Analyzing Limitations in Exposure Estimates Based on Self-Reported Dietary Intake of Caffeinated Beverages in the Baltimore-Washington Infant Study, 1981-1989

Caffeine, a mild central nervous system stimulant, is a natural component of common hot and cold beverages like coffee, tea, sodas and cocoa. Animal studies have demonstrated caffeine’s teratogenic effects when administered at high concentrations; however, epidemiologic studies have yielded inconsistent results in humans. Because caffeine containing beverages are commonly consumed by pregnant women, we examined the prevalence of use and explored possible associations of maternal caffeine consumption with cardiovascular malformations in 3,274 cases matched with 3,519 controls enrolled in the 1981-89 “Baltimore-Washington Infant Study,” a population-based case-control investigation. We explored several key aspects of the quality of and distribution of measurements of caffeine consumption among mothers in the study population. We concluded with recommendations for refining data collection to reduce potential bias associated with assessing both caffeine content and changes in caffeine consumption during pregnancy in order to inform future research studies and birth defects/adverse birth outcomes surveillance programs