In two large population-based studies of UK individuals, better adherence to the MIND diet was associated with higher levels of beneficial metabolites, including unsaturated fats and related lipoprotein characteristics, while levels of potentially detrimental metabolites such as glycoprotein acetyls and phenylalanine were lower. The overall metabolic signature of the MIND diet was independently associated with better global and domain-specific cognitive function, and partially mediated the associations between MIND diet and cognitive function. Overall, the MIND diet could potentially improve metabolic health, which later contributes to better cognitive function.
Previous studies have shown the beneficial associations of the MIND diet with cognitive outcomes, including cognitive function9, cognitive decline8, and dementia10. For example, in our previous meta-analysis of eleven cohort studies, the highest adherence to the MIND diet is associated with 17% lower risk of all-cause dementia10. Similarly, in a meta-analysis of eight cohort studies, each 3-point increment in the MIND diet score is associated with 0.11 higher z-score of global cognitive function9. Findings in the current study from two large-scale population-based studies confirmed the previous evidence and supported the potential benefits of the MIND diet to cognitive health, especially memory function. These findings also highlight that metabolic health may play an important role in the observed associations between the MIND diet and cognitive function. Specifically, participants with greater adherence to the MIND diet had favorable metabolomic profiles, such as higher concentrations of DHA, omega-3 fatty acids, and unsaturation degree of fatty acids and lower concentrations of glycoprotein acetyls and phenylalanine. They are also important characteristics of the Mediterranean diet12 and the DASH diet13, on which the MIND diet was based. Moreover, evidence has suggested their roles in cognitive aging. For example, omega-3 fatty acids, especially DHA, is important for central nervous system function35. On the contrary, several systemic inflammation markers (such as glycoprotein acetyls36) and amyloid-formation-related amino acids (phenylalanine37) have been individually associated with higher dementia risk.
Besides the discoveries on the metabolomic profiles of the MIND diet, we explored to which extent the metabolomic features mediated the association between the MIND diet with cognitive function in later life. We constructed a metabolomic signature score of the MIND diet consisting of 28 metabolites using machine learning-based methods and assessed the proportion of mediation in the diet-cognitive association in the WHII. Overall, the metabolomic signature score moderately correlated with the self-reported adherence to the MIND diet (with a correlation coefficient in 0.20-0.40). We found that the metabolomic signature score mediated the associations with global cognitive function by around 31%, and the proportions varied by cognitive domains from 24% to 36%. The findings are relatively consistent across major subgroups of the study participants. Our research thus contributes to the growing body of evidence supporting the potential of lifestyle interventions in mitigating the risk of dementia through metabolic pathways and may serve as a hint for future studies into the targets of dietary intervention for cognitive health. The study findings also have important clinical and public health implications. From a clinical standpoint, our results underscore the importance of considering metabolic health alongside dietary modifications when managing patients at risk of cognitive decline or neurodegenerative disorders. In public health practices, promoting the adoption of dietary patterns, like the MIND diet, that are not only beneficial for metabolic health but also associated with better cognitive outcomes can have a substantial impact on population-level brain health. As the 3-year randomized control trial of the MIND diet among older adults with overweight or obesity did not observe additional beneficial effect for prevention of cognitive decline as compared with a mild weight-loss control diet38, more studies are needed to look into whether specific sub-populations may benefit from the diet.
Although evidence is still accumulating, our study identified certain specific metabolic mediators between MIND diet and cognitive function, which also shed light on several hypotheses on both direct and indirect pathways underlying associations of dietary factors with brain health39. Firstly, the blood-brain barrier serves multiple functions, including the delivery of various nutrients such as lutein40. Another hypothesis is that effects of dietary factors on cognition primarily occur through the reduction of other risk factors, such as obesity, diabetes, and cardiovascular disease15. Our findings indicate that the relation between the MIND diet and cognition is explained to some extent by metabolic health. For example, circulating unsaturated fatty acids appears to be protective against a wide range of diseases41. Conversely, GlyA, indicating chronic inflammation, have been associated with a spectrum of cardiometabolic diseases42, which are also risk factors for dementia. However, a large portion of the association remains unexplained, which is likely a mixed result of other biological pathways and errors in measuring diet, metabolites, and cognitive function. Further studies are needed to clarify and understand the underlying mechanisms.
Despite the large sample size and the long-term follow-up for metabolic and outcome measurements, our study has several limitations that need to be carefully account for in interpretation. First, despite of the temporal sequences of the dietary assessment (basically recalling the past behaviors), metabolomics assays, and the cognitive function measurements, our findings may not fully represent causal relations due to the observational nature. While we have attempted to account for potential confounding factors, residual and unmeasured confounders as well as reverse causality may still have influenced our findings. Therefore, randomized controlled trials are warranted to provide stronger evidence and establish causal relationships. Secondly, the study populations in WHII and UK Biobank were both from the UK and mostly consisted of White individuals. Whether our findings could be generalized to other populations remains to be further examined. Another limitation of our study pertains to the metabolomic assay, which was conducted only once. This single measurement may not fully capture the dynamic nature of an individual's metabolic status over time. Furthermore, the dietary assessments had certain levels of measurement errors, thus diluting the potential associations and masking potential metabolites influenced by the diet. Finally, we used cognitive function as the study outcome rather than ‘hard endpoints’ such as clinical diagnosis of dementia. There warrant future investigations to further elucidate the associations of the metabolic signature of the MIND diet with dementia risk.
In conclusion, the circulating metabolite profile of the MIND diet was associated with better global and domain-specific cognitive function, which partially mediated the observed associations between MIND diet and cognitive function. The study findings support the potentially beneficial role of the MIND diet in cognitive health through metabolic health maintenance. Future investigations are required to further comprehend the underlying mechanisms.