4 ms·

I'm not a big fan of the overweight argument either. If I can find it, I read an article just the other day[1] that said there was no expectation of untimely d

• sliverstorm 14y ago
  In regards to drinks, it's probably indirect. E.g., the sort of person who has several drinks every night is probably the sort of person to have alcohol abuse issues at some point in their life. Alternatively, the data was averaged, and you are seeing the negative impact of binge drinking on weekends.
• dredmorbius 14y ago
  BMI is an absolute crap metric. It's vaguely useful for populations (and even there serious objections have been raised). It's utterly inappropriate for individuals, particularly when ready alternatives exist (waist size, for example, is a more robust predictor, despite its own vagaries). For individuals, BMI suffers in that it generally understates obesity in unfit, but "BMI-compliant" individuals, while overstating obesity in fit, BMI-noncompliant individuals. Which means that interventions are not indicated for those who would benefit and are indicated for those who would not. Generally, there's also the issue that there are large natural variations in healthy human weights, and no single target or designator will be appropriate for all. Fitness is multidimensional, and any attempt to reduce it to a single metric will likely fail. My lay recollection is that slightly higher-than-prescribed BMI is positively associated with greater life expectancy. Rationales vary, but among them, sick or ill individuals typically have lower-than-average bodyweights (skewing life expectancy down for low BMI scores), and athletes typically have elevated BMIs. Studies of Scandinavian Olympic athletes shows a pronounced lifetime mortality benefit to having competed in games (which tends to balance out the notoriously poor life expectancies of US professional athletes, particularly football, many associated with cumulative injuries sustained).