DSM-IV alcohol dependence: a categorical or dimensional phenotype?

Area of Science:

• Psychiatry
• Epidemiology
• Genetics

Background:

• Complex disorders like alcohol dependence require informative phenotypes for accurate etiologic research.
• Categorical variables can obscure information when representing inherently dimensional conditions.
• Investigating the validity of the DSM-IV alcohol dependence diagnosis as a dimensional phenotype is crucial.

Purpose of the Study:

• To examine evidence for linearity and thresholds in associations between the DSM-IV alcohol dependence diagnosis and validating variables.
• To determine if alcohol dependence is better represented as a dimensional or categorical construct.

Main Methods:

• Analysis of current drinkers from the National Longitudinal Alcohol Epidemiologic Survey (NLAES) and National Epidemiologic Survey of Alcohol and Related Conditions (NESARC).
• Utilized validating variables: family alcoholism, early-onset drinking, and alcohol treatment.
• Employed logistic or Poisson regression to model relationships, defining severity by the number of DSM-IV alcohol dependence criteria.

Main Results:

• No evidence supported distinct boundaries between alcohol dependence categories.
• Associations between alcohol dependence and validating variables increased linearly with the number of dependence criteria.
• Dimensional linear predictor models best described the association between dependence criteria and validating variables in both NLAES and NESARC datasets.

Conclusions:

• Large sample sizes in NLAES and NESARC provided robust statistical power for the study.
• While diagnoses are essential for clinical practice, a dimensional indicator of alcohol dependence offers greater utility for research.
• Findings suggest a dimensional approach to alcohol dependence enhances research insights.