MODERATE OBESITY IN LATER LIFE: IS IT REALLY PROTECTIVE FOR HEALTHY AGING?

摘要

There is very contradictory evidence on whether being overweight or moderately obese (BMI 30 to 35) is protective or risky in later life for both survival and ageing well. With the increasing prevalence of obesity world-wide, there is a need to clarify this so-called obesity paradox of later life, including addressing suggested confounders and measurement errors. We aimed to extend the current evidence using very large-scale electronic medical record data and volunteer cohort data, plus improved measures of adiposity, co-morbidity and confounding. We also aimed to quantify trajectories of BMI change in the years before death, to clarify the effect of reverse causation.We used Clinical Practice Research Datalink (England) linked primary care, hospital and death certificate electronic medical record data on nearly 1 million patients aged ≥60 years in England from 1 January 2000 onwards. Cox survival models for each age-subgroup were adjusted for age, gender, alcohol use, smoking, calendar year, and socioeconomic status. We also used data from the UK Biobank, a volunteer study of 500,000 people aged 40 to 69, followed for up to 8 years.Our program of work has shown that 1) In the 14 years before death there are progressive declines in BMI over the entire period, which accelerate in the last 2 years of life. 2) After accounting for the decline in BMI before death, smoking and major disease, most of the BMI risk paradoxes with mortality and cardiovascular disease (CVD) incidence in the ≥60s disappear or reverse. 3. Misclassification of central adiposity in later life by the conventional BMI measures explains more of the risk paradox. Measures combining BMI and Waist Hip ratio are associated with major excess risks for mortality and CVD incidence from both being overweight or moderately obese in otherwise healthy 60 to 69 year olds. 4. Accounting for the above issues, levels of exercise and muscle strength remain important independent predictors but do not interact with risks due to adiposity. In conclusion, when adiposity is measured appropriately, being overweight or obese is predictive of substantially worse outcomes in older people who do not have confounding conditions. Calls for changing policies on obesity prevention because of the claimed risk paradoxes are misplaced.