Cigarette smoking during pregnancy: independent associations with religious participation.

OBJECTIVE Data from a national health survey were used to test the hypothesis of a negative association of smoking in pregnancy and three measures of religious participation and importance. METHODS The 2002 National Survey of Family Growth included 2395 women aged 15 to 44 years with a history of at least one pregnancy in the five years before interview. An association between religious participation and cigarette smoking during the last pregnancy was assessed in bivariate and multivariate analyses. RESULTS The rate of smoking during the last pregnancy was 4% (95% confidence limit [CL] 2-7%) among those who attended service more than once weekly and 24% (95% CL 20-30%) among those who never attended (chi-square 68, P < 0.0001). In logistic regression models compared with those who never attended, those attending once a week or more were only one-fifth as likely to smoke during pregnancy among European Americans (adjusted odds ratio with 95% confidence limits of 0.22, 0.12-0.39) and Hispanics (0.28 95% CL, 0.11-0.73), and one-half as likely to smoke among African Americans (0.53 95% CL, 0.16-1.69). Significant associations were also observed for affiliation and importance of religion. CONCLUSION The frequency of attendance at religious services, affiliation, and importance were independently inversely associated with smoking during pregnancy in American women. The strength of these associations varied among ethnic groups.