Sub-Saharan Africa has become associated in the minds of many with news stories about famine, civil war, or hospitals overwhelmed by terminally ill AIDS patients. The continent’s success stories receive less news attention. Delaunay et al. present results from a study documenting the reductions in infant and child mortality achieved in rural Senegal since the 1960s. 1 In the Niakhar area, the probability of a child dying in the first five years of life in the mid-1960s remained 485 per 1000. Nationally, this proportion was around 300 per 1000. 2,3 The former statistic, at least, represents a level of mortality among infants and children more characteristic of 19th century Asia or 18th century Europe than of the second half of the 20th century. The improvement in child survival achieved since the 1960s is impressive. By the late 1990s under-5 mortality in Niakhar had fallen to about 213 per 1000, while estimates of the proportion of all Senegalese children dying before their fifth birthday fall in the range 115‐130 per 1000. 4,5 Senegal is now one of the lowest mortality countries in tropical Africa. Despite its relative proximity to the capital, Niakhar lies in the region of Senegal with the highest mortality. 3 In particular, in the 1960s the study area combined high mortality in infancy with exceptionally high mortality of children aged 1‐4. This pattern has been found in many parts of Africa and some other tropical regions but it seems to be most pronounced in the Sahel. 6 It was more extreme in Niakhar in the 1960s than in any other population that has been investigated in detail. 3,6,7 In this population, mortality rates rose in the second half of infancy, reaching a peak at about age 12 months, but remained very high right through the second year of life and into the third. This unusual age pattern of mortality was underlain by seasonal variation in numbers of births and deaths. Detailed analysis of the early data from Niakhar demonstrates that the age at which the risk of dying was highest differed according to season of birth: children were experiencing heavy excess mortality during their first two rainy seasons. 8 The overall pattern arose because the peak in mortality was particularly severe for those children aged 9‐15 months during the rains and because this was a relatively large group of children, as the birth rate also peaked at this time of the year. Delaunay et al. propose that under-5 mortality has been declining at a steady rate throughout the 37 years under study. Visual inspection of the relevant Figure in their paper suggests that one could argue equally plausibly that much of the decline in the risk of death occurred during the 1980s, with slower advances being made before and since. It would perhaps be asking too much of the data from a relatively small study, however, to attempt to differentiate between these accounts. It is clear though that it was children aged 6‐30 months that benefited from the early part of the mortality decline. Although mortality in this age range is still higher than is usual, the peak in mortality around one year of age had disappeared by the end of the 1970s. In contrast, the survival of children of all ages has improved during the past two decades. Thus, the explanations of the age pattern of mortality and of mortality decline in this population are closely inter-linked and the causes of mortality decline may differ between the two periods. Two diseases are emphasized in accounts of the age pattern of mortality in childhood in Senegal, of seasonal variations in mortality and of the reduction of mortality: measles and malaria. 1,3,9 The former disease figures prominently in the discussion by Delaunay et al. of the changing age pattern of mortality and the latter in their discussion of the disappearance and re-emergence of the seasonal peak in mortality. Apart from isolated campaigns in the 1960s, however, vaccination coverage in Niakhar began to rise only with the establishment of an EPI programme in the 1980s. 3 It therefore seems unlikely that measles and its sequellae played a significant role in the elimination of the 6‐30 month peak in the risk for dying in Niakhar, which occurred before this date. Moreover, others have argued that more than half the drop in mortality up to 1980 resulted from declining rainfall rather than from human intervention. 9 It was deaths from malaria, and perhaps also diarrhoea, that fell. During the 1980s though, when mortality may have fallen particularly rapidly, the crucial factor does seem to have been improvements in health care, perhaps buttressed by improve
Read full abstract