Pubertal Onset in Children With Perinatal HIV Infection in the Era of Combination Antiretroviral Treatment

Paige L. Williams; Mark J. Abzug; Denise L. Jacobson; Jiajia Wang; Russell B. Van Dyke; Rohan Hazra; Kunjal Patel; Linda A. Dimeglio; Elizabeth J. Mcfarland; Margarita Silio; William Borkowsky; George R. Seage III; James M. Oleske; Mitchell E. Geffner


AIDS. 2013;27(12):1959–1970 

In This Article


Description of Protocols and Study Population

This investigation included children born to HIV-infected women who had pubertal staging assessed at age 7 years or older in the 219C or AMP studies. The 219C study was a large prospective study conducted at over 80 US clinical research sites between 1993 and 2007 to evaluate the long-term effects of HIV infection and in-utero antiretroviral exposure.[30,31] AMP is a smaller ongoing prospective cohort study which opened in March 2007.[32] Individuals were eligible for AMP (whether in 219C or not) if they were PHIV or HIV-exposed, and 7 to below 17 years old. Both studies were approved by the site Institutional Review Boards and written informed consent was obtained from each parent or legal guardian, with assent from children as appropriate.

At each study visit, we obtained medical histories through chart reviews and ascertained health status through physical and laboratory evaluations. HIV disease severity information for HIV-infected youth was collected at scheduled visits (every 3 months for 219C, every 6–12 months for AMP). Race and ethnicity were self-reported at study entry and categorized as white non-Hispanic, black non-Hispanic, Hispanic, or 'other'. We excluded participants of PACTG 219 who did not participate in 219C and youth who were judged to be at later stages of pubertal maturity (Tanner stage 3–5) at their first study visit.

Pubertal Staging Measures

Study clinicians trained to evaluate child growth and development assessed pubertal staging annually in 219C and at each study visit in AMP (initially every 6 months and then annually after August 2010) by visual inspection according to criteria of Tanner and Whitehouse,[33] ranging from 1 (prepubertal) to 5 (sexually mature). Pubertal onset was defined for each Tanner measure as attaining stage 2 or higher, with separate indicators for breast development and pubic hair in girls, and for genitalia and pubic hair in boys.

Statistical Methods

The date of the first pubertal staging assessment at the age of at least 7 years was considered to define the start of longitudinal follow-up for pubertal onset. The last available measures of height and BMI (kg/m2) prior to or on this index date were obtained, and Centers for Disease Control and Prevention (CDC) 2000 growth standards were used to calculate age and sex-adjusted Z-scores.[34–38] For HIV-infected youth, latest measures of CD4 T-lymphocyte percentage (CD4%), CD4 cell count, HIV-1 RNA viral load, and CDC clinical classification were identified prior to or at the index date. Nadir CD4% and peak viral load were based on the lowest CD4% and highest viral load measure, respectively, prior to or at the first pubertal assessment. For those with at least two viral load measures before their first pubertal assessment, a cumulative measure of copy-years viremia was calculated as described by Cole et al..[39] cART was defined as concurrent use of at least three drugs from at least two drug classes, and cumulative antiretroviral history was reflected by years of receipt of combination treatment, protease inhibitors, and non-nucleoside reverse transcriptase inhibitors (NNRTIs) prior to the index date. We summarized characteristics of the participants by HIV infection status. The percentage of children with delayed pubertal onset, defined as not attaining Tanner stage 2 by age 12 in girls (by both breast and pubic hair staging) or by age 13 in boys (by both genitalia and pubic hair staging) was assessed by birth cohort and HIV infection status.

Interval-censored approaches under an assumed normal distribution were used to estimate the mean age at pubertal onset separately for each Tanner measure, by HIV infection status, race/ethnicity, and birth cohort (pre-1990, 1990–1992, 1993–1996, 1997 or later). The interval-censoring approach accounts for whether pubertal onset occurred prior to the first pubertal assessment (left-censored), between study visits (interval-censored), or had not occurred by the last study visit (right-censored); this is the most appropriate statistical approach for evaluating pubertal onset based on longitudinal data.[40] The mean age at pubertal onset was compared between HIV-infected and HEU youth in unadjusted models and with adjustment for race/ethnicity and birth cohort. Further adjustment for BMI and height Z-scores was performed, but since these measures may be on the causal pathway between HIV infection and pubertal onset, results of these sensitivity analyses are presented only in online supplemental tables. Caregiver education level was considered but not found to be associated with timing of pubertal onset. We also qualitatively compared the estimated mean ages at pubertal onset with the general US population as reflected by the National Health and Nutrition Examination Survey (NHANES) III (1988–1994).[4,5,40]

For PHIV youth, interval-censored models were fit to evaluate each HIV disease severity measure and antiretroviral treatment characteristic separately, in unadjusted models and after adjustment for race/ethnicity and birth cohort. Antiretroviral regimens were classified as combination treatment with protease inhibitor, combination treatment without protease inhibitor, or not on combination treatment (not on antiretroviral drugs, or on other regimen). Sensitivity analyses were conducted to examine effects of further adjustment for potential intermediates between cART and pubertal onset, including BMI and height Z-scores and CD4% less than 15%. Duration of prior use of combination treatment and specific drug classes (protease inhibitors and NNRTIs) were considered as continuous predictors after evaluation of linearity assumptions. A sensitivity analysis was also conducted to examine consistency of results when restricting the study population to only prepubertal youth.