Handgrip strength and all-cause dementia incidence and mortality: findings from the UK Biobank prospective cohort study

Abstract

Background: This study aimed to investigate the associations of grip strength with incidence and mortality from dementia and whether these associations differ by sociodemographic and lifestyle factors.

Methods: A total of 466 788 participants of the UK Biobank (median age 56.5 years, 54.5% women). The outcome was all-cause dementia incidence and mortality and the exposure was grip strength. Grip strength was assessed using a Jamar J00105 hydraulic hand dynamometer.

Results: Excluding the first 2 years of follow-up (landmark analysis), mean follow-up was 9.1 years (inter-quartile range: 8.3; 9.7) for incidence and 9.3 (inter-quartile range: 8.7; 10.0) for mortality. During this time, 4087 participants developed dementia, and 1309 died from it. Lower grip strength was associated with a higher risk of dementia incidence and mortality independent of major confounding factors (P < 0.001). Individuals in the lowest quintile of grip strength had 72% [95% confidence interval (CI): 1.55; 1.92] higher incident dementia risk and 87% [95% CI: 1.55; 2.26] higher risk of dementia mortality compared with those in the highest quintile. Our PAF analyses indicate that 30.1% of dementia cases and 32.3% of dementia deaths are attributable to having low grip strength. The association between grip strength and dementia outcomes did not differ by lifestyle or sociodemographic factors.

Conclusions: Lower grip strength was associated with a higher risk of all-cause dementia incidence and mortality, independently of important confounding factors.

Keywords: Adults; Alzheimer; Mortality; Muscular strength; Prevention; Vascular dementia.

Figure 1

Association of handgrip strength with all‐cause dementia incidence (top graphs) and mortality (bottom graphs). Data are presented as hazard ratios (HRs) and 95% confidence interval (CI). Handgrip strength was expressed in absolute terms. All analyses were conducted using a 2 years landmark. Model 1 was adjusted for age, sex, ethnicity, and deprivation index. Model 2 was additionally adjusted for health‐related factors including body mass index categories, multimorbidity (prevalent diabetes, hypertension, cardiovascular disease, and cancer), and long‐standing illness. Model 3 was additionally adjusted by lifestyle behaviours including walking pace, sleep duration, watching TV, smoking, and dietary intake (alcohol, fruits and vegetables, red meat, processed meat, and oily fish intake).