A Posteriori Data-Derived Dietary Patterns and Incident Coronary Heart Disease: Making Sense of Inconsistent Findings

Inconsistent findings have been reported from numerous prospective studies for the relations of the “Western” (unhealthy) and “Prudent” (healthy) diet patterns, derived using factor, principle components, or cluster analysis methods, with incident coronary heart disease (CHD). Among contemporary prospective studies, the Prudent diet pattern was inversely related to CHD risk in 7 of 12 studies, while the Western diet pattern positively related to risk in only 3 of 11 studies. To explain these inconsistent findings, we compared the methods and results from these prospective studies conducted in the USA, Europe, and Asia. A Prudent diet pattern was consistently related to 18–65 % lower risk of incident CHD in 7 studies conducted in the USA, Europe, and Asia. In 3 of 4 US studies, but not cohorts in Europe or Asia, the Western diet pattern was related to 37–64 % greater CHD risk. In Asian cohorts, the Western diet pattern was not related to increased CHD risk, which may be partially explained by the overall higher fish intake among Asians. The “a posteriori,” or data driven, approach to diet patterns is based on reported dietary intake and we found the components of each dietary pattern differed by geographic location and diet assessment instrument. We discuss how the non-standardized methods used to discern diet patterns from the dietary data may contribute to discrepant results. Further, the disparate findings may also be explained by differing sample characteristics, follow-up period, and CHD ascertainment. In summary, a posteriori derived Prudent diet pattern was related to cardiovascular health.