Files in this item



application/pdfFOX-DISSERTATION-2020.pdf (11MB)
(no description provided)PDF


Title:The biomechanical consequences of body size differences in humans
Author(s):Fox, Maria Christine
Director of Research:Polk, John D; Hsiao-Wecksler, Elizabeth T
Doctoral Committee Chair(s):Polk, John D; Hsiao-Wecksler, Elizabeth T
Doctoral Committee Member(s):Kersh, Mariana E; Konigsberg, Lyle W; Shackelford, Laura L
Department / Program:Anthropology
Degree Granting Institution:University of Illinois at Urbana-Champaign
gait biomechanics
postural sway
bone microstructure
human variation
body size
Abstract:Humans vary widely in body size and shape, but all humans are bipedal, and they are assumed to perform bipedalism in essentially the same way. Body size affects nearly all aspects of any animal’s ecology, physiology, and behavior, including locomotor behavior. Much of the theory underlying the consequences of size variation on locomotion was developed from broad interspecific studies of mammals and birds, but researchers have focused relatively little on the consequences of size within species, or within humans in particular. This dissertation examined how body size influences human morphology and behavior in four specific areas: 1) scaling of linear anthropometric dimensions; 2) scaling of stiffness, force, displacement, and leg spring angle during running; 3) quiet standing postural sway and stance characteristics; and 4) variability and scaling patterns in bone microstructure of the femur. Additionally, this dissertation explored different statistical approaches for examining each of these areas to determine whether results were robust to the chosen statistical approach and whether novel statistical methods would offer new insights into the data. These studies represent a holistic analysis of human size variation from macro-scale locomotor behavior to micro-scale bone microstructure characteristics both within and between sexes. The first study examined the scaling of human linear anthropometric dimensions from motion capture data in a sample of 104 size-varying adults using two statistical methods: log-log regression and principal component analysis (PCA). Results of this study indicated agreement between patterns of observed allometry in the two methods in the pooled sex sample, but not in magnitude, where PCA resulted in exaggerated allometry coefficients. Both methods produced isometric scaling of the upper arm, thigh, and shoulder in the pooled sex analysis, but results differed between methods in the sex-specific analysis. When sexes were pooled, departures from isometry were minimal in all but pelvis width. When sexes were analyzed separately, all dimensions except for shoulder width adhered to isometry in the log-log regressions. The second study analyzed how leg and vertical stiffness, force, displacement, and leg spring angle scaled with body mass in a sample of 69 adults when running at the same relative (leg-length adjusted) speeds. Allometric scaling exponents were calculated for stiffness, force, displacement, and leg spring angle via kinematic and kinetic data using three different types of log-log regressions (ordinary least squares, linear mixed models, and Bayesian linear mixed models) with 95% confidence/highest density intervals. In this sample, regardless of model choice, all variables scaled according to the isometric expectations, suggesting that humans do not violate the assumptions of dynamic similarity. Sex-specific analyses revealed similar patterns of isometry in all variables except for vertical stiffness at the slow running speed. The third study examined how 1) anthropometrics and sex influence stance characteristics, and 2) how anthropometrics, sex, and stance characteristics influence postural sway metrics during quiet standing in 69 adults. A multivariate regression technique, partial least squares regression (PLSR), was employed to evaluate highly-correlated and complex relationships between variables. Stance characteristics were strongly related to sex and to some extent body size. Postural sway metrics were dominated primarily by stance characteristics, but did display some effects of body size, where shorter subjects had less anteroposterior sway magnitude and variability. In a subset of subjects with less variable stance width, body size characteristics became more influential predictors of postural sway. Sex-specific analyses primarily suggested the common influence of stance width on postural sway. The fourth study evaluated variability and scaling patterns in subchondral and trabecular bone microstructure of the femoral medial condyle in 24 specimens. Coefficients of variation (CV) were calculated to describe variability and Bayesian linear mixed models were employed to examine size differences and scaling patterns in the sample. Results indicated high variability in subchondral and trabecular properties both across specimens and across the joint surface. The scaling analysis revealed no real size dependency in the properties when all volumes of interest (VOIs) were analyzed together, although trabecular spacing and bone volume fraction did trend towards negative and positive allometry, respectively. It appeared that humans in this sample may not follow interspecific scaling patterns observed in other animals, likely due the large variability observed in the properties analyzed. Results from these scaling studies indicated that intraspecific scaling patterns match those found interspecifically for movement patterns and to some extent limb dimensions, but not for bone microstructure. Each study revealed slightly different results for how body size influenced the parameters tested, but all indicated a large amount of variability in the observed patterns. Humans are highly variable in morphology and behavior, but this variability appears to affect relationships with body size differently on the macro- and micro-scale.
Issue Date:2020-05-06
Rights Information:Copyright 2020 Maria C Fox
Date Available in IDEALS:2020-08-26
Date Deposited:2020-05

This item appears in the following Collection(s)

Item Statistics