In Haiti, differences in the prevalence of anemia between urban and rural areas have been observed.
To identify moderating factors that may help explain the difference in the prevalence of anemia in children from poor urban vs. rural areas of Haiti.
This cross-sectional study used secondary data from urban and rural school-based trials that assessed the effectiveness of a nutrition intervention. The study was registered at ClinicalTrials.gov as NCT02747524. A total of 300 rural- and 981 urban- children between 2.5-13 years of age were included in this analysis. Effect modification in a binary logistic generalized linear mixed model was conducted using sample weights in SPSS® version 26. Models were adjusted for age and income. School cluster was included as random effect.
In rural areas, stunting was more prevalent in children with anemia vs. no anemia, (16.6%, and 6.3%, P = 0.008), respectively. Also, rural children with anemia lived with fewer adults vs. rural children with no anemia, ([Formula: see text] = 2.83±1.29, and 3.30±1.54, P = 0.005), respectively. In poor urban areas, helminth morbidities were more frequent in children with anemia vs. no anemia, (21.9% vs. 13.9, P = 0.011), respectively. In the combined sample, stunting, [AOR = 2.05; 95%CI (1.32-3.18)], age [AOR = 0.89; 95%CI (0.85-0.93)], and households with more adults [AOR = 0.77; 95%CI (0.67-0.87)] were associated with anemia. Effect modification by place of residence was observed in households with more adults (t = 3.83, P<0.001). No other nutritional, dietary, sanitation or morbidity factors or effect modifiers were observed.
In this sample, factors associated with anemia differed in poor urban and rural children from Haiti including family structure and helminth morbidities. Stunting and lower age increased the odds of anemia in the combined sample. Family structure appears to have an important role in anemia, and further research understanding the influence of family structures in anemia is needed.