By Kilmer McCully, M.D.
The longest running study of cardiovascular disease in a population was initiated in Framingham, Massachusetts in 1948 and continues to this day. This important longitudinal study identified important major risks for disease, especially smoking, lack of exercise, age, male gender, and elevated cholesterol levels in younger men.
In spite of the great emphasis on cholesterol levels, the Framingham study made several critical observations that refute the "diet-heart" hypothesis. In the first place, dietary cholesterol has no relation to cholesterol levels in the blood, and dietary cholesterol has no relation to the risk of developing cardiovascular disease.
This observation was confirmed by multiple large studies from Chicago, Puerto Rico, Honolulu, Netherlands, Ireland, and the massive Lipid Research Clinics study of US citizens. The next astounding finding is that elevated cholesterol is not a risk factor for women of any age or for men over age 47.
Furthermore, both total mortality and cardiovascular mortality in Framingham participants increases in those with LOW cholesterol levels. This finding has been confirmed by multiple studies from Canada, Sweden, Russia, and New Zealand. These contradictory findings have been ignored, distorted, and incorrectly reported by supporters of the "diet-heart" hypothesis.
The massive Multiple Risk Factor Intervention Trial (MRFIT) screened 360,000 men to find those with the highest risk of developing cardiovascular disease. Approximately 12,000 overweight, hypertensive, smokers with elevated cholesterol levels were recruited for this 7 year trial, involving consuming a low fat diet, smoking cessation, exercise and anti-hypertensive drugs.
At the end of the trial, blood pressure was down, smoking decreased, and average cholesterol levels were down 7%. When the results of this $100M trial were analyzed, 115 in the treatment group had died of heart disease, compared with 124 in the control group, an insignificant difference. Looking at mortality from all causes, there were 265 deaths in the treatment group, compared with 260 in the control group. In looking at the failure of this massive and expensive $100M trial, the investigators found minor benefits of smoking cessation, no benefit of lowering blood pressure, and no effect of lowering cholesterol levels by 2% compared with the control group.
In the even more massive Lipid Research Clinics (LRC) trial, 4000 participants with very high cholesterol levels were selected from almost half a million men. After significant lowering of cholesterol levels for 7 years by the resin cholestyramine, 190 men had suffered nonfatal heart attacks in the treatment group, compared with 212 in the treated group. As for fatal heart attacks, the figures were 1.7% compared with 2.3%, a difference of 0.6%, or 12 individuals. The investigators expressed these differences as relative risk reductions of 19% and 30% by throwing out the denominators of their fractions.
In the later trials with statin drugs that lower cholesterol levels more effectively than the unpleasant resin cholestyramine, a similar statistical approach was taken to increasing the apparent effect on reducing cardiovascular mortality and adverse events.
In an analysis of 6 major statin trials (EXCEL, 4S, WOSCOPS, CARE, AFCAPS, LIPID), the reduction of cardiovascular mortality ranged from -19% to -41% when expressed as relative risk reduction, but from -0.12% to -3.5% when expressed as absolute risk reduction. This statistical manipulation to make the results more impressive illustrates Mark Twain's aphorism: There are lies, damn lies, and statistics. Thus a multi-billion dollar drug industry depends upon using misleading interpretations of statistics showing trivial differences between treated and control groups.
The gigantic MONICA study, sponsored by the World Health Organization, analyzed the relation between cardiovascular mortality and blood cholesterol in 27 countries, in much the same way as the Seven Countries Study. The results are similar, showing that countries like Japan and China have low mortality and low cholesterol levels, and countries like Finland have high mortality and high cholesterol levels.
Yet countries like France, Germany, Switzerland, and Luxembourg have a low mortality rate and yet a high blood cholesterol value. This so-called "French paradox" is not a paradox at all, when examination of the data reveals great disparities in mortality between different regions with the same cholesterol levels.
Similarly the residents of Corfu have a 5 fold greater mortality than residents of Crete, despite identical dietary practices and identical cholesterol levels. Residents of the North Karelia regions of Finland have mortality rate of 493/100,000 and those in Fribourg France have mortality rate of 102/100,000, yet the cholesterol levels are identical at 245 mg/dl in both regions.
The National Cholesterol Education Program is a quasi-governmental body sponsored by members of the National Institutes of Health, American Heart Association, and other supporters of the "diet-heart" hypothesis. This body recommends a low fat, high carbohydrate diet to prevent heart disease, in spite of the increasing incidence of diabetes, obesity, and hypertension that is linked to consumers of this diet.
They consistently advocate programs of extreme lowering of cholesterol levels by drug therapy, in spite of evidence of increased risk of mortality from heart failure, cancer, cirrhosis, and other diseases in older subjects with low cholesterol levels. They also recently recommended lowering the acceptable level of Low Density Lipoprotein (LDL) in the population by statin therapy, in spite of the fact that 8 of the 9 members of the advisory panel had a direct conflict of interest by accepting payments from the drug industry.
This body has popularized the concept that LDL is "bad cholesterol" and HDL is "good cholesterol" in spite of the marginal and sometimes contradictory data distinguishing these fractions from total blood cholesterol. This body also advocates "aggressive cholesterol lowering" in the population in spite of the fact that no cholesterol lowering trials have demonstrated reduced mortality or sudden death from such treatments in the otherwise normal population.
Kilmer S. McCully, M.D.
Chief, Pathology and Laboratory Medicine Service
West Roxbury Veterans Affairs Medical Center
Kilmer McCully M.D. - Cholesterol - Part 1