Table 1 depicts the percent distribution of children under five years of age along with underweight prevalence, crude, and adjusted odds ratio by background characteristics of the child and the mother in Northeast states, India.
The mothers in the age group of 25-34 years had the most infants, followed by those in the age group of 15-24 years. 61.2 percent of mothers have a secondary or higher level of education. As shown in the table 80 percent of the children were having an average size or more than average. A birth order of one is seen in 68.9 percent of the children. 71.1 percent of children were also receiving breast milk at the time of the study. About 85 percent of the 32,005 children in the survey lived in rural regions. Hindu children accounted for 46.9% of all children, the highest percentage, led by Muslims (32.1%), Christians (18.1%), and children of other religions (2.9%). Males accounted for nearly 89 percent of household heads. 56 percent of the households had 5 to 8 members. 25.3 percent of the children in the study come from the poorest families, 21.2 percent from the poorer family, 19.8 percent from the middle-income family, 19.8 percent from the middle-income community, 18.7 percent from the wealthier family, and the remaining children come from the richest family. Table 1 shows that the Schedule tribe has the highest percentage of children (28.7%), led by OBC (18.6%), Schedule caste (8.6%), and Others (8.6%). (44.0%). In addition, 38.7percent of households said they don't have upgraded toilet facilities, 73.3 percent said they don't have clean drinking water, and 21.2 percent reported that they don't use safe cooking fuel.
Column 2 shows the prevalence of underweight. The prevalence of underweight has increased with an increase in the age group of the mother, i.e., from 15-24 years to 35-49 years. The prevalence of underweight in children has declined dramatically when the mother's educational degree grows, from uneducated mothers (41.1%) to those of at least secondary schooling (21.2%), which is still substantial. The chi square value is considerably high (1.93.30, p=0.001), suggesting a high relationship between the mother's level of schooling and status of underweight among children. Children with small size at the time of birth have shown more prevalence of underweight (34.8%) than those with the average or above (25.5%) the size at birth. The chi square value is again very high (228.42, p=0.001) hence, a very strong association between the size of child at the time of birth and the underweight can be seen. Male child has shown more prevalence of underweight than female child. As the age of child increases the prevalence of underweight increases significantly from 18.2 percent to 31.7 percent. Chi square value (328.10, p=0.001) shows a very high association between the age and underweight prevalence up-to age five. Children with birth order one and birth interval one have a higher prevalence of underweight (36.5%), whereas those with a birth order of less than two have the lowest prevalence (24.7%). As a consequence, there is a strong positive association between birth order and underweight (ꭓ2 value 288.50, p=0.001). Those children who were never breastfed show a higher prevalence of underweight than those who stopped or are still breastfeeding. The table also indicates that children living in rural areas have a higher prevalence (28.9%) of underweight than those living in urban areas (18.6%). Indicates that children living in rural areas have a higher level of starvation than those living in urban areas. As a result, a child's residency has a significant effect on the prevalence of underweight with chi square value 216.79 (p=0.001). Muslims have shown highest prevalence underweight with 36.5 percent, led by Hindus (23.6%). Also, the sex of the household head and household members does not show any significant impact on the prevalence of underweight as the chi square value is very low and insignificant. The prevalence is lowest amongst the children belonging to the richest family (13.2%) and highest among poorer wealth quantile with 37.2 percent, hence there is a very strong and negative association between the wealth index of the household and underweight prevalence (ꭓ2 value 1038.92, p=0.001). Schedule Caste has the lowest prevalence of underweight (25.5%). Similarly, children that do not have improved toilet facilities or clean drinking water have higher prevalence of underweight than those who do, indicating that sanitation and cleanliness plays a substantial role in a child’s underweight status. Before applying the multivariable regression analysis, we check the collinearity of all predictor variables with variance inflation factor and result shows less than 5.7, and it could be considered that there was no multicollinearity, and therefore, all predictor variables were included in the binary logistic regression model.
Column 3 and 4 depicts binary logistic regression (crude and adjusted, respectively) for underweight among children under five years of age in the Northeast States, India (2015-16). It shows the crude and adjusted odds ratios for the association between underweight by various background characteristics.
Children whose mother's level of education was no education and primary were 1.30 (p=0.001) times and 1.74 (p=0.001) times more likely to be underweight than those whose mother's level of education was secondary or above. Children with small size at birth have a 1.32 (p=0.001) times higher risk of being underweight as compared to those having size average and above size at birth. Male children were having a 0.13 (p=0.001) times lower risk of being underweight as compared its counterpart. The risk of being underweight was 1.71 (p=0.001), 2.01 (p=0.01), 2.01 (p=0.001), and 2.24 (p=0.001) times higher in the ages 1, 2, 3, 4 respectively as compared to those at the age 0. Children who were ever breastfed and then stopped were at 0.11 (p=0.05) times less risk of becoming underweight as compared to those who were never breastfed. Muslims had 1.31 (p=0.001) times, and Christians had 1.19 (p=0.05) times higher risk of being underweight as compared to Hindus. Children belonging to poorest had 2.12 (p=0.001), poorer had 1.87 (p=0.001), middle had 1.89 (p=0.001), richer had 1.57 (p=0.001) times higher risk of being underweight as compared to children belonging to the richest wealth index. Scheduled Tribes had a 0.42 (p=0.001) times lesser risk of being underweight as compared to that of others. Households with unimproved toilet facilities had 1.11 (p=0.01) times higher risk of being underweight than those with improved one.
Odds Ratio Graphs
Table 2 shows the results of multilevel analysis of the variables related to underweight among children under five years of age in Northeast states, India, NFHS-4 (2015-16). Underweight risk was 0.07 (p=0.05) times lesser for children whose mothers were in the age group 25-34 years relative to those whose mothers were in the age group 15-24 years. Children whose mother's level of education was no education or primary were 1.13 (p=0.01) times and 1.34 (p=0.001) times more likely to be underweight respectively than those whose mother's level of education was secondary and above. As compared to children who are average or above in size at birth, those who are small in size at birth have a 1.41 (p=0.001) times higher chance of being underweight. Female children were having 1.15 (p=0.001) times higher risk of underweight as compared to that of male children. The risk of being underweight was 1.77 (p=0.001), 2.17 (p=0.01), 2.28 (p=0.001), and 2.44 (p=0.001) times higher in the ages of 1, 2, 3, and 4 respectively as compared to those in the age 0. Hence, as age increases risk of being underweight increases up-to age five which means, age 1-5 are crucial years of life. Children with a birth order two and interval one had a 1.29 (p=0.001) times greater chance of being underweight as compared to those with a birth order of one. Muslims had 1.36 (p=0.001), Christians had 1.15 (p=0.05) times higher risk of being underweight as compared to those of Hindus. Children in the richer wealth quintile had 1.58 (p=0.001), middle wealth quantile had 1.89 (p=0.001), poorer wealth index children had 2.05 (p=0.001), and poorest had 2.45 (p=0.001) times higher chance of being underweight as compared to that of the richest wealth quintile. Schedule Tribe had 0.40 (p=0.001) times lower risk of being underweight as compared to that of others. Those children with no safe cooking fuel had in their household had shown a 1.12 (p=0.05) times higher risk of being underweight than those with safe cooking fuel.
Present study shows that in northeastern states the risk of underweight is the lowest among Schedule Tribes with the lowest prevalence and risk as compared to its counterparts whereas in India, the overall prevalence of underweight is highest among STs unlike the results shown in some studies already being done (Usmani G, 2017). In present study, the prevalence and risk of becoming underweight was lower among boys than in girls, like in another study, underweight was found to be more common in girls because they were fed less than boys (Chen, 1981). Study also suggests that age 1-5 years are crucial age and hence need and extra care.
Table 3 depicts the risk difference between underweight of children under five years of age belonging to poor and non-poor household in the northeast, India (2015-16). Three of the eight states studied, Assam, Arunachal Pradesh, and Meghalaya, have shown higher risk differences. In Assam, children from poor households, on average, are underweight 11.1 percentage points more often than those of non-poor households. Similarly, in Arunachal Pradesh and Meghalaya, those belonging to poor household are 9.3 percentage and 8.3 percentage points more often to be underweight, respectively than those to a non-poor household.