BackgroundInterventions to improve the food environment are often implemented at the population level. The health effects of these interventions can be difficult to observe but can be predicted by models. Prediction models require evidence from multiple sources if changes to diet affect sodium, fat, and carbohydrate consumption. We aimed to develop a model to estimate how multidimensional changes in individuals' diets affect metabolic risk factors. MethodsWe developed an evidence review to identify mathematical models describing relationships between calories and weight in adults, and systematic reviews estimating relationships between fats, carbohydrates, and sodium on systolic blood pressure (SBP), cholesterol, and HbA1c from randomised controlled trials. We searched the MEDLINE, Embase, Cochrane Library, and Science Citation Index databases for peer-reviewed systematic reviews published in English between 1946 and May 26, 2020 using keywords for metabolic risks and nutrients . A systems map combining the School for Public Health Research (SPHR) Diabetes Prevention model, background literature, and dietary public health policies was used to develop the review protocol. Data extracted from the review were used to develop a conceptual map to illustrate the causal relationships between dietary measures and metabolic risk factors, including the effects of food substitutions (carbohydrates vs fats). The map was discussed with two senior lecturers in nutrition to identify evidence gaps and additional evidence. The map was programmed into a deterministic simulation in R to estimate how changes in diet affect metabolic risk factors. FindingsWe estimated that reducing calorie intake by 50 kcal per day would reduce population mean weight by 1·3 kg over 12 months. Reducing sodium intake by 1 g per day would decrease population mean SBP by 0·212 mm Hg. Increasing fibre intake by 1 g per day would decrease SBP by 0·200 mm Hg over 12 months, with marginal decreases in weight, HbA1c, and LDL cholesterol. Substituting carbohydrates for polyunsaturated fat (2·5% of energy) would decrease population mean total cholesterol by 0·12 mmol/L over 12 months, with marginal reductions in SBP and weight. InterpretationThis study provides a tool to estimate effects of dietary changes on individual physiological indicators of health. The research will be most beneficial to extrapolate health effects of population-level food interventions, including those leading to multiple, and possibly, competing nutritional effects. FundingNational Institute for Health Research School for Public Health Research (grant reference number PD-SPH-2015).