Arm-span/height correlation in the Luo community-a tool to improve admission criteria based on BMI for Nutritional Rehabilitation HIV Programs
Ngetich, Kiprono Weldon
MetadataAfficher la notice complète
Body mass index (BMI) is one of the main criteria used during targeting for therapeutic feeding. For bed ridden HIV-positive people, height cannot be accurately taken because they cannot stand straight due to gross muscle weakness resulting in inaccurate BMl calculation. Arm span has been shown to correlate with height and is not affected by sickness or aging but varies with ethnicity. This study was to explore the correlation between arm span/demi span and height among the Luo community so as to come up with a correlation equation that can be used to approximate the height of bedridden Hl V positive people. A cross sectional study was done and the main objective was to determine the correlation between arm span and height among the Luo community. The study location was Homabay MSF HIV clinic with the sample size of 500 Luo adults between the age IS and 60 years. At the study site, anthropometric measurements were taken using the standard techniques. Partial correlation between arm span/demi arm span and height with control of sex was done. Stepwise forward analysis was done to enable construction of a correlation equation that best estimate height from arm span. Bland-Altman assessment of agreement was done to check the difference in the use of either predicted height from arm span or demi arm span to calculate BMI. Scatter plot diagrams were used to show the correlation between arm span/demi span and height. Tables showing anthropometric characteristics of the sample, arm span! demi span and height correlation by sex, prediction models, partial correlation between MUAC and BMl, sex specific regression equation relating arm span/demi span and height and BMI-as cut offs equivalent to BMI-height cut offs for classifying nutritional status among adults were used to present data. The study found that there were differences in all variables between males and females from the Luo community. There was a high arm span/demi span with height correlation (0.S6) in both sexes among the Luo community. Arm span alone accounted for 83.6% of the variability in height with sex and age having no impact on the variability. The prediction equations for estimating height are: H eight=O. 71 (arm span)+40.S6 and Height=I.44(demi arm span)+43.2S. Partial correlation between BMl-asIBMI-ht with MUAC (controlled by sex) showed the lowest (0.57) and the highest (0.65) correlations respectively. In conclusion, the height estimated from arm span can be used to calculate BMI as a way of improving admission criteria for therapeutic feeding programs for the HIV positive adults in the Luo community. More research is recommended to determine the trend in other communities in Kenya and other African countries where there is high HIV prevalence.