Relative Validity and Reliability of Isometric Lower Extremity Strength Assessment in Older Adults by Using a Handheld Dynamometer

Abstract

Background:

Handheld dynamometry (HHD) is a practical alternative to traditional testing of lower extremity strength. However, its reliability and validity across different populations and settings are not clear.

Hypothesis:

We hypothesize that HHD is a valid and reliable device to assess lower extremity strength in a population of older adults.

Study Design:

Cross-sectional/cohort.

Level of Evidence:

Level 3.

Methods:

This study included 258 older adults (≥65 years). Isometric knee extension and flexion force were measured by 1 examiner, using an HHD (n = 222), including 3 repetitions to calculate within-day intrarater reliability. These measurements were repeated by the examiner in a subgroup (n = 23) to analyze intrarater reliability over a test-retest period of on average 8 weeks. In addition, HHD force measures were performed by a second examiner (n = 29) to analyze interrater reliability. In another subgroup (n = 77), isometric knee extension and flexion torque were measured by 1 examiner using both the HHD and Biodex System 4 to assess relative validity.

Results:

HHD and Biodex measurements were highly correlated and showed excellent concurrent validity. HHD systematically overestimated torque as compared with Biodex by 8 N·m on average. Same-day intrarater intraclass correlation coefficients (ICCs) ranged from 0.97 to 0.98. Interrater reliability ICCs ranged from 0.83 to 0.95.

Conclusion:

HHD represents a reliable and valid alternative to Biodex to rank individuals on leg strength, or to assess within-person changes in leg strength over time, because of the high validity and reliability. The HHD is less suited for absolute strength assessment because of significant systematic overestimations.

Clinical Relevance:

Clinicians are encouraged to use HHD to rank older adults on leg strength, or to assess within-person changes in leg strength over time, but not to compare readings with cut-offs or normative values.