"To obtain an unbiased relationship between BMI and mortality, it is essential to analyze individuals who never smoked and had no existing chronic diseases at the start of the study", said Frank Hu, professor of nutrition and epidemiology at Harvard Chan School and a co-leader of the collaboration.
"We still have more work to do to better understand how weight, weight gain, and weight loss influence mortality", said Barry Graubard, a senior investigator in the biostatistics branch of the U.S. National Cancer Institute. The new analysis involves participants who lived at least 5 years; of the 3,951,455 participants, including 69% women, 385,879 died.
This study overcomes numerous limitations of earlier efforts and "renders a clear and emphatic verdict - obesity increases the risk of premature death around the globe", Katz said.
Being overweight shaves about a year off one's life expectancy, a price which soars to about 10 years for the severely obese, a large-scale study reported Thursday.
Using data from nearly four million adults on four continents, the study in The Lancet medical journal found that overweight people lost about a year of life expectancy on average, and "moderately obese" people about three years.
The risk increased steadily and steeply as BMI increased, which is measured by calculating height and weight.
Looking at specific causes of death, the study found that, for each 5-unit increase in BMI above 25 kg/m2, the corresponding increases in risk were 49% for cardiovascular mortality, 38% for respiratory disease mortality, and 19% for cancer mortality.
Why obese men are at greater risk for premature death than women isn't clear.
A team of Cambridge and Oxford researchers has found that being too fat is far more harmful for men than women. For this study, a person with a BMI of 20 to 25 was normal weight; a BMI of 25 to 29.9 was considered overweight, and 30 or higher was considered obese.
The World Health Organization estimates that 1.3 billion adults worldwide are overweight, and 600 million more are obese. The reasons for the gender difference isn't clear, but other studies have pointed to differences in things like how men and women respond to insulin and variations in the amount of harmful fat in the liver that could be affected by obesity.
The results showed that participants with a normal BMI had the lowest premature mortality risk - of dying before the age of 70 - during the time they were followed.
Prof Sir Richard Peto, said: "Smoking causes about a a quarter of all premature deaths in Europe and North America, and smokers can halve their risk of premature death by stopping". The risk of mortality increased significantly throughout the overweight range: a BMI of 25- 27.5 kg/m2 was associated with a 7% higher risk of mortality; a BMI of 27.5- 30 kg/m2 was associated with a 20% higher risk; a BMI of 30.0- 35.0 kg/m2 was associated with a 45% higher risk; a BMI of 35.0- 40.0 kg/m2 was associated with a 94% higher risk; and a BMI of 40.0- 60.0 kg/m2 was associated with a almost three-fold risk.
Those underweight, with a BMI of 15 to 18.5, had an elevated risk of early death as well, with an HR of 1.51.
If Australia's overweight and obese population were within the normal weight range, about one in six premature deaths could potentially be avoided, according to the authors. "Actually, it's not. I think here we are reinforcing this message, and we are reinforcing the need to have public-health measures to try to implement a strategy to reduce body weight", said Dr Di Angelantonio.
The new study brings together information on the causes of any deaths in 3.9 million adults from 189 previous studies in Europe, North America and elsewhere.