# Evaluating alcohol consumption in adolescents and young adults: a meta-analysis of the psychometric properties of measurement instruments

## TL;DR

This study evaluates the reliability of tools used to measure alcohol consumption in young people, finding that some tools are highly reliable and valid.

## Contribution

The study provides a meta-analysis of psychometric properties of alcohol consumption instruments in adolescents and young adults.

## Key findings

- The pooled reliability coefficient of alcohol consumption instruments was α = 0.88, indicating high internal consistency.
- AUDIT and BYAACQ showed the strongest psychometric robustness across cultural contexts.
- CLASS, PRQ, and ACQ-SF-R are viable alternatives for measuring alcohol-related beliefs and norms.

## Abstract

Early alcohol consumption is a public health concern among young people in most Western countries. However, psychometric evidence on the tools used to assess alcohol consumption and its associated factors is limited. This meta-analysis aimed to evaluate the overall reliability and measurement quality of instruments assessing alcohol consumption, following standardized methodological frameworks.

The protocol was registered in PROSPERO (ID: CRD4202424533078) and followed PRISMA guidelines for systematic review and meta-analysis. The search was conducted in PubMed, Scopus, PsycINFO, and Web of Science using the PECOS strategy. Methodological characteristics of the studies and the reported psychometric properties of the instruments were documented. Reliability estimates and confidence intervals were analyzed using a restricted maximum likelihood random-effects model. Methodological quality and risk of bias were evaluated based on the COSMIN checklist for systematic reviews of outcome measurement instruments and the Terwee quality criteria, ensuring the inclusion of studies with acceptable internal consistency (α ≥ 0.70) and appropriate model fit (CFI ≥ 0.90; RMSEA ≤ 0.08).

Twenty-seven studies met the inclusion criteria, reporting Cronbach’s alpha coefficients between 0.71 and 0.98. The pooled reliability coefficient was α = 0.88 (SE = 0.0088; 95% CI = 0.86–0.89), indicating high internal consistency. Significant heterogeneity (I2 = 99.6%) was observed, primarily related to differences in sample size and item number. In addition, according to COSMIN and Terwee assessments the Alcohol Use Disorders Identification Test (AUDIT) and the Brief Young Adult Alcohol Consequences Questionnaire (BYAACQ) demonstrated the strongest psychometric robustness, supported by consistent reliability and validity across cultural contexts.

This meta-analysis provides strong evidence of reliability and adequate structural validity for instruments measuring alcohol consumption in adolescents and young adults. The AUDIT and BYAACQ emerge as the most reliable and valid options, while the CLASS, PRQ, and ACQ-SF-R also represent promising, psychometrically viable alternatives for assessing beliefs, parental norms, and craving associated with alcohol use. These instruments can guide early detection and preventive interventions, supporting more consistent measurement standards in public health and behavioral research.