Handgrip strength in elderly with dementia: study of reliability

Background: Measuring instruments should have their scientific and clinical value evaluated in different populations. The handgrip strength test is widely used, however little has been investigated about its reliability when used in elderly with dementia and the right stage wich its use should be avoided. Objectives: To evaluate the test-retest reliability of the handgrip strength test in elderly with different ratings of dementia. Method: The cognitive function of 76 elderly subjects with dementia was measured, and the caregivers were interviewed to allow classification by the Clinical dementia rating (CDR). For these assessments the Mini-Metal State Examination and the Pfeffer, Lawton, and Katz scales were used. Twenty subjects were classified as borderline (83.4 +/- 5.8 years), 19 as mild (82.4 +/- 6.8 years), 19 as moderate (85.8 +/- 5.6 years) and 18 as severe dementia (84.0 +/- 5.1 years). Handgrip strength was assessed with a JAMAR hydraulic dynamometer and after one week it was reevaluated. Reliability was analyzed by Intraclass Correlation Coefficient (ICC). The significance level was set at alpha=0.05. Results: Test-retest reliability was excellent for groups with borderline (ICC=0.975; p=0.001), mild (ICC=0.968; p=0.002), and moderate (ICC=0.964; p=0.001) dementia. The analysis of the group with a severe CDR showed no statistical significance and a low ICC (ICC=0.415; p=0.376). Conclusion: The handgrip strength test has excellent reliability when used in elderly with borderline, mild, and moderate dementia, which enables its use in research. However, its use is not recommended in elderly classified with severe dementia due to the measure's low reliability and subsequent irrelevance in clinical practice.