TABLE 2.

Cognitive function, cognitive impairment, and dementia assessment outcomes¹

Source	Cognitive function	Cognitive impairment	Dementia	Confounders adjusted for
Chan et al., 2013 (29)	NA	No significant association between MeDi Score dietary pattern and cognitive impairment in men or women (P-trend > 0.05). Highest tertile of the MD adherence: Men, OR: 0.89 (95% CI: 0.57, 1.42); Women, OR: 1.02 (95% CI: 0.76, 1.43) vs. lowest tertile.	NA	Age, BMI, PASE, energy intake, education level, Hong Kong ladder, community ladder, smoking status, alcohol use, number of activities of daily living, GDS, and GDS category.
Crichton et al., 2013 (30)	No significant associations between absolute MeDi Score and any of the self-appraised cognitive function or psychological well-being tests.	NA	NA	Age, sex, education, BMI, exercise, smoking, and total energy intake.
Gardener et al., 2012 (31)	Significant correlation between baseline MeDi score and change in MMSE score in HC (r = 0.098; P = 0.014). No significant correlation between MeDi score and the other neuropsychological tests in HC. Correlations: LM II (P = 0.779); D-KEFS (P = 0.294); CVLT II Long Delay (P = 0.472).	Significant association between high adherence to MD and reduced risk of MCI (OR: 0.866; 95% CI: 0.75, 1.00; P < 0.05). Significant 13–19% reduction in odds of being in the MCI category for each additional unit on the MeDi score.	Significant association between high adherence* to MD and reduced risk of AD (OR: 0.806; 95% CI: 0.71, 0.92; P < 0.01). Significant 19–26% reduction in risk of being in the AD category with each additional unit of the MeDi score vs. the reference HC category.	Age at assessment; sex; country of birth; education; apoE4 allele status; current smoking status; caloric intake; BMI; and history of stroke, diabetes, hypertension, angina and heart attack.
Katsiardanis et al., 2013 (32)	NA	Significant association between MD adherence and risk of MCI in men (OR: 0.88; 95% CI: 0.80, 0.98; P = 0.02) and women (OR: 1.11; 95% CI: 1.00, 1.22; P = 0.04).	NA	Age, education, social activity, smoking, depression symptomatology (with the use of the GDS), MedDietScore (range: 0–55), and metabolic syndrome.
Ye et al., 2013 (33)	Each MeDi score point was associated with a 0.14 point higher MMSE (P = 0.012). After adjustment for all confounders MD adherence was not significantly associated with executive function, memory, or attention.	Significantly reduced risk of cognitive impairment in those in the highest vs. lowest quintile of MeDi score (OR: 0.51; 95% CI: 0.33, 0.79). Significant negative association between each point of MeDi score and risk of cognitive impairment (OR: 0.87; 95% CI: 0.80, 0.94; P-trend < 0.001).	NA	Age, sex, educational attainment, household income below threshold, acculturation score, smoking status, physical activity score, supplement use, taking >5 types of medications within the previous 12 mo, BMI, hypertension, diabetes, total cholesterol, HDL cholesterol, and TGs.
Scarmeas et al., 2006 (34)	NA	NA	Significant reduction in prevalence of AD per unit increase in the MeDi score (OR: 0.76; 95% CI: 0.66, 0.86; P < 0.001) and for highest vs. lowest tertile of the MeDi score (OR: 0.31: 95% CI: 0.16, 0.58; P < 0.001).	Age, sex, education, ethnicity, cohort, caloric intake, apoE4, BMI, smoking, and comorbidity.
Cherbuin and Anstey, 2012 (35)	NA	No significant association between each unit increase in the MeDi score and CDR (OR: 1.18; 95% CI: 0.88, 1.57) or MCI (OR: 1.41; 95% CI: 0.95, 2.10) or any MCD (OR: 1.20; (95% CI: 0.98, 1.47).	NA	Age, sex, education, apoE4 genotype, BMI, physical activity, stroke, diabetes, hypertension, and total caloric intake.
Corley et al., 2013 (36)	No significant association between the MD and IQ, processing speed, or memory in multivariate model. Significant positive correlation between “Mediterranean-style” pattern and verbal ability (NART, mean ± SD: 37.1 ± 7.5 vs. 33.0 ± 7.3, P = 0.024; WTAR, 43.4 ± 6.1 vs. 39.9 ± 6.5, P = 0.001), upper vs. lower tertile.	NA	NA	Sex, age at testing in later life, occupational social class, and IQ at age 11 y from the MHT.
Féart el al., 2009 (37)	Each unit increase in the MeDi score corresponds to 0.006 (95% CI: 0.0003, 0.01; P = 0.04) less cognitive decline per year on the MMSE.	NA	No significant association between MeDi score and dementia (HR: 1.12; 95% CI: 0.60, 2.10; P = 0.72 for highest vs. lowest on MeDi score tertile).	Age, sex, education, marital status, caloric intake, apoE4, physical exercise, 5 medications/d, depression score, BMI, diabetes, hypertension, tobacco use, hypercholesterolemia, and stroke.
Kesse-Guyot et al., 2013 (38)	No significance for most associations in the fully adjusted model, except for association between low MSDPS and poor phonemic fluency performance [−1.00 (95% CI: −1.85, −0.15); P-trend = 0.048], lowest vs. highest tertile.	NA	NA	Age, sex, education, follow-up time between baseline and cognitive evaluation, supplementation group during the trial phase, number of 24-h dietary records, energy intake, BMI, occupational status, tobacco use status, physical activity, memory difficulties at baseline, depressive symptoms concomitant with the cognitive function assessment, history of diabetes, hypertension, and cardiovascular disease.
Koyama et al., 2015 (39)	In the fully adjusted model, lower MedDietScores were associated with a significantly slower rate of cognitive decline on the 3MS score (mean: 0.22 points/y; 95% CI: 0.05, 0.39 points/y; P = 0.01) vs. those with high MedDietScores (only in black participants).	NA	NA	Age, sex, education, BMI, current smoking, physical activity, depression, diabetes, total energy intake, and socioeconomic status.
Olsson et al., 2015 (40)	NA	NA	No significant association between the mMDS and all-type cognitive impairment (OR: 0.82; 95% CI: 0.65, 1.05; P-trend = 0.41). No significant association between the mMDS and risk of AD (HR: 0.99; 95% CI: 0.44, 2.26) or all-type dementia (HR: 0.85; 95% CI: 0.44, 1.62) in highest vs. lowest tertile.	Energy intake as a continuous variable, educational level, physical activity, smoking, single household, and apoE genotype (absence of any E4 allele vs. presence of ≥1 E4 allele).
Psaltopoulou et al., 2008 (41)	Significant association for each unit increase in the MeDi score and 0.05 (95% CI: 0.09, 0.19; P = 0.49) higher cognitive function on the MMSE at follow-up.	NA	NA	Age, sex, education, marital status, caloric intake, height, physical activity, alcohol intake, smoking, depression, BMI, diabetes, and hypertension.
Samieri et al., 2013 (42)	Significant association between 5th quintile of MeDi score and “no limitation of mental health” prevalence ratio: 1.12 (95% CI: 1.04, 1.20); P < 0.001.	NA	NA	Age; education; marriage status; median income; median house value; family history of diabetes, cancer, and MI; physical activity; energy intake; smoking; multivitamin use; aspirin use; BMI; history of high blood pressure; and history of hypercholesterolemia.
Samieri et al., 2013 (43)	No significant association for alternate MeDi score and trajectories of repeated cognitive scores in the multivariate model (P-trend across quintiles = 0.26 and 0.40 for global cognition and verbal memory, respectively), nor with overall global cognition and verbal memory at older ages, assessed by averaging the 3 cognitive measures (P-trend = 0.63 and 0.44, respectively).	NA	NA	Age at the start of cognitive testing, race, higher education, annual household income, energy intake, Women’s Health Study treatment assignment (aspirin and/or vitamin E), regular vigorous exercise, BMI, current smoking, history of T2DM, self-reported history of hypertension, use of antihypertensive medications or elevated systolic blood pressure, self-reported history of elevated cholesterol, use of lipid-lowering medications or elevated blood cholesterol, postmenopausal hormone use, or self-reported history of depression.
Samieri et al., 2013 (44)	Long-term MD exposure was estimated by averaging all repeated measures of diet (>13 y, on average). During examination of cognitive status in older age, each higher quintile of long-term MeDi score was linearly associated with better mean z scores [differences in mean z scores between highest and lowest quintiles of MD: 0.06 (95% CI: 0.01, 0.11); 0.05 (95% CI: 0.01, 0.08); and 0.06 (95% CI: 0.03, 0.10) standard units; P-trend = 0.004, 0.002, and <0.001 for TICS, global cognition, and verbal memory, respectively].	NA	NA	Age, education, long-term physical activity and energy intake, BMI, smoking, multivitamin use, and history of depression, diabetes, hypertension, hypercholesterolemia, or MI.
Scarmeas et al., 2009 (45)	NA	NA	Significant association between MeDi score and risk of AD (HR: 0.60, 95% CI: 0.42, 0.87; P = 0.007) for highest vs. lowest tertile on MeDi score.	Age, sex, education, ethnicity, cohort, caloric intake, apoE4, BMI, smoking, comorbidity, depression, leisure activities, and CDR score.
Scarmeas et al., 2009 (46)	NA	Significant decrease in MCI risk with each MeDi score increase (HR: 0.92; 95% CI: 0.85, 0.99; P = 0.04). Significant decrease in MCI risk for highest vs. lowest tertile of MeDi score (HR: 0.72; 95% CI: 0.52, 1.00; P = 0.05).	No significant association between MCI conversion to AD per unit increase in 0- to 9-point MeDi score (HR: 0.89, 95% CI: 0.78, 1.02; P = 0.09). Significant association between MCI to AD conversion (HR: 0.52; 95% CI: 0.30, 0.91; P = 0.02) for highest vs. lowest tertile on MeDi score.	Age, sex, education, ethnicity, cohort, caloric intake, apoE4, BMI, and time between first dietary assessment and baseline diagnosis.
Scarmeas et al., 2006 (47)	Significant association between each unit increase in the MeDi score and 0.003 (95% CI: 0, 0.006; P = 0.05) less cognitive decline per year on the composite cognitive z score.	NA	Significant association between MeDi score and reduced risk of AD (HR: 0.91; 95% CI: 0.83, 0.98; P = 0.015 per unit increase in MeDi score); (HR: 0.60; 95% CI: 0.42, 0.87; P-trend = 0.007 for highest vs. lowest tertile).	Age, sex, education, ethnicity, cohort, caloric intake, apoE4, BMI, smoking, and comorbidity.
Tangney et al., 2014 (48)	Significant association for each unit increase in MedDietScore and a slower rate of global cognitive decline by 0.002 standardized units (P = 0.01) in mixed models adjusted for covariates. Only the upper tertile of MedDietScore was associated with rates of global cognitive change.	NA	NA	Energy, age, sex, education, and cognitive activities.
Titova et al., 2013 (49)	No significant association between MeDi score and the 7MS score; β value = 0.11; P = 0.13.	NA	NA	Sex, energy intake, education, self-reported physical activity, serum concentration of LDL cholesterol, BMI, systolic blood pressure, and HOMA-IR.
Tsivgoulis et al., 2013 (50)	NA	Significant reduction in likelihood of ICI with increased MeDi score (OR 0.87; 95% CI: 0.76, 1.00) with the use of a median split for MeDi score (0–4 vs. 5–9).	NA	Demographics, environment, vascular risk factors, antihypertensive medications, depressive symptoms, self-reported health status, incident stroke, and diabetes.
Vercambre et al., 2012 (51)	No significant associations across any MDS categories and any cognitive function tests. All P values > 0.05	NA	NA	Age, education, energy from diet, marital status, and physical activity.
Wengreen et al., 2013 (52)	Significant association between highest quintile of MD adherence and 0.94 higher score on the 3MS vs. those in the lowest quintile (P = 0.001). Differences consistent over 11 y.	NA	NA	Age; sex; education; BMI; frequency of moderate physical activity; multivitamin and mineral supplement use; history of drinking and smoking; and history of diabetes, heart attack, and stroke.
Gu et al., 2010 (53)	Better adherence to the MeDi score was marginally associated with significantly better cognitive performance at baseline: after adjusting for age, gender, race, and education, β = 0.013 (p = 0.05) for each unit increase of MeDi score.	NA	Longitudinal analysis: significant association between MeDi score and reduction in risk of AD (HR: 0.87; 95% CI: 0.78, 0.97; P = 0.01 per unit increase of MeDi score; HR: 0.68; 95% CI: 0.42, 1.08; P-trend = 0.1 for highest vs. lowest tertile of the MeDi score).	Age, sex, education, race, caloric intake, apoE4, BMI, smoking, comorbidity, insulin, and adiponectin.
Roberts et al., 2010 (54)	NA	Longitudinal analysis: no significant reduction in MCI risk with increased MeDi score (HR: 0.75, 95% CI: 0.46, 1.21; P = 0.24 for highest vs. lowest MeDi score tertile). Cross-sectional analysis: no significant reduction in prevalence of MCI (OR: 0.80; 95% CI: 0.52, 1.25; P = 0.33) in highest vs. lowest MeDi score tertile).	No significant association between MeDi score and risk of dementia, HR: 0.75 (95% CI: 0.46, 1.21; P = 0.24) for highest vs. lowest tertile on MeDi score.	Age, sex, education, caloric intake, apoE4, stroke, CHD, and depressive symptoms.
Tangney et al., 2011 (55)	Significant association between each unit increase in the MedDietScore and 0.007 (95% CI: 0.003, 0.011; P < 0.001) increase on the global cognitive z score.	NA	NA	Age, sex, education, race, total energy intake, participation in cognitive activities, and interaction between time and dietary quality score.
Martínez-Lapiscina et al., 2013 (56)	Significant association between MedDiet+EVOO and higher mean MMSE (by 0.62; 95% CI: 0.18, 1.05; P = 0.005) and CDT (by 0.51; 95% CI: 0.20, 0.82; P = 0.001) vs. controls. Significant association between and MedDiet+Nuts and higher MMSE scores (by 0.57; 95% CI: 0.11, 1.03; P = 0.015) and CDT scores (by 0.33; 95% CI: 0.003, 0.67; P = 0.048) vs. controls.	NA	NA	Sex, age, education, family history of cognitive impairment or dementia, apoE4 genotype, hypertension, dyslipidemia, diabetes, smoking status, alcohol intake, BMI, physical activity, and total energy intake.
Martínez-Lapiscina et al., 2013 (57)	Significant association between MedDiet+EVOO group and better post-trial cognitive performance in all cognitive tests vs. the control group. These crude differences were not statistically significant after correcting for multiple comparisons (all P values > 0.05).	Significantly reduced MCI risk in participants allocated to the MedDiet+EVOO vs. the control group (OR: 0.34; 95% CI: 0.12, 0.97).	NA	Sex, age, education, apoE genotype, family history of cognitive impairment or dementia, smoking, physical activity, BMI, hypertension, dyslipidemia, diabetes, alcohol, and total energy intake.
Martínez-Lapiscina et al., 2014 (58)	Significant association between MD interventions and MMSE scores in non-apoE4 carriers (0.56; 95% CI: 0.15, 0.97; P = 0.007) and apoE4 carriers (1.61; 95% CI: 0.10, 3.13; P = 0.037). Significant association between MD interventions and CDT scores in non-apoE4 carriers (0.55; 95% CI: 0.25, 0.85; P < 0.001) and apoE4 carriers (0.33; 95% CI: −0.6, 1.27; P = 0.477).	NA	NA	Sex, age, education, family history of cognitive impairment or dementia, hypertension, dyslipidemia, diabetes, smoking status, alcohol intake, BMI, physical activity, and total energy intake.
McMillan et al., 2011 (59)	Significant improvement in cognitive function speed in the “Diet Change” group vs. the “No Change” group (P = 0.002) in a post hoc comparison.	NA	NA	Not reported.
Wardle et al., 2000 (60)	Those consuming the MD had a 1.5 (95% CI: 0.7, 2.3) reduction in confusion vs. a 0.5 (95% CI: −0.4, 1.3) reduction in the control group at 12 wk after baseline.	NA	NA	Weight loss only.

¹*Definition of high adherence is not clear. AD, Alzheimer disease; CDR, clinical dementia rating; CDT, clock-drawing test; CHD, coronary heart disease; CVLT, California Verbal Learning Test; D-KEFS, Delis–Kaplan Executive Function System; GDS, Geriatric Depression Scale; HC, healthy controls; ICI, incident cognitive impairment; IQ, intelligence quotient; LM II, Logical Memory II; MCD, mild cognitive disorder; MCI, mild cognitive impairment; MD, Mediterranean diet; MDS, Mediterranean diet score; MedDietScore, Mediterranean diet score developed by Panagiotakos et al. (61); MedDiet+EVOO, Mediterranean Diet enriched with extra-virgin olive oil; MedDiet+Nuts, Mediterranean Diet enriched with nuts; MeDi score, Mediterranean diet score developed by Trichopolou et al. (13); MHT, Moray House Test; MI, myocardial infarction; mMDS, modified Mediterranean Diet Score; MMSE, Mini-Mental State Examination; MSDPS, Mediterranean-style dietary pattern score; NA, not applicable; NART, National Adult Reading Test; PASE, Physical Activity Scale for the Elderly, TICS, Telephone Interview for Cognitive Status; T2DM, type 2 diabetes mellitus; WTAR, Wechsler Test of Adult Reading; 3MS, Modified Mini-Mental State Examination; 7MS, 7-min screen.