Association between moderate alcohol consumption during pregnancy and spontaneous abortion, stillbirth and premature birth: A meta-analysis

OBJECTIVE: To examine the effect of moderate alcohol consumption during pregnancy on the risk of spontaneous abortion, stillbirth and premature birth. DATA SOURCES: A structured MEDLINE search from 1966 to December 1993 using the key words 'alcoholic beverage' or 'alcohol drinking', 'pregnancy outcome' or 'pregnancy complication' and 'analytic studies (epidemiologic)'; a structured search of PsycLit from 1974 to 1994 using the keywords 'alcohol abuse','alcohol drinking attitudes', 'alcohol drinking patterns', 'alcohol intoxication', 'alcoholic beverage' or 'alcoholism' and 'pregnancy', 'prenatal', 'neonatal development' or 'neonatal development'; and a bibliographic search of selected review articles. STUDY SELECTION: Case-control or cohort studies of pregnant human subjects that included a moderate alcohol consumption group defined as consumption of more than two alcoholic drinks per week up to and including two drinks per day and an abstainer group defined as consumption of two or fewer alcoholic drinks per week (alcoholic drink defined as 15 mL or 10 g of absolute alcohol), and that examined the outcome of spontaneous abortion, stillbirth or premature birth. DATA EXTRACTION: Two independent reviewers assessed the methodological quality of each study and abstracted data. The heterogeneity of individual study odds ratios was assessed, and data were pooled using a random effects model. RESULTS: Eight studies were included in the analysis. The trials had a mean methodological quality score of 0.74 (range 0.67 to 0.83, maximum 1.0). Heterogeneity existed among the individual odds ratios for spontaneous abortion (P= 0.007). Significant heterogeneity did not exist for the other outcomes. For the pooled outcomes, the odds ratio for spontaneous abortion was 1.35 (95% CI 1.09 to 1.67), the odds ratio for stillbirth was 0.65 (CI 0.46 to 0.91) and the odds ratio for premature birth was 0.95 (CI 0.79 to 1.15). CONCLUSIONS: Moderate maternal alcohol consumption is associated with an increased risk of spontaneous abortion and a decreased risk of stillbirth, and has no effect on premature birth.