Factors contributing to the reduction in childhood stunting in Bangladesh: a pooled data analysis from the Bangladesh demographic and health surveys of 2004 and 2017–18

Table 4 Decomposition estimates for stunting among children aged under five in Bangladesh, 2004 to 2017–18

Background characteristics	Coef.	Std. error	p-value	% Contribution
Mother characteristics
Age at first birth
Less than 18 years	Ref.
18 years or more	0.003	0.002	0.063	1.6
Educational status
Not educated	Ref.
Educated	0.020	0.004	0.000	10.5
Working status
Not working	Ref.
Working	0.007	0.002	0.006	3.6
Low BMI
No	Ref.
Yes	0.026	0.003	0.000	13.7
Father characteristics
Educational status
Not educated	Ref.
Educated	0.017	0.003	0.000	9.0
Working status
Not working	Ref.
Working	0.000	0.000	0.969	0.0
Child characteristics
Children age (in months)
0–23	Ref.
23–47	0.000	0.000	0.000	0.2
48–59	0.001	0.000	0.000	0.6
Sex
Male	Ref.
Female	0.000	0.000	0.971	0.0
Birth order
1	Ref.
2	0.001	0.001	0.117	0.6
3	0.000	0.000	0.721	0.0
4 and above	0.007	0.002	0.000	3.9
Household characteristics
Protected water source
No	Ref.
Yes	0.004	0.002	0.003	2.3
Improved toilet facility
No	Ref.
Yes	0.001	0.000	0.155	0.3
Wealth Index
Poorest	Ref.
Poorer	0.000	0.000	0.694	−0.1
Middle	−0.001	0.000	0.405	−0.5
Richer	0.001	0.000	0.000	0.6
Richest	0.008	0.001	0.000	4.0
Religion
Islam	Ref.
Others	0.000	0.000	0.201	0.1
Residential status
Urban	Ref.
Rural	0.000	0.001	0.952	0.0
Region
Barisal	Ref.
Chittagong	0.000	0.000	0.740	0.0
Dhaka	0.001	0.000	0.045	0.4
Khulna	0.001	0.000	0.005	−0.4
Rajshahi	0.000	0.000	0.143	0.2
Sylhet	0.000	0.000	0.563	0.0
				50.6
N (2004)	6003
N (2018)	7849
Predictive probability of being stunted in 2004 (p1)	0.498
Predictive probability of being stunted in 2017–18 (p2)	0.307
Difference (p1-p2)	0.191
Total explained	0.097

ISSN: 1471-2458