Heuristic cluster analysis of alcoholics according to biographic and personality features

Abstract

Aims: In spite of the impact of biographical factors on alcoholism and on the therapy of alcoholics, there are few studies that statistically investigate the biographical data of alcoholics. This is the reason why biographical variables in addition to personality variables are taken into consideration in our investigation. Design: At the end of detoxification the “Biographic Questionnaire for Alcoholics” (BIOQUESTAL) was administered. The BIOQUESTAL was used because its biographic scale and its personality scales represent a biographic questionnaire and a personality questionnaire as well. Setting: The patients were consecutively admitted to in-patient treatment. The treatment consisted of clinical management, detoxification from alcohol, and the administration of tranquiliser and clomethiazol, if necessary. What was very important in our procedure was the motivation for long-term treatment in a special clinic for alcoholics. Participants: The random sample consisted of 924 alcoholics (ICD-10: F 10.2). The modal value for age was 38 years. In the sample, 75% was males and 25% females, and 31.7% were single, 42.5% married, 23.5% divorced and 2.3% widowed. Measurements: The BIOQUESTAL is appropriate for the assessment of the biography of alcoholics and for their clinical classification. The dimensions of the scales were constructed by factor analysis, by multidimensional scaling and by cluster analysis of the items. From the data, three oblique rotated factors were extracted: Scale 1: “neuroticism,” 15 items; Scale 2: “favourable versus unfavourable primary socialisation,” 12 items; Scale 3: “unspecific motivation and orientation to future,” 11 items. The reliability of these scales is high. Findings: The five-cluster solution was accepted, yielding an easily interpretable classification. The least impaired cluster is cluster 2 (named: normovalent or syntonous cluster) followed by cluster 4 (named: neurotic aim-related cluster), cluster 1 (named: sociodeprivative cluster), cluster 5 (named: neurotic aimless cluster) and cluster 3 (named: socioneurotic cluster). The syntonous cluster 2 is significantly (F-test; p < 0.001) the eldest cluster and the socioneurotic cluster 3 the youngest one. Female subjects prevail significantly in clusters with higher severity of impairment, i.e. in the socioneurotic cluster 3 and in the neurotic aimless cluster 5. Subjects of the syntonous cluster 2 originate significantly less often from broken homes, whereas subjects of the socioneurotic cluster 3 originate significantly more often from such situations. Subjects of the syntonous cluster 2 were significantly more often brought up by their parents, whereas subjects of the socioneurotic cluster 3 and the neurotic aimless cluster 5 were significantly less often brought up by their parents. Subjects of the syntonous cluster 2 and the neurotic aim-related cluster 4 significantly less often spent their life as singles than the rest of the clusters. Conclusions: The questionnaire is also suitable for the selection of alcoholics and for allocation to appropriate therapy settings. Thus, the reported clusters of alcoholics have not only a descriptive or classificatory importance but also, considering the severity of impairment attributed to the clusters, an impact on differential therapy with regard to the severity of impairment.