OBJECTIVE:
To compare the growth patterns in the first year of life between children born to perinatally HIV-infected (PHIV) vs. nonperinatally HIV-infected (NPHIV) women in the United States.

DESIGN:
Retrospective cohort study of HIV-infected pregnant women who received care and delivered a live-born at two urban tertiary centers from January 2004 to March 2012.

METHODS:
We collected data via chart review on demographics, behavioral risk factors, HIV clinical markers, combination antiretroviral therapy (cART), mode of HIV acquisition, pregnancy outcomes, and infant anthropometrics on study participants. Mixed-effects models were used to assess the association between maternal mode of HIV acquisition and weight-for-age z-score (WAZ), length-for-age z-score (LAZ), and weight-for-length z-score (WLZ).

RESULTS:
Of the 152 pregnancies evaluated, 32 and 120 infants were born to 25 PHIV and 99 NPHIV women, respectively. Infants of PHIV women exhibited lower mean WAZ and LAZ throughout the first year of life in unadjusted analyses. After adjusting for potential confounders, the relationship between PHIV women and LAZ persisted (β = -0.54, P = 0.026). Small-for-gestational age for each birth anthropometric parameter (birth length, birth weight, and both birth length and weight) was associated with decreased LAZ (β = -0.48, P = 0.007), WAZ (β = -0.99, P < 0.001), and WLZ (β = -0.36, P = 0.027), respectively. A delivery HIV RNA level below 400 copies/ml was associated with increased WAZ and WLZ (β = 0.43, P = 0.015 and β = 0.38, P = 0.021, respectively).

CONCLUSIONS:
Infants of PHIV women may remain at persistently decreased lengths throughout the first year of life. Further studies aimed at understanding intrauterine and environmental factors in PHIV women are warranted.