Despite years of advisories against the behavior, smoking among pregnant women remains a persistent public health issue in the USA. Recent estimates suggest that 9.4% of women smoke before pregnancy and 7.1% during pregnancy in the USA. Epidemiological research has attempted to pinpoint individual-level and neighborhood-level factors for smoking during pregnancy, including educational attainment, employment status, housing conditions, poverty, and racial demographics. However, most of these studies have relied upon self-reported measures of smoking, which are subject to reporting bias. To more accurately and objectively assess smoke exposure in mothers during pregnancy, we used Bayesian index models to estimate a neighborhood deprivation index (NDI) for block groups in Durham County, North Carolina, and its association with cotinine, a marker of smoke exposure, in pregnant mothers (n = 887 enrolled 2005-2011). Results showed a significant positive association between NDI and log cotinine (beta = 0.20, 95% credible interval = [0.11, 0.29]) after adjusting for individual covariates (e.g., race/ethnicity and education). The two most important variables in the NDI according to the estimated index weights were percent females without a high school degree and percent Black population. At the individual level, Hispanic and other race/ethnicity were associated with lowered cotinine compared with non-Hispanic Whites. Higher education levels were also associated with lowered cotinine. In summary, our findings provide stronger evidence that the socio-geographic variables of educational attainment and neighborhood racial composition are important factors for smoking and secondhand smoke exposure during pregnancy and can be used to target intervention efforts.
Keywords: Bayesian analysis; Neighborhood; Smoke exposure; Socioeconomic status.
© 2022. Society for Prevention Research.