Caffeinated Beverage Intake and Serum Caffeine Metabolites and Risk of Pregnancy Loss (OR17-04-19)

Abstract

OBJECTIVE: According to current ACOG recommendations, moderate intake of caffeine (<200 mg/day) during pregnancy does not increase risk of pregnancy loss, though epidemiologic evidence to support this recommendation is controversial. Our objective was to address limitations of prior studies by evaluating associations of both self-reported intake of caffeinated beverages and preconception and early pregnancy serum caffeine biomarkers and risk of pregnancy loss, while accounting for nausea and vomiting, smoking, and alcohol intake in early pregnancy. DESIGN: This was a secondary analysis of the EAGeR trial, which included 1,228 reproductive-age women attempting pregnancy during 2007-2011. MATERIALS AND METHODS: Questionnaires administered at baseline assesed self-reported intake of caffeinated beverages and other demographic and lifestyle variables. During pregnancy, daily questionnaires assessed caffeinated beverage intake, nausea and vomiting, alcohol intake, and smoking. Serum caffeine, paraxanthine, and theobromine were measured at preconception and during the 8th week of gestation. HCG-detected losses occurred prior to ultrasound confirmation and clinical pregnancy losses occurred after ultrasound visualization of a gestation sac. We used Cox proportional hazards regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for pregnancy loss according to: 1) baseline caffeinated beverage intake and caffeine biomarkers measured at preconception and the 8th week of gestation, adjusted for baseline factors and, 2) time-varying caffeinated beverage intake during pregnancy adjusted for baseline factors and time-varying nausea and vomiting, alcohol intake, and smoking. RESULTS: 67%, 28%, and 9% of women reported any preconception intake of caffeinated sodas, coffee, and tea, respectively. Preconception total caffeinated beverage intake of≥2 vs. 0 servings/d was marginally associated with increased risk of any loss (HR = 1.51, 95% CI = 0.98, 2.34), and associations were stronger for hCG-detected losses than for clinical losses. Soda intake was more strongly associated with hCG-detected losses, whereas coffee intake was more strongly associated with clinical losses. Any detectable level of serum caffeine (>0.2 vs.≤0.2 ng/mL) at preconception was strongly associated with hCG-detected loss (HR = 4.51; 95% CI: 1.36, 14.91), but biomarkers measured during the 8th week of gestation were not associated with loss. Further adjustment for nausea and vomiting and other factors that change during early pregnancy showed that caffeinated beverage intake at levels lower than those corresponding to current medical recommendation was positively associated with risk of loss (0 vs.≥1 servings/day HR=1.73; 95% CI = 1.02, 2.94), particularly among hCG-detected losses (HR = 2.83; 95% CI = 1.08, 7.39). CONCLUSIONS: Our findings suggest that any level of caffeine intake during pregnancy may increase risk of pregnancy loss, particularly in the first 8 weeks' gestation. Women attempting to conceive may benefit from eliminating caffeine intake during preconception and early pregnancy.