Comparison of weight gain to age- and sex-specific norms in children 2 to 10 years old on highly active anti-retroviral treatment

Abstract

Background: Growth charts are essential tools used for the evaluation of children's health and nutritional status. Growth monitoring has been used to identify children who may require highly active anti-retroviral therapy (HAART), especially in resource-limited settings where treatment decisions are often made on growth data alone. Growth reference data that is used to establish growth charts are most often obtained from populations where growth was optimal, however, growth failure is a hallmark of human immunodeficiency virus (HIV) infection in the peadiatric patient. Inadequate weight gain might warn of clinical deterioration in children infected by HIV, but existing references for optimal weight gain and determining of response to treatment in children initiated on HAART at different ages are not being widely implemented. Interpretations from growth chart evaluations will ultimately have important implications for the treatment of individual children and for child health programmes. Objectives: The objectives of the study were to assess and analyse the weight gain and weight gain patterns of children younger than ten years old, from initiation of HAART to 6, 12, 18 and 24 months' follow-up after HAART initiation. This study also compares the interpretations of weight gain patterns of the same group of children according to two different weight monitoring reference charts: age- and sex-specific charts developed to assess the growth and response to treatment of children on HAART and weight pattern interpretations according to current World Health Organization (WHO) weight-for-age z-scores (WAZ). Methods: This project was approached in a quantitative, descriptive-comparative manner with a retrospective design. Weight and other data relating to HIV were captured from patient records kept from the time that an infant/child was initiated on HAART. The weight gain recorded of boys and girls younger than ten years old, during the 24 months following HAART initiation, was assessed and analysed. Descriptive statistics were used to describe the baseline and follow-up characteristics of the boys and girls. Mixed model analysis was also used to test for significance of increases in weight, WAZ, serum-haemoglobin (Hb) and percentage T-lymphocyte-bearing CD4 receptor (CD4%). Mixed methods analysis of longitudinal data was performed, using the restricted maximum likelihood (REML) function with an unstructured covariance type. The quality of fit was estimated by Akaike's information criterion (AIC). Repeated comparisons were made to test for changes between six-monthly follow-up visits, with Sidak adjustment for multiple comparisons. A Kappa test for agreement between the four identified growth pattern categories according to the two growth norms was performed. The Kappa test was also repeated in a subgroup analysis, where only children with a low weight at HAART initiation were included, as defined by WAZ-score < -1. The range of deviations from the norms is presented and the effect sizes of the differences were calculated as mean difference divided by standard deviation of the mean weight gain at 24-month follow-up for each age group of boys and girls. Results: The total number of baseline and follow-up data points that formed part of the statistical analyses was N = 363, which was derived from 98 infants/children. More than half of the children in this study were underweight and stunted for their age by the time that HAART was initiated. There were statistically significant improvements in weight gain over the 24-month research period and at each six-month follow-up visit since HAART was initiated. Weight gain improved significantly from as early as six months and linear growth started improving significantly after six months. The children in our study did not reach complete catch-up growth after 24 months. The interpretations of weight gain patterns between the two reference charts that were used: according to the WHO charts, 69% of the children had an increase in rate of weight gain versus only 16% according to the age- and sex-specific weight gain charts. These interpretations were comparatively statistically different, as proven by the poor agreement between the two growth patterns. The results of the subgroup analysis also indicated that the two growth charts were very different in terms of agreement between interpretation outputs. Discussion, conclusion and recommendations: Even though the children in this study were severely immunocompromised when HAART was started, they showed rapid weight and height improvements. The children did not manage to reach complete catch-up growth within the 24-month research period, which indicated that the unique environment and socio-economic setting of the cohort affected the rate of growth of infants and children. Regarding the weight gain interpretations; the poor agreement between the WHO- and the age- and sex-specific weight gain charts established by Yotebieng et al., (2015) prove that children's weight gain, and growth should be interpreted by using appropriate references, especially if they are available, otherwise we risk making invalid interpretations. Timing is important, especially when it comes to the care and monitoring of young infants/children and particularly in settings where blood cannot be drawn or analysed. A simplified version of expected weight gain in infants/children on HAART, as established by Yotebieng et al., (2015) should be created and provided to parents/caretakers, together with education, so that they can also monitor their infants/children at home. It might be necessary also to create monthly or three-monthly weight gain references for healthcare professionals, so that weight can be monitored more often and not just six monthly on average. Regular weight monitoring could aid in improving the infant's/child's outcome through timeous intervention decisions, whether these are social interventions, feeding programmes, special counselling or the improvement of interdisciplinary treatment in any setting. More nutritional research is needed to determine the impact of nutrition interventions, especially during the early stages of improper weight gain, in order to assess the impact on HAART treatment success and immunity. Length/height gain references have not been established and future research should investigate its association with HIV progression.