Alcohol and drug use and attainment of pregnancy preferences in the southwestern United States: A longitudinal cohort study

Abstract

Aims: Addressing design and measurement limitations of prior studies, we examined whether individuals who use alcohol or drugs are less likely to attain their pregnancy preferences over one year.

Design: Longitudinal cohort study.

Setting: Five southwestern US states.

Participants: 2015 individuals ages 15-34 years, capable of pregnancy and non-pregnant at enrollment from 23 primary and reproductive healthcare facilities (2019-2022).

Measurements: Data were collected over 13.5 months. Past-month alcohol, cannabis and other drug use were self-reported at baseline; prospective pregnancy preferences (prior to pregnancy) were reported quarterly using the Desire to Avoid Pregnancy (DAP) scale (range 0-4, 4 representing higher desire to avoid pregnancy); and incident pregnancy was reported every 6 weeks. We used adjusted mixed effects linear regression models to assess associations between baseline substance use and quarterly DAP scores and Cox proportional hazard models to investigate differential attainment of pregnancy preferences by substance use.

Findings: At baseline, 40% (795/1968) of participants reported drinking heavily, 16% (308/1968) reported cannabis use and 3% (53/1968) reported other drug use. Heavy (vs. moderate, low or no) drinking and daily cannabis (vs. no use) were associated with higher desire to avoid pregnancy [coefficient = 0.06, 95% confidence interval (95% CI) = 0.01-0.12], P = 0.02; coefficient = 0.14, 95% CI = 0.07-0.22, P < 0.001], but DAP scores did not differ by other drug use (coefficient = 0.13, 95% CI = -0.07 to 0.33, P = 0.21). 282 participants reported incident pregnancies; of these, 46% (130/282), 29% (81/282) and 25% (71/282) were among participants with low, mid-range and high DAP scores. Among those with high DAP scores, heavy drinking was positively associated with pregnancy [adjusted hazard ratio (aHR) = 1.51, 95% CI = 1.12-2.04, P < 0.01]. Among those with low DAP scores, less than daily versus no cannabis use was associated with pregnancy (aHR = 1.64, 95% CI = 1.13-2.38, P = <0.01). Among those with mid-range DAP scores, heavy drinking was inversely associated with pregnancy (aHR 0.62, 95% CI = 0.42-0.90, P = 0.01). Other drug use was not associated with pregnancy regardless of DAP score.

Conclusions: Heavy drinking, but not cannabis use or other drug use, appears to be associated with elevated pregnancy risk among those who most desire to avoid pregnancy.

Keywords: alcohol use during pregnancy; desire to avoid pregnancy; drug use during pregnancy; pregnancy preferences; reproductive autonomy; substance use.

FIGURE 1

Flow chart of participants. Retained in study refers to the number of participants who remain in the study sample up until or beyond the time point. Completed survey refers to the number of participants who completed the specific survey at the time point. LTFU, lost to follow up.

FIGURE 2

Proportion of low, mid‐range and high desire to avoid pregnancy (DAP) pregnancies with alcohol, cannabis and other drug use at baseline. Gray bars represent missing substance use data.–

FIGURE 3

Predicted 1‐year marginal pregnancy hazards, by DAP score category and heavy drinking. Predicted hazards from imputed Cox proportional hazard models of incident pregnancy with interactions between DAP score and baseline substance use are presented in per 100 person‐years. All models account for clustering by participant and site and are adjusted for: age, race and ethnicity, parity, religiosity, relationship cohabitation and quality, schooling status, employment status, food insecurity and difficulty meeting housing expenses. Missing data are imputed. DAP, Desire to Avoid Pregnancy scale.

FIGURE 4

Predicted 1‐year marginal pregnancy hazards, by DAP score category and binge drinking. Predicted hazards from imputed cox proportional hazard models of incident pregnancy with interactions between DAP score and baseline substance use are presented in per 100 person‐years. All models account for clustering by participant and site and are adjusted for: age, race and ethnicity, parity, religiosity, relationship cohabitation and quality, schooling status, employment status, food insecurity and difficulty meeting housing expenses. Missing data are imputed. DAP, Desire to Avoid Pregnancy scale.