Food consumption and the actual statistics of cardiovascular diseases: an epidemiological comparison of 42 European countries

Background The aim of this ecological study was to identify the main nutritional factors related to the prevalence of cardiovascular diseases (CVDs) in Europe, based on a comparison of international statistics. Design The mean consumption of 62 food items from the FAOSTAT database (1993–2008) was compared with the actual statistics of five CVD indicators in 42 European countries. Several other exogenous factors (health expenditure, smoking, body mass index) and the historical stability of results were also examined. Results We found exceptionally strong relationships between some of the examined factors, the highest being a correlation between raised cholesterol in men and the combined consumption of animal fat and animal protein (r=0.92, p<0.001). The most significant dietary correlate of low CVD risk was high total fat and animal protein consumption. Additional statistical analyses further highlighted citrus fruits, high-fat dairy (cheese) and tree nuts. Among other non-dietary factors, health expenditure showed by far the highest correlation coefficients. The major correlate of high CVD risk was the proportion of energy from carbohydrates and alcohol, or from potato and cereal carbohydrates. Similar patterns were observed between food consumption and CVD statistics from the period 1980–2000, which shows that these relationships are stable over time. However, we found striking discrepancies in men's CVD statistics from 1980 and 1990, which can probably explain the origin of the ‘saturated fat hypothesis’ that influenced public health policies in the following decades. Conclusion Our results do not support the association between CVDs and saturated fat, which is still contained in official dietary guidelines. Instead, they agree with data accumulated from recent studies that link CVD risk with the high glycaemic index/load of carbohydrate-based diets. In the absence of any scientific evidence connecting saturated fat with CVDs, these findings show that current dietary recommendations regarding CVDs should be seriously reconsidered.


Supplementary
. Actual total CVD mortality (Men): Parsimonious models of the ridge regression, LASSO regression and elastic net regression, computed by the bootstrapping method and sorted according to the absolute beta coefficients of the ridge regression and elastic net regression. In the ridge regression, all food items that reach a non-zero value in the elastic net regression are listed.   Table 6a. Actual total CVD mortality (Women): Parsimonious models of the ridge regression, LASSO regression and elastic net regression, computed by the bootstrapping method and sorted according to the absolute beta coefficients of the ridge regression and elastic net regression. In the ridge regression, all food items that reach a non-zero value in the elastic net regression are listed.  Supplementary Table 8. Actual total CVD mortality in women: The first row displays results of the dependent t-test (standard deviations of the mean difference between pairs of r-values). The lower the number, the closer the temporal relationship between two trend lines. The second row displays results of the regression slope test (probability p-values expressing the similarity between linear slopes of two trend lines  Table 9. Raised blood pressure in men: The first row displays results of the dependent t-test (standard deviations of the mean difference between pairs of r-values). The lower the number, the closer the temporal relationship between two trend lines. The second row displays results of the regression slope test (probability p-values expressing the similarity between linear slopes of two trend lines). The higher the p-value, the higher the probability that two linear trend lines are running parallel to each other. ( Table 10. Raised blood pressure in women: The first row displays results of the dependent t-test (standard deviations of the mean difference between pairs of r-values). The lower the number, the closer the temporal relationship between two trend lines. The second row displays results of the regression slope test (probability p-values expressing the similarity between linear slopes of two trend lines). The higher the p-value, the higher the probability that two linear trend lines are running parallel to each other. ( Supplementary Fig. 1. Correlation between the prevalence of raised blood pressure and the prevalence of raised cholesterol levels in men (r= -0.55; p<0.001).

Supplementary
Supplementary Fig. 2. Correlation between the actual total CVD mortality and the prevalence of raised cholesterol levels in men (r= -0.69; p<0.001). Actual total CVD mortality in men (per 100,000 people)

Prevalence of raised cholesterol in men (%) (2008)
Supplementary Fig. 3. Correlation between the mean daily consumption of fruits and the prevalence of raised blood pressure in women (r= -0.69; p<0.001).
Supplementary Fig. 4. Correlation between the mean daily consumption of oranges & mandarins and the prevalence of raised blood pressure in women (r= -0.75; p<0.001).
Supplementary Fig. 5. Correlation between the mean consumption of fruits and the actual total CVD mortality in women (r= -0.71; p<0.001).
Supplementary Fig. 6. Correlation between the mean daily consumption of oranges & mandarins and the actual total CVD mortality in women (r= -0.76; p=0.001).  Fig. 7. Correlation between the mean daily consumption of meat fat and the prevalence of raised blood pressure in women (r= -0.71; p=0.001).
Supplementary Fig. 8. Correlation between the mean daily consumption of dairy fat and the prevalence of raised blood pressure in women (r= -0.62; p=0.001).
Supplementary Fig. 9. Correlation between the mean daily consumption of meat fat and the actual total CVD mortality in women (r= -0.70; p<0.001).
Supplementary Fig. 10. Correlation between the mean daily consumption of dairy fat and the actual total CVD mortality in women (r= -0.57; p=0.001). Actual total CVD mortality in women (per 100,000 people) Supplementary Fig. 11. Correlation between the mean daily consumption of cheese and the prevalence of raised blood pressure in women (r= -0.79; p=0.001).
Supplementary Fig. 12. Correlation between the mean daily consumption of cheese and the actual total CVD mortality in women (r= -0.72; p<0.001).
Supplementary Fig. 13. Correlation between the mean daily consumption of tree nuts and the prevalence of raised blood pressure in men(r= -0.65; p<0.001).
Supplementary Fig. 20. Correlation between the mean daily consumption of sunflower oil (2007) and the actual total CVD mortality in men (r=0.55; p<0.001).
Supplementary Fig. 21. Correlation between the mean daily consumption of cereals (1993) and the actual total CVD mortality in women (r=0.73; p<0.001).
Supplementary Fig. 22. Correlation between the mean daily consumption of onions (2008) and the actual total CVD mortality in women (r=0.58; p=0.001).  Fig. 27. Correlation between the mean prevalence of smoking in men and women (r= -0.33; p=0.032).
Supplementary Fig. 28. Relationship between the mean BMI in men and women.
Supplementary Fig. 29. Correlation between the mean prevalence of smoking and the mean BMI in men (r= -0.47; p=0.002).
Supplementary Fig. 30. Correlation between the mean prevalence of smoking and the mean BMI in women (r= -0.53; p<0.001). Supplementary Fig. 31. Correlation between the mean consumption of vegetables and the prevalence of raised blood pressure in men (r= -0.21; p=0.033).
Supplementary Fig. 32. Correlation between the mean vegetable carbohydrates/PC CARB energy ratio and the prevalence of raised blood pressure in men (r= -0.58; p<0.001).
Supplementary Fig. 33. Correlation between the mean consumption of vegetables and the actual total CVD mortality in men (r=0.01; p=0.96).
Supplementary Fig. 34. Correlation between the mean vegetable carbohydrates/PC CARB energy ratio and the actual total CVD mortality in men (r= -0.44; p=0.004).