# ZIP code-level alcohol outlet density and nonfatal overdose among people who inject drugs in 22 US metropolitan statistical areas: a multilevel modeling analysis

## TL;DR

This study examines if alcohol outlet density in US cities affects nonfatal overdoses among people who inject drugs, finding no significant link.

## Contribution

The study is the first to investigate alcohol outlet density's impact on overdose risk among people who inject drugs at a ZIP code level across 22 US cities.

## Key findings

- No significant association was found between alcohol outlet density and nonfatal overdose risk among people who inject drugs.
- The relationship did not vary significantly by race or ethnicity among participants.
- Regulating alcohol outlet density alone may not reduce overdose risk for this population.

## Abstract

Alcohol outlet density (AOD) is associated with drinking behaviors and related harms across several populations. As alcohol consumption compounds the depressive effects of opioids to increase overdose risk among people who inject drugs (PWID), this study investigated (1) whether off-premise AOD is associated with the risk of nonfatal overdose among a large sample of PWID, and (2) whether this relationship varies by individual race/ethnicity.

We linked individual-level 2018 National HIV Behavioral Surveillance (NHBS) data with ZIP code-level data on off-premise AOD in 2016 from the US Census Bureau’s ZIP Code Business Pattern survey. NHBS surveys PWID across 22 metropolitan statistical areas (i.e., urban cores and surrounding counties with ≥ 50,000 residents, defined by commuting patterns) using respondent-driven sampling. Hierarchical generalized linear models quantified the association between AOD and nonfatal overdose, overall and by PWID race/ethnicity.

The sample comprised 9,764 PWID who used opioids, 45% of whom were non-Hispanic/Latine White, 38% non-Hispanic/Latine Black, and 18% Hispanic/Latine. Over a quarter (28%) experienced an overdose in the past year. The median outlet density was 4.0 outlets per square mile. The adjusted model did not show a significant relationship between AOD and odds of overdose (OR: 1.01, 95% CI 1.00–1.02), and the relationship remained null in the race/ethnicity interaction model.

Regulating AOD alone may not effectively mitigate overdose risk among PWID. Future research should explore why overdoses among PWID are not sensitive to changes in AOD. Possible explanations to consider are whether (1) AOD is unassociated with alcohol consumption among PWID, or (2) PWID stagger the timing of their alcohol consumption so it does not coincide with opioid consumption.