https://www.nap.edu/read/13089...
"In 1999, Allison and colleagues calculated that in 1991, between 280,000 and 325,000 deaths of U.S. adults were due to overweight and obesity (Allison et al., 1999a, 1999b). Six years later, Mokdad and colleagues announced that overweight and obesity (together with physical inactivity) had been responsible for 365,000 excess deaths among U.S. adults in 2000, making it the second-leading preventable cause of death in the United States, behind only smoking (Mokdad et al., 2005)."
"Those numbers, which were publicized widely, now appear likely to have been major overestimates. For example, one study published shortly after the Mokdad et al. (2005) article that used more recent data and took into account how mortality risk varies by age yielded much smaller numbers (Flegal et al., 2005). According to that study, obesity, defined as a BMI of 30 or above, caused approximately 112,000 excess deaths among U.S. adults in 2000, while being overweight had a protective effect and led to 86,000 fewer deaths than would have been expected if all of those people had had a BMI in the normal range. The net result was that overweight and obesity together resulted in an excess of 26,000 deaths in 2000, the authors concluded, which was less than a tenth of the earlier estimate."
Body Mass Index (BMI) and life expectancy follow a curve, and the curve peak - the longest lifespans on average - are for BMI between 27 and 28, which is qualified as overweight. An individual athlete can have BMI 28 and ultra low body fat, but these were population studies and
across populations BMI 27 or 28 carries lots of fat on average.
But quoting research and statistics like that cuts into profits for the diet industry and takes away all the fun of mocking the fatties, so it doesn't get much press attention.