Earlier this week, Colby Cosh rounded up some recent re-evaluations of “settled food science”:
Their first target was the Sydney Diet Heart Study (1966-73), in which 458 middle-aged coronary patients were split into a control group and an experimental group. The latter group was fed loads of “healthy” safflower oil and safflower margarine in place of saturated fats. Even at the time it was noticed that the margarine-eaters died sooner, although their total cholesterol levels went down: the investigators sort of shrugged and wrote that heart patients “are not a good choice for testing the lipid hypothesis.” Their data, looked at now, shows that the increased mortality in the margarine group was attributable specifically to heart problems.
The team’s reanalysis of the Minnesota Coronary Experiment (1968-73) is more hair-raising. This study involved nearly 10,000 Minnesotans at old-age homes and mental hospitals. The investigators had near-complete control of the subjects’ diets, and were able to autopsy the ones who died. But much of their data, including the autopsy results, ended up misplaced or ignored. Some of it disappeared into a master’s thesis by a young statistician, now a retired older chap, who helped with the 2016 paper and is named at its head as one of the authors.
In the Minnesota study, replacement of saturated fats with corn oil led, again, to reductions in total cholesterol. This finding was touted at major conferences, and it became one of the key moments in the creation of the classic diet-heart myth. This time nobody but the guy who wrote the thesis even noticed that the patients in the corn oil group were, overall, dying a little faster. The 2016 re-analysis uncovered a dose-response relationship: the more the patients’ total cholesterol decreased, the faster they died.
The Sydney and Minnesota studies themselves may have caused a few premature deaths, which is a possibility we accept as the price of science. But the limitations and omissions of the researchers, and the premature commitment of doctors to a total-cholesterol model, helped create a suspicion of saturated fats. This flooded into frontline medical advice and the wider culture, and it put margarine on millions of tables, pushed consumers toward deadly trans fats, and put millions of people with innately high cholesterol levels through useless diet austerity. The scale of the error is numbing, unfathomable.