Identifying risk and response to nutritional supplementation in severely underweight children missed by standard treatment criteria in Mali: A secondary prognostic analysis of data from a randomised controlled trial

Severely underweight children (weight-for-age Z-score [WAZ] <-3) are ineligible for wasting treatment unless they have a mid-upper arm circumference (MUAC) <125 mm or a weight-for-height Z-score < -3. However, the risk of negative outcomes without nutritional supplementation and the likelihood of positive outcomes with supplementation are unknown. We aimed to investigate this using a secondary analysis of randomised controlled trial data from children in Mali with WAZ < -3 and MUAC ≥ 125 mm. We assessed baseline anthropometric, clinical, and social factors associated with negative outcomes (sustained WAZ change < 0 at any point during 0–12 weeks of follow-up or death or hospitalisation, or development of MUAC < 125 mm) in the absence of nutritional supplementation, and positive outcomes (sustained WAZ change ≥ 0.15 and no death and no hospitalisation, and sustained MUAC ≥ 125 mm) with nutritional supplementation. Unadjusted and adjusted generalised linear models were fitted separately for the supplemented and unsupplemented children, with significant exposures from the unadjusted models then adjusted for sex, season, age, and distance from the health centre. A total of 1378 children were included. Less than 2% of children experienced hospitalisation or death; thus, the negative outcomes were mostly anthropometric. Among children who were not nutritionally supplemented, 35% (151/432) experienced negative anthropometric outcomes, while 67% (643/946) of children who were supplemented experienced positive anthropometric outcomes. In the unadjusted analysis only, younger age (< 2 years), lower body weight and height, and currently breastfeeding children were associated with an increase in negative anthropometric outcomes. Among those receiving nutritional support, having an illness prior to the start of supplementation was associated with positive anthropometric outcomes in both unadjusted and adjusted analyses. Different factors were associated with negative and positive anthropometric outcomes, with no overlap, highlighting the difficulty of targeting nutritional supplementation among severely underweight children.