Introduction Little quantitative data are available on the global impact of diet on chronic diseases, including cardiovascular diseases, type 2 diabetes and cancer.

Methods We used a comparative risk assessment framework to develop systematic and comparable methods to establish for selected dietary risk factors the effect sizes of probable or convincing causal diet-disease relationships; the alternative minimum risk exposure distribution; and the exposure distribution. These inputs, together with disease-specific mortality rates, allow computation of the numbers of events attributable to each dietary factor globally and in 21 world regions.

Results Using WHO evidence criteria for convincing / probable causal effects, we identified potential causal diet-disease relationships. Effect sizes and ranges of uncertainty were derived from published or de-novo systematic reviews and meta-analyses of trials or high-quality observational studies. Alternative minimum-risk distributions were identified based on amounts corresponding to lowest disease rates in existing populations. Optimal and alternative definitions for each diet exposure were established based on the data used to quantify the harmful or protective effects. We developed methods for identifying and obtaining comparable data from nationally representative diet surveys throughout the world (see separately submitted Abstract). Multi-level hierarchical models will be developed and will be presented to handle the diversity of diet metrics, units of measure, energy-adjustment techniques, and to impute missing data.

Conclusions These new methods will allow, for the first time, comparable and quantitative assessment of the global impact of specific dietary factors on chronic disease mortality. Such global assessment is not only possible, but imperative for priority setting and policy making.