Eight men in a sealed metabolic chamber showed the mechanism plainly: alcohol speeds your metabolism up and stops your fat-burning cold. Then thirty-one population studies failed to find the weight gain.
Everyone trained the same arm the same way for twelve weeks, with their other arm as the control. Biceps size changed by anywhere from minus two percent to plus fifty-nine. That spread is the answer.
The paper the ranges trace back to opens by saying no accepted ranges exist. It derived them from BMI cut-offs, needed your age and ethnicity to do it, and called the result provisional. That was 2000.
Cut 500 a day and the rule says 52 lb in a year. The measured response to a change in intake has a half-time of about a year, because your expenditure falls as you shrink. Same arithmetic, completely different trajectory.
572 studies and 486,538 cancer cases say yes, with a clear dose-response. The protective effect of moderate drinking disappeared once studies stopped counting sick quitters as non-drinkers.
Fifteen randomised trials, 56,675 people. Cutting saturated fat cut cardiovascular events by 17% and did nothing to how long anyone lived. And 21 cohort studies found no link at all. All three are true.
Thirty-one percent of American adults take one. After 84 randomised trials the task force could not say whether it helps or harms. There is one small real signal, and one ingredient they tell you to stop.
Across 16 randomised trials, protein source did not change strength or absolute lean mass. The one real edge for animal protein was 0.90 lb, in under-50s, and one of its intervals touches zero.
Across 121 randomised trials the strength effect was large. The effect on everyday physical ability, from the same review, was small. Both numbers are real and you should have been given both.
Across 4,871 tests, submaximal equations put people in the right fitness third 51% of the time. Guessing gets 33%. Here is what a real test is, and when you actually need one.