This story in the March 14 Washington Post captures pretty well the issues underlying the discussion about whether American consumers should believe the massive new federal study that found no health outcome benefit of reducing dietary fat.
Authors Lisa Schwartz, Steven Woloshin and Gilbert Welch capture public anxiety over what to believe when headlines trumpet stories that diet can prevent cancer and counter-punch with other stories saying just the opposite.
In the recent Women's Health Initiative (WHI), researcher reported "the latest apparent flip-flop...about low-fat diet and breast cancer. The story describes the new 50,000, eight-year WHI trial and the 2005 Women's Intervention Nutrition Study (WINS) with about 5% that number. Helpfully, the authors point out that the two studies "differed in two major ways. First, they studied different groups of women: one with prior breast cancer (WINS) and one without (WHI). Women with breat cancer are substantially more likely todevelop a new breast cancer than women without breast cancer....And this is exactly what happened....The second major difference ...was in how statisticians judged the findings. This difference was profound. Based on conventiaionl statistical practices, the effect of the low-fat diet in women with breast concer was deemed to be real--the result of more than chance. But in women without breast cancer, the effect of diet was judged to be statistical noise."
The authors go on to explain the importance of statistical significance, the difficulty of compliance -- especially by the subgroup that derives no benefit and, importantly, the crucial difference between observational studies (that show diet affecting cancer risk) and intervention trials which don't.
"In fact, most of the diet flip-flops come from observational studies that suggest a benefit (or harm) from diet that is not subsequently confirmed in randomized trials." See where I'm headed? With regard to salt, we have policy based on an intermediate variable (blood pressure) and only observational studies of health outcomes. While the observational studies offer no comfort to low-salt interventionistsl, we can only insist on a controlled trial of low sodium diets to see if they will reduce risk. The early betting is that they won't.
For more, see the author's website, the VA Oucomes Group .
