Abstract With obesity becoming a major health issue in recent years and a topic of interest in vehicle occupant safety, there is a need for detailed descriptions of body fat distribution in order to develop accurate models of the human body. In this study subcutaneous fat measurements were obtained from over 17,000 CT scans using planar distance maps from skin surface locations to the fascial envelope. Measures were taken as a map registered by vertebral level (from T6 down to L5, extended to the sacrum) and body location around the body perimeter. Multivariate regression maps were calculated showing the individual effect of demographics on the subcutaneous layer thickness. Regression coefficients were statistically significant, with results showing the progression of 20 years of age producing a migration of 3mm of fat from mid‐abdominal to lower‐abdominal regions and female occupants having thicker subcutaneous regions compared to males by between 6 and 24mm with accumulation peaks near the buttocks and breasts. Each 20kg of added body weight produce regional fat increases between 2 and 14mm across the torso, with a primary peak near the hips and secondary peak near the lower abdomen. Keywords fat, human body, model, obesity, subcutaneous I. INTRODUCTION Concern for increasing obesity has gained international interest since obesity will soon become the world's biggest health problem (NIH 2003 [1]). In motor vehicle crashes, obesity has been shown to influence the risks of serious and fatal injury [2‐4]. Jehle et al. [5] found that overweight occupants had the lowest risk of fatality while obese and morbidly obese occupants had the highest risks. The authors suggested that subcutaneous fat may provide a cushioning effect for overweight drivers involved in a vehicle crash. The authors also indicated that a large amount of subcutaneous fat may force drivers to sit in close proximity to the steering wheel increasing their injury risks. Parenteau et al. [6] found that the amount of subcutaneous fat differed based on location, vertebra level, BMI and gender. Crash test dummies and post‐mortem human subjects (PMHS) are generally used to assess occupant responses in a simulated crash environment. They are however representative of normal‐sized occupants. Different methods have been used to assess the effect of weight on occupant response. For example, McGowan [7] ballasted the weight of a 95th percentile male Hybrid III dummy to 300 lb. to study the effect of seat back stiffness in rear impacts. Additional weights were added to the abdomen, thorax and extremities. Kent et al. [8] tested two obese and three normal weight, lap‐shoulder belted PMHS to study the effect of obesity on the biomechanical responses in frontal crashes. The authors found increased excursion with the obese occupants in the test. Zhu et al. [9] developed computer models of various sized male and female drivers. They varied the amount of torso subcutaneous fat in the form of elliptical masses added axially around the body based on limited MRI data. Other studies [10] have also used digital models to assess the role of obesity on occupant seated posture. While considerable clinical research (see [11] for review) has gone into using medical imaging modalities to assess body fat content, published studies are generally focused on the association of fat content with particular clinical outcomes, rather than detailing the distribution of the fat itself. This study aims to enhance the available data for human body modeling by providing 3‐dimensional torso subcutaneous fat measures registered by underlying vertebral landmarks and skin perimeter locations. A large underlying population of CT scans drives regression analyses to highlight local changes in subcutaneous thickness by demographic factors of age, gender and weight. Sven A. Holcombe, Stewart C. Wang Subcutaneous Fat Distribution in the Human TorsoSven Holcombe (svenho@umich.edu, +1‐734‐647‐5444) is a PhD student in Biomedical Engineering at the University of Michigan in Ann Arbor, USA. S. C. Wang is director of the International Center for Automotive Medicine at the University of Michigan Hospital. IRC-14-43 IRCOBI Conference 2014- 389 -
