Child Anthropometrics and Geographic Differences in Malnutrition in Uganda Malnutrition Looking at the data further, we can identify In 2009/2010, Uganda’s Bureau of Statistics, in geographic areas with particularly high rates of collaboration with the World Bank, conducted the malnutrition. The data show that rural areas face a first wave of the Uganda National Panel Survey higher burden of malnutrition than do urban areas. (UNPS), which collects detailed data on household Thirty-seven percent of under-5 children in rural welfare and income-generating activity. The UNPS’ Uganda are stunted, compared to 15 percent of final sample includes 2,9751 households that are children in urban areas. Similarly, the representative at the national level; waves 2 and 3 underweight prevalence of 17 percent in rural are currently available online and wave 4 will run areas is more than double that of urban areas, from 2013 to 2014. This note summarizes the which stands at 8 percent (see Figure 1). anthropometric data and resulting malnutrition indicators from UNPS-wave 1.2 Figure 1: Malnutrition estimates, by rural and urban Background: Child Anthropometry 40   The three anthropometric indicators most often 35   referenced for monitoring malnutrition in children 30   Rural   are: stunting, or low height-for-age; underweight, 25   Urban   %   low weight-for-age; and wasting, low weight-for- 20   height. More specifically, these figures represent 15   children whose height-for-age, weight-for-age, and 10   weight-for-height fall more than two standard 5   deviations below the median of internationally 0   accepted growth standards. Thus, a child is Stunted   Underweight   Wasted   labeled stunted if he or she has a height-for-age z- score that is less than -2. The UNPS was stratified regionally and by rural Table 1 shows the stunting, underweight, and and urban areas, and has six domains of analysis. wasting prevalence estimates for Uganda. The These domains include Kampala, all other urban data reveal that 34 percent, 15 percent, and 5 areas, Central Rural, Eastern Rural, Northern percent of children 6-59 months old, are stunted, Rural, and Western Rural. Table 2 shows the underweight, and wasted, respectively. stunting and underweight prevalence estimates for these six domains of analysis. There is substantial Table 1: Malnutrition estimates variation in malnutrition between the six domains.   Prevalence  (%)   The stunting prevalence in Western Rural is more  (Std.  Error)   than three times the prevalence in other urban Stunted   34    (2)   areas and approximately 50 percent more than in Underweight   Central and Northern Rural areas. However, we 15    (1)   find that underweight prevalence rates in the four Wasted   5    (1)   rural domains of analysis are more or less                                                                                                                 comparable, though significantly greater than 1 Intended sample size was 3,123 households underweight prevalence in Kampala and other 2 The final sample for this analysis included 2,086 children aged 6- urban areas. 59 months.   Table 2: Stunting and Underweight, by domain of Table 3: Malnutrition and household head’s analysis education Stunted  (%)   Underweight  (%)   HH  head  has  been   HH  head  has  never   Domain   (SE)   (SE)     to  school  (SE)     been  to  school  (SE)   Kampala   18  (7)      9  (4)   Stunted                              33              (2)   40                  (4)   Other  urban   13  (3)      6  (2)   Underweight                            14***    (1)   25***      (4)   Central  Rural   32  (4)   15  (2)   Wasted                                  5**        (1)   3**            (1)   Eastern  Rural   37  (4)   16  (2)   Note: ** Difference is significance at p<0.05; *** Difference is significance at p<0.01 Northern  Rural   31  (3)   18  (3)   Western  Rural   45  (4)   19  (3)   Household ownership of certain assets, particularly those that increase connectivity and access to technology, are correlated with better nutrition Identifying Vulnerable Sub- outcomes. Children living in households without a Populations television are more than three times as likely to be In developing countries, boys typically exhibit stunted and more than twice as likely to be higher rates of malnutrition than girls. Uganda underweight than those without (see Table 4). proves to be no exception; 39 percent of under-5 Similar positive effects are observed for radio and boys are stunted, compared to only 30 percent of cell phone ownership at the household level. under-5 girls (see Figure 3). Table 4: Malnutrition and asset ownership Asset   Stunted     Underweight     Wasted     Figure 3: Malnutrition estimates, by gender TV  (%)                 45   Owns  a  tv   11***      7***    4     40   No  tv   36***      17***    5   Boys   35   Radio  (%)         30   Girls   Owns  a  radio   33   14**   4   25   %   No  radio   36   21**   6   20   15   Cell  phone  (%)         10   Owns  a  cell  phone   28***   13***   5   5   No  cell  phone   40***   19***   5   0   Note: ** Difference is significance at p<0.05; *** Difference is significance at p<0.01 Stunted   Underweight   Wasted   Analysis of the UNPS wave 1 anthropometric data suggests that Uganda faces a high burden of The data show that a household head’s level of malnutrition, a conclusion that matches that of the education also plays a role in child malnutrition. On current literature. Identifying particularly vulnerable average, children living with literate household groups, such as children in rural areas, boys, and heads exhibit lower rates of stunting and those living in households without a television or underweight3. Similarly, children living with phone, can help policy makers target nutrition educated household heads are less likely to be programs more effectively. underweight than those whose heads’ have never This brief was prepared by Ilana Seff, World Bank, attended school (14 vs. 25 percent). Somewhat based on data collected by Uganda’s Bureau of surprisingly however, living with an educated Statistics (UBOS) as part of the Living Standards household head is correlated with higher rates of Measurement Study – Integrated Surveys on wasting in under-5 children, though the magnitude Agriculture (LSMS-ISA) project. The full dataset is of the difference is not large (see table 3). available for download at UBOS via http://www.ubos.org.                                                                                                                   3 Difference is only significant for underweight prevalence