Cardiovascular risk prevention: One size does not fit all

Abstract

In his seminal article of 1965, Prof. Bradford Hill proposed nine criteria to establish causal relationship between a given risk factor and a disease. The eighth criterion was termed ‘Experiment’, where the suppression (or decrease) of the risk factor level would translate into the suppression or decrease of the disease of interest. Interestingly, Bradford Hill stated that ‘Here, the strongest support for the causation hypothesis may be revealed.’ Since then, most researchers have relied on those nine criteria to establish causation, and several randomized controlled trials, conducted mostly in Caucasian populations, have consistently shown that decreasing the levels of cardiovascular risk factors leads to a decrease in the incidence of cardiovascular disease. Cardiovascular risk factors (hypertension, high cholesterol, smoking and diabetes) are thought to be universal, that is, to exert their effect by the same magnitude for a given increase or decrease. This has been the basis for the recalibration of most cardiovascular risk equations, where the coefficients are held constant while the averages are calibrated to the population the risk equation should be applied to. Still, data from two Australian studies suggest that the prevalence and possibly the effect of each risk factor differ by ethnicity. In this issue of European Journal of Preventive Cardiology, Pais et al. reanalysed the data from the HOPE-3 trial, stratifying it according to Asian and non-Asian self-reported ethnicity. They observed several striking features. First, despite higher compliance to treatment levels, decreases in both blood pressure and low-density lipoprotein (LDL)-cholesterol levels were lower among Asians compared with non-Asians. Second, despite similar benefits regarding the reduction of both co-primary endpoints, the reduction was stronger for strokes inAsians and formyocardial infarction in non-Asians. Third, treatment by rosuvastatin increased the risk of incident diabetes in non-Asians only. Although the authors have no clear explanations for the differences between Asians and non-Asians, some results merit investigation. First, while in the candesartan/HCTZ arm the decrease in systolic blood pressure was similar for Asians and non-Asians, in the placebo arm the decrease was greater for Asians (– 5.7mmHg) than for non-Asians (–2.4mmHg). This stronger decrease in the placebo arm among Asians would reduce the absolute effect of the treatment among Asians. Second, and contrary to the results for blood pressure, in the rosuvastatin arm, changes in total and LDL-cholesterol were much smaller for Asians compared with non-Asians, while in the placebo arm the differences were stronger for Asians than for non-Asians. Overall, the results of Pais et al. suggest that, for a given dose of rosuvastatin, Asians and nonAsians respond differently, while the response to antihypertensive treatment might be similar. The differential effect of treatment on cardiovascular risk factor levels and outcomes according to ethnicity has been suggested previously. In the blood pressurelowering arm of the Anglo-Scandinavian Cardiac Outcomes Trial, the response to the combination atenololþdiuretic was similar among Whites, Blacks and South Asians, while the response to the combination perindoprilþ amlodipine was lower in Blacks and higher in South Asians when compared with Whites. In a recent paper including over 200,000 patients living in Canada, Ke at al. also showed that effectiveness of cardiovascular disease risk reduction therapy on mortality varied considerably by ethnicity. Still, both studies were conducted in high-income countries (Scandinavia, the UK and Canada) and not in Asian countries. Thus, the results of Pais et al. add further strength to the existence of ethnic differences in response to antihypertensive and hypolipidaemic drug therapy.