Our study population of 4,586 women contributed 14,489 menstrual cycles of attempt time and 2,604 pregnancies (73.4% conceived during 12 cycles of follow-up after accounting for censoring). At baseline, 17% of women ever used an e-cigarette, and 4% were current users. Comparing baseline and last follow-up questionnaire, 10% of current e-cigarette users became former users over follow-up, 4% of former users became current users, and 0.5% of never users became current or former users.
Table 1 presents baseline characteristics stratified by baseline e-cigarette use. E-cigarette use was associated with lower education, income, Healthy Eating Index, use of multivitamins, higher body mass index, and greater caffeine intake, marijuana use, physician-diagnosed depression and anxiety, Major Depression Inventory score, parity, and intercourse frequency. E-cigarette use was strongly associated with combustible cigarette smoking history. Approximately 83% of never e-cigarette users were never cigarette smokers, compared with 37% of former and 23% of current e-cigarette users. Current e-cigarette use was positively associated with pack-years of combustible cigarette smoking (Table 1).
Women retained in the study (n = 3,427) and those lost to follow-up (n = 1,115) were similar according to mean age (29.8 vs. 29.5 years), alcohol intake (3.1 vs. 2.8 drinks/week), having partners who smoke combustible cigarettes (4.2% vs. 5.4%), intercourse frequency (20.6% vs. 21.2%, <1 time/week), and depression (27.9% vs. 27.8%) or anxiety (28.3% vs. 26.4%) diagnoses. They differed by e-cigarette use (84.7% vs. 77.7%, never users), current combustible cigarette smoking (5.4% vs. 14.3%), mean body mass index (calculated as weight (kg)/height (m)2; 28.2 vs. 31.0), use of multivitamins/prenatal supplements (83.6% vs. 67.4%), marijuana use (7.2% vs. 9.1%, use ≥1 time/week), education (4.6% vs. 14.8%, ≤12 years), and annual income (18.0% vs. 33.7%, less than $50,000).
E-cigarette use was associated with small reductions in fecundability (Table 2). Baseline current and former e-cigarette use were associated small reductions in fecundability after confounder adjustment (current-use fecundability ratio (FR) = 0.85, 95% confidence interval (CI): 0.68, 1.07; former-use FR = 0.89, 95% CI: 0.78, 1.00). The association did not get stronger with greater intensity of e-cigarette use (<3 mL per day, FR = 0.82, 95% CI: 0.58, 1.15; ≥3 mL per day, FR = 0.89, 95% CI: 0.65, 1.21). Time-varying estimates for current and former users were similar to baseline estimates (current-use FR = 0.84, 95% CI: 0.67, 1.06; former-use FR = 0.89, 95% CI: 0.79, 1.01). Confidence intervals for current-use and intensity estimates were wide and consistent with a broad range of associations, including none. Among never cigarette smokers, among whom residual confounding by cigarette smoking is unlikely, associations between e-cigarette use and fecundability were slightly attenuated and more imprecise (current-use FR = 0.87, 95% CI: 0.57, 1.32) (Table 2).
Compared with noncurrent users of both e-cigarettes and combustible cigarettes (Table 3), there was essentially no association among exclusive cigarette smokers (FR = 1.01, 95% CI: 0.85, 1.20); exclusive current e-cigarette use was associated with slightly lower fecundability (FR = 0.91, 95% CI: 0.70, 1.18) and dual current use of e-cigarettes and cigarettes was associated with reduced fecundability (FR = 0.83, 95% CI: 0.54, 1.29), all with wide confidence intervals. Compared with never users of e-cigarettes and cigarettes (Web Table 1, available at https://academic.oup.com/aje), estimates remained imprecise due to small numbers within strata. Fecundability ratios suggested a nearly null association for current cigarette smokers who were never e-cigarette users (FR = 1.01, 95% CI: 0.83, 1.24) and small but imprecise inverse associations for current e-cigarette users who were never smokers (FR = 0.87, 95% CI: 0.57, 1.32) and dual current users of e-cigarettes and cigarettes (FR = 0.82, 95% CI: 0.52, 1.27).
Adjusted fecundability ratios for current e-cigarette use were slightly stronger among women with shorter pregnancy attempt times at study entry (<3 cycles, FR = 0.75, 95% CI: 0.56, 1.02; ≥3 cycles, FR = 1.04, 95% CI: 0.69, 1.57) and women <30 years old at baseline (<30 years, FR = 0.78, 95% CI: 0.55, 1.10; ≥30 years, FR = 0.89, 95% CI: 0.64, 1.25) (Web Tables 2 and 3). Excluding e-cigarette users with devices containing 0 mg nicotine (n = 28 women; Web Table 4), the adjusted fecundability ratio for time-varying current e-cigarette use was consistent with our primary analysis (FR = 0.83, 95% CI: 0.66, 1.04). Reclassifying pregnancies to include only viable pregnancies led to slightly attenuated results (Web Table 5; current-use FR: 0.87, 95% CI: 0.68, 1.10). Models that did not adjust for parity were similar to models adjusting for parity (data not shown).
Am J Epidemiol. 2021;190(3):353-361. © 2021 Oxford University Press