Because America is a nation of immigrants, and so many of us have lost our connections to the original cuisines of our homelands, we are probably more susceptible to the guidance of nutrition experts.
His observation that a diet low in saturated fat was consistent with minimal heart disease was possibly accurate in 1960 but was no longer true in 1990. And this original mistake seems to have been compounded a thousand times over during the next decades by scientists who inherited Keys's dietary biases. No doubt a Cretan or Calabrian peasant might find it ironic that New York socialites and Hollywood movie stars -indeed, nearly all the wealthy peoples on the planet- are now trying to replicate the diet of an impoverished post-war population desperate to improve its lot.
Keys had published a paper only a few years before the pyramid came out, stating that the average Cretan consumed 8 ounces (1 cup) of fresh milk every day, mainly from goats but also from cows, which was more than the US cohort was drinking. Why did this information not make it into the pyramid? I asked. Willett even cited this paper by Keys but then explained that he is nevertheless excluding milk because it is so "high in saturated fatty acids, which are believed to cause CHD." A fear of saturated fat appeared to trump all other considerations, even the actual data on milk consumption itself.
The two sources of dietary data gave different results that could not be reconciled. So Keys assumed that the Cretan men must have been giving inaccurate replies to the questionnaires and he did a rather astonishing thing. Although you have to read carefully between the lines of his papers to figure it out, Keys ended up simply getting rid of the survey data he had collected from the 655 men on Corfu and Crete. That left only one source of dietary data for his calculations: the food collected from the smaller group of men. These meals were gathered on three separate occasions on Crete and once on Corfu. Keys went to Corfu twice, actually, but had to throw out one set of data because some of the fats had been "destroyed in processing." Other fats were absorbed into the clay containers used to carry the food samples. In the end, it turned out that only thirty to thirty-three men were sampled on Crete and thirty-four on Corfu.
The choice to favor LDL-cholesterol over HDL-cholesterol was also probably fueled by the megabillion-dollar pharmaceutical industry, which heavily favored LDL-cholesterol as a target for therapy. Drug companies had made quite a few attempts to find a drug that raised HDL-cholesterol, but those efforts had all failed. Lowering LDL-cholesterol, however, was something they could do -very well.