TSGMA: identification of macro associations from global data to build global MA networks

The growing availability of globally harmonized datasets offers unprecedented opportunities to identify population-level risk factors, yet systematic tools for macro-scale association analysis remain scarce. Here, I introduce the concept of “macro association” (MA) and propose a three-tiered framework: Three Stars Global Macro Association-analysis (TSGMA), which integrates correlation, partial correlation adjusted for confounders, and temporal lag analysis to rapidly and robustly rank global associations. Applying TSGMA to dietary and health data from 159 countries or regions, I identify strong links between diet and three cardiometabolic markers. Animal fats group, red meat, and eggs show robust, time-lagged associations with elevated non-HDL cholesterol; sugar, cereals, and poultry meat are associated with increased diabetes prevalence; while starchy roots and pulses consistently exhibit protective associations. TSGMA not only confirms established patterns but also reveals overlooked signals, offering a scalable approach to construct integrative global “MA networks” and enabling hypothesis generation from open-access “macro data”.