A significant increased risk of mortality over the 12 years of follow-up was observed for underweight (BMI 18.5-35; RR = 1.36, P l.t. 0.05) and obesity class II+ (BMI to 35; RR = 1.36, P l.t. 0.05). Overweight (BMI 25 to 30) was associated with a significantly decreased risk of death (RR = 0.83, P l.t. 0.05). The RR was close to one for obesity class I (BMI 30-35; RR = 0.95, P l.t. 0.05). Our results are similar to those from other recent studies, confirming that underweight and obesity class II+ are clear risk factors for mortality, and showing that when compared to the acceptable BMI category, overweight appears to be protective against mortality."Overweight appears to be protective against mortality." Then why is it called "overweight?"
"Public health," like "public education," was imported to America from Prussia. The Prussian state was founded by a military order of armed monks, who imposed on the people they conquered an order of Christian discipline similar to their own. Their ideal subject was a man optimally suited for military service. Their ideal soldier was a dragoon, that is, a mounted infantryman: Dragons could be used either as highly mobile infantry or as light cavalry. This meant that the ideal soldier, and therefore the ideal Prussian subject, had to be light enough to ride all day without exhausting the horse. The acceptable weight for conscripting a Prussian dragoon is still with us as the range of "acceptable weight" used in public health studies. Adapted to America's greater variation of human height by substituting height-adjusted BMI for weight, the old Prussian standard of "acceptable weight" remains in world-wide "public health" use to this day.
An objective science of human health would set ideal weight to the weight at which the likelihoods of disease and death from disease are minimized. The corresponding measurement is the relative risk of death: the ideal weight is the weight at which the long term (say 12 year) risk of death is at its local minimum. In other words, the real, objective ideal weight has nothing to do with the desiderata of the Prussian General Staff. It ought to be set by measuring the facts of reality. And, from the facts measured to date, it is clear that the objectively optimal weight is nothing like the "acceptable weight" found in "public health" directives. It is almost certainly somewhere in the range that "public health" professionals call "overweight:" BMI between 25.1 and 29.9.
From the perspective of objective scientific methodology there is much wrong with BMI as the independent variable in health research. Optimal weight should be measured by plotting long-term (e.g. 12-year) mortality versus actual weight in the context of sex/gender, age and height. Unfortunately, I do not have access to the raw data that I would need to set an objective target range for my own weight. In the absence of such data, I use a target of BMI 27.5, the midpoint of the BMI range with the lowest observed mortality risk in nearly all quantitative studies to date.
The continuing use of the Prussian "acceptable weight" ranges, objectively known to be sub-optimal for human life and health, should be an epistemic scandal. It is a public folly with political uses. It permits "public health" authoritarians to claim that individual choice must be restricted to save us from the supposed epidemic of fat. Because if one accepts the Prussian pseudo-standard, 68% of Americans are overweight or obese. And this Prussian pseudo-standard is seldom challenged, because Americans "educated" in Prussian-standard public schools are so concept-deprived that they will believe anything, as long as it comes with a number and a percent sign somewhere - and will submit to the authority of the hoax.