Soyka M, Bottlender R, Möller H-J;
Epidemiological Evidence for a
Low Abuse Potential of Zolpidem; Pharmacopsychiatry 2000, 33:
138 - 141

I read Soyka's publication with great interest. Due to the importance of the subject, I would like to make a few comments. Since the introduction of zolpidem, I have regularly been conducting enquiries on its abuse and dependence potential using databases such as Medline and Embase.

Whenever new findings on zolpidem's abuse and dependence potential are published, I am always surprised at how wide the divergences are between my enquiry results and those of other authors. Soyka has found 15 cases of zolpidem abuse and dependence for the period before 1997; I, on the other hand, found 10 further cases for the same period from the literature. I have another 17 case reports at my disposal referring to the period between 1997 and 2000, the time Soyka's publication was being reviewed. The 1999 WHO Drug Information mentions another 13 cases of abuse, 71 cases of dependence and 36 cases of with- drawal syndrome associated with zolpidem. The many case reports on the psychological, paradoxical, hepatotoxic reactions and deaths since zolpidem's registration have not been taken into account, either.

I am also astonished about Soyka's discussion of the 1997 MMW Newsletter. As far as I know, this newsletter covers a press conference organized by Synthelabo. I am not aware of any data from this having been scientifically processed or published.

No differences can be found from comparing the type and the severity of withdrawal syndromes described in the zolpidem case reports with those described in the literature on benzodiazepines. The doses abusively used (many times the therapeutically recommended dose) are also comparable. The diagnostic criteria for substance dependence as well as the diagnostic abuse criteria according to DSM-III-R and DSM-IV are met by zolpidem and benzodiazepines alike. This all suggests the same abuse and dependence potential of benzodiazepine receptor agonists (benzodiazepines and newer non-benzodiazepines).

Reasons for differences in the prevalence rates might be diverse; these may consist of treated indication - anxiety disorder, sleep disorder; age of the patient; duration and severity of the disease; concomitant diseases; addiction history; taking interval; kind of application; doses; duration of treatment; personality structure of the patient; peristatic factors.

The newer non-benzodiazepines, of which only hypnotics but no anxiolytics are available (alpidem, an anxiolytic of the imidazopyridine substance class, had to be taken off the market due to its severe hepatic side effects) were introduced at a time when all previous health policies were designed to foster a compliant use of benzodiazepine receptor agonists were showing first signs of success. These policies include, above all, the recommendations for the use of lower doses and to limit the time of treatment.

In this context, it is striking that it is always pointed out in almost all review articles and in the summary of product characteristics and patient's information leaflet about zolpidem that the “advantages” of zolpidem will only come to the fore when the drug is used as advised. This information would surely be valid for all benzodiazepine receptor agonists.

In conclusion, I would like to point out that most of the national health authorities in Europe have so far followed the 1993 recommendations of the Commission of the European Communities, requiring the pharmaceutical companies distributing zolpidem and zopiclone to include information on tolerance, dependence, rebound insomnia, psychic and paradoxical reactions and amnesia with the summary of product characteristics and a patient's information leaflet. For benzodiazepines, these recommendations were already valid before 1993.

Correspondence