Aims
To compare the diagnostic value of relative sit-to-stand muscle power with grip strength or gait speed for identifying a history of recurrent falls and fractures in older adults.
Methods
Data from an outpatient clinic included anthropometry (height/weight), bone density, 5 times sit-to-stand time (s), handgrip strength (kg) and gait speed (m/s). Relative sit-to-stand muscle power (W.kg-1; normalised to body mass) was calculated using validated equations. Outcomes of falls (past 1 year) and fractures (past 5 years) were self-reported and verified by medical records. Binary logistic regression considering for potential confounders (age, sex, BMI, Charlson comorbidity index, femoral neck bone density) and receiver operating characteristics (ROC) curves were used in statistical analysis.
Results
508 community-dwelling older adults (median age: 78 years, interquartile range: 72, 83, 75.2% women) were included. Compared to greater relative sit-to-stand muscle power (1.62-3.78W.kg-1 for women; 2.03-3.90W.kg-1 for men), those with those with extremely low relative sit-to-stand muscle power were 2.35 (95% CI 1.54, 3.60, p<0.001) and 2.41 (95% CI 1.25, 4.65, p= 0.009) times more likely to experience recurrent falls and fractures (respectively) in fully-adjusted model. Compared to grip strength or gait speed, relative sit-to-stand muscle power showed the highest area under the ROC curve for identifying recurrent falls (AUC: 0.64) and fractures (AUC: 0.62). All tests showed low diagnostic power (AUC: <0.7).
Conclusion
Relative sit-to-stand muscle power performed slightly (but not statistically) better than grip strength or gait speed for identifying a history of recurrent falls and fractures in older adults. However, all tests showed low diagnostic power.