Can Men Provide Accurate Confounder Data About their Partners for Time-to-Pregnancy Studies?

Purpose: In studies of time to pregnancy (TTP), women's health-related behaviors may confound other determinants of TTP. In many occupation-based TTP studies, all information is collected through the male partner. There are no data on the validity of the man's report of his partner's fertility-related behavior. Methods: We studied 202 men and their partners from the most recent pregnancy. Validity of men's reporting on their partner's use of oral contraceptives (OCs) as the last birth control method and her smoking around the beginning of TTP and agreement of coital frequency were assessed. Results: The index pregnancy was an average of 6 years before interview. Overall percentage of agreement was 81% for OCs as the last contraceptive method (κ agreement = 0.44). Ninety-five percent of men accurately reported whether their partner smoked (κ agreement = 0.83). Among couples agreeing on smoking status, 90% agreed on the categorical cigarette number (weighted κ = 0.60). Reporting accuracy was not influenced by men's characteristics. Median coital frequency was eight times per month, with a weighted κ = 0.34 after categorization. Conclusions: Our data generally justify the use of men's reports of potential confounders in TTP studies when women's reports are not available.