Research Press Release

Aging: Rethinking optimal BMI for octogenarians in China

Nature Aging

April 26, 2022

Being overweight or mildly obese is associated with reduced mortality in those aged over 80 in China, according to an observational study involving over 27,000 individuals published in Nature Aging. The findings suggest that public health recommendations for optimal weights may need to be revised upward in this age group. Although further studies involving people from different geographies are needed.

The effects of age-related changes in body mass index (BMI) on health are not well understood. There is some evidence for an ‘obesity paradox’ in adults over the age of 65, in which those with a BMI in the overweight-to-mildly obese range tend to experience better health outcomes and lower mortality than their leaner counterparts. However, optimal BMI recommendations have been established on the basis of large-scale epidemiological studies of the general population that tend to poorly represent older individuals, and in particular the oldest old — those aged 80 and over.

To better understand the relationship between BMI, health outcomes and mortality in older adults, Xiaoming Shi and colleagues analysed data collected from 27,026 individuals (mean age of 92.7 years) in their ninth decade or older (aged 80 and over), who were followed for up to 20 years as part of an observational study in China. The authors reveal that in this age group, the risk of mortality — in particular from non-cardiovascular disorders — decreased with increasing BMI values and that the optimal BMI was situated in the overweight-to-mildly obese range. A similar relationship was observed for waist circumference, which is a proxy for the accumulation of abdominal fat.

The authors conclude that these findings support the existence of an obesity paradox in the oldest in China, suggesting that public health recommendations on what is an optimal BMI may not be appropriate for this section of the population.


Return to research highlights

PrivacyMark System