Revisiting the Coronary Primary Prevention Trial (CPPT) – 1973 to 1984
Launched in 1973 by the National Institutes of Health, the Coronary Primary Prevention Trial (CPPT) set out to prove that lowering blood cholesterol with a drug and a low cholesterol, low saturated fat diet would reduce the risk of coronary heart disease and extend the lives of the study participants.
(As part of the study, twelve new Lipid Research Clinics were set up by the Heart Institute at large universities throughout the country including Baylor, Stanford, Johns Hopkins and the University of Washington in Seattle.)
The researchers were looking for middle-aged men with total cholesterol levels exceeding those of 95 percent of Americans. (Only men with the highest 0.8 percent total cholesterol qualified.) The CPPT researchers screened 480,000 applicants in order to select 3,806 high risk men between the ages of 35 and 59.
This meant that many of the participants had familial hypercholesterolemia, a rare genetic defect in cholesterol metabolism present in about 1 percent of the population. The trials chance of success was therefore greatly increased by focusing on this particular group of presumably vulnerable men.
In their preliminary report, CPPT researchers announced that they would study two separate outcomes: (1) Nonfatal heart attacks and (2) fatal heart attacks or deaths from coronary heart disease. The CPPT directors emphasized that they would be satisfied with nothing less than the strongest statistical proof of their findings; they had to be “99 percent certain that the results were not due to chance.”
The researchers also announced their goal of reducing blood cholesterol in the treatment group by 25 percent and reducing the risk of heart disease in the treatment group by at least 50 percent.
Approximately half of the 3,806 men were provided low cholesterol, low saturated fat dietary advice and were treated with cholestyramine, a cholesterol-lowering bile acid resin (Questran). (Cholestyramine lowers cholesterol by interfering with digestion. Statins such as Lipitor, Mevacor and Zocor were not available yet.)
The control group was provided the same dietary advice and an unpleasant tasting placebo – an indigestible mixture of sand, sugar and food coloring. Both trial groups suffered with moderate to severe gastrointestinal distress. There were eight gastrointestinal cancer deaths in the treatment group (out of 21 cases) and one in the placebo group (out of 11 cases).
(There were more deaths from cancer, intestinal disease, stroke, violence and suicide in the group taking the cholesterol-lowering drug, and overall mortality was essentially the same for both groups.)
In 1984, the disappointing results were tabulated. Cholesterol levels in the treatment group had decreased by no more than 7 percent. Cholestyramine and the low cholesterol, low saturated fat diet had failed to lower cholesterol enough to prove that lowering cholesterol would reduce the risk of heart disease and extend the lives in the treatment group.
The difference in nonfatal heart attacks was not statistically significant. In the treatment group, 130 participants (6.8 percent) had a heart attack versus 158 in the placebo group (8.3 percent). After 7 years, the fraction of the treatment group that had benefited was less than 2 percent.
Nonfatal heart attacks Fatal heart attacks/coronary deaths
1,900 Control Group: 158 or 8.3 percent 38 or 2.0 percent
1,906 Treatment Group: 130 or 6.8 percent 30 or 1.6 percent
The difference in fatal heart attacks was not significant either. In the treatment group, 30 participants (1.6 percent) suffered a fatal heart attack compared to 38 in the placebo group (2.0 percent) Again – after 7 years of taking an unpleasant drug (and following a low fat diet), the fraction of the treatment group that benefited was less than 1 percent.
However, by applying relative risk statistics (a percentage of a percentage), the CPPT researchers improved their results. They took the number of people who presumably didn’t have a heart attack because of taking the drug (28) and looked at it as a percentage of the people who did have heart attacks (158) but didn’t take the drug:
The less than 2 percent absolute difference in nonfatal heart attacks rose to a reported 19 percent reduction in risk of a heart attack!
In similar statistical fashion, the researchers announced a 24 percent reduction in the risk of dying from a heart attack. The 8 men or 1.6 percent out of 1,900 who presumably did not have a fatal heart attack because they took the drug became the same 24 percent who reduced their risk of mortality compared to those in the control group who did die (38) but did not take the drug.
Additional study design changes
To prop up their victory, the CPPT researchers decided to exclude “uncertain” nonfatal heart attacks from the treatment group while including “uncertain” fatal heart attacks in the placebo group. Also, using the original 99 percent standard, the small favorable trend in either group could only be explained by chance (as defined by the researchers themselves at the start of the trial.)
By applying the less stringent 95 percent standard and by combining the two groups into one (nonfatal and fatal heart attacks), the CPPT researchers improved their results – declared victory – while the press responded with unbridled enthusiasm.
In 1984, the press and medical journals portrayed CPPT as the long sought proof that animal fats were the cause of heart disease. It was widely reported that for the first time:
“It had been proven that lowering cholesterol would reduce the mortality from heart disease and lower the risk of having a heart attack.”
Much of what we hear today about diet and heart disease can be traced back to this notorious failed study. When other scientists voiced their objections to the trial’s design changes, the CPPT directors simply denied that they had ever embraced the original more stringent standards.
In January 1984, the Journal of the American Medical Association (JAMA) dutifully reported:
“The trial’s implications…could and should be extended to other age groups and women, and to others with more modest elevations of cholesterol levels. The benefits that could be expected from cholestyramine treatment are considerable.”
George Mann, M.D., professor in medicine and biochemistry at Vanderbilt University, severely criticized the CPPT directors and the trial’s unsupportable results:
“The managers at the National Institutes of Health have used Madison Avenue hype to sell this failed trial in the way the media people sell an underarm deodorant…”
Giving cholestyramine for over seven years to 1,906 middle age men – many with a genetic predisposition to atherosclerosis – had only saved the lives of eight but the Heart Institute was now recommending that cholesterol-lowering drug treatment be extended to patient groups that had not been part of the trial.
Even without the solid evidence they sought, the medical elite in the American Heart Association and the National Institute of Health decided to push ahead with cholesterol-lowering drugs and the still unproven low cholesterol, low saturated fat diet:
“Now we have proved that it is worthwhile to lower blood cholesterol; no more trials are necessary. Now is the time for treatment.”
The long War on Cholesterol had begun