The prevalence of anemia among pre-school children in Bangladesh was noted to be 52.10% (rural 53.70% and urban 51.70%). This was considerably higher than the global prevalence of anemia of 24.8% [2]. WHO consider child anemia as one of the severe public health problem with a prevalence of greater than 40% in Bangladesh [2]. Higher prevalence rates of child anemia have been observed in other countries such as Indonesia (58.7%) [21], Benin (82%), Mali (83%) [22] and Ghana (78.4%) [23]. However, the prevalence rate in Bangladesh is higher those of neighboring countries Pakistan (33.2%) [24] and India (31.4%) [25] located in the South Asia region. It is also higher than other developing nations like Haiti (38.8%) [26] and Brazil (32.8%) [27].
Most previous studies on child anemia among pre-school children in Bangladesh were limited to the rural areas [13,14,15]. There was a published study on anemia in pre-school children in Bangladesh that was based nationally representative hierarchy structural dataset (BDHS dataset) [16]. The study applied a single level logistic regression analysis to determine the effects of socio-economic and demographic factors on child anemia. In our opinion, the use of single level statistical model is not appropriate for analyzing this type of nested dataset [18]. We feel that a multilevel regression models should be preferable, and this model has been used for several other studies that were based on BDHS datasets [9, 17, 28, 29]. In this study, two-level logistic regression models were used to determine the effects of socio-economic, demographic, and behavioral factors with anemia among pre-school children in Bangladesh. Moreover, the value of median odds ratio (MOR) was found to be 1.293, indicating that there was a variation anemia status among the 600 enumeration areas (geographical clusters).
This study showed that under-nourished children were more likely to have anemia compared to normal or over-nourished child. This is expected as anemia is one of the clinical and investigation indicators for mal-nourishment. Previous study that was based on BDHS-2011 dataset reported that prevalence of stunted, wasted and underweight children under age 5 was 41, 16 and 36% respectively, and that most of these children were from poor family environment [17]. We also observed that wealth index was an important predictor of child anemia, there was increasing tendency to have anemia with increasing wealth index of the families. We would expect that children from poor family were more likely to be under-nourished. These results were in agreement with other studies from Mali [22] and Indonesia [21]. Our study also showed that the anemic mother was an important risk factor for child anemia, and this has been reported in other studies from Pakistani [24] and Haiti [26]. In Bangladesh, 42% of ever-married women age 15–49 were found to be anemic, with most of them living in poor family environments [17].
Our study noted that age was another factor that was associated with anemia. Young children have limited body reserve, and would be more dependent on their parents for adequate nutrition from daily food intake. Studies on children in Pakistani [24], Haiti [26] and Brazil [27] reported that younger children were more vulnerable to anemia compared to older children. Our study also demonstrated that children living in the Dhaka division were less likely to be anemia than children in other divisions such as Barisal and Rangpur. Being the capital of the nation, the general standard of living would be better since more people staying here would be from the higher wealth quintile compared to those living in the other geographical divisions of the country [17].
The results of the multivariate logistic models in the Khan et al. study [16] that was based on the same dataset showed that water source, wealth index, maternal anemia, age of the children, stunting and division were risk factors for anemia in Bangladeshi pre-school children. When we analyze the data after removing the cluster effect, we were able to identify nutrition level of the child and religion as additional risk factors for this condition.
Most of the factors associated with anemia among pre-school children in Bangladesh were related to poverty. Over the decade, family wealth quintiles, childhood nutrition status and school attendance rate in Bangladesh have generally improved. The level of stunting among children under age 5 has declined from 51% in 2004 to 36% in 2014, while the level of underweight has declined from 43% in 2004 to 33% in 2014 [30]. The authorities should focus on improving the nutritional status both children and mothers, and also eradication of poverty. Based on the observation that prevalence of anemia was highest in children below age of 2 years old, remedial measures should target young children to prevent irreversible adverse effect on the growth and development of these children. We do not have subsequent information on anemia to compare since blood test was not part of the variable in the subsequent BDHS 2014.
Limitation of this study
Since this study was based on secondary data, we were not able to investigate all factors that may be relevant to anemia in children, including eating habits, parasite infestations, previous hospitalization, availability of filtered water, use of nutritional supplements and gestational birth weight. Prospective study focusing on more specific and relevant variables would be able to generate more useful information.
Despite general improvements of various nutritional and health parameters over the last few years, we were not able to provide any information in the trend since blood Hb level was not part of subsequent BDHS 2014. Blood Hb level is less likely to be influenced by short term alterations in the external environments, therefore trend of change would better reflect the nutritional or health status of the children.