The path less traveled: Using structural equation modeling to investigate factors influencing bone functional morphology.

Abstract

The relationship between an organism's mechanical environment and its bone strength has been long established by experimental research. Multiple intrinsic and extrinsic factors, including body mass, muscle strength, genetic background, and nutritional and/or hormonal status, are likely to influence bone deposition and resorption throughout the lifespan, complicating this relationship. Structural equation modeling (SEM) is uniquely positioned to parse this complex set of influences. Data from the Third National Health and Nutrition Examination Survey, including sex, total body mass, lean body mass, exercise frequency, peak body mass, and age, were analyzed using SEM to determine how they affect bone strength both individually and combined. Body mass is typically the driver of cross-sectional area, but body mass and lean mass have similar effects on the polar moment of area (J). Peak body mass had a strong direct effect on J, despite decreasing strongly with increases in lean mass. Exercise also did not confer a large direct effect on cross-sectional area or J but did modify body mass and lean mass. In females, intentional weight loss was associated with decreased exercise levels. SEM is a useful tool for parsing complex systems in bone functional morphology and has the potential to uncover causal links in the study of skeletal remodeling, including factors like weight loss or exercise that may have secondary effects.