Chapter 6
Getting Off Sleeping Pills

As I have explained, because of mortality, cancer, depression, infection, and behavioral risks, I cannot recommend circumstances when anybody should take zolpidem, eszopiclone, zaleplon, temazepam, triazolam, flurazepam, estazolam, quazepam, barbiturates, or diphenhydramine as hypnotics.

The manufacturers generally claim that a person taking only the recommended dosage each night should safely be able to stop the pill immediately. Many experts feel that patients who have been taking higher doses or a modest regular dosage for a long time may need to slowly taper off the medication, reducing their dosage by a small portion every week or two. Even with slow tapering, withdrawal from sleeping pills can cause at least a few nights of insomnia, anxiety (both day and night), tremulousness, and other symptoms. People will have much less difficulty withdrawing from sleeping pills if they first begin CBT-I treatment as described in Chapter 3 above, or obtain CBT-I from a therapist or web site.

It is always recommended that a patient consult the prescribing doctor before discontinuing a prescribed sleeping pill. A doctor’s supervision is particularly important for patients withdrawing from higher doses.

For most patients, it will not be necessary to replace a sleeping pill with any other drug merely for treatment of insomnia. If related illnesses such as depression, anxiety, etc. are involved, an approved medication, CBT-I, psychotherapy, or bright light treatment for those conditions may be needed. Even people with no intrinsic depression or anxiety are likely to become anxious when withdrawing from a sleeping pill. It helps to understand that this anxiety and fear of insomnia is usually a drug withdrawal reaction which will go away in time, often within a day or two, so starting a replacement drug may not be advisable. People withdrawing from sleeping pills may become filled with the idea that they can never do without their pill, when a few days later, they do perfectly well without it.

There are some drugs which could be substituted for those sleeping pills that I have advised discontinuing because of mortality and cancer risks. I do not say that I recommend such substitution. Certainly, I would not recommend substituting in ordinary circumstances, but I recognize that physicians will encounter some patients for whom at least short-term substitution seems a good idea. I do not think that the possible substitutes have been shown to be associated with mortality. The relationship to cancer for these drugs seems to me uncertain.

The most reasonable substitute drugs might be trazodone, Silenor (doxepin 3 or 6 mg.) and melatonin, but I say this without recommending substitution. Trazodone and melatonin are not FDA-approved as hypnotics as of September, 2018.

Trazodone has been shown to be somewhat effective as a hypnotic in low doses (in higher doses, it is an effective antidepressant), but trazodone has worrisome side effects. Trazodone has recently been very popular in the United States for off-label use as a hypnotic, which seems to indicate that patients and doctors like it. I have seen good results myself, but I have also seen some bad side effects such as falls and excessive daytime sedation. Use of trazodone as a hypnotic is not FDA-approved, and little is known about trazodone’s long-term safety.

Silenor in early reports seems to be somewhat effective for maintaining sleep, though of less use for helping people fall asleep. I am not convinced that we have enough experience with use of Silenor to be sure of its safety, and I have not personally seen patients who are doing well with Silenor. It has been claimed that Silenor is lacking in significant side effects at doses of 6 mg. and below, which might be believable, since we formerly prescribed up to 300 mg. of doxepin to treat depression. Time will tell. Silenor is currently FDA-approved as a hypnotic.

Melatonin in an immediate release form sometimes has a benefit in slightly reducing the time to fall asleep, but it is less effective or ineffective in prolonging sleep later in the night, so its benefits for total sleep time are often weak or absent. Melatonin may accelerate sleep onset best when given an hour or more before bedtime. Melatonin is a night-timing drug, not a hypnotic as such. Night-active rodents have the highest melatonin blood concentrations when they are wide awake. There is evidence that melatonin has a variety of minor side effects such as headache and nightmares, and some little-studied effects on the reproductive endocrine system, but little or no evidence in humans of serious side effects. A sustained release melatonin preparation (Circadin) has been approved as a sleeping pill in Europe. Initial published reports suggest that Circadin has a favorable benefits/risks ratio. However, there seems to have been a trend to leave the less favorable studies of sustained release melatonin unpublished.[60]  I confess I am skeptical of drugs whose manufacturers tend not to publish the less favorable studies, although it is a common failing of pharmaceutical manufacturers. As of this writing, sustained release melatonin is not yet FDA approved in the U.S. as a hypnotic. Our research suggested a trend for older women who secreted more natural melatonin to have higher mortality, but this trend was not statistically significant.[61]  I think we need more long-term studies of melatonin safety.

A specific use for melatonin is treating people with delayed sleep phase disorder (nightowls who have trouble falling asleep and trouble getting up in the morning). There is considerable evidence that very low doses of melatonin (50-500 micrograms) may be useful for these patients. The recommended dosage is much lower than the 1-5 mg. (1000-5000 micrograms) usually sold over the counter. As I have mentioned, I agree with the European Committee for Medicinal Products for Human Use (CHMP), which thought that melatonin would have a better benefits/risks ratio than ramelteon.

Suvorexant is an approved hypnotic, but not enough is known about its interactions with other hypnotics for me to recommend its substitution for patients withdrawing from other hypnotics. Suvorexant seems to have particularly dramatic withdrawal effects.[62] 

Endnotes for Chapter 6

60. Mattila, T, et al.  Insomnia medication: do published studies reflect the complete picture of efficacy and safety?  Eur Neuropsychopharmacol. 2011 Jul; 21(7):500-7. [return]

61. Kripke, DF, et al.  Mortality Related to Actigraphic Long and Short Sleep.  Sleep Med. 2011 January; 12(1)Link to a website outside this eBook: 28-33. [return]

62. Kripke, DF  Is suvorexant a better choice than alternative hypnotics? F1000Res. 4, 456. 2015.Link to a website outside this eBook. [return]