TABLE 1.
Summary of studies on association of Dietary Patterns with cognitive decline or Incident AD that have at least > 20% minority population included.
| Cohort and study population | Minority population% | Years Follow-up | N; Mean Age (SD) | Exposure | Outcome | Findings | Exposure Interaction with Race |
| WHICAP- Washington Heights-Inwood Columbia Aging Project (Multi-racial) 68% Female | African American (34%) and Hispanics (34%) | Longitudinal, 4–5 years | N = 2258; 77.6 (6.6) | Mediterranean diet | Incident MCI Progression of MCI to AD | ↓ MCI risk and ↓ risk for MCI conversion to AD (Scarmeas et al., 2009) | Not reported |
| Healthy dietary pattern for study population * | Incident AD | ↓ AD risk (Gu et al., 2010) | No stratified analysis by race | ||||
| Total calories and Fat intake | Incident AD | ↑ AD risk (Luchsinger et al., 2002) | |||||
| Antioxidant vitamin | Incident AD | No association (Luchsinger et al., 2003) | |||||
| CHAP-Chicago Health and Aging Project | African American (63%) | Longitudinal, 6–9 years | N = 3790; 75.4 (6.2) | Mediterranean diet, HEI-2005 | Cognitive decline OR Incident AD | Mediterranean diet ↓cognitive decline (Tangney et al., 2011) | Diet*race not significant No stratified analysis by race |
| Fruits and Vegetables | Vegetable intake ↓ cognitive decline (Morris et al., 2006) | Diet*race not significant | |||||
| Fish intake | n-3 FA ↓ AD risk (Morris et al., 2003) | Diet*race not significant | |||||
| Antioxidant vitamins | Vitamin E from foods ↓ cognitive decline (Morris et al., 2002b) ↓ AD risk (Morris et al., 2002a) | Diet*race not significant | |||||
| Folate Vitamin B12 | ↑ Cognitive decline (Morris et al., 2005) ↓ Cognitive decline (Morris et al., 2005) | Diet*race not significant | |||||
| Dietary fats | Saturated fats ↑cognitive decline (Morris et al., 2004) | Animal fat*race was not significant. | |||||
| Serum Vitamin B12 | ↓ Cognitive decline (Tangney et al., 2009) | Vitamin B12* race was not significant | |||||
| Homocysteine | No association | Homocysteine* race was not significant | |||||
| Methyl malonic acid | ↑ Cognitive decline (Tangney et al., 2009) | Methyl malonic acid* race was not significant. | |||||
| Health ABC – Health Aging and Body Composition Study (Biracial) | African American (38%) | Longitudinal, 8.0 years | N = 2326; 74.6 (2.9) | Mediterranean diet | Cognitive decline | ↓ Cognitive decline in Blacks not Whites (Koyama et al., 2015) | Mediterranean diet*race was significant Stratified analysis by race |
| Health and Retirement study 60% women | African American (22%) | Cross-sectional | N = 5907; 68 (10.8) | Mediterranean diet | Cognitive Scores | “ + ” Cognition (McEvoy et al., 2017) | Not reported |
| MIND diet | “ + ” Cognition | ||||||
| Coronary Artery Risk Development in Young Adults (CARDIA) | African American (45%) | Longitudinal, 8.0 years | N = 2621; 25 (3.5) | Mediterranean diet | Cognitive Scores assessed 25 and 30 years later | ↑ Cognitive function in midlife (McEvoy et al., 2019) | Not reported |
| 57% female | DASH diet | No association | |||||
| A Priori Dietary Quality Index | ↑ Cognitive function in midlife (McEvoy et al., 2019) | ||||||
| REGARDS- REasons for Geographic And Racial Differences in Stroke | African American (31%) | Longitudinal 4–7 years | N = 18,080; 64.4 (9.1) | Plant-based diet Southern diet | Incident cognitive Impairment | ↓ Incident cognitive impairment (Pearson et al., 2016) ↑ Incident cognitive impairment (Pearson et al., 2016) | Diet*race not significant No stratified analysis by race |
| Mediterranean diet | ↓ Incident cognitive Impairment in non-diabetic participants (Tsivgoulis et al., 2013) | Only Diet*diabetes significant. Stratified by diabetes status | |||||
| HANDLS (Healthy Aging in the Neighborhood of Diversity Across Lifespan) Participants (Biracial) 57% female | African American (51%) | Cross-sectional | N = 2090; 47.9 (9.2) | HEI-2010 | Cognitive Scores | “ + ” Cognition only in those below the poverty line (Beydoun et al., 2018a) | Diet*race not significant No stratified analysis by race |
| Dietary Antioxidant vitamins | Vitamin E “ + ” Cognition (Beydoun et al., 2015) | Vitamin E*race not significant No stratified analysis by race | |||||
| Longitudinal, 4–5 years | Dietary Vitamin D | Cognitive decline | ↓ Cognitive decline (visual memory) (Beydoun et al., 2018b) | Vitamin D* race interaction significant: Improved visual memory only in Whites and not in Blacks | |||
| Nutrient adequacy score (NAS) Caffeine Alcohol | “ + ” Cognition ↓ Attention decline (Beydoun et al., 2014) “ + ” Cognition “ + ” Attention and Working memory (Beydoun et al., 2014) | NAS*race not significant No stratified analysis by race | |||||
| Two Boston based cohorts (Boston Puerto Rican Health Study (BPRHS) and Nutrition, Aging and Memory in Elders (NAME) study) | African American (37%) BPRHS- Hispanics (100%) | Cross-sectional | N = 1956, | Plasma Vitamin B12 Vitamin B6 Folate Homocysteine | Cognitive Scores | “ + ” Cognition (Moorthy et al., 2012) “ + ” Cognition No association No association | Not reported No stratified analysis by race |
| BPRHS (Boston Puerto Rican Health Study), 70% female | Hispanics (100%) | Cross-sectional | N = 1269, 57.3 (7.6) | Mediterranean diet HEI-2005 | Cognitive Impairment | “–”cognitive impairment (Ye et al., 2013b) | N/A |
| Fruits and Vegetables | “–” Cognitive Impairment (Ye et al., 2013a) | ||||||
| Longitudinal 2 years | Dietary n-3 and n-6 PUFA | Cognitive decline | EPA, DHA and n3VLCFA ↑ Executive Function (Bigornia et al., 2018) | ||||
| Plasma vitamin B-6 | ↓ Cognitive decline (Palacios et al., 2019a) | ||||||
| Serum vitamin D | No association (Palacios et al., 2019b) |
↑ Upward arrow indicates statistically significant increased risk in longitudinal analysis; ↓ downward arrow indicates statistically significant decreased risk in longitudinal analysis; “ + ” Plus indicates statistically significant positive association in the cross-sectional analysis; “−” Minus indicates statistically significant negative association in the cross-sectional analysis. *Dietary pattern for study population identified based on Reduced Rank Regression.