Using AI to control energy for indoor agriculture
30 September 2024
Published online 28 February 2018
A new statistical modelling technique could show decision-makers where to focus resources to improve achievements of sustainable development targets.
Many African countries have improved women’s education and child nutrition at the national level, but new analyses show that communities within countries are still lagging.
Simon Hay, professor of global health at the University of Washington’s Institute for Health Metrics and Evaluation (IHME), and his team gathered data from household surveys on years of education and child nutrition conducted in 51 African countries from 2000 to 2015. The data was incorporated into statistical computer models to predict patterns that were then used to fill in information gaps where data was missing.
The method enabled the team to map changes across Africa in women’s education and child nutrition over time at the level of communities living within 5x5 kilometre-scale boundaries.
The researchers found that nearly all countries in Africa have at least one region where children’s health is improving, yet not a single country is expected to put an end to childhood malnutrition by 2030, the United Nations’ second Sustainable Development Goal.
They also found that national and provincial maps on education could be misleading, as residents in many communities were found to have significantly fewer years of schooling than national averages.
Previous estimates had shown that many African countries are expected to reduce educational inequalities based on income, gender and location by 2030, the UN’s fourth Sustainable Development Goal.
Studies have found that improvements in education for women of reproductive age are associated with improvements in child nutrition and decreases in child mortality. Understanding the distribution of child growth failure and women’s education across Africa is important for improving child health, explains Hay.
“It is the first time researchers have looked at child growth failure and years of education on a continental scale at such a fine, local resolution,” says Hay. “Generally we found urban populations are better educated and nourished than rural populations.”
The team noted that many communities that did well in 2000 continued to do well in 2015, while many communities in which residents had significantly lower levels of nutrition and basic schooling in 2000, continued to have low nutrition and education levels in 2015 — a common scenario in northern Nigeria and Chad, says Hay.
Stunting, a low height for age, was the most prevalent form of child growth failure, an indicator of child under-nutrition, with some of the highest stunting rates in 2000 and 2015 present in southern Niger and northern Zambia. Still, many countries showed improvements. Large areas in Algeria, Mozambique, Burkina Faso and Ghana, for example, showed reductions in stunting prevalence from 2000 to 2015.
Child wasting, a low weight for height, which is correlated with environmental factors such as crop yields and food availability, showed high prevalence in a band across the continent, with concentrations in Niger, South Sudan, and Burkina Faso in 2000. The Afar region in Ethiopia showed a high prevalence of wasting both in 2000 (16.7%) and in 2015 (21.7%). Sites of high prevalence remained even in countries with low national rates, such as Kenya, which had a national prevalence of 5.7% but rates as high as 28.2% in the western Rift Valley province.
Prevalence of wasting in southern Africa remained consistently low across the study period.
High prevalence of underweight children, defined as a low weight for age, was found in a band stretching across Africa’s Sahel region, from southern Mali in the west to the Horn of Africa in the east. The prevalence was low in the northern coast of Africa, in countries near the Gulf of Guinea, and in southern Africa. In Nigeria, rates were below 10% in some regions and more than 30% in northern areas in 2015. Rwanda reduced their national prevalence of underweight children from 22.3% in 2000 to 8.7% in 2015. Angola and the Democratic Republic of Congo showed substantial improvements at the national level, but remaining hot spots remained at local levels.
Persistently low levels of educational attainment for women were found across the Sahel region, particularly in northern Nigeria, South Sudan and northern Kenya. For example, Ekiti State in western Nigeria had the highest average number of years of education (11.3) among women in the country in 2015, whereas many states in the north of the country had averages below two years. The lowest four regions across all of Africa, with an average of less than six months of education for women of reproductive age, were in rural areas of Chad. Particular improvement was found for 20- to 24-year-old women in urban centres in Nigeria, Kenya Ghana, Sudan and South Africa.
Men have significantly more years of educational attainment compared to women in the Sahel region and in Central Africa, particularly in northern regions of Nigeria and Kenya.
There were no significant differences by sex for any district within South Africa, Botswana, Zimbabwe, and Rwanda.
Evaluating the success of global health goals, especially in low-income countries, is a constant challenge, requiring the right tools for measuring outcome changes in target populations, says Kelley Lee of Canada’s Simon Fraser University, whose research considers how globalization affects disease, and the implications for strengthening global governance.
Lee says the important contribution of Hay’s team’s research is that it improves how we map global health needs.
“The results reveal patterns of need that can remain hidden by using national-level data alone. More granular data can reveal hidden deficits, enable finer targeting of populations, and hence provide more accurate measures of success,” she says.
Global health mapping needs to become even more sophisticated, adds Lee, so that researchers can analyse patterns of health and disease tied not only to physical location but also to variables such as gender, race or ethnicity, occupation and socioeconomic status.
Hay and his team have made their maps widely accessible to decision-makers and researchers through online interactive visualization tools. They plan to analyse a wider range of risk factors and diseases that impact under-five-mortality, such as vaccine coverage, respiratory infections, and diarrhoea, first in Africa and then in other low- and middle-income countries.
doi:10.1038/nmiddleeast.2018.24
Stay connected: