Population-attributable fractions for risk factors for childhood anaemia: findings from the 2017 Togo Malaria Indicator Survey

Anaemia remains a significant public health concern in developing countries. This study estimated the proportion of childhood anaemia cases that could be potentially prevented in Togo using data from the 2017 National Malaria Indicator Survey. Maternal, child, and household data were collected through standardized face-to-face interviews. Haemoglobin (Hb) levels were measured in children and their mothers. A total of 2796 children were included in the analyses. The prevalence of anaemia was 75·0 % (95 % CI, CI: 72·5, 88·0). Factors associated with childhood anaemia were age ((adjusted prevalence ratio, aPR = 1·46 (CI: 1·37, 1·56) for 6-23 months and aPR = 1·23 (1·14, 1·32) for 24-42 months, ref: 43-59 months), a later birth order (≥ 4th position) (aPR = 1·11 (1·03, 1·19), ref: 1st-2nd position), malaria in children (aPR = 1·30 (1·22, 1·38)), maternal age ≤ 25 (aPR = 1·17 (1·08, 1·27), ref: ≥ 35 years), maternal anaemia (aPR = 1·13 (1·07, 1·19)), lack of maternal education (aPR = 1·10 (1·02, 1·18), ref: ≥ secondary), number of children under 5 in household (aPR = 1·07 (1·00, 1·14) for ≥ 3, ref: 0-1), unimproved sanitation facilities (aPR = 1·12 (1·02, 1·22)) and low/middle household incomes (aPR = 1·16 (1·04, 1·30) and aPR = 1·13 (1·01, 1·26), respectively, ref: high). The population-attributable fraction was estimated at 8·2 % (6·3, 10·1 %) for child-related modifiable factors, 11·1 % (5·7, 16·3 %) for maternal-related factors, 15·8 % (8·6, 22·5 %) for household-related factors and 30·9 % (24·0, 37·2 %) for the combination of all modifiable factors. This study highlighted a high prevalence of childhood anaemia in Togo and showed that a high proportion of this could be prevented.