Reliability and validity of measurements of knee extension strength obtained from nursing home residents with dementia

Am J Phys Med Rehabil. 2009 Nov;88(11):924-33. doi: 10.1097/PHM.0b013e3181ae1003.


Objective: To describe the reliability of strength measurement by hand-held dynamometer and to identify the prediction of gait and stand-up ability by the strength measurement in people with dementia.

Design: Correlational study in which 60 persons with Alzheimer disease were enrolled. The strength of knee extensor was measured twice separated by a 3-min interval using a hand-held dynamometer. The presence or absence of impaired gait and sit-to-stand was also determined.

Results: Intraclass correlation coefficient was 0.97. Bland-Altman plots showed the distribution of test-retest differences in the subjects with Mini-Mental State Examination score of 10 points or fewer, increased with isometric knee extensor strength. Logistic regression analysis showed strength of the knee extensor muscles to be a significant predictor of gait performance (odds ratio, 443.0; 95% confidence interval, 9.20-21325.7) and sit-to-stand performance (odds ratio, 47.32; 95% confidence interval, 3.31-675.8).

Conclusions: The strength measurement with a hand-held dynamometer was reliable in people with dementia. Furthermore, normalized knee extensor strength was indicated to be a significant predictor of gait and sit-to-stand performance. However, caution is needed when interpreting the result of strength measurement when it is carried out on subjects with a Mini-Mental State Examination score of 10 points or fewer and with a high strength level.

Publication types

  • Validation Study

MeSH terms

  • Aged
  • Aged, 80 and over
  • Dementia / diagnosis
  • Dementia / physiopathology
  • Dementia / rehabilitation*
  • Female
  • Gait
  • Geriatric Assessment / methods*
  • Humans
  • Knee Joint / physiopathology*
  • Male
  • Muscle Strength Dynamometer
  • Muscle Strength*
  • Nursing Homes
  • Range of Motion, Articular
  • Reproducibility of Results