key: cord-0324941-cda0rfte authors: Zhang, J.; Cao, X.; Li, X.; Sun, X.; Yang, G.; Hao, M.; Sun, C.; Xia, Y.; Huang, H.; HJ, T. S.; Agogo, G. O.; Wang, L.; Zhang, X.; Gao, X.; Allore, H.; Liu, Z. title: Associations of Midlife Diet Quality with Incident Dementia and Brain Structure: Findings from the UK Biobank Study date: 2022-05-09 journal: nan DOI: 10.1101/2022.05.06.22274696 sha: b124cad1d82dba053a7fcdf1fe36c0355c2b25ed doc_id: 324941 cord_uid: cda0rfte Objective: To investigate the associations of midlife diet quality with incident dementia and brain structure. Design: Population-based prospective study and cross-sectional study. Setting: UK Biobank. Participants: In total, 187,783 participants (mean age 56.8 years, 54.9% women) who completed the 24-hour recall dietary questionnaire were included in the prospective study. A subgroup of 25,380 participants (mean age 55.7 years, 52.9% women) with brain structure data were included in the cross-sectional study. Main exposure and outcome measures: Cox proportional hazards models and linear regression models were used to examine the associations of seven diet quality scores, i.e., hPDI (Healthful Plant-based Diet index), MDS (Mediterranean Diet score), aMED (alternate Mediterranean diet), RFS (Recommended Food Score), DASH (Dietary Approaches to Stop Hypertension), MIND (Mediterranean-DASH Intervention for Neurodegenerative Delay diet) and AHEI-2010 (the Alternative Healthy Eating Index-2010), with incident dementia and brain structure (estimated using magnetic resonance imaging), respectively. Results: During a total follow-up of 1,969,993 person-years, 1,363 (0.73%) participants developed dementia. Higher diet quality scores (except for hPDI) were consistently associated with a lower incidence risk of dementia (all P for trend<0.001). For instance, for RFS, the hazard ratios of the intermediate tertile group and the highest tertile group relative to the lowest tertile group were 0.85 (95% confidence interval [95%CI]=0.75 to 0.97) and 0.70 (95%CI=0.61 to 0.81), respectively. Moreover, higher diet quality scores were significantly associated with larger regional brain volumes including volumes of grey matter (GM) in the parietal and temporal cortex and volumes of the hippocampus and thalamus. For instance, higher RFS was associated with larger volumes of GM in the postcentral gyrus (Beta=16.05, SD=4.08, P<0.001) and the hippocampus (Beta=5.87,SD=1.26, P<0.001). A series of sensitivity analyses confirmed the main results. Conclusion: Greater adherence to MDS, aMED, RFS, DASH, MIND, and AHEI-2010 were individually associated with lower risk of incident dementia and larger brain volumes in specific regions. This study shows a comprehensive picture of the consistent associations of midlife diet quality with dementia risk and brain health, providing mechanistic insights into the role of healthy diet in the prevention of dementia. c n What is already known on this topic 1. Previous prospective studies and meta-analyses suggested significant associations between a few diet quality scores (i.e., MDS, DASH, MIND, and AHEI-2010) and the risk of dementia in different populations; however, the results did not reach agreement. 2. Nutrient intakes or very few diet quality scores have been demonstrated to be associated with brain volumes derived from MRI. There is limited research on the associations of various diet quality scores with the risk of dementia and brain structures in the same population. 1. Greater adherence to MDS, aMED, RFS, DASH, MIND, and AHEI-2010, but not hPDI was individually associated with lower risks of incident dementia. 2. Greater adherence to MDS, aMED, DASH, and AHEI-2010, especially RFS, was individually associated with larger brain volumes in special regions (e.g., parietal and temporal cortex, and hippocampus). 3. This study shows a comprehensive picture of the consistent associations of midlife diet quality with dementia risk and brain health, providing mechanistic insights into the role of healthy diets in the prevention of dementia. Dementia, mainly including Alzheimer's Disease (AD) and vascular dementia, is a pivotal public health issue with immense economic burdens at both individual and national levels. According to the World Health Organization, more than 55 million people live with dementia worldwide in 2021, and this number may increase by 10 million per year 1 . Dementia increases the risk of poor patient-centered outcomes, such as cardiovascular diseases 2 , disturbed emotions 3 , and death [4] [5] [6] , increases risk of depression among caregivers 7 8 , and limits social interactions 9 10 . Thus, proper preventive and management strategies are critical due to the lack of effective treatments of dementia. Healthy diet, as a modifiable lifestyle factor, may help prevent incident dementia and dementia progression. Previous prospective studies and meta-analyses suggested significant associations of the following diet quality scores with the risk of dementia: the AHEI-2010 (the Alternative Healthy Eating Index-2010) 11 , DASH (Dietary Approaches to Stop Hypertension) 12 , aMED (alternate Mediterranean diet) 11 , MDS (Mediterranean Diet Score) [13] [14] [15] , and MIND (Mediterranean-DASH Intervention for Neurodegenerative Delay Diet) 12 . Yet, nonsignificant associations of MDS 16 Pathological changes in brain structures due to onset or progression of neurodegenerative disorders can be detected using advanced diagnostic techniques including Magnetic resonance imaging (MRI) 19 . Previous studies have demonstrated the potential effects of diet quality or nutrient intakes on brain volumes derived from MRI. The Prospective Investigation of the Vasculature in Uppsala Senior's (PIVUS) study showed that lower meat intake was associated with greater total brain volume 20 . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint Columbia Aging Project (WHICAP) showed that higher adherence to the Mediterranean diet was associated with less brain atrophy 21 . However, the association of midlife diet quality as assessed by other diet quality scores, such as hPDI, with total brain volumes, and regional volumes of brain, remain largely unknown. The aim of this study was twofold. First, we investigated the prospective associations of midlife diet quality, as measured by seven existing diet quality scores, namely, hPDI, AHEI-2010, DASH, aMED, MDS, MIND, and RFS (Recommended Food Score) with incident dementia using data from the large, population-based UK Biobank Study with a total follow-up time of 1,969,993 person-years (n=187,783). Second, we examined the associations of the above seven diet quality scores with total brain volumes and brain regional volumes from a subgroup of 25,380 participants with complete, high quality MRI images (Figure 1 ). The UK Biobank is a large-scale and population-based study that recruited more than half a million participants between 40 and 69 years from 2006 to 2010 in the UK 22 . The UK Biobank collected multi-dimension data, including biological samples (e.g., blood, urine, and saliva), physical measurements data (e.g., blood pressure, weight, and height), questionnaires on health, and genetic data. In addition, it also has invited some original participants back to collect the body, brain, and heart imaging from 2014. More details of UK Biobank are available online (http://www.ukbiobank.ac.uk/). This study included two analytic samples (analysis 1 for incident dementia and analysis 2 for brain structure, see Figure 1) . First, as shown in Figure 2 , among 210,971 participants who completed the 24-hour recall dietary questionnaire at least once, 23,188 participants were excluded due to: 1) dementia at baseline (n=75); 2) non-British white (n=9,682, to reduce the ethnic influence on genetic data); 3) lack of . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint APOE genotypes (n=4,169); 4) implausible energy intake (n=4,956, men <800 or >4200 kcal/day, and women <500 or >3500 kcal/day 23 ); and 5) the lack of other covariates (n=4306, e.g., education level, smoking status, alcohol consumption), resulting in 187,783 participants (analytic sample 1) included in the prospective study of associations between diet quality and incident dementia. Second, a subgroup of 25,380 participants (analytic sample 2) with brain MRI data was included in the cross-sectional study of associations between diet quality and brain volumes. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint components that categorized into "healthy plant", "less healthy plant" and "animal" based food. Higher intake of foods from the "healthy plant" food component was given positive scores, while higher intake of foods from the other two food component were given reverse scores. Total scores of these 17 components were summed to obtain the hPDI. MDS is a food-based and nutrient-based nine-item score developed and validated by Trichopoulou et al 25 to reflect adherence to a Mediterranean style diet. It consists of nine components that are scored either 1 or 0 based on sex-specific median intake as cut-offs. Total MDS score ranged from 0 to 9, with higher scores representing better adherence to the Mediterranean diet. aMED score developed by Fung et al 26 was an adaptation of the MDS. It includes nine food components that are common in the Mediterranean-style diet adapted to the US population. The aMED ranged from 0 to 9, with 0 representing the minimum alignment and 9 representing the maximum alignment of the Mediterranean-style diet. The DASH score was created based on foods that were emphasized or discouraged in the DASH trial 28 . It consists of eight food components that are scored 1 to 5 based on quintiles classified according to intake ranking. The modified DASH score developed by previous research within the UK Biobank study 26 consists of 1 0 seven components due to lack of sodium intake information. Total score of DASH ranged from 7 to 35, with higher scores indicating better adherence to the DASH diet. MIND diet is a combination of the Mediterranean diet and the DASH diet that specifically focuses on brain health developed by Dr. Morris et al 12 . It consists of 10 brain healthy food components and five unhealthy food components. Each food component is scored 1 to 5 based on quintiles classified according to intake ranking. Since olive oil consumption information was not collected in the UK Biobank Study, it was not included in this study. The total MIND score was computed by summing all the 14 components, and ranged from 14 to 70. AHEI-2010 includes foods components that have been found to be associated with risks of major chronic diseases 29 . The original AHEI-2010 score consists of 11 food components that each range from 0 to 10 based on consumption of each food portion. This study included a modified AHEI-2010 consists of 8 components instead due to the lack of information on sodium, trans fat, and long-chain fatty acid intakes in the UK Biobank Study. The total AHEI-2010 score ranged from 0 to 80 in this study. The diagnosis of all-cause dementia was obtained by integrating data from two resources: hospital inpatient records and death register data recorded by the International Classification of Diseases (ICD) coding system 22 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 1 1 Brain volumes were extracted from T1-structural brain MRI images, which were provided by an ongoing research that began in 2014, with the aim to acquire high-quality imaging data from 100,000 predominantly healthy participants in UK Biobank 30 . In this study, we used imaging-derived phenotypes (IDPs) according to a previous study, in which the brain MRI imaging processing pipelines were described in detail 31 . In total, 19 IDPs were involved, including volume of grey matter (GM), the volume of white matter (WM), volume of brain (GM+WM), regional grey matter There were several covariates considered in our study, including chronological age, is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint special occasions only, one to three times per month, one to four times per week, and daily or almost daily. RPA was defined as meeting the current global health recommendations for physical activity (150 minutes of moderate activity or 75 minutes of vigorous activity or an equivalent combination), which equated to ≥ 500 MET-minutes/week), or no RPA (<500 MET-minutes/week) 34 . APOE genotype was identified based on two single nucleotide polymorphisms (SNPs) including rs7412 and rs429358 (Supplementary Table S3) , and was classified into E2 (low risk, including ε 2/ε2 and ε 2/ε3), E3 (neutral risk, ε 3/ε3), and E4 (high risk, including ε 3/ε4 and ε 4/ε4) 35 . In particular, ε 2/ε4 was excluded because of its ambiguous genetic risk of dementia 35 . The baseline characteristics of study population were presented by incident dementia. Continuous and categorical variables were described by mean ± standard deviation (SD) and number (percentages) and compared using ANOVA analysis and Chi-square test, respectively. To estimate the association of diet quality (as tertiles) with incident dementia, Cox proportional hazards regressions were conducted (analysis 1). Three models were applied: model 1 adjusted for chronological age and sex; model 2 further adjusted for education level, TDI, BMI, smoking status, alcohol consumption, RPA, time on watching TV, sleep duration, family history of AD, and APOE genotype; and model 3 additionally adjusted for three chronic diseases at the baseline, including CVD, cancer, and diabetes. The hazard ratios (HRs) and their 95% confidence intervals (CIs) were documented, as well as the P for trend. To estimate the associations of diet quality with brain volumes, linear regression models were conducted (analysis 2). Three same models as above were applied and the coefficient and SD were documented. To test the robustness of the associations, several sensitivity analyses were conducted. For associations of diet quality with incident dementia and brain volumes (analysis 1 and 2), four sensitivity analyses were performed: 1) to evaluate whether the associations differed by subgroup, we performed stratified analyses by each . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint covariate except for age and chronic diseases; 2) we excluded participants aged less than 60 years as most dementia occur among the older population; 3) we repeated the analyses among relatively healthier participants without CVD, cancer, and diabetes at baseline to minimize the effects of these diseases; 4) we repeated the analyses with additional adjustment for CVD biomarkers including systolic blood pressure (SBP), diastolic blood pressure (DBP), glycosylated hemoglobin (HbA1c), and high-density lipoprotein cholesterol (HDL-C) (model 4). For associations of diet quality with incident dementia (analysis 1), three additional sensitivity analyses were conducted: 1) we repeated the analyses after setting the end of follow-up at the occurrence of COVID-19 (censor date: Nov 30, 2019) to minimize the effects of COVID-19; 2) we repeated the analyses by type of dementia (i.e., Alzheimer's or vascular dementia); and 3) we repeated the analyses while considering death as a competing risk. All statistical analyses were conducted by SAS 9.4 (SAS Institute Inc., Cary, NC) and R (version 4.0.3). A two-tailed p < 0.05 was identified as statistically significant. Patients and the public were not involved in setting the research question or the outcome measures, and developing plans for recruitment, design, or implementation of the study. No plans exist to involve patients in dissemination. The analytic sample 1 included 187,783 participants, and the mean age was 56. Table 1 ). The . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint 1 4 characteristics of analytic sample 2 are exhibited in Supplementary Table S4 . As shown in As shown in Supplementary Table S5 For subgroup analyses (Supplementary Table S6) , the associations between most . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. For subgroup analyses (Supplementary Table S8) , the patterns of associations between diet quality and brain volumes were maintained, and RFS was still associated with most IDPs across subgroups. For the additional three sensitivity analyses (Supplementary Table S9), we found that the associations between diet quality and brain volumes were maintained when excluding participants with chronic diseases at baseline (Sensitivity analysis 2) or further adjusting for CVD biomarkers (Sensitivity analysis 3). However, after excluding participants less than 60 years old (Sensitivity analysis 1), AHEI-2010 became positively associated with larger volumes of brain and WM after adjusting for all covariates. Based on a large sample of over 180,000 middle-aged and older adults from UK . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint Biobank, we found that six diet quality scores (i.e., MDS, aMED, RFS, DASH, MIND, and AHEI-2010) but not hPDI were significantly associated with the higher risk of incident dementia after an average follow-up of 10.5 years. Furthermore, all of the diet quality scores were positively associated with larger brain volumes in specific regions, particularly the parietal and temporal cortex, as well as the hippocampus in over 25,000 adults. This study shows a comprehensive picture of the consistent associations of diet quality with dementia risk and brain health, providing further evidence and mechanistic insights into the role of diet quality in slowing down the progression of dementia. The beneficial effect of healthy diets on cognitive function has previously been explored 36 37 . However, limited research has estimated the associations between diet quality and the risk of incident dementia 11 12 16 17 38-45 . Most studies focused on MDS 16 17 39-41 , with a few studies considering aMDS 11 45 , DASH 11 12 17 42 , MIND 12 42-44 , and AHEI-2010 11 . Moreover, the results of these studies were inconsistent. Significant associations of MDS 39-41 , DASH 12 , and MIND 12 43 44 with reduced dementia risk were observed in short follow-up time (i.e., 3 to 7 years) studies. However, the associations diminished with longer follow-up time (i.e., 12 to 27 years) 11 17 38 45 . After a median follow-up time (i.e., 10.5 years), we observed significant associations of diet quality with the risk of incident dementia in the current study. The inconsistency across studies may be partly explained by the differences in follow-up time and study population. Dietary habits may be affected by the cognitive decline during the long preclinical period before dementia 46 ; and thus, reverse causation may exist in these short-term follow-up studies. Additionally, previous studies only assessed one diet quality score 16 40 or a few diet quality scores simultaneously 11 12 17 , thus limiting the comparability across diet quality scores. In contrast, we systematically assessed diet quality with various diet quality scores in the same large sample of UK middle-aged and older adults. Although most diet quality scoring systems showed that higher quality diets were protective of incident dementia while hPDI was not associated. A potential explanation might be that when more food items were considered in hPDI . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint (188 kinds of food in this study), the influence of some critical food on dementia was attentued 24 . Moreover, the absence of data on vegetable oil to construct hPDI in UK Biobank may have influenced our results, as vegetable oil is a major source of dietary fatty acids, which could help maintain cognitive functioning 14 . In addition, the results raised an intriguing question of whether the hPDI is suitable for the UK population. More studies are required to assess the true effect of hPDI on dementia. Since the neuropathologic changes of dementia may begin 20 years before dementia diagnosis 47 , exploring the brain structures may help understand potential mechanisms that relate diet quality to dementia risk. Accordingly, we found that all considered diet quality scores were associated with larger volumes of several specific regions (e.g., parietal and temporal cortex, and hippocampus), but not with total volumes of the total brain (i.e., GM and WM), GM, and WM. To date, limited studies have explored the associations of diet quality scores including MDS 21 48-50 , RFS 50 , MIND 44 51 52 , and AHEI-2010 53 with brain structures, mainly among older adults with a relatively small sample size 21 44 48 49 54 , and reported mixed results. With over 25,000 middle-aged and older adults, this study revealed discrepancies in various diet quality scores. RFS presented to be the most sensitive diet quality scores to brain volumes, with significant associations with almost all regional grey matter volumes and volumes of subcortical areas. However, hPDI and MIND were the least sensitive diet quality scores to brain volumes. These results are in line with differences in the magnitude of the associations between diet quality scores and the risk of incident dementia, confirming the differences in various diet quality scores and providing evidence for the necessity of comprehensive evaluation of diet quality. We speculate that a potential mechanism for the benefit of healthy diets in preventing dementia is that adhering to healthy diets may promote brain volumes. Although one or more diet quality scores are significantly associated with almost all regional brain volumes in our study, the most extensively influenced regions include the parietal and temporal cortex, and hippocampus. Similarly, these brain regions are also found to be largely affected by alcohol intake 55 56 , which is an item in the MDS, . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint aMED and AHEI-2010 diets. Volume loss in the temporal lobe and hippocampus have been acknowledged as predictive biomarkers of incident dementia 57 . More recently, the role of the parietal lobe in the development of dementia has attracted attention 58 . The close connectivity between the parietal lobe and other brain areas, as well as a wide range of cognitive function that relied on the parietal lobe, may account for the involvement of the parietal lobe in dementia 58 . Indeed, the development of dementia should be a consequence of changes in multiple brain regions. Nevertheless, due to the cross-sectional design for the second aim of our study, we were unable to draw a temporal association. More longitudinal and experimental studies are needed to verify the underlying mechanisms. Major strengths of this study include the larger sample size, the prospective design with long-term follow-up, available APOE genotypes, seven considered diet quality scores, diagnosis of incident dementia through linked hospital data, and assessment of brain structure in a subgroup, providing a comprehensive exhibition of associations of different diet quality scores with incident dementia and brain volumes in the same population, a valuable data resource for future researchers. Other strength includes the series of sensitivity analyses to reinforce our findings. Nevertheless, this study has a few limitations. First, the incidence of dementia in our study was lower than that in other cohort studies 59 , and this may be explained by the fact that UK Biobank participants are healthier than the general UK population. However, such ascertainment of dementia cases has been demonstrated to be in agreement with primary care records 60 . Second, when constructing these diet quality scores, some food items were unavailable in UK Biobank. For example, total sodium intake, as an important food component for assessing the DASH diet, was not available in the UK Biobank dataset. Thus, the diet quality scores we constructed may not be able to fully reflect the diet quality scores per se. Third, we used a single 24-hour dietary recall to assess diet quality, and thereby, recall bias and a lack of representation of habitual diet quality were inevitable. However, previous studies in the UK Biobank showed good . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint correlations of these scores during a period 24 . Finally, although we estimated the associations between diet quality and brain volumes assessed by MRI, the cross-sectional design hampered the ability to explore underlying mechanisms. Meanwhile, the limited events of incident dementia (42/25380, 0.17%) among participants with MRI data led to the difficulty in estimating potential effects of brain volumes in the association between diet quality and incident dementia. In this large sample of UK Biobank, we demonstrated that greater adherence to MDS, aMED, RFS, DASH, MIND, and AHEI-2010 at midlife were independently associated with lower risk of incident dementia and larger brain volumes in specific regions, regardless of social-economic status and APOE genotypes. Our study presents a full picture of the potential comprehensive beneficial effect of midlife healthy diets on dementia risk and brain health. These findings underscore the importance of midlife diet quality in maintaining brain health and provide mechanistic insight into the role of diet quality in the prevention of dementia. From a public health perspective, preventive and interventional dietary strategies could counter the growing burden of dementia in aging populations, which may be effective even in resource-limited settings. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint This study was based on the Application 61856 in UK Biobank. We acknowledge the UK Biobank participants who provided the sample that made the data available. The UK Biobank study was approved by the North West Multi-Centre Research Ethics Committee (11/NW/0382) and written informed consent was provided by each participant before the study. UK Biobank also has approval from the North West CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint 1 design; data collection, analysis, or interpretation; in the writing of the report; or in the decision to submit the article for publication. The funders had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. None declared. This study was based on the Application 61856 in UK Biobank. The lead author (ZL) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as originally planned have been explained. The results of the research will be disseminated to the public through broadcasts and popular science articles. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. MET-minutes/week, or no regular physical activity (<500 MET-minutes/ week). g APOE genotypes were defined as E2 (protective), E3 (mild), or E4 (risky) by two single-nucleotide polymorphisms, rs429358 and rs7412, the details were described in supplementary table S1. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 9, 2022. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 9, 2022. ; https://doi.org/10.1101/2022.05.06.22274696 doi: medRxiv preprint Dementia: The World Health Organization Cardiovascular risk factors and dementia Behavioral and psychological symptoms of dementia Burden of mortality and morbidity from dementia Dementia prevention, intervention, and care: 2020 report of the Lancet Commission Mortality risk after dementia diagnosis by dementia type and underlying factors: a cohort of 15,209 patients based on the Swedish Dementia Registry Quality of life in dementia: more than just cognition. An analysis of associations with quality of life in dementia Depression Among Caregivers of Patients With Dementia Individual quality of life factors distinguishing low-burden and high-burden caregivers of dementia patients Change in Social Engagement among Incident Caregivers and Controls: Findings from the Caregiving Transitions Study Adherence to Dietary Patterns and Risk of Incident Dementia: Findings from the Atherosclerosis Risk in Communities Study MIND diet associated with reduced incidence of Alzheimer's disease Recent Findings in Alzheimer Disease and Nutrition Focusing on Epigenetics Dietary fatty acids in dementia and predementia syndromes: epidemiological evidence and possible underlying mechanisms Mediterranean diet and preserved brain structural connectivity in older subjects Adherence to a Mediterranean diet, cognitive decline, and risk of dementia The Role of Lifestyle Factors and Sleep Duration for Late-Onset Dementia: A Cohort Study Healthful and Unhealthful Plant-Based Diets and the Risk of Coronary Heart Disease in U.S. Adults Mapping progressive brain structural changes in early Alzheimer's disease and mild cognitive impairment Mediterranean diet habits in older individuals: associations with cognitive functioning and brain volumes Mediterranean diet and brain structure in a multiethnic elderly cohort Association of Lifestyle and Genetic Risk With Incidence of Dementia Genetic analysis of dietary intake identifies new loci and functional links with metabolic traits Healthful plant-based dietary patterns, genetic risk of obesity, and cardiovascular risk in the UK biobank study Modified Mediterranean diet and survival: EPIC-elderly prospective cohort study Diet Quality and the Risk of Impaired Speech Reception Threshold in Noise: The UK Biobank cohort A comparison of three dietary pattern indexes for predicting biomarkers of diet and disease Adherence to a DASH-style diet and risk of coronary heart disease and stroke in women Alternative dietary indices both strongly predict risk of chronic disease Multimodal population brain imaging in the UK Biobank prospective epidemiological study Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank The relationship of parental longevity with the aging brain-results from UK Biobank Combined association of general and central obesity with incidence and mortality of cancers in 22 sites Healthy lifestyle and life expectancy in people with multimorbidity in the UK Biobank: A longitudinal cohort study Longitudinal association of apolipoprotein E and sleep with incident dementia Dietary patterns in middle age: effects on concurrent neurocognition and risk of age-related cognitive decline The Mediterranean, Dietary Approaches to Stop Hypertension (DASH), and Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) Diets Are Associated with Less Cognitive Decline and a Lower Risk of Alzheimer's Disease-A Review Association of Midlife Diet With Subsequent Risk for Dementia Physical activity, diet, and risk of Alzheimer disease Mediterranean diet and risk for Alzheimer's disease Mediterranean diet, inflammatory and metabolic biomarkers, and risk of Alzheimer's disease Dietary Habits and Risk of Early-Onset Dementia in an Italian Case-Control Study MIND diet and risk of dementia in a population based cohort Association of a MIND diet with the risk of dementia and brain structure in a French older population Dietary patterns and cognitive dysfunction in a 12-year follow-up study of 70 year old men Modeling Risk-Factor Trajectories When Measurement Tools Change Sequentially During Follow-up in Cohort Studies: Application to Dietary Habits in Prodromal Dementia Tracking pathophysiological processes in Alzheimer's disease: an updated hypothetical model of dynamic biomarkers Mediterranean diet and preserved brain structural connectivity in older subjects Mediterranean diet and white matter hyperintensity volume in the Northern Manhattan Study Associations of Diet Quality with Midlife Brain Volume: Findings from the UK Biobank Cohort Study Adherence to a MIND-Like Dietary Pattern, Long-Term Exposure to Fine Particulate Matter Air Pollution, and MRI-Based Measures of Brain Volume: The Women's Health Initiative Memory Study-MRI Mind Diet Adherence and Cognitive Performance in the Framingham Heart Study Association of Long-Term Diet Quality with Hippocampal Volume: Longitudinal Cohort Study Mediterranean diet habits in older individuals: associations with cognitive functioning and brain volumes Associations between alcohol consumption and gray and white matter volumes in the UK Biobank Alcohol and adult hippocampal neurogenesis: promiscuous drug, wanton effects Volume loss of the hippocampus and temporal lobe in healthy elderly persons destined to develop dementia Parietal cortex matters in Alzheimer's disease: an overview of structural, functional and metabolic findings