Background:  Data on lipoatrophy in children from Sub-Saharan Africa are extremely limited, where stavudine (d4T) is commonly used and 2.3 million are estimated to be HIV⁺. In 2008 to 2009, 88% of children on HAART took d4T. We investigated the prevalence, accurate diagnosis, and risk factors for lipoatrophy in pre-pubertal African children.

Methods:  In this cross-sectional study, the first 100 of 300 children on HAART at a Family HIV Clinic in South Africa were recruited, along with 34 controls. An expert HIV pediatrician graded visually obvious lipoatrophy. A dietician performed anthropometric measurements of trunk and limb fat. DEXA was performed on 42 patients and 34 controls. Duration of ARV use was recorded. Linear regression models were used to compare fat distributions captured by DEXA and anthropometrics among children who were HIV^–, HIV⁺ with lipoatrophy and HIV⁺ without lipoatrophy, adjusting for age and sex. Logistic regression was used to study the risk factors associated with clinical lipoatrophy.

Results:  Prevalence of clinical lipoatrophy was 37% (CI 28.2% to 46.8%). Adjusting for age and sex, DEXA and anthropometrics confirmed significant, substantial extremity fat loss in children with clinical lipoatrophy:  Mean adjusted DEXA total extremity fat was 2.7 kg in HIV^–, 1.7 kg in HIV⁺ with lipoatrophy, 2.3 kg in HIV⁺ without lipoatrophy (p = 0.001) and extremity fat vs extremity lean was 0.63 in HIV^–, 0.36 in HIV⁺ with lipoatrophy, 0.62 in HIV⁺ without lipoatrophy (p = 0.0001). Mean adjusted anthropometrics showed that biceps skin-fold thickness was 5.5 in HIV^–, 4.2 in HIV⁺ with lipoatrophy and 5.3 in HIV⁺ without lipoatrophy (p <0.0001), and triceps skin-fold thickness was 8.7 in HIV^–, 7.1 in HIV⁺ with lipoatrophy and 8.9 in HIV⁺ without lipoatrophy (p <0.0001). Multivariable analysis, controlling for age, sex and CD4 percentage, identified duration of d4T use (p = 0.0008) as the predominant risk factor for clinical lipoatrophy. Cumulative d4T was also associated with reductions in biceps and triceps skin-fold thickness (p =0.008).

Conclusions:  The objective measures of DEXA and anthropometrics confirmed the clinical assessment of visually obvious lipoatrophy. Thus, clinical assessment of lipoatrophy in children is reliable and while it requires objective training, no additional investigations are needed in a developing country context. The prevalence of lipoatrophy is higher in our cohort than non-African cohorts. Our data identify cumulative d4T exposure as the greatest risk factor for lipoatrophy, highlighting the urgent need for all children to transition to alternative medication.