Polygynous family system, neighbourhood contexts and under-five mortality in sub-Saharan Africa

Figures & data

Table 1. Distribution of respondents by family type and according to selected individual-level characteristics.

Characteristic	Total, N = 54 842 (%)	Monogamous union, N = 39 171 (%)	Polygynous union, N = 15 671 (%)	p value
Maternal age				< 0.001
15 to 24 years	25.9	28.8	18.7
25 to 34 years	50.5	50.4	50.7
35+ years	23.6	20.8	30.6
Maternal age at birth				< 0.001
<18 years	6.4	6.9	5.0
18 to 24 years	35.0	37.6	28.5
25 to 34 years	44.1	42.9	47.2
35+ years	14.6	12.7	19.4
Education				< 0.001
None	49.7	40.9	71.6
Primary	23.1	25.1	17.9
Secondary	22.7	28.1	9.4
Tertiary	4.5	5.9	1.1
Birth interval				< 0.05
<24 months	21.4	21.3	21.9
24 to 35 months	38.1	37.6	39.2
36+ months	40.4	41.1	38.9
Parity				< 0.001
<3	29.3	34.2	17.1
3 to 4	30.9	32.4	27.1
5+	39.8	33.4	55.8
Wealth index				< 0.001
Poorest	23.1	20.6	29.4
Poorer	22.0	20.9	24.7
Middle	18.9	18.7	19.4
Richer	18.9	19.9	16.2
Richest	17.1	19.9	10.4
Religion				< 0.001
Christianity	31.7	39.2	13.2
Islam	36.9	30.3	53.2
Others	31.4	30.5	33.6

Table 2. Distribution of respondents by family type and according to selected contextual characteristics.

Characteristic	Total, N = 54 842 (%)	Monogamous union, N = 39 171 (%)	Polygynous union, N = 15 671 (%)	p value
Country				< 0.001
Kenya	13.8	17.1	5.5
Niger	23.5	21.1	29.4
Nigeria	54.5	51.5	61.9
Zimbabwe	8.2	10.3	3.2
Place of residence				< 0.001
Urban	29.1	33.5	18.2
Rural	70.9	66.5	81.8
Community poverty level				< 0.001
Poor	75.1	71.4	84.5
Middle	20.6	23.6	13.1
Rich	4.3	5.0	2.4
Community level of education				< 0.001
Low	34.0	28.4	47.9
Middle	35.4	35.3	35.6
High	30.6	36.2	16.5

Figure 1. Percentage distribution of the study sample by family system.

Figure 2. Child survival plot by family structure.

Table 3. Results from Cox proportional hazard model showing the effects of polygyny and selected variables on child survival in selected sub-Saharan African countries.

Characteristic	Model 1	Model 2	Model 3	Model 4	Model 5
Family structure
Monogamous	1.0	1.0	1.0	1.0
Polygynous	1.37 (1.27 to 1.48)***	1.15 (1.05 to 1.25)**	1.17 (1.08 to 1.29)***	1.13 (1.03 to 1.23)**
Respondents’ age
15 to 24 years	1.0	1.0		1.0	1.0
24 to 34 years	0.85 (0.78 to 0.92)***	0.89 (0.77 to 1.03)		0.88 (0.76 to 1.03)	0.88 (0.76 to 1.02
35+ years	1.01 (0.92 to 1.11)	0.88 (0.72 to 1.08		0.87 (0.71 to 1.07)	0.87 (0.71 to 1.06)
Respondents’ age at birth
<18 years	1.0	1.0		1.0	1.0
18 to 24 years	0.71 (0.63 to 0.81)***	0.79 (0.61 to 1.02)		0.80 (0.62 to 1.04)	0.81 (0.62 to 1.03
25 to 34 years	0.69 (0.61 to 0.79)***	0.74 (0.55 to 0.98)*		0.74 (0.56 to 0.99)*	0.74 (0.55 to 0.98)*
35+ years	0.85 (0.74 to 0.99)*	0.87 (0.63 to 1.18)		0.87 (0.63 to 1.19)	0.86 (0.63 to 1.18)
Respondents’ educational level
No education	1.0	1.0		1.0
Primary	0.69 (0.63 to 0.76)***	0.94 (0.83 to 1.05)		0.96 (0.85 to 1.09)
Secondary	0.63 (0.56 to 0.69)***	1.07 (0.91 to 1.26)		0.98 (0.82 to 1.17)
Higher	0.52 (0.40 to 0.67)***	0.95 (0.67 to 1.34)		0.88 (0.62 to 1.25)
Birth interval
<24 months	1.0	1.0		1.0	1.0
24 to 35 months	0.68 (0.62 to 0.74)***	0.69 (0.63 to 0.76)***		0.69 (0.63 to 0.76)***	0.69 (0.63 to 0.76)***
36+ months	0.50 (0.45 to 0.55)***	0.53 (0.48 to 0.59)***		0.53 (0.48 to 0.59)***	0.53 (0.48 to 0.59)***
Parity
<3	1.0	1.0		1.0	1.0
3 to 4	1.20 (1.09 to 1.33)***	0.69 (1.45 to 2.07)***		1.73 (1.44 to 2.07)***	1.74 (1.45 to 2.08)***
5 or more children	1.58 (1.45 to 1.73)***	2.35 (1.92 to 2.88)***		2.36 (1.92 to 2.90)***	2.39 (1.94 to 2.93)***
Wealth index
Poorest	1.0	1.0		1.0	1.0
Poorer	1.11 (1.01 to 1.23)*	1.16 (1.04 to 1.30)**		1.15 (1.03 to 1.29)**	1.14 (1.02 to 1.28)*
Middle	0.82 (0.73 to 0.91)***	0.88 (0.78 to 1.00)		0.90 (0.79 to 1.03)	0.89 (0.78 to 1.06)
Richer	0.85 (0.75 to 0.95)**	0.96 (0.84 to 1.11)		1.05 (0.90 to 1.23)	1.02 (0.87 to 1.18)
Richest	0.58 (0.50 to 0.67)***	0.68 (0.57 to 0.81)***		0.80 (0.64 to 1.00)*	0.76 (0.62 to 0.94)
Religion
Christianity	1.0	1.0		1.0	1.0
Islam	1.67 (1.52 to 1.84)***	1.31 (1.15 to 1.49)***		1.02 (0.87 to 1.18)	1.09 (0.95 to 1.26)
Others	1.40 (1.26 to 1.54)***	1.16 (1.01 to 1.34)*		0.92 (0.71 to 1.18)	0.97 (0.75 to 1.24)
Country
Kenya	1.0		1.0	1.0	1.0
Nigeria	2.04 (1.83 to 2.28)***		1.73 (1.41 to 2.13)***	1.68 (1.34 to 2.11)***	1.69 (1.35 to 2.12)***
Niger	1.82 (1.60 to 2.07)***		1.45 (1.16 to 1.82)***	1.56 (1.13 to 2.16)**	1.57 (1.14 to 2.17)**
Zimbabwe	1.64 (1.40 to 1.91)***		1.35 (1.04 to 1.73)**	2.20 (1.58 to 3.05)***	2.15 (1.55 to 3.00)***
Place of residence†
Urban	1.0		1.0	1.0
Rural	1.39 (1.27 to 1.54)***		1.35 (1.21 to 1.51)***	1.20 (1.03 to 1.38)**
Community poverty level†
High	1.0		1.0	1.0
Middle	1.68 (1.40 to 2.03)***		1.56 (1.14 to 2.14)**	1.78 (1.26 to 2.50)***
Low	2.06 (1.75 to 2.42)***		1.54 (1.12 to 2.13)**	1.62 (1.13 to 2.32)**
Community level of education†
Low	1.0		1.0	1.0
Middle	0.83 (0.75 to 0.92)***		0.90 (0.81 to 1.00)*	0.90 (0.81 to 1.01)
High	0.65 (0.59 to 0.72)***		0.76 (0.68 to 0.86)***	0.84 (0.73 to 0.96)**
Interaction effects
Family structure and education
Monogamy and primary/no education	1.0				1.0
Monogamy and secondary/higher education	0.73 (0.65 to 0.82)***				0.98 (0.83 to 1.17)
Polygyny and primary/no education	1.29 (1.19 to 1.40)***				2.24 (1.46 to 3.45)***
Polygyny and secondary/higher education	0.90 (0.70 to 1.15)				1.95 (1.19 to 3.21)**
Family structure and place of residence
Monogamy × urban	1.0				1.0
Monogamy × rural	1.39 (1.24 to 1.58)***				1.17 (1.02 to 1.36)*
Polygyny × urban	1.23 (1.00 to 1.50)*				0.73 (0.57 to 0.92)**
Polygyny × rural	1.85 (1.63 to 2.10)***				†
Family structure and community poverty
Monogamy × rich	1.0				1.0
Monogamy × Middle	2.0 (1.47 to 2.69)***				1.75 (1.20 to 2.56)**
Monogamy × poor	2.24 (1.69 to 2.98)***				1.64 (1.10 to 2.44)*
Polygyny × rich	1.11 (0.62 to 1.99)				0.76 (0.44 to 1.32)
Polygyny × middle	2.89 (2.07 to 4.03)***				1.13 (0.91 to 1.40)
Polygyny × poor	2.98 (2.23 to 3.95)***				†

Note: Data presented as hazard ratio (95% confidence interval).

*p < 0.05.

**p < 0.01.

***p < 0.001.

†Omitted in Model 5 to prevent multi-collinearity.

Table 4. Results from multilevel Cox proportional hazard model showing the effects of polygyny and selected variables on child survival in selected sub-Saharan African countries.

Characteristic	Model 0	Model 1	Model 2	Model 3	Model 4
Fixed effects
Family structure
Monogamous		1.0	1.0	1.0	1.0
Polygynous		1.30 (1.22 to 1.39)*	1.14 (1.05 to 1.23)*	1.24 (1.16 to 1.33)*	1.14 (1.05 to 1.23)*
Respondents’ age
15 to 24 years			1.0		1.0
24 to 34 years			0.72 (0.65 to 0.80)*		0.72 (0.65 to 0.79)*
35+ years			0.75 (0.66 to 0.86)*		0.75 (0.66 to 0.85)*
Birth interval
<24 months			1.0		1.0
24 to 35 months			0.64 (0.58 to 0.70)*		0.64 (0.58 to 0.70)*
36+ months			0.64 (0.59 to 0.70)*		0.64 (0.59 to 0.70)*
Parity
<3			1.0		1.0
3 to 4			1.26 (1.13 to 1.41)*		1.26 (1.13 to 1.41)*
5 children or more			1.58 (1.39 to 1.79)*		1.58 (1.39 to 1.80)*
Wealth index
Poorest			1.0		1.0
Poorer			1.09 (0.99 to 1.20)		1.11 (1.01 to 1.22)*
Middle			0.83 (0.74 to 0.93)*		0.86 (0.76 to 0.97)*
Richer			0.79 (0.70 to 0.89)*		0.82 (0.72 to 0.95)*
Richest			0.62 (0.53 to 0.71)*		0.65 (0.55 to 0.78)*
Religion
Islam			1.0		1.0
Christianity			0.91 (0.84 to 0.99)*		0.91 (0.83 to 1.00)*
Others			0.80 (0.62 to 1.03)		0.80 (0.62 to 1.03)
Place of residence
Urban				1.0	1.0
Rural				1.42 (1.30 to 1.55)*	1.07 (0.96 to 1.20)
Community poverty level
Low				1.0	1.0
Middle				0.98 (0.86 to 1.12)*	0.99 (0.86 to 1.14)
High				0.96 (0.82 to 1.13)	0.91 (0.76 to 1.09)
Community level of education
Low				1.0	1.0
Middle				0.89 (0.80 to 1.00)*	0.90 (0.79 to 1.02)
High				0.91 (0.80 to 1.04)	0.91 (0.79 to 1.06)
Random effects	Empty	Polygyny covariate	Individual-level variables	Community-level variables	Full model
Community level  Variance (SE)  ICC	0.08 (0.05) 2.4	0.03 (0.01) 0.9	0 (0.0) 0.0	0.08 (0.04) 2.4	0.04 (0.02) 1.2
Model fit statistics  AIC  BIC	31 182.3 31 217.9	28 840.6 28 884.7	21 737.8 21 883.1	28 746.7 28 861.4	21 740.0 2195.7

Note: Data presented as hazard ratio (95% confidence interval).

*p < 0.05.

SE = standard error, ICC = intra-class correlation, AIC = Akaike information criterion, BIC = Bayesian Information criterion.