Objective: To determine the relationship between maternal heterosexual
activity during pregnancy and perinatal transmission of HIV-1. Design
: A retrospective analysis of 175 New York City HIV-1-seropositive wom
en enrolled during pregnancy or immediately post-partum from 1986 to 1
994 in a prospective cohort study. Methods: Frequency of heterosexual
intercourse and condom use during pregnancy was determined from self-r
eport measures. Unprotected intercourse was defined as follows: 'none'
, consistent condom use or abstinence; 'moderate', inconsistent condom
use and fewer than 80 episodes of intercourse; and 'high', inconsiste
nt condom use and 80 or more episodes. Results: The rate of perinatal
HIV-1 transmission was 9.1% (four out of 44) among women with no unpro
tected intercourse during pregnancy, 22.2% (20 out of 90) among those
with moderate frequency, and 39.0% (16 out of 41) among those with hig
h frequency (P < 0.01). The relative risk (RR) of perinatal transmissi
on was higher among women with moderate [RR, 2.4; 95% confidence inter
val (Cl), 0.9-6.7] and high frequency of unprotected sexual intercours
e (RR, 4.3; 95% Cl, 1.6-11.8) compared with women with no unprotected
sexual intercourse. When potential covariates (maternal injecting drug
use, CD4 lymphocyte count, AIDS, zidovudine use, pelvic inflammatory
disease or sexually transmitted disease during pregnancy, delivery mod
e, and extreme prematurity) were included in a logistic regression mod
el (n = 128), the rate of perinatal transmission remained significantl
y higher among women with any unprotected sexual intercourse during pr
egnancy. Conclusions: Data suggest that unprotected sexual intercourse
during pregnancy influences perinatal HIV-1 transmission.