Abstract
Proportional Mortality Indicator (PMI) is defined as percentage of deaths at the ages 50 years-old and over to the total deaths in a population. This is essentially an indicator of age-distribution of deaths. Nevertheless, it reflects the levels of health of the population in developing countries sensitively. However, there have been no way to link such an indicator of age-distribution of deaths with other mortality indicators based on age-specific death rates such as life expectancy at birth. In this paper, the author tried to find out the missing link between these two types of mortality indicators. Considering that the values of l(x) function of a life table correspond to the proportions of deaths at the ages x year-old and above in the life table population, the author examined the relationships between the values of l(x) at various ages and life expectancy at birth. Using 90 pairs of reliable male/female empirical life tables representing a wide variety of mortality schedules in human populations, the values of l(x) at the ages 50 to 60 were shown to have very strong linear relationship between life expectancy at birth. The coefficients of correlation were as high as 0.99, even when male and female life tables were pooled together. The results were also shown to be in agreement with the life table theory which is known as Brass logit system. Based on these results and considering the effect of aging of population which follows a demographic transition, a theoretical interpretation of levels of PMI in actual populations was proposed.
