# Participant ascertainment is differentially related to phenotypic characteristics and alcohol‐related genetic liability in a sample with severe alcohol use disorder

## TL;DR

This study shows that how people with severe alcohol use disorder are recruited affects their reported traits and genetic risk factors.

## Contribution

The study reveals that recruitment methods influence phenotypic and genetic characteristics in severe AUD samples.

## Key findings

- Clinic-recruited participants reported more severe alcohol-related outcomes and antisocial behaviors.
- Online participants reported more depression symptoms and had different genetic risk profiles for AUD.
- Polygenic risk score differences were observed only in participants of European descent.

## Abstract

Alcohol use disorder (AUD) is a common substance use disorder associated with a range of sociodemographic, behavioral, and genetic factors. The current study characterizes variation in such factors as a function of ascertainment strategy in a sample of individuals with a lifetime history of severe AUD.

Participants (N = 10,804) were recruited through substance use treatment facilities or through online outreach/advertisement in the United States and completed a survey that assessed a range of alcohol‐related variables, sociodemographics, psychopathology, and personality. Participants were asked to provide a saliva sample for DNA extraction and genotyping. Participants' survey responses and their polygenic risk for multiple alcohol outcomes were compared as a function of ascertainment strategy (“clinic” vs. “online”) using t and chi‐square tests.

Ascertainment strategy was significantly associated with many phenotypic variables, although effect sizes were generally small. In general, clinic participants reported more adverse outcomes, such as higher AUDIT‐C scores, longer duration of alcohol problems, antisocial behavior symptom counts, and four of five impulsivity facets. However, online participants reported more problems with depression. Polygenic risk scores differed by ascertainment strategy only for participants of European descent. Clinic participants' scores were higher for AUD, AUDIT‐C, drinks per week, and problematic alcohol use. The groups' scores did not significantly differ for typical maximum drinks in 24 h.

Individuals with severe AUD exhibit heterogeneity across many risk domains, particularly for alcohol‐related measures. This heterogeneity can be captured through the ascertainment of study participants via diverse modalities, improving representativeness and potentially facilitating gene identification efforts.

Using multiple recruitment modalities for studies of individuals with alcohol use disorder can expand the phenotypic and genetic heterogeneity of a sample, potentially making it more representative.