Timing and patterns of exposures during pregnancy and their implications for study methods.
Hertz-Picciotto-I; Pastore-LM; Beaumont-JJ
Am J Epidemiol 1996 Mar; 143(6):597-607
The influence of timing and patterns of maternal exposures on pregnancy outcome were analyzed. Data were obtained via questionnaire from a case/control study of stillbirths and live births of female residents in ten agricultural counties in California. Cases were identified from fetal death certificates occurring after 20 weeks gestation or infant death within 24 hours after birth. Controls consisted of randomly selected women who delivered live infants. The overall sample size was 332 cases and 357 controls. Questions were asked on various pregnancy risk factors including medications, illnesses, and antenatal medical procedures; home exposures; lifestyle factors; and occupational exposures. The questionnaire data were used to determine the prevalence of these exposures over the entire pregnancy and for each trimester. More than half the subjects had used acetaminophen and had undergone ultrasonic examination. These two items accounted for the highest reported overall prevalences, as well as the highest prevalences by trimester; each was reported by at least 33% of the subjects in at least 1 trimester. Other relatively common exposure sources reported by at least 10% of the subjects in at least two trimesters included alcohol consumption, smoking, use of sprayed insecticides in the home, residing within 1 mile of commercially grown crops, and use of a sauna or hot tub. The overall prevalence of video display terminal (VDT) use during a pregnancy was 10.4%. Inter trimester concordance of exposures, as assessed by kappa statistics, was highest for smoking, VDT use, and living less than 0.25 mile from crops. Variation in intra trimester concordance indicated that women did not tend to recall their exposures in trimester blocks. The authors note that misclassification bias will occur if there is a vulnerable time window during which the exposure exerts its effect. The authors urge that measures which assume constant exposure throughout pregnancy not be used unless the data supports such findings.
NIOSH-Publication; NIOSH-Cooperative-Agreement; Reproductive-hazards; Fetus; Risk-factors; Pregnancy; Reproductive-system-disorders; Genotoxic-effects; Cytotoxic-effects; Pesticides; Agricultural-chemicals; Agricultural-workers;
Author Keywords: bias; epidemiology; environmental exposure; epidemiologic methods; fetal death; life style; occupational exposure; pesticides; pregnancy
Dr. Irva Hertz-Picciotto, Department of Epidemiology, School of Public Health, CB 7400, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599
American Journal of Epidemiology
University of California - Davis