Background: The physiologic effects and common use of caffeine during pregnancy call for examination of maternal caffeine consumption and risk of birth defects. Epidemiologic studies have yielded mixed results, but such studies have grouped etiologically different defects and have not evaluated effect modification.
Methods: The large sample size and precise case classification of the National Birth Defects Prevention Study allowed us to examine caffeine consumption and specific cardiovascular malformation (CVM) case groups. We studied consumption of caffeinated coffee, tea, soda, and chocolate to estimate total caffeine intake and separately examined exposure to each caffeinated beverage. Smoking, alcohol, vasoactive medications, folic acid supplement use, and infant gender were evaluated for effect modification. Maternal interview reports for 4,196 CVM case infants overall and 3,957 control infants were analyzed.
Results: We did not identify any significant positive associations between maternal caffeine consumption and CVMs. For tetralogy of Fallot, nonsignificant elevations in risk were observed for moderate (but not high) caffeine intake overall and among nonsmokers (ORs of 1.3 to 1.5). Risk estimates for both smoking and consuming caffeine were less than the sum of the excess risks for each exposure. We observed an inverse trend between coffee intake and risk of atrial septal defect; however, this single significant pattern of association might have been a chance finding.
Conclusions: Our study found no evidence for an appreciable teratogenic effect of caffeine with regard to CVMs.
(c) 2007 Wiley-Liss, Inc.