We investigated agreement between self-reported prenatal alcohol exposure (PAE) and objective meconium alcohol markers to determine the optimal meconium marker and threshold for identifying PAE.

Meconium fatty acid ethyl esters (FAEE), ethyl glucuronide (EtG), and ethyl sulfate (EtS) were quantified by LC-MS/MS in 0.1 g meconium from infants of Safe Passage Study participants. Detailed PAE information was collected from women with a validated timeline follow-back interview. Because meconium formation begins during weeks 12–20, maternal self-reported drinking at or beyond 19 weeks was our exposure variable.

Of 107 women, 33 reported no alcohol consumption in pregnancy, 16 stopped drinking by week 19, and 58 drank beyond 19 weeks (including 45 third-trimester drinkers). There was moderate to substantial agreement between self-reported PAE at ≥19 weeks and meconium EtG ≥30 ng/g (κ = 0.57, 95% CI 0.41–0.73). This biomarker and associated cutoff was superior to a 7 FAEE sum ≥2 nmol/g and all other individual and combination marker cutoffs. With meconium EtG ≥30 ng/g as the gold standard condition and maternal self-report at ≥19 weeks' gestation as the test condition, 82% clinical sensitivity (95% CI 71.6–92.0) and 75% specificity (95% CI 63.2–86.8) were observed. A significant dose–concentration relationship between self-reported drinks per drinking day and meconium EtG ≥30 ng/g also was observed (all P < 0.01).

Maternal alcohol consumption at ≥19 weeks was better represented by meconium EtG ≥30 ng/g than currently used FAEE cutoffs.