Objectives: To investigate the reversal in adult mortality trend from declining to rising in some segments of population in India, and to use an indirect demographic method to examine if this increase could be due to AIDS mortality. Also, to estimate the total excess deaths.

Design: Cross-sectional data on age-specific death rate in 5-year age-intervals from 25 to 44 years for the years 1970 to 1998 for rural/urban and male/female segments for each of 16 major states of India obtained from the government reports, and their projections till the year 2001.

Methods: In view of reversal of trend in some areas, we tried to fit a parabola to the observed rates in each segment. A statistically significant fit in some segments revealed the year with least mortality rate when the reversal started. Another fit was obtained by projection of the previously declining trend. Excess deaths were estimated by applying the excess death rate to the population of the segment where reversal in trend was significant.

Results: Reversal in declining mortality trend was detected in 65 of a total 256 age-sex-area (urban/rural) segments that we examined. Fourteen of the 16 States revealed reversal in at least one segment. The year of reversal in most segments coincides fairly well with the anticipated year of start of substantial AIDS mortality. At the national level, a total of at least 214,390 deaths till 2001 were revealed as excess by this method. This number is quite low relative to the deaths otherwise attributed to AIDS in the country. Contrary to belief, increase in mortality due to AIDS was seen more commonly in rural areas than in urban areas, and more in females than in males.

Conclusions: The indirect demographic method of estimating AIDS deaths in India yields an apparently low number of deaths, and does not confirm the belief that AIDS in India is spreading from urban to rural, and from male to female populations.

Keywords: AIDS deaths, Mortality trend reversal, States of India, Age-specific death rates, Demographic method


Vital and Health Statistics