LOW-CARB LOW-DOWN

The claim that a dietary pattern that improves health also

shortens life is extraordinary, and should be supported by a high standard of evidence ( Nutrition, December 15).

Yet the average person in the study referred to reported eating an implausibly low number of calories, indicating considerable inaccuracy. Unfortunately, this sort of thing is par for the course in associational studies.

A more serious defect is the fact that alcohol consumption was not reported or adjusted for anywhere in the study paper. We can be sure that excessive drinking shortens life expectancy.

Researchers have also found that alcohol, as a source of energy, displaces carbohydrates in the diet more than it displaces other nutrients. Yet even in studies where alcohol has been reported, people have been found to under-report its consumption by a factor of two to four.

Accuracy could be improved if the association of diet with accidental death was always calculated, as no food properly chewed causes accidental death except for alcohol.

In a 2009 study, eating red meat was associated with a 26% higher risk of accidental death, much the same as the risk of the other adverse health outcomes that were associated with eating red meat.

This raises the question whether much of the observational research used to scare us off food would, properly analysed, be better employed in scaring us sober. George Henderson (Huia, Auckland)