Health Analysis:
Risk of dying and income inequalities

Death is the event of higher social cost and continues to constitute a fundamental element in the health situation analysis of populations. Mortality rates are not only an indicator of the magnitude of this event, but also an indicator of the absolute risk of dying, by cause, and for the age group, population and time that are considered. As a result, it is not surprising that the effect and impact of socioeconomic inequalities on the mortality of the population remain an area of attention in the study of socioeconomic health inequalities. Much empirical evidence has established the direct relationship between socioeconomic inequality and mortality, and enriched the discussion on the matter of equity in health. It is generally considered that the level of economic development is a determinant of the health situation in a defined space-population and as such, it is assumed that the risk of dying is a condition—and mortality a result - dependent on the quantity of available resources for social investment. From the methodological standpoint, this implies observing the health phenomenon in the population, i.e., mortality, at the various levels of a hierarchical socioeconomic scale.

In line with this methodological approach, this article presents a summary of mortality trends in the Americas between 1980 and 1994, in two opposite and differentiated age groups: children under 1 year of age and adult from 45 to 64 years, in relation to a hierarchical socioeconomic scale constructed with the values of the 1995 per capita Gross National Product (GNP) of the countries and territories of the Region, and previously published by PAHO/1 (Table 1). By means of the mortality ratio, i.e., an indicator of the relative risk of dying, the ecological approach incorporates an analytical dimension that contrasts different causes of death in the study of mortality trends in the Americas.

Income inequalities and risk of dying in infancy
Infant mortality presents a sustained downward trend in all the countries and territories of the Americas. In the last 50 years this trend has represented, in general, a reduction of 300% in the infant mortality rate in each of the five groups of countries differentiated by their level of economic income. The slope of the decline slows down from the threshold value of 40-45 deaths per 1,000 live births. The dispersion of this indicator—given by the interquartile range–in every group of economic income is being reduced progressively, which suggests an increase in the homogeneity between groups, i.e., a reduction of the inequalities in the risk of dying in infancy within each group (Figure 1). By contrast, the presence of inequalities in infant mortality among the groups of economic income persists in time; for example, the ratio of the median values of infant mortality rate for group I and group V has remained constant in the last three 5-year periods: 6.3 (72.5/11.5), 6.1 (60.5/10.0) and 6.1 (49.0/8.0). This implies that a child under one year of age living in a country of Group V continues to exhibit 6 times more risk of dying before 1 year of age than another child living in a country of Group I.

Viewed broadly, infant mortality can be considered as having two major groups of contributing causes: communicable diseases and perinatal conditions. The impact of the first is associated with broad changes in living conditions (safe water, sanitation, education, etc.), while the impact on the second component is more directly related to the access to high quality services and technology. A high rate of infant mortality can usually be reduced significantly first by means of interventions that address the communicable disease component. Subsequently, efforts to further reduce infant mortality through interventions that address the perinatal component will have less of an impact and entail higher costs. The trend of this behavior in the Region of the Americas can be observed by means of the ratio between mortality rates from perinatal conditions and from communicable causes. A ratio of perinatal to communicable causes of 1 implies that each component has equal weight in the infant mortality rate, thus indicating a higher potential for reduction. The greater the ratio, the greater the weight of the perinatal component compared to the communicable disease component and the more dependent further efforts to reduce that rate will be on investing in high-quality technology. In the Americas, a very clear pattern can be seen in this indicator among income groups: the ratio of perinatal to communicable causes diminishes—and, in fact, gets closer to 1 - as the group of countries exhibits smaller economic income, which indicates a greater relative weight of the communicable diseases in the higher infant mortality rates in the Region. Thus, for the 5-year period between 1990 and 1994, the median value of the perinatal/communicable ratio was 14.7 in Group I, while it was 5.4 in Group II, 1.8 in Group III, 1.7 in Group IV and 1.3 in Group V (Figure 2). On the other hand, a long-term trend is also confirmed in this indicator, i.e., the ratio of perinatal causes to communicable diseases causes has increased in recent decades in all the economic income groups, which confirms the significant reduction in the infant mortality rates, especially as a result of interventions intended to reduce the communicable disease component observed through the countries of the Region in the same period. This upward trend is more clearly discernible in the higher-income countries, i.e., the ones that had curtailed communicable diseases as a cause of infant mortality in 5-year periods. Thus, while in Group I the median value of the perinatal/communicable ratio went from 9.9 in 1980-85 to 14.7 in 1990-95, it went from 1.6 to 1.8 in Group III, and from 1.4 to 1.3 in Group V in the same 5-year periods.

Income inequalities and risk of dying in adulthood
Unlike the situation with infant mortality, the trend of mortality in the 45-64 age group in the Region of the Americas shows differences by sex. In general terms, the magnitude of mortality, and thus the risk of dying, in males of this age group (aprox. 1,200±400 per 100,000) is twice that of women (aprox. 700±200 per 100,000) in the period studied. Among males, there is no clear trend neither to the increase nor to a progressive reduction between 1980 and 1994 in any of the groups of countries: although the 5-year median values tend to the decline, the ranges of its distribution tend to the increase, which somewhat stabilizes the overall trend (Figure 3). In addition, the absence of a group-to-group pattern in mortality in males from 45 to 64 years is noticeable. The risk of dying is relatively similar in males of this age, regardless of their pertainance to a group of countries and, thus, of their per capita income. This contrasts with the mortality pattern observed among women from 45 to 64 years who, in addition to a confirmed moderate long-term downward trend within each country group—in particular Groups IV and V - display a clear gradient between groups: mortality is systematically higher in the groups of countries with smaller per capita income (Figure 4). It is possible to suggest that differences in the risk profiles by specific mortality causes, such as the relative impact of prevention measures, prevalence of healthy habits and access to health services among others, can help explain this inequality.

Recent decades have seen a kind of epidemiological polarization in the Americas, with chronic noncommunicable diseases becoming more prominent in mortality profiles as a result of demographic changes — especially the aging of the population - and of lifestyle changes in the various social groups. The trend of this behavior in the Region of the Americas can be observed by means of the ratio between mortality from noncommunicable causes (neoplasms and diseases of the circulatory system) and communicable causes in the adult population. A ratio of noncommunicable to communicable causes of 1 implies that the two components of mortality in adults have a equal weight. The higher the ratio, the greater the relative weight of the noncommunicable component and the more advanced the stage of epidemiological polarization, with the resultant implications in terms of investments, service access, and opportunity for health care. As did the pattern observed in the ratio of perinatal to communicable causes in children under 1 year of age, the ratio of noncommunicable to communicable causes for the age group 45-64 years old in both sexes in the Americas presents a gradient across groups, i.e., this indicator of relative risk decreases as the group of countries exhibits smaller economic income, which indicates the greater relative contribution of noncommunicable diseases in adult mortality in the most developed countries of the Region. Thus, in the 5-year period 1990-94, the median of the noncommunicable/communicable ratio in Group I was 31.5, while it was 14.8 in Group II, 8.5 in Group III, 9.4 in Group IV and 7.0 in Group V (Figure 5).

On the other hand, a long-term trend is also confirmed in this indicator, i.e., the ratio of noncommunicable to communicable causes has risen over the past few decades in all country groups, being more pronounced among lower-income countries which are precisely the ones that continue to have a larger communicable component in their adult mortality profile and are the ones that are reducing this component more rapidly. Thus, while the noncommunicable/communicable ratio in Group I went from 30.7 in 1980-85 to 31.5 in 1990-95, it went from 3.3 to 9.4 in Group IV, and from 5.9 to 7.0 in Group V in the same 5-year periods. The patterns of this indicator did not present significant variations by sex in this age group.

The ecological analysis presented here illustrates the gains in health realized by the countries of the Americas with regard to the reduction of the risk of dying, particularly in infancy. At the same time, it shows the considerable socioeconomic health inequalities that persist in the Region and emphasizes the importance of the mortality structure, i.e., the distribution of mortality by cause, age, and sex, in the analysis of health inequalities and, accordingly, in the identification of possible specific interventions that can positively affect this situation .

Return to Index
Epidemiological Bulletin, Vol. 20 No. 4, December 1999