For some chronic benzodiazepine users, withdrawal can be a long, drawn-out process. A sizable minority, perhaps 10% to 15%^ develop a Apost-withdrawal syndrome,"2 which may linger for months or even years. This syndrome is clearly not a disease entity; it probably represents an amalgam of pharmacological and psychological factors directly and indirectly related to benzodiazepine use. The syndrome includes (1) pharmacological withdrawal symptoms involving the slow reversal of receptor changes directly induced in the brain by benzodiazepines,3,5 and (2) psychological symptoms resulting indirectly from long-term benzodiazepine use, including exposure of poor stress-coping abilities and other personal difficulties. These symptoms merge into a complex clinical picture that may be further complicated by the reappearance of underlying anxiety or depression and possibly also by illunderstood long-term neurological effects of benzodiazepines.4
Thus, the totality of the benzodiazepine withdrawal syndrome is as difficult to define or demarcate as a bout of influenza, which may include the overlapping pathologies of acute viral toxemia, secondary bacterial infection, and prolonged post-viral depression. Nevertheless, an awareness that symptoms may be protracted is important for clinicians supervising benzodiazepine withdrawal; proper management of the post-withdrawal phase can decrease its severity and duration and improve the prospects for eventual recovery.
ACUTE WITHDRAWAL PHASE
The acute "pharmacological" benzodiazepine withdrawal syndrome is classically described as lasting 5 to 28 days, with a peak in severity around 2 weeks post-withdrawal, after which most symptoms return to prewithdrawal levels.6"12 The symptom constellation includes symptoms common to all anxiety states, but some features are unusual and considered to be relatively specific to benzodiazepine withdrawal (Table 1).
However, the duration of this acute phase has probably been underestimated. First, most clinical studies terminate 4 to 8 weeks after withdrawal, and the progress of remaining symptoms is not monitored. Second, most reports do not include the later experience of dropouts, although the reason for dropout is often the continuation of symptoms. Indeed, persistence of high anxiety levels beyond 28 days post-withdrawal is usually interpreted not as a withdrawal effect, but as reemergence of an underlying anxiety state previously controlled by the benzodiazepine12'13 and often results in reinstatement of benzodiazepine treatment. Third, it has been assumed that return to pre-withdrawal symptom levels in those who complete withdrawal represents the end of the withdrawal syndrome.
Some clinical reports do not support these assumptions. Observations of patients followed up for longer periods suggest that, at least in some individuals, typical benzodiazepine withdrawal symptoms, including paresthesia, sensory hypersensitivity, muscle spasms, and tinnitus as well as less specific symptoms such as anxiety, insomnia, and depression, can take 6 to 12 months to subside completely.14'21 A further problem is the interpretation of "baseline" symptoms. Patients presenting for withdrawal often have many "typical" benzodiazepine withdrawal symptoms, as well as high levels of anxiety, even while still taking the drugs.4'15 Although these symptoms may return to pre-withdrawal levels in a matter of weeks following an acute withdrawal peak, follow-up observations show that such symptoms may continue to improve over subsequent months. Even without specific treatment, they may decline to levels well below "baseline," sometimes enabling patients to resume their normal lives after years of incapacity.4,18,20 In a few patients, somatic symptoms persist for years in the absence of psychopathological signs of anxiety or hysteria, suggesting a pharmacological causation.3,4,20
Some Common Acute Benzodiazepine Withdrawal Symptoms
PROTRACTED WITHDRAWAL PHASE
The acute withdrawal phase may merge imperceptibly into a more protracted phase in which symptoms gradually decline but may be punctuated by wave-like recurrences14'20 interspersed with windows of normality that gradually extend in frequency and duration toward eventual, but occasionally incomplete, recovery. From current evidence, symptoms that are most likely to be long-lasting are anxiety and insomnia, depression, various sensory and motor phenomena, and gastrointestinal disturbances (Table 2).
Anxiety persisting after the acute phase of withdrawal may be partly due to the uncovering of a learning defect caused by the benzodiazepines. These drugs cause cognitive deficits see this issue22) and specifically impair the learning of new skills, including stresscoping strategies. For example, Gray23 and others showed that behavior treatments for anxiety, including those for agoraphobia, are generally ineffective while patients are still taking benzodiazepines, but become more effective when the drugs are stopped. There may be an extended period after benzodiazepine cessation when patients have a decreased ability to cope with stressful situations; full recovery may require the learning of new strategies to replace the years of coping by pharmacological means.
In addition, drug withdrawal may uncover problems in the patient's life that have never been fully addressed. Tyrer24 points out, for example, that the amnesic effects of benzodiazepines may prevent the resolution of personal stresses such as bereavement. Such buried or half-forgotten stresses may have to be faced after withdrawal and may prolong both anxiety and depression.
Hence, persistence or worsening of anxiety after withdrawal does not necessarily imply the reemergence of an anxiety state existing before withdrawal. Indeed, some patients experience major panic attacks and agoraphobia for the first time after withdrawal and may, for a time, develop a more severe degree of anxiety than was present when the drugs were first prescribed. Nevertheless, these symptoms tend gradually to subside over several months even without formal treatment, although the process may be hastened by appropriate psychological support.
Some Protracted Benzodiazepine Withdrawal Symptoms
Benzodiazepines disrupt normal sleep patterns, suppressing slow wave sleep, rapid eye movement sleep, and dreaming. Cessation of benzodiazepines commonly leads to rebound insomnia25 and sometimes to nightmares and other sleep disturbances, including "restless legs," nocturnal myoclonus, and hypnagogic hallucinations. Sleep disturbance may persist as part of a protracted anxiety state, but occasionally as an isolated symptom. Several months may elapse before a normal sleep profile is reestablished.
Although depression is common in longterm benzodiazepine users and may be aggravated by the drugs (see this issue22), it is also a consistent feature of the withdrawal syndrome.6'20 Depressive symptoms may appear for the first time after withdrawal, often some weeks later, and may be severe and protracted for some months. Suicide has been reported in some studies. For example, in a series of 50 consecutive patients undergoing benzodiazepine withdrawal,20 one patient committed suicide, three developed major depressive disorder, and 17 had depression severe enough to require antidepressant medication. It is not clear whether post-withdrawal depression results from a direct pharmacological action of benzodiazepines, such as central serotonin depletion,19 but it responds to antidepressant drugs and usually remits within a few months. There have been no studies to establish whether or not it recurs in later years.
Perceptual and Motor Symptoms
A variety of perceptual and motor symptoms are characteristic of benzodiazepine withdrawal (Table 1). These usually subside during the acute withdrawal phase, but may sometimes be prolonged.
Tinnitus. Tinnitus may initially result from the generalized sensory hypersensitivity seen in early withdrawal, but may persist after the symptoms have disappeared. Busto et al26 describe two cases of tinnitus persisting for 6 and 12 months after benzodiazepine withdrawal, and they mention a third patient who was unable to withdraw because of severe tinnitus at each attempt. Ashton4 reported four cases of intractable unilateral or bilateral tinnitus first appearing during benzodiazepine withdrawal and persisting over several years. Three of these patients had symmetrical bilateral hearing defects, which may have been a causative or aggravating factor. Other case reports have appeared sporadically in the literature. Tinnitus, which is often experienced as being precisely localized, can reach almost intolerable levels. One patient described her tinnitus as "a needle of sound" piercing somewhere deep inside her head.
Paresthesia. Tingling or numbness in the extremities, scalp, or face is common in benzodiazepine withdrawal and in anxiety states and may be associated with hyperventilation. It may persist as continuing "pins and needles"5 or as severe burning pain without demonstrable neurological cause.3
Other perceptual disturbances, including sensations of movement, inner vibration, and bizarre skin sensations, may persist in the absence of any detectable psychopathology.
Motor Symptoms. Increased muscle tension, hyperreflexia, tremor, fasciculation, and muscle jerking are common in early withdrawal. These signs may represent rebound from the myorelaxant effects of benzodiazepines. However, a range of these symptoms may persist for months or years, including tension, weakness, muscle cramps, tremor, shaking attacks, muscle jerks, and blepharospasm.4'5 As with sensory disturbances, longlasting motor symptoms are not necessarily associated with high levels of anxiety or other mood disorder.
Possible mechanisms. Such protracted perceptual and muscular disturbances raise the possibility that benzodiazepines are capable of inducing long-term hyperexcitability of central sensory and motor neural pathways. A pharmacological basis is suggested by the observation that the benzodiazepine receptor antagonist flumazenil can alleviate some of these symptoms.
In an exploratory single-blind trial in 11 patients, Lader and Morton5 found that intravenous infusion of flumazenil, but not saline, brought rapid relief of protracted symptoms (including muscle tension, pins and needles, weakness, muscle cramps or jerks, burning, tremor, or shaking) that had been present for 5 to 42 months after benzodiazepine withdrawal. These symptoms were improved by 27% to 82%. Anxiety, depression, and poor concentration were also alleviated, but this may have been secondary to somatic improvement. The authors considered it unlikely that the effects of flumazenil were due to expectation because there was no response to saline injections. Furthermore, it was notable that the greatest response occurred in patients with the lowest anxiety ratings.
As a practical measure, the use of flumazenil for benzodiazepine withdrawal remains limited because of its rapid elimination, the necessity for intravenous infusion, and the risk of precipitating acute withdrawal reactions. However, these intriguing preliminary results suggest that benzodiazepines can cause longer-lasting pharmacological effects than previously believed. The results are in keeping with animal work showing that flumazenil can, in certain circumstances, prevent benzodiazepine withdrawal reactions in rats and primates.5
There is equivocal evidence that benzodiazepines might cause structural brain damage. A number of computer tomography studies have given contradictory results. However, Schmauss and Krieg27 found a significant dose-related increase in cerebral ventricular size in 17 long-term benzodiazepine users who had never abused alcohol, compared with control subjects. Those who took higher doses of benzodiazepines (above 50 mg diazepam equivalent daily) had the greatest changes, but abnormalities were also present in those who took less than 50 mg diazepam equivalent daily. Other studies have suggested that highdose benzodiazepine abusers may show signs of cerebral atrophy along with neuropsychological impairments, but these changes appear to be largely reversible after a period of abstinence. Present evidence suggests that the risk of cortical atrophy, if it occurs, applies mainly to long-term, high-dose benzodiazepine abusers.
Gastrointestinal symptoms are common during chronic benzodiazepine use and in withdrawal and usually fit into such categories as "nervous diarrhea" or "irritable bowel syndrome," which may be aggravated by hyperventilation.28 Such symptoms may disappear after withdrawal, but are prominent in some patients who develop a post-withdrawal syndrome. Common complaints are of gaseous abdominal distension, lower abdominal pain, and alternating diarrhea and constipation. Symptoms often appear to be exacerbated by certain foods, and patients may be convinced that they have food allergies or intestinal candidiasis despite negative laboratory findings. However, the incidence of food intolerance and pseudo-allergic reactions is reported to be high in chronic hyperventilators, possibly associated with histamine release,28 and the effect of benzodiazepine withdrawal on gastrointestinal function and on immune responses perhaps merits further attention.
IMPLICATIONS FOR MANAGEMENT OF WITHDRAWAL
"Treatment for benzodiazepine dependence should take account of the postwithdrawal syndrome, and therapeutic contact should not be discontinued too early."2 The two essential pillars of a successful benzodiazepine withdrawal strategy are gradual dosage reduction and anxiety management. Of these, dosage reduction is by far the easiest, but long-term psychological support is equally important for successful outcome, particularly for reducing the incidence and severity of postwithdrawal syndromes.
It is generally agreed that dosage should be tapered gradually in long-term benzodiazepine users. The rate of withdrawal should be individually tailored to the patient's lifestyle, personality, environmental stresses, reasons for taking benzodiazepines, and amount of support available. For most patients on therapeutic doses of benzodiazepines, withdrawal is best completed on an outpatient basis. Such an approach, in the patient's own environment, allows time for both pharmacological and psychological adjustments and permits the patient to continue with his normal life while building up alternative coping strategies.
It is usually easier to withdraw from slowly eliminated benzodiazepines, such as diazepam, and it may be advisable (although not always essential) at some stage to transfer patients taking more rapidly eliminated benzodiazepines, such as alprazolam, onto this drug. At present, there is no conclusive evidence that any adjuvant drugs are generally helpful in alleviating withdrawal symptoms. However, in individual cases, antidepressants, beta-adrenoceptor antagonists (for palpitation and tremor), or carbamazepine (for withdrawal from high doses of benzodiazepines) may be indicated. For high-dose benzodiazepine abusers, admission to a hospital may be indicated where drug reduction can be initiated at a faster rate. On reaching more moderate dosage levels, withdrawal in these patients can be continued as for "therapeutic" dose users.
However careful the dosage reduction, patients dependent on benzodiazepines may develop numerous symptoms (Table 1), and a withdrawal plan should include provision for some form of psychological support. The degree of support required varies individually, ranging from simple encouragement and information to formal cognitive, behavioral, or other therapies. Polydrug and high-dose benzodiazepine abusers may need special treatment for drug addiction problems, but anxiety and other withdrawal symptoms are similar to those of therapeutic dose users. Appropriate support should be available not only during dosage reduction, but for a prolonged period afterward.
Patients who have the greatest difficulties in withdrawing, and who are most vulnerable to protracted withdrawal symptoms, are those with high anxiety levels before withdrawal, those with personality disorder or difficulty, and those with continuing life stresses, although none of these factors contraindícate withdrawal in motivated patients. In a few cases, it may be necessary to reinstate benzodiazepines temporarily, preferably for a short period (e.g., a few weeks only), but this measure need not necessarily lead to continued dependence.2
Unfortunately, no treatment has yet proved generally effective for the few patients who develop protracted neurological symptoms, although these tend to lessen in severity over time. For the majority, the outcome of slow benzodiazepine withdrawal coupled with long-term sympathetic support is good, and most patients feel better than when they were taking benzodiazepines.20
SUMMARY AND CONCLUSIONS
In a substantial minority of patients, benzodiazepine withdrawal is followed by a protracted post-withdrawal syndrome lasting many months. Both pharmacological and psychological factors may be involved and the symptoms include anxiety, depression, and a variety of perceptual, motor, and gastrointestinal disturbances. Treatment for benzodiazepine dependence should take into account prolonged symptoms, which may be minimized by gradual dosage reduction and long-term therapeutic contact with appropriate psychological support.
1. Dupont RL, Saylor KE. Sedatives/hypnotics and benzodiazepines. In: Frances RJ, Miller SI, eds. Clinical Textbook of Addictive Disorders. New York: Guilford Press; 1991:69-102.
2. Tyrer P, The benzodiazepine post-withdrawal syndrome. Stress Medicine. 1991; 7:1-2.
3. Nutt DJ. Benzodiazepine dependence: new insights from basic research. In: Hindmarch I, Beaumont G1 Brandon S, Leonard BE, eds. Beiuodiazepin.es: Current Concepts, New York: John Wiley & Sons; 1990:19-41.
4. Ashton H. Protracted withdrawal syndromes from benzodiazepines. JSubstAbuse Treat. 1991; 8:19-28.
5. Lader MH, Morton SV. A pilot study of the effects of flumazenil on symptoms persisting after benzodiazepine withdrawal. J Psychopharmacol. 1992; 6:357-363.
6. Busto U, Sellers EM, Naranjo CA, Cappell HP, Sánchez CM, Sykora K. Withdrawal reaction after long-term therapeutic use of benzodiazepines. N Engl J Med. 1986; 315:654-659.
7. Murphy SM, Owen RT, Tyrer PJ. Withdrawal symptoms after six weeks' treatment with diazepam. Lancet. 1984; 2:1389.
8. Owen RT, Tyrer P. Benzodiazepine dependence: a review of the evidence. Drugs. 1983; 25:385-398.
9. Petursson H, Lader MH. Withdrawal from long-term benzodiazepine treatment. British Medical Journal. 1981; 283:634-635.
10. Petursson H5 Lader MH. Benzodiazepine dependence. British Journal of Addiction. 1981; 76:133-145.
11. Tyrer P, Rutherford D, Higgett T. Benzodiazepine withdrawal symptoms and propranolol. Lancet. 1981; 1:520-522.
12. Tyrer P, Owen R, Dawling S. Gradual withdrawal of diazepam after long-term therapy. Lancet. 1983; 1:1402-1406.
13. Marriott S, Tyrer P. Benzodiazepine dependence: avoidance and withdrawal. Drug Safety. 1993; 9:93-103.
14. Smith DE, Wesson DR. Benzodiazepine dependency. J Psychoactive Drugs. 1983; 15:85-95.
15. Ashton H, Rawlins MD, Tyrer SP. A double blind placebo controlled study of buspirone in diazepam withdrawal in chronic benzodiazepine users. British Journal of Addiction. 1990; 157:232-238.
16. Busto J, Sellers EM, Naranjo CA, Cappell HP, Sánchez CM, Sykora K. Withdrawal reaction after long-term therapeutic use of benzodiazepines. N Engl J Med. 1986; 315:654-659.
17. Tyrer P, Murphy S, Riley P. The benzodiazepine withdrawal symptom questionnaire. J Affect Disord. 1989; 19:53-61.
18. Hallstrom C, Lader M. Benzodiazepine withdrawal phenomena. International Pharmacopsychiatry. 1981; 16:235-244.
19. Olajide D, Lader M. Depression following withdrawal from long-term benzodiazepine use: a report of four cases. Psychol Med. 1984; 14:937-940.
20. Ashton H. Benzodiazepine withdrawal: outcome in 50 patients. British Journal of Addiction. 1987; 82:665-671.
21. Higgitt A, Fonagy P, Toone B, Shine P. The prolonged benzodiazepine withdrawal syndrome: anxiety or hysteria? Acta Psychiatr Scand. 1990; 89:165-168.
22. Ashton H. Toxicity and adverse consequences of benzodiazepine use. Psychiatric Annals. 1995; 25:158-165.
23. Gray JA. The neuro psychology of emotion and personality. In: Stahl SM, Iverson SD, Goodman EC, eds. Cognitive Neurochemistry. Oxford University Press; 1987:171-190.
24. Tyrer P. Choice of treatment in anxiety. In: Tyrer P, ed. Psychopharmacology of Anxiety. Oxford University Press; 1990:255-282.
25. Kales A, Scharf MB, Kales JD. Rebound insomnia: a new clinical syndrome. Science. 1978; 201:1039-1041.
26. Busto U, Fornazzari L, Naranjo CA. Protracted tinnitus after discontinuation of long-term therapeutic use of benzodiazepines. N Engl J Med. 1988; 315:654-659.
27. Schmauss C, Krieg J-C. Enlargement of cerebrospinal fluid spaces in long-term benzodiazepine abusers. PsycholMed. 1987; 17:869-873.
28. Lum LC. Hyperventilation syndromes in medicine and psychiatry: a review. JR Soc Med. 1987; 80:229-231.
Some Common Acute Benzodiazepine Withdrawal Symptoms
Some Protracted Benzodiazepine Withdrawal Symptoms