Could it really be that fat doesn’t increase the risk of cancer — or even of heart disease? The answer seems to be Yes — according to a large-scale study reported in three articles in the current Journal of the American Medical Assocation. The Women’s Health Initiative Dietary Modification Trial studied 48,835 (!) post-menopausal women — that’s a huge number of people for a medical study — with 40% on a low-fat diet and 60% eating whatever they wanted — and found no significant effect of the low-fat diet on heart disease (!), breast cancer, or colorectal cancer. JAMA has the full text of the breast cancer article free on its web site; for the others you need a subscription for the full text, but the abstracts of the colorectal cancer and cardiovascular disease are free.
- For cardiovascular disease, there is basically no effect whatsoever from the low-fat diet. LDL cholesterol drops a tiny bit (3 mg/dL) — but there is no effect on heart attacks, strokes, or mortality. And of course, the usual goal of reducing LDL cholesterol is to reduce the change of heart attacks, strokes, and mortality, and this doesn’t accomplish that.
- For breast cancer, there are small reductions in the incidence of cancer (too small to be statistically significant by the usual criterion), but there basically no reduction in mortality. There was, however, a significant reduction in two specific types of tumors, which indicates that fat consumption may have some role in breast cancer even if it doesn’t really affect mortality. Then again, they were measuring so many things it’s likely that something would come out “significant” just by change — kind of like how if you get 100 people to flip 5 coins, there’s a 96% somebody is going to get 5 heads in a row.
- For colorectal cancer, there is basically no effect whatsoever from the low-fat diet. In the low-fat diet group, some kinds of colorectal cancers occurred at higher rates, and some at lower rates. In fact, the overall rate of death due to colorectal cancer was higher in the low-fat diet group — though not high enough to be statistically significant.
The only real caveat is that this study was only of postmenopausal women between the ages of 50 and 79, it might not be the same for men, or for younger women. But it was a study of 48,835 of them, so it’s quite likely the results are valid for at least that group. (Of course, the rates of breast cancer for men are vanishingly small compared to those for women!)
Note also that while this implies it is OK to eat fat — it does not imply that it is OK to be fat. What it basically means is that if you eat the same number of calories but switch some of the calories from fat to other things (protein, carbohydrates) you don’t get any reductions in cardiovascular disease, breast cancer, or colorectal cancer. However, it doesn’t mean you wouldn’t benefit from reducing fat and not replacing it with something else; that is, reducing your total caloric intake. To put it another way, the study shows that the composition of your calories doesn’t matter, but the total number of calories still might.
For an interesting article on the impact of the study, and for a more non-technical explanation of the results, see this article by Gina Kolata in the New York Times.
I normally don’t find science articles in the general media to be very good, or even true, but Gina Kolata has always seemed to me to be one of the few science reporters who actually knows some science. (Her book on the 1918 flu pandemic is excellent, and is a great read for someone who wants to understand what’s going on with the avian flu that might or might not break out.)
Having read both her article, and the JAMA articles, for once I think the newspaper got a science story basically right.
The lead authors of two of the three articles were originally trained in mathematics and statistics, not medicine. Remember when you were sitting in some math class wondering what that stuff was good for? Well, now you know!