Table 4

Decomposition of rural-urban differences in child nutrition outcomes – Bangladesh
Quantiles
Q10 Q25 Q50 Q75 Q90
Rural HAZ scores −3.5188 −2.7013 −1.8577 −1.0179 −0.1760
Urban HAZ scores −3.1753 −2.4321 −1.5227 −0.6166 0.2152
Observed raw gap in HAZ scoresa −0.3436*** −0.2692*** −0.3350*** −0.4013*** −0.3912***
Covariate effect (% contribution) −0.2694*** (78.43%) −0.2555*** (94.95%) −0.2493*** (74.39%) −0.2816*** (70.17%) −0.2434*** (62.19%)
Co-efficient effect (% contribution) −0.0741 (21.57%) −0.0136 (5.05%) −0.0858 (25.58%) −0.1197 (29.83%) −0.1478* (37.78%)
Contribution of individual characteristics to rural-urban differences in HAZ scoresb
Covariate effect Co-efficient effect
Q10 Q25 Q50 Q75 Q90 Q10 Q25 Q50 Q75 Q90
Size of effect −0.2694*** −0.2555*** −0.2493*** −0.2816*** −0.2434*** −0.0741 −0.0136 −0.0858 −0.1197 −0.1478*
Explained
Female gender of child −0.0001 (0.05%) 0.0003 (-0.15%) −0.0002 (0.11%) −0.0004 (0.22%) −0.0014 (0.67%) −0.0001 (0.14%) 0.0010 (-1.42%) 0.0024 (-2.20%) 0.0023 (-2.78%) −0.0022 (2.09%)
Age of child −0.0060* (3.16%) −0.0085* (4.38%) −0.0127* (6.69%) −0.0171* (9.59%) −0.0246* (11.76%) −0.0014 (2.00%) −0.0026 (3.69%) −0.0035 (3.21%) −0.0048 (5.81%) −0.0051 (4.84%)
Child vaccinated −0.0039 (2.06%) −0.0004 (0.21%) 0.0095*** (-5.00%) 0.0161*** (-9.03%) 0.0199*** (-9.52%) 0.0003 (-0.43%) −0.0002 (0.28%) −0.0005 (0.46%) −0.0007 (0.85%) −0.0005 (0.47%)
Age of mother −0.0021 (1.11%) −0.0009 (0.46%) 0.0000 (0.00%) 0.0004 (-0.22%) 0.0015 (-0.72%) 0.0016 (-2.29%) 0.0011 (-1.56%) −0.0035 (3.21%) −0.0080 (9.69%) −0.0106 (10.07%)
Mother’s working status −0.0027 (1.42%) −0.0027 (1.39%) −0.0018 (0.95%) −0.0021 (1.18%) −0.0026 (1.24%) −0.0030 (4.29%) −0.0004 (0.57%) −0.0007 (0.64%) 0.0006 (-0.73%) 0.0005 (-0.47%)
Mother’s education (yrs) −0.0495*** (26.09%) −0.0312*** (16.08%) −0.0194** (10.22%) −0.0214** (12.00%) 0.0065 (-3.11%) −0.0066 (9.43%) −0.0094 (13.33%) −0.0096 (8.81%) −0.0076 (9.20%) −0.0087 (8.26%)
Spouse’s education (yrs) −0.0424*** (22.35%) −0.0394*** (20.31%) −0.0339*** (17.85%) −0.0423*** (23.72%) −0.0453*** (21.66%) −0.0103 (14.71%) −0.0127 (18.01%) −0.0116 (10.64%) −0.0141 (17.07%) −0.0078 (7.41%)
Wealth index −0.0748*** (39.43%) −0.1026*** (52.89%) −0.1122*** (59.08%) −0.1019*** (57.15%) −0.1404*** (67.14%) −0.0497** (71.00%) −0.0482*** (68.37%) −0.0948*** (86.97%) −0.0643*** (77.85%) −0.0757*** (71.89%)
Extended family dummy 0.0122* (-6.43%) 0.0023 (-1.19%) −0.004 (2.11%) −0.0047 (2.64%) −0.0225*** (10.76%) −0.0008 (1.14%) 0.0011 (-1.56%) 0.0123** (-11.28%) 0.0140** (-16.95%) 0.0055 (-5.22%)
Dependency ratio −0.0151 (7.96%) −0.0079 (4.07%) −0.0196*** (10.32%) −0.0130** (7.29%) −0.0098 (4.69%) −0.0001 (0.14%) −0.0003 (0.43%) 0.0005 (-0.46%) 0.0003 (-0.36%) −0.0004 (0.38%)
No. of children < 5 yrs −0.0051 (2.69%) −0.0031 (1.60%) 0.0043 (-2.26%) 0.0082** (-4.60%) 0.0097** (-4.64%) −0.0001 (0.14%) 0.0002 (-0.28%) 0.0000 (0.00%) −0.0003 (0.36%) −0.0003 (0.28%)
Total −0.1897*** (100%) −0.1940*** (100%) −0.1899*** (100%) −0.1783*** (100%) −0.2091*** (100%) −0.0700*** (100%) −0.0705** (100%) −0.1090*** (100%) −0.0826*** (100%) −0.1053*** (100%)
Unexplained
Residual −0.0797 −0.0615 −0.0593 −0.1033* −0.0343 −0.0041 0.0569 0.0233 −0.0371 −0.0426

aRaw gap in HAZ scores computed as rural HAZ scores- urban HAZ scores. The negative sign of the observed raw gap figures reflects the fact that rural HAZ scores are lower than urban HAZ scores.

bFigures in brackets show percentage contribution of individual characteristics to the total explained effect.

Significant effects are in bold. Asterisks denote level of significance - ***, ** and * denote significance at the 1%, 5% and 10% level of significance respectively.

Srinivasan et al.

Srinivasan et al. BMC Public Health 2013 13:581   doi:10.1186/1471-2458-13-581

Open Data