Objective: To examine factors associated with perinatal HIV-1 transmis
sion among twins. Methods: We identified twins delivered by a populati
on-based cohort of HIV-infected mothers on New York State Medicaid. Te
sted algorithms were applied to Medicaid files to identify HIV infecti
on in mothers and twins. The HIV transmission rate 3 years after deliv
ery was assessed from Kaplan-Meier curves. Proportional hazards models
with adjustment for twin clustering were used to determine the relati
ve hazard (RH) of transmission. Results: In 35 twin pairs, transmissio
n was 20.5%. The risk of transmission was increased significantly for
advanced maternal HIV infection (rh = 10.8, 95% confidence interval 2.
11, 54.9). We found no association of birth order with twin HIV status
. Conclusions: These data suggest that maternal stage of disease plays
a greater role in vertical HIV transmission than birth order. To prev
ent material-child HIV transmission, reducing maternal viral load is l
ikely to have a greater impact than modifying delivery factors.