Part IGlobal Burden ofDisease and RiskFactors17Health status is both a determinant of population change,largely through population aging, and a consequence ofpopulation growth, with smaller family size associated withlower mortality, and of economic and social development.Studies of the interrelationship between demographictrends and health have typically focused on health as theindependent or determining variable. Indeed, a population’shealth status influences all components of populationchange. In addition to the obvious direct effect of individualhealth status on mortality and morbidity, it has a directimpact on fertility, largely through improved child survival,but also through the biological capability of a sick womanto bear children. Processes such as screening potentialmigrants for disease are also mechanisms whereby healthstatus exerts a direct impact on population change, and thuson population size and composition.In contrast, demographic variables influence healththrough two interrelated phenomena. First, a population’ssize, composition by age and sex, and geographical distribu-tion have a direct influence on overall health status. Age hasa particularly marked effect on the pattern and extent of ill-health in populations because of the strong relationshipbetween age and mortality and morbidity. Second, eachof the dynamic processes influencing population size andgrowth, structure, and distribution, namely, fertility,mortality, and migration, will also affect health status. Thus,any discussion of disease control priorities and of thehealth system for delivering interventions requires anunderstanding of the demographic context and how it ischanging.This chapter begins by providing an overview of globalpopulation trends in each major region of the world and thecurrent size and composition of the population. Given thisvolume’s focus on the descriptive epidemiology of diseases,injuries, and risk factors, we then examine trends in mortal-ity over the past decade in more detail as backgroundagainst which the current assessment of the disease burdenmight be more usefully interpreted. This includes both anassessment of trends in age-specific mortality and summarymeasures of the age schedule of mortality, such as lifeexpectancy and the probability of dying within certain ageranges, as well as a specific discussion of trends in the maincauses of child mortality. The focus on child mortality isentirely appropriate because (a) the fact that at the end ofChapter 2Demographic andEpidemiological Characteristicsof Major Regions, 1990–2001Alan D. Lopez, Stephen Begg, and Ed Bosthe 20th century, we remained woefully ignorant of its lev-els, let alone its causes, is highlighted; (b) the reduction ofchild mortality should remain a priority for global healthdevelopment efforts, and the moral imperative to do soremains as relevant today as it was 30 years ago, when effortsto improve child survival became increasingly organizedand focused; and (c) the resulting emphasis by the globalpublic health community on reducing child mortality hasyielded vastly more epidemiological information that can beused to assess trends in levels and causes. Nevertheless, weargue later in the chapter that large and unacceptable uncer-tainties about trends in cause-specific child mortality ratespersist, with important implications for program planningand evaluation.REGIONAL DEMOGRAPHIC CHARACTERISTICSThe key characteristics of regional demography of concernfor health services provision include the size, age structure,and sex structure of the population and its rate of growthand comparative measures of fertility and mortality.Sources of Population Data and MethodologyThe population and mortality estimates for various regionssummarized here are based on different data sources andmethods, and thus are not strictly comparable. This prima-rily concerns the impact of different estimates of deaths byage and sex on population size and structure. Because theeffect of mortality on population size and structure is gen-erally modest, such differences have little impact on thefindings reported in this chapter. The population estimatesare based on data the United Nations (UN) PopulationDivision compiled and analyzed for its biennial assessmentof global population trends and regional demographic pat-terns (United Nations 2003). The UN Population Divisionestimates population size and vital rates (births and deaths)from censuses, vital registration, and demographic andhealth surveys and evaluates the data for completeness,accuracy, and consistency. Where necessary, it adjusts thedata to achieve internal consistency and cross-country com-parability. The baseline from which the UN projections aremade is mid-2003. Because the 2002 revision was producedwithout complete data for 2001 for all countries, the base-line estimates are also projections, and the population fig-ures in this chapter are therefore a mixture of both observedand projected data.1The UN Population Division assesses a number of demo-graphic parameters to produce country projections. In addi-tion to total population, the baseline assessment includes abreakdown of population by sex and age (in five-year aggre-gates). Fertility is specified as age-specific fertility rates forfemales and mortality rates are based on survival probabili-ties from life tables. Age-specific patterns of migration arealso incorporated for countries in which migration flows areobserved or are thought to occur. When these inputs are notavailable from any of the sources listed earlier, the UN usesdemographic models, such as model life tables or indirectmortality estimation techniques, to generate the informa-tion. Additional modeling is applied to estimate mortalitypatterns in countries with significant HIV/AIDS prevalencelevels.The UN Population Division provides a limited amountof information about the data in its reports, including thedates of censuses, the adjustment factors applied to totalcensus populations, and the type and year of the latest sur-veys that contained mortality and fertility estimates. It doesnot provide information about the adjustments made toreported fertility rates, age and sex structures, or mortalityrates. Basic information on population size and composi-tion is available for most countries for 1990, and with theexception of Sub-Saharan Africa, for 2000 (or thereabouts)as well (table 2.1). Around both dates, censuses coveredmore than 90 percent of populations