Commentary: Ecologic studies in identifying dietary risk factors for coronary heart disease and cancer.

The hypothesis study by Jeffrey Segall, 1 suggesting that milk consumption was an important risk factor for ischaemic heart disease [IHD, or coronary heart disease (CHD)], was based on considerations of lactose intolerance and IHD rates for 23 populations. However, a recent meta-analysis of cohort studies found that milk consumption was actually correlated with a small but significant reduced risk, 2 suggesting that the work by Segall was somehow in error. In the well-known Seven Countries Study, 3 Ancel Keys established that animal fat was an important risk factor for IHD on the basis of a study of dietary factors for men between the ages of 55 and 59 years and incidence of CHD in seven countries. However, when the Seven Countries Study was being conducted , John Yudkin, a British diabetes expert, suggested that dietary sucrose was a more important risk factor than was animal fat. 4 Keys and Yudkin had a spirited debate, which Keys eventually won. 5 Examining the data in the work of Segall 1 and other studies is worthwhile to evaluate which dietary factors are most strongly correlated with risk of IHD. Therefore, I obtained dietary supply data for 1961–63 from the Food and Agriculture Organization 6 for use with data for IHD circa 1972 in the work of Segall. 1 The data were used in linear regression analyses with SPSS version 13 (Chicago, IL, USA). Table 1 gives the results of that analysis. Although milk and its components have a high correlation coefficient, that of animal fat is higher. Had Segall considered animal fat in addition to milk, he might not have concluded that milk was a risk factor for IHD. After I found that total energy and fat supply were the primary risk factors for Alzheimer's disease, with fish and cereals reducing the risk, 7 I evaluated dietary risk factors for CHD. I used mortality rate data for acute myocardial infarction and other IHDs for males and females for 33 countries in 1986 and dietary supply data for earlier periods. In the first study, 8 animal fat supply for 1973 had the highest correlation for males of all age ranges, whereas added sugar for 1983 had the highest correlation for females of all ages; both factors together gave higher regression coefficients. Had women been included in the Seven Countries Study, Yudkin might have been able to make a better case for the role …