Jayson E. Tipp

An evaluation of Type A and B alcoholics

Evaluations of 1539 alcohol-dependent subjects (including 512 women) were carried out in an attempt to replicate the Type A/B dichotomy suggested by Babor et al. (1992). The subjects are participants in the Collaborative Study on the Genetics of Alcoholism (COGA), and each was evaluated using a face-to-face structured interview. Following the procedure of Babor et al. (1992), data were used to create 17 domains, and a k-means clustering method was invoked to generate a two-cluster solution. Thirty-one per cent of the males and 25% of the females fell into the Type B group, with overall R2 of 0.22 and 0.24 for males and females, respectively. The scores in each of the 17 domains and the analyses of the clinical characteristics for Type A and B subjects were, in general, consistent with the earlier onset and more severe course for Type B men and women. The ability of the domains to identify subgroups of alcoholics remained robust even after the exclusion of alcohol dependent subjects with antisocial personality disorder (ASPD) and those with an onset of alcohol dependence before age 25 years. The present analyses suggest that five of the 17 domains might be especially useful in identifying Type A and B groups.

Are Daughters of Alcoholics More Likely to Marry Alcoholics

This study evaluates the relationship between having an alcoholic (i.e., alcohol-dependent) parent and the presence of a spouse with a similar diagnosis. Data relating to 708 men and 708 women, the parents of the questionnaire respondents, revealed that even after controlling for the increased rate of alcohol-dependent spouses among alcoholics, assortative mating appears to be associated with positive family histories of alcoholism. Within this sample, nonalcoholic daughters of alcoholics were more than twice as likely to marry an alcoholic as nonalcoholic daughters of nonalcoholics, irrespective of the alcoholic parents gender. In contrast, in the same sample daughters of alcoholics did not demonstrate a higher rate for having a spouse with another of the more common psychiatric syndromes, a major depressive episode. In this sample, sons of alcoholics did not demonstrate an increased rate for marrying an alcoholic when compared to sons of nonalcoholics.

A Comparison of ICD-10 and DSM-III-R Criteria for Substance Abuse and Dependence

As part of DSM-IV field trials for substance use disorders, 100 inpatients from two psychiatric substance abuse units were interviewed using a modified version of the Substance Abuse Module (SAM) to ascertain substance use diagnoses according to ICD-10 and DSM-III-R criteria. Both criteria sets developed from the theoretical framework presented by Gross and Edwards (1976) and thus, they should demonstrate close concurrence in diagnoses of dependence and abuse/harmful use. The kappa scores obtained in these analyses demonstrate good to excellent agreement on the diagnoses of dependence across substances. There was poor agreement between DSM-III-R and ICD-10 for abuse/harmful use diagnoses. Although there is generally good agreement between DSM-III-R and ICD-10 for substance dependence diagnoses, important differences exist between the two criteria sets both for the diagnoses of abuse and harmful use, and for the diagnosis of marijuana dependence. These differences are primarily due to the inclusion of social problems and repeated use of substances in hazardous situations as DSM-III-R criteria.

A bootstrapped commingling analysis of platelet monoamine oxidase activity levels corrected for cigarette smoking

Monoamine oxidase (MAO) activity levels have been suggested as a possible biological marker for alcohol dependence and abuse, as well as for schizophrenia and other psychiatric conditions. Using platelet MAO activities in the Collaborative Study on the Genetics of Alcoholism data set, we applied bootstrapping methods as a novel way to test for admixture in families. This bootstrapping involved resampling in family units and hypothesis testing of the resampled datasets for commingling in the distribution of MAO activity levels. Prior to commingling analysis, we used linear models to find covariates of greatest effect on MAO activity levels. While an alcoholism diagnosis was significant in men (n = 1151, P < 0.0001) and women (n = 1254, P = 0.0003), the effect lost significance after controlling for cigarette smoking, indicating alcoholism and smoking behavior to be highly confounded. When smoking histories were compared, former smokers had levels (mean = 7.1) closer to those who never smoked (mean = 7.0) than to current smokers (mean = 5.4). Furthermore, current daily smoking and time since smoking cessation were significantly related to MAO levels, indicating smoking probably has a direct effect on MAO levels, rather than the reverse. These results suggest that studies using MAO levels as a biological marker should consider smoking as an important covariate. Finally, admixture was found in MAO levels controlled for smoking and sex, possibly indicating a major genetic locus; this confirms previous evidence for admixture.