A 10-Year Follow-Up of Adiposity and Dementia in Swedish Adults Aged 70 Years and Older.

ABSTRACT

BACKGROUND: Adiposity measured in mid- or late-life and estimated using anthropometric measures such as body mass index (BMI) and waist-to-hip ratio (WHR), or metabolic markers such as blood leptin and adiponectin levels, is associated with late-onset dementia risk. However, during later life, this association may reverse and aging- and dementia-related processes may differentially affect adiposity measures. OBJECTIVE: We explored associations of concurrent BMI, WHR, and blood leptin and high molecular weight adiponectin levels with dementia occurrence. METHODS: 924 Swedish community-dwelling elderly without dementia, aged 70 years and older, systematically-sampled by birth day and birth year population-based in the Gothenburg city region of Sweden. The Gothenburg Birth Cohort Studies are designed for evaluating risk and protective factors for dementia. All dementias diagnosed after age 70 for 10 years were identified. Multivariable logistic regression models were used to predict dementia occurrence between 2000-2005, 2005-2010, and 2000-2010 after excluding prevalent baseline (year 2000) dementias. Baseline levels of BMI, WHR, leptin, and adiponectin were used. RESULTS: Within 5 years of baseline, low BMI (<20 kg/m2) was associated with higher odds of dementia compared to those in the healthy BMI category (≥ 20-24.9 kg/m2). Compared to the lowest quartile, leptin levels in the second quartile were associated with lower odds of dementia in women (p < 0.05). CONCLUSION: In late-life, anthropometric and metabolic adiposity measures appear to be differentially associated with dementia risk. While BMI and leptin levels are highly positively correlated, our results show that their association with dementia at age ≥70 years, is asynchronous. These data suggest that with aging, the complexity of the adiposity exposure may increase and suggests metabolic dysregulation. Additional studies are needed to better understand this complexity.