Abstract
OBJECTIVE: This study aims to explore the role of nutrients with pro-inflammatory or anti-inflammatory potential in the risk of mild cognitive impairment in the elderly. METHODS: We evaluated the intake of 22 nutrients in the baseline survey data of 612 elderly people in northern China. Meanwhile, the least absolute contraction and selection operator (LASSO) regression analysis was used to screen the nutrients with strong correlation with MCI. Bayesian kernel machine regression (BKMR) was devoted to explore the possible associations between various nutrients with different inflammatory potentials and the risk of mild cognitive impairment. RESULTS: A total of 253 people (41.3%) were diagnosed with mild cognition. Ten nutrients are significantly related to the risk of MCI and were screened by a lasso regression model, including 5 pro-inflammatory nutrients (inflammation effect score >0) and 5 anti-inflammatory nutrients (inflammation effect score <0). We incorporated the inflammatory effect scores of 10 nutrients into the BKMR model, and the results showed that the inflammatory effect of 10 nutrients continued to rise with the increase in inflammation scores, proposing that the overall effect is pro-inflammatory. The BKMR analysis results of the pro-inflammatory group and the anti-inflammatory group showed that multiple nutrients in the two groups had a significant combined effect on mild cognitive impairment. We found that by comparing the overall effect of inflammation and the effect of a single group, we found that the inflammation effect of the pro-inflammatory diet and the anti-inflammatory diet had a certain offsetting effect (P < 0.005). CONCLUSION: In the elderly population in northern China, pro-inflammatory diets are associated with an increased risk of mild cognitive impairment. However, these results need to be further evaluated and verified in more prospective studies.