Perspective: Big Data and Machine Learning Could Help Advance Nutritional Epidemiology

Dietary intake of potassium, vitamin E, and vitamin C emerges as the most significant predictors of cardiovascular disease risk in adults

Prediction models were developed to assess the risk of cardiovascular disease (CVD) based on micronutrient intake, utilizing data from 90,167 UK Biobank participants. Four machine learning models were employed to predict CVD risk, with performance evaluation metrics including area under the receiver operating characteristic curve (AUC), accuracy, recall, specificity, and F1-score. The eXtreme Gradient Boosting (XGBoost) model was utilized to rank the importance of 11 micronutrients in cardiovascular health. Results indicated that vitamin E, calcium, vitamin C, and potassium intake were associated with a reduced risk of CVD. The XGBoost model demonstrated the highest performance with an AUC of 0.952, highlighting potassium, vitamin E, and vitamin C as key predictors of CVD risk. Subgroup analysis revealed a stronger correlation between calcium intake and CVD risk in older adults and those with higher BMI, while vitamin B6 intake showed a link to CVD risk in women. Overall, the XGBoost model emphasized the significance of potassium, vitamin E, and vitamin C intake as primary predictors of CVD risk in adults, with age, sex, and BMI potentially influencing the importance of micronutrient intake in predicting CVD risk.

Advances in methods for characterizing dietary patterns: A scoping review

There is a growing focus on better understanding the complexity of dietary patterns and how they relate to health and other factors. Approaches that have not traditionally been applied to characterize dietary patterns, such as machine learning algorithms and latent class analysis methods, may offer opportunities to measure and characterize dietary patterns in greater depth than previously considered. However, there has not been a formal examination of how this wide range of approaches has been applied to characterize dietary patterns. This scoping review synthesized literature from 2005-2022 applying methods not traditionally used to characterize dietary patterns, referred to as novel methods. MEDLINE, CINAHL, and Scopus were searched using keywords including machine learning, latent class analysis, and least absolute shrinkage and selection operator (LASSO). Of 5274 records identified, 24 met the inclusion criteria. Twelve of 24 articles were published since 2020. Studies were conducted across 17 countries. Nine studies used approaches that have applications in machine learning to identify dietary patterns. Fourteen studies assessed associations between dietary patterns that were characterized using novel methods and health outcomes, including cancer, cardiovascular disease, and asthma. There was wide variation in the methods applied to characterize dietary patterns and in how these methods were described. The extension of reporting guidelines and quality appraisal tools relevant to nutrition research to consider specific features of novel methods may facilitate complete and consistent reporting and enable evidence synthesis to inform policies and programs aimed at supporting healthy dietary patterns.

Predicting daily recovery during long-term endurance training using machine learning analysis

PURPOSE

The aim of this study was to determine if machine learning models could predict the perceived morning recovery status (AM PRS) and daily change in heart rate variability (HRV change) of endurance athletes based on training, dietary intake, sleep, HRV, and subjective well-being measures.

METHODS

Self-selected nutrition intake, exercise training, sleep habits, HRV, and subjective well-being of 43 endurance athletes ranging from professional to recreationally trained were monitored daily for 12 weeks (3572 days of tracking). Global and individualized models were constructed using machine learning techniques, with the single best algorithm chosen for each model. The model performance was compared with a baseline intercept-only model.

RESULTS

Prediction error (root mean square error [RMSE]) was lower than baseline for the group models (11.8 vs. 14.1 and 0.22 vs. 0.29 for AM PRS and HRV change, respectively). At the individual level, prediction accuracy outperformed the baseline model but varied greatly across participants (RMSE range 5.5-23.6 and 0.05-0.44 for AM PRS and HRV change, respectively).

CONCLUSION

At the group level, daily recovery measures can be predicted based on commonly measured variables, with a small subset of variables providing most of the predictive power. However, at the individual level, the key variables may vary, and additional data may be needed to improve the prediction accuracy.

Artificial intelligence in preventive cardiology

Artificial intelligence (AI) is a field of study that strives to replicate aspects of human intelligence into machines. Preventive cardiology, a subspeciality of cardiovascular (CV) medicine, aims to target and mitigate known risk factors for CV disease (CVD). AI's integration into preventive cardiology may introduce novel treatment interventions and AI-centered clinician assistive tools to reduce the risk of CVD. AI's role in nutrition, weight loss, physical activity, sleep hygiene, blood pressure, dyslipidemia, smoking, alcohol, recreational drugs, and mental health has been investigated. AI has immense potential to be used for the screening, detection, and monitoring of the mentioned risk factors. However, the current literature must be supplemented with future clinical trials to evaluate the capabilities of AI interventions for preventive cardiology. This review discusses present examples, potentials, and limitations of AI's role for the primary and secondary prevention of CVD.

Development of a Machine Learning Model for Classifying Cooking Recipes According to Dietary Styles

To complement classical methods for identifying Japanese, Chinese, and Western dietary styles, this study aimed to develop a machine learning model. This study utilized 604 features from 8183 cooking recipes based on a Japanese recipe site. The data were randomly divided into training, validation, and test sets for each dietary style at a 60:20:20 ratio. Six machine learning models were developed in this study to effectively classify cooking recipes according to dietary styles. The evaluation indicators were above 0.8 for all models in each dietary style. The top ten features were extracted from each model, and the features common to three or more models were employed as the best predictive features. Five well-predicted features were indicated for the following seasonings: soy sauce, miso (fermented soy beans), and mirin (sweet cooking rice wine) in the Japanese diet; oyster sauce and doubanjiang (chili bean sauce) in the Chinese diet; and olive oil in the Western diet. Predictions by broth were indicated in each diet, such as dashi in the Japanese diet, chicken soup in the Chinese diet, and consommé in the Western diet. The prediction model suggested that seasonings and broths could be used to predict dietary styles.

Precise prediction of metabolites patterns using machine learning approaches in distinguishing honey and sugar diets fed to mice

As a natural sweetener produced by honey bees, honey was recognized as being healthier for consumption than table sugar. Our previous study also indicated thatmetaboliteprofiles in mice fed honey and mixedsugardiets aredifferent. However, it is still noteworthy about the batch-to-batch consistency of the metabolic differences between two diet types. Here, the machine learning (ML) algorithms were applied to complement and calibrate HPLC-QTOF/MS-based untargeted metabolomics data. Data were generated from three batches of mice that had the same treatment, which can further mine the metabolite biomarkers. Random Forest and Extra-Trees models could better discriminate between honey and mixed sugar dietary patterns under five-fold cross-validation. Finally, SHapley Additive exPlanations tool identified phosphatidylethanolamine and phosphatidylcholine as reliable metabolic biomarkers to discriminate the honey diet from the mixed sugar diet. This study provides us new ideas for metabolomic analysis of larger data sets.

Triangulating nutrigenomics, metabolomics and microbiomics toward personalized nutrition and healthy living

The unique physiological and genetic characteristics of individuals influence their reactions to different dietary constituents and nutrients. This notion is the foundation of personalized nutrition. The field of nutrigenetics has witnessed significant progress in understanding the impact of genetic variants on macronutrient and micronutrient levels and the individual's responsiveness to dietary intake. These variants hold significant value in facilitating the development of personalized nutritional interventions, thereby enabling the effective translation from conventional dietary guidelines to genome-guided nutrition. Nevertheless, certain obstacles could impede the extensive implementation of individualized nutrition, which is still in its infancy, such as the polygenic nature of nutrition-related pathologies. Consequently, many disorders are susceptible to the collective influence of multiple genes and environmental interplay, wherein each gene exerts a moderate to modest effect. Furthermore, it is widely accepted that diseases emerge because of the intricate interplay between genetic predisposition and external environmental influences. In the context of this specific paradigm, the utilization of advanced "omic" technologies, including epigenomics, transcriptomics, proteomics, metabolomics, and microbiome analysis, in conjunction with comprehensive phenotyping, has the potential to unveil hitherto undisclosed hereditary elements and interactions between genes and the environment. This review aims to provide up-to-date information regarding the fundamentals of personalized nutrition, specifically emphasizing the complex triangulation interplay among microbiota, dietary metabolites, and genes. Furthermore, it highlights the intestinal microbiota's unique makeup, its influence on nutrigenomics, and the tailoring of dietary suggestions. Finally, this article provides an overview of genotyping versus microbiomics, focusing on investigating the potential applications of this knowledge in the context of tailored dietary plans that aim to improve human well-being and overall health.

Multi-nutrient interventions and cognitive ageing: are we barking up the right tree?

As we continue to elucidate the mechanisms underlying age-related brain diseases, the reductionist strategy in nutrition–brain function research has focused on establishing the impact of individual foods. However, the biological processes connecting diet and cognition are complex. Therefore, consideration of a combination of nutritional compounds may be most efficacious. One barrier to establishing the efficacy of multi-nutrient interventions is that the area lacks an established set of evidence-based guidelines for studying their effect on brain health. This review is an output of the International Life Sciences Institute (ILSI) Europe. A multi-disciplinary expert group was assembled with the aim of developing a set of considerations to guide research into the effects of multi-nutrient combinations on brain functions. Consensus recommendations converged on six key issues that should be considered to advance research in this area: (1) establish working mechanisms of the combination and contributions of each individual compound; (2) validate the relevance of the mechanisms for the targeted human condition; (3) include current nutrient status, intake or dietary pattern as inclusion/exclusion criteria in the study design; (4) select a participant population that is clinically and biologically appropriate for all nutritional components of the combination; (5) consider a range of cognitive outcomes; (6) consider the limits of reductionism and the ‘gold standard’ randomised controlled trial. These guiding principles will enhance our understanding of the interactive/complementary activities of dietary components, thereby strengthening the evidence base for recommendations aimed at delaying cognitive decline.

Advancing sustainability in the food and nutrition system: a review of artificial intelligence applications

Promoting sustainability in food and nutrition systems is essential to address the various challenges and trade-offs within the current food system. This imperative is guided by key principles and actionable steps, including enhancing productivity and efficiency, reducing waste, adopting sustainable agricultural practices, improving economic growth and livelihoods, and enhancing resilience at various levels. However, in order to change the current food consumption patterns of the world and move toward sustainable diets, as well as increase productivity in the food production chain, it is necessary to employ the findings and achievements of other sciences. These include the use of artificial intelligence-based technologies. Presented here is a narrative review of possible applications of artificial intelligence in the food production chain that could increase productivity and sustainability. In this study, the most significant roles that artificial intelligence can play in enhancing the productivity and sustainability of the food and nutrition system have been examined in terms of production, processing, distribution, and food consumption. The research revealed that artificial intelligence, a branch of computer science that uses intelligent machines to perform tasks that require human intelligence, can significantly contribute to sustainable food security. Patterns of production, transportation, supply chain, marketing, and food-related applications can all benefit from artificial intelligence. As this review of successful experiences indicates, artificial intelligence, machine learning, and big data are a boon to the goal of sustainable food security as they enable us to achieve our goals more efficiently.

Predicting Cardiovascular Disease Mortality: Leveraging Machine Learning for Comprehensive Assessment of Health and Nutrition Variables

Cardiovascular disease (CVD) is one of the primary causes of death around the world. This study aimed to identify risk factors associated with CVD mortality using data from the National Health and Nutrition Examination Survey (NHANES). We created three models focusing on dietary data, non-diet-related health data, and a combination of both. Machine learning (ML) models, particularly the random forest algorithm, demonstrated robust consistency across health, nutrition, and mixed categories in predicting death from CVD. Shapley additive explanation (SHAP) values showed age, systolic blood pressure, and several other health factors as crucial variables, while fiber, calcium, and vitamin E, among others, were significant nutritional variables. Our research emphasizes the importance of comprehensive health evaluation and dietary intake in predicting CVD mortality. The inclusion of nutrition variables improved the performance of our models, underscoring the utility of dietary intake in ML-based data analysis. Further investigation using large datasets with recurring dietary recalls is necessary to enhance the effectiveness and interpretability of such models.

Non-traditional data sources in obesity research: a systematic review of their use in the study of obesogenic environments

BACKGROUND

The complex nature of obesity increasingly requires a comprehensive approach that includes the role of environmental factors. For understanding contextual determinants, the resources provided by technological advances could become a key factor in obesogenic environment research. This study aims to identify different sources of non-traditional data and their applications, considering the domains of obesogenic environments: physical, sociocultural, political and economic.

METHODS

We conducted a systematic search in PubMed, Scopus and LILACS databases by two independent groups of reviewers, from September to December 2021. We included those studies oriented to adult obesity research using non-traditional data sources, published in the last 5 years in English, Spanish or Portuguese. The overall reporting followed the PRISMA guidelines.

RESULTS

The initial search yielded 1583 articles, 94 articles were kept for full-text screening, and 53 studies met the eligibility criteria and were included. We extracted information about countries of origin, study design, observation units, obesity-related outcomes, environment variables, and non-traditional data sources used. Our results revealed that most of the studies originated from high-income countries (86.54%) and used geospatial data within a GIS (76.67%), social networks (16.67%), and digital devices (11.66%) as data sources. Geospatial data were the most utilised data source and mainly contributed to the study of the physical domains of obesogenic environments, followed by social networks providing data to the analysis of the sociocultural domain. A gap in the literature exploring the political domain of environments was also evident.

CONCLUSION

The disparities between countries are noticeable. Geospatial and social network data sources contributed to studying the physical and sociocultural environments, which could be a valuable complement to those traditionally used in obesity research. We propose the use of information available on the Internet, addressed by artificial intelligence-based tools, to increase the knowledge on political and economic dimensions of the obesogenic environment.

Evaluating the prebiotic effect of oligosaccharides on gut microbiome wellness using in vitro fecal fermentation

We previously proposed the Gut Microbiome Wellness Index (GMWI), a predictor of disease presence based on a gut microbiome taxonomic profile. As an application of this index for food science research, we applied GMWI as a quantitative tool for measuring the prebiotic effect of oligosaccharides. Mainly, in an in vitro anaerobic batch fermentation system, fructooligosaccharides (FOS), galactooligosaccharides (GOS), xylooligosaccharides (XOS), inulin (IN), and 2'-fucosyllactose (2FL), were mixed separately with fecal samples obtained from healthy adult volunteers. To find out how 24 h prebiotic fermentation influenced the GMWI values in their respective microbial communities, changes in species-level relative abundances were analyzed in the five prebiotics groups, as well as in two control groups (no substrate addition at 0 h and for 24 h). The GMWI of fecal microbiomes treated with any of the five prebiotics (IN (0.48 ± 0.06) > FOS (0.47 ± 0.03) > XOS (0.33 ± 0.02) > GOS (0.26 ± 0.02) > 2FL (0.16 ± 0.06)) were positive, which indicates an increase of relative abundances of microbial species previously found to be associated with a healthy, disease-free state. In contrast, the GMWI of samples without substrate addition for 24 h (-0.60 ± 0.05) reflected a non-healthy, disease-harboring microbiome state. Compared to the original prebiotic index (PI) and α-diversity metrics, GMWI provides a more data-driven, evidence-based indexing system for evaluating the prebiotic effect of food components. This study demonstrates how GMWI can be applied as a novel PI in dietary intervention studies, with wider implications for designing personalized diets based on their impact on gut microbiome wellness.

Artificial Intelligence in Food Safety: A Decade Review and Bibliometric Analysis

Artificial Intelligence (AI) technologies have been powerful solutions used to improve food yield, quality, and nutrition, increase safety and traceability while decreasing resource consumption, and eliminate food waste. Compared with several qualitative reviews on AI in food safety, we conducted an in-depth quantitative and systematic review based on the Core Collection database of WoS (Web of Science). To discover the historical trajectory and identify future trends, we analysed the literature concerning AI technologies in food safety from 2012 to 2022 by CiteSpace. In this review, we used bibliometric methods to describe the development of AI in food safety, including performance analysis, science mapping, and network analysis by CiteSpace. Among the 1855 selected articles, China and the United States contributed the most literature, and the Chinese Academy of Sciences released the largest number of relevant articles. Among all the journals in this field, PLoS ONE and Computers and Electronics in Agriculture ranked first and second in terms of annual publications and co-citation frequency. The present character, hot spots, and future research trends of AI technologies in food safety research were determined. Furthermore, based on our analyses, we provide researchers, practitioners, and policymakers with the big picture of research on AI in food safety across the whole process, from precision agriculture to precision nutrition, through 28 enlightening articles.

Food composition databases in the era of Big Data: Vegetable oils as a case study

Understanding the population's dietary patterns and their impacts on health requires many different sources of information. The development of reliable food composition databases is a key step in this pursuit. With them, nutrition and health care professionals can provide better public health advice and guide society toward achieving a better and healthier life. Unfortunately, these databases are full of caveats. Focusing on the specific case of vegetable oils, we analyzed the possible obsolescence of the information and the differences or inconsistencies among databases. We show that in many cases, the information is limited, incompletely documented, old or unreliable. More importantly, despite the many efforts carried out in the last decades, there is still much work to be done. As such, institutions should develop long-standing programs that can ensure the quality of the information on what we eat in the long term. In the face of climate change and complex societal challenges in an interconnected world, the full diversity of the food system needs to be recognized and more efforts should be put toward achieving a data-driven food system.

Implementation of Nutrigenetics and Nutrigenomics Research and Training Activities for Developing Precision Nutrition Strategies in Malaysia

Nutritional epidemiological studies show a triple burden of malnutrition with disparate prevalence across the coexisting ethnicities in Malaysia. To tackle malnutrition and related conditions in Malaysia, research in the new and evolving field of nutrigenetics and nutrigenomics is essential. As part of the Gene-Nutrient Interactions (GeNuIne) Collaboration, the Nutrigenetics and Nutrigenomics Research and Training Unit (N2RTU) aims to solve the malnutrition paradox. This review discusses and presents a conceptual framework that shows the pathway to implementing and strengthening precision nutrition strategies in Malaysia. The framework is divided into: (1) Research and (2) Training and Resource Development. The first arm collects data from genetics, genomics, transcriptomics, metabolomics, gut microbiome, and phenotypic and lifestyle factors to conduct nutrigenetic, nutrigenomic, and nutri-epigenetic studies. The second arm is focused on training and resource development to improve the capacity of the stakeholders (academia, healthcare professionals, policymakers, and the food industry) to utilise the findings generated by research in their respective fields. Finally, the N2RTU framework foresees its applications in artificial intelligence and the implementation of precision nutrition through the action of stakeholders.

Personalized nutrition: A review of genotype-based nutritional supplementation

Nutritional disorders have become a major public health issue, requiring increased targeted approaches. Personalized nutrition adapted to individual needs has garnered dramatic attention as an effective way to improve nutritional balance and maintain health. With the rapidly evolving fields of genomics and nutrigenetics, accumulation of genetic variants has been indicated to alter the effects of nutritional supplementation, suggesting its indispensable role in the genotype-based personalized nutrition. Additionally, the metabolism of nutrients, such as lipids, especially omega-3 polyunsaturated fatty acids, glucose, vitamin A, folic acid, vitamin D, iron, and calcium could be effectively improved with related genetic variants. This review focuses on existing literatures linking critical genetic variants to the nutrient and the ways in which these variants influence the outcomes of certain nutritional supplementations. Although further studies are required in this direction, such evidence provides valuable insights for the guidance of appropriate interventions using genetic information, thus paving the way for the smooth transition of conventional generic approach to genotype-based personalized nutrition.

MilkyBase, a database of human milk composition as a function of maternal-, infant- and measurement conditions

This study describes the development of a database, called MilkyBase, of the biochemical composition of human milk. The data were selected, digitized and curated partly by machine-learning, partly manually from publications. The database can be used to find patterns in the milk composition as a function of maternal-, infant- and measurement conditions and as a platform for users to put their own data in the format shown here. The database is an Excel workbook of linked sheets, making it easy to input data by non-computationally minded nutritionists. The hierarchical organisation of the fields makes sure that statistical inference methods can be programmed to analyse the data. Uncertainty quantification and recording dynamic (time-dependent) compositions offer predictive potentials.

Measurement(s)	Concentration of biochemical compounds in human milk or/and derived quantities, like their sums or ratios.
Technology Type(s)	Data mining, by means of Machine Learning and targeted manual literature search within available scientific publications in the internet.
Factor Type(s)	Georgaphical region • Cohort size • Measurement Method • Various characteristics (including history) of mother, child, breast milk and measurement
Sample Characteristic - Organism	Human milk
Sample Characteristic - Environment	Standard birth environment
Sample Characteristic - Location	Various regions of the world

Determining the effective factors in predicting diet adherence using an intelligent model

Adhering to a healthy diet plays an essential role in preventing many nutrition-related diseases, such as obesity, diabetes, high blood pressure, and other cardiovascular diseases. This study aimed to predict adherence to the prescribed diets using a hybrid model of artificial neural networks (ANNs) and the genetic algorithm (GA). In this study, 26 factors affecting diet adherence were modeled using ANN and GA(ANGA). A dataset of 1528 patients, including 1116 females and 412 males, referred to a private clinic was applied. SPSS Ver.25 and MATLAB toolbox 2017 were employed to make the model and analyze the data. The results showed that the accuracy of the proposed ANN and ANGA models for predicting diet adherence was 93.22% and 93.51%, respectively. Also, the Pearson coefficient showed a significant relationship among the factors. The developed model showed the proper performance for predicting adherence to the diet. Moreover, the most effective factors were selected using GA. Some important factors that affect diet adherence include the duration of the marriage, the reason for referring to the clinic, weight, body mass index (BMI), weight satisfaction, lunch and dinner times, and sleep time. Therefore, applying the proposed model can help dietitians identify people who need more support to adhere to the diet.

Comparing human milk macronutrients measured using analyzers based on mid-infrared spectroscopy and ultrasound and the application of machine learning in data fitting

OBJECTIVE

Fat, carbohydrates (mainly lactose) and protein in breast milk all provide indispensable benefits for the growth of newborns. The only source of nutrition in early infancy is breast milk, so the energy of breast milk is also crucial to the growth of infants. Some macronutrients composition in human breast milk varies greatly, which could affect its nutritional fulfillment to preterm infant needs. Therefore, rapid analysis of macronutrients (including lactose, fat and protein) and milk energy in breast milk is of clinical importance. This study compared the macronutrients results of a mid-infrared (MIR) analyzer and an ultrasound-based breast milk analyzer and unified the results by machine learning.

METHODS

This cross-sectional study included breastfeeding mothers aged 22-40 enrolled between November 2019 and February 2021. Breast milk samples (n = 546) were collected from 244 mothers (from Day 1 to Day 1086 postpartum). A MIR milk analyzer (BETTERREN Co., HMIR-05, SH, CHINA) and an ultrasonic milk analyzer (Honɡyanɡ Co,. HMA 3000, Hebei, CHINA) were used to determine the human milk macronutrient composition. A total of 465 samples completed the tests in both analyzers. The results of the ultrasonic method were mathematically converted using machine learning, while the Bland-Altman method was used to determine the limits of agreement (LOA) between the adjusted results of the ultrasonic method and MIR results.

RESULTS

The MIR and ultrasonic milk analyzer results were significantly different. The protein, fat, and energy determined using the MIR method were higher than those determined by the ultrasonic method, while lactose determined by the MIR method were lower (all p < 0.05). The consistency between the measured MIR and the adjusted ultrasound values was evaluated using the Bland-Altman analysis and the scatter diagram was generated to calculate the 95% LOA. After adjustments, 93.96% protein points (436 out of 465), 94.41% fat points (439 out of 465), 95.91% lactose points (446 out of 465) and 94.62% energy points (440 out of 465) were within the LOA range. The 95% LOA of protein, fat, lactose and energy were - 0.6 to 0.6 g/dl, -0.92 to 0.92 g/dl, -0.88 to 0.88 g/dl and - 40.2 to 40.4 kj/dl, respectively and clinically acceptable. The adjusted ultrasonic results were consistent with the MIR results, and LOA results were high (close to 95%).

CONCLUSIONS

While the results of the breast milk rapid analyzers using the two methods varied significantly, they could still be considered comparable after data adjustments using linear regression algorithm in machine learning. Machine learning methods can play a role in data fitting using different analyzers.

Development of machine learning prediction models to explore nutrients predictive of cardiovascular disease using Canadian linked population-based data

Machine learning may improve use of observational data to understand the nutritional epidemiology of cardiovascular disease (CVD) through better modelling of non-linearity, non-additivity, and dietary complexity. Our objective was to develop machine learning prediction models for exploring how nutrients are related to CVD risk and to evaluate their predictive performance. We established a population-based cohort from the Canadian Community Health Survey and measured CVD incidence and mortality from 2004 to 2018 using administrative databases of national hospital discharges and deaths. Predictors included 61 nutrition variables and fourteen socioeconomic, demographic, psychological, and behavioural variables. Conditional inference forest models were interpreted and evaluated by permutation feature importance, accumulated local effects, and predictive discrimination and calibration. A total of 12 130 individuals were included in the study. Use of supplements, caffeine, and alcohol were the most important nutrition variables for prediction of CVD. Supplement use was associated with decreased risk, caffeine was associated with increasing risk, and alcohol had a u-shaped association with risk. The model had an out-of-sample c-statistic of 0.821 (95% confidence interval = 0.801-0.842). Exploratory findings included both known and novel associations and predictive performance was competitive, suggesting that further application of machine learning to nutritional epidemiology may help elucidate risks and improve predictive models. Novelty: Machine learning prediction models were developed for CVD using dietary data. Models were interpreted with interpretable machine learning techniques, revealing diverse associations between diet and CVD. Models achieved comparable or superior predictive performance to existing CVD risk prediction models.

Prospects and Pitfalls of Machine Learning in Nutritional Epidemiology

Nutritional epidemiology employs observational data to discover associations between diet and disease risk. However, existing analytic methods of dietary data are often sub-optimal, with limited incorporation and analysis of the correlations between the studied variables and nonlinear behaviours in the data. Machine learning (ML) is an area of artificial intelligence that has the potential to improve modelling of nonlinear associations and confounding which are found in nutritional data. These opportunities notwithstanding, the applications of ML in nutritional epidemiology must be approached cautiously to safeguard the scientific quality of the results and provide accurate interpretations. Given the complex scenario around ML, judicious application of such tools is necessary to offer nutritional epidemiology a novel analytical resource for dietary measurement and assessment and a tool to model the complexity of dietary intake and its relation to health. This work describes the applications of ML in nutritional epidemiology and provides guidelines to avoid common pitfalls encountered in applying predictive statistical models to nutritional data. Furthermore, it helps unfamiliar readers better assess the significance of their results and provides new possible future directions in the field of ML in nutritional epidemiology.

Exploration of Machine Learning and Statistical Techniques in Development of a Low-Cost Screening Method Featuring the Global Diet Quality Score for Detecting Prediabetes in Rural India

BACKGROUND

The prevalence of type 2 diabetes has increased substantially in India over the past 3 decades. Undiagnosed diabetes presents a public health challenge, especially in rural areas, where access to laboratory testing for diagnosis may not be readily available.

OBJECTIVES

The present work explores the use of several machine learning and statistical methods in the development of a predictive tool to screen for prediabetes using survey data from an FFQ to compute the Global Diet Quality Score (GDQS).

METHODS

The outcome variable prediabetes status (yes/no) used throughout this study was determined based upon a fasting blood glucose measurement ≥100 mg/dL. The algorithms utilized included the generalized linear model (GLM), random forest, least absolute shrinkage and selection operator (LASSO), elastic net (EN), and generalized linear mixed model (GLMM) with family unit as a (cluster) random (intercept) effect to account for intrafamily correlation. Model performance was assessed on held-out test data, and comparisons made with respect to area under the receiver operating characteristic curve (AUC), sensitivity, and specificity.

RESULTS

The GLMM, GLM, LASSO, and random forest modeling techniques each performed quite well (AUCs >0.70) and included the GDQS food groups and age, among other predictors. The fully adjusted GLMM, which included a random intercept for family unit, achieved slightly superior results (AUC of 0.72) in classifying the prediabetes outcome in these cluster-correlated data.

CONCLUSIONS

The models presented in the current work show promise in identifying individuals at risk of developing diabetes, although further studies are necessary to assess other potentially impactful predictors, as well as the consistency and generalizability of model performance. In addition, future studies to examine the utility of the GDQS in screening for other noncommunicable diseases are recommended.

Artificial intelligence in nutrition research: perspectives on current and future applications

Artificial intelligence (AI) is a rapidly evolving area that offers unparalleled opportunities of progress and applications in many healthcare fields. In this review, we provide an overview of the main and latest applications of AI in nutrition research and identify gaps to address to potentialize this emerging field. AI algorithms may help better understand and predict the complex and non-linear interactions between nutrition-related data and health outcomes, particularly when large amounts of data need to be structured and integrated, such as in metabolomics. AI-based approaches, including image recognition, may also improve dietary assessment by maximizing efficiency and addressing systematic and random errors associated with self-reported measurements of dietary intakes. Finally, AI applications can extract, structure and analyze large amounts of data from social media platforms to better understand dietary behaviours and perceptions among the population. In summary, AI-based approaches will likely improve and advance nutrition research as well as help explore new applications. However, further research is needed to identify areas where AI does deliver added value compared with traditional approaches, and other areas where AI is simply not likely to advance the field. Novelty: Artificial intelligence offers unparalleled opportunities of progress and applications in nutrition. There remain gaps to address to potentialize this emerging field.