FoodNetBD: A Bilingual, Disease-Annotated Nutritional Dataset for Bangladesh

FoodNetBD is an openly licensed, machine-readable nutritional dataset of 600 food items commonly consumed across the Bangladeshi diet. Each item is described by 33 nutrient variables per 100 g edible portion, a bilingual English–Bengali (Unicode) name, and a structured Disease_Tags field assigning the item to one of three constraint tiers (safe, restrict, avoid) across twelve diet-sensitive clinical conditions (type 2 diabetes, PCOS, hypertension, hypotension, chronic kidney disease, anemia, gout, irritable bowel syndrome, asthma, hypothyroidism, hyperthyroidism, and celiac disease). Nutrient values were compiled from the Food Composition Table for Bangladesh, the Indian Food Composition Tables 2017, USDA FoodData Central, and supplementary national and regional composition databases, with every row carrying source citation, source-type, data-completeness, and energy-calculation-method fields. The dataset is directly reusable for disease-aware meal-planning and clinical decision-support tools, Bengali-language health applications, algorithmic benchmarking on tabular nutritional data, and epidemiological research on the Bangladeshi diet. File: - FoodNetBD.csv — the primary dataset (600 rows × 56 columns)