Smartphone camera based assessment of adiposity: a validation study

Maternal obesity and offspring metabolism: revisiting dietary interventions

Maternal obesity increases the risk of metabolic disorders in offspring. Understanding the mechanisms underlying the transgenerational transmission of metabolic diseases is important for the metabolic health of future generations. More research is needed to elucidate the mechanisms underlying the associated risks and their clinical implications because of the inherently complex nature of transgenerational metabolic disease transmission. Diet is a well-recognized risk factor for the development of obesity and other metabolic diseases, and rational dietary interventions are potential therapeutic strategies for their prevention. Despite extensive research on the physiological effects of diet on health and its associated mechanisms, little work has been devoted to understanding the effects of early-life dietary interventions on the metabolic health of offspring. In addition, existing dietary interventions are insufficient to meet clinical needs. Here, we discuss the literature on the effects of maternal obesity on the metabolic health of offspring, focusing on the mechanisms underlying the transgenerational transmission of metabolic diseases. We revisit current dietary interventions and describe their strengths and weaknesses in ameliorating maternal obesity-induced metabolism-related disorders in offspring. We also propose innovative strategies, such as the use of precision nutrition and fecal microbiota transplantation, which may limit the vicious cycle of intergenerational metabolic disease transmission.

Comparing digital anthropometrics from mobile applications to reference methods: a scoping review

This scoping review aimed to assess the repeatability and accuracy of Digital Anthropometry by Mobile Application (DAM) compared to reference methods for estimating anthropometric dimensions, body volume (BV), and body composition. A comprehensive search was conducted on December 8th, 2024, without restrictions on language, time, sex, ethnicity, age, or health condition. We found 14 different DAMs across the 23 included studies. Reference methods for each estimated variable were: (a) Body circumferences-tape measure; (b) body mass-calibrated scale; (c) body length-stadiometer; (d) BV-Underwater Weighing; (e) percentage of body fat-Dual energy x-ray absorptiometry (DXA), BOD POD, 3, 4, and 5-compartment models; (f) fat mass and fat-free mass-DXA, 3 and 4-compartment models; (g) appendicular Lean Mass-DXA. DAMs demonstrated high repeatability and accuracy at a mean level in most studies. However, their accuracy is lower at individual-level analysis and for tracking changes over time. Estimated BV showed high accuracy compared to UWW (SEE = 0.68; MD = 0.04 to 0.1; LoA = 2.86), including the BV-derived DAMs integrated into alternative multi-compartment models compared to reference methods. As relatively new methods, DAMs offer numerous possibilities and areas for exploration in future studies. However, caution is advised due to their potentially low or unknown accuracy at the individual level.

American Society for Metabolic and Bariatric Surgery Review of Body Composition

Although the body mass index (BMI) has been used as a measure of obesity for decades, it is now possible to measure adiposity more directly with technologies that can quantitate body fat and other tissues. The purpose of this review is to understand body composition, describe the different ways to measure it, review changes in body composition after metabolic and bariatric surgery (MBS), and provide guidance on how providers can introduce measurements of body composition into their everyday practice.

Core Components of Cardiac Rehabilitation Programs: 2024 Update: A Scientific Statement From the American Heart Association and the American Association of Cardiovascular and Pulmonary Rehabilitation: Endorsed by the American College of Cardiology

The science of cardiac rehabilitation and the secondary prevention of cardiovascular disease has progressed substantially since the most recent American Heart Association and American Association of Cardiovascular and Pulmonary Rehabilitation update on the core components of cardiac rehabilitation and secondary prevention programs was published in 2007. In addition, the advent of new care models, including virtual and remote delivery of cardiac rehabilitation services, has expanded the ways that cardiac rehabilitation programs can reach patients. In this scientific statement, we update the scientific basis of the core components of patient assessment, nutritional counseling, weight management and body composition, cardiovascular disease and risk factor management, psychosocial management, aerobic exercise training, strength training, and physical activity counseling. In addition, in recognition that high-quality cardiac rehabilitation programs regularly monitor their processes and outcomes and engage in an ongoing process of quality improvement, we introduce a new core component of program quality. High-quality program performance will be essential to improve widely documented low enrollment and adherence rates and reduce health disparities in cardiac rehabilitation access.

Advances in the estimation of body fat percentage using an artificial intelligence 2D-photo method

There is a growing need to evaluate the agreement between the field methods and integrate artificial intelligence (AI) using two-dimensional (2D) photos for enhanced real-world analysis. This study evaluated the agreement between AI-2D photos and the clinical reference method, dual-energy X-ray absorptiometry (DXA) to estimate the body fat percentage (BFP). Other methods were also investigated, including skinfolds, A-mode ultrasound, and bioelectrical impedance analysis (BIA). This cross-sectional study was conducted on 1273 adults of both sexes. The Bland-Altman plots, Lin's Correlation Coefficient of Agreement (CCC), and error analyses were calculated. AI-2D photos demonstrated substantial agreement with DXA presenting the highest agreement (CCC ≥ 0.96) among all the investigated methods. InBody-270 and Omron HBF-514 BIA devices showed moderate agreement (CCC = 0.90 to 0.95) for all participants, age groups >30 years, and body mass index >25 kg/m2. AI-2D photos can be interchangeable with DXA, providing a practical, accessible alternative and an easy-to-use system for BFP estimation.

Updates on Methods for Body Composition Analysis: Implications for Clinical Practice

BACKGROUND

Recent technological advances have introduced novel methods for measuring body composition, each with unique benefits and limitations. The choice of method often depends on the trade-offs between accuracy, cost, participant burden, and the ability to measure specific body composition compartments.

OBJECTIVE

To review the considerations of cost, accuracy, portability, and participant burden in reference and emerging body composition assessment methods, and to evaluate their clinical applicability.

METHODS

A narrative review was conducted comparing traditional reference methods like dual-energy X-ray absorptiometry (DXA), magnetic resonance imaging (MRI), and computed tomography (CT) with emerging technologies such as smartphone camera applications, three-dimensional optical imaging scanners, smartwatch bioelectric impedance analysis (BIA), and ultrasound.

RESULTS

Reference methods like CT and MRI offer high accuracy and the ability to distinguish between specific body composition compartments (e.g., visceral, subcutaneous, skeletal muscle mass, and adipose tissue within lean mass) but are expensive and non-portable. Conversely, emerging methods, such as smartwatch BIA and smartphone-based technologies, provide greater accessibility and lower participant burden but with reduced accuracy. Methods like three-dimensional optical imaging scanners balance portability and accuracy, presenting promising potential for population-level applications.

CONCLUSIONS

The selection of a body composition assessment method should be guided by the clinical context and specific application, considering trade-offs in cost, accuracy, and portability. Emerging methods provide valuable options for population-level assessments, while reference methods remain essential for detailed compartmental analysis.

Predicting Bone Mineral Content from Smartphone Digital Anthropometrics: Evaluation of an Existing Application and the Development of New Prediction Models

INTRODUCTION/BACKGROUND

Bone mineral content (BMC) is most commonly evaluated using dual-energy X-ray absorptiometry (DXA), but there are several challenges that limit use of DXA during routine care. Breakthroughs in digital imaging now allow smartphone applications to automate important anthropometrics that can predict several body composition components. However, it is unknown whether the anthropometrics automated using smartphone applications can predict DXA-derived BMC.

METHODOLOGY

A total of 214 participants (129 F, 85 M) had BMC measurements collected from an existing proprietary prediction equation, embedded within a smartphone application (MeThreeSixty), and evaluated against DXA. LASSO regression was then used to develop a new BMC prediction equation using the anthropometric estimates produced by the smartphone application in a portion of the participants (n = 174), which was subsequently evaluated against DXA in the remaining sample (n = 40). BMC z-scores were calculated and used to identify the prevalence of low BMC for the existing and newly developed smartphone prediction equations and evaluated against DXA-derived z-scores.

RESULTS

Neither BMC estimates (R2: 0.72; RMSE: 376 g) nor BMC z-scores (R2: 0.55; RMSE: 1.09 SD) produced from the existing propriety prediction equation demonstrated equivalence with DXA in the combined sample. Moreover, the existing prediction equation had a 69.6 % accuracy of identifying low BMC. LASSO regression for the newly developed smartphone prediction model produced the following equation: BMC (g) = -2020.769 + 60.902(Black=1, 0=all other races) - 180.364(Asian=1, 0=all other races) + 24.433(height) + 1.702(weight) + 2.92(shoulder circumference) + 0.258(arm surface area) - 715.29(waist circumference/(BMI2/3 x height1/2)). BMC (R2: 0.91; RMSE: 209 g) and BMC z-scores (R2: 0.85; RMSE: 0.61) produced from the newly developed equation in the testing sample demonstrated equivalence with DXA and had a 92.5 % accuracy of identifying low BMC.

CONCLUSIONS

Smartphone anthropometrics provide accurate and clinically relevant BMC measurements outside of an advanced setting through the use of our newly-developed smartphone prediction model.

Differences in Perceptual and Attitudinal Body Image Between White and African-American Adults Matched for Sex, Age, and Body Composition

PURPOSE

The purpose of this study was to evaluate the differences in perceptual and attitudinal body image between White and African-American males and females matched for sex, age, BMI, and other body composition components using a combination of 3-dimensional mobile digital imaging analysis (DIA) and the Multidimensional Body-Self Relations Questionnaire-Appearance Scale (MBSRQ-AS).

METHODS

One-hundred non-Hispanic White (n=50) and non-Hispanic African-American (n=50) adults (M=34, F=66) matched for sex, age, BMI, and body composition components completed this cross-sectional study. Participants underwent several anthropometric assessments, completed the MBSRQ-AS, and rated their perceived appearance, ideal appearance, and the appearance they believed a partner would find societally attractive using a state of the art mobile 3-dimensional DIA produced using broad developmental populations. Body image distortion was measured as the perceived minus actual appearance, and body image dissatisfaction was defined as the ideal appearance and appearance a partner would find attractive minus the perceived appearance.

RESULTS

Using the DIA, only African-American females demonstrated significant body image distortion (p<0.001); reporting perceived appearances significantly lower their than their actual. Further, AA females demonstrated significantly larger differences between their ideal and perceived appearance (p=0.009), perceived larger bodies as more attractive to a potential partner (p=0.009), and reported higher ratings of appearance evaluation (p=0.001) and body area satisfaction (p=0.011) compared to White females.

CONCLUSIONS

After accounting for all anthropometric determinants of body image, perceptual and attitudinal body image differs between White and African-American adults with differences supporting larger body size acceptance for African-American individuals, particularly African-American females.

Beyond Body Mass Index: Accurate Metabolic Disease-Risk Phenotyping With 3D Smartphone Application

Objective

Smartphone applications (apps) with optical imaging capabilities are transforming the field of physical anthropometry; digital measurements of body size and shape in clinical settings are increasingly feasible. Currently available apps are usually designed around the capture of two-dimensional images that are then transformed with app software to three-dimensional (3D) avatars that can be used for digital anthropometry. The aim of the current study was to compare waist circumference (WC), hip circumference (HC), four other circumferences (right/left upper arm, thigh) and WC/HC evaluated with a novel high-precision 3D smartphone app to ground-truth measurements made with a flexible tape by a trained anthropometrist.

Methods

Forty-four participants aged 20-78 years and body mass index 18.5-48.5 kg/m2 completed digital and manual circumference evaluations and dual-energy X-ray absorptiometry for visceral adipose tissue mass (VAT).

Results

3D-digital and ground-truth tape WC, HC, and WC/HC estimates were highly correlated (R 2s, 0.90-0.97, p < 0.001), mean 3D and tape group means at each site did not differ significantly, mean absolute (± SD) and root-mean square errors were low (e.g., WC, 3.4 ± 2.6 and 4.4 cm), and strong concordance correlations were present (0.90-0.99); bias with Bland-Altman analyses was small but significant (p < 0.001) for WC and WC/HC. Comparable results were observed for the four other circumferences. VAT was equally well-correlated with 3D and tape WC measurements (R 2s 0.70, 0.69, both p < 0.001); comparable tape-3D VAT-WC/HC associations were also observed in males (R 2s, 0.85, 0.73, both p < 0.001) and females (R 2s, 0.43, p < 0.01; 0.73, p < 0.001).

Conclusions

Digital anthropometry, with accessible technology such as the evaluated novel 3D app, has reached a sufficiently developed stage to go beyond body mass index for phenotyping patient's metabolic disease risks.

Accurate prediction of three-dimensional humanoid avatars for anthropometric modeling

OBJECTIVE

To evaluate the hypothesis that anthropometric dimensions derived from a person's manifold-regression predicted three-dimensional (3D) humanoid avatar are accurate when compared to their actual circumference, volume, and surface area measurements acquired with a ground-truth 3D optical imaging method. Avatars predicted using this approach, if accurate with respect to anthropometric dimensions, can serve multiple purposes including patient body composition analysis and metabolic disease risk stratification in clinical settings.

METHODS

Manifold regression 3D avatar prediction equations were developed on a sample of 570 adults who completed 3D optical scans, dual-energy X-ray absorptiometry (DXA), and bioimpedance analysis (BIA) evaluations. A new prospective sample of 84 adults had ground-truth measurements of 6 body circumferences, 7 volumes, and 7 surface areas with a 20-camera 3D reference scanner. 3D humanoid avatars were generated on these participants with manifold regression including age, weight, height, DXA %fat, and BIA impedances as potential predictor variables. Ground-truth and predicted avatar anthropometric dimensions were quantified with the same software.

RESULTS

Following exploratory studies, one manifold prediction model was moved forward for presentation that included age, weight, height, and %fat as covariates. Predicted and ground-truth avatars had similar visual appearances; correlations between predicted and ground-truth anthropometric estimates were all high (R2s, 0.75-0.99; all p < 0.001) with non-significant mean differences except for arm circumferences (%Δ ~ 5%; p < 0.05). Concordance correlation coefficients ranged from 0.80-0.99 and small but significant bias (p < 0.05-0.01) was present with Bland-Altman plots in 13 of 20 total anthropometric measurements. The mean waist to hip circumference ratio predicted by manifold regression was non-significantly different from ground-truth scanner measurements.

CONCLUSIONS

3D avatars predicted from demographic, physical, and other accessible characteristics can produce body representations with accurate anthropometric dimensions without a 3D scanner. Combining manifold regression algorithms into established body composition methods such as DXA, BIA, and other accessible methods provides new research and clinical opportunities.

Core Components of Cardiac Rehabilitation Programs: 2024 Update: A Scientific Statement From the American Heart Association and the American Association of Cardiovascular and Pulmonary Rehabilitation

The science of cardiac rehabilitation and the secondary prevention of cardiovascular disease has progressed substantially since the most recent American Heart Association and American Association of Cardiovascular and Pulmonary Rehabilitation update on the core components of cardiac rehabilitation and secondary prevention programs was published in 2007. In addition, the advent of new care models, including virtual and remote delivery of cardiac rehabilitation services, has expanded the ways that cardiac rehabilitation programs can reach patients. In this scientific statement, we update the scientific basis of the core components of patient assessment, nutritional counseling, weight management and body composition, cardiovascular disease and risk factor management, psychosocial management, aerobic exercise training, strength training, and physical activity counseling. In addition, in recognition that high-quality cardiac rehabilitation programs regularly monitor their processes and outcomes and engage in an ongoing process of quality improvement, we introduce a new core component of program quality. High-quality program performance will be essential to improve widely documented low enrollment and adherence rates and reduce health disparities in cardiac rehabilitation access.

Prediction of total and regional body composition from 3D body shape

Accurate assessment of body composition is essential for evaluating the risk of chronic disease. 3D body shape, obtainable using smartphones, correlates strongly with body composition. We present a novel method that fits a 3D body mesh to a dual-energy X-ray absorptiometry (DXA) silhouette (emulating a single photograph) paired with anthropometric traits, and apply it to the multi-phase Fenland study comprising 12,435 adults. Using baseline data, we derive models predicting total and regional body composition metrics from these meshes. In Fenland follow-up data, all metrics were predicted with high correlations (r > 0.86). We also evaluate a smartphone app which reconstructs a 3D mesh from phone images to predict body composition metrics; this analysis also showed strong correlations (r > 0.84) for all metrics. The 3D body shape approach is a valid alternative to medical imaging that could offer accessible health parameters for monitoring the efficacy of lifestyle intervention programmes.

Body composition estimation from mobile phone three-dimensional imaging: evaluation of the USA army one-site method

Within the USA military, monitoring body composition is an essential component of predicting physical performance and establishing soldier readiness. The purpose of this study was to explore mobile phone three-dimensional optical imaging (3DO), a user-friendly technology capable of rapidly obtaining reliable anthropometric measurements and to determine the validity of the new Army one-site body fat equations using 3DO-derived abdominal circumference. Ninety-six participants (51 F, 45 M; age: 23·7 ± 6·5 years; BMI: 24·7 ± 4·1 kg/m2) were assessed using 3DO, dual-energy X-ray absorptiometry (DXA) and a 4-compartment model (4C). The validity of the Army equations using 3DO abdominal circumference was compared with 4C and DXA estimates. Compared with the 4C model, the Army equation overestimated BF% and fat mass (FM) by 1·3 ± 4·8 % and 0·9 ± 3·4 kg, respectively, while fat-free mass (FFM) was underestimated by 0·9 ± 3·4 kg (P < 0·01 for each). Values from DXA and Army equation were similar for BF%, FM and FFM (constant errors between -0·1 and 0·1 units; P ≥ 0·82 for each). In both comparisons, notable proportional bias was observed with slope coefficients of -0·08 to -0·43. Additionally, limits of agreement were 9·5-10·2 % for BF% and 6·8-7·8 kg for FM and FFM. Overall, while group-level performance of the one-site Army equation was acceptable, it exhibited notable proportional bias when compared with laboratory criterion methods and wide limits of agreement, indicating potential concerns when applied to individuals. 3DO may provide opportunities for the development of more advanced, automated digital anthropometric body fat estimation in military settings.

Modeling health risks using neural network ensembles

This study aims to demonstrate that demographics combined with biometrics can be used to predict obesity related chronic disease risk and produce a health risk score that outperforms body mass index (BMI)-the most commonly used biomarker for obesity. We propose training an ensemble of small neural networks to fuse demographics and biometrics inputs. The categorical outputs of the networks are then turned into a multi-dimensional risk map, which associates diverse inputs with stratified, output health risk. Our ensemble model is optimized and validated on disjoint subsets of nationally representative data (N~100,000) from the National Health and Nutrition Examination Survey (NHANES). To broaden applicability of the proposed method, we consider only non-invasive inputs that can be easily measured through modern devices. Our results show that: (a) neural networks can predict individual conditions (e.g., diabetes, hypertension) or the union of multiple (e.g., nine) health conditions; (b) Softmax model outputs can be used to stratify individual- or any-condition risk; (c) ensembles of neural networks improve generalizability; (d) multiple-input models outperform BMI (e.g., 75.1% area under the receiver operator curve for eight-input, any-condition models compared to 64.2% for BMI); (e) small neural networks are as effective as larger ones for the inference tasks considered; the proposed models are small enough that they can be expressed as human-readable equations, and they can be adapted to clinical settings to identify high-risk, undiagnosed populations.

Smartphone three-dimensional imaging for body composition assessment using non-rigid avatar reconstruction

Background

Modern digital anthropometry applications utilize smartphone cameras to rapidly construct three-dimensional humanoid avatars, quantify relevant anthropometric variables, and estimate body composition.

Methods

In the present study, 131 participants ([73 M, 58 F] age 33.7 ± 16.0 y; BMI 27.3 ± 5.9 kg/m2, body fat 29.9 ± 9.9%) had their body composition assessed using dual-energy X-ray absorptiometry (DXA) and a smartphone 3D scanning application using non-rigid avatar reconstruction. The performance of two new body fat % estimation equations was evaluated through reliability and validity statistics, Bland-Altman analysis, and equivalence testing.

Results

In the reliability analysis, the technical error of the measurement and intraclass correlation coefficient were 0.5-0.7% and 0.996-0.997, respectively. Both estimation equations demonstrated statistical equivalence with DXA based on ±2% equivalence regions and strong linear relationships (Pearson's r 0.90; concordance correlation coefficient 0.89-0.90). Across equations, mean absolute error and standard error of the estimate values were ~ 3.5% and ~ 4.2%, respectively. No proportional bias was observed.

Conclusion

While continual advances are likely, smartphone-based 3D scanning may now be suitable for implementation for rapid and accessible body measurement in a variety of applications.

The challenges of assessing adiposity in a clinical setting

To tackle the burden of obesity-induced cardiometabolic disease, the scientific community relies on accurate and reproducible adiposity measurements in the clinic. These measurements guide our understanding of underlying biological mechanisms and clinical outcomes of human trials. However, measuring adiposity and adipose tissue distribution in a clinical setting can be challenging, and different measurement methods pose important limitations. BMI is a simple and high-throughput measurement, but it is associated relatively poorly with clinical outcomes when compared with waist-to-hip and sagittal abdominal diameter measurements. Body composition measurements by dual energy X-ray absorptiometry or MRI scans would be ideal due to their high accuracy, but are not high-throughput. Another important consideration is that adiposity measurements vary between men and women, between adults and children, and between people of different ethnic backgrounds. In this Perspective article, we discuss how these critical challenges can affect our interpretation of research data in the field of obesity and the design and implementation of clinical guidelines.

Prediction of presence and severity of metabolic syndrome using regional body volumes measured by a multisensor white-light 3D scanner and validation using a mobile technology

Aims

To test whether an index based on the combination of demographics and body volumes obtained with a multisensor 3D body volume (3D-BV) scanner and biplane imaging using a mobile application (myBVI®) will reliably predict the severity and presence of metabolic syndrome (MS).

Methods and results

We enrolled 1280 consecutive subjects who completed study protocol measurements, including 3D-BV and myBVI®. Body volumes and demographics were screened using the least absolute shrinkage and selection operator to select features associated with an MS severity score and prevalence. We randomly selected 80% of the subjects to train the models, and performance was assessed in 20% of the remaining observations and externally validated on 133 volunteers who prospectively underwent myBVI® measurements. The mean ± SD age was 43.7 ± 12.2 years, 63.7% were women, body mass index (BMI) was 28.2 ± 6.2 kg/m2, and 30.2% had MS and an MS severity z-score of -0.2 ± 0.9. Features β coefficients equal to zero were removed from the model, and 14 were included in the final model and used to calculate the body volume index (BVI), demonstrating an area under the receiving operating curve (AUC) of 0.83 in the validation set. The myBVI® cohort had a mean age of 33 ± 10.3 years, 61% of whom were women, 10.5% MS, an average MS severity z-score of -0.8, and an AUC of 0.88.

Conclusion

The described BVI model was associated with an increased severity and prevalence of MS compared with BMI and waist-to-hip ratio. Validation of the BVI had excellent performance when using myBVI®. This model could serve as a powerful screening tool for identifying MS.

Digital Biometry as an Obesity Diagnosis Tool: A Review of Current Applications and Future Directions

Obesity is a chronic relapsing disease and a major public health concern due to its high prevalence and associated complications. Paradoxically, several studies have found that obesity might positively impact the prognosis of patients with certain existing chronic diseases, while some individuals with normal BMI may develop obesity-related complications. This phenomenon might be explained by differences in body composition, such as visceral adipose tissue (VAT), total body fat (TBF), and fat-free mass (FFM). Indirect measures of body composition such as body circumferences, skinfold thicknesses, and bioelectrical impedance analysis (BIA) devices are useful clinically and in epidemiological studies but are often difficult to perform, time-consuming, or inaccurate. Biomedical imaging methods, i.e., computerized tomography scanners (CT scan), dual-energy X-ray absorptiometry (DEXA), and magnetic resonance imaging (MRI), provide accurate assessments but are expensive and not readily available. Recent advancements in 3D optical image technology offer an innovative way to assess body circumferences and body composition, though most machines are costly and not widely available. Two-dimensional optical image technology might offer an interesting alternative, but its accuracy needs validation. This review aims to evaluate the efficacy of 2D and 3D automated body scan devices in assessing body circumferences and body composition.

Accurate Prediction of Three-Dimensional Humanoid Avatars for Anthropometric Modeling

Objective

To evaluate the hypothesis that anthropometric dimensions derived from a person's manifold-regression predicted three-dimensional (3D) humanoid avatar are accurate when compared to their actual circumference, volume, and surface area measurements acquired with a ground-truth 3D optical imaging method. Avatars predicted using this approach, if accurate with respect to anthropometric dimensions, can serve multiple purposes including patient metabolic disease risk stratification in clinical settings.

Methods

Manifold regression 3D avatar prediction equations were developed on a sample of 570 adults who completed 3D optical scans, dual-energy X-ray absorptiometry (DXA), and bioimpedance analysis (BIA) evaluations. A new prospective sample of 84 adults had ground-truth measurements of 6 body circumferences, 7 volumes, and 7 surface areas with a 20-camera 3D reference scanner. 3D humanoid avatars were generated on these participants with manifold regression including age, weight, height, DXA %fat, and BIA impedances as potential predictor variables. Ground-truth and predicted avatar anthropometric dimensions were quantified with the same software.

Results

Following exploratory studies, one manifold prediction model was moved forward for presentation that included age, weight, height, and %fat as covariates. Predicted and ground-truth avatars had similar visual appearances; correlations between predicted and ground-truth anthropometric estimates were all high (R2s, 0.75-0.99; all p < 0.001) with non-significant mean differences except for arm circumferences (%D ~ 5%; p < 0.05). Concordance correlation coefficients ranged from 0.80-0.99 and small but significant bias (p < 0.05 - 0.01) was present with Bland-Altman plots in 13 of 20 total anthropometric measurements. The mean waist to hip circumference ratio predicted by manifold regression was non-significantly different from ground-truth scanner measurements.

Conclusions

3D avatars predicted from demographic, physical, and other accessible characteristics can produce body representations with accurate anthropometric dimensions without a 3D scanner. Combining manifold regression algorithms into established body composition methods such as DXA, BIA, and other accessible methods provides new research and clinical opportunities.

Relationship of Fat Mass Ratio, a Biomarker for Lipodystrophy, With Cardiometabolic Traits

Familial partial lipodystrophy (FPLD) is a heterogenous group of syndromes associated with a high prevalence of cardiometabolic diseases. Prior work has proposed DEXA-derived fat mass ratio (FMR), defined as trunk fat percentage divided by leg fat percentage, as a biomarker of FPLD, but this metric has not previously been characterized in large cohort studies. We set out to 1) understand the cardiometabolic burden of individuals with high FMR in up to 40,796 participants in the UK Biobank and 9,408 participants in the Fenland study, 2) characterize the common variant genetic underpinnings of FMR, and 3) build and test a polygenic predictor for FMR. Participants with high FMR were at higher risk for type 2 diabetes (odds ratio [OR] 2.30, P = 3.5 × 10-41) and metabolic dysfunction-associated liver disease or steatohepatitis (OR 2.55, P = 4.9 × 10-7) in UK Biobank and had higher fasting insulin (difference 19.8 pmol/L, P = 5.7 × 10-36) and fasting triglycerides (difference 36.1 mg/dL, P = 2.5 × 10-28) in the Fenland study. Across FMR and its component traits, 61 conditionally independent variant-trait pairs were discovered, including 13 newly identified pairs. A polygenic score for FMR was associated with an increased risk of cardiometabolic diseases. This work establishes the cardiometabolic significance of high FMR, a biomarker for FPLD, in two large cohort studies and may prove useful in increasing diagnosis rates of patients with metabolically unhealthy fat distribution to enable treatment or a preventive therapy.

ARTICLE HIGHLIGHTS

Advantages and Limitations of the Body Mass Index (BMI) to Assess Adult Obesity

Obesity reflects excessive fat deposits. At-risk individuals are guided by healthcare professionals to eat fewer calories and exercise more, often using body mass index (BMI; weight/height2) thresholds for screening and to guide progress and prognosis. By conducting a mini-narrative review of original articles, websites, editorials, commentaries, and guidelines, we sought to place BMI in the context of its appropriate use in population health, clinical screening, and monitoring in clinical care. The review covers studies and publications through 2023, encompassing consensus reviews and relevant literature. Recent consensus reviews suggest that BMI is a valuable tool for population surveys and primary healthcare screening but has limitations in predicting the risk of chronic diseases and assessing excess fat. BMI can guide nutritional and exercise counseling, even if it is inadequate for reliable individual risk prediction. BMI cut-offs must be reconsidered in populations of varying body build, age, and/or ethnicity. Since BMI-diagnosed overweight persons are sometimes physically and physiologically fit by other indicators, persons who are overweight on BMI should be more fully evaluated, diagnosed, and monitored with combined anthropometric and performance metrics to better clarify risks. The use of combined anthropometric and performance metrics involves integrating measurements of body composition with assessments of physical function and fitness to provide a more comprehensive evaluation of an individual's health and fitness status. Eligibility for bariatric surgery or semaglutide satiety/appetite-reduction medications should not be determined by BMI alone. Awareness of the advantages and limitations of using BMI as a tool to assess adult obesity can maximize its appropriate use in the context of population health and in rapid clinical screening and evaluation.

Emergence of imaging technology beyond the clinical setting: Utilization of mobile health tools for at-home testing

Body composition assessment plays a pivotal role in understanding health, disease risk, and treatment efficacy. This narrative review explores two primary aspects: imaging techniques, namely ultrasound (US) and dual-energy x-ray absorptiometry (DXA), and the emergence of artificial intelligence (AI) and mobile health apps in telehealth for body composition. Although US is valuable for assessing subcutaneous fat and muscle thickness, DXA accurately quantifies bone mineral content, fat mass, and lean mass. Despite their effectiveness, accessibility and cost remain barriers to widespread adoption. The integration of AI-powered image analysis may help explain tissue differentiation, whereas mobile health apps offer real-time metabolic monitoring and personalized feedback. New apps such as MeThreeSixty and Made Health and Fitness offer the advantages of clinic-based imaging techniques from the comfort of home. These innovations hold the potential for individualizing strategies and interventions, optimizing clinical outcomes, and empowering informed decision-making for both healthcare professionals and patients/clients. Navigating the intricacies of these emerging tools, critically assessing their validity and reliability, and ensuring inclusivity across diverse populations and conditions will be crucial in harnessing their full potential. By integrating advancements in body composition assessment, healthcare can move beyond the limitations of traditional methods and deliver truly personalized, data-driven care to optimize well-being.

Validation of a Novel Perceptual Body Image Assessment Method Using Mobile Digital Imaging Analysis: A Cross-Sectional Multicenter Evaluation in a Multiethnic Sample

Given that mobile digital imaging analyses (DIA) are equipped to automate body composition and subsequently alter one's appearance at a given objective body fat percent (BF%), the purpose of this study was to validate the use of this tool for assessments of body image. Participants (f = 134, m = 89) from two separate centers underwent body composition scans using a mobile DIA and completed the Multidimensional Body Self-Relations Questionnaire-Appearance Scale (MBSRQ-AS). Using a DIA-generated avatar, participants altered their figure so that it represented their perceived body, ideal body, and what a partner would find attractive. Distortion was calculated as perceived minus actual BF% and dissatisfaction was calculated as either ideal or partner minus perceived BF%. The total sample and females (p < 0.050), but not males, had significantly lower perceived BF% compared to their actual. Ideal and partner BF% was significantly lower than the perceived BF% for all groups (all p < 0.050). Ideal and partner BF% mean differences (MD) from perceived were positively associated with appearance evaluation (AE) and body area satisfaction (BAS) and negatively associated with overweight preoccupation and self-classified weight for the total sample (all p < 0.050). PerceivedMD demonstrated negative associations with AE and BAS (p < 0.050), but only for females when separated by sex. Perceptual body image measured by DIA is significantly associated with attitudinal body image and may allow practitioners to better quantify this growing issue.

Advances in digital anthropometric body composition assessment: neural network algorithm prediction of appendicular lean mass

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-mean square errors (X ± SD, -0.17 ± 1.64 kg and 1.28 ± 1.04 kg; 1.64). These observations demonstrate the application of NN body composition prediction algorithms to rapidly emerging large and complex digital anthropometric datasets. Clinical Trial Registration: NCT03637855, NCT05217524, NCT03771417, and NCT03706612.

Body Composition in Anti-Obesity Medication Trials-Beyond Scales

This Viewpoint contends that focusing only on weight loss as the primary weight medication end point is an inaccurate measure of medication efficacy for both patients and clinicians.

A review of the application of deep learning in obesity: From early prediction aid to advanced management assistance

BACKGROUND AND AIMS

Obesity is a chronic disease which can cause severe metabolic disorders. Machine learning (ML) techniques, especially deep learning (DL), have proven to be useful in obesity research. However, there is a dearth of systematic reviews of DL applications in obesity. This article aims to summarize the current trend of DL usage in obesity research.

METHODS

An extensive literature review was carried out across multiple databases, including PubMed, Embase, Web of Science, Scopus, and Medline, to collate relevant studies published from January 2018 to September 2023. The focus was on research detailing the application of DL in the context of obesity. We have distilled critical insights pertaining to the utilized learning models, encompassing aspects of their development, principal results, and foundational methodologies.

RESULTS

Our analysis culminated in the synthesis of new knowledge regarding the application of DL in the context of obesity. Finally, 40 research articles were included. The final collection of these research can be divided into three categories: obesity prediction (n = 16); obesity management (n = 13); and body fat estimation (n = 11).

CONCLUSIONS

This is the first review to examine DL applications in obesity. It reveals DL's superiority in obesity prediction over traditional ML methods, showing promise for multi-omics research. DL also innovates in obesity management through diet, fitness, and environmental analyses. Additionally, DL improves body fat estimation, offering affordable and precise monitoring tools. The study is registered with PROSPERO (ID: CRD42023475159).

Smartphone derived anthropometrics: Agreement between a commercially available smartphone application and its parent application intended for use at point-of-care

BACKGROUND & AIMS

Smartphone applications can now automate body composition and anthropometric measurements remotely, prompting applications intended for use at point-of-care to provide commercially available smartphone applications intended for personal use. However, the agreement between such anthropometrics remain unclear.

METHODS

A total of 123 apparently healthy participants (F: 69; M: 54; age: 28.1 ± 11.3; BMI: 26.9 ± 5.9) completed consecutive body composition scans using a 3D smartphone application intended for personal use (MeThreeSixty; MTS) and it stationary counterpart intended for use in practice (Mobile Fit Booth; MFB). Agreement between devices were evaluated using root mean square error (RMSE), Bland-Altman analyses, and linear regression for all measurements, and additional equivalence testing was conducted for all circumference and limb length comparisons.

RESULTS

When evaluated against the MFB, MTS significantly overestimated all measurements other than waist circumference (p = 0.670) using paired t-tests. RMSE was 2.5 % for body fat percentage (BF%), 0.64-3.74 cm for all body circumferences, 0.71-2.3 kg for all lean mass estimates, and 126-659 cm2 and 608-4672 cm3 across all body surface area and body volume estimates, respectively. BF% was the only body composition estimate that did not demonstrate proportional bias (p = 0.221). Circumferences of the chest, shoulder, biceps, forearm, and ankle all demonstrated proportional bias (all coefficients: p < 0.050), but only chest, shoulder, and arm circumferences did not demonstrate equivalence. Arm surface area (p < 0.001) and arm (p = 0.002) and leg volumes (p = 0.004) were the only body surface area and volume estimates to reveal proportional biases.

CONCLUSIONS

These findings demonstrate the agreement between 3D anthropometric applications intended for clinical and personal use, particularly for whole-body composition estimates and clinically meaningful body circumferences. Given the advantages of commercially available remote applications, practitioners and consumers may consider using this method in place of those intended for clinical practice, but should express caution when overestimation is a concern.

Methods for Assessing Health Outcomes Associated with Food Insecurity in the United States College Student Population: A Narrative Review

In the United States, college students experience disproportionate food insecurity (FI) rates compared to the national prevalence. The experience of acute and chronic FI has been associated with negative physical and mental health outcomes in this population. This narrative review aims to summarize the current methodologies for assessing health outcomes associated with the experience of FI in college students in the United States. To date, assessing the health outcomes of FI has predominately consisted of subjective assessments, such as self-reported measures of dietary intake, perceived health status, stress, depression, anxiety, and sleep behaviors. This review, along with the emergence of FI as an international public health concern, establishes the need for novel, innovative, and objective biomarkers to evaluate the short- and long-term impacts of FI on physical and mental health outcomes in college students. The inclusion of objective biomarkers will further elucidate the relationship between FI and a multitude of health outcomes to better inform strategies for reducing the pervasiveness of FI in the United States college student population.

Equations for smartphone prediction of adiposity and appendicular lean mass in youth soccer players

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 showed percent standard error of measurement values < 10% for all anthropometric and body composition measurements, thus indicating high agreement between the measurements obtained for the two avatars. Mobile Fit app overestimated the body fat percentage with respect to DXA (average overestimation of + 3.7% in males and + 4.6% in females), while it underestimated the total lean mass (- 2.6 kg in males and - 2.5 kg in females) and the appendicular lean mass (- 10.5 kg in males and - 5.5 kg in females). Using data of the soccer players, we reparameterized the equations previously proposed to estimate the body fat percentage and the appendicular lean mass and we obtained new equations that can be used in youth athletes for body composition assessment through conventional anthropometrics-based prediction models.

Fluid-Dependent Single-Frequency Bioelectrical Impedance Fat Mass Estimates Compared to Digital Imaging and Dual X-ray Absorptiometry

The need for a practical method for routine determination of body fat has progressed from body mass index (BMI) to bioelectrical impedance analysis (BIA) and smartphone two-dimensional imaging. We determined agreement in fat mass (FM) estimated with 50 kHz BIA and smartphone single lateral standing digital image (SLSDI) compared to dual X-ray absorptiometry (DXA) in 188 healthy adults (69 females and 119 males). BIA underestimated (p < 0.0001) FM, whereas SLSDI FM estimates were not different from DXA values. Based on limited observations that BIA overestimated fat-free mass (FFM) in obese adults, we tested the hypothesis that expansion of the extracellular water (ECW), expressed as ECW to intracellular water (ECW/ICW), results in underestimation of BIA-dependent FM. Using a general criterion of BMI > 25 kg/m2, 54 male rugby players, compared to 40 male non-rugby players, had greater (p < 0.001) BMI and FFM but less (p < 0.001) FM and ECW/ICW. BIA underestimated (p < 0.001) FM in the non-rugby men, but SLSDI and DXA FM estimates were not different in both groups. This finding is consistent with the expansion of ECW in individuals with excess body fat due to increased adipose tissue mass and its water content. Unlike SLSDI, 50 kHz BIA predictions of FM are affected by an increased ECW/ICW associated with greater adipose tissue. These findings demonstrate the validity, practicality, and convenience of smartphone SLSDI to estimate FM, seemingly not influenced by variable hydration states, for healthcare providers in clinical and field settings.

Relevance of body composition in phenotyping the obesities

Obesity is the most extended metabolic alteration worldwide increasing the risk for the development of cardiometabolic alterations such as type 2 diabetes, hypertension, and dyslipidemia. Body mass index (BMI) remains the most frequently used tool for classifying patients with obesity, but it does not accurately reflect body adiposity. In this document we review classical and new classification systems for phenotyping the obesities. Greater accuracy of and accessibility to body composition techniques at the same time as increased knowledge and use of cardiometabolic risk factors is leading to a more refined phenotyping of patients with obesity. It is time to incorporate these advances into routine clinical practice to better diagnose overweight and obesity, and to optimize the treatment of patients living with obesity.

Development and validation of an accurate smartphone application for measuring waist-to-hip circumference ratio

Waist-to-hip circumference ratio (WHR) is now recognized as among the strongest shape biometrics linked with health outcomes, although use of this phenotypic marker remains limited due to the inaccuracies in and inconvenient nature of flexible tape measurements when made in clinical and home settings. Here we report that accurate and reliable WHR estimation in adults is possible with a smartphone application based on novel computer vision algorithms. The developed application runs a convolutional neural network model referred to as MeasureNet that predicts a person's body circumferences and WHR using front, side, and back color images. MeasureNet bridges the gap between measurements conducted by trained professionals in clinical environments, which can be inconvenient, and self-measurements performed by users at home, which can be unreliable. MeasureNet's accuracy and reliability is evaluated using 1200 participants, measured by a trained staff member. The developed smartphone application, which is a part of Amazon Halo, is a major advance in digital anthropometry, filling a long-existing gap in convenient, accurate WHR measurement capabilities.

Accuracy and Precision of 3-dimensional Optical Imaging for Body Composition by Age, BMI, and Ethnicity

BACKGROUND

The obesity epidemic brought a need for accessible methods to monitor body composition, as excess adiposity has been associated with cardiovascular disease, metabolic disorders, and some cancers. Recent 3-dimensional optical (3DO) imaging advancements have provided opportunities for assessing body composition. However, the accuracy and precision of an overall 3DO body composition model in specific subgroups are unknown.

OBJECTIVES

This study aimed to evaluate 3DO's accuracy and precision by subgroups of age, body mass index, and ethnicity.

METHODS

A cross-sectional analysis was performed using data from the Shape Up! Adults study. Each participant received duplicate 3DO and dual-energy X-ray absorptiometry (DXA) scans. 3DO meshes were digitally registered and reposed using Meshcapade. Principal component analysis was performed on 3DO meshes. The resulting principal components estimated DXA whole-body and regional body composition using stepwise forward linear regression with 5-fold cross-validation. Duplicate 3DO and DXA scans were used for test-retest precision. Student's t tests were performed between 3DO and DXA by subgroup to determine significant differences.

RESULTS

Six hundred thirty-four participants (females = 346) had completed the study at the time of the analysis. 3DO total fat mass in the entire sample achieved R2 of 0.94 with root mean squared error (RMSE) of 2.91 kg compared to DXA in females and similarly in males. 3DO total fat mass achieved a % coefficient of variation (RMSE) of 1.76% (0.44 kg), whereas DXA was 0.98% (0.24 kg) in females and similarly in males. There were no mean differences for total fat, fat-free, percent fat, or visceral adipose tissue by age group (P > 0.068). However, there were mean differences for underweight, Asian, and Black females as well as Native Hawaiian or other Pacific Islanders (P < 0.038).

CONCLUSIONS

A single 3DO body composition model produced accurate and precise body composition estimates that can be used on diverse populations. However, adjustments to specific subgroups may be warranted to improve the accuracy in those that had significant differences. This trial was registered at clinicaltrials.gov as NCT03637855 (Shape Up! Adults).

Validity and reliability of a mobile digital imaging analysis trained by a four-compartment model

BACKGROUND

Digital imaging analysis (DIA) estimates collected from mobile applications comprise a novel technique that can collect body composition estimates remotely without the inherent restrictions of other research-grade devices. However, the accuracy of the artificial intelligence used in DIA is reliant on the accuracy of the developmental methods. Few DIA applications are trained by multicompartment models, but this developmental strategy may be most accurate. Thus, the aim of the present study was to assess the precision and agreement of a DIA application with developmental software trained by a four-compartment (4C) model using an actual 4C model as the criterion method.

METHODS

For this cross-sectional study, body composition estimations were collected from 102 participants (63 females, 39 males) using the methods necessary for a rapid 4C model and a DIA application using two different smartphones.

RESULTS

Intraclass correlation coefficients (0.96-0.99; all p < 0.001) and root mean square coefficients of variation (0.5%-3.0%) showed good reliability for body fat percentage, fat mass and fat-free mass. There were no significant mean differences between the 4C model or the DIA estimates for the total sample, by sex, and for non-Hispanic White (n = 61) and Black/African-American (n = 32) participants (all p > 0.050). DIA estimates demonstrated equivalence with the 4C model for all variables but revealed proportional biases that underestimated body fat percentage (both β = -0.25; p < 0.001) and fat mass (both β = -0.07; p < 0.010) at higher degrees of each variable.

CONCLUSIONS

DIA applications trained by a 4C model are reliable and produce body composition estimates equivalent to an actual 4C model.

A Women's health perspective on managing obesity

While the prevalence of obesity in US men and women is nearly equivalent, obesity management in women requires a different approach that considers age and life stage in development including sexual maturation/reproduction, menopause and post-menopause. In this review, the diagnosis and treatment of obesity using lifestyle modification, pharmacotherapy and metabolic and bariatric surgery are discussed from a women's health perspective, with emphasis on management during pregnancy and post-partum.

Smartphone prediction of skeletal muscle mass: model development and validation in adults

BACKGROUND

Skeletal muscle is a large and clinically relevant body component that has been difficult and impractical to quantify outside of specialized facilities. Advances in smartphone technology now provide the opportunity to quantify multiple body surface dimensions such as circumferences, lengths, surface areas, and volumes.

OBJECTIVES

This study aimed to test the hypothesis that anthropometric body measurements acquired with a smartphone application can be used to accurately estimate an adult's level of muscularity.

METHODS

Appendicular lean mass (ALM) measured by DXA served as the reference for muscularity in a sample of 322 adults. Participants also had digital anthropometric dimensions (circumferences, lengths, and regional and total body surface areas and volumes) quantified with a 20-camera 3D imaging system. Least absolute shrinkage and selection operator (LASSO) regression procedures were used to develop the ALM prediction equations in a portion of the sample, and these models were tested in the remainder of the sample. Then, the accuracy of the prediction models was cross-validated in a second independent sample of 53 adults who underwent ALM estimation by DXA and the same digital anthropometric estimates acquired with a smartphone application.

RESULTS

LASSO models included multiple significant demographic and 3D digital anthropometric predictor variables. Evaluation of the models in the testing sample indicated respective RMSEs in women and men of 1.56 kg and 1.53 kg and R2's of 0.74 and 0.90, respectively. Cross-validation of the LASSO models in the smartphone application group yielded RMSEs in women and men of 1.78 kg and 1.50 kg and R2's of 0.79 and 0.95; no significant differences or bias between measured and predicted ALM values were observed.

CONCLUSIONS

Smartphone image capture capabilities combined with device software applications can now provide accurate renditions of the adult muscularity phenotype outside of specialized laboratory facilities. Am J Clin Nutr 2023;x:xx. This trial was registered at clinicaltrials.gov as NCT03637855 (https://clinicaltrials.gov/ct2/show/NCT03637855), NCT05217524 (https://clinicaltrials.gov/ct2/show/NCT05217524), and NCT03771417 (https://clinicaltrials.gov/ct2/show/NCT03771417).

Visual body composition assessment methods: A 4-compartment model comparison of smartphone-based artificial intelligence for body composition estimation in healthy adults

BACKGROUND & AIMS

Visual body composition (VBC) estimates produced from smartphone-based artificial intelligence represent a user-friendly and convenient way to automate body composition remotely and without the inherent geographical and monetary restrictions of other body composition methods. However, there are limited studies that have assessed the reliability and agreement of this method and thus, the aim of this study was to evaluate VBC estimates compared to a 4-compartment (4C) criterion model.

METHODS

A variety of body composition assessments were conducted across 184 healthy adult participants (114 F, 70 M) including dual-energy X-ray absorptiometry and bioimpedance spectroscopy for utilization in the 4C model and automated assessments produced from two smartphone applications (Amazon Halo®, HALO; and myBVI®) using either Apple® or Samsung® phones. Body composition components were compared to a 4C model using equivalence testing, root mean square error (RMSE), and Bland-Altman analysis. Separate analyses by sex and racial/ethnic groups were conducted. Precision metrics were conducted for 183 participants using intraclass correlation coefficients (ICC), root mean squared coefficients of variation (RMS-%CV) and precision error (PE).

RESULTS

Only %fat produced from HALO devices demonstrated equivalence with the 4C model although mean differences for HALO were <±1.0 kg for FM and FFM. RMSEs ranged from 3.9% to 6.2% for %fat and 3.1-5.2 kg for FM and FFM. Proportional bias was apparent for %fat across all VBC applications but varied for FM and FFM. Validity metrics by sex and specific racial/ethnic groups varied across applications. All VBC applications were reliable for %fat, fat mass (FM), and fat-free mass (FFM) with ICCs ≥0.99, RMS-%CV between 0.7% and 4.3%, and PEs between 0.3% and 0.6% for %fat and 0.2-0.5 kg for FM and FFM including assessments between smartphone types.

CONCLUSIONS

Smartphone-based VBC estimates produce reliable body composition estimates but their equivalence with a 4C model varies by the body composition component being estimated and the VBC being employed. VBC estimates produced by HALO appear to have the lowest error, but proportional bias and estimates by sex and race vary across applications.

Digital Single-Image Smartphone Assessment of Total Body Fat and Abdominal Fat Using Machine Learning

Background: Obesity is chronic health problem. Screening for the obesity phenotype is limited by the availability of practical methods. Methods: We determined the reproducibility and accuracy of an automated machine-learning method using smartphone camera-enabled capture and analysis of single, two-dimensional (2D) standing lateral digital images to estimate fat mass (FM) compared to dual X-ray absorptiometry (DXA) in females and males. We also report the first model to predict abdominal FM using 2D digital images. Results: Gender-specific 2D estimates of FM were significantly correlated (p < 0.001) with DXA FM values and not different (p > 0.05). Reproducibility of FM estimates was very high (R2 = 0.99) with high concordance (R2 = 0.99) and low absolute pure error (0.114 to 0.116 kg) and percent error (1.3 and 3%). Bland−Altman plots revealed no proportional bias with limits of agreement of 4.9 to −4.3 kg and 3.9 to −4.9 kg for females and males, respectively. A novel 2D model to estimate abdominal (lumbar 2−5) FM produced high correlations (R2 = 0.99) and concordance (R2 = 0.99) compared to DXA abdominal FM values. Conclusions: A smartphone camera trained with machine learning and automated processing of 2D lateral standing digital images is an objective and valid method to estimate FM and, with proof of concept, to determine abdominal FM. It can facilitate practical identification of the obesity phenotype in adults.

Silhouette images enable estimation of body fat distribution and associated cardiometabolic risk

Inter-individual variation in fat distribution is increasingly recognized as clinically important but is not routinely assessed in clinical practice, in part because medical imaging has not been practical to deploy at scale for this task. Here, we report a deep learning model trained on an individual’s body shape outline—or “silhouette” —that enables accurate estimation of specific fat depots of interest, including visceral (VAT), abdominal subcutaneous (ASAT), and gluteofemoral (GFAT) adipose tissue volumes, and VAT/ASAT ratio. Two-dimensional coronal and sagittal silhouettes are constructed from whole-body magnetic resonance images in 40,032 participants of the UK Biobank and used as inputs for a convolutional neural network to predict each of these quantities. Mean age of the study participants is 65 years and 51% are female. A cross-validated deep learning model trained on silhouettes enables accurate estimation of VAT, ASAT, and GFAT volumes (R²: 0.88, 0.93, and 0.93, respectively), outperforming a comparator model combining anthropometric and bioimpedance measures (ΔR² = 0.05–0.13). Next, we study VAT/ASAT ratio, a nearly body-mass index (BMI)—and waist circumference-independent marker of metabolically unhealthy fat distribution. While the comparator model poorly predicts VAT/ASAT ratio (R²: 0.17–0.26), a silhouette-based model enables significant improvement (R²: 0.50–0.55). Increased silhouette-predicted VAT/ASAT ratio is associated with increased risk of prevalent and incident type 2 diabetes and coronary artery disease independent of BMI and waist circumference. These results demonstrate that body silhouette images can estimate important measures of fat distribution, laying the scientific foundation for scalable population-based assessment.