P1‐098: Minutes of Walking Per Day at Age 50 is Associated with Dementia Risk, and This Association is Partially Mediated by Whole Brain Volume in Women: SCOBHI‐P

Amy Borenstein,James Mortimer,Ding,Qianhua Zhao,Shugang Chu,Qihao Guo,Charles DeCarli,Zhen Hong
DOI: https://doi.org/10.1016/j.jalz.2012.05.374
2012-01-01
Abstract:Physical exercise is a promising strategy for preventing dementia. However, the mechanisms through which it reduces dementia risk are largely unknown. We examined the association of exercise in midlife with whole brain volume (WBV) and dementia in later life, as well as the possible role of WBV in mediating the association between exercise and dementia risk. Thirty-four cases of dementia were frequency-matched to 32 controls from the community in Shanghai, China. All participants were clinically and neuropsychologically evaluated, had MRIs, and were APOE genotyped. Minutes of walking per day at age 50 (MWD50) were obtained from proxy interviews. The association between MWD50 and risk of dementia, adjusted for age, sex, education, APOE-e4, hippocampal volume (HV), volume of white matter hyperintensities (WMH) and proxy type, was evaluated by logistic regression. To test whether normalized WBV mediated this association we added WBV to the logistic models for MWD50. The association of WBV with MWD50, adjusted for age, sex and education, was assessed by linear regression. Cases' and controls' mean ages were 74.6 ( ± 5.7) and 73.4 ( ± 5.6) with 50% women among cases and 48.4% of controls. Case proxies reported 22 minutes of walking per day at age 50, compared with 33.9 in control proxies (P = 0.11). Adjusted for age, sex, education, proxy type, HV, WMH and APOE-e4, the OR for dementia per ten minutes per day of walking was 0.61 (95%CI = 0.43–0.88,p = 0.008). An association with MWD50 was seen in women (P = 0.02) but not men (P = 0.97). In addition, MWD50 was associated with WBV in women (P = 0.04) but not men (P = 0.84). When WBV was added to the adjusted model predicting dementia, MWD50 was no longer significant (P = 0.54) in the whole sample or in women (P = 0.24). Exercise in mid-life reduces the risk for dementia in women independently of APOE-e4, hippocampal volume and the volume of white matter hyperintensities. This association is partially mediated by late-life brain volume, which is related to minutes of walking per day at age 50 in women, but not men. Estrogen in earlier adult and midlife may interact with exercise to upregulate BDNF, increasing brain reserve in late life (Eur J Neurosci 2001;14(12):1992–2002).
What problem does this paper attempt to address?