Normalization alters the interpretation of hip strength in established unilateral lower limb prosthesis users

Valid comparisons of muscle strength between individuals or legs that differ in size requires normalization, often by simple anthropometric variables. Few studies of muscle strength in lower-limb prosthesis users have normalized strength data by any anthropometric variable, potentially confounding our understanding of strength deficits in lower-limb prosthesis users. The objective of this pilot study was to determine the need for as well as effectiveness and impact of normalizing hip strength in lower-limb prosthesis users.

Peak isometric hip extension and abduction torques were collected from 28 lower-limb prosthesis users. Allometric scaling was used to determine if hip torque values were significantly associated with, and therefore needed to be adjusted for, body mass, thigh length, or body mass x thigh length, and whether normalization was effective in reducing any associations. Between limb differences in peak hip torque, and correlations with balance ability, were inspected pre- and post-normalization.

Hip torques were consistently and significantly associated with body-mass x thigh length. Associations between peak hip torque and body-mass x thigh length were reduced by normalization. After normalization by body-mass x thigh length, between limb differences in hip extension torque, as well as the correlation between hip abduction torque and balance ability, changed from non-significant to significant.

In the absence of normalization, hip strength (i.e., peak torque) in lower-limb prosthesis users remains dependent on basic anthropometric variables, masking relationships between hip strength and balance ability, as well as between limb differences.