Analytical and experimental models of the musculoskeletal system often
assume single values rather than ranges for anatomical input paramete
rs. The hypothesis of the present study was that anatomical variabilit
y significantly influences the results of biomechanical analyses, spec
ifically regarding the moment arms of the various thigh muscles. Inser
tions and origins of muscles crossing or attaching to the femur were d
igitized in six specimens. Muscle volumes were measured; muscle attach
ment area and centroid location were computed. To demonstrate the infl
uence of inter-individual anatomic variability on a mechanical modelin
g parameter, the corresponding range of muscle moment arms were calcul
ated. Standard deviations, as a percentage of the mean, were about 70%
for attachment area and 80% for muscle volume and attachment centroid
location. The resulting moment arms of the m. gluteus maximus and m.
rectus femoris were especially sensitive to anatomical variations (SD
65%). The results indicate that sensitivity to anatomical variations s
hould be analyzed in any investigation simulating musculoskeletal inte
ractions. To avoid misinterpretations, investigators should consider u
sing several anatomical configurations rather than relying on a mean d
ata set. Copyright (C) 1996 Elsevier Science Ltd.