John A Shepherd

Background & aimsThe multicompartment approach to body composition modeling provides a more precise quantification of body compartments in healthy and clinical populations. We sought to develop and validate a simplified and accessible multicompartment body composition model using 3-dimensional optical (3DO) imaging and bioelectrical impedance analysis (BIA).MethodsSamples of adults and collegiate-aged student-athletes were recruited for model calibration. For the criterion multicompartment model (Wang-5C), participants received measures of scale weight, body volume (BV) via air displacement, total body water (TBW) via deuterium dilution, and bone mineral content (BMC) via dual energy x-ray absorptiometry. The candidate model (3DO-5C) used stepwise linear regression to derive surrogate measures of BV using 3DO, TBW using BIA, and BMC using demographics. Test-retest precision of the …

BackgroundBreast density (BD) is a strong risk factor for breast cancer. Little is known about how BD develops during puberty. Understanding BD trajectories during puberty and its determinants could be crucial for promoting preventive actions against breast cancer (BC) at early ages. The objective of this research is to characterize % fibroglandular volume (%FGV), absolute fibroglandular volume (AFGV), and breast volume (BV) at different breast Tanner stages until 4-year post menarche in a Latino cohort and to assess determinants of high %FGV and AFGV during puberty and in a fully mature breast.MethodsThis is a longitudinal follow-up of 509 girls from low-middle socioeconomic status of the Southeast area of Santiago, recruited at a mean age of 3.5 years. The inclusion criteria were singleton birth born, birthweight between 2500 and 4500 g with no medical or mental disorder. A trained dietitian measured …

BackgroundArtificial intelligence (AI) large language models (LLMs) such as ChatGPT have demonstrated the ability to pass standardized exams. These models are not trained for a specific task, but instead trained to predict sequences of text from large corpora of documents sourced from the internet. It has been shown that even models trained on this general task can pass exams in a variety of domain-specific fields, including the United States Medical Licensing Examination. We asked if large language models would perform as well on a much narrower subdomain tests designed for medical specialists. Furthermore, we wanted to better understand how progressive generations of GPT (generative pre-trained transformer) models may be evolving in the completeness and sophistication of their responses even while generational training remains general. In this study, we evaluated the performance of two versions …

Purpose of the studyBreast density is an established risk factor for breast cancer. However, little is known about metabolic influences on breast density phenotypes. We conducted untargeted serum metabolomics analyses to identify metabolic signatures associated with breast density phenotypes among young women.MethodsIn a cross-sectional study of 173 young women aged 25–29 who participated in the Dietary Intervention Study in Children 2006 Follow-up Study, 449 metabolites were measured in fasting serum samples using ultra-high-performance liquid chromatography-tandem mass spectrometry. Multivariable-adjusted mixed-effects linear regression identified metabolites associated with magnetic resonance imaging measured breast density phenotypes: percent dense breast volume (%DBV), absolute dense breast volume (ADBV), and absolute non-dense breast volume (ANDBV). Metabolite results …

Background Folate is the primary methyl donor and B vitamins are cofactors for one-carbon metabolism that maintain DNA integrity and epigenetic signatures implicated in carcinogenesis. Breast tissue is particularly susceptible to stimuli in early life. Only limited data are available on associations of one-carbon metabolism–related vitamin intake during youth and young adulthood with breast density, a strong risk factor for breast cancer. Methods Over 18 years in the DISC and DISC06 Follow-up Study, diets of 182 young women were assessed by three 24-hour recalls on five occasions at ages 8 to 18 years and once at 25 to 29 years. Multivariable-adjusted linear mixed-effects regression was used to examine associations of intakes of one-carbon metabolism-related vitamins with MRI-measured percent dense breast volume (%DBV) and absolute dense breast volume (ADBV …

Accurate assessment of body composition is essential for evaluating the risk of chronic disease. Access to medical imaging methods is limited due to practical and ethical constraints. 3D body shape, which can be obtained using smartphones, correlates strongly with body composition. However, large-scale datasets containing 3D body shapes with paired anthropometric and metabolic traits are scarce. Here, we present a novel method that fits a 3D body mesh to a dual-energy X-ray absorptiometry (DXA) silhouette which is paired with anthropometric traits (height, waist and hip circumferences), using a large dataset from the UK population-based Fenland study (12,435 adults, age 30–65 years at baseline phase 1). We predict total and regional body composition metrics using these meshes, and monitor changes in body composition between the baseline (phase 1) and a follow-up assessment (phase 2). We also evaluate a 3D body shape smartphone app which reconstructs a 3D body mesh from phone images to predict body composition metrics, and compare the results against the reference methods DXA and air plethysmography. In the Fenland validation dataset (follow up), total and regional body composition metrics were predicted accurately, achieving r> 0.86 for all metrics. Absolute mean bias expressed as percentage of the mean was less than 2% for all metrics except for visceral fat mass and subcutaneous abdominal fat mass (4%, 7% respectively). Predictions for changes achieved r> 0.60 for all metrics. The predicted metrics from the smartphone generated avatars also showed strong correlations r> 0.84 for all metrics. The 3D body …

BackgroundBody shape, an intuitive health indicator, is deterministically driven by body composition. We developed and validated a deep learning model that generates accurate dual-energy X-ray absorptiometry (DXA) scans from three-dimensional optical body scans (3DO), enabling compositional analysis of the whole body and specified subregions. Previous works on generative medical imaging models lack quantitative validation and only report quality metrics.MethodsOur model was self-supervised pretrained on two large clinical DXA datasets and fine-tuned using the Shape Up! Adults study dataset. Model-predicted scans from a holdout test set were evaluated using clinical commercial DXA software for compositional accuracy.ResultsPredicted DXA scans achieve R2 of 0.73, 0.89, and 0.99 and RMSEs of 5.32, 6.56, and 4.15 kg for total fat mass (FM), fat-free mass (FFM), and total mass, respectively …

Total and regional body composition are strongly correlated with metabolic syndrome and have been estimated non-invasively from 3D optical scans using linear parameterizations of body shape and linear regression models. Prior works produced accurate and precise predictions on many, but not all, body composition targets relative to the reference dual X-Ray absorptiometry (DXA) measurement. Here, we report the effects of replacing linear models with nonlinear parameterization and regression models on the precision and accuracy of body composition estimation in a novel application of deep 3D convolutional graph networks to human body composition modeling.

Whether trace metals modify breast density, the strongest predictor for breast cancer, during critical developmental stages such as puberty remains understudied. Our study prospectively evaluated the association between trace metals at Tanner breast stage B1 (n = 291) and at stages both B1 and B4 (n = 253) and breast density at 2 years post‐menarche among Chilean girls from the Growth and Obesity Cohort Study. Dual‐energy x‐ray absorptiometry assessed the volume of dense breast tissue (absolute fibroglandular volume [FGV]) and percent breast density (%FGV). Urine trace metals included arsenic, barium, cadmium, cobalt, cesium, copper, magnesium, manganese, molybdenum, nickel, lead, antimony, selenium, tin, thallium, vanadium, and zinc. At B1, a doubling of thallium concentration resulted in 13.69 cm3 increase in absolute FGV (β: 13.69, 95% confidence interval [CI]: 2.81, 24.52), while a …

BackgroundPubertal growth patterns correlate with future health outcomes. However, the genetic mechanisms mediating growth trajectories remain largely unknown. Here, we modeled longitudinal height growth with Super-Imposition by Translation And Rotation (SITAR) growth curve analysis on ~ 56,000 trans-ancestry samples with repeated height measurements from age 5 years to adulthood. We performed genetic analysis on six phenotypes representing the magnitude, timing, and intensity of the pubertal growth spurt. To investigate the lifelong impact of genetic variants associated with pubertal growth trajectories, we performed genetic correlation analyses and phenome-wide association studies in the Penn Medicine BioBank and the UK Biobank.ResultsLarge-scale growth modeling enables an unprecedented view of adolescent growth across contemporary and 20th-century pediatric cohorts. We identify …

Combining classification systems potentially improves predictive accuracy, but outcomes have proven impossible to predict. Similar to improving binary classification with fusion, fusing ranking systems most commonly increases Pearson or Spearman correlations with a target when the input classifiers are "sufficiently good" (generalized as "accuracy") and "sufficiently different" (generalized as "diversity"), but the individual and joint quantitative influence of these factors on the final outcome remains unknown. We resolve these issues. Building on our previous empirical work establishing the DIRAC (DIversity of Ranks and ACcuracy) framework, which accurately predicts the outcome of fusing binary classifiers, we demonstrate that the DIRAC framework similarly explains the outcome of fusing ranking systems. Specifically, precise geometric representation of diversity and accuracy as angle-based distances within rank …

Purpose. Screening mammography is unavailable in many low-resource areas. We ask if the state-of-the-art in artificial intelligence (AI)-enhanced breast ultrasound (BUS) is sufficiently accurate to be used for primary breast cancer screening in low-resource regions. Background. Since the 1980s, high-income countries have implemented mammographic screening programs, leading to breast cancer mortality reduction in screened women.1 Mammography is unavailable in many low-resource regions, such as the USAPI. Furthermore, travel difficulties and lack of radiologists hinder implementation. AI combined with portable BUS may address limitations of the high-income paradigm. In this systematic review, we ask if AI-enhanced BUS can detect/segment lesions (Objective 1) and classify lesions as cancerous (Objective 2). Methods. Two reviewers independently assessed articles from 1/1/2016 to 8/6/2023 from …

Background There are established links between the accumulation of body fat as visceral adipose tissue (VAT) and the risk of developing obesity‐associated metabolic disease. Previous studies have suggested that levels of intake of specific foods and nutrients are associated with VAT accumulation after accounting for total energy intake. Objective This study assessed associations between a priori selected dietary factors on VAT quantified using abdominal magnetic resonance imaging. Methods The cross‐sectional Multiethnic Cohort Adiposity Phenotype Study included n = 395 White, n = 274 Black, n = 269 Native Hawaiian, n = 425 Japanese American and n = 358 Latino participants (mean age = 69 years ± 3 SD). Participants were enrolled stratified on sex, race, ethnicity and body mass index. General linear models were used to estimate the mean VAT area (cm2) for participants categorized into quartiles …

BackgroundAthletes vary in hydration status due to ongoing training regimes, diet demands, and extreme exertion. With water being one of the largest body composition compartments, its variation can cause misinterpretation of body composition assessments meant to monitor strength and training progress. In this study, we asked what accessible body composition approach could best quantify body composition in athletes with a variety of hydration levels.MethodsThe Da Kine Study recruited collegiate and intramural athletes to undergo a variety of body composition assessments including air-displacement plethysmography (ADP), deuterium-oxide dilution (D2O), dual-energy X-ray absorptiometry (DXA), underwater-weighing (UWW), 3D-optical (3DO) imaging, and bioelectrical impedance (BIA). Each of these methods generated 2- or 3-compartment body composition estimates of fat mass (FM) and fat-free mass …

Background and aimsBody fat distribution, i.e., visceral (VAT), subcutaneous adipose tissue (SAT) and intramuscular fat, is important for disease prevention, but sex and ethnic differences are not well understood. Our aim was to identify anthropometric, demographic, and lifestyle predictors for these outcomes.Methods and resultsThe cross-sectional ShapeUp!Kids study was conducted among five ethnic groups aged 5–18 years. All participants completed questionnaires, anthropometric measurements, and abdominal MRI scans. VAT and SAT areas at four lumbar levels and muscle density were assessed manually. General linear models were applied to estimate coefficients of determination (R2) and to compare the fit of VAT and SAT prediction models. After exclusions, the study population had 133 male and 170 female participants. Girls had higher BMI-z scores, waist circumference (WC), and SAT than boys but …

PurposeAge and body mass index (BMI) are critical considerations when assessing individual breast cancer risk, particularly for women with dense breasts. However, age- and BMI-standardized estimates of breast density are not available for screen-aged women, and little is known about the distribution of breast density in women aged < 40. This cross-sectional study uses three different modalities: optical breast spectroscopy (OBS), dual-energy X-ray absorptiometry (DXA), and mammography, to describe the distributions of breast density across categories of age and BMI.MethodsBreast density measures were estimated for 1,961 Australian women aged 18–97 years using OBS (%water and %water + %collagen). Of these, 935 women had DXA measures (percent and absolute fibroglandular dense volume, %FGV and FGV, respectively) and 354 had conventional mammographic measures (percent and …

Purpose/AimsIdentify visceral adipose tissue (VAT) thresholds associated with increased cardiometabolic disease risk.Rationale/BackgroundBeyond overall obesity, VAT storage is associated with adverse metabolic parameters that increase the risk of heart disease, stroke, and type II diabetes. Computed tomography (CT) and magnetic resonance imaging (MRI) technologies have been used to define thresholds of VAT associated with MetS, however these techniques are of limited availability for clinical risk assessment. Dual energy X-ray absorptiometry (DXA) is accurate compared to VAT measures from CT and MRI, however differences in scanning region and algorithms results in device-specific VAT estimates. We recently generated cross-calibration equations for DXA systems, allowing DXA measures to be compared to CT and MRI and providing more access to VAT assessments for the determination of VAT …

BackgroundExcess adiposity in children is strongly correlated with obesity-related metabolic disease in adulthood, including diabetes, cardiovascular disease, and 13 types of cancer. Despite the many long-term health risks of childhood obesity, body mass index (BMI) Z-score is typically the only adiposity marker used in pediatric studies and clinical applications. The effects of regional adiposity are not captured in a single scalar measurement, and their effects on short- and long-term metabolic health are largely unknown. However, clinicians and researchers rarely deploy gold-standard methods for measuring compartmental fat such as magnetic resonance imaging (MRI) and dual X-ray absorptiometry (DXA) on children and adolescents due to cost or radiation concerns. Three-dimensional optical (3DO) scans are relatively inexpensive to obtain and use non-invasive and radiation-free imaging techniques to …

Currently available anthropometric body composition prediction equations were often developed on small participant samples, included only several measured predictor variables, or were prepared using conventional statistical regression methods. Machine learning approaches are increasingly publicly available and have key advantages over statistical modeling methods when developing prediction algorithms on large datasets with multiple complex covariates. This study aimed to test the feasibility of predicting DXA-measured appendicular lean mass (ALM) with a neural network (NN) algorithm developed on a sample of 576 participants using 10 demographic (sex, age, 7 ethnic groupings) and 43 anthropometric dimensions generated with a 3D optical scanner. NN-predicted and measured ALM were highly correlated (n = 116; R2, 0.95, p < 0.001, non-significant bias) with small mean, absolute, and root …

Digital anthropometry by three-dimensional optical imaging systems and smartphones has recently been shown to provide non-invasive, precise, and accurate anthropometric and body composition measurements. To our knowledge, no previous study performed smartphone-based digital anthropometric assessments in young athletes. The aim of this study was to investigate the reproducibly and validity of smartphone-based estimation of anthropometric and body composition parameters in youth soccer players. A convenience sample of 124 male players and 69 female players (median ages of 16.2 and 15.5 years, respectively) was recruited. Measurements of body weight and height, one whole-body Dual-Energy X-ray Absorptiometry (DXA) scan, and acquisition of optical images (performed in duplicate by the Mobile Fit app to obtain two avatars for each player) were performed. The reproducibility analysis …