Development of a Personalized Diet Quality Framework for Chronic Disease Prevention

Thesis Abstract

The rising burden of chronic diseases linked to suboptimal diet quality necessitates a move beyond one-size-fits-all nutrition guidance toward personalized frameworks that account for individual biological, behavioral, and environmental determinants of diet. This study aims to develop a Personalized Diet Quality Framework (PDQF) for chronic disease prevention by integrating diet quality assessment, behavioral theory, and precision nutrition concepts to tailor recommendations to individual risk profiles. Specific objectives include (1) to identify core determinants of diet quality across functional, metabolic, and psychosocial domains; (2) to construct a multidimensional Diet Quality Index aligned with personalized risk stratification for cardiometabolic diseases; (3) to validate the framework in diverse populations using a mixed-methods approach; (4) to evaluate the predictive validity of the PDQF against incident cardiometabolic outcomes over 12 months; and (5) to develop practical decision-support tools for clinicians and dietitians. The methodology adopts a sequential explanatory mixed-methods design underpinned by the Capability, Opportunity, Motivation-Behavior (COM-B) model and the Social Ecological Model to ensure comprehensiveness of determinants. A two-phase study will be implemented. Phase I employs a cross-sectional survey of 1,200 adults aged 25–65 years recruited from urban, peri-urban, and rural clinics, with stratified sampling by BMI category and ethnicity. Data collection uses validated instruments to quantify diet quality (modified Alternative Healthy Eating Index with micronutrient and fiber enhancements), metabolic biomarkers (fasting glucose, HbA1c, lipid panel, hs-CRP), and psychosocial variables (food literacy, health beliefs, self-efficacy). Phase II uses longitudinal follow-up of 600 participants for 12 months to capture cardiometabolic events and changes in risk factors, with dietary intake monitored quarterly by 24-hour recalls and mobile food diaries. Instrument validity and reliability are evaluated through confirmatory factor analysis (CFA) for the PDQF items, Cronbach’s alpha for internal consistency, and test–retest reliability for key scales. Data analysis proceeds in three integrated strands. First, exploratory factor analysis identifies latent dimensions of diet quality and determinants; subsequently, multiple regression and structural equation modeling (SEM) examine associations between latent determinants and diet quality, and then the extent to which PDQF scores predict incident cardiometabolic outcomes, adjusting for confounders. Model specification follows guidance from Akaike information criterion (AIC) and Bayesian information criterion (BIC) for comparison, with cross-validation to guard against overfitting. Second, qualitative data from a purposive subsample of 40 participants (from Phase I) will be analyzed using thematic analysis to elucidate perceived barriers and enablers of personalized nutrition adherence, informing refinement of the PDQF. Third, a decision-support prototype, informed by the framework, will be pilot-tested with 10 clinicians for usability and feasibility, employing think-aloud protocols and thematic summation to assess practical applicability. Expected findings include (a) identification of key diet quality determinants at the biological (genetic risk, metabolic status), behavioral (self-efficacy, nutrition literacy), and environmental (food access, cultural norms) levels; (b) development of a validated PDQF that integrates a refined Diet Quality Index with personalized risk strata, capable of predicting 12-month cardiometabolic outcomes with improved discrimination (AUC improvement of at least 0.05 over conventional indices); (c) evidence that personalized recommendations, guided by PDQF, yield greater adherence to improved dietary patterns and favorable changes in risk biomarkers than non-personalized guidance. The study expects to reveal differential effects across demographic subgroups, underscoring equity considerations in personalized nutrition. The contribution to knowledge includes (1) a theoretically grounded, empirically validated framework for personalized diet quality assessment and guidance that merges precision nutrition with behavioral and environmental determinants; (2) a robust, implementable Diet Quality Index embedded within a PDQF that can be operationalized in clinical and public health settings; and (3) practical decision-support tools for dietary counseling, enabling rapid stratification and tailored recommendations. The study concludes that a PDQF enhances predictive accuracy for cardiometabolic risk and supports more effective behavior change strategies; recommendations propose integration of PDQF into electronic health records, routine monitoring of diet quality in primary care, and further longitudinal studies to assess long-term health outcomes and cost-effectiveness across diverse populations.