Comparison of two methods of measuring body composition in HIV-infected children (9-36 months) at initiation of highly active antiretroviral therapy (HAART)
Date
2014-09-11
Authors
Haupt, Carey Anne
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The relationship between HIV and body composition is not well documented. No description of Fat Mass (FM), Fat Free Mass (FFM) or body fat percentage (BF%) has been reported in children infected with HIV (9-36 months), before initiation of antiretroviral treatment (ART). This is mainly due to the lack of a validated in-field method for measuring body composition in HIV infected children. Regression equations to estimate FM or BF% have not been developed for young South African children. Guidelines however recommend that equations developed and validated elsewhere be used in this population. For the recommended guidelines to be used with confidence in South African children they need to be validated.
Objectives
To describe the body composition of 36 treatment naïve HIV-infected children (9-36 months) and to compare the BF% and FM (g) results of four regression equations to those obtained from Dual Energy X-ray’s (DXA).
Methods
This is a retrospective, observational, cross sectional study. It includes descriptive data on body composition of HIV infected children and comparative data between published regression equations and DXA results. The 37 participants were enrolled from Harriet Shezi Children’s Clinic and all were aged 9-36 months, had parental consent for the study and had confirmed HIV infection but where ART naïve. Anthropometric measurements and DXA scan were done within a 10 day window period. Secondary analysis of FM and BF%
estimates from DXA scans and four regression equations were tested for agreement using Bland-Altman method.
Results
The DXA results (with head) showed a median BF% 25% (IQR22:32) and FM 1952g (IQR1594:2847) with no significant difference between genders. The fat patterning showed most of the FM in the legs 810g (IQR560:1220), then the trunk 689g (IQR520:930) and arms 13g (1.2:42.1). The trunk had the highest FFM 2700g (IQR2030:2960) of the three regions.
The calculated BF% from the equations were: Slaughter 10% (IQR9:13) and Shaikh & Mahalanabis 19% (IQR18:20). The FM (g) calculated from the equations were: Dezenberg -1823g (IQR-2450:-1160) and Goran 296g (IQR-83:1116). The results of the Bland-Altman analysis showed a high mean biases with wide limits of agreement between DXA and all four of the equations.
Conclusion
This is the first study to describe body composition in HIV-infected South African children (9-36 month) using DXA. The study has shown the importance in validating a regression equation before using it in a new study population, as the agreement between all 4 regression equations and DXA was very poor. The need for further studies in body composition and HIV has been highlighted. Further topics to be studied should include: the use of skinfold sites and girth measurements as in field monitoring of fat patterning, the development of South African specific regression equations for use in HIV infected children, the study of body composition using DXA in healthy and HIV infected children and the role of body composition with regards to absorption and metabolism of ART.