Abstract
Background and Aims:
The Adherence to the Dietary Approaches to Stop Hypertension (DASH) diet has been associated with better cognitive function in studies of predominantly White participants; few studies have examined this association in diverse cohorts. Our objective was to examine the association between the DASH diet and cognitive function in the diverse Multi-Ethnic Study of Atherosclerosis (MESA) cohort.
Methods:
Among 4,169 MESA participants, we evaluated prospectively, the association between DASH diet adherence and cognitive function. Participants completed a food frequency questionnaire at baseline (2000–2002) and cognitive function was assessed using the Cognitive Abilities Screening Instrument (CASI), Digit Symbol Coding (DSC), and Digit Span (DS) at Exam 5 in 2010–2012 and Exam 6 (2016–2019). Regression analyses were used to evaluate the association between quintiles of DASH diet adherence with CASI, DSC, and DS performance and decline, adjusting for potential confounders. Effect modification by hypertension, diabetes, race/ethnicity, acculturation, and exercise were evaluated.
Results:
DASH diet adherence was not associated with cognitive performance or decline for any of the measures. There were no differences by racial/ethnic groups, with the exception that Hispanic participants reporting greater DASH diet adherence, performed worse on DS at Exam 5 (p=0.05). Components of the DASH diet were differentially correlated with test performance: increased consumption of nuts/legumes was associated with better performance on the CASI at Exam 5 (p=0.003) and Exam 6 (p=0.007). Increased consumption of whole grains was associated with better DSC performance at Exam 5 (p=0.04) and better DS performance at Exam 6 (p=0.01).
Conclusions:
DASH diet adherence was nominally associated with cognitive function with a suggestion of differences by race/ethnicity. Future work should examine more closely, the relationships between racial and ethnic groups and the impact of diet on cognitive function.
Keywords: Dietary Approaches To Stop Hypertension, Nutrition Assessment, Cognition, Race Factors, Aging
Introduction
Dietary modification as a strategy to prevent hypertension and cardiovascular disease (CVD) is well recognized.1–4 Diet has been also recognized as a modifiable factor that may help protect the brain in later life. The Dietary Approaches to Stop Hypertension (DASH) diet was developed to lower blood pressure and prevent hypertension; and it has been endorsed by the National Heart, Lung, and Blood Institute for the prevention and management of hypertension.5 The importance of reduced blood pressure for cognitive health was demonstrated by the recent SPRINT trial where participants randomized to a systolic blood pressure (SBP) goal of less than 120 mmHg compared to standard treatment (<140 mmHg) had a significantly reduced risk of cognitive impairment.6 The DASH eating pattern helps to reduce blood pressure by emphasizing the intake of several components including: fruits, vegetables, whole grains, poultry, fish, nuts, and low-fat dairy products, and reduced intake of red meat, sweets, sugar-sweetened beverages, and sodium.5,7–9
The relationship between DASH diet adherence and cognitive decline has been a focus of several epidemiological studies.10–14 For example, a longitudinal study in 2013 found that increased adherence to the DASH diet was associated with higher levels of cognitive function and psychomotor speed over an 11-year period in a mostly White, non-Hispanic elderly cohort.14 However, the relationship between DASH diet adherence and cognition has not been adequately examined in multi-ethnic cohorts. This is particularly relevant, as African- and Hispanic-Americans may be at increased risk for chronic cardiovascular and cognitive disorders like hypertension and dementia.15–17 To better understand the potential impact of diet on cognition, it is crucial to examine diet in individuals with a broad range of demographic and clinical characteristics. MESA provides an opportunity to advance previous investigations in this area, given its large and diverse sample size and detailed characterization of participants, including dietary intake. The objectives of this study are to evaluate whether greater adherence to the DASH diet predicts slower cognitive decline, both as a whole diet or by individual components of the diet, and whether this relationship differs in different racial and ethnic groups, as well as according to important health characteristics.
Methods and Materials
Participants
The primary objective of MESA is to track characteristics related to the prevalence and progression of subclinical cardiovascular disease (CVD) to clinical CVD, focusing on ethnic, age, and sex differences in subclinical disease prevalence, progression risk, and rates of CVD.18 MESA recruited 6,814 participants from six communities in the United States in 2000–2002: Baltimore, MD; Chicago, IL; Forsyth County, NC; Los Angeles County, CA; Northern Manhattan and the Bronx, NY; and St. Paul, MN. All participants provided informed consent, and institutional review board approval was received at all sites and reading centers. Participants were between 45 and 84 years old at baseline and self-reported their race/ethnicity as Chinese, Hispanic, non-Hispanic Black, or non-Hispanic White. Individuals with clinically apparent cardiovascular disease (CVD) were excluded from participation in MESA at baseline. In the current analysis, participants with incomplete dietary information at Exam 1 or cognitive data at Exam 5 were excluded.
Dietary Assessment
At the baseline examination (2000–2002), participants completed the modified Block-style 120-item food frequency questionnaire (FFQ).19 Comparative validity of the questionnaire was previously established in demographic groups including non-Hispanic Black, Hispanic, and non-Hispanic White participants.20 Criterion validity of the FFQ was also evaluated in MESA.21 FFQ forms were processed by the DietSys Nutrient Analysis Program, which assigns for each participant, average daily intakes of nutrients based on responses to the items on the FFQ. Energy intake was calculated using weighted recipes from the Nutrition Data System for Research (NDSR) and estimated per 100g of food. Extreme energy intakes, defined as <500 or >5000 kcal/d, were excluded.14,22 Values were then multiplied by individual intake frequency and age-, sex- and portion- size specific gram weights for each food. Data were unavailable on the intake of 14 foods in ~25% of MESA participants, the majority of which were from two field centers. To minimize bias due to this error, imputation was conducted using sequential chained regression, implemented in Stata with the multiple imputation program specifying multinomial or ordinal regression, with one model for all 14 items and 14 intakes (servings/day and serving size). Imputation models accounted for basic demographics and other variables selected on the basis of high correlation with the FFQ or serving size. Corrected data were used in the DASH component scoring paradigm which has been described previously.23,24 Scores were determined according to quintile ranks of 8 food groups. For favorable food groups (fruits, vegetables, whole grains, nuts and legumes, and low-fat dairy products), the highest quintile was assigned 5 points, and the lowest quintile is assigned 1 point. Reverse scoring was applied for unfavorable food groups (red and processed meats, sweetened beverages, and sodium). Quintile ranks were summed to obtain DASH diet component score (expected range 8 – 40), with higher scores indicating greater adherence.
Cognitive Assessment
The cognitive evaluation included the Cognitive Abilities Screening Instrument (CASI)25,26 which measured global cognitive function and includes items assessing attention, concentration, orientation, short- and long-term memory, language, visual construction, list-generating fluency, abstraction, and judgement. The Digit Symbol Coding (DSC) task from the Wechsler Adult Intelligence Scale-III27 measured processing speed. Working memory was assessed using the Digit Span test (DS, forwards and backwards), also from the Wechsler Adult Intelligence Scale-III.27 These cognitive measures were treated as continuous variables (higher scores indicate better performance) and administered at MESA Exams 5 (2010–2012) and 6 (2016–2019) allowing assessment of cognitive decline. Examination of both performance level and cognitive decline is important in this multi-ethnic cohort, as baseline cognitive performance can be biased by sociodemographic variables like race, ethnicity, language, and education, while a number of studies have shown little or no difference by racial and ethnic groups in the examination of cognitive decline.28
Covariates
Demographic variables included age, sex, race/ethnicity, education (less than high school, high school graduate, some college, college graduate, or post graduate), and income (<$20k/yr, $20k-34,999k/yr, $35k/yr-49,999k/yr, $50k/yr-$99,999k/yr, ≥$100k/yr). Acculturation status (high vs low) was defined by whether the participant was born in the US and whether English was the language spoken at home, as previously described.29 Clinical risk factors included APOE ε4 status (SNPs rs7412 and rs429358; Applied Biosystems TaqMan SNP system), body mass index (BMI; weight in kg/height m2), smoking (current vs. not), alcohol intake (never, former, current drinker), Center for Epidemiological Studies Depression (CES-D) scale score (<16 vs ≥16),30–32 total intentional exercise (MET-minutes/week), total energy intake, diabetes categories (normal (fasting glucose <100 mg/dl), pre-diabetes (100–125 mg/dl), and diabetes (≥126mg/dl)), diabetes medication use, antihypertensive medication use, Alzheimer’s medication use at Exam 6, and stroke diagnosis prior to Exam 5 (ICD-9 documented histories). Resting, seated systolic and diastolic blood pressure (DBP) were measured three times using a Dinamap automated oscillometric sphygmomanometer. The last two of the three measures were averaged. Hypertension was defined as SBP over 140 mmHG and/or DBP over 90 mmHG, or use of antihypertensive medications. Fasting lipid levels (triglycerides, high density lipoprotein cholesterol, low density lipoprotein cholesterol, and total cholesterol) were analyzed at a central laboratory. Covariates were selected based on prior previous research and known associations with dietary patterns, cognitive outcomes, or both.10,14,33,34
Statistical Analysis Plan
Differences across DASH quintiles of adherence were evaluated using χ2 tests for categorical characteristics and analysis of variance for continuous characteristics. The association between overall adherence and cognitive function was examined using multivariable linear regression analyses adjusting for covariates (age, sex, race/ethnicity, education level, income level, APOE ε4 carrier status, BMI, smoking, alcohol intake, depressive symptoms, physical activity, energy intake, diabetes, diabetes medications, SPB, DPB, antihypertensive medications, stroke history, HDL, LDL, and total cholesterol). Exam 6 outcomes and cognitive change analyses were also adjusted for Alzheimer’s medication use at Exam 6. Least squares means for average cognitive test scores for DASH adherence quintiles were reported.
We planned a priori to address confounding by indication by examining effect modification by hypertension and diabetes as participants with these conditions may have received nutritional counseling. Similarly, we sought to examine effect modification by race/ethnicity and APOE ε4 carrier status, as it was unclear whether associations would be consistent across racial and ethnic groups or in the presence of APOE ε4. High vs. low acculturation and total intentional exercise were examined in post-hoc analyses. Stratified analyses for diabetes, hypertension, race/ethnicity, APOE ε4 status, acculturation, and intentional exercise were conducted where significant interaction terms are identified (p<.15). Sensitivity analyses were conducted using inverse probability of attrition weighting (IPW) to address potential confounding due to participant dropout or mortality at Exam 5, as predicted by primary outcome and exposure variables, and fully adjusted model covariates. Similar to overall DASH adherence, associations between DASH food group components and cognitive domains were examined using multivariable linear regression analyses, adjusted for the same covariates.
Results
General Characteristics
Of the 6,814 MESA participants, 6,531 completed the baseline FFQ. Participants with extreme energy intake were excluded (n=323), leaving 6,208 participants. A total of 4,599 participants were seen at Exam 5; 4,199 completed the cognitive assessment, and 30 participants were excluded for use of Alzheimer’s medications, which may have modified their performance leaving 4,169 participants (Figure 1). A total of 2,048 completed the Exam 6 cognitive assessment. Compared with those excluded from our analysis, those who were included were ~5 years younger, had lower SBP (4.3mmHg lower), and were less likely to have treated hypertension or diabetes.
Figure 1.
Sample Selection Flowchart
Participant characteristics by DASH quintiles are shown in Table 1. A total of 2,204 women and 1,965 men were included in the final sample. They had a mean age of 60.4 (9.5) years and 42% were non-Hispanic White, 11% were Chinese, 25% were non-Hispanic Black, and 21% were Hispanic. Participants scoring in higher quintiles were generally older, and were more likely to be White, female, and have higher education and income. They had lower BMIs on average, reported more physical activity, and were less likely to have or be treated for diabetes. SBP, hypertension, and antihypertensive medication use were not associated with DASH adherence at baseline.
Table 1.
Baseline Characteristics of n=4,169 Participants by DASH Diet Score Quintiles
| DASH Diet Quintiles | |||||||
|---|---|---|---|---|---|---|---|
| Characteristic | Q1 n=939 |
Q2 n=586 |
Q3 n=954 |
Q4 n=869 |
Q5 n=821 |
Total n=4169 |
p value |
| Higher quintiles indicate greater DASH adherence | |||||||
| DASH Score Range | 10–20 | 21–22 | 23–25 | 26–28 | 29–39 | ||
| Age, years Mean (SE) | 57.6 (8.8) | 58.9 (9.1) | 59.9 (9.4) | 61.9 (9.7) | 63.5 (9.4) | 60.35 (9.5) | <0.001 |
| Sex | <0.001 | ||||||
| Male | 560 (59.6) | 338 (57.7) | 471 (49.4) | 350 (40.3) | 246 (30.0) | 1965 (47.1) | |
| Female | 379 (40.3) | 248 (42.3) | 483 (50.6) | 519 (59.7) | 575 (70.0) | 2204 (52.9) | |
| Race | <0.001 | ||||||
| White | 307 (32.7) | 212 (36.2) | 391 (50.0) | 404 (46.5) | 442 (53.8) | 1756 (42.1) | |
| Chinese American | 119 (12.7) | 89 (15.2) | 136 (14.3) | 81 (9.3) | 51 (6.2) | 476 (11.4) | |
| African American | 323 (34.4) | 151 (25.8) | 225 (23.6) | 199 (22.9) | 158 (19.2) | 1056 (25.3) | |
| Hispanic | 190 (20.2) | 134 (22.9) | 202 (21.2) | 185 (21.3) | 170 (20.7) | 881 (21.1) | |
| Less than College degree | 640 (68.2) | 392 (66.9) | 566 (59.3) | 481 (55.4) | 407 (49.6) | 2486 (59.6) | <0.001 |
| Household Income ($/yr) | 0.002 | ||||||
| <$20k | 139 (15.2) | 110 (19.2) | 183 (19.7) | 143 (16.9) | 148 (18.4) | 723 (17.8) | |
| $20,000–$34,999 | 183 (20.0) | 129 (22.6) | 193 (20.8) | 158 (18.6) | 140 (17.4) | 803 (19.8) | |
| $35,000–$50,000 | 157 (17.2) | 84 (14.7) | 167 (18.0) | 144 (17.0) | 131 (16.3) | 683 (16.8) | |
| $50,000–$99,999 | 303 (33.2) | 164 (28.7) | 251 (27.1) | 243 (28.7) | 225 (28.0) | 1186 (29.2) | |
| ≥$100,000 | 131 (14.4) | 85 (14.9) | 133 (14.4) | 160 (18.9) | 161 (20.0) | 670 (16.5) | |
| High Acculturation | 671 (71.5) | 380 (64.8) | 620 (65.0) | 592 (68.1) | 556 (67.8) | 2818 (67.6) | 0.02 |
| APOE ε4 (≤ 1 allele) | 248 (28.0) | 153 (27.9) | 234 (26.2) | 224 (27.5) | 188 (24.2) | 1047 (26.7) | 0.41 |
| BMI, mean (SE) | 29.3 (5.9) | 28.6 (5.2) | 28.1 (5.2) | 28.2 (5.1) | 27.4 (5.0) | 28.29 (5.3) | <0.001 |
| Current Smoker | 199 (21.2) | 74 (12.6) | 97 (10.2) | 83 (9.6) | 40 (4.9) | 493 (11.8) | <0.001 |
| Alcohol intake | 0.03 | ||||||
| Never | 146 (15.6) | 109 (18.7) | 195 (20.4) | 169 (19.5) | 152 (18.6) | 770 (18.6) | |
| Former | 206 (22.1) | 142 (24.3) | 210 (22.1) | 165 (19.0) | 158 (19.3) | 881 (21.2) | |
| Current | 582 (62.3) | 333 (57.0) | 546 (57.5) | 533 (61.5) | 507 (62.1) | 2501 (60.2) | |
| CES-D (≥16) | 124 (13.2) | 70 (12.0) | 90 (9.4) | 88 (10.2) | 91 (11.1) | 463 (11.1) | 0.08 |
| Physical Activity (MET min/week), mean (SE) | 1456.1 (2631.8) | 1438.5 (2007.5) | 1569.6 (2080.2) | 1746.9 (2543.8) | 1956.6 (2641.9) | 1638.8 (2323.1) | <0.001 |
| Total Energy intake (kcal), mean (SE) | 1731.6 (775.1) | 1690.1 (749.9) | 1613.7 (779.4) | 1671.0 (771.9) | 1597.0 (628.7) | 1659.7 (746.6) | <0.001 |
| Glucose Levels | <0.001 | ||||||
| Normal | 696 (74.2) | 446 (76.5) | 734 (77.3) | 647 (74.7) | 678 (82.9) | 3201 (77.0) | |
| Pre-diabetes | 155 (16.5) | 77 (13.2) | 113 (11.9) | 110 (12.7) | 79 (9.7) | 534 (12.9) | |
| Diabetes Mellitus | 87 (9.3) | 60 (10.3) | 103 (10.8) | 109 (12.6) | 61 (7.5) | 420 (10.1) | |
| Diabetes Medications | 61 (6.5) | 44 (7.5) | 83 (8.7) | 91 (10.5) | 43 (5.2) | 322 (7.7) | <0.001 |
| Systolic BP (mm Hg), mean (SE) | 124.3 (19.7) | 124.5 (19.9) | 123.3 (20.2) | 124.9 (20.0) | 125.5 (21.4) | 124.5 (20.24) | 0.23 |
| Diastolic BP (mm Hg), mean (SE) | 73.9 (10.0) | 72.9 (10.0) | 71.8 (10.0) | 70.9 (10.1) | 69.9 (10.1) | 71.9 (10.1) | <0.001 |
| Hypertension (Yes) | 378 (40.3) | 239 (40.8) | 372 (39.0) | 373 (42.9) | 340 (41.4) | 1702 (40.8) | 0.53 |
| Antihypertensive Medication | 308 (32.8) | 200 (34.1) | 304 (31.9) | 328 (37.7) | 278 (33.9) | 1418 (34.0) | 0.09 |
| Alzheimer’s Medication (Exam 6) | 5 (0.77) | 5 (1.19) | 15 (2.3) | 8 (1.44) | 9 (1.66) | 42 (1.5) | 0.23 |
| History of Stroke | 8 (0.85) | 5 (0.9) | 17 (1.8) | 20 (2.3) | 11 (1.3) | 61 (1.5) | 0.06 |
| LDL Cholesterol (mg/dl), mean (SE) | 119.7 (31.1) | 118.2 (29.1) | 117.2 (30.5) | 116.6 (32.2) | 115.3 (30.0) | 117.4 (30.8) | 0.04 |
| HDL Cholesterol (mg/dl), mean (SE) | 48.8 (14.2) | 48.3 (12.6) | 50.6 (14.0) | 52.5 (15.6) | 56.0 (16.4) | 51.3 (15.0) | <0.001 |
| Total Cholesterol (mg/dL), mean (SE) | 195.0 (35.8) | 194.0 (32.7) | 193.6 (34.1) | 194.5 (37.6) | 195.4 (34.1) | 194.5 (35.0) | 0.84 |
Numbers are No. (%) unless indicated as mean (SE).
Abbreviations: APOE ε4, Apolipoprotein E4 allele; BMI, Body Mass Index; BP, Blood Pressure; CES-D, Center for Epidemiologic Studies- Depression Scale; DASH, Dietary Approaches to Stop Hypertension; HDL, high density lipoprotein; LDL, low density lipoprotein; MET, Metabolic Equivalent; Q1-Q5, DASH Quintiles 1–5; SE, standard error.
DASH Diet Adherence and Cognitive Function
Table 2 presents associations between baseline DASH adherence and global cognitive function at Exams 5 and 6, and the change between them. DASH adherence was not significantly associated with performance on the CASI, Digit Span, or Digit Symbol Coding tests before or after adjusting for covariates.
Table 2.
Baseline DASH Diet score quintiles predicting Cognitive Performance
| DASH Diet Quintiles | ||||||
|---|---|---|---|---|---|---|
| Cognitive Performance Score (SE) | Q1 | Q2 | Q3 | Q4 | Q5 | p value |
| Higher quintiles indicate greater DASH adherence | ||||||
| CASI | ||||||
| Exam 5 | 84.13 (0.61) |
83.66 (0.63) |
83.59 (0.60) |
83.93 (0.60) |
84.07 (0.62) |
0.41 |
| Exam 6 | 82.15 (1.25) |
82.37 (1.28) |
82.08 (1.24) |
82.35 (1.27) |
82.74 (1.28) |
0.73 |
| Change | −0.24 (0.16) |
−0.21 (0.16) |
−0.25 (0.16) |
−0.29 (0.16) |
−0.16 (0.16) |
0.37 |
| DSC | ||||||
| Exam 5 | 41.38 (1.31) |
40.47 (1.36) |
40.63 (1.29) |
40.55 (1.30) |
40.31 (1.35) |
0.66 |
| Exam 6 | 38.54 (2.82) |
38.51 (2.88) |
39.24 (2.79) |
37.91 (2.84) |
37.02 (2.87) |
0.35 |
| Change | −0.30 (0.14) |
−0.33 (0.14) |
−0.30 (0.14) |
−0.36 (0.14) |
−0.34 (0.14) |
0.75 |
| Digit Span | ||||||
| Exam 5 | 14.83 (0.35) |
14.56 (0.37) |
14.55 (0.35) |
14.80 (0.35) |
14.71 (0.37) |
0.52 |
| Exam 6 | 14.51 (0.73) |
14.40 (0.74) |
14.29 (0.72) |
14.72 (0.74) |
14.55 (0.74) |
0.67 |
| Change | −0.05 (0.14) |
−0.01 (0.14) |
−0.11 (0.14) |
−0.06 (0.14) |
−0.03 (0.14) |
0.44 |
Abbreviations: CASI, Cognitive Abilities Screening Instrument; DSC, Digit Symbol Coding; SE, standard error.
Models adjusted for age, sex, race/ethnicity, education level, income level, APOE ε4 carrier status, BMI, smoking, alcohol intake, depressive symptoms, physical activity, energy intake, diabetes, diabetes medications, SPB, DPB, antihypertensive medications, stroke history, HDL, LDL, and total cholesterol
Pre-specified subgroup analyses included stratification by self-reported hypertension, diabetes, race/ethnicity, and APOE ε4 (data not shown). There were no significant associations in analyses stratified by hypertension, diabetes, or APOE ε4. Among Hispanic participants, there was a significant inverse association between DASH adherence and working memory (DS, p=0.05). Acculturation was identified post-hoc as an important covariate for stratification as ~33% of our sample was either born in another country or spoke a language other than English at home, and were classified in the lower acculturation group. We also conducted post-hoc evaluation of intentional exercise. Interaction terms for diet by acculturation groups and by intentional exercise groups were not significant.
Sensitivity analyses of the relationship between DASH adherence and Exam 5 cognition using IPW produced similar results (not shown) after accounting for systematic exclusion based on study attrition prior to Exam 5, or missing dietary or cognitive information. While age, BMI, and race/ethnicity significantly predicted sample group membership (n=4,169), total DASH adherence was unrelated to participant attrition.
DASH Food Group Components and Cognitive Function
Significant associations between DASH components and global cognitive function were seen for nuts/legumes, meats, and sodium after covariate adjustment (Table 3). Higher intake of nuts/legumes was associated with better Exams 5 and 6 performances (CASI; p=0.003, p=0.007). Lower meat intake was associated with lower global cognitive scores at Exam 5 (CASI; p=0.04). Similarly, lower sodium intake was associated with poorer global cognitive function at Exam 5 (CASI; p=0.007). No DASH components were associated with global cognitive change from Exams 5 to 6.
Table 3.
Association between baseline DASH Food Group Components and Cognitive Performance on the CASI
| Exam 5 CASI Score (SE) | Exam 6 CASI Score (SE) | CASI Change (SE) | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Dash Component Quintiles | Dash Component Quintiles | Dash Component Quintiles | ||||||||||||||||
| Q1 | Q2 | Q3 | Q4 | Q5 | p value | Q1 | Q2 | Q3 | Q4 | Q5 | p value | Q1 | Q2 | Q3 | Q4 | Q5 | p value | |
| Food Group | Higher quintiles indicate greater DASH adherence | Higher quintiles indicate greater DASH adherence | Higher quintiles indicate greater DASH adherence | |||||||||||||||
| Fruit | 84.02 (0.60) | 83.89 (0.60) | 83.81 (0.61) | 83.35 (0.62) | 83.97 (0.61) | 0.29 | 82.44 (1.24) | 82.04 (1.26) | 82.00 (1.28) | 82.28 (1.27) | 81.93 (1.27) | 0.84 | −0.25 (0.16) | −0.24 (0.16) | 0.23 (0.16) | −0.17 (0.16) | −0.26 (0.16) | 0.63 |
| Vegetables | 83.56 (0.62) | 83.67 (0.61) | 84.18 (0.60) | 84.04 (0.61) | 83.77 (0.62) | 0.37 | 81.87 (1.25) | 82.55 (1.26) | 82.04 (1.25) | 83.10 (1.26) | 82.11 (1.26) | 0.09 | −0.26 (0.16) | −0.13 (0.16) | −0.28 (0.16) | −0.17 (0.16) | −0.24 (0.16) | 0.11 |
| Whole Grains | 83.59 (0.60) | 83.73 (0.60) | 83.87 (0.61) | 84.16 (0.61) | 84.35 (0.62) | 0.24 | 82.22 (1.26) | 82.31 (1.25) | 81.71 (1.28) | 81.98 (1.26) | 82.28 (1.26) | 0.71 | −0.19 (0.16) | −0.22 (0.16) | −0.29 (0.16) | −0.28 (0.16) | −0.29 (0.16) | 0.43 |
| Nuts/Legumes | 83.14 (0.61) | 83.64 (0.60) | 83.75 (0.61) | 84.53 (0.61) | 84.22 (0.61) | 0.003 | 81.29 (126) | 82.18 (125) | 82.17 (125) | 82.36 (125) | 83.40 (1.27) | 0.007 | −0.22 (0.16) | −0.27 (0.16) | −0.23 (0.16) | −0.26 (0.16) | −0.17 (0.16) | 0.57 |
| Low-fat Dairy | 83.84 (0.62) | 83.66 (0.61) | 83.75 (0.61) | 84.21 (0.60) | 83.82 (0.62) | 0.56 | 81.87 (1.26) | 81.82 (1.26) | 82.35 (1.26) | 82.62 (1.25) | 82.60 (1.27) | 0.38 | −0.25 (0.16) | −0.27 (0.16) | −0.23 (0.16) | 0.22 (0.16) | −0.19 (0.16) | 0.79 |
| Meats | 84.11 (0.62) | 84.28 (0.61) | 83.84 (0.60) | 83.46 (0.62) | 83.21 (0.63) | 0.04 | 82.66 |(126) | 82.69 (125) | 82.70 (125) | 81.97 (1.27) | 80.99 (126) | 0.006 | −0.22 (0.16) | −0.20 (0.16) | −0.19 (0.16) | −0.30 (0.16) | −0.30 (0.16) | 0.30 |
| Sweetened Beverages | 83.62 (0.62) | 84.08 (0.62) | 83.88 (0.61) | 84.62 (0.70) | 83.75 (0.58) | 0.28 | 81.56 (1.28) | 81.87 (1.26) | 82.39 (1.24) | 82.76 (1.35) | 82.38 (1.24) | 0.34 | −0.29 (0.16) | −0.25 (0.16) | −0.23 (0.16) | −0.28 (0.17) | −0.21 (0.16) | 0.79 |
| Sodium | 84.30 (0.67) | 84.12 (0.61) | 84.35 (0.61) | 83.12 (0.62) | 83.23 (0.65) | 0.007 | 82.50 (1.32) | 81.78 (1.27) | 82.79 (1.25) | 81.72 (1.28) | 81.95 (1.29) | 0.14 | −0.32 (0.17) | −0.27 (0.16) | −0.20 (0.16) | −0.19 (0.16) | −0.18 (0.17) | 0.68 |
Abbreviations: CASI, Cognitive Abilities Screening Instrument; SE, standard error.
Models adjusted for age, sex, race/ethnicity, education level, income level, APOE ε4 carrier status, BMI, smoking, alcohol intake, depressive symptoms, physical activity, energy intake, diabetes, diabetes medications, SPB, DPB, antihypertensive medications, stroke history, HDL, LDL, and total cholesterol
Associations between DASH components, processing speed (DSC), and working memory (DS) are shown in supplemental tables S1 and S2. Higher consumption of whole grains predicted better processing speed at Exam 5 (DSC; p=0.04). Lower meat consumption was associated with lower processing speed at Exams 5 (DSC; p<0.001) and 6 (DSC; p=0.001). Lower sodium consumption was associated with poorer processing speed at Exam 5 (DSC; p=0.01). On Table S2, higher consumption of whole grains (DS; p=0.01) and lower consumption of sweetened beverages (DS; p=0.04) were associated with better working memory at Exam 6. Higher fruit consumption was associated with poorer working memory at Exam 5 (DS; p=0.01). Higher meat consumption was associated with more decline in working memory (p=0.04). No other DASH components were associated with cognitive change from Exams 5 to 6.
Discussion
We investigated associations between overall DASH diet adherence and DASH food group components with cognitive function in a large, multi-ethnic cohort. We also examined potential differences by health and demographic characteristics that could influence diet. DASH adherence in MESA was not associated with global cognitive function, processing speed, or working memory. Adherence with components of the DASH diet were individually, but differentially correlated with performance on different tests. Higher DASH adherence for components like vegetables, nuts/legumes, whole grains, and sweetened beverages was predictive of better cognitive function across multiple tests, which aligns with similar studies of cognitive function and decline.10,14 However, we observed that increased DASH adherence for components including sodium, meats, and fruits were associated with poorer performance on multiple tests. Differences in cohort characteristics such as age, health factors, and diversity may offer some insight into deviation from expected results in our study.
Age
Our results are inconsistent with previous studies showing a relationship between DASH adherence, better cognitive function, and less decline.10,13,14,35 The Cache County Study, the Nurse’s Health Study, and the Memory and Aging Project all showed that increased overall adherence to the DASH diet was associated with better global cognitive functioning or reduced cognitive decline.10,13,14 The MESA cohort was younger at baseline than these prior studies. The average age of participants was ~15 years younger at baseline and 5 years younger at initial cognitive assessment (~69 years old) than participants in similar studies.10,13,14 Better DASH adherence was also seen in older participants in our study. Despite adjusting for age and lacking evidence of an interaction effect between age and DASH adherence, it is possible that this association could mask effects, as age is a powerful risk factor for cognitive decline.
Health Factors
The age difference in our study may explain the lower baseline prevalence of cardio-metabolic risk factors for cognitive dysfunction, such as hypertension and diabetes, compared with other studies. Roughly 40% of the sample was hypertensive at baseline, compared to 55%−65% in other studies.10,13 Hypertension and diabetes have been identified as independent risk factors for cognitive dysfunction in MESA36 and would be expected to modify the relationship between DASH adherence and cognitive function. Additionally, a previous study showed that less than 30% of MESA participants met DASH nutrient targets.37 Reported DASH adherence varies between studies and is generally higher in clinical trials than in cohort studies.38 This lack of adherence may be reflected in the lack of association between DASH adherence with hypertension and SBP in Table 1, and could be important as higher adherence may be necessary to observe significant differences in cardiovascular and cognitive health.39
Diversity/Acculturation
The MESA cohort is diverse, and the aforementioned studies were mainly comprised of either predominantly (~90%) White participants or did not report racial/ethnic group characteristics.10,13,14,35 In MESA, suboptimal intake of DASH nutrient targets was particularly high in ethnic minorities.37 Our stratified analyses showed a similar pattern of results in White, Black, and Chinese participants, i.e., that overall baseline DASH adherence was not predictive of later life cognition. Interestingly, better DASH adherence was associated with poorer working memory among Hispanic participants. It would be inappropriate to draw conclusions from this as Hispanic populations are very heterogeneous, with differences in origins and acculturation levels,40 which may be a source of bias for these results. Higher levels of acculturation in Hispanic participants have been associated with increased dementia risk factors including diabetes, hypertension, and subclinical CVD in MESA29,41,42 and with better cognitive test performance.43 Our results may be indicative of higher acculturated individuals having both poorer diets and better cognitive scores, due to factors like education and language. Post hoc analyses stratified by acculturation demonstrated that increased DASH adherence was not associated with cognition in either high or low acculturation groups. Further study of this relationship in ethnically diverse samples is needed to better understand the complex relationships between diet, cognitive aging, and the influence of acculturation.
Counterintuitive Findings
We found a counterintuitive association between lower sodium intake and cognitive function. Others have shown potential associations between both low and high sodium levels and worse cerebrovascular and cognitive outcomes.44–47 One cross-sectional study found an association between lower sodium intake and poorer cognitive function in an older cohort, citing the potential for a J-shaped association which may indicate that older adults may need to avoid low sodium diets for the maintenance of cognitive health.46 Nonetheless, average Americans consume at least 1000 mg/d more sodium than recommended48 so this is an unlikely explanation. Alternatively, participants with health problems and/or risk factors for cognitive decline, may strive to consume healthier diets. Those with less healthy diets, i.e., higher meat and sodium intake, may be younger or free of conditions that influence diet like hypertension or diabetes. While Table 1 does not indicate that those with adverse health characteristics have better diets, it is possible that indication bias may have been missed during the long interval between the FFQ and cognitive assessment.
The interval between the FFQ and cognitive assessments, a mean of 9.5 years, could explain the lack of significant associations in this study. Supporting this, studies with positive associations between DASH adherence and cognitive function have less than a 5-year difference between the FFQ and cognitive assessments,10,12,13 while longer intervals have failed to show this same relationship.33,49 As diet-related benefits to cardiovascular and cognitive health are likely dependent upon the duration of a healthy diet, assessment of dietary patterns over time may be essential for identifying DASH-related cognitive changes in prospective cohort studies such as this. Finally, a recent report from the Chicago Health and Aging Project noted that some diet scores fail to consider the impact of unhealthy foods, which may serve to attenuate the association between a healthy diet like DASH and cognitive function.50
Strengths and Limitations
The strengths of this study are the prospective cohort design, detailed longitudinal cardiovascular data, and racially and ethnically diverse cohort. There are some limitations; our exposure variable, DASH adherence, was based on FFQ responses and likely has some measurement error, as it not designed to provide precise measurements of some components, like sodium, and may omit or misclassify foods like fast-foods and whole dairy products. Different food groups are unweighted and thus contribute equally to overall DASH scores, potentially attenuating the effect of certain components.23 No corrections were made for multiple comparisons as our investigation of dietary components was determined a priori and is closely related to our main analyses, following previously described conventions.51 Adjustments for multiple comparisons would likely eliminate any significant findings reported. This sample was susceptible to biases that are common in nutritional studies, i.e., selection and reporting biases that generally lead to healthier participants compared to non-participants, and under-reporting of negative behaviors related to food intake.52 However, bias due to selective attrition is unlikely, as demonstrated by inverse probability weighting analysis results. We were unable to include in this analysis, examination of dietary supplements such as Vitamins D, B12, and folic acid due to missing data. We evaluated whether multivitamin use or thyroid agents differed across DASH quintiles using a χ2 analysis and thyroid agent use was significantly higher among those in higher DASH quintiles, but no difference was observed in multivitamin use across DASH quintiles. Sensitivity analyses with thyroid agents as an additional covariate in our regression models resulted in no significant changes (data not shown). Therefore, this variable was not included in the final regression model. Consideration of medications that may affect cognition, such as anti-depressants or neuroleptics, were beyond the scope of this study.
Conclusion/Future Directions
We examined the association between mid-to-late life DASH adherence on cognitive functioning in MESA. Overall DASH adherence was not associated with cognitive function in this relatively healthy, and multi-ethnic, sample. However, analysis of food group components partially supported the hypotheses that greater adherence would result in better late life cognitive function, particularly through increased consumption of whole grains, nuts, legumes, and decreased intake of sweetened beverages. The relationship between DASH adherence and cognition may vary by ethnicity. Future studies should investigate this association in groups with increased cardiovascular risk factors and ethnic diversity.
Supplementary Material
Acknowledgements
The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.
Funding Sources
This research was supported by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168 and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences (NCATS).
KMH, TMH, and SRR were supported in part by P30AG049638, which funds the Wake Forest Alzheimer’s Disease Core Center and the MESA Core. KMH, TMH, and JAL were supported in part by R01AG058969. JAL was also supported by K24AG045334; and TMH was also supported by R01AG054069.
Footnotes
Conflict of Interest Statement:
The authors report no conflicts of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Bazzano LA, Green T, Harrison TN, Reynolds K. Dietary Approaches to Prevent Hypertension. Curr Hypertens Rep. 2013;15(6):694–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ozemek C, Laddu DR, Arena R, Lavie CJ. The role of diet for prevention and management of hypertension. Curr Opin Cardiol. 2018;33(4):388–393. [DOI] [PubMed] [Google Scholar]
- 3.Retelny VS, Neuendorf A, Roth JL. Nutrition Protocols for the Prevention of Cardiovascular Disease. In. Vol 23. United States: SAGE Publications; 2008:468–476. [DOI] [PubMed] [Google Scholar]
- 4.Whelton PK, He J, Appel LJ, et al. Primary prevention of hypertension: clinical and public health advisory from The National High Blood Pressure Education Program. JAMA. 2002;288(15):1882–1888. [DOI] [PubMed] [Google Scholar]
- 5.National Heart Lung and Blood Institute. Your guide to lowering your blood pressure with DASH. Bethesda, Md. : U.S. Department of Health and Human Services, National Institutes of Health, National Heart, Lung, and Blood Institute;2006. [Google Scholar]
- 6.Williamson JD, Pajewski NM, Auchus AP, et al. Effect of Intensive vs Standard Blood Pressure Control on Probable Dementia: A Randomized Clinical Trial. JAMA. 2019;321(6):553–561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Appel LJ, Champagne CM, Harsha DW, et al. Effects of comprehensive lifestyle modification on blood pressure control: main results of the PREMIER clinical trial. JAMA. 2003;289(16):2083–2093. [DOI] [PubMed] [Google Scholar]
- 8.Appel LJ, Moore TJ, Obarzanek E, et al. A clinical trial of the effects of dietary patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med. 1997;336(16):1117–1124. [DOI] [PubMed] [Google Scholar]
- 9.Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. N Engl J Med. 2001;344(1):3–10. [DOI] [PubMed] [Google Scholar]
- 10.Berendsen AAM, Kang JH, van de Rest O, Feskens EJM, de Groot L, Grodstein F. The Dietary Approaches to Stop Hypertension Diet, Cognitive Function, and Cognitive Decline in American Older Women. J Am Med Dir Assoc. 2017;18(5):427–432. [DOI] [PubMed] [Google Scholar]
- 11.Chen X, Maguire B, Brodaty H, O’Leary F. Dietary Patterns and Cognitive Health in Older Adults: A Systematic Review. J Alzheimers Dis. 2019;67(2):583–619. [DOI] [PubMed] [Google Scholar]
- 12.Smith PJ, Blumenthal JA, Babyak MA, et al. Effects of the dietary approaches to stop hypertension diet, exercise, and caloric restriction on neurocognition in overweight adults with high blood pressure. Hypertension. 2010;55(6):1331–1338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tangney CC, Li H, Wang Y, et al. Relation of DASH- and Mediterranean-like dietary patterns to cognitive decline in older persons. Neurology. 2014;83(16):1410–1416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wengreen H, Munger RG, Cutler A, et al. Prospective study of Dietary Approaches to Stop Hypertension- and Mediterranean-style dietary patterns and age-related cognitive change: the Cache County Study on Memory, Health and Aging. Am J Clin Nutr. 2013;98(5):1263–1271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen C, Zissimopoulos JM. Racial and ethnic differences in trends in dementia prevalence and risk factors in the United States. Alzheimer’s & Dementia : Translational Research & Clinical Interventions. 2018;4:510–520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gu A, Yue Y, Desai RP, Argulian E. Racial and Ethnic Differences in Antihypertensive Medication Use and Blood Pressure Control Among US Adults With Hypertension: The National Health and Nutrition Examination Survey, 2003 to 2012. Circulation Cardiovascular Quality and Outcomes. 2017;10(1). [DOI] [PubMed] [Google Scholar]
- 17.Redmond N, Baer HJ, Hicks LS. Health behaviors and racial disparity in blood pressure control in the national health and nutrition examination survey. Hypertension (Dallas, Tex : 1979). 2011;57(3):383–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bild DE, Bluemke DA, Burke GL, et al. Multi-Ethnic Study of Atherosclerosis: objectives and design. Am J Epidemiol. 2002;156(9):871–881. [DOI] [PubMed] [Google Scholar]
- 19.Nettleton JA, Steffen LM, Mayer-Davis EJ, et al. Dietary patterns are associated with biochemical markers of inflammation and endothelial activation in the Multi-Ethnic Study of Atherosclerosis (MESA). Am J Clin Nutr. 2006;83(6):1369–1379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mayer-Davis EJ, Vitolins MZ, Carmichael SL, et al. Validity and reproducibility of a food frequency interview in a Multi-Cultural Epidemiology Study. Ann Epidemiol 1999;9(5):314–324. [DOI] [PubMed] [Google Scholar]
- 21.Nettleton JA, Rock CL, Wang Y, Jenny NS, Jacobs DR. Associations between dietary macronutrient intake and plasma lipids demonstrate criterion performance of the Multi-Ethnic Study of Atherosclerosis (MESA) food-frequency questionnaire. Br J Nutr. 2009;102(8):1220–1227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hayden KM, Beavers DP, Steck SE, et al. The association between an inflammatory diet and global cognitive function and incident dementia in older women: The Women’s Health Initiative Memory Study. Alzheimers Dement. 2017;13(11):1187–1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fung TT, Chiuve SE, McCullough ML, Rexrode KM, Logroscino G, Hu FB. Adherence to a DASH-style diet and risk of coronary heart disease and stroke in women. Arch Intern Med. 2008;168(7):713–720. [DOI] [PubMed] [Google Scholar]
- 24.Levitan EB, Wolk A, Mittleman MA. Consistency with the DASH diet and incidence of heart failure. Arch Intern Med. 2009;169(9):851–857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Tsai RC, Lin KN, Wang HJ, Liu HC. Evaluating the uses of the total score and the domain scores in the Cognitive Abilities Screening Instrument, Chinese version (CASI C-2.0): results of confirmatory factor analysis. International psychogeriatrics. 2007;19(6):1051–1063. [DOI] [PubMed] [Google Scholar]
- 26.Teng EL, Hasegawa K, Homma A, et al. The Cognitive Abilities Screening Instrument (CASI): a practical test for cross-cultural epidemiological studies of dementia. International psychogeriatrics. 1994;6(1):45–58; discussion 62. [DOI] [PubMed] [Google Scholar]
- 27.Wechsler D Wechsler Adult Intelligence Scale-III (WAIS-III) New York: Psychological Corporation/Harcourt, Inc.; 1996. [Google Scholar]
- 28.Weuve J, Barnes LL, de Leon CFM, et al. Cognitive aging in black and white Americans: cognition, cognitive decline, and incidence of Alzheimer disease dementia. Epidemiology (Cambridge, Mass). 2018;29(1):151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kandula NR, Diez-Roux AV, Chan C, et al. Association of acculturation levels and prevalence of diabetes in the multi-ethnic study of atherosclerosis (MESA). Diabetes Care. 2008;31(8):1621–1628. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Beekman AT, Deeg DJ, Van Limbeek J, Braam AW, De Vries MZ, Van Tilburg W. Criterion validity of the Center for Epidemiologic Studies Depression scale (CES-D): results from a community-based sample of older subjects in The Netherlands. Psychological medicine. 1997;27(1):231–235. [DOI] [PubMed] [Google Scholar]
- 31.Radloff LS, Locke BZ. Center for Epidemiologic Studies Depression Scale (CES-D). 1st ed. Washington, DC: American Psychiatric Association; 2000. [Google Scholar]
- 32.Radloff LS. The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement. 1977;1(3):385–401. [Google Scholar]
- 33.Haring B, Wu C, Mossavar-Rahmani Y, et al. No Association between Dietary Patterns and Risk for Cognitive Decline in Older Women with 9-Year Follow-Up: Data from the Women’s Health Initiative Memory Study. J Acad Nutr Diet. 2016;116(6):921–930.e921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Nguyen HT, Bertoni AG, Nettleton JA, Bluemke DA, Levitan EB, Burke GL. DASH eating pattern is associated with favorable left ventricular function in the multi-ethnic study of atherosclerosis. J Am Coll Nutr. 2012;31(6):401–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Abbatecola AM, Russo M, Barbieri M. Dietary patterns and cognition in older persons. Curr Opin Clin Nutr Metab Care. 2018;21(1):10–13. [DOI] [PubMed] [Google Scholar]
- 36.Hughes TM, Craft S, Baker LD, et al. Changes in metabolic risk factors over 10 years and their associations with late-life cognitive performance: The Multi-Ethnic Study of Atherosclerosis. In: Alzheimers Dement (Amst). Vol 8.2017:18–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Gao SK, Fitzpatrick AL, Psaty B, et al. Suboptimal nutritional intake for hypertension control in 4 ethnic groups. Arch Intern Med. 2009;169(7):702–707. [DOI] [PubMed] [Google Scholar]
- 38.Kwan MW-M, Wong MC-S, Wang HH-X, et al. Compliance with the Dietary Approaches to Stop Hypertension (DASH) diet: a systematic review. PLoS One. 2013;8(10):e78412–e78412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Folsom AR, Parker ED, Harnack LJ. Degree of Concordance With DASH Diet Guidelines and Incidence of Hypertension and Fatal Cardiovascular Disease. American Journal of Hypertension. 2007;20(3):225–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Pontón MO, Ardila A. The future of neuropsychology with Hispanic populations in the United States. Arch Clin Neuropsychol. 1999;14(7):565–580. [PubMed] [Google Scholar]
- 41.Lutsey PL, Diez Roux AV, Jacobs DR Jr., et al. Associations of acculturation and socioeconomic status with subclinical cardiovascular disease in the multi-ethnic study of atherosclerosis. Am J Public Health. 2008;98(11):1963–1970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Moran A, Diez Roux AV, Jackson SA, et al. Acculturation Is Associated With Hypertension in a Multiethnic Sample*. American Journal of Hypertension. 2007;20(4):354–363. [DOI] [PubMed] [Google Scholar]
- 43.Herrera LP, Ponton MO, Corona M, Gonzales J, Higareda I. Acculturation impact on neuropsychological test performance in a Hispanic population. Archives of Clinical Neuropsychology. 1998;13(1):27–28. [Google Scholar]
- 44.Faraco G, Brea D, Garcia-Bonilla L, et al. Dietary salt promotes neurovascular and cognitive dysfunction through a gut-initiated TH17 response. Nat Neurosci. 2018;21(2):240–249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Nowak KL, Yaffe K, Orwoll ES, et al. Serum Sodium and Cognition in Older Community-Dwelling Men. Clin J Am Soc Nephrol. 2018;13(3):366–374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Rush TM, Kritz-Silverstein D, Laughlin GA, Fung TT, Barrett-Connor E, McEvoy LK. Association between Dietary Sodium Intake and Cognitive Function in Older Adults. J Nutr Health Aging. 2017;21(3):276–283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Strazzullo P, D’Elia L, Kandala NB, Cappuccio FP. Salt intake, stroke, and cardiovascular disease: meta-analysis of prospective studies. BMJ. 2009;339:b4567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Anderson CAM, Johnson RK, Kris-Etherton PM, Miller EA. Commentary on Making Sense of the Science of Sodium. Nutr Today. 2015;50(2):66–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Shakersain B, Rizzuto D, Larsson SC, Faxén-Irving G, Fratiglioni L, Xu W-L. The Nordic Prudent Diet Reduces Risk of Cognitive Decline in the Swedish Older Adults: A Population-Based Cohort Study. Nutrients. 2018;10(2):229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Agarwal P, Dhana K, Barnes LL, et al. Unhealthy foods may attenuate the beneficial relation of a Mediterranean diet to cognitive decline. Alzheimers Dement. 2021;17(7):1157–1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Greenland S, Hofman A. Multiple comparisons controversies are about context and costs, not frequentism versus Bayesianism. European Journal of Epidemiology. 2019;34(9):801–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Lissner L, Heitmann BL, Bengtsson C. Population studies of diet and obesity. Br J Nutr. 2000;83 Suppl 1(S1):S21–24. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.