Incorporating locally available food using linear programming to improve food-based recommendation for preelderly diabetic patients

Background: Dietary adherence is essential for glycemic control in diabetic patients, yet cultural preferences, availability, and affordability often hinder acceptance of recommended foods. Thus, it requires food-based recommendations (FBRs) tailored to local contexts. Aim: This study identified locally available foods to develop FBR using linear programming (LP) in WHO Optifood for preelderly diabetic patients. Methods: A descriptive cross-sectional study was conducted on 174 type 2 diabetic patients aged 50-64 years in Malang, Indonesia, selected through multistage random sampling. Dietary data were collected over seven days using a 24-h recall, a one-day estimated food record, and a five-day food tally to capture habitual intake patterns. Optifood analysis incorporated demographic data, commonly consumed local foods, and nutrient targets based on Perkeni and Indonesian RDA. Dietary adequacy assessment used a modeled scenario using the 65 Recommended Nutrient Intake (RNI) threshold in the worst-case intake analysis. Results: Several nutrients showed dietary inadequacy in the modeled analysis, rather than actual individual deficiencies. Protein, fiber, vitamins A, C, E, B12, iron, and omega-3 could be optimized to exceed 65 of the RNI using local foods such as tempeh, corn rice, moringa leaves, tongkol, kepok banana, and koro beans. However, calcium, vitamin B6, and zinc did not meet the 65 adequacy threshold. Conclusion: Locally available food was successfully incorporated into dietary recommendations, fulfilling eight key nutrients. Nevertheless, vitamin B6, calcium, and zinc remained below the recommended thresholds in worst-case modeling, highlighting the need for additional interventions such as food fortification or dietary diversification. © The Author(s) 2025