Data describing lipodystrophy in children from sub-Saharan Africa are extremely limited. However use of d4T is widespread among children receiving ART in the region: in 2008 to 2009 approximately 90% of children on ART were taking d4T.
Two posters at CROI 2012 described substantial rates of lipodystrophy in South African children aged 3-12 and less than 2 years respectively.
Steve Innes and colleagues from the Children's Infectious Diseases Clinical Research Unit (KID-CRU), Tygerberg Children's Hospital and Stellenbosch University Cape Town performed a cross sectional study of 100 of 300 children on ART at the Tygerberg Family HIV Clinic and 34 HIV-negative controls. 
In this study, two clinicians graded fat changes visually using a standardised scale and a dietician took anthropometric measurements of trunk and limb fat. DXA was performed in a sub study of 42 patients and 34 controls. The duration of ART use was recorded.
Using linear regression models the investigators compared fat distribution captured by DXA and anthropometrics among children who were HIV negative, HIV positive with lipoatrophy and HIV positive without lipoatrophy, adjusted for age and sex. The risk factors for clinical lipoatrophy were evaluated by logistic regression.
This reported prevalence of lipoatrophy was 36% (95% CI 26 - 45%). DXA and anthropometrics confirmed significant, substantial extremity fat loss in children with clinical lipoatrophy.
Mean adjusted DXA total limb fat was 2.7 kg (95% CI 2.4 2.9), 1.7 kg (95% CI 1.4 -2.1) and 2.3 kg (95% CI 2.1 2.6) in HIV negative, HIV positive with lipoatrophy and HIV positive children without lipoatrophy respectively, p = 0.001. Limb fat vs limb lean ratios were respectively, 0.63 (0.56 0.7), 0.36 (0.25 0.46) and 0.62 (0.54 0.7), p = 0.0001.
Mean adjusted anthropometrics found biceps skin-fold thicknesses of 5.5 mm (5.0 5.9), 4.2 mm (3.6 - 4.7) and 5.3 mm (4.9 5.7), in HIV negative, HIV positive with lipoatrophy and HIV positive children without lipoatrophy respectively, p <0.0001. Triceps skin-fold thicknesses were respectively 8.7 mm (8.1 9.4), 7.1 mm (6.2 7.9) and 8.9 mm (8.3 9.6), p <0.0001.
The investigators noted that diagnosis by visual grading correlated well with anthropometry and DXA, which are not commonly available in developing countries.
In multivariate analysis, controlling for age, sex and CD4 percentage, the greatest risk factor for clinical lipoatrophy was duration of d4T use, p=0.0008. Cumulative d4T use was also associated with reductions in biceps and triceps skin-fold thickness, p=0.008. Each additional year of accumulated d4T exposure gave an odds ratio of 1.9 (95% CI 1.3 2.9), p=0.002.
The investigators wrote: "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."
Stephanie Shiau and colleagues from the NEVEREST study team described the prevalence of lipodystrophy and associated patterns of regional fat distribution and metabolic alterations in young children who had started ART at less than 2 years of age. 
They performed an evaluation of 156 vertically infected children who started ART at Rahima Moosa Mother and Child Hospital, Johannesburg with lopinavir/ritonavir (LPV/r) + 3TC + d4T, and were randomised to either continue LPV/r (n = 85) or switch to nevirapine (NVP) (n = 71), while continuing 3TC + d4T. This was done on exit from the NEVEREST 2 trial after approximately 4 years on ART.
Clinicians assessed the children visually for signs of lipodystrophy, including lipoatrophy and lipohypertrophy. Anthropometrics, bio-impedance analysis, viral load, CD4, fasting total cholesterol, HDL, LDL, and triglycerides were measured. Measurements of regional fat -- including trunk-extremity skin-fold ratios were estimated. Outcomes were compared across lipodystrophy groups defined as, lipodystrophy, possible lipodystropy and no lipodystrophy.
The investigators used multiple linear regression to access differences in arm, trunk and leg fat across lipodystrophy groups, adjusted for total fat, sex and age.
They found, of 156 children with a mean age 5.1 who initiated ART at a mean age of 10.7 months, 13 (8.4%) children were defined as having lipodystrophy, 18 (11.5%) as having possible lipodystophy and 125 (80.1%) as no lipodystrophy. All 13 children defined as having lipodystrophy had lipoatrophy and 6 also had signs of lipohypertrophy.
There were no differences in age, sex, age at ART initiation, duration of ART, weight-for-age z-scores, height-for-age z-scores, body mass index, or proportion of children with a viral load <50 copies/mL among the three lipodystrophy groups.
There was no difference in the proportion of children with lipodystrophy between those who remained on LPV/r and those who switched to NVP, respectively 7.1% vs 9.9%, p=0.51.
The children with lipodystrophy had significantly less body fat than children with no lipodystrophy, measured by mean (+SD) skin-fold sum, 34.1 mm (+5.7) vs 42.0 mm (+11.1), p=0.0016. Children with lipodystrophy had greater trunk-arm 0.53 mm (+0.07) vs 0.50 mm (+0.05), p=0.028 and trunk-leg skin-fold ratios 0.61 mm (+0.07) vs 0.55 mm (+0.06), p=0.004, than children without lipodystrophy.
Lipid concentrations were similar across groups, except for mean triglycerides level which was greater for children with lipodystrophy compared to those without, 101 (+45) vs 80 (+34) mg/dL, p= 0.045. The proportion of children with triglycerides >150 mg/dL was greater for children with lipodystrophy and those with possible lipoystrophy compared to those without, respectively 23.1% vs 4.8% and p=0.04 22.2% vs 4.8%, p=0.023.
"A substantial portion of young children who initiated d4T-containing ART before two years of age have lipodystrophy as classified by clinical criteria..." the investigators concluded, adding: "Lipodystrophy can be cosmetically stigmatising and adversely affect adherence to ART. Finding a substantial proportion of young children with lipodystrophy has implications for future adherence, especially during adolescence when awareness of physical appearance is greatly heightened."
These reports are concerning and the rate reported by Innes et al particularly is high compared to other (generally anecdotal) reports from other parts of Africa. This may be because children were properly evaluated, although it is not clear whether there was blinding to laboratory results when the clinical diagnosis was made but visual grading correlated well with anthropometry and DXA.
In South Africa, where FDCs are not generally used, the 1 mg/kg doses of d4T will usually be rounded up using stand alone products resulting in a dose at least equivalent 40 mg in adults (Steve Innes, personal communication), so the effects might be less or occur over a longer duration of exposure with a lower dose.
This possibility in children has been used to argue for a controversial study in adults, of lower dose (20 mg) d4T, which, the investigators hope, will mitigate the drug's toxicity. As is usual in adult studies, Innes et al indentified "cumulative d4T exposure as the greatest risk factor for lipoatrophy" and since d4T toxicity is both dose and time dependent -- as we have stated before -- it seems most unlikely that this could be reduced to acceptable levels in this way and there are better solutions on the horizon. [3, 4, 5]
A Thai study suggests that some lipodystrophy in children may be reversible after substitution with another NRTI and results from CHAPAS 3 will clarify whether this occurs.  But for the proportion for whom it is not reversible, as Shiau et al wrote, this can be stigmatising and adversely affect adherence to ART, particularly during adolescence when awareness of physical appearance is particularly sensitive (though when is it not?). Adult guidelines have for many years stressed that lipoatrophy is better to avoid than to treat. Children should also therefore be protected from this side effect.
Links to external websites are current at time of posting but not maintained.