Abstract
ABSTRACT
Background
Wrist‐worn transdermal alcohol concentration (TAC) sensors enable continuous, objective measurement of alcohol use. This study validated a newer generation alcohol sensor (BACtrack Skyn) in a laboratory setting among persons with and without HIV. We examined: (1) correspondence between TAC and breath alcohol concentration (BrAC) overall and by HIV status and (2) factors associated with TAC (peak and area under the curve [AUC]) controlling for peak BrAC.
Methods
Twenty‐four participants (32% persons with HIV [PWH], 55% male,
M
age
= 34.4) completed 40 laboratory alcohol administration sessions while wearing the sensor and providing BrACs. All participants underwent the same dosing procedures, consuming three 4.5% beers. Correlations, multilevel models, and generalized estimating equations models examined relationships between TAC, BrAC, and individual factors.
Results
In this single‐dose study, peak TAC and TAC AUC significantly correlated with peak BrAC (
r
s
=
0.36–0.39). Correlations between peak TAC and BrAC were similar by HIV status (
r
s
=
0.5) but varied between TAC AUC and peak BrAC (persons without HIV
r
= 0.57, PWH
r
= 0.33), with nonsignificant associations among PWH. Full time‐series analyses showed stronger TAC‐BrAC correlations (
r
= 0.64) with no HIV differences. Age (
b
= 0.01, 95% CI: 0.01, 0.02), race/ethnicity other than non‐Hispanic White (
b
= −0.44, 95% CI: −0.79, −0.08), positive HIV status (
b
= 0.67, 95% CI: 0.24, 1.09), and AUDIT score (
b
= 0.09, 95% CI: 0.04, 0.14) were associated with peak TAC. Positive HIV status (
b =
1.33, 95% CI: 0.48, 2.18), AUDIT score (
b =
0.11, 95% CI: 0.02, 0.19), and BMI (
b
= −0.06, 95% CI: −0.11, −0.00) were associated with TAC AUC.
Conclusion
In light of the fixed dosing procedure, peak‐based correlations reflect comparisons that limit BrAC variability to a narrow band, thereby constraining TAC‐BrAC correlations. TAC and BrAC were significantly correlated, with the strongest concordance observed using full time‐series data. Findings suggest similarities and differences in TAC‐BrAC relationships by HIV status depending on metric. Individual factors including age, race/ethnicity, HIV, and AUDIT score were associated with TAC. Future research should confirm these findings in larger, diverse samples.