Dependent and problem drinking over 5 years: a latent class growth analysis.

Area of Science:

• Alcohol research
• Health services research
• Longitudinal studies

Background:

• Understanding the long-term trajectory of problematic drinking is crucial for effective alcohol research and health services.
• The stability and change in drinking patterns, particularly heavy drinking, have significant theoretical and applied implications.

Purpose of the Study:

• To apply latent class growth modeling to survey data to identify prototypical longitudinal drinking patterns among dependent and problem drinkers.
• To uncover distinct trajectories of alcohol consumption over a five-year period.

Main Methods:

• Utilized latent class growth modeling on five years of survey data from individuals with alcohol dependence and problem drinking.
• Analyzed baseline factors and significant influences on drinking patterns, including social networks and treatment history.

Main Results:

• Identified five distinct longitudinal drinking patterns: early quitters, light/non-drinkers, gradual improvers, moderate drinkers, and heavy drinkers.
• Baseline factors such as drug severity, dependence symptoms, and marital status were significant predictors.
• Social network characteristics, past treatment, and medical contacts influenced drinking trajectories, with heavy drinkers having larger heavy drinking and drug-using social networks.

Conclusions:

• Latent class growth modeling effectively identified distinct, prototypical longitudinal drinking patterns.
• Findings underscore the heterogeneity of drinking trajectories and the importance of social and clinical factors in shaping these patterns.
• The study supports the utility of semi-parametric, group-based approaches for analyzing complex alcohol-related behaviors.