Millennium Development Goal shortfalls in Zimbabwe: Analysing the impact of access to water and sanitation on early childhood morbidity

Figures & data

Table 1. Overview of selected MDG indicators and targets, Zimbabwe 1990–2013.

Zimbabwe MDG estimates	1990	1995	2000	2005	2010	2012	1990 minus 2012	Results	2015 target
Urban % improved drinking water^a	100	99	99	98	98	97	−3.00	Net reduction	100
Rural % improved drinking water^a	71	70	70	69	69	69	−2.00	Net reduction	100
Total % improved drinking water^a	79	79	80	80	80	80	1.00	Slight increase	100
Urban % improved sanitation^a	54	53	53	52	52	52	−2.00	Net reduction	100
Rural % improved sanitation^a	35	35	34	33	33	32	−3.00	Net reduction	100
Total % improved sanitation^a	41	41	40	40	40	40	−1.00	Net reduction	100
Under-five child mortality rate^a	74.6	94.6	102.6	96.6	96	88.5	13.90	Net increase	50.0
Infant mortality rate^a	50.4	59.2	61	57.1	58.7	55	4.60	Net increase	33.8
Under-five % moderately underweight^a	13	–	–	14	10.1	9.7	−3.30	Reduction	6.5
Under-five % severely underweight^a	1.5	–	–	4.1	2.2	1.9	0.40	Increase	0.8
% Stunted (reflecting chronic malnutrition)^b	29.8	–	–	35	32	32	2.20	Increase	14.9
Gender Parity Index: primary education^a	0.99	0.97	0.97	–	–	1	0.01	improvement	1
Gender Parity Index: secondary education^a	0.88	0.84	0.88	–	–	1	0.12	improvement	1
Gender Parity Index: tertiary education^a		–	–	–	0.8	0.78	n/a	insufficient data	1
Total adult literacy rate^c	62	–	–	91		99.6	37.60	improvement	100
Male adult literacy rate^d	70	–	–	–	–	89.6	19.60	improvement	100
Female adult literacy rate^d	56	–	–	–	–	92.1	36.10	improvement	100

Sources:

^aUnited Nations Statistics Division (2014a).

^bUNICEF (2013).

^cUnited Nations Statistics Division (2014b).

^dUNESCO Institute for Statistics (2013).

Figure 1. Conceptual model.

Table 2. Prevalence of children under age five experiencing growth faltering in Zimbabwe from 1988 to 2010/11.

ZDHS survey year	Severe chronic malnourishment (height for age) −3 SD	Chronic malnourishment (height for age) −2 SD	Current nutritional status (very low weight for height) −3 SD	Current nutritional status (low weight for height) −2 SD	Composite growth faltering (weight for age) −3 SD	Composite growth faltering (weight for age) −2 SD
1988	8.7	20.3	0.2	1.1	1.6	9.9
1999	9.4	26.5	1.6	6.4	1.5	13
2005/06	11.4	29.4	1.3	6.4	3.2	16.6
2010/11	10.7	31.9	0.7	3	1.9	9.7

Source: ZDHS Country Reports (Zimbabwe Central Statistical Office & Demographic and Health Surveys, 1989:96, 98 and 100; Zimbabwe Central Statistical Office & Macro International Inc., 2007:170; Zimbabwe National Statistics Agency & ICF International, 2012:142).

Note: Unit of analysis is all children age three to 59 months. SD = standard deviation.

Table 3. Descriptive statistics for analysis variables, Zimbabwe 1988, 1999, 2005/06, and 2010/11.

	2010/11		2005/06		1999		1988
Variable	N	Proportion	N	Proportion	N	Proportion	N	Proportion
Child morbidity
Stunted, age three to 59 months^a	4 279	0.25	4 127	0.26	2 754	0.25	2 467	0.29
Household contamination
Safer drinking water sources^b	4 280	0.34	4 127	0.38	2 754	0.35	2 469	0.31
Improved toilets (JMP)^c	4 280	0.29	4 127	0.34	Not available		Not available
Unshared type of toilet facility^d					2 644	0.31	Not available
Unshared flush toilet only					2 644	0.10	Not available
Shared or unshared facility^e							2 469	0.58
Shared or unshared flush toilet							2 469	0.27
Type of place of residence:
Child lives in Shona region	4 280	0.47	4 127	0.41	2 754	0.42	2 469	0.39
Child lives in Ndebele region	4 280	0.14	4 127	0.17	2 754	0.18	2 469	0.19
Child lives in urban area	4 280	0.27	4 127	0.27	2 754	0.32	2 469	0.24
Mother's characteristics:
Currently married (1 = yes)	4 280	0.84	4 127	0.83	2 754	0.80	2 469	0.91
Literacy (1 = yes, reads)^f	4 280	0.86	4 127	0.78	Not available		2 469	0.77
Incomplete primary education					2 644	0.29
Completed primary education					2 644	0.45
Valid N (listwise)	4 279		4 127		2 754		2 467

Source: Analysis of ZDHS 1988, 1999, 2005/06, and 2010/11 (unit of analysis is children age three to 59 months for years 1994–2011 and age three to 39 months for 1988).

^aChild morbidity data are only available for age three to 39 months in 1988.

^bSafer drinking water sources are defined by the Joint Monitoring Program (JMP) for water supply and sanitation recommendations (UNICEF.org).

^cImproved toilets are defined in 2010/11 and 2005/06 as unshared and flush, ventilated improved pit latrine, and pit latrine with slab (as recommended by JMP).

^dUnshared toilet facility in 1999 includes flush, traditional pit toilet, and ventilated pit latrine (with no information on whether pit latrine has a slab).

^eShared or unshared toilet facility in 1988 includes flush, traditional pit toilet, and ventilated pit latrine (not JMP improved category; no pit latrine slab info).

^fLiteracy data are not available in the 1999 dataset; therefore mother’s level of education is substituted.

Table 4. Comparing percentages of young children in households with water on premises in Zimbabwe 1988, 1999, 2005/06, and 2010/11.

	1988	1999	2005/06	2010/11
Water on premises for all young children’s households	30.8	34.9	37.7	33.7
Water on premises in urban young children’s households	95.8	95.1	94.6	75.7
Water on premises for young children’s households in Shona regions	39.3	48.9	47.5	37.0
N (children age three to 59 months with anthropometric data)	2469	2644	4127	4280

Source: Analysis of ZDHS (Zimbabwe Central Statistical Office & Demographic and Health Surveys, 1989; Zimbabwe Central Statistical Office & Macro International Inc., 2000, 2007; Zimbabwe National Statistics Agency & ICF International, 2012).

Note: Unit of analysis is children age three to 59 months with missing cases excluded.

Table 5. Survey year differences for safe drinking water and improved sanitation models, Zimbabwe 2005/06 and 2010/11 (pooled).

	Drinking water models				95% CI for Exp(B)		Improved sanitation models				95% CI for Exp(B)
	B	Wald	Odds ratio		Lower	Upper	B	Wald	Odds ratio		Lower	Upper
Survey year (1 = 2010–11)	−0.059	1.628	0.943	NS	0.861	1.032	0.490	94.049	1.632***		1.478	1.801
Constant	−0.647	388.584	0.524***				−0.665	408.572	0.514***
−2 Log Likelihood	10742.73						9068.48
Probability > χ^2	0.000						0.000
Total sample size	8407						6868

Source: Author analysis of ZDHS data 2005/06 and 2010/11 (Zimbabwe Central Statistical Office & Macro International Inc., 2007; Zimbabwe National Statistics Agency & ICF International, 2012).

Notes: ‘Safe drinking water’ denotes water in residence and Joint Monitoring Program recommendation for piped water, water from communal

tap, covered well, boreholes, and bottled water.

Unit of analysis is all children age three to 59 months.

B = coefficient and Exp(B) = odds ratio.

Survey year attributes: 0 = year 2005/06 and 1 = year 2010/11.

***p<0.001.

Table 6. Determinants of access to safe drinking water for children age zero to 59 months, Zimbabwe 2005/06 and 2010/11 (pooled).

	Shona region model			95% CI for Exp(B)			Ndebele and Shona model			95% CI for Exp(B)			Maternal characteristic model			95% CI for Exp(B)			Saturated model			95% CI for Exp(B)
	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper
Survey year (1 = 2010/11)	−0.086	3.447	0.917	NS	0.837	1.005	−0.096	4.23	0.908*		0.829	0.995	−0.202	17.618	0.817***		0.744	0.898	−0.256	16.967	0.774***		0.686	0.875
Region of residence
Child lives in Shona region	0.53	127.624	1.699***		1.55	1.862	0.638	141.184	1.893***		1.704	2.103	0.686	155.906	1.986***		1.784	2.212	−0.087	1.536	0.917	NS	0.799	1.052
Child lives in Ndebele region							0.272	18.362	1.313***		1.159	1.487	0.321	24.098	1.378***		1.213	1.567	−0.628	49.943	0.534***		0.448	0.635
Mother's characteristics
Currently married (1 = yes)													0.03	0.228	1.031	NS	0.91	1.168	−0.038	0.214	0.963	NS	0.82	1.131
Mother's literacy (1 = reads well)													1.424	323.926	4.154***		3.557	4.851	0.616	47.149	1.852***		1.553	2.209
Type of place of residence
Family lives in urban area																			3.612	2 060.285	37.05***		31.699	43.304
Constant	−0.857	499.292	0.425***				−0.96	440.0	0.383***				−2.181	463.225	0.113***				−1.851	253.737	0.157***
−2 log likelihood	10 615.00						10 596.73						10 186.92						7 024.55
Probability > χ^2	0.000						0.000						0.000						0.000
Total sample size	8 407						8 407						8 407						8 407

Sources: Author analysis of ZDHS data 2005/06 and 2010/11 (Zimbabwe Central Statistical Office & Macro International Inc., 2007; Zimbabwe National Statistics Agency & ICF International, 2012).

Notes: ‘Safe drinking water’ denotes water in residence and Joint Monitoring Program recommendation for piped water, water from communal tap, covered well, boreholes, and bottled water.

Unit of analysis is all children age three to 59 months.

B = coefficient and Exp(B) = odds ratio.

Survey year attributes: 0 = year 2005/06 and 1 = year 2010/11.

*p<0.05.

***p<0.001.

Table 7. Determinants of access to safe drinking water for children age zero to 59 months in 1999 and age zero to 39 months in 1988.

	Shona region model				95% CI for Exp(B)		Ndebele and Shona model				95% CI for Exp(B)		Saturated model				95% CI for Exp(B)
1999	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper
Shona/Ndebele region
Child lives in Shona region	1.067	160.121	2.905***		2.463	3.427	1.544	230.843	4.685***		3.839	5.718	0.934	25.88	2.545***		1.776	3.647
Child lives in Ndebele region							1.269	104.938	3.557***		2.79	4.534	0.394	2.495	1.483	NS	0.909	2.42
Mother's characteristics
Currently married (1 = yes)													−0.202	0.923	0.817	NS	0.541	1.234
Incomplete primary education													−0.646	8.504	0.524**		0.339	0.809
Completed primary education													−0.808	15.819	0.446***		0.299	0.664
Type of place of residence
Family lives in urban area													5.359	744.692	212.44***		144.574	312.165
Constant	−1.112	352.727	0.329***				−1.59	375.318	0.204***				−2.323	77.948	0.098***
−2 log likelihood	3 255.54						3 149.87						1 181.84
Probability > χ^2	0.000						0.000						0.000
Total sample size	2 733						2 733						2 733
1988
Shona/Ndebele region
Child lives in Shona region	0.651	53.907	1.917***		1.611	2.28	1.021	97.987	2.776***		2.268	3.398	0.273	2.973	1.313	NS	0.963	1.791
Child lives in Ndebele region							1.020	67.856	2.772***		2.175	3.533	0.030	0.021	1.031	NS	0.681	1.56
Mother's characteristics
Currently married (1 = yes)													−0.482	4.336	0.617*		0.392	0.972
Literate (1 = yes)													−0.908	34.606	0.404***		0.298	0.546
Type of place of residence
Family lives in urban area													5.515	572.185	248.273***		158.014	390.088
Constant	−1.085	332.810	0.338***				−1.456		0.233***				−1.245	22.88	0.288***
−2 log likelihood	2 994.443						2 927.014						1 422.275
Probability > χ^2	0.000						0.000						0.000
Total sample size	2 469						2 469						2 469

Source: Author analysis of ZDHS data 1988 and 1999 (Zimbabwe Central Statistical Office and Demographic and Health Surveys, 1989; Zimbabwe Central Statistical Office and Macro International Inc., 2000).

Notes: ‘Safe drinking water’ denotes water in residence and Joint Monitoring Program recommendation for piped water, water from communal

tap, covered well, boreholes, and bottled water.

Unit of analysis is all children age three to 59 months.

B = coefficient and Exp(B) = odds ratio.

*p<0.05.

**p<0.01.

***p<0.001.

Table 8. Determinants of modern toilets in households with children age zero to 59 months, Zimbabwe 2005/06 and 2010/11 (pooled).

	Shona model				95% CI for Exp(B)		Ndebele and Shona model				95% CI for Exp(B)		Maternal Characteristics model				95% CI for Exp(B)		Saturated model				95% CI for Exp(B)
	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper	B	Wald	Odds Ratio		Lower	Upper
Survey year (1 = 2010/11)	0.486	90.966	1.625***		1.471	1.796	0.477	86.896	1.61***		1.457	1.78	0.39	55.42	1.473***		1.33	1.632	0.35	44.21	1.418***		1.279	1.572
Region of residence
Child lives in Shona region	0.028	0.305	1.028	NS	0.931	1.136	0.213	13.685	1.238***		1.106	1.386	0.24	17.021	1.271***		1.134	1.425	0.146	5.952	1.157*		1.029	1.302
Child lives in Ndebele region							0.481	49.327	1.617***		1.414	1.85	0.47	45.659	1.6***		1.396	1.833	0.363	25.865	1.438***		1.25	1.654
Mother's characteristics
Currently married (1 = yes)													−0.263	15.949	0.769***		0.676	0.875	−0.278	17.764	0.757***		0.665	0.862
Mother's literacy (1 = reads well)													0.742	100.35	2.101***		1.817	2.429	0.627	68.148	1.872***		1.613	2.173
Type of place of residence
Family lives in urban area																			0.402	48.543	1.495***		1.335	1.674
Constant	−0.676	314.106	0.509***			−0.857	335.921	0.425***				−1.239	166.953	0.29***				−1.176	149.342	0.309***
−2 log likelihood	9 068.17						9 018.89						8 894.28						8 845.83
Probability > χ^2	0.000						0.000						0.000						0.000
Total sample size	6 868						6 868						6 868						6 868

Sources: Author analysis of ZDHS data 2005/06 and 2010/11 (Zimbabwe Central Statistical Office & Macro International Inc., 2007; Zimbabwe National Statistics Agency & ICF International, 2012).

Notes: ‘Modern toilet’ denotes Joint Monitoring Program recommendation for 2005/06 and is modified in 2010/11 to include unshared flush toilets, following Humphrey's (2009) suggestion of flush toilets.

Unit of analysis is all children age three to 59 months.

B = coefficient and Exp(B) = odds ratio.

Survey year attributes: 0 = year 2005/06 and 1 = year 2010/11.

*p<0.05.

***p<0.001.

Table 9. Determinants of modern toilets in young children's households, Zimbabwe 1999 and 1988.

	Ndebele and Shona model				95% CI for Exp(B)		Ndebele/Shona and mother's characteristics model				95% CI for Exp(B)		Saturated model				95% CI for Exp(B)
1999	B	Wald	Odds ratio		Lower	Upper					Lower	Upper	B	Wald	Odds ratio		Lower	Upper
Region of residence
Child lives in Shona region	0.779	22.224	2.180***		1.577	3.015	0.583	11.711	1.792**		1.283	2.502	−0.529	7.245	0.589**		0.401	0.866
Child lives in Ndebele region	1.452	66.938	4.271***		3.016	6.048	1.498	60.750	4.471***		3.068	6.515	0.634	7.881	1.885**		1.211	2.934
Mother's characteristics
Currently married (1 = yes)							0.353	4.027	1.423*		1.008	2.01	0.258	1.716	1.295	NS	0.88	1.905
Incomplete primary education							−2.124	94.533	0.120***		0.078	0.183	−0.316	1.436	0.729	NS	0.435	1.223
Completed primary education							−1.413	92.514	0.243***		0.183	0.325	−0.840	26.818	0.432***		0.314	0.593
Child lives in urban area													4.418	168.370	82.935***		42.551	161.643
Constant	−2.848	443.289	0.058***				−2.062	90.683	0.127***				−4.895	152.116	0.007***
−2 log likelihood	1 703.59						1 541.19						1 112.5
Probability > χ^2	0.000						0.000						0.000
Total sample size	2 733						2 733						2 733
	Shona region model				95% CI for Exp(B)		Ndebele and Shona model				95% CI for Exp(B)		Saturated Model				95% CI for Exp(B)
1988	B	Wald	Odds ratio		Lower	Upper	B	Wald	Odds ratio		Lower	Upper	B	Wald	Odds ratio		Lower	Upper
Region of residence
Child lives in Shona region	0.524	32.232	1.688***		1.409	2.022	1.134	99.188	3.108***		2.487	3.886	−0.99	0.183	0.906	NS	0.575	1.426
Child lives in Ndebele region							1.301	97.162	3.674***		2.836	4.758	0.661	7.191	1.937**		1.195	3.14
Mother's characteristics
Currently married (1 = yes)							0.103	0.41	1.109	NS	0.808	1.521	−0.197	0.418	0.822	NS	0.453	1.491
Mother's literacy (1 = yes, reads)							1.378	89.624	3.968***		2.983	5.278	0.639	6.106	1.895*		1.141	3.145
Child lives in urban area													6.063	616.065	429.837***		266.296	693.813
Constant	−1.24	400.628	0.289***				−3.016	183.894	0.049***				−3.412	75.403	0.033***
−2 log likelihood	2 824.585						2 605.673						916.593
Probability > χ^2	0.000						0.000						0
Total sample size	2 469						2 469						2 469

Sources: Author analysis of ZDHS data 1988 and 1999 (Zimbabwe Central Statistical Office & Demographic and Health Surveys, 1989; Zimbabwe Central Statistical Office & Macro International Inc., 2000)

Notes: Unit of analysis is all children age three to 59 months (1999) and age three to 39 months (1988).

‘Modern toilet’ denotes only flush unshared toilets in 1999, and only flush toilets in 1988.

B = coefficient and Exp(B) = odds ratio.

*p<0.05.

**p<0.01.

***p<0.001.

Table 10. Impact of water and sanitation on growth stunting, Zimbabwe 1988–2011.

	Water model						Water and sanitation model
	B	Wald	Odds ratio		95% CI for Exp(B)		B	Wald	Odds ratio		95% CI for Exp(B)
					Lower	Upper					Lower	Upper
2010/11
Water at home	−0.169	4.913	0.845*		0.728	0.981	−0.136	3.06	0.873	NS	0.749	1.017
Modern toilet (Joint Monitoring							−0.162	3.956	0.851*		0.725	0.998
Program recommendation)
Constant	−1.056	605.876	0.348***				−1.021	488.123	0.36***
−2 log likelihood	4 785.48						4 781.48
Probability > χ^2	0.026						0.011
Total sample size	4 288						4 288
Children experiencing stunting	1 060						1 060
1999
Water at home	−0.447	20.446	0.639***		0.527	0.776	−0.430	18.816	0.65***		0.535	0.79
Modern toilet (Joint Monitoring							−0.322	9.929	0.724**		0.593	0.885
Program recommendation)
Constant	−0.965	319.958	0.381***				−0.877	213.841	0.416***
−2 log likelihood	2 940.99						2 930.78
Probability > χ^2	0.000						0.000
Total sample size	2 733						2 733
Children experiencing stunting	656						656
2005–06
Water at home	−0.305	16.681	0.737***		0.637	0.853	−0.264	11.261	0.768**		0.659	0.896
Modern toilet (Joint Monitoring							−0.135	2.863	0.873	NS	0.747	1.022
Program recommendation)
Constant	−0.923	445.394	0.397***				−0.893	360.763	0.41***
−2 log likelihood	4 735.28						4 732.39
Probability > χ^2	0.000						0.000
Total sample size	4 117						4 117
Children experiencing stunting	1 084						1 084
1988
Water at home	−0.797	56.041	0.450691***		0.366	0.555	0.073	0.161	0.688	NS	0.753	1.537
Modern toilet (Joint Monitoring							−1.13	31.359	0.323***		0.218	0.48
Program recommendation)
Constant	−0.679	175.976	0.507***				−0.666	168.771	0.513629***
−2 log likelihood	2 911.20						2 880.51
Probability > χ^2	0.000						0.000
Total sample size	2 467						2 467
Children experiencing stunting	716						716

Sources: Author analysis of ZDHS data 1988, 1999, 2005/06 and 2010/11 (Zimbabwe Central Statistical Office & Demographic and Health Surveys, 1989; Zimbabwe Central Statistical Office & Macro International Inc., 2000, 2007; Zimbabwe National Statistics Agency & ICF International, 2012).

Notes: Unit of analysis is all children age three to 59 months (2010/11, 2005/06, and 1999) and age three to 39 months (1988).

‘Modern toilet’ denotes Joint Monitoring Program recommendation for 1999 through 2011. ‘Modern toilet’ denotes only flush toilets in 1988.

B = coefficient and Exp(B) = odds ratio.

*p<0.05.

**p<0.01.

***p<0.001.

United Nations Statistics Division, 2014a. Millennium development goals indicators. http://mdgs.un.org/unsd/mdg/Default.aspx.

 Google Scholar

UNICEF, 2013. Zimbabwe Statistics. http://www.unicef.org/infobycountry/zimbabwe_statistics.html.

 Google Scholar

United Nations Statistics Division, 2014b. Millennium development goals indicators – Literacy rates of 15–24 years old, both sexes, percentage. http://mdgs.un.org/unsd/mdg/SeriesDetail.aspx?srid = 656&crid = 716.

 Google Scholar

UNESCO Institute for Statistics, 2013. Adult and youth literacy: National, regional and global trends, 1985–2015. http://www.uis.unesco.org/Education/Documents/literacy-statistics-trends-1985-2015.

 Google Scholar

Zimbabwe Central Statistical Office & Demographic and Health Surveys, 1989. Zimbabwe demographic and health survey, 1988. Zimbabwe Central Statistical Office and Demographic and Health Surveys, Columbia, MD. http://www.dhsprogram.com/pubs/pdf/FR43/FR43.pdf.

 Google Scholar

Zimbabwe Central Statistical Office & Macro International Inc, 2000. Zimbabwe demographic and health survey, 1999. Zimbabwe Central Statistical Office and Macro International Inc, Calverton, MD. http://www.dhsprogram.com/pubs/pdf/FR116/FR116.pdf.

 Google Scholar

Zimbabwe National Statistics Agency (ZIMSTAT) & ICF International, 2012. Zimbabwe demographic and health survey, 2010–11. ZIMSTAT and ICF International Inc, Calverton, MD. https://dhsprogram.com/pubs/pdf/FR254/FR254.pdf.

 Google Scholar

Zimbabwe Central Statistical Office (CSO) & Macro International Inc, 2007. Zimbabwe demographic and health survey, 2005–06. CSO and Macro International Inc, Calverton, MD. http://www.dhsprogram.com/pubs/pdf/FR186/FR186.pdf.

 Google Scholar

Humphrey, JH, 2009. Child undernutrition, tropical enteropathy, toilets, and handwashing. The Lancet 374, 1032–5. doi: 10.1016/S0140-6736(09)60950-8

 PubMed Web of Science ®Google Scholar