Abstract
Purpose: We aimed to assess the associations of handgrip strength (HS) with cardiovascular and all-cause mortality and whether adding data on HS to cardiovascular disease (CVD) risk factors is associated with improvement in CVD mortality prediction.
Design: Handgrip strength was assessed in a population-based sample of 861 participants aged 61-74 years at baseline. Relative HS was obtained by dividing the absolute value by body weight.
Results: During a median (interquartile range) follow-up of 17.3 (12.6-18.4) years, 116 fatal coronary heart diseases (CHDs), 195 fatal CVDs, and 412 all-cause mortality events occurred. On adjustment for several risk factors, the hazard ratios (95% CIs) for fatal CHD, fatal CVD, and all-cause mortality were 0.59 (0.37-0.95), 0.59 (0.41-0.86), and 0.66 (0.51-0.84) respectively comparing extreme tertiles of relative HS. Adding relative HS to a CVD mortality risk prediction model containing established risk factors did not improve discrimination or reclassification using Harrel’s C-index (C-index change: 0.0034; p=0.65), integrated-discrimination-improvement (0.0059; p=0.20), and net-reclassification-improvement (-1.31%; p=0.74); however, there was a significant difference in in -2 log likelihood (p<0.001).
Conclusion: Relative HS is inversely associated with CHD, CVD and all-cause mortality events. Adding relative HS to conventional risk factors improves CVD risk assessment using sensitive measures of discrimination.
Design: Handgrip strength was assessed in a population-based sample of 861 participants aged 61-74 years at baseline. Relative HS was obtained by dividing the absolute value by body weight.
Results: During a median (interquartile range) follow-up of 17.3 (12.6-18.4) years, 116 fatal coronary heart diseases (CHDs), 195 fatal CVDs, and 412 all-cause mortality events occurred. On adjustment for several risk factors, the hazard ratios (95% CIs) for fatal CHD, fatal CVD, and all-cause mortality were 0.59 (0.37-0.95), 0.59 (0.41-0.86), and 0.66 (0.51-0.84) respectively comparing extreme tertiles of relative HS. Adding relative HS to a CVD mortality risk prediction model containing established risk factors did not improve discrimination or reclassification using Harrel’s C-index (C-index change: 0.0034; p=0.65), integrated-discrimination-improvement (0.0059; p=0.20), and net-reclassification-improvement (-1.31%; p=0.74); however, there was a significant difference in in -2 log likelihood (p<0.001).
Conclusion: Relative HS is inversely associated with CHD, CVD and all-cause mortality events. Adding relative HS to conventional risk factors improves CVD risk assessment using sensitive measures of discrimination.
Original language | English |
---|---|
Journal | Annals of Medicine |
Early online date | 30 Mar 2020 |
DOIs | |
Publication status | Published - 18 May 2020 |
Keywords
- handgrip strength
- cardiovascular disease
- mortality
- risk prediction