Professional Documents
Culture Documents
6 (1999) S399–S405
www.elsevier.com / locate / euroneuro
Abstract
The benzodiazepines are still extensively used in psychiatry, neurology and medicine in general. Anxiety disorder and severe insomnia
are important syndromal indications, but these drugs are widely prescribed at the symptomatic level, resulting in potential overuse. The
official data sheets recommend short durations of usage and conservative dosage. Although short-term efficacy is established, long-term
efficacy remains controversial, as relevant data are scanty and relapse, rebound and dependence on withdrawal not clearly distinguished.
The risks of the benzodiazepines are well-documented and comprise psychological and physical effects. Among the former are subjective
sedation, paradoxical release of anxiety and / or hostility, psychomotor impairment, memory disruption, and risks of accidents. Physical
effects include vertigo, dysarthria, ataxia with falls, especially in the elderly. Dependence can supervene on long-term use, occasionally
with dose escalation. The benzodiazepines are now recognised as major drugs of abuse and addiction. Other drug and non-drug therapies
are available and have a superior risk benefit ratio in long-term use. It is concluded that benzodiazepines should be reserved for short-term
use — up to 4 weeks — and in conservative dosage. 1999 Elsevier Science B.V. All rights reserved.
0924-977X / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved.
PII: S0924-977X( 99 )00051-6
S400 M.H. Lader / European Neuropsychopharmacology 9 Suppl. 6 (1999) S399 –S405
of this improvement was placebo-related, i.e. non-specific. disorder) were treated with either diazepam, dothiepin (a
Similarly, many short-term trials (up to 28 nights) show sedative tricyclic antidepressant), placebo, cognitive / be-
benzodiazepines to be effective treatments for insomnia haviour therapy, or a self-help procedure. All treatments
(Nicholson, 1986). These hypnotics shorten time to sleep were given for 6 weeks and withdrawn by week 10. By 6
onset, usually prolong sleep time (mainly longer-acting weeks, the initial efficacy of diazepam had waned and by
ones, and reduce the number of arousals in the night. Such the end of the study period, patients treated with diazepam
effects can be quantified both with objective EEG and were actually more symptomatic than those on placebo and
other recordings in the sleep laboratory (polysomnogram) the other treatments.
and subjectively with rating scales completed by the Benzodiazepines are used, often at higher dose ranges
patient each morning. Although these two sets of data and for months at a time, in the management of panic
correlate at a disappointingly weak level, the rating of ‘a patients. The panics may be spontaneous (PD) or occur in
good night’s sleep’ usually reflects infrequent nocturnal agoraphobic or social phobic situations. The drug quickly
arousals. The effects generally wane beyond 28 nights and suppresses the panics and anxiety and facilitates the
even before that time. application of non-drug measures such as psychotherapy
and cognitive behavioural therapy (CBT). Efficacy can
3.2. Long-term effects usually be established in the short-term although again
placebo responses can be appreciable (Ballenger et al.,
Controversy surrounds the issue of whether these un- 1988). As with GAD, discontinuation of the benzodiaze-
equivocal short-term benefits as anxiolytic and hypnotic pine can be complicated by the panics recrudescing often
agents continue beyond a few weeks. A further complica- with rebound to higher levels than experienced before,
tion is the problem of relapse on discontinuation, which and / or with withdrawal phenomena (Pecknold et al., 1988;
can be misread as continued efficacy. This must also be Dager et al., 1992). Direct comparisons with other effec-
differentiated from rebound or withdrawal, evidence that tive anti-panic agents such as imipramine, or an SSRI,
the benzodiazepines were acting merely to suppress dis- indicates that such withdrawal after several months is often
continuation effects (Task Force Report of the American an important problem requiring clinical resources for its
Psychiatric Association, 1990). management.
Of course, apparent long-term efficacy might be, and in Studying the long-term efficacy of the hypnotic benzo-
many cases probably is, a combination of them both. But diazepines is objectified because of the availability of the
the implication is that the frequent observation that many polysomnogram, PSG. Fairly consistently, sleep EEG
long-term benzodiazepine users claim continuing benefit measures of hypnotic effect revert to pre-treatment values
cannot be taken at its face value. after 1–4 weeks (Kales and Kales, 1983). Despite this,
One important study invoked chronically anxious pa- most patients claim that some benefit remains and may
tients being treated with benzodiazepines for 6, 14 or 22 insist on continuing medication. This demonstrates the
weeks. They were then transferred to placebo for 18, 10 mismatch between subjective and objective measures. In
and 2 weeks, respectively (Rickels et al., 1983). In some one study, patients given a benzodiazepine overestimated
patients, the switch to placebo was accompanied by their sleep duration by an average of 72 min as compared
withdrawal reactions. The incidence ranged from 43% in with the EEG recordings (Schneider-Helmert, 1988). The
patients who had been taking a benzodiazepine for more benzodiazepine was then abruptly withdrawn, and patients
than a year before entering the study to as low as 5% in the slept poorly but underestimated their duration of sleep by
short-term users. Even in those patients who did not about 1 h. Thus, the symptomatic improvement on a
develop withdrawal reactions, the switch to placebo was hypnotic and worsening on withdrawal are magnified in
usually followed by a worsening of anxiety symptoms. The opposite directions by these subjective discrepancies.
variability between patients is high but these data can be To summarise the benefits, benzodiazepine anxiolytic,
interpreted as showing that long-term use, although proba- antipanic and hypnotic effects are certainly useful, indeed
bly maintaining some efficacy, also involves a definite risk often invaluable, in the short-term, say up to 2 weeks for
of inducing dependence. hypnotic use, 4 for anxiolytic use, and 12 for antipanic
Another study from this group evaluated the long-term effects. But placebo effects and general measures of
treatment of chronic anxiety with the benzodiazepine support and reassurance are important elements in those
clorazepate as compared with the effects of the non- responses. Beyond these periods, efficacy wanes but to a
benzodiazepine, buspirone (Rickels et al., 1988). After variable extent, both patient-to-patient, and probably
double-blind placebo substitution, the clorazepate-treated among the individual benzodiazepines. Some patients
patients developed increased anxiety levels and some had develop tolerance with the possibility of rebound or a full
typical withdrawal syndromes: no such phenomena were withdrawal syndrome on discontinuation. The efficacy of
detected in the buspirone-treated group. the benzodiazepine as a symptomatic remedy then be-
In a more complex study (Tyrer et al., 1988), 210 comes tangled up with its effectiveness in suppressing
psychiatric outpatients (71 GAD; 74 PD; 65 dysthymic possible withdrawal reactions on discontinuation. The
S402 M.H. Lader / European Neuropsychopharmacology 9 Suppl. 6 (1999) S399 –S405
contribution of these constituents to the apparent therapeu- pine. Tasks that are ‘overlearned’ such as mental arith-
tic actions of the benzodiazepine varies among patients, metic are less affected, if at all. Effects are strongly
and at different phases in the course of treatment. Careful dose-related and also vary from drug to drug.
clinical questioning and observation is needed to dis- On repeated dosing, normal subjects show the well-
tinguish the beneficial elements of benzodiazepine treat- known phenomena of tolerance so that impairments wane
ment from the potential problems of rebound and with- over a week or so. The situation in patients is more
drawal. complex. Anxious or insomniac patients are already im-
paired in psychological performance because of the effects
of the disorders themselves: anxiety and sleeplessness
4. Risks interfere with attention, concentration and motivation
(Bond et al., 1974). Consequently, by reducing anxiety and
These will be reviewed under several headings. insomnia the anxiolytic drug tends to improve psychologi-
cal performance. This improvement outweighs any drug-
4.1. Psychological effects related impairment except when high doses are given.
Exceptions to this include the elderly who are typically
The psychological effects of the benzodiazepines can be sensitive to psychomotor-impairing drugs and some tests
divided into the subjective, behavioural, psychomotor and of very skilled performance where impairments tend to be
cognitive. Subjectively, the benzodiazepines reduce anxie- both detectable and persistent. Thus, withdrawal from
ty and induce sleepiness, torpor and relaxation. This is the long-term benzodiazepine use may be followed by some
wanted effect when the medication is taken as a hypnotic improvements.
in a single dose at night to induce sleep but is an unwanted
side-effect when administered during the day for general-
ised or panic anxiety. Sedation occurs in about a third of 4.3. Cognitive effects
people given an anxiolytic benzodiazepine (Salzman,
1992), and euphoria may be induced in some. These In memory tasks, benzodiazepines disrupt the consolida-
effects are most noticeable during the first week but lessen tion process in semantic (verbal) memory whereby materi-
until the patient can detect no sedation at all. The severity al in short-term stores is transferred to long-term stores
of the sedation depends on dose, individual susceptibility (Curran, 1991). Thus, an individual given a benzodiaze-
and also on the particular benzodiazepine: diazepam, for pine can remember immediate information and that re-
example, is more sedative than lorazepam or oxazepam. membered before the benzodiazepine was initiated but may
Long-acting hypnotics are likely to produce residual have difficulty recalling material given after taking the
sedative effects for much of the next day. benzodiazepine was given (see Table 1). As with psycho-
Although uncommon, paradoxical effects can present a motor tests, single-dose studies in normals detect these
major management problem (Dietch and Jennings, 1988). effects much more readily than repeated dose studies in
The patient may become more anxious, panicky or sleep- patients. Moreover, tolerance to these effects may be
less rather than less. Abnormal affects may develop such incomplete so that memory difficulties can persist.
as hostility or depression; antisocial behaviour may super- Periods of total amnesia may follow the use of a
vene with rare cases of violence to persons or property. benzodiazepine, particularly when taken in high dose and /
The interaction with alcohol is particularly dangerous. or in combination with alcohol. This type of problem with
Uncharacteristic acts such as petty theft or sexual im- triazolam forced a reevaluation of its risk / benefit ratio and
proprieties may occur, or the patient break down into strong recommendations towards conservative dosage.
uncontrollable weeping. Reduction in dosage or total Memory problems in the elderly taking a benzodiazepine
discontinuation of the benzodiazepine is usually needed to may lead to a mis-diagnosis of dementia (Ancill et al.,
remove the reaction. 1987).
Benzodiazepines neither induce nor lessen depression.
The euphoriant effects of a benzodiazepine may appear to
Table 1
ameliorate depression. Also, many depressed patients are Effects of benzodiazepines on memory
anxious so that decreasing the anxiety may allow the
Type of memory Function BZD effect
depressive features greater expression.
Working Short-term: Hold Nil
Episodic Autobiographical: Impairment
4.2. Psychomotor effects Remember
Semantic Shared knowledge: Nil
Benzodiazepines can impair psychomotor functions such Know
as those involving speed and accuracy (Koelega, 1989). Procedural Skills: Know how Nil
Tasks requiring sustained attention and concentration can PRS Representational: Depends on
Identity BZ
be markedly disrupted by administration of a benzodiaze-
M.H. Lader / European Neuropsychopharmacology 9 Suppl. 6 (1999) S399 –S405 S403
5. Other therapies
evidence is accruing that some forms of such combined
A range of other treatments are available, some being therapy produce responses distinctly superior to either
appropriate for long-term treatment. For GAD and PD, treatment alone.
these alternatives include tricyclic antidepressants (TCAs)
such as imipramine and clomipramine, selective serotonin
reuptake inhibitors (SSRIs) like citalopram, fluoxetine, 6. Conclusions
sertraline and paroxetine, and monoamine oxidase in-
hibitors (MAOIs) both non-selective and irreversible, like A wide spectrum of opinions is expressed concerning
phenelzine, and selective and reversible (moclobemide). the benzodiazepines. As this symposium shows this ranges
Such use of antidepressants is formally approved by the from protestations that they are valuable drugs grossly
Regulatory Authorities for panic and phobia indications for underused and therefore depriving suffering patients of
some SSRIs. The advantages of antidepressants are definite effective symptomatic relief to judgements that their risk
efficacy and easy withdrawal, and the disadvantages are benefit ratios as anxiolytics and hypnotics is adverse under
the various types of side-effect profiles which may com- all circumstances. The middle ground, held by most
promise compliance. A meta-analysis concluded that prescribers is that short-term use is acceptable but that
SSRIs are superior to imipramine and alprazolam in questions hang over long-term use.
alleviating panic attacks (Boyer, 1995). Buspirone, a non- Thus:
benzodiazepine, acts on serotonin mechanisms and has
some efficacy in GAD. Cautious initial dosage is essential
‘‘ even if BZs still represent a powerful treatment for
to minimise side-effects such as dizziness; withdrawal is
several psychopathological conditions accompaniedby
almost always uneventful suggesting little or no depen-
acute anxiety and agitation, doubts about their long-
dence potential. Beta-blockers can help patients with
term use in chronic and recurrent mental disorders
performance anxiety with tremor and palpitations.
have arisen. In addition to dependence and withdrawal
For insomniac patients, alternatives include, again, the
phenomena, the presence of chronic subtle toxicity and
antidepressants, with a preference for sedative agents such
interference with underlying psychopathology appears
as amitriptyline and trazodone. Nefazodone is believed to
to suggest that a more careful evaluation of the risk /
have beneficial effects on sleep patterns and may prove to
benefit ratio in the long-term administration of BZs is
be particularly useful to help manage insomnia associated
needed’’ ( Michelini et al., 1996).
with depression. Alcohol is a poor hypnotic as it causes
rebound insomnia later in the night.
Many non-pharmacological treatments are available to Until these questions are satisfactorily addressed, the
help anxiety and insomnia (Durham and Allan, 1993). wise prescriber will limit his prescriptions in number to
Those for anxiety are listed in Table 4 and for insomnia in those patients who are severely anxious or insomniac; in
Table 5. dosage to the lowest effective; and in duration to a few
These interventions are quite effective (Morin et al., weeks rather than months or years.
1994). It is particularly important to appreciate that drug
and non-drug treatments can be given in combination as
well as in succession. Care is needed with this strategy but 7. Summary
manage patients suffering from anxiety, panics or insom- Lader, M., 1994. Benzodiazepines. A risk-benefit profile. CNS Drugs 1,
377–387.
nia.
MacDonald, J.B., 1985. The role of drugs in falls in the elderly. Clin.
Geriatr. Med. 1, 621–632.
Maczaj, M., 1993. Pharmacological treatment of insomnia. Drugs 45,
References 44–55.
Morin, C.M., Culbert, J.P., Schwartz, M.S., 1994. Nonpharmacological
interventions for insomnia: A meta-analysis of treatment efficacy. Am.
Ashton, H., 1994. Guidelines for the rational use of benzodiazepines.
J. Psychiatry 151, 1172–1180.
When and what to use. Drugs 48, 25–40.
Neutel, C.I., 1995. Risk of traffic accident injury after a prescription for a
Ancill, R.J., Embury, G.D., MacEwan, G.W. et al., 1987. Lorazepam in
benzodiazepine. Ann. Epidemiol. 5, 239–244.
the elderly—A retrospective study of the side-effects in 20 patients. J.
Nicholson, A.N., 1986. Hypnotics. Their place in therapeutics. Drugs 31,
Psychopharmacol. 2, 126–127.
164–176.
Ballenger, J.C., Burrows, G.D., Dupont, R.L. et al., 1988. Alprazolam in
Pecknold, J.C., Swinson, R.P., Kuch, K. et al., 1988. Alprazolam in panic
panic disorder and agoraphobia: Results from a multicenter trial. Arch.
disorder and agoraphobia: Results from a multicenter trial. Arch. Gen.
Gen. Psychiatry 45, 413–422.
Psychiatry 45, 429–435.
Bond, A.J., James, D.C., Lader, M., 1974. Physiological and psychologi-
Quality Assurance Project, 1985. Treatment outlines for the management
cal measures in anxious patients. Psychol. Med. 43, 364–373.
of anxiety states, Aust. NZ. J. Psychiatry 19, 138–151.
Boyer, W., 1995. Serotonin uptake inhibitors are superior to imipramine
Ray, W.A., Fought, R.L., Decker, M.D., 1992. Psychoactive drugs and the
and alprazolam in alleviating panic attacks: A meta-analysis. Int. Clin.
risk of injurious motor vehicle crashes in elderly drivers. Am. J.
Psychopharmacol. 10, 45–49.
Epidemiol. 136, 873–883.
Calhoun, S.R., Wesson, D.R., Galloway, G.P., Smith, D.E., 1996. Abuse
Rickels, K., Case, G., Downing, R.W. et al., 1983. Long-term diazepam
of flunitrazepam (Rohypnol) and other benzodiazepines in Austin and
therapy and clinical outcome. J. Am. Med. Assoc. 250, 767–771.
South Texas. J. Psychoactive Drugs. 28, 183–189.
Rickels, K., Schweizer, S., Casanalosi, I. et al., 1988. Long-term
Curran, H.V., 1991. Benzodiazepines, memory and mood: A review.
treatment of anxiety and risk of withdrawal. Arch. Gen. Psychiatry 45,
Psychopharmacology 105, 1–8.
444–450.
Dager, S.R., Roy-Byrne, P., Hendrickson, H. et al., 1992. Long-term
Rosenbaum, J.F., 1982. The drug treatment of anxiety. New Engl. J. Med.
outcome of panic states during double-blind treatment and after
306, 401–404.
withdrawal of alprazolam and placebo. Ann. Clin. Psychiatry 4,
Salzman, C., 1992. Behavioural side effects of benzodiazepines. In: Kane,
251–258.
J.M., Lieberman, J.A. (Eds.), Adverse Effects of Psychotropic Drugs,
Dietch, J.T., Jennings, R.K., 1988. Aggressive dyscontrol in patients
Guilford Press, New York, pp. 139–152.
treated with benzodiazepines. J. Clin. Psychiatry 49, 184–187.
Schneider-Helmert, D., 1988. Why low-dose benzodiazepine-dependent
Durham, R.C., Allan, T., 1993. Psychological treatment of generalised
insomniacs can’t escape their sleeping pills. Acta Psychiatr. Scand. 78,
anxiety disorder. A review of the clinical significance of results in
706–711.
outcome studies since 1980. Br. J. Psychiatry 163, 19–26.
Serfaty, M., Masterton, G., 1993. Fatal poisonings attributed to benzo-
Fifer, S.K., Mathias, S.D., Patrick, D.L., Mazonson, P.D., Lubeck, D.P.,
diazepines in Britain during the 1980s. Br. J. Psychiatry 163, 386–393.
Buesching, D.R., 1994. Untreated anxiety among adult primary care
Sheehan, M.F., Sheehan, D.V., Torres, A., Coppola, A., Francis, F., 1991.
patients in a health maintenance organization. Arch. Gen. Psychiatry
Snorting benzodiazepines. Am. J. Drug Alcohol Abuse 17, 457–468.
51, 740–750.
¨ C. (Ed.), 1993. Benzodiazepine Dependence, Oxford Universi- Strang, J., Seivewright, N., Farrell, M., 1992. Intravenous and other novel
Hallstrom,
abuses of benzodiazepines: The opening of Pandora’s box? Br. J.
ty Press, Oxford.
Addict. 87, 1373–1375.
Kales, A., Kales, J., 1983. Sleep laboratory studies of hypnotic drugs:
Task Force Report of the American Psychiatric Association, 1990.
Efficacy and withdrawal effects. J. Clin. Psychopharmacol 3, 40–150.
Benzodiazepine dependence, toxicity and abuse, American Psychiatric
Koelega, H.S., 1989. Benzodiazepines and vigilance performance: A
Association, Washington, DC.
review. Psychopharmacology 98, 145–156.
Tyrer, P., Murphy, S., Kingdon, D. et al., 1988. The Nottingham study of
Kroenke, K., Spitzer, R.L., Williams, J.B.W. et al., 1994. Physical
neurotic disorder: Comparison of drug and psychological treatments.
symptoms in primary care. Predictors of psychiatric disorders and
Lancet 2, 235–240.
functional impairment. Arch. Fam. Med. 3, 774–779.
Woods, J.H., Katz, J.L., Winger, G., 1987. Abuse liability of benzo-
Lader, M., 1992. Rebound insomnia and newer hypnotics. Psycho-
diazepines. Pharmacol. Rev. 39, 251–419.
pharmacology 108, 248–255.