Sedative, hypnotic, anxiolytic and opioid medicament use and its co-occurrence with tobacco smoking and alcohol risk drinking in a community sample.(Research article)

Authors: Ulrich John (corresponding author) [1]; Sebastian E Baumeister [1]; Henry V�lzke [1]; Christian Meyer [1]; Sabina Ulbricht [1]; Dietrich Alte [1]

Background

Little is known about (1) sedative, hypnotic, and anxiolytic (SHA), particularly benzodiazepine, and about (2) opioid medicament use behavior in general adult populations. Both substance groups, SHA and opioids taken together (SO) are related to health disorders, including dependence, abuse, intoxication, and withdrawal [1]. Benzodiazepines have been shown to be the most prevalent substance taken among mental health-related drugs in a general population sample aged 65 or older in Canada [2]. In France, a telephone survey study of the general population aged 18 or older revealed that 7.5% used benzodiazepines and altogether 11.5% used drugs against "anxiety, stress, to sleep or to relax [3]."

Generally, there are two alternative approaches to collect data on medicament use in general population studies. The first is to ask for medicament use for specific purposes, e. g. as a sleeping aid. The second approach involves asking for any medicament use irrespective of its purpose. Limitations of asking for medicament use by purpose are that (1) those consumers are not considered who take psychotropic medicines without having knowledge of the indication or effect of the substance and the potential dependence-related consequences of its intake and (2) consumers who might intentionally conceal non-medical use of substances. Asking for any medicament consumption or collecting medicament package data has the advantage to potentially reduce reporting bias, although this approach does not allow to separate non-medical from medical use.

Evidence is needed about the co-occurrence of SO use with tobacco smoking and alcohol risk drinking in the adult general population since data revealed higher morbidity risks for co-occurrent smoking and risk drinking than might be expected from the sum of risks of single substance use [4, 5]. The common addictive nature of the substances makes it useful to describe whether there is co-occurrent use and to what extent and in which subgroups of the sample to a particular high degree. Co-use of these substances may be relevant for explaining common diseases. A co-occurrence might be explained by intentions to regulate body sensations such as upward and downward regulation of mood or use of different drugs for the same end. Alcohol and SO consumption might be interdependent because some of their effects, such as sedation, are similar [cf. [6]]. But SO consumption might be preferred because it is less obvious in the public. In the 1991 to 1993 US National Household Survey on Drug Abuse respondents with daily alcohol use had higher odds for prescription drug nonmedical use than individuals with less than daily alcohol use [7]. In a region of France, a postal survey of the general population aged 18 to 74 revealed that 28.6% smoked, 8.6% had excessive alcohol consumption, and 16.8% used medicaments for sleeping, against tiredness, against nervousness or anxiety [8]. In the US National Survey on Drug Use and Health more current smokers and more individuals with frequent binge drinking in the past two weeks were among benzodiazepine users than among non-users [9].

Contrary to SO use, tobacco smoking and alcohol risk drinking is less frequent in older than in younger adult age according to survey data, both among male and female users [10]. Data from Germany revealed that the proportion of current smokers among ever smokers is lower among older adult age groups than among younger adult age groups [11]. Older respondents in France included lower proportions of smokers but higher proportions of psychotropic medicament users [8]. Major shortcomings of this study included that medicaments were assessed with respect to their purpose; no data about the co-occurrence of different substance use behaviors were provided and that the survey had a response proportion of only 44%. Taken together, we lack evidence about SO use and its co-occurrence with tobacco smoking and alcohol risk drinking from general population samples that include young and older adult age.

The aim of this paper is to present, firstly, prevalence data about SO use stratified by gender, age, education, income, utilization of medical care, and a screening of psychiatric diagnoses. Secondly, the co-occurrence of SO use with current tobacco smoking and alcohol risk drinking will be analyzed. We hypothesized that (1) among smokers and alcohol risk drinkers, particularly among subjects who smoke and additionally drink alcohol in a risky way, more SO users may be found than among individuals who are nonsmokers and non-alcohol problem drinkers, and (2) in young and middle adult age there is a predominance of cigarette smoking and alcohol risk drinking whereas in older adult age SO use is more prevalent than current cigarette smoking and alcohol risk drinking. Accordingly, the use of any of the three substances is expected to be equally distributed over young, middle and older adult age.

Methods

Sample

The sample comprised all subjects from a random population sample drawn in a north-eastern German region. The sample is representative for the population of 212157 residents aged 20 - 79 in an area of West Pomerania in North-Eastern Germany. There were 7008 eligible individuals within this age range, stratified by 5-year strata and gender from residents' registration office files, in which every resident has to be registered by law, in three cities of 20,000 …