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

Discussion

We confirmed a significant delay in the mean age at pubertal onset for perinatally HIV-exposed children compared to uninfected but PHIV youth, ranging from a 6 to 8-month later mean age at onset in girls and a 10 to 11-month later onset in boys. The later average age at onset corresponded to an increased prevalence of delayed onset for PHIV vs. HEU youth, but clinical delay among youth born after 1997 was rare regardless of HIV status. The PHIV youth in our cohort were more often born in earlier years than the HEU youth, which is notable given the secular trends in timing of pubertal onset;^[1–5] however, the later pubertal onset for PHIV youth persisted even after adjustment for race/ethnicity and birth cohort for all measures except pubic hair in girls. PHIV girls and boys with more advanced HIV disease were at greatest risk of delay in pubertal onset, and these associations persisted for most staging measures after adjustment for birth cohort. Boys had stronger associations with past measures of disease severity than did girls, which could be attributable to either poorer initial immunological status or increased sensitivity of hormonal pathways in boys.

Whereas both prior cART with protease inhibitors and longer duration of combination treatment appeared to be protective in restoring earlier timing of pubertal onset, these associations did not persist after adjustment for birth cohort. Descriptively, the mean age at pubertal onset was earlier for those on a combination regimen compared to those not on combination regimens born in 1997 or later, whereas the reverse was generally true for those born earlier. We did not observe a significant benefit of combination regimens in models adjusting for birth cohort, possibly due to confounding by indication resulting from only the sickest HIV-infected children receiving combination treatment in the earlier birth cohorts (33% of those in the pre1990 cohort), whereas combination treatment was widespread among HIV-infected youth in the later birth cohorts (88% among those born 1997 or later). Although there were no statistically significant interaction effects between combination treatment and birth cohort for any of the four staging measures, power for testing interaction may have been limited.

The mean ages at pubertal onset across the four staging measures and by race/ethnicity in our HEU youth from 219C and AMP were generally similar to other studies using NHANES III (1988–1994) data. Among PHIV youth, our estimated mean ages at onset are similar to those reported by Buchacz et al.^[14] of 10.7 years for girls and 11.8 years for boys (their study population was 52% black non-Hispanic, 33% Hispanic), but substantially earlier than mean ages at onset of 12.1–12.9 years among Caucasian HIV-infected children in an Italian study reported by De Martino et al..^[13]

The cause of delayed puberty in adolescents with HIV infection is not well understood. It has been ascribed to the general effects of chronic illness mediated through cytokine-induced inhibition of gonadotropin secretion.^[26,44,45] It has also been suggested that HIV infection directly or indirectly affects production or secretion of hormones that regulate or control pubertal initiation and tempo (e.g. leptin produced in adipose tissue).^[14] Delayed pubertal development in HIV-infected children has been attributed in part to reduced adrenal androgen secretion.^[11,23]

Implications of altered pubertal timing in the general population have received more attention for early maturation, which has been associated with increased incidence of antisocial behaviors and substance use. A general trend of earlier pubertal onset is thus not necessarily desirable, given the potential adverse social and clinical consequences of early puberty noted in the literature.^[1] However, whereas youth in many developed countries are attaining pubertal onset earlier than in previous decades, those with perinatally acquired HIV still tend to have later onset than US norms based on NHANES. Later maturation may also be associated with risk for psychosocial problems, including lower self-esteem and depression, and may have implications for reproductive health.^[15,46] Thus, the implications of our study findings in perinatally infected youth focus at the other end of the spectrum, in the benefits of reducing the risk for delayed pubertal onset along with associated psychosocial and reproductive consequences. Finally, our study findings may have particular relevance for low-resource settings such as sub-Saharan Africa, where rates of vertical HIV transmission remain relatively high and thus the population of youth with perinatally acquired HIV remains large.^[47] Despite wider availability and earlier initiation of antiretroviral treatment in South Africa and other African countries over the past decade, it has been documented that the majority of children still have severe immunodeficiency before starting treatment, increasing the risk of delayed pubertal onset.^[48] Similar studies are warranted to evaluate the impact of early antiretroviral treatment initiation on pubertal onset and maturation in low-resource settings.

We recognize several limitations in our analysis. Our cohort of HEU youth was relatively small and included few white non-Hispanic youth and no uninfected youth born before 1990. Like all studies utilizing Tanner staging measures, there is potential for misclassification, particularly for breast staging, which may be confounded by increased body fat deposition.^[1,40] Although orchidometers were not used in 219C, they were used in AMP and may provide better accuracy in future evaluations of pubertal progression and sexual maturation. We lacked information on birth weight or other early life exposures because our cohorts were not followed from birth. As an observational study, our analysis was subject to confounding by indication, particularly for evaluating the association of combination treatment with pubertal onset for the earlier birth cohorts; lack of information for most participants on viral load or CD4 prior to combination treatment initiation precluded our ability to evaluate and adjust for such confounding.

Despite these limitations, this is the largest study to date evaluating the timing of pubertal onset, with more than 2000 PHIV youth. We found that pubertal onset occurs significantly later in HIV-infected than in uninfected youth, with the greatest delays among those with more advanced HIV disease. Importantly, combination treatment may result in more normal timing of pubertal onset, as suggested for youth born since 1997 who were receiving protease inhibitor-containing combination regimens. Further evaluation of pubertal onset and sexual maturation in the current era of widespread treatment with combination treatment will be needed to fully understand the impact of antiretroviral treatment on sexual development of youth with HIV infection.

AIDS. 2013;27(12):1959–1970 © 2013 Lippincott Williams & Wilkins
Lippincott Williams & Wilkins