To the Editor: Wang and Hoy1 deserve much credit for highlighting yet again the poor state of health of Indigenous Australians. However, their conclusion that the Framingham equation underestimated risk and that better prediction equations are needed may miss the point.

The Framingham equations work well in other populations if the aim is to rank groups of individuals into higher or lower risk categories. Box 4 in the article by Wang and Hoy shows that they do this pretty well across increasing age groups. Framingham equations fall down when they are used to estimate absolute risk in populations whose coronary heart disease (CHD) rates are different from those in the Framingham study. Some years ago, we showed that adjusting the Framingham risk estimates in line with the overall incidence of CHD in the population modestly improved their performance.2

This is all very nice, but is better risk estimation the solution? We don’t estimate risk in other high-risk groups (eg, patients with CHD), because all are at high risk and all need risk factor reduction. A brief look at the risk factor profile in Box 2 of Wang and Hoy’s article reveals an alarming picture of uncontrolled CHD risk factors in a relatively young population (average age, 33–36 years). Cigarette smoking, dyslipidaemia, diabetes and overweight prevail.

Perhaps, rather than concentrating on quantifying the exact risk in such a high-risk population, we should look at the reasons for the high rates of risk factors. What motivates some Indigenous people to smoke more, be more overweight and have a higher incidence of dyslipidaemia and diabetes than other Australians?3,4 Do they feel disenfranchised when governments infer they are “dirty” by tying financial aid to face-washing?5 Do they have attractive employment opportunities? Do they have enough sense of control over their lives to reduce their need to indulge in cigarettes and other short-term pleasures? Are there adequate supplies of healthy foods that they like? These factors may differ, as some rural Abori-ginal communities have much lower rates of smoking, overweight and diabetes6 than others. Exploring these issues will aid preventive methods aimed at the whole community.

In the meantime, I would suggest that the Framingham equation does rank members of this community — into modest, high, and very high risk (the average 45–54-year-old has a 20% risk of a CHD event over 10 years1). This may help guide the medical treatment of risk factors and the pursuit of the medical model of prevention while social changes dictated by Aboriginal communities take effect.

In reply: We agree with Kinlay that it is important to prevent risk factors at the population level (a population strategy). However, there is also a need to properly identify high-risk individuals who require immediate medical intervention (a high-risk strategy) and to understand the full spectrum of factors that determine such risk.

The primary focus of our study was to assess whether the widely used Framingham risk functions were applicable to Aboriginal people in remote communities. Our data show that the Framingham functions significantly underestimated the risk of coronary heart disease (CHD). 1 The high CHD risk in Aboriginal people cannot be fully explained by traditional risk factors. Some major risk factors such as abnormal total cholesterol level and obesity in the study population are actually not as prevalent as those in the general Australian population. 2 Evaluation of traditional risk factors and identification of novel factors in this population are useful for the development of intervention strat-egies. Novel factors such as infection, inflammation, albuminuria and low birthweight have been suggested as predictors of CHD risk in this population. 3,4

Kinlay suggests that Framingham functions should be used to predict CHD risk in Aboriginal people. We disagree. Guidelines for the management of Aboriginal people need to recognise the serious underestimation of risk that the Framingham formulas provide.

We agree that some high-risk groups, such as patients with established CHD, do not need additional risk estimates. With our current knowledge, however, we can not say whether the whole Aboriginal community should be treated as a very high-risk population.