DOI: 10.1002/alz.077415 ISSN: 1552-5260

Associations between balance, strength and incident dementia in older men: the Health in Men Study

Michelle L Trevenen, Jack Paterson, Keith Hill, Osvaldo P Almeida, Bu Yeap, Jonathan Golledge, Graeme Hankey, Leon Flicker
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience
  • Geriatrics and Gerontology
  • Neurology (clinical)
  • Developmental Neuroscience
  • Health Policy
  • Epidemiology



Increased amount of physical activity has been associated with reduced incidence of dementia. It is less clear whether the discrete physical performance measures of balance and strength are associated with dementia incidence. We investigated whether baseline measures of balance and strength were associated with incident dementia.


A cohort of 1,314 men with no recorded history of dementia, aged 79‐97 underwent physical performance assessments between 2011 and 2013. These assessments were performed during wave 4 of the Health in Men Study, initially recruiting participants 1996‐9 in Perth, Australia. Balance was assessed with a modified Balance Outcome Measure for Elder Rehabilitation (mBOOMER) score: a combination of three tests, the Timed Up and Go, Step and Functional Reach Tests. Strength was tested using knee extension with a spring dynamometer, measured in kilograms. Incident dementia was identified up to December 2017 from the Western Australian Data Linkage System and the national Aged Care Assessment Program. Data were analysed using Cox proportional hazards models. Balance and strength measures were entered as restricted cubic splines to allow for non‐linear relationships. Models were adjusted for sociodemographic, lifestyle and health variables, including hours of physical activity.


A total of 187 (14.2%) incident cases of dementia were identified. Higher (better) mBOOMER scores were associated with decreased risk of incident dementia, with greater changes in risk occurring at higher scores (1 vs 0: HR 0.91, 95% CI 0.81‐1.01, 6 vs 5: HR 0.91, 95% CI 0.81‐1.01, 9 vs 8: HR 0.77, 95% CI 0.70‐0.84, 12 vs 11: HR 0.51, 95% CI 0.37‐0.69). Higher knee extension strength was associated with a decreased risk of incident dementia, with greater changes occurring at lower levels then plateauing at around 25 kilograms (5kg vs 4kg: HR 0.93, 95% CI 0.89‐0.97, 15kg vs 14kg: HR 0.93, 95% CI 0.89‐0.97, 25kg vs 24kg: HR 0.99, 95% CI 0.95‐1.02, 35kg to 34kg: HR 1.00, 95% CI 0.95‐1.06). Relationships remained after excluding those with a dementia diagnosis within two years of baseline physical performance assessment.


Better balance and strength are both associated with decreased risk of developing dementia, thus representing potentially modifiable risk factors.

More from our Archive