Calcium and health outcomes

Today's New England Journal of Medicine reports the second major health outcomes study in a week from the massive Women's Health Initiative, exploding another cherished axiom of nutrition, this one about dietary calcium. Higher intakes of calcium, the study concluded, do not protect against hip fractures, although consuming more calcium does increase bone mineral density slightly (unexamined in the study is the imporant contribution calcium likely plays in blood pressure regulation). No doubt these results, like those last week finding that low-fat diets do not improve health outcomes, will rekindle debates among nutritionists and the medical establishment. Good. No matter how we have all felt about the wisdom of eating less fat or drinking more milk, a couple other "health outcomes" seem worth mentioning.

• First, these studies offer a great and humbling reminder that our confident embrace of logical explanations for population data demands that population health advisories should be based on randomized trials, not observational studies. They should reinforce our insistence on true evidence-based public health nutrition recommendations.
• Second, randomized health outcomes trials are possible. Many have argued that the importance of the health threat targeted for dietary therapy is so compellingly urgent that waiting for health outcomes trials -- expensive and long-term to be sure -- would be unethical. How ethical is it to tell the population to reconfigure its diet to obtain health benefits that are supported only by population studies whose conclusions are extrapolated into headlines claiming improved health outcomes? Without health outcomes trials examining the validity of the assumptions underlying these extrapolations, the dietary recommendation becomes a house of cards. Let's not gamble with cardsharps.

The salt connection, I hope, is obvious here. Logically, since salt is related to blood pressure and blood pressure is related to the incidence of cardiovascular events, some countries like the U.S. and the U.K. have embraced universal sodium reduction. The theory is clean and neat. But the argument is without evidentiary support. The only health outcomes data today are observational (and those data are only about a decade old, produced long after nutrition know-it-alls confidently concluded that lowering population salt intakes would improve public health). The health outcomes data we do have today shows no improvement in heart attack rate or increased morbidity/mortality for those on lower-sodium diets. In fact, 30% of the studies have identified an increased risk while none have identified a population benefit.

Before 2000, the U.S. Dietary Guideline suggested consuming salt or sodium in moderation. Reasonable advice, if somewhat ambiguous; it matched the ambiguity of the evidence. More recent Guidelines have been more strident even as the observational health outcomes studies accumulated and consistently showed no benefit. We should go back to the pre-2000 Guidelines until we can conduct a controlled trial of the health outcomes of reducing population sodium intake levels.