Sammendrag på engelsk (abstract)
The relationship between body mass index (BMI) and dementia is complex and controversial. This study investigates the association of weight change during midlife and later dementia-related mortality.
Two BMI measurements (average of 9.0 years apart) were available for 43,721 participants in the Norwegian Counties Study (NCS), with mean age 42 years at first BMI measurement and 51 at the final measurement. NCS was linked with the Cause of Death Registry until year 2015 (mean follow-up time 25.9 years). Cox regression with a conditional growth model was used.
Our study comprised 1,205 dementia-related deaths. Weight loss was associated with increased dementia-related mortality, irrespectively of baseline BMI and confounders; those with 10% or more loss had hazard ratio (HR) = 1.52 (95% confidence interval [CI]: 1.09, 2.12) compared to those being stable (0%-2.5% BMI gain), and those with 5%-10% loss had HR = 1.38 (95% CI: 1.08, 1.76). Gaining weigh was associated with reduced dementia-related mortality. Associations with BMI change did not vary by baseline BMI.
Weight loss during midlife was associated with increased dementia-related mortality risk more than 3 decades later, while weight gain was associated with reduced risk. These associations held both for low and high baseline BMI. Weight loss was an independent risk factor for dementia-related mortality and more strongly related with dementia-related mortality than stable BMI (stable high or low). Overweight and obesity were associated with an increased risk for nondementia-related mortality, which was far more common than dementia-related mortality
The Journals of Gerontology Series A Biological Sciences and Medical Sciences, 2016