Editor—Many non-communicable diseases are set to increase dramatically during the 21st century. In particular, the prevalence of diabetes mellitus may double from 124 million people worldwide in 1997 to 220 million by 2010.1 Regions with the greatest potential increases are Asia and Africa—precisely the areas with the greatest potential increases in the prevalence of HIV infection and AIDS. The impact of HIV infection and AIDS on the diabetes epidemic is difficult to assess, but we have tried to estimate it for South Africa using population data from the United Nations2 and estimates of the prevalence of diabetes (table).1,3
The population will continue to increase but at a slower rate because of the HIV/AIDS epidemic. South Africa’s annual growth rate, which was 1.9% in 1995, is expected to decrease to 0.3% in 2010. Without HIV/AIDS it would be 1.5%. The highest age specific prevalence of HIV infection is forecast for those aged 20-34 years.4 Amos et al predict that the age adjusted prevalence of type 2 diabetes will increase from 1.7% to 3.7%, resulting in 1 624 000 cases in 2010.1 A less conservative prevalence of 4% rising to 8% gives an estimated 3 482 000 cases, occurring mainly in those aged 50-59 years. When the effect of HIV/AIDS on population growth is calculated this number is predicted to decrease to about 3 380 000 cases, a 3% reduction representing over 100 000 fewer cases of type 2 diabetes.
Although the interaction of two diseases that might affect various subgroups of the South African population differently is difficult to model, the burden of both HIV/AIDS and type 2 diabetes is likely to fall on the lower socioeconomic classes. The peak age specific prevalence occurs earlier in HIV/AIDS than type 2 diabetes, but correspondingly fewer infected people will survive to middle age. Thus our calculation of 3% fewer cases, reflecting the expected decrease in population growth due to HIV/AIDS, represents a small but substantial reduction, irrespective of the prevalence of diabetes selected. This figure may actually be higher if the indirect effects of AIDS are considered, such as the decrease in obesity because of chronic infection and wasting. This will ameliorate insulin resistance and better preserve residual β cell function.5 The toll of HIV/AIDS on mortality, population loss, and diabetes prevalence may even be greater elsewhere in Africa.
Our model highlights the need to adjust for the impact of HIV/AIDS when projecting the prevalence of chronic diseases and national health budgets into the next century.
Table.
Demographic impact of HIV infection and AIDS on projected prevalence of type 2 diabetes in South Africa
| Reference | 1995 | 2010
|
|
|---|---|---|---|
| Without HIV/AIDS | With HIV/AIDS | ||
| United Nations2 | |||
| Population (×1000) | 41 464 | 43 529 | 42 256 |
| Growth rate (%) | 1.9 | 1.5 | 0.3 |
| Amos et al1 | |||
| Type 2 diabetes: | |||
| Prevalence (%) | 1.7 | 3.7 | 3.7 |
| No of cases (×1000) | 717.8 | 1624.0 | 1576.1 |
| Levitt and Mollentze3 | |||
| Type 2 diabetes: | |||
| Prevalence (%) | 4.0 | 8.0 | 8.0 |
| No of cases (×1000) | 1658.5 | 3482.3 | 3380.5 |
References
- 1.Amos AF, McCarty DJ, Zimmet P. The rising global burden of diabetes and its complications: estimates and projections to the year 2010. Diabet Med. 1997;14(suppl 5):S7–84. [PubMed] [Google Scholar]
- 2.United Nations Population Division. World population prospects: the 1998 revision. New York: UNPD; 1999. [Google Scholar]
- 3.Levitt N, Mollentze WF. Diabetes and impaired glucose tolerance: a review of South African studies. In: Fourie J, Steyn K, editors. Chronic diseases of lifestyle in South Africa. Cape Town: Medical Research Council of South Africa; 1995. pp. 99–126. . (MRC technical report.) [Google Scholar]
- 4.Whiteside AW, ed. South Africa’s antenatal results give mixed messages. AIDS analysis Africa. Southern Africa ed. Vol 9. Cape Town: Whiteside and van Niftrik Publications:1-2.
- 5.Joffe BI, Seftel HC. Diabetes mellitus in the black communities of Southern Africa. J Intern Med. 1994;235:137–142. doi: 10.1111/j.1365-2796.1994.tb01047.x. [DOI] [PubMed] [Google Scholar]