Current issues of ACP Journal Club are published in Annals of Internal Medicine


Cognitive behavior therapy, temazepam, or both improved short-term outcomes for older adults with persistent insomnia

ACP J Club. 1999 Sept-Oct;131:35. doi:10.7326/ACPJC-1999-131-2-035

Source Citation

Morin CM, Colecchi C, Stone J, Sood R, Brink D. Behavioral and pharmacological therapies for late-life insomnia. A randomized controlled trial. JAMA. 1999 Mar 17:281:991-9.



In older adults with persistent insomnia, is cognitive behavior therapy (CBT) or temazepam, alone or together, effective for improving sleep patterns and habits?


Randomized, placebo-controlled, 8-week trial with follow-up at 3, 12, and 24 months.


The outpatient clinic of a Canadian academic medical center.


78 older adults (mean age 65 y, 64% women, 90% white, average duration of insomnia 17 y) were studied. Inclusion criteria were age ≥ 55 years, sleep onset or maintenance insomnia, duration of insomnia > 6 months, and ≥ 1 negative effects of insomnia during waking hours. Exclusion criteria were insomnia caused by disease or medication, sleep apnea, regular use of hypnotic or psychotropic medication, current psychotherapy, presence of major depression or other serious psychiatric conditions, or confirmed cognitive impairment. Follow-up ranged from 92% at 8 weeks to 63% at 2 years.


18 adults were allocated to CBT alone, 20 to temazepam alone, 20 to both, and 20 to placebo. CBT was designed to change habits of and beliefs about sleep by using weekly 90-minute small-group sessions with behavioral, cognitive, and educational components. Individual action plans were implemented. Temazepam therapy was started at 7.5 mg/night and was gradually increased to 30 mg/night as needed. All treatments were given for 8 weeks.

Main outcome measures

Sleep diary reports, polysomnography, and sleep impairment index.

Main results

Groups receiving CBT initially improved more than those receiving temazepam or placebo, as measured by sleep diaries (Table). Polysomnography showed that sleep improved in the short term in all active treatment groups; the combined therapy group improved more than the others. CBT alone maintained improvement in sleep measures better than combined treatment, which was better than temazepam alone, although follow-up beyond 8 weeks was < 80%.


Cognitive behavior therapy, temazepam, or both improved short-term outcomes in older adults with persistent insomnia.

Source of funding: National Institute of Mental Health Clinical Research Center.

For correspondence: Dr. C.M. Morin, Université Laval, Ecole de Psychologie, Pavillion FAS, Sainte-Foy, Quebec G1K 7P4, Canada. FAX 418-656-5152.

Table. Older adults with persistent insomnia who achieved > 85% sleep efficiency (time asleep/time in bed) after 8 weeks of cognitive behavior therapy (CBT), temazepam, or both*

Group Treatment Placebo RBI (95% CI) NNT (CI)
CBT 55.6% 22.2% 150% (3.8 to 561) 3 (2 to 96)
Temazepam 47.1% 22.2% 112% (-17 to 477) Not significant
Both 68.4% 22.2% 208% (36 to 689) 3 (2 to 7)

*Abbreviations defined in Glossary; RBI, NNT, and CI calculated from data in article.


This exciting, exquisitely done study from an academic medical center has several implications for front-line clinicians. First, many older adults with sleep problems have underlying clinical problems that should be treated directly: medical disease states, medication that causes insomnia, depression and other psychiatric disorders, sleep apnea or periodic limb movements during sleep, and inability to decrease the dose of a presently taken hypnotic. These patients (85 of 168 screened) were excluded from this study.

Second, long-term control of later-life sleep problems is probably best achieved through a psychological educational approach rather than a pharmacologic approach. Although this implication is based on results with < 80% follow-up, I feel it is sensible and justified.

Third, short-term (< 2 mo) control of insomnia may be treated with similar efficacy with educational efforts, medicines, or both. However, this trial was not large enough to address important concerns about benzodiazepine therapy, including falls, cognitive impairment, and dependence.

One shortcoming is that the extensive CBT classes used in the study are not generally available. Excellent teaching of sleep hygiene has long been encouraged by most texts as the preferred method for promoting healthy sleep. An aid for teaching sleep hygiene is the patient education page that appears in the same issue of JAMA (1); this aid, the article itself, and an accompanying editorial (2) are also available on the AMA's Web site ( Unfortunately, substituting the teaching of sleep hygiene for CBT and expecting similar results involves a leap of faith. I would like to see a parallel study using an abbreviated form of CBT, such as the teaching of sleep hygiene. This teaching would probably be more practical for busy clinicians who are not psychiatrists. In the meantime, we can discuss the common-sense principles of CBT with our patients as outlined in the article.

Robert Lebow, MD
University of Massachusetts-HarringtonSouthbridge, Massachusetts, USA


1. JAMA patient page: insomnia. JAMA. 1999;281:1056.

2. Reynolds CF 3d, Buysse DJ, Kupfer DJ. Treating insomnia in older adults: taking a long-term view [Editorial]. JAMA. 1999;281:1034-5.