BACKGROUND: Cardiometabolic risk factors - including diabetes, hypertension, and obesity - have long been linked with adverse health outcomes such as strokes, but more subtle brain changes in regional brain volumes and cortical thickness associated with these risk factors are less understood. Computer models can now be used to estimate brain age based on structural magnetic resonance imaging data, and subtle brain changes related to cardiometabolic risk factors may manifest as an older-appearing brain in prediction models
thus, we sought to investigate the relationship between cardiometabolic risk factors and machine learning-predicted brain age. METHODS: We performed a systematic search of PubMed and Scopus. We used the brain age gap, which represents the difference between one's predicted and chronological age, as an index of brain structural integrity. We calculated the Cohen RESULTS: We identified 185 studies, of which 14 met inclusion criteria. Among the 3 cardiometabolic risk factors, diabetes had the highest effect size (12 study samples
LIMITATIONS: Our study tested effect sizes of only categorically defined cardiometabolic risk factors and is limited by inconsistencies in diabetes classification, a smaller pooled sample in the obesity analysis, and limited age range reporting. CONCLUSION: Our findings show that each of the cardiometabolic risk factors uniquely contributes to brain structure, as captured by brain age. The effect size for diabetes was more than 2 times greater than the independent effects of hypertension and obesity. We therefore highlight diabetes as a primary target for the prevention of brain structural changes that may lead to cognitive decline and dementia.