Height-Age as An Alternative to Height-For-Age z-Scores to Assess the Effect of Interventions on Child Linear Growth in Low- and Middle-Income Countries.

Background: Assessments of the efficacy of interventions to improve child growth are often based on differences in mean height-for-age Objectives: The objectives of this study were to develop and apply a new method using height-age to express linear growth effects in RCTs. Methods: Longitudinal individual participant data (IPD) from a Bangladeshi trial cohort were used to compare height-age estimates derived from individual-level heights, mean raw height, or mean HAZ. Then, using mean height-age as a proxy for skeletal age, we developed the "proportion of maximal benefit" (PMB) metric to quantify intervention effects relative to optimal growth for children's starting skeletal age. Optimal growth occurs when height-age increases in parallel with chronologic age (i.e., PMB = 100%), whereas no effect (compared with control) corresponds to a PMB of 0%. Linear growth outcomes in 4 published RCTs of nutrition-specific interventions were re-expressed as mean height-age and PMB and compared with effects conventionally expressed as intervention-compared with-control mean differences (MD) in HAZ. Results: Mean height-age could be derived from any published estimate of mean raw height or mean HAZ; however, to calculate the PMB, height or HAZ data were required at both the beginning and end of the observation period. Interpretations of intervention effects were consistent when expressed as either the height-age MD or HAZ MD. In contrast, the PMB does not have a corresponding metric on the HAZ scale and, therefore, provided a new way to quantify intervention efficacy. Conclusions: Height-age can be used as an alternative to HAZ to express intervention effects. The PMB has the advantage of conveying the extent to which an intervention improved average linear growth in relation to a biologically-defined benchmark.