Interactive Tree Analysis Identifies Dietary Fiber and Magnesium Adequacy as Exploratory Screening Markers for...

Background/Objectives: The POLA index is a comprehensive tool for evaluating the nutrient-dense, immune-supportive, and anti-inflammatory properties of the diet, but its multi-component structure may limit routine use. We aimed to identify simple dietary markers associated with a lower follow-up incidence of COVID-19 or influenza, as well as the anti-inflammatory properties of the diet, and to compare a simplified screening tool with the full POLA index. Methods: This prospective observational study included 146 healthy adults aged 25-45 years from two Polish cohorts examined in 2020 and 2022 (cohort/year adjusted). Habitual diet was assessed using at least 5-day food records, and nutrient adequacy was expressed relative to Polish dietary reference values. Classification and regression tree analyses were used to identify the most informative dietary predictors of the reduction in risk of infection, and logistic regression models were used to evaluate associations after adjustment for sex, diet type, physical activity, marital status, year of cohort and waist-to-height ratio. Results: During follow-up, 39/146 participants (26.7%) reported COVID-19 or influenza. Interactive tree analysis identified dietary fiber in g per kg/m² of BMI ≥ 1, and magnesium adequacy as the key discriminators. In StrongPOLA, participants not meeting the cut-offs of ≥1 g fiber per kg/m² of BMI and ≥130% of the magnesium reference value had a higher incidence of COVID-19 or influenza than those meeting both of those cut-offs (34.9% vs. 2.7%); however, this estimate was large and imprecise, with a wide confidence interval (the adjusted OR = 14.9 (95% CI: 1.89-118.06)), and should, therefore, be interpreted cautiously. In RapidPOLA, the participants not meeting the cut-offs of ≥1 g fiber per kg/m² of BMI and ≥110% of the magnesium reference value (i.e., 352 mg/day for women and 462 mg/day for men) had a higher observed incidence of COVID-19 or influenza than those meeting both of those cut-offs (36.4% vs. 12.1%); the adjusted OR was 3.4 (95% CI: 1.18-8.75). RapidPOLA showed good agreement with the favorable result of the POLA classification (κ = 0.65). Conclusions: Dietary fiber in g per kg/m² of BMI and magnesium adequacy appear to be practical markers of a broader nutrient-dense, immune-supporting, and anti-inflammatory dietary pattern associated with a lower follow-up incidence of COVID-19 or influenza in young adults without comorbidities. RapidPOLA may be useful as a simple screening tool for a nutrient-dense, immune-supportive, and anti-inflammatory (NUTRIDIMAF) diet in young people without obesity and comorbidities, whereas StrongPOLA may serve as a stricter reference profile. The proposed cut-offs require external validation in independent and more diverse cohorts.