CLINICAL QUESTION: Is maternal consumption of caffeine associated with an increased risk of spontaneous abortion?
BACKGROUND: Some previous studies have reported a doubling of the risk of fetal loss with caffeine use during pregnancy, some reported a risk only with large amounts of caffeine, and others have not shown any increased risk. Those studies were limited because of small sample size, suboptimal study design, and reliance on self-reporting for the quantity of caffeine consumed. The authors of this study measured the primary metabolite of caffeine, paraxanthine, to see if it is associated with the risk of spontaneous abortions.
POPULATION STUDIED: This study used data from the Collaborative Perinatal Project, a prospective study of pregnancy, labor, and child development at 12 sites in the United States. More than 42,000 women were enrolled, and 55,000 births were tracked between 1959 and 1966. The authors of this study identified 591 women in the cohort who experienced a spontaneous abortion before 140 days’ gestation. Each woman with a spontaneous abortion was matched with at least 4 control patients who were from the same site and had serum obtained on the same day of gestation.
STUDY DESIGN AND VALIDITY: This was a nested case-control study in which serum paraxanthine levels were measured in 591 women who experienced spontaneous abortions at less than 140 days’ gestation and in 2558 matched control patients who gave birth to live infants at 28 weeks’ gestation or later. Laboratory personnel were blinded to the outcome of the pregnancy. The women who had spontaneous abortions were different from those in the control group in several ways. They were older, smoked more, and were less likely to have vomited or used medications with caffeine. Serum paraxanthine levels were higher in older women who were white, smokers, and those who did not vomit during pregnancy. Therefore, the odds ratios were adjusted for smoking status, age, and race. Because this was a retrospective case-control study, there is the potential for bias-caused selection of the controls; however, the design is acceptable to answer the question.
OUTCOMES MEASURED: The primary outcome was the risk of an early spontaneous abortion (as estimated by the odds ratio) for different serum paraxanthine levels. Patients were divided into 3 groups on the basis of their paraxanthine levels: less than 50 ng/mL, 50 to 1845 ng/mL, and greater than 1845 ng/mL.
RESULTS: The women who experienced an early spontaneous abortion had higher mean serum paraxanthine concentrations than the control group (752 ng/mL vs 583 ng/mL, P <.001). However, the odds ratio for spontaneous abortion was not significantly elevated at paraxanthine levels of 1845 ng/mL or lower. When the lowest level group (<50 ng/mL) was compared with the highest level group (>1845 ng/mL), the odds ratio indicated a higher risk for the latter group (odds ratio=1.9; 95% confidence interval, 1.2-2.8). This paraxanthine level corresponds to a caffeine in take of 1100 mg, or roughly 11 cups of coffee per day in a smoker and 6 cups of coffee per day in a nonsmoker.
The authors of this study used a metabolite of caffeine to evaluate the risk of spontaneous abortions from caffeine intake during pregnancy. It has an adequate sample size and does not rely on self-reported caffeine use, which gives it an advantage over previous studies. It does not definitively establish a causal relationship between caffeine and spontaneous abortions but does suggest that if such a relationship exists only very high levels of caffeine pose a threat. With these new data physicians can more confidently discourage high levels of caffeine intake during pregnancy and may tolerate occasional caffeine use in some pregnant patients.