MODERATE ACUTE MALNUTRITION IN A PROSPECTIVE COHORT OF REFUGEE CHILDREN: COMMON, BUT OFTEN TRANSIENT OR UNRESPONSIVE...

Moderate acute malnutrition (MAM) may be identified using low Weight-For-length-Z scores (WLZ) or Mid-Upper-Arm-Circumference (MUAC), but increasingly MUAC alone is used.

To examine how identification as MAM using WLZ compared to MUAC, related to prior and subsequent growth and survival, and to explore a more selective screening approach.

A longitudinal cohort was constructed using routine clinic anthropometric data from 4131 children, monitored monthly in refugee camps in Bangladesh. Children with WLZ >-3 to <-2 or MUAC >115 to <125 mm were identified as MAM and managed with counselling and food supplements.

Between ages 6-24 months, 1586 (38%) children met criteria for MAM (40.3% isolated low MUAC, 41.3% isolated low WLZ, 18.5% with BOTH). Of these 44.1% had Persistent MAM (low mean values across two 3-month periods) but 29.3% were Transient (screen positive only at one measurement point) particularly those identified by WLZ (48.2%). A prior decline in Weight-for Age (WAZ) of more than >0.2SD was observed in 87.5% of children with low WLZ, compared to 56.0% with low MUAC, who also on average showed no gain in WAZ after treatment and were more likely to already be stunted. Children with BOTH, or with persistent MAM were more likely to die than never MAM (relative risk BOTH 3.2, 95%CI 1.5-6.8; persistent 2.51, 1.4-4.7). MUAC only screening would have missed 25% with persistent MAM.

Screening for MAM identifies a high proportion of children, of whom many have only transient MAM or do not appear to benefit from treatment. A staged approach using increasingly stringent screening criteria could reduce the work of screening and the need for supplementary foods. This approach needs to be formally trialled.