Syndicate content

Africa's success story: Infant mortality down

Gabriel Demombynes's picture

There is a tremendous success story in Sub-Saharan Africa that has only barely been recognized. Infant and under-5 mortality has plummeted in many countries in the region in recent years.

The under-5 mortality (U5MR) measure captures the number of children per 1000 live births who die before their 5th birthday. One of the Millennium Development Goals is a two-thirds decline in U5MR between 1990 and 2015, which would require an annual decline of 4.4 percent per year.


In the 20 countries for which recent data is available, 12 show rates of decline above this “MDG rate.” In particular, Senegal, Rwanda, Kenya, Uganda, and Ghana have experienced extremely large drops at a rate of more than 6 percent per year. This does not necessarily indicate that any particular country will meet the MDG. But it does tell us that the African Renaissance is bringing tangible benefits to the continent’s citizens. Because of this miracle, hundreds of thousands of parents will be spared the agony of the loss of a child.


The acceleration in the decline of child mortality in Sub-Saharan Africa was evident in a 2010 Lancet article, but that article was completed before much of the recent data showing the most dramatic declines was available.


(Click on the graph to see it larger)


One of us (Ritva) used to work on post-conflict Uganda in the 1990s analyzing household survey data on under-5 mortality.  The figures were 200 per 1000 live births and even higher.  Uganda has witnessed a huge improvement in survival of under 5 year-olds—today Uganda is at 90.  In Ethiopia, the second most populous country in Africa, U5MR fell from 123 to 88 between 2005 and 2010.  Its 35,000 health extension workers have brought basic preventive and curative health packages to the rural population en masse.  Case control studies have shown significant expansion of coverage of malaria interventions (artemisinin-based combination therapy and insecticide-treated bednets) and improvement in sanitation.


In a recent co-authored working paper, one of us (Gabriel) explores what may have driven part of this decline in Kenya, which had the fastest rate of infant (under age 1) mortality, with a drop of 7.6 percent per year between recent Demographic and Health Surveys. A decomposition analysis suggests that the huge increase in ownership of insecticide-treated bednets (ITNs) played a substantial role, by reducing malaria deaths. The paper applies some statistical machinery to the question, but the basic logic is simple.



The infant mortality decline took place chiefly in malaria high-risk zones and was entirely confined to postneonatal mortality (deaths after the 1st month.) Postneonatal mortality is much higher in malaria high-risk zones and much lower among households that own an ITN in malaria high-risk zones (both unconditionally and controlling for a wide variety of other factors). ITN ownership skyrocketed in Kenya between 2003 and 2008, going from 8 to 60 percent of households nationally and reaching 75 percent of households in high-risk zones.  Malaria is known not to have any substantial effect on neonatal mortality (deaths during the 1st month), and neonatal mortality does not vary across malaria risk zones and has remained constant over time in Kenya. This collection of facts suggests that ITNs played an important role in the decline of infant mortality in Kenya. This conclusion is also compatible with the findings of small-scale experimental and panel studies on the effects of ITNs on malaria prevalence and mortality.


In the Kenya decomposition analysis, the growth in ITN ownership explains just over half of the infant mortality decline. That still leaves a large chunk of the drop not explained by the factors we have analyzed. In future work, we will examine the broader story of infant and under-5 mortality declines across many countries in the region, which allow us to consider the role of country-wide factors like the broad improvement in living standards that have come with renewed economic growth.



Submitted by Rob Yates on
Has anyone spotted that the big fall in infant mortality in Africa over the last five years has coincided with the period when more and more African governments removed user fees for health services - especially for pregnant women and children? The Kenyan example is illustrative because bed net use only sky-rocketed post 2003 when the Government started distributing them free of charge. In these papers I charted the beginning of this trend to remove user fees In Africa and noted how many governments were starting this process by exempting women and children first 2012 marks an important anniversary in this area, as it is 25 years ago since a number of influential development agencies started encouraging African governments to charge user fees - including for young children. Following these excellent recent results one really thinks that these agencies owe it to the developing world to acknowledge that this policy was wrong.

Submitted by Nick Ericsson on
Rob - I'm from BBC Focus on Africa magazine. Can you email me your contact details: - it would be good to get in touch thanks Nick

Submitted by Robert Liebenthal on
Is there any evidence from other countries, apart from Kenya, showing the importance of ITN for the IMR? And, if so, what does it tell us about other factors that might have been important, like clean water, immunization etc?

Submitted by swati singh on
Sub- Saharan Africans have its bad record in past in case of poverty,hunger,IMR,MMR etc.But now it has changed its story.Sub-Saharan countries like Senegal,Rwanda,Kenya...experienced extremely large drops at a rate of more than 6%annually,which is higher than the MDG target of 4.4%per year till 2015.It was possible because of up to mark health facilities in rural areas, control of diseases like malaria and improvement in sanitation.Well-done Africans...........

Submitted by Gabriel on
Thanks for all the very good comments. Rob: Thanks for the link to your paper. I am not deeply familiar with the history and research on user fees for health, and I look forward to learning more. Robert: I haven't seen any research that examines the influence of ITNs for IMR using national data. This is something I would like to do using data for several countries. The Kenya paper does mention several relevant references. The most important is this review of evidence on the effect of ITNs on malaria infections: Lengeler, Christian. 2004. ―Insecticide-treated bed nets and curtains for preventing malaria.‖ In Cochrane Database of Systematic Reviews, ed. The Cochrane Collaboration and Christian Lengeler. Chichester, UK: John Wiley & Sons, Ltd, April Swati: I share your enthusiasm for recognizing that the widespread decline in infant mortality is a substantial achievement, for which the citizens and governments of the countries in questions deserve congratulations.

Dear, Would like to be member of your organization (WHO), and like to be inform/ updated. Our organization Worldview Mission ( please visit) focus in programs about Midwives & Infant. Our organization complies with the 8th major UN MDG'Ls to reduce the poverty by: 2015. Regard, Ms H. H. Oord Chair & Founder Worldview Mission

I sense some over-exuberance in this blog. I've graphed the top 8 declines indicated in this article using the World Bank's WDI data for under-5 mortality (Q5) from 1980 to 2010. The trends suggest that half of these 8 countries are simply following a better-than-average path of decline that goes back at least three decades; the other 4 appear to be recovering from a disaster in the 1990s that disrupted childhood mortality declines (HIV/AIDS for Uganda, Zambia and Kenya; Rwanda's genocide/state failure). Certainly, these declines are indeed a "good news story". They are steeper than the sub-Saharan average -- but graphing them produces trends that, for half of the "top-8" (Mozambique, Ethiopia, Ghana, Senegal) seem to be an extension of the "long slog" of broad, incremental improvements in public health, in education (particularly for women), income and urbanization that promoted declines in under-5 mortality elsewhere. The remaining 4 countries are recovering from disaster that affected childhood mortality. Zambia, Uganda and Kenya seem to be declining from upsurges in Q5 during the 1990s, likely the influence of mother-to-child transmission of HIV. Rwanda's declines in Q5 are steepest, but if one draws a line from Rwanda's pre-conflict period (1985), when Q5 was ~160 per thousand, and connects it with the 2010 level, ~90, the decline's slope is much like Ghana's; still impressive, but not as steep as indicated in the article. Yes, treated bed nets are likely contributing to these incremental declines; yes, improving access to health services works. But the underlying secular trends haven't dramatically changed -- childhood mortality in SSA has been declining for more than 3 decades. Four of these African states have, without disruption, done better than most SSA states; four of the 8 have shown the strength to recover from a significant disruption. It is great news ... just not so exciting. On the other hand, can you imagine a journalist picking up the news of ongoing secular declines and transforming it into the newspaper headline, "Sub-Saharan Africa: Long-term childhood mortality declines continue, even after disruptions" ? How exciting! (not really, unless you're in the public health game)

Submitted by Nachiket Mor on
Dear Dr. Cincotta, I think it is indeed insightful that these countries have been showing secular declines. But, while I am no expert in this, it is my understanding that as the rates get lower further declines get to be harder to achieve and that for each decrease of 10% different factors are responsible. Many Indian States for example are "bottoming out" having reach the magic number of 50 IMR because the steps needed now are very different. Work that we supported in Bombay (where IMR is below 30) failed to produce impact when identical work had produced some very dramatic impact in regions with IMR of 50 plus. Sincerely, Nachiket Mor

Having worked in Sudan, Angola and parts of Africa over the past 30 years on malaria control, it is gratifying to see your attribution of lower Infant Mortality to decreases in malaria. As an engineer I would like to offer an additional piece of good news, especially related to water resource development and rehabilitation. Current strategies for malaria control - although apparently effective - are using unsustainable methods such as bednets which last only 3-5 years, and spraying houses with biocides which last only a few months. Drugs in use are quickly provoking resistance in the malaria parasite. As long as mothers force infants to sleep under bednets, the IMR goes down. That is a nice statistic, however adults avoid sleeping under these hot and irritating nets, so the average usage in a community is less than half, and malaria transmission continues, despite the good showing among infants. Thus purchase of nets, drugs and biocides at increasingly expensive prices, will have to continue forever. The good news is that a more sustainable approach, which we demonstrated in the Gezira Irrigation System in central Sudan in the 1980's, is improving the irrigation and drainage system in ways which reduce malaria mosquito breeding. This protects everyone in the community, not just infants for a few years. Also, in a new irrigation project or in a resettlement project related to dam and reservoir construction, investment in screened housing is a much better way than bednets to protect a family from the malaria mosquitoes, and more economical in the long run for a host of reasons. For 10 years we held the malaria prevalence near 0.1% in the 2 million people of the Gezira, under the Blue Nile Health Project. Previous prevalences had exceeded 20%, and had made malaria a crippling public health scourge. We have published this in great detail, and I commend it to you as a way of expanding and sustaining the good news you have reported on malaria. So in construction of new irrigation systems, and in rehab of existing ones, malaria control can be made sustainable and economical, as we showed with the $100 million World Bank loan we received in 1984 for the Gezira Rehabilitation Program. It was developed with our Blue Nile staff along with Graham Clarkson and Nick Prescott from the Bank, and was a singular success is creating sustainable suppression of malaria.

Add new comment