This article is part of a certified activity. The complete activity is available at:

Long-term Pharmacologic Management of Insomnia

Andrew D. Krystal, MD, MS

Posted: 9/7/2006


The pharmacologic treatment of chronic insomnia has long been a daunting clinical challenge. For many years, practice was impeded by a number of factors that limited the capacity of physicians to effectively prescribe medications for the durations required for the treatment of many individuals with long-lasting insomnia. These included (1) the absence of studies of the pharmacotherapy of insomnia of more than 3-4 weeks duration[1,2] and (2) clinical guidelines that discouraged the pharmacotherapy of insomnia with the approved agents for more than a few weeks on the basis of assumptions that tolerance (loss of therapeutic effect over time) was inevitable, and chronic insomnia is a symptom of an underlying medical or psychiatric condition and therefore insomnia-specific therapy is superfluous.[3]

As a result, practice has been non-research-based and had obvious limitations. A shortcoming of targeting treatment to a medical/psychiatric condition is that 15% to 20% of insomnia patients with primary insomnia do not have an associated medical/psychiatric condition.[4] The use of sedating antidepressants or antipsychotics "off-label" is problematic due to the lack of data on their use in treating insomnia.[5] The use of approved sedative-hypnotic agents occurs in the face of guidelines discouraging this practice and in the absence of data on longer-term treatment.[6]

Relatively recent studies provided data that are inconsistent with many of the core beliefs of the problematic, longstanding clinical model. As a result, there has been a change in clinical guidelines, and there is a need for a corresponding change in clinical practice.[7] In this column, we review the data that form the basis for the change in perspective and discuss how best to apply these data in clinical practice.

The Need to Treat Chronic Insomnia

A number of observations contradict the symptom model of chronic insomnia. There is evidence that insomnia may predate and increase the risks for future medical/psychiatric disorders.[4,8,9] Furthermore, insomnia often does not resolve with otherwise effective treatment for the associated condition.[10] In these instances, persistent insomnia also predicts relapse of those disorders.[11-13] Even more convincing are studies that have examined administering insomnia treatment along with usual therapy for the associated condition and found that this improves outcome, not only in terms of insomnia, but also in terms of the associated medical/psychiatric conditions.[14-17]

These studies demonstrated that chronic insomnia is not best viewed as a symptom of an underlying medical or psychiatric condition and established the need to administer insomnia-specific treatment. However, they do not resolve another long-standing impediment to the pharmacologic treatment of chronic insomnia: the lack of data on long-term pharmacotherapy and the assumption that medications would inevitably lose their therapeutic effect and be associated with prohibitive risks when administered long term.

Studies of the Long-term Pharmacologic Management of Insomnia

In the last 5 years, the first large-scale, placebo-controlled, open-label studies of pharmacotherapy for durations of more than 3-4 weeks were carried out. These studies, although few in number, contradicted the traditional assumptions of the inevitability of tolerance and adverse effects and provided evidence that the pharmacotherapy of chronic insomnia may be safely and effectively carried out for periods of up to 1 year. They consisted of 2 large, 6-month, placebo-controlled studies of nightly treatment with eszopiclone 3 mg (one of which was followed by 6 months of open-label treatment); one 6-month, placebo-controlled study of intermittent (3-7 nights per week as needed) treatment with zolpidem extended release 12.5 mg; and pooled data from open-label 1-year extension trials of 2 randomized, double-blind trials of nightly treatment with zaleplon 5 and 10 mg in the elderly.[18-22] There was no evidence for tolerance in any of these studies, nor was there evidence for clinically significant health risks occurring with "cold-turkey" discontinuation following up to 1 year of nightly treatment. This set of studies provided the beginning of an empirical foundation for the pharmacologic management of chronic insomnia in clinical practice. Although some important issues have not been addressed by the research carried out to date, with the help of some commonsense strategies, it is possible to apply the newly emerging foundation to improve the quality of clinical care.

Patient Evaluation: Who Should Receive Long-term Pharmacotherapy?

We currently have a limited capacity to predict which patients presenting with insomnia will suffer long-term difficulties with their sleep. Predictive factors that have been identified include greater insomnia severity and the coexistence of psychiatric diagnoses.[23,24] However, the expected duration of insomnia is not necessarily an indicator of the duration of needed pharmacotherapy. It may be that a relatively brief period of pharmacotherapy could eliminate what might have been a long course of sleep difficulties if untreated. This remains an unresolved issue, although there is preliminary evidence in a study of the treatment of insomnia comorbid with depression that 8 weeks of insomnia cotherapy leads to improvement in sleep that is sustained for at least 2 weeks post discontinuation of the anti-insomnia agent, eszopiclone.[17] Because of the lack of a means to determine when to optimally discontinue pharmacotherapy in any individual, it is necessary to reassess patients regularly in order to determine whether treatment should be continued. Each reassessment, such as the decision to initiate treatment, should be based on whether the risks and benefits of treatment outweigh the risks and benefits of other options, which include no treatment or another therapy. In order to get these data, it is necessary to turn to the available literature on the attributes of available therapies as well as carry out a careful medication taper to compare the function of the individual on vs off of his or her pharmacotherapeutic agent.

As such, optimal management can be viewed as deriving from a string of risk-benefit decisions as to whether to initiate or continue treatments. According to this model, no patient is actually selected for long-term pharmacotherapy when initial treatment decisions are made. Lacking empirical guidance, it is only via a series of assessments that regularly confirm the need for ongoing therapy that optimally, long-term pharmacotherapy should be administered. The overall principle that should guide these decisions derives from the emerging empirical foundation: Any individual with significant impairment occurring in association with insomnia, whether that be primary or comorbid or prior to or during a course of treatment, should receive adequate insomnia-specific therapy.

Assessing Risks and Benefits of Therapies

In order to effectively gauge the risks and benefits that are associated with a therapy under consideration for the treatment of insomnia in an individual, placebo-controlled studies are needed that were carried out: with that agent, at the dosage being considered, over the same treatment duration as the intended use, with the planned dosing schedule (nightly vs intermittent use), and in subjects of a comparable age to the patient in question. The current database of available studies that might be used in this process is relatively limited. However, this structure provides a method for applying new research findings as they become available to improving clinical practice. Most helpful in this regard would be studies comparing the effects of available therapies, including studies of pharmacologic and nonpharmacologic therapies.

Further discussion of assessing risks in individuals is merited. The paucity of available data forces practitioners to extrapolate data across dosages, age groups, agents within a class, etc, when assessing risks and benefits. Although this reduces the capacity to make effective assessments in general, assessment of risks should be undertaken with particular caution. Studies of the pharmacotherapy of older adults are particularly lacking, and studies in younger adults are likely to underestimate the adverse effects and problems with drug interactions that will be seen in the elderly.

Another risk factor that should be considered is abuse potential. Although this issue is not dependent on duration of treatment, it is important to consider when instituting therapy. Abuse of benzodiazepine-related insomnia agents is relatively infrequent and is largely limited to those with a history of multisubstance abuse.[25,26] The problem in clinical practice is that it is not possible to identify such individuals definitively among those who present for treatment of insomnia. However, it is often possible to identify who is not an at-risk individual. In this case, because the risk is low, it should not be a consideration in choosing therapy.

Choosing a Dosing Regimen: Nightly or Intermittent Dosing

No data currently exist to indicate when to administer pharmacotherapy nightly vs intermittently. Although intermittent dosing may be preferable because of less drug exposure and cost, it is not clear whether it provides therapeutic advantages. Data in regard to intermittent dosing for up to 6 months with zolpidem extended release suggested that this can be an efficacious and safe practice.[18] From a practical point of view, intermittent dosing is only possible in the subset of individuals who can predict when they are going to have a bad night before going to bed, who have relatively infrequent sleep difficulties, and who are not strongly affected by experiencing sleep difficulties on the nights that they chose not to take medication but had problems with their sleep.

Follow-up, Medication Tapering, and Discontinuation

Because of the lack of data on optimal methods for scheduling follow-up and tapering medications, it is necessary to develop an individualized plan that should take into account symptom severity and ongoing function. In order to achieve successful treatment, it is necessary to decide on a plan before initiating therapy that should include the agreement that periodic tapers will occur, as well as when and how they will occur. In all cases, even though these are unlikely, patients should be warned about the potential for a loss of therapeutic effect or for withdrawal symptoms, and a taper should be instituted. Although there are no data to guide when to institute discontinuation, a trial of tapering every 2-4 months has been suggested.[3] The decisions about risk and benefit should only be made following resolution of whatever transient effects that the patient may experience in the first few days following discontinuation. At some point, hopefully, the decision will be made that the patient does as well or better without treatment with their medication, which will constitute the end of their pharmacotherapy.


Although the pharmacologic management of chronic insomnia has long been problematic, relatively recent data have begun providing an empirical foundation from which to improve clinical practice. These findings can best be applied in practice by using them to improve the effectiveness of clinical decision making by providing the capacity to define the risks and benefits of the range of alternative therapies. Most importantly, they suggest the need to target treatment specifically to individuals in whom insomnia occurs in association with significant impairment.


  1. Krystal AD. The changing perspective of chronic insomnia management. J Clin Psychiatry. 2004;65:20-25.
  2. Nowell PD, Mazumdar S, Buysse DJ, Dew MA, Reynolds CF III, Kupfer DJ. Benzodiazepines and zolpidem for chronic insomnia: a meta-analysis of treatment efficacy. JAMA. 1997;278:2170-2177. Abstract
  3. NIH. Consensus conference. Drugs and insomnia. The use of medications to promote sleep. JAMA. 1984;11:2410-2414.
  4. Ford DE, Kamerow DB. Epidemiologic study of sleep disturbance and psychiatric disorders: an opportunity for prevention? JAMA. 1989;262:1479-1484.
  5. Walsh JK, Schweitzer PK. Ten-year trends in the pharmacological treatment of insomnia. Sleep. 1999;22:371-375. Abstract
  6. Balter MB, Uhlenhuth EH. New epidemiologic findings about insomnia and its treatment. J Clin Psychiatry. 1992;53(suppl):34-39. Abstract
  7. National Institutes of Health. National Institutes of Health State of the Science Conference statement on Manifestations and Management of Chronic Insomnia in Adults, June 13-15, 2005. Sleep. 2005;28:1049-1057. Abstract
  8. Breslau N, Roth T, Rosenthal L, Andreski P. Sleep disturbance and psychiatric disorders: a longitudinal epidemiological study of young adults. Biol Psychiatry. 1996;39:411-418. Abstract
  9. Livingston G, Blizard B, Mann A. Does sleep disturbance predict depression in elderly people? A study in inner London. Br J Gen Pract. 1994;44:445-448.
  10. Nierenberg AA, Keefe BR, Leslie VC, et al. Residual symptoms in depressed patients who respond acutely to fluoxetine. J Clin Psychiatry. 1999;60:221-225. Abstract
  11. Currie SR, Clark S, Rimac S, Malhotra S. Comprehensive assessment of insomnia in recovering alcoholics using daily sleep diaries and ambulatory monitoring. Alcohol Clin Exp Res. 2003;27:1262-1269. Abstract
  12. Brower KJ, Aldrich MS, Robinson EAR, Zucker RA, Greden JF. Insomnia, self-medication, and relapse to alcoholism. Am J Psychiatry. 2001;158:399-404. Abstract
  13. Reynolds CF III, Frank E, Houck PR, et al. Which elderly patients with remitted depression remain well with continued interpersonal psychotherapy after discontinuation of antidepressant medication? Am J Psychiatry. 1997;154:958-962.
  14. Walsh J, Muehlbach MJ, Lauter SA, Hilliker NA, Schweitzer PK. Effects of triazolam on sleep, daytime sleepiness, and morning stiffness in patients with rheumatoid arthritis. J Rheumatol. 1996;23:245-252. Abstract
  15. Fava M, McCall WV, Krystal A, et al. Eszopiclone co-administered with fluoxetine in patients with insomnia co-existing with major depressive disorder. Biol Psychiatry. 2006;59:1052-1060. Abstract
  16. Soares C, Rubens R, Caron J, et al. Eszopiclone treatment during menopausal transition: sleep effects, impact on menopausal symptoms, and mood. Sleep. 2006;29:A239.
  17. Krystal A, Fava M, Rubens R, et al. Evaluation of eszopiclone discontinuation after co-therapy with fluoxetine for insomnia with co-existing depression. J Clin Sleep Med. In press.
  18. Erman M, Krystal A, Zammit G, Soubrane C, Roth T; Zolong Study Group. Zolpidem extended-release 12.5-mg, taken for 24 weeks "as needed" up to 7 nights/week improves subjective measures of therapeutic global impression, sleep onset, and sleep maintenance in patients with chronic insomnia. Int J Neuropsychopharmacol. 2006;9(suppl2):S251.
  19. Krystal AD, Walsh JK, Laska E, et al. Sustained efficacy of eszopiclone over six months of nightly treatment: results of a randomized, double-blind, placebo controlled study in adults with chronic insomnia. Sleep. 2003;26:793-799. Abstract
  20. Krystal A, Walsh J, Rubens R, et al. Efficacy and safety of six-months of nightly eszopciclone in patients with primary insomnia: a second long-term placebo-controlled study. Sleep. 2006;29:A249-A250.
  21. Ancoli-Israel S, Richardson GS, Mangano RM, Jenkins L, Hall P, Jones WS. Long-term use of sedative hypnotics in older patients with insomnia. Sleep Med. 2005;6:107-113. Abstract
  22. Roth T, Walsh JK, Krystal A, Wessel T, Roehrs TA. An evaluation of the efficacy and safety of eszopiclone over 12 months in patients with chronic primary insomnia. Sleep Med. 2005;6:487-495. Abstract
  23. Hohagen F, Rink K, Kappler C, et al. Prevalence and treatment of insomnia in general practice. A longitudinal study. Eur Arch Psychiatry Clin Neurosci. 1993;242:329-336. Abstract
  24. Benca RM, Ancoli-Israel S, Moldofsky H. Special considerations in insomnia diagnosis and management: depressed, elderly, and chronic pain populations. J Clin Psychiatry. 2004;65(suppl8):26-35.
  25. Hajak G. A comparative assessment of the risks and benefits of zopiclone: a review of 15 years' clinical experience. Drug Saf. 1999;21:457-469. Abstract
  26. Roehrs T, Pedrosi B, Rosenthal L, Roth T. Hypnotic self administration and dose escalation. Psychopharmacology (Berl). 1996;127:150-154. Abstract

This article is part of a certified activity. The complete activity is available at: