Figure.

Year-to-year prevalence^a of past-month marijuana use^b, pregnant and non-pregnant women, overall and by age, 2002–2014

^a“Year-to-year” and “Linear predicted” adjusted prevalence estimates are from log-Poisson regressions. Models controlled for race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, and other non-Hispanic minorities), family income ($0–19,999, $20,000–49,999, $50,000–74,999, $75,000+), age (18–25, 26–34, 35–44), education (less than high school, high school, at least some college), year (year was categorical in the “year-to-year” model, and continuous in the “linear predicted model”), pregnancy status, pregnancy*year interaction, covariate*pregnancy interactions and complex survey design. 95% confidence bars are shown for “Year-to-year” adjusted prevalence estimates for the overall group. Percent of variability in dichotomous marijuana use explained by the model with year as a continuous variable was 6% (McFadden’s pseudo-R²); the ratio of the pseudo-R² statistics for the models with year as a continuous vs. categorical variable is 0.98, indicating strong evidence for a linear trend.

^bPast-month marijuana use was defined as responding “within the past 30 days” to the question, “How long has it been since you last used marijuana or hashish?”.

^cData are from the U.S. National Survey on Drug Use and Health (NSDUH). Sample size across all years combined: Pregnant women (n=10,587), non-pregnant women (n=189,923). N’s represent actual number of participants, not weighted values.