Influence of sexual activity on cytomegalovirus seroprevalence in the United States, 1988-1994.

BACKGROUND

Sexual and nonsexual transmission of cytomegalovirus (CMV) occurs, but the frequency of sexual transmission in the general population of the United States is unknown.

METHODS

Using data from 15- to 44-year-old (n = 7883) participants of the Third National Health and Nutrition Examination Survey (1988-1994), we examined the association between CMV seroprevalence and sexual activity markers. Using logistic regression, we calculated standardized prevalence differences (PDs)-the weighted average CMV prevalence among higher sexual risk groups minus CMV prevalence among the lowest sexual risk group-for each of several sexual activity markers (ever had sex, number of sex partners [lifetime and past year], age at first intercourse, potential years of sexual activity, ever use oral contraceptives, herpes simplex virus type 2 antibody, and a calculated composite marker).

RESULTS

Even after controlling for covariates, we found associations between CMV seroprevalence and sexual activity among non-Hispanic black [all PDs for sexual activity markers were positive and composite PD = 8.5%, 95% confidence interval (CI) = 4.0%-13.1%] and non-Hispanic white women (15 of 18 PDs for sexual activity markers were positive and composite PD = 10.8%, 95% CI = 3.1%-18.5%). We found a borderline significant association among Mexican American women (13 of 18 PDs for sexual activity markers were positive and composite PD = 3.5%, 95% CI = -0.7% to 7.6%). We found little or no association within each racial/ethnic group of men.

CONCLUSIONS

Sexual activity measurably influences CMV seroprevalence among women of childbearing age, indicating that congenital CMV prevention messages should include strategies to reduce sexual transmission of CMV among pregnant women.