Urbanization, malaria transmission and disease burden in Africa

Abstract

Many attempts have been made to quantify Africa's malaria burden but none has addressed how urbanization will affect disease transmission and outcome, and therefore mortality and morbidity estimates. In 2003, 39% of Africa's 850 million people lived in urban settings; by 2030, 54% of Africans are expected to do so. We present the results of a series of entomological, parasitological and behavioural meta-analyses of studies that have investigated the effect of urbanization on malaria in Africa. We describe the effect of urbanization on both the impact of malaria transmission and the concomitant improvements in access to preventative and curative measures. Using these data, we have recalculated estimates of populations at risk of malaria and the resulting mortality. We find there were 1,068,505 malaria deaths in Africa in 2000 - a modest 6.7% reduction over previous iterations. The public-health implications of these findings and revised estimates are discussed.

Figure 1. Population totals from 1950–2050

The bar plots are stratified by urban (red) and rural (green) for the world (a) and Africa (b). Note that urban and rural population projections stop in 2030, after which total population is shown (blue). Data from the World Population Prospects Database (see the Online links box).

Figure 2. Mean population density (persons per km²) in 2000 of the 37 urban agglomerations (UA) in Africa

a | The bar plot shows the mean population density in 2000 of the 37 UA with more than one million inhabitants in 2003 at successive 5 km buffers from the UA edge. The Global Rural–Urban Mapping Project (GRUMP) urban mask^, defined the spatial extent of the 37 UA. Overlaying this on the gridded population of the world version 3 (GPWv3) allowed the population associated with these classes to be shown in the bar plot: urban (red), peri-urban (yellow), rural 1 (light green) and rural 2 (dark green, off the scale). The core UA (defined by its GRUMP extent) has a population density >1,000 persons per km²; peri-urban areas (defined as between 5–15 km from the UA edge) have population densities 250–1,000 persons per km² and rural areas (defined as >20 km from UA edge) have densities of <250 persons per km². To prevent this rural stratification being biased towards high population densities (as we conservatively trained population density on the largest African UAs), the rural class was further divided into rural 1 (100–250 persons per km² at 20–55 km from UA edge) and rural 2 (<100 persons per km² at >55 km from the UA edge). The vertical lines indicate the standard errors. b | The map shows these population densities plotted across Africa using GPWv3 (REF. 97). The map is exaggerated vertically to help resolve small-area urban and peri-urban classes across Africa.

Figure 3. Bar plots of APfEIR by land-use category (a) and APfEIR by dry or wet ecozone (b)

The red bars in part b are pooled urban and peri-urban data and the green bars are pooled rural 1 and rural 2 data. Dry and wet areas were defined by satellite sensor measurements of <0.35 and >0.35 normalized difference vegetation index (NDVI)^,, respectively. The NDVI is a measure of photosynthetic activity or ‘greenness’ of a region^, and is often used as a surrogate for moisture availability. In both plots the solid vertical lines indicate the standard error. APfEIR, annual P. falciparum entomological inoculation rates; ib/p/a, infected bites per person per annum; NDVI, normalized difference vegetation index.

Figure 4. Plot of PR by log₁₀ APfEIR

The red dots are pooled urban and peri-urban data (n = 33) and the green dots are pooled rural 1 and rural 2 data (n = 88). The fit line is drawn for all data. APfEIR, annual P. falciparum entomological inoculation rates; ib/p/a, infected bites per person per annum; PR, parasite prevalence ratio.

Figure 5. Childhood mortality in Africa

a | A box-and-whisker plot of all-cause childhood mortality in children (0–4 years of age), shown as deaths per 1,000 children per annum. b | A box-and-whisker plot of malaria-specific mortality in children (0–4 years of age), shown as deaths per 1,000 children per annum. The median (thick central line) is shown with the 25% and 75% percentile ranges (box depth) and the maximum and minimum (T-bars). These data are plotted against categorical classes of the PR (class 2, marginal risk areas, PR 0–0.25, n = 8; class 3, acute seasonal transmission PR 0.25–0.75, n = 24; class 4, stable endemic transmission, PR >0.75, n = 15)^, (see FIG. 6). PR, parasite prevalence ratio.

Figure 6. The MARA model of fuzzy climate suitability (FCS) for P. falciparum malaria transmission

FCS values vary between zero (totally unsuitable) to 1 (totally suitable) in an average year. The data are grouped into class 1 zero risk (FCS = 0), class 2 marginal risk (FCS 0–0.25), class 3 acute seasonal transmission (FCS 0.25–0.75) and class 4 stable endemic transmission (FCS >0.75)^,.

Figure 7. Plot of PR by log₁₀ APfEIR

The blue circles and blue fit-line are the original data (see FIG. 4). The red (urban), yellow (peri-urban) and green (rural 1) fit-lines show the change in the relationship between APfEIR and PR predicted by the decrease APfEIR with urban intensity outlined in FIG. 3a. Class 1–4 emphasize the categorical separation of PR across Africa (FIG. 6). The solid arrows show levels of urbanization that result in movement between classes and dashed arrows those that do not. These categorical changes are then used in the population at risk estimates to calibrate burden estimates (TABLE 2). APfEIR, annual P. falciparum entomological inoculation rates; ib/p/a, infected bites/person/annum; PR, parasite prevalence ratio.