Search

BACKGROUND: When a lifestyle intervention combines caloric restriction and increased physical activity energy expenditure (PAEE), there are two components of energy balance, energy intake (EI) and physical activity energy expenditure (PAEE), that are routinely misreported and expensive to measure. Energy balance models have successfully predicted EI if PAEE is known. Estimating EI from an energy balance model when PAEE is not known remains an open question. OBJECTIVE: The objective was to evaluate the performance of an energy balance differential equation model to predict EI in an intervention that includes both calorie restriction and increases in PAEE. DESIGN: The Antonetti energy balance model that predicts body weight trajectories during weight loss was solved and inverted to estimate EI during weight loss. Using data from a calorie restriction study that included interventions with and without prescribed PAEE, we tested the validity of the Antonetti weight predictions against measured weight and the Antonetti EI model against measured EI using the intake-balance method at 168 days. We then evaluated the predicted EI from the model against measured EI in a study that prescribed both calorie restriction and increased PAEE. RESULTS: Compared with measured body weight at 168 days, the mean (±SD) model error was 1.30 ± 3.58 kg. Compared with measured EI at 168 days, the mean EI (±SD) model error in the intervention that prescribed calorie restriction and did not prescribe increased PAEE, was −84.9 ± 227.4 kcal/d. In the intervention that prescribed calorie restriction combined with increased PAEE, the mean (±SD) EI model error was −155.70±205.70 kcal/d. CONCLUSION: The validity of the newly developed EI model was supported by experimental observations and can be used to determine EI during weight loss.

Background Nutrition research is relying more on artificial intelligence and machine learning models to understand, diagnose, predict, and explain data. While artificial intelligence and machine learning models provide powerful modeling tools, failure to use careful and well-thought-out modeling processes can lead to misleading conclusions and concerns surrounding ethics and bias. Methods Based on our experience as reviewers and journal editors in nutrition and obesity, we identified the most frequently omitted best practices from statistical modeling and how these same practices extend to machine learning models. We next addressed areas required for implementation of machine learning that are not included in commercial software packages. Results Here, we provide a tutorial on best artificial intelligence and machine learning modeling practices that can reduce potential ethical problems with a checklist and guiding principles to aid nutrition researchers in developing, evaluating, and implementing artificial intelligence and machine learning models in nutrition research. Conclusion The quality of AI/ML modeling in nutrition research requires iterative and tailored processes to mitigate against potential ethical problems or to predict conclusions that are free of bias.

A long-standing question in nutrition and obesity research involves quantifying the relationship between body fat and anthropometry. To date, the mathematical formulation of these relationships has relied on pairing easily obtained anthropometric measurements such as the body mass index (BMI), waist circumference or hip circumference to body fat. Recent advances in 3D body shape imaging technology provides a new opportunity for quickly and accurately obtaining hundreds of anthropometric measurements within seconds, however, there does not yet exist a large diverse database that pairs these measurements to body fat. Herein, we leverage 3D scanned anthropometry obtained from a population of United States Army basic training recruits to derive four sub-populations of homogenous body shape archetypes using a combined principle components and cluster analysis. While the Army database was large and diverse, it did not have body composition measurements. Therefore, these body shape archetypes were paired to an alternate smaller sample of participants from the Pennington Biomedical Research Center in Baton Rouge, LA that were not only similarly imaged by the same 3D scanning machine, but also had concomitant measures of body composition by dual energy X-ray absorptiometry body composition . With this enhanced ability to obtain anthropometry through 3D scanning quickly of large populations, our machine learning approach for pairing body shapes from large datasets to smaller datasets that also contain state of the art body composition measurements can be extended to pair other health outcomes to 3D body shape anthropometry.

Body mass in humans and animals is strongly associated with the rate of heat production as defined by resting energy expenditure (REE). Starting with the ancient Greeks up to the present time philosophers and scientists have endeavored to understand the nature and sources of body heat. Today we recognize that body mass consists of organs and tissues, each of which produces a specified amount of heat at rest. An individual organ’s REE can now be estimated in vivo as the product of its assumed mass-specific metabolic rate and its imaging-derived mass; whole-body REE reflects the sum of organ and tissue metabolic rates. The sizes of organs present and total body mass in adults are governed by two main factors, a person’s stature or height, and their level of adiposity. With greater body size, as represented by adult height independent of adiposity, organs remain stable or increase in mass according to distinct “scaling” patterns. Similarly, with greater relative adiposity organs adaptively accommodate to the increase in imposed mechanical and metabolic loading conditions. Through a detailed analysis of these stature and adiposity effects we show how classical statistical REE prediction models can be mechanistically understood at the anatomic body composition level.

The objective for percent body fat standards in the United States Army Body Composition Program (ABCP) is to ensure soldiers maintain optimal well-being and performance under all conditions. However, conducting large-scale experiments within the United States Army to evaluate the efficacy of the thresholds is challenging. A receiver operating characteristic (ROC) analysis with corresponding area under the curve (AUC) was performed on body mass index (BMI) and waist circumference to determine optimal gender-specific age cohort thresholds that meet ABCP percent body fat standards in the National Health and Nutrition Examination Survey (NHANES) III. A second dataset consisting of a cohort of basic training recruits (N = 20,896 soldiers, 28% female) with BMI and waist circumference measured using a 3D body image scanner was applied to calculate what percent of basic training recruits meet the ABCP percent body fat standards. Regression models to determine the contribution of different circumference sites to the predictions of percent body fat were developed using a database compiled at the New York Obesity Research Center (N = 500). Optimal BMI thresholds ranged from 23.65 kg/m2 (17–21-year-old cohort) to 26.55 kg/m2 (40 and over age cohort) for males and 21.75 to 24.85 kg/m2 for females. The AUC values were between 0.86 and 0.92. The waist circumference thresholds ranged 81.35 to 97.55 cm for males and 77.05 to 89.35 cm for females with AUC values between 0.90 and 0.91. These BMI thresholds were exceeded by 65% of male and 74% of female basic training recruits and waist circumference thresholds were exceeded by 73% of male and 85% of female recruits. The single circumference that contributed most to prediction of body fat was waist circumference in males and mid-thigh circumference in females. The ABCP percent body fat thresholds yield BMI thresholds that are below the United States Army BMI standards, especially in females which suggests the ABCP percent body fat standards may be too restrictive. The United States Army percent body fat standards could instead be matched to existing national health guidelines.

Validated thermodynamic energy balance models that predict weight change are ever more in use today. Delivery of model predictions using web-based applets and/or smart phones has transformed these models into viable clinical tools. Here, we provide the general framework for thermodynamic energy balance model derivation and highlight differences between thermodynamic energy balance models using four representatives. Energy balance models have been used to successfully improve dietary adherence, estimate the magnitude of food waste, and predict dropout from clinical weight loss trials. They are also being used to generate hypotheses in nutrition experiments. Applications of thermodynamic energy balance weight change prediction models range from clinical applications to modify behavior to deriving epidemiological conclusions. Novel future applications involve using these models to design experiments and provide support for treatment recommendations.

Background Body mass index (BMI) represents a normalization of weight to height and is used to classify adiposity. While the capacity of BMI as an adiposity index has been experimentally validated in Caucasians, but there has been little testing Asian populations. Methods To determine whether weight scales to height squared in Asian Indians across the general population and in Asian Indian tribes an allometric analysis on the power law model, W = αHβ, where W is weight (kg) and H is height (m) was performed on cross-sectional weight and height data from India (N = 43,880) collected through the Anthropological Survey of India. The database contained males 18–84 years of age spanning 161 districts of 14 states and including 33 different tribes (N = 5,549). Models were developed that were unadjusted and adjusted for tribe membership. The Korean National Health and Nutrition Examination Survey (KNHANES) was used to compare to height–weight data from the Anthropological Survey of India and to calculate BMI thresholds for obesity status using a receiver operating characteristic. Results The unadjusted power was β = 2.08 (s = 0.02). The power for the general population (non-tribal) was β = 2.11 (s = 0.02). Powers when adjusted for tribe ranged from 1.87 to 2.35 with 24 of the 33 tribes resulting in statistically significant (p 2 (AUC = 0.89) for males 18 kg/m2 (AUC = 0.97) for females. Obesity classification was calculated as BMI = 26 kg/m2 (AUC = 0.81) and 23 kg/m2 (AUC = 0.83) for females. Conclusions Our study confirms that weight scales to height squared in Asian Indian males even after adjusting for tribe membership. We also demonstrate that optimal BMI thresholds are lower in a Korean population in comparison to currently used BMI thresholds. These results support the application of BMI in Asian populations with potentially lower thresholds.

Conclusions regarding bite count rates and body mass index (BMI) in free-living populations have primarily relied on self-report. The objective of this exploratory study was to compare the relationship between BMI and bite counts measured by a portable sensor called the Bite Counter in free-living populations and participants eating in residence. Two previously conducted studies were analyzed for relationships between BMI and sensor evaluated bite count/min, and meal duration. Participants from the first study (N = 77) wore the bite counter in a free-living environment for a continuous period of 14 days. The second study (N = 214) collected bite count/min, meal duration, and total energy intake in participants who consumed one meal in a cafeteria. Linear regression was applied to examine relationships between BMI and bite count/min. There was no significant correlation in the free-living participants average bite counts per second and BMI (R2 = 0.03, p = 0.14) and a significant negative correlation in the cafeteria participants (R 2 = 0.04, p = 0.03) with higher bite count rates observed in lean versus obese participants. There was a significant correlation between average meal duration and BMI in the free-living participants (R 2 = 0.08, p = 0.01). Total energy intake in the cafeteria participants was also significantly correlated to meal duration (R 2 = 0.31, p

Background Experimentally establishing a group’s body weight maintenance energy requirement is an important component of metabolism research. At present, the reference approach for measuring the metabolizable energy intake (MEI) from foods required for body weight maintenance in non-confined subjects is the doubly-labeled water (DLW)–total energy expenditure (TEE) method. In the current study, we evaluated an energy-intake weight balance method as an alternative to DLW that is more flexible and practical to apply in some settings. Methods The hypothesis was tested that MEI from foods observed in a group of subjects maintaining a constant energy intake while keeping their weight within ±1 kg over 10 days is non-significantly different from DLW-measured TEE (TEEDLW). Six non-obese subjects evaluated as part of an earlier study completed the inpatient protocol that included a 3-day initial adjustment period. Results The group body weight coefficient of variation (X ± SD) during the 10-day balance period was 0.38 ± 0.10% and the slope of the regression line for body weight versus protocol day was non-significant at 1.8 g/day (R2, 0.002, p = 0.98). MEI from foods observed during the 10-day balance period (2390 ± 543 kcal/day) was non-significantly different (p = 0.96) from TEE measured by DLW (2373 ± 713 kcal/day); the MEI/TEEDLW ratio was 1.03 ± 0.15 (range 0.87–1.27) and the correlation between MEI from foods and TEEDLW was highly significant (R2, 0.88, p = 0.005). Conclusions A carefully managed 10-day protocol that includes a constant MEI level from foods with weight stability (±1 kg) will provide a group’s body weight maintenance energy requirement similar to that obtained with DLW. This approach opens the possibility of conducting affordable weight balance studies, shorter in duration than those previously reported, that are needed to answer a wide range of questions in clinical nutrition. Trial registration The study is registered at http://www.clinicaltrials.gov (NCT01672632; August 20, 2012).

Background/objectives Remarkable improvements in mathematical methodology combined with knowledge and data on the response of the human body to changes in nutrition, activity and environment have led to a rapid expansion of mathematical models that predict, describe and aggregate conclusions in nutrition. Although mathematical models in nutrition have made significant advances in predictive accuracy and physiological descriptions, these advances have compromised model simplicity, introducing obstacles to their widespread application and contribution to clinical care. The challenge of model complexity is moderated by delivery through well-designed software. Subjects/methods We reviewed several recent and novel web-based mathematical models related to nutrition and describe the successful application of a dynamic mathematical model to patient care implemented through counseling software in a recent weight-loss intervention. To illustrate the power of model transfer through software, we designed a Visual Basic macro within Microsoft Excel to deliver predictions from six well-established and validated resting energy expenditure formulas in children and adults. Results The six resting energy expenditure models that were deployed using the Visual Basic Application developer ranged in technical complexity requiring decision trees, calculation of nonlinear terms or inclusion of multiple covariates. The developed software allows users to select specific models and desired units. After input of individual height, weight, age and sex data through a user form, individuals can effortlessly view predictions. Conclusions Advances in web-based and widely accessible software provide the capacity to deliver more accurate and physiologically realistic nutrition-related models, and ultimately translating model results to patient care.