This study analyzed the increase of anemia in children under-5 in Ethiopia between 2011 and 2016. Further analysis of this increase in anemia is required to understand which socio-demographic segments were more affected. More than half, 56.9%, of Ethiopian children under-5 are anemic according to the 2016 EDHS. This is a little lower than the average anemia prevalence of 59.9% in children under-5 from 27 Sub-Saharan African countries . Anemia in children is a severe public health problem in Ethiopia and needs urgent attention to address its underlying causes.
Anemia increased across all age categories and children aged 36–47 months had the highest increase (42.5%) of anemia (Table 1). Age had the highest (Cramer’s V = 0.25) strength of association with anemia than the rest of the predictor variables studied (Table 2). Similar to findings from Armenia  and India , infants aged 9–11 months had the highest prevalence (72.7%) of anemia. At the same time except for 9–11 month infants, all forms of anemia increased across all age categories (Fig. 1). The lowest affected group was 48–59 months, this is a time range when the child’s growth rate decreases, a better food variety is available for the child, and the child is physiologically more equipped to take part in the family dining table, for example mastication of meat . The highest increase of severe anemia was among children 6–8 months (Fig. 1). The increase of anemia after reaching 6 months could be due to the depletion of fetal iron stores  and the increased iron requirements placed on the body around 6 months of exclusive breastfeeding after birth [1, 23].
Regardless of their educational level, anemia increased in children born to mothers of all educational levels. The highest prevalence of anemia is among children born from mothers with no formal education (Table 1). The highest increase in anemia (65.8%) was seen in children born to mothers with higher than secondary education (Table 1). Although the smaller sample size and lower baseline prevalence of anemia could be a possible reason, this finding is counter-intuitive because mothers with higher levels of education are more likely to be employed and have better knowledge about infant feeding. It’s important to note that having an above secondary level education does not necessarily mean the mothers are employed. Moreover, educated mothers could be practicing poor dietary habits themselves and for their children because they spend many hours at work. Between 2011 and 2016, the gap in the prevalence of anemia between children born to mothers with no education and above secondary education has narrowed from 15 to 8% (Table 1). Figure 2 shows that mild and moderate anemia showed the highest increase among mothers above secondary level education (Fig. 2).
Anemia increased across all the wealth quintile groups and the highest increase (41.5%) was among children born from the lowest wealth quintile (Table 1). Severe anemia decreased among children born in the second and middle wealth quintile (Fig. 3). Household wealth index was significantly associated with the anemia status of children (p < 0.0001). This association is supported by some studies [13, 24, 25], however, other studies in Ethiopia [14, 26, 27] have reported that there was no statistically significant relationship between household wealth and anemia. The strength of association between anemia and wealth quintile increased from weak (χ2 = 61.6, V = 0.079) in 2011 to moderate (χ2 = 166.9, V = 0.134), moreover, the gap between the highest and lowest quintiles increased from 12% in 2011 to 20% in 2016 (Table 1). The highest wealth quintile showed the second highest increase (33.4%) of anemia (Table 1). This increase in the highest wealth quintile is unexpected because families with the highest income are more likely to be food secure and provide better healthcare to their children. Food secure households might lack adequate dietary practices and engaged in eating less nutritious, undiversified foods. Furthermore, the highest quintile doesn’t necessarily mean they are rich; it simply represents the highest quintile in comparison with the rest of the population. Furthermore, most of the increase of anemia in the highest wealth quintile is mostly mild anemia (Fig. 3).
Anemia in children shows significant differences across Ethiopian regions. Except for Benishangul Gumuz region, anemia increased in all the regions of Ethiopia (Table 1). The highest increase was in Tigray (42.9%) and South Nations Nationalities People’s (SNNP) regions (35.5%). Somali region had the highest decrease in mild anemia and the highest increase in moderate anemia from its baseline in 2011. Moderate and mild anemia decreased in Affar and Benishangul Gumuz regions but Benishangul Gumuz is the only region where all forms of anemia declined (Fig. 4). Because regional administrations in Ethiopia are ethnically based, anemia might be affected by cultural and dietary practices of populations living in a given region.
Regional health and nutrition reports from Benishangul Gumuz exhibit indicators which are better than most regions and possibly explain why anemia could have decreased in this region. For example, Benishangul Gumuz has the second-highest minimum dietary diversity score (eating from at least four food groups) and the second-highest minimum acceptable diet when compared with the other regions. Additionally, Benishangul Gumuz has the second-highest median breastfeeding duration, the lowest prevalence of severe anemia and the second lowest prevalence of anemia in children. It also has the lowest percentage of children who described as “very small” after birth and the lowest percentage of children under-5 with fever in the 2 weeks before the survey. When compared with normal practice, children in Benishangul Gumuz receive the highest increase in food during diarrhea . A national nutrition survey in 2015 reported that Benishangul Gumuz had the second lowest vitamin A deficiency after the capital, Addis Ababa, and the highest percentage of children who drunk thin porridge (semisolid food) before the survey . Thus, the decrease in anemia in children in Benishangul Gumuz could be partly due to the dietary and childcare practices of the people living in this region.
Ethiopia is off course in meeting its objective of decreasing anemia in children under-5 as outlined in its NNPs. The increase of anemia attests that the implementation of the nutrition programs needs to be strengthened and coordinated more efficiently. The other regions of Ethiopia could benefit from the successful experience of Benishangul Gumuz in reducing anemia in children. Further studies regarding infant feeding practices and dietary customs in Benishangul Gumuz region are required to identify the reasons behind this exceptional decrease of anemia in this region. Because this study was a secondary analysis of a cross-sectional study, it is associated with the cause-effect problem of such cross-sectional studies. Hemoglobin was the only indicator used to measure the anemia status of children, thus the specific type of anemia could not be determined.