Researchers from Oxford Brookes University examined data on patients whose BMI and waist to height ratio were measured in the 1980s.
Twenty years later, death rates among the group were much more closely linked to participants' earlier waist-to-height ratio than their BMI, suggesting it is a more useful tool for identifying health risks at an early stage.
By comparing the life expectancies of various groups of people at different waist-to-height ratios, they were able to calculate how many years of life were lost as people's waistlines increased.
For example, a man aged 30 with a waist-to-height ratio of 0.8, representing the largest one in 500 men, stood to lose 16.7 years of life due to his size.
A 50-year-old woman with the same ratio, accounting for about one in 150 women of the same age, would lose 8.2 years of life on average.