DMP4P: A Daily Meal Plan Dataset for Individuals with Chronic Conditions

DMP4P (Daily Meal Plan Dataset for Individuals with Chronic Conditions) is a dedicated open-access dataset designed to provide targeted daily meal plan resources for adults with chronic conditions, constructed using data from 23,152 adult participants of the National Health and Nutrition Examination Survey (NHANES, 2003–2020). The dataset contains 215,135 daily meal plans, each tailored to the individual health profiles of participants with nine prevalent chronic conditions. These meal plans were generated using a systematic three-tiered computational pipeline that integrates direct matching, tensor-based querying, and a genetic algorithm. All nutrient targets underpinning DMP4P are formulated in strict alignment with authoritative dietary guidelines including the World Health Organization (WHO), Dietary Reference Intakes (DRIs), and relevant clinical practice standards, ensuring the scientific rigor of nutritional specifications for each meal plan. The three-tiered pipeline ensures intrinsic coherence across all meal plans: Gold and Silver level plans are directly derived from real-world dietary intake records in NHANES, while Bronze level plans—which constitute the primary component of the dataset—are generated from the NHANES food library under constraints that reflect real dietary patterns, ingredient co-occurrence, and nutritional consistency with Gold/Silver level plans. The dataset is provided in machine-readable CSV and PostgreSQL formats, with each meal plan linked to detailed participant health profile, nutrient intake, and food composition information. DMP4P addresses the lack of large-scale, condition-specific daily meal plan resources for chronic disease research and practice, serving as a foundational data resource for researchers, clinicians, and public health practitioners working on dietary interventions for individuals with chronic conditions. The complete database archived on Figshare (DOI:10.6084/m9.figshare.30153778), with generating code on GitHub, an interactive website, and APIs provided for broad access.NHANES original data is uploaded to NHANES_raw_data folder for replay the data generation process only so it is not mandatory for usage.For data regeneration, and web-based recommendation system, refer to Methodology.pdf in the repository and GitHub: https://github.com/tyrone1979/DMP4P