Melatonin Therapy: From Benzodiazepine-Dependent Insomnia to Authenticity and Autonomy

by Harold J. Bursztajn, M.D.
Archives of Internal Medicine
Volume:159, page: 2393, November 8, 1999

For patients for whom the potential short-term benefits of insomnia relief by benzodiazepine therapy are offset by their specific vulnerability to both acute and long-term adverse effects or dependency, the study by Garfinkel and colleagues [1] using melatonin therapy for discontinuation offers a hopeful adjunct in the weaning process and a sleep-maintenance alternative to benzodiazepine therapy. Alternatives such as melatonin therapy can be also helpful in maintaining a therapeutic alliance by offering a treatment enabling focus instead of a do-or-die crisislike atmosphere that can surround the encounter between a concerned clinician and a vulnerable, anxious patient seeking or already dependent on benzodiazepines. Thus, for clinicians who treat patients who are benzodiazepine-dependent and suffer from insomnia, an informed consent process that offers melatonin therapy as an alternative to benzodiazepine dependence integrates good clinical care with effective risk management by carefully navigating between the Scylla of addiction and the Charybdis of abandonment.

To Sleep Perchance To Dream To Remember Perchance To Be Sleepless

Sleep disturbance has a variety of causes. Although unreported by Garfinkel et al, it is reasonable to hope that the patients included in their study had an adequate diagnostic workup for such physiological causes of disturbed sleep in the elderly as early Alzheimer syndrome. But a complete diagnostic workup also needs to include the far more difficult to diagnose and often overlooked neuropsychiatric and psychosocial causes of disturbed sleep. Still, a managed care–influenced skeptic may ask, "What difference does a comprehensive diagnostic workup make if melatonin treatment can obviate the need for benzodiazepines?" However, a clinically informed pragmatist can answer the skeptic by pointing to the value of not overlooking often neglected yet treatable causes of sleeplessness. Thus, it would be helpful to know whether such a complete workup was undertaken for the patients in the reported study. If not, the workup should still proceed, even for patients who can now sleep with melatonin therapy.

Among the treatable causes of insomnia in some elderly are chronic bereavement and chronic depression (dysthymia), as well as disorders in the posttraumatic stress spectrum and their complications. The social- and work-impairing symptoms of these disorders can include a mixture of hyperarousal; irritability; difficulty in concentration; demoralization; fatigue; and loss of pleasure (anhedonia), including a lack of sexual appetite, obsessive rumination, increased superstitions, phobias, a foreshortened sense of future, and a sense of the meaninglessness of life. Undiagnosed and untreated, these disorders and often their unvoiced and impairing symptoms can persist even in the face of reports of a good night's sleep with benzodiazepine therapy. In the elderly, the above symptoms are all too often and all too easily written off as simply "normal aging" and "this is what happens when you get old."

The reported study also raises the question whether that group of patients who remained dependent on benzodiazepines in spite of the authors' careful efforts at weaning represents a subgroup for which the memory-suppressing properties of benzodiazepines are particularly important. Intrusive memories are a hallmark of a variety of disorders, including posttraumatic stress disorder and pathological grief. On the other hand, some kinds of memory, such as autobiographical memory, are important for a meaningful life. Other kinds of memory, such as procedural memory (eg, how to drive), are essential to many tasks of daily living. The inhibition of each by benzodiazepines is far too high a price to pay for sleep. [2-4] While benzodiazepines have a role to play in treating the acute complications of trauma (acute stress disorder) by providing relief from agitation, its potential inhibition of autobiographical and procedural memory formation and retrieval is a major roadblock in treating patients with posttraumatic stress disorder by helping those who are suffering find renewed meaning and exercise renewed skill in coping with life. [5]

The alternative of using sedating antidepressants, such as trazodone, for sleep can result in dream suppression or suppressed dream recall, although overly vivid and disturbing dreams have been reported as well with other antidepressants. [6] Although disturbing dreams or nightmares are the hallmark of posttraumatic stress disorder, [7] their occurrence is even more widespread. In the Israeli sample of elderly sleepless patients, a relevant question to ask is how many of the patients who remained dependent on benzodiazepines were traumatized Holocaust survivors with continued sleep disturbance that is manifested by ongoing frightening dreams. [8] Benzodiazepine therapy may have been symptomatically helpful for this group by both suppressing the night terrors of stages 3 and 4 of sleep and decreasing the anticipatory anxiety heightened by prior dream recall when falling asleep. However, as noted above, the price of such tranquilization in terms of inhibiting meaningful therapy for chronic posttraumatic stress disorder is high. Even in the course of normal aging, as life continues and the slings and arrows of outrageous fortune find their mark, what we seek to set aside during the days of our lives often returns in the form of frightening, sad, and disturbing dreams whose remembrance and anticipation make for sleepless nights. Once heard, such sadness can be often helped by relatively time-limited supportive therapy or by simply helping the patients to restore their sense of social connection. Unfortunately, except for the most psychologically minded patients, the useful information of hidden sadness and trauma contained in dreams will not be volunteered unless specific inquiry is made.

Even patients who are not psychologically minded and who deny persistent sad and anxious moods may offer physicians clues to hidden depressive or anxiety disorders by reporting sad or frightening dreams when questioned tactfully. Adding the simple question "How well do you remember your dreams?" to the standard clinical "review of systems" is a useful and long overdue modification that may be used prior to or at least concurrently with prescribing memory-inhibiting or dream-suppressing agents. Even patients who, out of shame and fear, deny any memory or sleep impairment will feel free to volunteer that they do not remember their dreams as well as they used to or, to the contrary, that sleeplessness has resulted in a greater recall of disturbing dreams. The price of treating one's sleepless nights no longer needs to be dreamless nights or frightening or meaningless days.

Why Not Let Sleeping Dogs Lie?

The major impediment to the clinical evaluation and treatment of insomnia, especially in the elderly, is the attitude of "if it ain't broke, don't fix it," which, in its more extreme form, reflects the fear of raising patient anxieties and fears by well-motivated albeit ill-conceived or poorly timed inquiry. In light of the managed care–influenced shrinkage of time available for talking with patients, the subheading above accurately reflects clinician resistance to additional clinical responsibilities for addressing quality-of-life issues, such as sleeplessness in the elderly. However, there are a variety of clinical, ethical, and legal reasons why managed health care organizations need to reconsider current disincentives to clinicians talking with patients and, when indicated, consulting with psychiatrists.

Clinically, sleeplessness in the elderly, although often unreported, is a common problem. Recent studies suggest that insomnia is rarely a volunteered problem, and patients with "closet" insomnia tend to have a much higher rate of interaction with their health care providers. In one study, approximately half of such patients had disorders in the depressive spectrum. [9] Although such patients are often dismissed as "hypochondriacs" or even derided as "crocks," shying away from inquiring about sleep and dreams saves neither time nor cost and undermines the treatment of these problems. While reliance on benzodiazepine treatment may initially restore sleep for such patients, it does not treat other occult but significant neuropsychiatric disorders or the symptoms or underlying causes of depressive and stress disorders. Not only does benzodiazepine therapy not treat the root causes of depressive and stress disorders, but it may exacerbate depressive states as well as foster dependency.

From an ethical perspective, the prescription of benzodiazepines for insomnia in the elderly can proceed only when accompanied by an informed consent process. This process must begin by informing the patient of the need for a comprehensive diagnostic evaluation and the availability of alternative primary approaches, including using melatonin and promising behavioral interventions. [10] Moreover, the patient must also be informed of the well-recognized risks of benzodiazepine treatment for patients with neuropsychiatric disorders. [11] Even when a patient asks for benzodiazepine therapy or is already dependent on it, an informed consent process needs to be initiated. By this, we do not mean "pro forma" informed consent. [12] Simply signing a form does not count for a meaningful informed consent process. Rather, given the negative impact of anxiety, depression, and sleeplessness on reading, comprehension, and retention, physicians need the time and motivation to talk with patients and review the differential diagnosis, needed workup, treatment alternatives, and relevant benefits and risks each time a benzodiazepine prescription is first written to treat insomnia in the elderly. Once initiated, such an informed consent process needs to continue even after the initial encounter. Thus, when an elderly patient's benzodiazepine therapy is being monitored, physicians need to continue to talk with the patient regarding alternative approaches for insomnia. An added benefit of such an informed consent process is that it allows physicians to perform an ongoing informal Mini-Mental Status Examination–like screen that is sensitive to early signs of diminished capacity or fluctuating competency in the face of progressive cognitive impairment exacerbated by benzodiazepine therapy. [13]

From a legal perspective, the informed consent process is also crucial. In some patients, such as those with Alzheimer disease and sleep disturbance, this process may include referral for a forensic psychiatric consultation to evaluate competency. Where significant diminished capacity exists, recommending that the patient's family petition the court for appointment of a guardian for treatment purposes may be in order. Melatonin itself is a promising treatment for patients with sleep disturbances related to Alzheimer disease, [14] when an increased sleep latency and a decrease in melatonin production seem to coincide.

Patients with early-stage Alzheimer disease represent a particularly vulnerable subpopulation among elderly sleepless persons. On one hand, patients with minimal dementia are particularly sensitive to having their cognitive capacity degraded by sleeplessness and most likely will eventually exhaust their families by "wandering" sleepless. On the other hand, patients with dementia are also especially sensitive to the sedative depressant effects of benzodiazepines. All too many premature nursing home placements as well as accidents, such as slips and falls, occur as a result of patients with early-stage Alzheimer disease being undertreated or overmedicated for insomnia. Given the significant risks of dependency and the likelihood of significant albeit occult neuropsychiatric comorbid conditions in patients who are long-term (more than 3 months of constant use) benzodiazepine users for sleep, a psychiatric referral for this subgroup is most often merited. Also, even prior to the conclusion of ongoing clinical trials of melatonin therapy for sleeplessness related to Alzheimer disease, to the extent that the reported findings are generalizable, any patient with Alzheimer disease who is sleepless and dreamless deserves the opportunity for a therapeutic trial of melatonin. [15, 16]

Sleep, Memory, And Autonomy

Melatonin treatment is no panacea. The current study, although promising, raises the question: Were the patients in this study an especially motivated subgroup? Even so, this does not diminish the importance of the process described, since motivation plus medication are often the needed synergy for addiction recovery. In the managed care era, we see a triage mentality as a major obstacle to appropriately treating quality-of-life symptoms, such as sleeplessness in the elderly. [17] Nonetheless, both good clinical practice and ethics require treatment for sleeplessness beyond the benzodiazepine therapy solution. Untreated or overmedicated, silent yet sleep-deprived or sedated patients often overuse clinical resources, become candidates for accidents, and suffer significant social and work impairment. Thus, not only direct health care costs but disability and work accidents leading to worker's compensation claims can be controlled and sleeplessness can be treated without reliance on benzodiazepines.

Melatonin therapy may not improve all stages of sleep for all sleepless elderly. Moreover, the generalizability of the study by Garfinkel et al must also be considered in light of the relatively small number of patients and clinical reports (ie, that melatonin can increase the vividness of dreams). In at least some recent studies, melatonin therapy was found to decrease sleep latency without improving the total time a patient was asleep. [18-20] However, even reducing the time it takes to fall asleep, without improving total sleep time, can have significant positive meaning to an elderly patient seeking to maintain an internal locus of control and sense of autonomy as age advances. Patients need to be informed that we do not know the long-term consequences of melatonin therapy. Different physician-patient relationships may also result in wide variability in clinical judgment as to benzodiazepine therapy indications. However, with the now-demonstrated usefulness of melatonin therapy to wean elderly sleep-disturbed patients from benzodiazepines and as America ages and "grays," clinicians have the opportunity and health care organizations have the responsibility to facilitate making available as a choice effective benzodiazepine-free treatment for insomnia. Choosing benzodiazepine-free treatment for insomnia can help patients sleep, dream, remember, and continue to have access to both the continuity of autobiographical memories relevant to authenticity and the procedural memory essential to autonomy. [21]


  1. Garfinkel D, Zisapel N, Wainstein J, Laudon M. Facilitation of benzodiazepine discontinuation by melatonin: a new clinical approach. Arch Intern Med. 1999;159:2456-2460.
  2. Buffett-Jerrott SE, Stewart SH, Bird S, Teehan MD. An examination of differences in the time course of oxazepam's effects on implicit vs. explicit memory. J Psychopharmacol. 1998;12:338-347.
  3. Bareggi SR, Ferini-Strambi L, Pirola R, Smirne S. Impairment of memory and plasma flunitrazepam levels. Psychopharmacology. 1998;140:157-163.
  4. Fossen A, Godlibsen OB, Loyning Y, Dreyfus JF. Effects of hypnotics on memory. Pharmacology. 1983;27(suppl 2):116-126.
  5. Kandel ER. Biology and the future of psychoanalysis: a new intellectual framework for psychiatry revisited. Am J Psychiatry. 1999;156:505-524.
  6. Schredl M, Schafer G, Weber B, Heuser I. Dreaming and insomnia: dream recall and dream content of patients with insomnia. J Sleep Res. 1998;7:191-198.
  7. Lansky MR, Bley CR. Posttraumatic Nightmares: Psychodynamic Explorations. Hillsdale, NJ: Analytic Press Inc; 1995.
  8. Kaminer H, Lavie P. Sleep and dreaming in Holocaust survivors: dramatic decrease in dream recall in well-adjusted survivors. J Nerv Ment Dis. 1991;179:664-669.
  9. Hatoum HT, Kania CM, Kong SX, et al. Prevalence of insomnia: a survey of the enrollees at five managed care organizations. Am J Manage Care. 1998;4:79-86.
  10. Morin CM, Colecci C, Stone J, Sood RK, Brink D. Behavioral and pharmacological therapies for late-life insomnia: a randomized controlled trial. JAMA. 1999;281:991-999.
  11. Bursztajn HJ, Brodsky A. Ethical and legal dimensions of benzodiazepine prescription. Psychiatr Ann. 1998;28:121-128.
  12. Bursztajn HJ, Gutheil TG, Cummins B. Legal issues in inpatient psychiatry. In: Sederer LI, ed. Inpatient Psychiatry. 3rd ed. Baltimore, Md: Williams & Wilkins; 1991:379-406.
  13. Bursztajn HJ, Harding HP, Gutheil TG, Brodsky A. Beyond cognition: the role of disordered affective states in impairing competence to consent to treatment. Bull Am Acad Psychiatry Law. 1991;19:383-388.
  14. Liu RY, Zhou JN, Van Heerikhuize J, et al. Decreased melatonin levels in postmortem cerebrospinal fluid in relation to aging, Alzheimer's disease and apolipoprotein E-epsilon 4/4 genotype. J Clin Endocrinol Metab. 1999;84:323-327.
  15. Maurizi CP. Dementia—the failure of hippocampal plasticity and dreams: is there a preventative role for melatonin? Med Hypotheses. 1987;24:59-68.
  16. Uchida K, Okamoto N, Ohara K, Morita Y. Daily rhythm of serum melatonin in patients with dementia of the degenerate type. Brain Res. 1996;717:154-159.
  17. Bursztajn HJ, Gutheil TG, Brodsky A. Ethics and the triage model in managed care hospital psychiatry. Psychiatric Times. 1998;15:33-38.
  18. Jean-Louis G, von Gizycki H, Zizi F. Melatonin effects on sleep, mood, and cognition in elderly with mild cognition impairment. J Pineal Res. 1998;25:177-183.
  19. Lushington K, Pollard K, Lack L, Kennaway DJ, Dawson D. Daytime melatonin administration in elderly good and poor sleepers: effects on core body temperature and sleep latency. Sleep. 1997;20:1135-1144.
  20. Avery D, Lenz M, Landis C. Guidelines for prescribing melatonin. Ann Med. 1998;30:122-130.
  21. Model AH. The Private Self. Boston, Mass: Harvard University Press; 1993.