Anderson, Barbara A., and Heston E. Phillips. 2008. "The Changing Pattern of Adult Mortality in South Africa, 1997-2005: HIV and Other Sources." PSC Research Report No. 08-649. July 2008.
Recent mortality changes in South Africa are interesting because it has the most plentiful and the highest quality mortality and cause of death data of any country with a major HIV epidemic. Also, South Africa exhibits a combination of first world and third world mortality patterns as predicted by Gwatkin (1980). Gwatkin saw developing countries as trapped in a situation in which death rates from non-communicable diseases associated with development, such as diabetes, would increase, while mortality from infectious diseases, such as tuberculosis, would still be substantial. Although HIV was unknown in 1980, Gwatkin also predicted that such countries would be vulnerable to new epidemics that might appear. We investigate adult mortality 1997-2005, paying special attention to sex differentials in mortality and the changing role of the three Global Burden of Disease categories.
• HIV mortality has increased in South Africa, especially since about 2000. • Adult mortality has increased enormously since 1997, especially among women age 20-39. • Between 1997 and 2005, overall adult male mortality increased by 1.8 times and overall adult female mortality by 2.6 times. • In 2005, female mortality exceeded male mortality at age 20-34, reversing a typical pattern. • Male age standardized death rates 15-64 remained higher than female rates, although the gap narrowed considerably over time. • Female mortality increased by more than 2.5 times for a wide variety of causes of death, both communicable and non-communicable diseases; Male mortality increased by more than 2.5 times for a smaller variety of causes, mainly communicable diseases. • Below age 20 and above age 55 mortality increases were modest. • Many deaths in which HIV is the underlying or a contributing cause of death are not identified as due to HIV. • The shape of the age-specific mortality schedule has often been used to identify hidden HIV mortality, which is a useful approach but needs to be applied with caution. • Rates for other important causes of death have been virtually unchanged (stroke, cancer), have declined (malaria) or have increased (diabetes), even though they are not directly related to HIV. • Higher female than male death rates at young adult ages have sometimes been interpreted as evidence of HIV, but female rates were higher than male rates for many causes of death in 1997, especially for those age 15-29. In 1997 HIV mortality was likely still low.
Country of focus: South Africa.