (marginally significant.)¹⁷ Mallon et al found habitual sleeping pill use significantly related to total mortality, with risk ratios (RR) of 3.0 in men and 3.8 in women.¹⁸ After multivariate adjustment, the relationships were significant for cancer deaths in men (RR=5.3) and noncancer noncoronary deaths in women (RR=3.3).¹⁸ In conclusion, a variety of studies from different countries have found a preponderance of evidence for increased mortality risk associated with the use of hypnotics. Lack of significant evidence of hypnotics risk in some of these studies could be attributed to insufficient power in small samples.

It appears that hypnotics may cause deaths through several mechanisms. The lethal effects of barbiturates and benzodiazepine agonists in overdose are extremely well understood on molecular, neuronal, and organismic levels. Hypnotics in high doses are reliably used to put animals to sleep and to execute the death sentence. In lower doses, hypnotics may be synergistic with alcohol and other drugs. Whether hypnotics occasionally cause respiratory arrest in recommended doses is not known, but there is evidence that hypnotics exacerbate sleep apnea in at least a percentage of patients. This might lead to chronic hypertension, cardiac ischemia, and consequent demise. Evidently, hypnotic use is associated with increased suicide, but it is unclear whether this is mediated through depressive effects of hypnotics, through heightened disinhibition or confusion, or through confounding with depression. A mechanism by which hypnotics would increase cancer deaths is not currently known, but the suggestion for some specificity in cancer deaths deserves more study.

Available studies indicate that not all mortality related to sleeping pill use can be explained by comorbidities. Although control for risk factors substantially reduced the mortality associated with use of prescription use in several studies, there may have been over control. For example, if sleeping pills cause deaths by increasing sleep apnea, thus increasing hypertension and heart disease, control for hypertension and heart disease may have biased the analysis to underestimate mortality caused hypnotics. Lapane et al found that benzodiazepine users had a 2-fold relative risk of ischemic heart disease.¹⁹ Pratt et al found that barbiturate use was associated with a 2.11 odds ratio for myocardial infarction; the odds ratio for benzodiazepines, 1.33, was not significant.²⁰ On the other hand, control may have been insufficient for comorbid factors such as depression. Depression was directly controlled in only a minority of the studies. In my opinion, the likelihood of over control and the likelihood of under control were about equal in the epidemiologic literature concerning hypnotic risks.

I know of no epidemiologic evidence that use of any hypnotic prolongs survival or improves general health in any way. If one looks at the weight of epidemiologic evidence (fig. 1), the weight is certainly on the side of hypnotics increasing mortality.

Figure 1

In epidemiologic evaluations of hypnotic risks for mortality, the evidence of 10 studies tilts heavily on the side associating hypnotics with increased mortality. There are two negative studies, but there is no suggestion that hypnotics lessen mortality. (more...)

It is true that controlled trials of long-term hypnotic administration would be necessary for absolute proof of causal effects of hypnotics (either negative or positive), and as new hypnotics are introduced, epidemiologic data on the latest drugs will always be sparse. In the absence of controlled trials, the epidemiologic evidence is the best evidence available on the mortality risks-to-benefits ratio of hypnotics. The burden is on advocates of chronic hypnotic prescribing to prove that any hypnotic might be safe for long term use.

Morbidity Associated with Chronic Hypnotic Use

As might be expected, the use of hypnotics “often” is associated with having insomnia “often_.⁶ In the 2001 Sleep In America Poll (national data kindly loaned by the National Sleep Foundation), use of prescription medicines for sleep was significantly correlated with trouble falling asleep, awakening during the night, awakening early, and awakening unrefreshed. Use of medications was not significantly correlated with self-reported total sleep time in the Sleep In America Poll or in French data concerning insomniacs.^21,22 We expect people to take hypnotics because they have insomnia symptoms, so it is even surprising that there is a substantial portion of those taking hypnotics who do not report insomnia.^6,23,24 Converse causality, that taking hypnotics causes insomnia symptoms, has rarely been considered. There are no data to exclude that hypnotics causing insomnia is one pathway contributing to the association of hypnotic use and insomnia. Indeed, there are preliminary reports that when chronic hypnotics users undergo withdrawal from these drugs, they eventually sleep better,^25,26 which would suggest that hypnotics have caused insomnia.

Certainly, hypnotics with short half lives cause early awakening when administered at bedtime, and they cause withdrawal insomnia and anxiety on nights when they are withheld.^27,28 It would appear that these withdrawal mechanisms are active with contemporary hypnotics. Patients taking zolpidem sleep significantly worse subjectively than patients taking placebo on nights when the drug is skipped.²⁹ Zaleplon does not consistently increase total nightly sleep time,³⁰ so its benefit of reducing sleep latency is seemingly counterbalanced by more awakening later in the night.³¹ These withdrawal effects of hypnotics demonstrate a complex interplay between their capacity to slightly increase sleep and their capacity to harm sleep.

There are some indications that the withdrawal effects and other sleep-damaging effects of hypnotics are more persistent than their benefits for sleep, to which rapid tolerance may develop. For example, the withdrawal effects of triazolam were worse than those of placebo at a time when, because of tolerance, the benefits did not exceed placebo.³² Early awakening effects of triazolam seemingly increased over several weeks.²⁷ In chronic hypnotic users, tolerance to flurazepam and midazolam developed within 2 weeks, as compared to placebo.³³ Zolpidem was ineffective in its recommended dose at 5 weeks.³⁴ After 8 weeks, intermittent use of zolpidem did not increase weekly subjectively-reported total sleep.²⁹ Although the subjective quality of sleep was not significantly improved by two benzodiazepines as compared to placebo after 24 weeks, significant deterioration of quality was demonstrated when the benzodiazepines were withdrawn.³⁵ This strange evidence that the sleep-harming effects of hypnotics might be more persistent than their benefits suggests a very negative assessment of the chronic effects of these drugs. The implication of these studies is that with chronic use, hypnotics may be of no benefit at all for sleep, but patients may continue dosing themselves because of the distress which is caused by drug withdrawal. This drug withdrawal is certainly a factor in the widely-reported habituation and dependence on hypnotics which contributes to the chronicity of hypnotic use.

Although hypnotics with short half-life cause much less impairment of daytime performance than those which are persistent during the day, even with short-half-life drugs, there is more evidence that they impair performance than that they improve performance, at least when the indication is insomnia rather than jet lag or shift work.^36,37 In a recent study, 8 weeks of intermittent use of zolpidem caused no improvement in health or performance, though the study had 80% power to detect a small-to-medium effect of 0.4 SD.³⁸ On the other hand, the more persistent hypnotics, such as flurazepam and quazepam, cause considerable impairment of daytime performance which augments with continuing daily use as the drug accumulates.³⁶ It also appears that the long-half-life hypnotics cause more problems with automobile accidents and falls, (but see data on zoplicone.)^39-45 Both short and long-half-life hypnotics can impair memory.⁴⁶

People who take hypnotics report an excess of a variety of somatic conditions and psychological symptoms.^47-50 Moreover, after adjusting for initial levels and confounding factors, Byles et al found that use of sleep medication predicted poor quality of life (SF-36), increased falls, increased doctor consultations, and increased days in the hospital upon 3-year follow-up.⁵⁰ The latter evidence would suggest that hypnotics caused these multiple aspects of morbidity and poor health.

Benefits of Chronic Hypnotic Use

In clinical trials of hypnotics, a frequent design has been to compare an initial drug-free or placebo baseline with results during several weeks of hypnotic administration. These sequential designs without cross-over are completely flawed and useless for judging drug effects, because parallel-placebo studies show that improvement occurs sequentially among patients given placebo. Therefore, the only scientifically meaningful designs would be counter-balanced cross-overs or better, parallel placebo-hypnotic contrasts.

There have been few parallel-placebo contrasts of hypnotic administration for more than 4 weeks, and no counter-balanced cross-overs of which I am aware. Of the parallel designs, as has been mentioned, several suggested that after 2-8 weeks of administration, hypnotics produce no more benefit than placebo at recommended doses, that is, there was no long-term benefit.^32-34 In 8 weeks of intermittent use, the benefits of zolpidem on drug nights was balanced by the harm on drug-withdrawal nights, resulting in no significant increase in weekly sleep.³⁸ The heroic study of Oswald and colleagues gave somewhat uncertain results for 24-week administration of two benzodiazepines, because only subjective measurements were made, because the benefit for sleep quality did not remain significant for 24 weeks, and because the apparent small benefit for sleep latency was not fully consistent.³⁵ In treatment of late-life insomnia for 8 weeks, temazepam was not significantly better than placebo on polysomnographic variables, though temazepam was superior on sleep diary scores.⁵¹ However, upon follow-up, patients receiving temazepam deteriorated and were slightly worse than the placebo group at 24 months, when some had resumed taking medication. To summarize the meager evidence of parallel studies of longer-term use of hypnotics, use of hypnotics for more than 4 weeks produces no objective improvement in sleep and little (if any) subjective benefit. I am not aware of any placebo-controlled parallel studies which show benefits of long-term use of hypnotics for performance, mood, health, or general quality of well-being.

Need for Research

To my knowledge, there has been no parallel-placebo-controlled trial of giving a hypnotic for more than 8 weeks which used objective sleep recording. Objective recording is important, because subjective and objective reports of sleep are so frequently discordant, particularly when addictive and memory-impairing properties of hypnotics may distort patient reports. Elsewhere, I have discussed the need for hypnotics trials in more detail.^52,53 In the absence of conclusive trials demonstrating the causal effects of chronic use, our best evidence comes from hypnotics epidemiology.

Conclusions

We must recognize that chronic use of hypnotics is associated with poor sleep, poor health, and reduced survival. We are faced with considerable evidence that hypnotics may create sleep and performance problems, memory disturbance, driving accidents, and falls. Lacking controlled evidence that chronic hypnotic use improves health or function in any way, and with evidence that there is no objective benefit even for sleep, we see that the weight of evidence favors risks over benefits in chronic hypnotic use (fig. 2). Several consensus conferences in the U.S. have reached the same conclusion,^54-57 along with the U.S. Food and Drug Administration and the regulatory agencies of several other countries.

Figure 2

For chronic hypnotic use, the balance of evidence tilts strongly toward risks, there being only a tiny cloud of evidence for subjective benefits for sleep.

Acknowledgment

Supported by National Institutes of Health AG12364, HL61280, and AG15763 and by the Sam and Rose Stein Institute for Research on Aging.

Note Added in Proof

Krystal and colleagues (Krystal AD, Walsh JK, Laska E et al. Sustained efficacy of eszopiclone over 6 months of nightly treatment: results of a randomized, double-blind, placebo-controlled study in adults with chronic insomnia. Sleep 2003; 26:793-799) found subjective evidence for improvements in average health in a placebo-controlled 6-month trial of eszopiclone. However, the eszopliclone group had almost 3 times the severe adverse effects, leaving the benefits/risks ratio uncertain.

References

1.

Kripke DF. Chronic hypnotic use: He neglected problem In: Koella WP, Ruther E, Schulz H, eds.Sleep _84Stuttgart: Gustav Fischer Verlag,1985338–340.

2.

Kripke DF, Klauber MR, Wingard DL. et al. Mortality hazard associated with prescription hypnotics. Biol Psychiatry. 1998;43:687–693. [PubMed: 9583003]

3.

Ohayon MM, Caulet M. Psychotropic medication and insomnia complaints in two epidemiological studies. Can J Psychiatry. 1996;41:457–464. [PubMed: 8884035]

4.

Balestrieri M, Bortolomasi M, Galletta M. et al. Patterns of hypnotic drug prescription in Italy. Brit J Psychiat. 1997;170:176–180. [PubMed: 9093510]

5.

Rayon P, Castro-Serrano M, del Barrio H. et al. Hypnotic drug use in Spain: A cross-sectional study based on a network of community pharmacies. Pharmacoepidemiology. 1996;30:1092–1100. [PubMed: 8893114]

6.

Kripke DF, Simons RN, Garfinkel L. et al. Short and long sleep and sleeping pills: Is increased mortality associated? Arch Gen Psychiatry. 1979;36:103–116. [PubMed: 760693]

7.

Kripke DF, Garfinkel L, Wingard DL. et al. Mortality associated with sleep duration and insomnia. Arch Gen Psychiatry. 2002;59:131–136. [PubMed: 11825133]

8.

Kripke DF, Marler MR. Specific causes of mortality associated with prescription sleeping pill usage [abstract] Sleep Research Online. 1999;2:144.

9.

Allgulander C, Nowak J, Rice JP. Psychopathology and treatment of 30,344 twins in Sweden. I. The appropriateness of psychoactive drug treatment. Acta Psychiatrica Scandinavica. 1990;82:420–426. [PubMed: 2291410]

10.

Allgulander C, Ljungberg L, Fisher LD. Long-term prognosis in addiction on sedative and hypnotic drugs analyzed with the Cox regression model. Acta Psychiatrica Scandinavica. 1987;75:521–531. [PubMed: 3604738]

11.

Thorogood M, Cowen P, Mann J. et al. Fatal myocardial infarction and use of psychotropic drugs in young women. Lancet. 1992;340:1067–1068. [PubMed: 1357456]

12.

Rumble R, Morgan K. Hypnotics, sleep, and mortality in elderly people. J Am Geriatr Soc. 1992;40:787–791. [PubMed: 1634722]

13.

Brabbins CJ, Dewey ME, Copeland RM. et al. Insomnia in the elderly: Prevalence, gender differences and relationships with morbidity and mortality. Int J Ger Psych. 1993;8:473–480.

14.

Sundquist J, Ekedahl A, Johansson S-E. Sales of tranquillizers, hypnotics/sedatives and antidepressants and their relationship with underprivileged area score and mortality and suicide rates. Eur J Clin Pharmacol. 1996;51:105–109. [PubMed: 8911872]

15.

Merlo J, Hedblad B, Ogren M. et al. Increased risk of ischaemic heart disease mortality in elderly men using anxiolytics-hypnotics and analgesics. Eur J Clin Pharmacol. 1996;49:261–265. [PubMed: 8857070]

16.

Hays JC, Blazer DG, Foley DJ. Risk of napping: Excessive daytime sleepiness and mortality in an older community population. J Am Geriatr Soc. 1996;44:693–698. [PubMed: 8642162]

17.

Kojima M, Wakai K, Kawamura T. et al. Sleep patterns and total mortality: A 12-year follow-up study in Japan. J Epidemiol. 2000;10:87–93. [PubMed: 10778032]

18.

Mallon L, Broman J-E, Hetta J. Sleep complaints predict coronary artery disease mortality in males: A 12-year follow-up study of a middle-aged Swedish population. J Int Med. 2002;251:207–216. [PubMed: 11886479]

19.

Lapane KL, Zierler S, Lasater TM. et al. Is the use of psychotropic drugs associated with increased risk of ischemic heart disease? Epidemiology. 1995;6:376–381. [PubMed: 7548344]

20.

Pratt LA, Ford DE, Crum RM. et al. Depression, psychotropic medication, and risk of myocardial infarction. Prospective data from the Baltimore ECA follow-up. Circulation. 1996;94:3123–3129. [PubMed: 8989119]

21.

Ohayon MM, Caulet M, Arbus L. et al. Are prescribed medications effective in the treatment of insomnia complaints? J Psychosom Res. 1999;47:359–368. [PubMed: 10616230]

22.

Leger D, Janus C, Pellois A. et al. Sleep, morning alertness and quality of life in subjects treated with zopiclone and in good sleepers. Study comparing 167 patients and 381 good sleepers. Eur Psychiatry. 1995;10:99s–102s. [PubMed: 19698404]

23.

Ohayon M, Caulet M, Lemoine P. Sujets ag_ habitudes de sommeil et consommation de psychotropes dan la population fran_se. L_Encephale. 1996;22:337–350. [PubMed: 9035990]

24.

Ohayon MM, Lader MH. Use of psychotropic medication in the general population of France, Germany, Italy, and the United Kingdom. J Clin Psychiatry. 2002;63:817–825. [PubMed: 12363124]

25.

Morin CM, Bastien CH, RadoucoThomas M. Late-life insomnia and chronic use of benzodiazepines: Medication tapering with and without behavioral intervention [abstract] Sleep. 1998;21S:99.

26.

Parrino L, Smerieri A, Boselli M. et al. Discontinuation of elevated doses of benzodiazepines used as hypnotics: Polysomnographic assessment of sleep parameters [abstract] Sleep. 1998;21S:101.

27.

Kales A, Soldatos CR, Bixler EO. et al. Early morning insomnia with rapidly eliminated benzodiazepines. Science. 1983;220:95–7. [PubMed: 6131538]

28.

Kales A, Soldatos CR, Bixler EO. et al. Rebound insomnia and rebound anxiety: A review. Pharmacology. 1983;26:121–37. [PubMed: 6132413]

29.

Walsh JK. Reply to Kripke. Sleep Medicine Reviews. 2003;7:195–196. [PubMed: 12628218]

30.

Elan Pharmaceuticals Sonata (zaleplon) Capsules 11/26/2001 [prescribing information] Philadelphia PA Wyeth Laboratories2002 .

31.

George CFP. Pyrazolopyrimidines. Lancet. 2001;357:1623–1626. [PubMed: 11716908]

32.

Mitler MM, Seidel WF, van den Hoed J. et al. Comparative hypnotic effects of flurazepam, triazolam, and placebo: A long-term simultaneous nighttime and daytime study. J Clin Psychopharmacol. 1984;4:2–15. [PubMed: 6141188]

33.

Kripke DF, Hauri P, Ancoli-Israel S. et al. Sleep evaluation in chronic insomniacs during 14-day use of flurazepam and midazolam. J Clin Psychopharmacol. 1990;10(Supplement 4):32S–43S. [PubMed: 2229462]

34.

Scharf MB, Roth T, Vogel GW. et al. A multicenter, placebo-controlled study evaluating zolpidem in the treatment of chronic insomnia. J Clin Psychiatry. 1994;55:192–199. [PubMed: 8071269]

35.

Oswald I, French C, Adam K. et al. Benzodiazepine hypnotics remain effective for 24 weeks. Bristish Medical Journal. 1982;284:860–863. [PMC free article: PMC1496323] [PubMed: 6121605]

36.

Judd LL, Ellinwood E, McAdams LA. Cognitive performance and mood in patients with chronic insomnia during 14-day use of flurazepam and midazolam. J Clin Psychopharmacol. 1990;10:56S–67S. [PubMed: 2229464]

37.

Johnson LC, Chernik DA. Sedative-hypnotics and human performance. Psychopharmacology (Berlin). 1982;76:101–113. [PubMed: 6805019]

38.

Walsh JK, Roth T, Randazzo A. et al. Eight weeks of nonnightly use of zolpidem for primary insomnia. Sleep. 2000;23:1087–1096. [PubMed: 11145323]

39.

Barbone F, McMahon AD, Davey PG. et al. Association of road-traffic accidents with benzodiazepine use. Lancet. 1998;352:1331–1336. [PubMed: 9802269]

40.

Greenblatt DJ, Allen MD. Toxicity of nitrazepam in the elderly: A report from the Boston collaborative drug surveillance program. Br J Clin Pharmacol. 1978;5:407–13. [PMC free article: PMC1429343] [PubMed: 656280]

41.

Hemmelgarn B, Suissa S, Huang A. et al. Benzodiazepine use and the risk of motor vehicle crash in the elderly. JAMA. 1997;278:27–31. [PubMed: 9207334]

42.

Oster G, Huse DM, Russell MW. et al. Benzodiazepine tranquilizers and the risk of accidental injury. Am J Public Health. 1990;80:1467–1470. [PMC free article: PMC1405109] [PubMed: 1978581]

43.

Ray WA, Griffin MR, Schaffner W. et al. Psychotropic drug use and the risk of hip fracture. N Engl J Med. 1987;316:363–9. [PubMed: 2880292]

44.

Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med. 1988;319(26):1701–1707. [PubMed: 3205267]

45.

Ray WA, Griffin MR, Downey W. Benzodiazepines of long and short elimination half-life and the risk of hip fracture. JAMA. 1989;262(23):3303–3307. [PubMed: 2573741]

46.

Greenblatt DJ, Harmatz JS, Engelhardt N. et al. Pharmacokinetic determinants of dynamic differences among three benzodiazepine hypnotics: Flurazepam, temazepam, and triazolam. Arch Gen Psychiatry. 1989;46:326–332. [PubMed: 2564763]

47.

Allgulander C, Nasman B. Regular hypnotic drug treatment in a sample of 32,679 Swedes: Associations with somatic and mental health, inpatient psychiatric diagnoses and suicide, derived with automated record-linkage. Psychosom Med. 1991;53(1):101–108. [PubMed: 2011645]

48.

Quera-Salva MA, Orluc A, Goldenberg F. et al. Insomnia and use of hypnotics: Study of a French population. Sleep. 1991;14(5):386–391. [PubMed: 1759090]

49.

Roehrs T, Hollebeek E, Drake C. et al. Substance use for insomnia in Metropolitan Detroit. J Psychosom Res. 2002;53:571–576. [PubMed: 12127173]

50.

Byles JE, Mishra GD, Harris MA. et al. The problems of sleep for older women: Changes in health outcomes. Age Ageing. 2003;32:154–163. [PubMed: 12615558]

51.

Morin CM, Colecchi C, Stone J. et al. Behavioral and pharmacological therapies for late-life insomnia. A randomized controlled trial. JAMA. 1999;281:991–999. [PubMed: 10086433]

52.

Kripke DF. Chronic hypnotic use: Deadly risks, doubtful benefit. Sleep Medicine Reviews. 2000;4:5–20. [PubMed: 12531158]

53.

Kripke DF.http://www.DarkSideOfSleepingPills.com2002.

54.

NIOM Sleeping Pills Insomnia and Medical Practice. Washington DC National Academy of Sciences. 1979:1–198.

55.

Consensus Conference Drugs and insomnia The use of medications to promote leep. JAMA. 1984;251(18):2410–2414. [PubMed: 6142971]

56.

National Institutes of Health Consensus development conference statement The treatment of sleep disorders of older people. Sleep. 1991;14(2):169–177. [PubMed: 2094480]

57.

Bunney JrWE, Azarnoff DL, Brown Jr BW. et al. Report of the institute of medicine committee on the efficacy and safety of Halcion. Arch Gen Psychiatry. 1999;56:349–352. [PubMed: 10197830]