Method: Data were derived from a population-based study (National Survey on Drug Use and Health) of adult women aged 18-44 in the U.S. between 2005 and 2014. At the time of interview, 7,627 adult respondents reported being pregnant and 52.3% (n=3,987) of the pregnant women were unmarried. For marijuana-specific protective factors, we examined self-reported marijuana disapproval (0=neither approve nor disapprove, 1=somewhat or strongly disapprove), perception of great risk (0=no/slight/moderate risk, 1=great risk), and limited access (0=very/fairly easy or fairly difficult, 1=very difficult or probably impossible). . Additionally, we examined past-year diagnosis of physical health (any respiratory-related disease, any cardiovascular disease, and sexually transmitted disease) and psychological health (i.e. anxiety, depression) correlates. Statistical analyses were conducted in three phases. First, demographic characteristics were examined for married and unmarried women. Second, the trend of past-month marijuana use and protective factors were examined separately for married and unmarried pregnant women, and the significance of the trends were tested using the method outlined by the CDC (2016). Third, logistic regression analyses were conducted to examine the associations between physical/psychological correlates and prenatal-marijuana use while stratifying the sample by marital status.
Results: Our study yielded a number of salient results. First, the increasing prenatal marijuana use trend (AOR=1.08, 95% CI=1.02-1.15) was observed only among unmarried women (5.5% in 2005 to 10.2% in 2014), while a leveling trend (mostly under 1%) was shown among married pregnant women. Second, lower perception of great risk explained the increase in marijuana use among unmarried pregnant women. Third, we found that past-year diagnosis of anxiety disorder (AOR=3.30, 95% CI=1.87-5.82) and depression (AOR=3.85, 95% CI=2.33-6.36) were significantly associated with marijuana use among unmarried pregnant women whereas no significant association was found for married pregnant women. Moreover, no association between physical health correlates and marijuana use was found for both marital groups.
Conclusions and Implications: Given the deleterious consequences of marijuana use during pregnancy, increased attention to reducing its prenatal use among unmarried women is critical. Our findings suggest that more liberal perception of great risks is especially problematic for prenatal marijuana use among unmarried women. Moreover, our study testifies to the need to pay attention to not only the health-risk behaviors among unmarried pregnant women, but also their recent mental health needs.