ABSTRACT Many epidemiological studies have shown the bene ficial effects of a largely plant-based diet, and the strong association between the consumption of a Mediterranean-type diet with healthy aging including a lower risk of cognitive decline. The Mediterranean diet is characterized by a high intake of olive oil, fruits and vegetables and is rich in dietary fiber and polyphenols – both of which have been postulated to act as important mediators of these bene fits. Polyphenols are large molecules produced by plants to protect them from environmental threats and injury. When ingested by humans, as little as 5% of these molecules are absorbed in the small intestine with the majority metabolized by the gut microbiota into absorbable simple phenolic compounds. Flavan-3-ols, a type of flavonoid, contained in grapes, berries, pome fruits, tea, and cocoa have been associated with many bene ficial e ffects on several risk factors for cardiovascular disease, cognitive function and brain regions involved in memory formation. Both preclinical and clinical studies suggest that these brain and heart bene fits can be attributed to endothelial vascular effects and anti-in flammatory properties among others. More recently the gut microbiota has emerged as a potential modulator of the aging brain and intriguingly polyphenols have been shown to alter microbiota composition and be metabolized by di fferent microbial species. However, there is a need for well controlled studies in large populations to identify predictors of response, particularly given the vast inter-individual variation of human gut microbiota. KEYWORDS Polyphenols; flavonoids; phenyl-γ-valerolactone; Mediterranean diet; brain gut microbiome system; neuroinflammation; Alzheimer’s disease; major depressive disorder Introduction It is estimated that more than 2 billion people will be over 60 years by 2050 [1] and this demographic shift will be accompanied by an increase in the prevalence of cogni- tive dysfunction and neurodegenerative disorders, ran- ging from mild cognitive impairment to more severe forms of neurocognitive disorders such as Alzheimer ’s disease (AD) and Parkinson’s disease (PD) [2]. To reduce the risk of developing neurodegenerative disorders, the World Health Organization has highlighted the need to understand the lifestyle factors underlying this trend. A predominantly plant-based diet o ffers several health benefits including reduction of disease risk, and pro- motion of healthy aging [3 –8]. The traditional Mediter- ranean diet is one of the best-studied examples of a predominantly plant-based diet [9, 10]. Consuming this diet is positively associated with reduced risk of several pathologies, including major depressive disorder (MDD), AD, PD, and cardiovascular diseases [ 11–16], as well as improved cognitive function [17 –20], based on epidemiological and observational studies. Regularly consuming the Mediterranean diet has also been associ- ated with reduced microvascular brain damage [ 21–23]. The Mediterranean-like DASH diet (Dietary Approaches to Stop Hypertension) has been associated with reduced cognitive decline in older subjects [ 24– 26] and has been reported to reduce several cardiovascu- lar risk factors in clinical studies [27 –30]. Likewise, the MIND diet (Mediterranean-DASH Intervention for Neurodegenerative Delay) has been associated with larger total brain volume, better cognitive function, decreased risk of dementia, slower cognitive decline, © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License ( http://creativecommons.org/licenses/by-nc- nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. CONTACT E. A. Mayer emayer@g.ucla.edu *(F. C. Ross, D. E. Mayer: Joint first authors)
NUTRITIONAL NEUROSCIENCE and fewer symptoms of MDD [17–20]. Even though cor- relational, these studies suggest that largely plant-based diets enhance brain function and may even improve cerebral vascular function [31 ].
Historically, nutrition science has focused on reduc- tionist accounts of essential nutrients needed for healthy aging, such as vitamins and minerals. However, in recent decades, the field has broadened its perspective to encompass the multifaceted roles of dietary bioac- tives, which are not considered essential nutrients but can play important roles in disease prevention and healthy aging. These bioactives, found in various plant-based foods, exert a range of benefi cial e ffects, including antioxidant and anti-in flammatory proper- ties. This shift highlights a more holistic approach to nutrition and acknowledges the importance of a varied diet rich in non-essential nutrients. Interestingly, the MIND diet maximizes nutrients with a positive effect on neurodegenerative disorders [26 ], with an emphasis on berries. Indeed, regular consumption of berries is associated with a reduced risk of AD, PD, and with delayed cognitive decline [32 –34]. Higher consumption of polyphenols (PPs) contained in these berries has been proposed as one of the responsible factors. Fruits including apples, grapes, pears, berries, and cherries harbor up to 200 –300 mg of PPs per 100 g of fresh weight [34 ]. The scienti fic consensus is that PP are most likely responsible for the health benefi ts of these foods, but knowledge gaps remain regarding the types of PP that exert such e ffects, the optimal doses, and their mechanism(s) of action.
The overall goal of this review is to explore the role of the gut microbiota in the modulation of dietary PPs into bioactive compounds with the potential for protection from age-related decline and neurodegeneration.
Polyphenols This is a large and diverse group of molecules with two significant roles in plant health: (1) defense mechanisms that respond to homeostatic perturbations such as drought, UV light, pests, and diseases [35 ], and (2) as part of a signaling code between the root system of the plant and soil microbes in close proximity . In vitro studies have reported antioxidant properties of PPs but this is in fact a limited bioactivity when they are ingested by humans [36 ]. Indeed, while PPs were initially endorsed for their antioxidant potential by the FDA in 2012, the endorsement was removed five years later. Even though various health benefi ts have been associated with the consumption of these molecules, they are not considered micronutrients as they have no intrinsic nutrient value [ 37]. The richest sources of PPs are vegetables, fruits, tea, pome fruits, and some berries. The PP content in these foods can vary signi fi- cantly due to factors like geographical location, climate, UV exposure, chemical fertilizers, fruit ripening, sto- rage, processing, and cooking. For example, cooking adds complexity by thermal degradation of PPs, altering concentrations and bioavailability, and thus impacting the nutritional pro file of plant-based foods [ 38]. Over 8,000 PP compounds have been discovered in various plant species, and these are classi fied into four sub- classes: phenolic acids ( ∼30% of all PP), stilbenes, lig- nans, and flavonoids ( ∼60% of all PP). The largest family of PP, the flavonoids, can be divided into six further subclasses: flavan-3-ols (also known as flava- nols), flavonols, flavanones, flavones, anthocyanins, and isofl avones [34 ]. Berries contain several subclasses of flavonoids including flavan-3-ols, anthocyanins, proanthocyanins, stilbenes, tannins, and non-flavonoids such as phenolic acids (for detailed reviews see [ 39,40]). Anthocyanins are generally the major PPs in berries and have received the most research interest for their potential benefits to human health, which include neuro- protection [41]. PP profiles in berries show huge variation depending on the species of berries, geographical location, and soil conditions, and PP profiles are also influenced by ripening levels, type of growing (for example convention- ally grown vs regenerative organic), processing of foods, and storage conditions [42], which have significant conse- qu ences on their health e ffects. Due to their large size, flavonoids are poorly absorbed in the small intestine (5 – 10%), reaching very low plasma levels following con- sumption, based on human bioavailability studies. Most PPs reach the ileum and colon where they interact with the gut microbiota resulting in extensive chemical modifi- cation of the parent compounds into small phenolic, absorbable metabolites [ 43–45]. But they also cause changes to gut microbiota composition [ 46,47]a n d potentially alter its metabolic activity and functioning. Animal studies have revealed that the gut microbiota influences brain structure, function, and behavior through several signaling pathways, including neural, neuroendocrine, and immune-mediated communi- cation channels [13 ,48,49]. Gut microbiota-generated metabolites and molecules from dietary, host, and de novo sources can indirectly infl uence brain function by modulating the mucosal barrier, the gut-associated immune system, and vascular health and function, or directly by modulating synaptic plasticity, neuroin flam- mation, oxidative stress, and Aβ pathology [50]. Preclin- cial evidence suggests that the kynurenine pathway is able to modulate N-methyl-D-aspartate receptor func- tion [51 ,52]. Activation of these receptors play an important role in excitotoxicity and neuroplasticity. NUTRITIONAL NEUROSCIENCE 1059 Some of the potential neuroactive candidates include phenolic catabolites such as phenyl-γ -valerolactones derived from flavan-3-ol-containing foods [53 –58]. Phenyl-γ-valerolactones have also been used as a bio- marker of flavan-3-ol intake. Thus, gut bacteria and the molecules they generate via the metabolism of PPs are crucial to understanding how PPs a ffect the brain. The last decade has seen a growing appreciation for the role of non-essential nutrient constituents or dietary bioactives in disease prevention and healthy aging [ 59]. As an example, several dietary intervention studies have reported that the flavan-3-ol, epicatechin, plays a sec- ondary role in the prevention of diabetes, and plays pri- mary and secondary roles in the prevention of cardiovascular disease [60 ]. However, the mechanisms of action underlying such health benefits remain incom- pletely understood.
This review will focus on the most studied family of PP, the flavonoids, presenting results of preclinical and clinical studies. The flavan-3-ols that reside within the family of flavonoids represent some of the more widely consumed PP, including epicatechin [61 ]. Results of human observational studies and clinical trials that have investigated the associations between flavonoid intake and brain health, as well as the prevention of these PPs in the development of neurodegenerative dis- orders, are critically reviewed. Further, key mechanistic studies that describe and conceptualize the role of flavo- noids in the brain-gut microbiome axis are discussed. Preclinical evidence for biological mechanisms mediating brain health bene fit The health bene fits of flavonoids in humans have been proposed to be primarily a result of gut microbial effects, vascular and endothelial e ffects, anti-in flamma- tory e ffects, and to a lesser degree antioxidant e ffects. Indirect brain e ffects of flavonoids through modulation of vascular function For healthy vascular function, endothelial cells produce a range of substances, including nitric oxide (NO) which plays an important vasodilatory role [ 62]. Reduced availability of NO as a result of oxidative stress and free radicals causes endothelial dysfunction [63]. Antho- cyanins can activate endothelial NO synthase, leading to increased NO production, resulting in vasodilation which has been observed in in vitro and in vivo studies [64,65]. This is most likely due to the modulation of different signal transduction pathways, such as the ade- nosine monophosphate-activated protein kinase and the phosphatidylinositol 3-kinase/Akt pathways [ 66]. In preclinical studies, flavonoids have demonstrated vaso- dilatory e ffects in the cardiovascular system [67 ,68], and vasorelaxant e ffects in the pulmonary artery [69 ]. In clinical studies, the intake of flavonoids have been associated with a reduction in blood pressure [ 70,71], arterial sti ffness [71 ], and increased vasodilation [70,72]. Additionally, flavonoids have been shown to regulate the overproduction and activation of cytokines, modulate the gene expression of several pro-in flamma- tory molecules, and inhibit the activity of enzymes associated with infl ammation [73 ].
Indirect flavonoid e ffects on the gut-brain-axis mediated by the gut microbiota The ability of flavonoids to infl uence gut microbiota composition, relative abundance, and functionality can potentially a ffect the brain through several gut-brain signaling pathways, including hormonal, neuronal, and immune-mediated channels ( Figure 1) (reviewed in [13]).
Several preclinical studies have demonstrated the anti-inflammatory e ffects of a range of flavonoids through the gut microbiota. For example, the adminis- tration of iso flavones [74], flavan-3-ols [75], and antho- cyanins [76 ] increased anti-in flammatory short-chain fatty acid (SCFA) levels in mice, which are generated by the microbial metabolism of microbiota accessible carbohydrates (MACs). The increase in the relative abundance of Akkermansia muciniphila in response to flavonoid administration was associated with enhanced intestinal barrier integrity and a reduction in gut-based and systemic infl ammation in preclinical studies [77 – 79]. Anthocyanin administration in healthy mice showed similar e ffects of microbial changes, which were associated with improved gut barrier function and a reduction in infl ammation, including reduced brain inflammation through supplementation of Aronia berry [76 ], acai berry [80 ], or bilberry [81 ]. Additional studies have demonstrated the neuroprotective effects of alterations in gut microbiota composition as a result of anthocyanin supplementation [82 ,83]. Moreover, the flavonoid, quercetin, has demonstrated anti-in flamma- tory properties [84 ], resulting in signi ficant reductions in systemic infl ammatory markers IL-4 and IL-5 in the bronchoalveolar lavage fluid (BALF) in mouse models. Quercetin was also shown to block the NF-kB activation pathway [85 ]. Rutin has been reported to reduce proin flammatory cytokines and decrease oxi- dative stress [86 ,87].
Kaempferol was shown to reduce oxidative stress and the infl ammatory response following spinal cord injury in a rat model, by down-regulating the ROS- 1060 F. C. ROSS ET AL.
dependent MAKs-NF-kB signaling pathway [ 88]. By inhibiting microglia activation in a rat model of cer- ebral ischemia and reperfusion, kaempferol was reported to reduce cerebral in flammation demonstrat- ing both blood –brain barrier protective e ffects and anti-inflammatory e ffects [89 ]. Catechin flavonoids have been reported to increase the relative abundance of the SCFA-producing bacterium Akkermansia in a mouse model [90 ]. The anti-in flammatory properties of this PP have been reported in rodent studies, observing attenuation of IL-1 beta production by the inhibition of pro-IL-1B expression via downregulation of NF-kB p38 MAPK and TLR signaling [91 ]. The catechin, epigallocat-echin-3-gallate (EGCG), has also been reported to attenuate neuroin flammation, speci fi- cally microglial infl ammation, as well as neurotoxicity, via the inhibition of canonical NLRP3 and non-cano- nical caspase-11-dependent in flammasome activation via the TLR4/NF-kB pathway [92 ].
Direct antioxidant effects of flavonoids in vitro The direct antioxidant e ffects of flavonoids include the chelation of metal ions and the scavenging of free radicals from reactive oxygen species [ 66,93,94] which have been observed in in vitro studies. This has resulted in the incorrect assumption that PPs exert their ben- eficial e ffects in humans through direct antioxidant effects. Indirect antioxidant e ffects include suppression of stress-related signaling pathways by modulation of ligand–receptor interactions such as tumor necrosis fac- tor α (TNF- α) and its receptor [95 ]. A signi ficant reduction of oxidative DNA damage has been observed in preclinical studies with blueberry flavonoids, strongly indicating the occurrence of PP-mediated antioxidant activity [96 ]. The size of the PP, and the absence of intestinal enzymes to break down the molecules into smaller, absorbable metabolites leads to poor absorption in the proximal small intestine. Only certain gut microbes perform this chemical transformation; thus, the generation and absorption of several phenolic acids occur with increasing microbial load towards the distal small intestine, particularly the colon. Microbial metabolism of PPs results in simple phenolic com- pounds that are suggested to have benefi cial cardiovas- cular and brain e ffects, presumably mediated by e ffects on endothelium-dependent vasodilation, a reduction of blood pressure, inflammation, and platelet activation. Figure 1. Gut interaction and health effects of polyphenols. The small intestine absorbs a limited amount of intact dietary polyphenols into the systemic circulation system where they probably exert a small transient antioxidant e ffect on the body and the brain. Most of the polyphenols ingested into the body interact with the gut microbiome (prebiotic, antibiotic e ffects) and go on to be metabolized into small, phenolic compounds. A quantity of these metabolites have exhibited bene ficial effects on the brain and the cardiovascular system (see text for details).
NUTRITIONAL NEUROSCIENCE 1061 Like other polyphenols, flavanols act as probiotics for several species in the microbial ecosystem in addition to the e ffects of their metabolites on the host. We hypothesize that similar to the fiber-induced increase in diversity and richness of the gut microbiome, eating a flavanol-rich diet will increase microbial diversity, including the increase in microbial strains that can metabolize them.
In summary, based on such mechanistic studies in animals, the evidence indicates that flavonoids may influence brain health through modulation of the gut microbiota, which may lead to increases in mucus and SCFA-producing gut microbes, improved gut barrier function, and a reduction in systemic and brain in flam- mation. Major limitations of the quoted studies include small sample size and the restriction to only male rodent animals. Future studies should be conducted in larger animal cohorts with different species and mixed genders to provide a more comprehensive understanding of the outcomes and to assess potential gender-speci fice ffects. Although the quoted studies provide valuable insight into the potential e ffects of PPs on the gut microbiome and their potential therapeutic e ffects in treating cogni- tive decline and neurodegenerative disorders, translat- ing animal studies to clinical human outcomes is often challenging, due to the disparities in gut microbial com- position, metabolism, and responses to interventions. Mirroring the true complexity and variability of human physiology and gut microbiota may not be poss- ible through mouse and/or rat models, while pigs may be a better animal model. Ultimately, well-controlled clinical trials are required to reproduce these findings in human cohorts and demonstrate clinically meaning- ful e ffects of PPs on the human gut microbiome and their subsequent impact on cognitive decline and neu- rodegenerative disorders.
Clinical evidence supporting the potential effect of PPs on cognitive function The suggested daily consumption of PP is approxi- mately 1 g/day [97 ]. Several factors infl uence the daily intake of PP among individuals, such as dietary habits, geographical location, sex, age, and socio-economic background. A 2018 review found that total PP intake based on geographical location was highest in Japan (1500 mg/day) in comparison to North America and South America, and European countries (800 and 900 mg/day, respectively). Total flavonoid intake was greatest in Australia and Poland (600 mg/day), com- pared with lower daily intakes in other countries, for example, the USA and South America (200– 400 mg/ day) [98 ]. The daily PP intake of the Mediterranean diet is approximately 664 mg/day, with flavonoids and phenolic acids being the primary subclasses (259 and 363 mg/day, respectively) [99 ].
Nutrition is key in preventing and slowing cognitive decline [100,101]. Evidence from studies evaluating the effects of the MIND and DASH diets indicates the pro- tective e ffects of PPs on cognitive decline [17 ,102,103] and lower AD risk in both men and women [104 – 106]. MIND concordance scores from zero to one were determined to evaluate the intake of PPs by the study participants. An annual clinical evaluation was conducted whereby foods from the previous 12 months were recorded. The low frequency of administration and the recall bias associated with this questionnaire, along with the self-report of diet, are signi ficant limitations of this study which likely biased these results [17,102,103]. Despite these limitations, PPs are likely important mediators of the reported dietary e ffects [107–109], with a 42% increase in flavonoid intake (from 296.8 mg/day to >421 mg/day), speci fically monomeric flavan-3-ols (or catechins), oligomeric flavan-3-ols (or procyanidins), and anthocyanins being associated with lower AD risk [110 ]. Based on mechan- istic studies in animal models, the observed PP e ffects are likely to be related to the anti-in flammatory and vasodilatory action of PP metabolites on the dentate gyrus and other brain regions [111 –113] implicated in AD pathogenesis [ 114,115]. Multimodal brain imaging, a biomarker of structural and functional brain integrity [116] has contributed to our understanding of the neu- robiology of cognitive function and disease progression in AD [ 115,117,118].
Cognitive decline and neurodegenerative disorders are associated with functional and structural alterations in the hippocampus, in particular the dentate gyrus, anterior/posterior cingulate, insula, striatum, brainstem, default mode, and executive control networks [ 119– 122]. In a study that included stroke-free male and female participants from the Framingham Heart Study Offspring Cohort ( n = 2086, mean age 60.6 years), white matter hyper-intensities volume, a measure of brain degeneration, was inversely associated with a 120.6% increase in total daily flavonoid consumption (>525 mg/day), including flavan-3-ols, monomeric flavan-3-ols (or catechins) and oligomeric flavan-3-ols (or procyanidins) [123 ].
Table 1 provides a list of some of the clinical studies supporting a potentially beneficial effect of PP on cogni- tive function. One PP intervention study conducted in 90 elderly individuals (aged 64– 82 years old) with mild cognitive impairment showed that high cocoa flavan-3-ol intake (993 mg/day, 8 weeks) improved cog- nitive function, as measured by a verbal fluency test 1062 F. C. ROSS ET AL.
Table 1. Description of the main findings of the clinical studies supporting a potential e ffect of PP on cognitive function. Main active polyphenol(s)
Study population Type of study Age range / gender split Dose Duration of treatment Findings Reference Flavonoids 2801 Observational 59.1 years / 52% female >421 mg/ day 19.7 years Lower risk of Alzheimer disease and related dementias Shishtar et al. [ 110] Flavonoids 2086 Observational 60.6 years / 53.7% female >525 mg/ day 10 years Reduced brain degeneration resulting in reduced risk of Alzheimer disease and related dementias Shishtar et al. [ 123]
Flavonoids 16010 Observational 30– 55 years / 100% female >684.1mg 6 years Reduced rates of cognitive decline Devore et al. [ 32] Flavonoids 12741 Observational 35– 60 years / 60.5% female >1280 mg/ day 13 years Improved long-term language ability and better verbal memory Kesse-Guyot et al. Flavonoids 77335 Observational 30– 55 years / 64% female >699 mg/ day 28 years Lower rates of subjective cognitive decline Yeh et al. [ 130] Flavan-3-ols 211 Parallel RCT 50– 75 years / 56.9% female >770 mg/ day 12 weeks Improved hippocampal-dependent list-learning performance Sloan et al. [ 61] Flavan-3-ols 34 Parallel RCT 50– 69 years / 73% female 900 mg/day 12 weeks Enhanced dentate gyrus function leading to delayed age-related cognitive decline, pre and post assessments were carried out through fMRI and cognitive testing Brickman et al.
Flavan-3-ols 90 Parallel RCT 61– 85 years / 58.9% female >993 mg/ day 8 weeks Reduced age-related cognitive dysfunction Mastroiacovo et al. Flavan-3-ols 30 Crossover RCT 18– 45 years / 100% male 681.5 mg/ day >1 week Preserved vascular health during stress through improved peripheral blood flow and prevention of stress-induced endothelial dysfunction Baynham et al.
Flavan-3-ols 18 Crossover RCT 18– 45 years / 100% male 185.5mg >2 weeks Improved brain oxygenation and higher cognitive performance when cognitive demands are high Gratton et al. [ 133]
Quercetin 96 Crossover RCT 25– 65 years / 54.8% female 150 mg/day 11 weeks Lowered blood pressure leading to reduced risk of cardiovascular disease Egert et al. [ 138] Quercetin 72 Parallel RCT 35– 55 years / 100% female 500 mg/day 10 weeks Signi ficantly reduced systolic blood pressure Zahedi et al. [ 139] Quercetin 61 Parallel RCT 60– 79 years / 55.7% female 50 mg/day 24 weeks Reduced age-related cognitive decline through reducing depressive symptoms and improving emotional conditions Nishihira et al.
Anthocyanins 146 Parallel RCT 40– 65 years / 58.2% female 640 mg/day 24 weeks Decrease in plasma IL-1 β levels, serum sVCAM-1 levels, and serum hsCRP levels, leading to lower risk of cardiovascular disease Zhu et al. [ 146]
Anthocyanins 49 Parallel RCT ≥70 years / 49% female 200 ml/day 12 weeks Improved verbal fluency, short-term memory, long-term memory, and reduced systolic blood pressure Kent et al. [ 147]
Anthocyanins 122 Parallel RCT 65– 80 years / 100% female 100 mg/ week 6 months Improvements in cognitive tasks related to working and episodic memory Gonçalves et al. Anthocyanins 37 Parallel RCT ≥70 years / 54% female >259.7 mg/ day 16 weeks Improved visual spatial memory and semantic access Krikorian et al. Anthocyanins 16 Parallel RCT 68– 92 years / 56.3% female 269mg 16 weeks Improved working memory by enhanced neuronal response Boespflug et al. Catechins 12 Observational 70– 98 years / 83.3% female 227mg 12 weeks Improved cognitive performance and reduced cognitive dysfunction Ide et al. [ 154] Catechins 1003 Observational ≥70 years / 57.1% female >2cups/day 12 weeks Reduced prevalence of cognitive impairment Kuriyama et al. NUTRITIONAL NEUROSCIENCE 1063 (VFT), Mini-Mental State Examination (MMSE), and Trail Making Test (TMT) A and B [124 ]. Similarly, sup- plementation with cocoa flavonoids (900 mg/day, 3 months) in a healthy cohort of 34 male and female indi- viduals aged 50 to 69 years old enhanced hippocampal dentate gyrus function in older adults. The authors pro- posed that this mechanism of mitigating age-related cognitive decline may be attributed to a specifi c aug- mentation in capillary density in the dentate gyrus [125]. These e ffects are likely mediated by gut microbial-derived PP metabolites rather than the parent flavonoids.
The NU-AGE study pro filed the gut microbiota in 612 elderly male and female subjects (aged 65– 79 years) on a year-long PP-rich dietary intervention com- pared to the control group. Adherence to the NU-AGE diet was associated with increased short/branched-chain fatty acid production and lower production of second- ary bile acids (BAs). In addition, overall speci fic micro- biome alterations and lower inflammatory markers were positively associated with markers of lower frailty and improved cognitive function [ 126].
In a longitudinal study, 16,010 female participants completed a food intake questionnaire every four years (these participants also participated in the Har- vard Nurses ’ Health study) [ 32]. Questions included how often certain foods, such as blueberries and strawberries were consumed on a weekly basis.
Through multiplication, the information was recalcu- lated into a quanti fiable amount of flavonoid intake. After 20 years, dietary patterns and cognitive tests were performed on the participants in two-year inter- vals. It was found that a higher intake of flavonoids (high flavonoid median of 684.1 mg/day, compared to low flavonoid median of 145.4 mg/day) through strawberries and blueberries was equivalent to approximately 1.5 to 2.5 years of delayed cognitive aging (after adjusting for income). The correlations observed between repeated measures of berry intake (p = 0.2–0.3) were modest, likely indicating genuine changes in berry consumption within individuals over time. These changes could be attributed to vari- ations in berry availability by year and region. Even though performed in a large sample, one major limit- ation of this study is that the questionnaire was com- pleted every four years, which is likely to introduce significant recall bias and results in less accurate or incomplete information due to participants ’ reliance on memory over an extended period.
In a cross-sectional study involving 1003 Japanese male and female subjects older than 70 years, MMSE results showed that improved cognitive function was associated with higher consumption of green tea (>2 cups/day), which is high in catechins [127 ]. In a ran- domized, double-blind, placebo-controlled trial invol- ving 12,741 participants (aged 35 –60 years) and spanning 13 years, intake of flavonoids, catechins, theaflavins, flavan-3-ols, and phenolic acids were associ- ated with long-term language ability and better verbal memory, as assessed using Forward and Backward Digit Span, the RI-48 Cued Recall test, the TMT and phonemic and semantic fluency [128 ]. A recent six- month double-blind randomized controlled trial of 131 participants (44, 42, and 45 participants in the wild blueberry group, placebo group, and reference group, respectively) aged 65 years and above showed that a flavonoid-rich wild blueberry intervention signifi- cantly improved the processing speed among the wild blueberry group. This improvement e ffectively restored their processing speed to a level on par with that of the reference group, as assessed by the Cambridge Neuro- logical Test Automated Battery (CANTAB) and event- related potentials (ERP) [129 ].
A study on flavonoid intake and its e ffects on cogni- tive function followed US female participants from the Harvard Nurses ’ Health study from 2001 –2006 ( n = 49,493 women), and men from a health professionals follow-up study from 1986 –2002 ( n = 27,842) [130 ]. Participants reported their dietary intake in seven semi- quantitative food frequency questionnaires as well as their own cognitive changes. It was reported that lower rates of subjective cognitive decline were associ- ated with self-reported consumption of several flavo- noid-rich foods. However, longitudinal studies have their limitations, especially with respect to the validity of each person ’s subjective evaluation of their dietary pattern and their cognitive function, the limited validity of food frequency questionnaires, as well as other con- founding variables. These can include di fferent concen- trations of PPs actually consumed due to di fferent foods containing varying concentrations. Those who eat more fruits and vegetables, or more PP-containing foods, may lead healthier lifestyles overall. Thus, despite the impressive number of participants, the results must be interpreted with caution and should not be misinter- preted as demonstrating causal e ffects between flavo- noid intake and cognitive function.
Flavan-3-ols Recent studies have replaced self-reported PP intake levels with blood levels of the flavan-3-ol metabolite phenyl-γ-valerolactone, providing for the first-time accurate estimates of PP intake. Cocoa flavan-3-ols have received attention for their e ffects on cognitive function in older adults. One study ( n = 90, aged 61 – 1064 F. C. ROSS ET AL.
85 years, males and females) conducted over a two- month period reported that high cocoa flavan-3-ol con- sumption (993 mg/day) reduced metrics of age-related cognitive dysfunction, compared to the low flavan-3-ol group (48 mg/day), as assessed through changes in sys- tolic blood pressure, homeostatic model assessment for insulin resistance, and plasma isoprostane concen- trations. Cognitive function was evaluated using the MMSE, the TMT A and B, and the VFT. MMSE scores found no di fference between groups, however signi fi- cant improvements ( p < 0.0001) in TMT A and B and VFT scores were found in the high compared with the low flavan-3-ol group, suggesting improved cognitive function through test factors such as greater processing speed, executive function, spatial awareness, and motor skills. The authors suggested these cognitive improve- ments could be attributed to improvements in insulin sensitivity, blood pressure, insulin resistance ( P < 0.0001, respectively), and lipid peroxidation ( P = 0.001) observed throughout the study [ 131]. Several mechanisms have been implicated in the antidiabetic effects of flavan-3-ols, including the glucose absorption rate in the gut, glucose peripheral uptake, glucose secretion, the modulation of beta cell function, the modulation of insulin secretion, and the modulation of the incretin [ 132].
High cocoa flavon-3-ol consumption (770 mg/day, 12 weeks) in 211 healthy male and female individuals (50–75 years) improved hippocampal-dependent list- learning performance in participants with low dietary intake at baseline, compared to the placebo group [61]. Blood samples were taken at the beginning and end of the intervention to test for phenyl-γ -valerolac- tone metabolites. The study also examined brain activity in a small subset of participants ( n = 12) and observed that following cocoa flavan-3-ol consumption, the den- tate gyrus was the only hippocampal region that showed an increase in basal metabolism. Another study also reported a similar finding in that intake of cocoa flavan-3-ols contributed to improved cognitive function associated with increased dentate gyrus function. Indeed, Brickman et al. found that daily consumption of 900 mg cocoa flavan-3-ol supplements for three months in 50 –69-year-old male and female participants (n = 37) enhanced dentate gyrus function, indexed by cerebral blood volume-fMRI measurement and cogni- tive testing [125 ].
Moreover, in a younger cohort ( n = 18, aged 18 –45 years) of male subjects, a high flavan-3-ol daily intake of 681.4 mg through supplementation resulted in not only greater, but faster cerebral tissue oxygenation, compared to the low flavan-3-ol group (4.1 mg/day). This study suggests that the flavan-3-ol-mediated increase in levels of circulating NO plays a role not only in improved peripheral endothelial function but also bene fits cerebrovascular reactivity. Due to the small cohort of only male participants within this study, further analysis in a larger cohort with both male and female participants is needed to validate these results [133 ]. Cerebrovascular reactivity (CVR) to CO 2, mediated by the NO pathway, is widely accepted as a key biomarker of cerebrovascular health and has been closely associated with cognitive function in health and disease states. While widely accepted as a key bio- marker of cerebrovascular health, the underlying mech- anism of CVR is unclear as the blood-oxygen-level- dependent signal is a ffected by multiple parameters, including cerebral blood volume, cerebral blood flow, and oxygen metabolism [134 ]. As reduced CVR may be associated with cognitive decline as a result of vascu- lar dysfunction [134], this study suggests that continued intake of flavan-3-ols may delay age-related decline in later life [133 ]. Similarly, 5-(3 ’.3’-dihydroxyphenyl-γ- Valerolactone), a major flavan-3-ol metabolite associ- ated with green tea intake, was reported to inhibit TNF-α-stimulated adhesion of monocytic cells in the range 7.5– 30 μM to human umbilical vein endothelial cells [135 ]. This suggests that metabolites of green tea PPs, speci fically some phenyl- γ-valerolactones, exert anti-inflammatory effects.
Moreover, a randomized, double-blind, controlled clinical trial also conducted in young healthy male adults (n = 30, aged 18 –45) showed that the physiologi- cal response to acute mental stress was reduced following a dose of 681.5 mg cocoa flavan-3-ol sup- plementation, compared to the low flavan-3-ol group (<4 mg). In this study, an 8-minute Paced-Auditory- Serial-Addition-Task was employed to induce mental stress, while blood pressure, forearm blood flow, cardi- ovascular activity, endothelial function, and brachial blood pressure were measured as physiological responses. Results showed improved peripheral blood flow during stress and associated prevention in stress- induced endothelial dysfunction, suggesting increased flavan-3-ol intake can attenuate stress-induced reduction in vascular integrity [136 ].
In summary, recent well-controlled studies, using the biomarker phenyl- γ-valerolactone as a more accurate measure of flavanol intake have signi ficantly improved the quality and validity of earlier studies using estimates based on food frequency questionnaires. This approach offers a more precise and objective means of assessing dietary flavanol consumption and its potential health effects, thus contributing to a more robust understand- ing of the relationships between flavanol intake and health outcomes. Even though the underlying NUTRITIONAL NEUROSCIENCE 1065 mechanism of action remains unknown, based on pre- clinical studies, it may be due to an indirect e ffect of dietary flavan-3-ols increasing vasodilation and improvement of endothelial function [137 ]. It has also been proposed that flavan-3-ol intake reduces the stress-induced infl ammatory response thereby improv- ing cognitive function, although this requires more research to verify [137 ]. Further research is also needed to determine why flavan-3-ol effects were limited to the dentate gyrus, a very small brain region.
Quercetin Quercetin, a flavonoid commonly found in tea, onions, grapes, and cranberries has been shown to be associated with a reduction in vascular in flam- mation and blood pressure. Supplementation given to 96 participants (42 male, 54 female) for six weeks with 150 mg of quercetin daily lowered the blood pressure of participants with a high risk of cardiovas- cular disease. No signi ficant disparities between sex groups were seen following supplementation [ 138]. It decreased systolic blood pressure by 2.9 mmHg ( P < 0.01) in the subgroup of hypertensive subjects and by 3.7 mmHg ( P < 0.001) in the subgroup of younger adults aged 25 to 50 years. A 500 mg daily dose of a quercetin supplement for ten weeks in women ( n = 72) with type 2 diabetes resulted in a decrease in plasma levels of TNF- α and interleukin compared to the placebo group [139 ], results that are consistent with preclinical studies [84 ,140–142]. Due to its anti- inflammatory e ffect, quercetin may be benefi cial in conjunction with other treatments for neuroin flamma- tory and cardiovascular diseases. A recent study of 70 male participants aged 60 to 79 years showed that increased quercetin intake (50 mg/day) over a period of 24 weeks signi ficantly reduced age-related cognitive decline. The authors speculated that this reduction was mediated by the improvement of depressive symptoms and increased motivation, compared to the placebo group. However, the mechanism by which quercetin- rich onion improves cognition remains unclear, although it may be a consequence of emotional improvement, as conducted by CADi2 (a dementia mass screening test) [143 ].
Anthocyanins Anthocyanins are a subset of flavonoids usually found in berries. Similar to cocoa flavan-3-ols, some studies [ 144,145] have reported that anthocya- nins increase vasodilation and reduce systemic inflammation, two e ffects that may lead to improved cognitive function. Participants with high plasma levels of low-density lipoprotein cholesterol ( n = 150 males, aged 40 to 65 years) who consumed a puri fied anthocyanin mixture (320 mg/d) twice daily for 24 weeks in a randomized, double-blind, controlled study showed decreases in serum levels of IL-1 β sol- uble vascular cell adhesi on molecule-1 (sVCAM-1) and high-sensitivity C-react ive protein (hsCRP) test. Reductions in these systemic in flammatory markers are changes associated with a lower risk of cardiovas- cular disease [ 146].
Several controlled human studies have provided evidence to support the bene ficial e ffect of regular blueberry consumption on cognitive function in older patients with cognitive decline. In a 12-week randomized-controlled study assessing cognitive out- comes in older adults ( n = 49 male and females aged 70 years and over) with mild-to-moderate dementia, consuming 200 ml/day of cherry juice, resulted in improvements in verbal fluency, short-term memory, and long-term memory, as well as a reduction in sys- tolic blood pressure, compared to the placebo group [147]. Numerous studies have also reported that increased anthocyanin intake through blueberry sup- plementation is associated with improved cognitive function [148 ]. A randomized controlled study invol- ving 122 male and female adults (65 to 80 years)
who consumed blueberry-rich extract supplements (1000 mg/day) over six months reported improve- ments in cognitive tasks related to working and episo- dic memory and decreased cardiovascular risk factors [149]. A randomized, placebo-controlled study over 16 weeks in male and female adults (aged 65 years and over) showed that visual-spatial memory and semantic access were signi ficantly improved following blueberry supplementation. Trends in increased pro- cessing speed were also identi fied but these findings were non-signi ficant ( p = 0.08) [150 ]. Consumption of blueberries (269 mg/day of anthocyanin) for 16 weeks in an elderly male and female cohort ( n = 16, aged 68 to 92 years) resulted in greater activation in the left pre-central gyrus, left middle frontal gyrus, and left inferior parietal lobe in participants when per- forming a task that involves the working memory as indexed by fMRI imaging and compared with control participants [151]. A six-month randomized controlled trial in 133 participants with mild cognitive impair- ment aged 65 –80 years investigating blueberry con- sumption (35 grams of powdered sachets/day?) on serum uric acids concentrations found that signi ficant sex-specificd i fferences prevailed during the study, whereby notable disparities between the placebo and blueberry groups between the three-month and six- 1066 F. C. ROSS ET AL.
month intervals were exclusively observed in male participants and not in their female counterparts, thus, highlighting that future studies should include as u fficient sample size that is large enough for analy- sis to be strati fied by gender [152 ].
Catechins Catechins are found in apples, persimmons, berries, tea, and cocoa. Epigallocatechin gallate, commonly found in green tea, is the most researched catechin [ 153]. It is thought that the metabolites derived from EGCG may contribute to the health bene fits associated with cate- chins, but few studies have investigated the independent function of catechins in humans in terms of cognitive function and reductions in infl ammation.
Three months of daily green tea consumption con- taining 227 mg catechins in 12 elderly, two male and ten female, participants (mean age, 88 years) with cogni- tive dysfunction was associated with improved cognitive performance, as measured by MMSE-J score changes [154]. However, the study is limited by its small sample size (n = 12) and should be replicated in a larger cohort for more reliable results. A meta-analysis of 70 male and female participants in 13 randomized controlled trials on the e ffects of catechin intake through green tea consumption improved blood pressure, particularly having a greater e ffect in those with systolic blood pressure ≥ 130 mmHg [ 155]. This is thought to be due to vasodilation induced by increasing plasma NO con- centration [ 155]. Thus, catechins may also improve cognitive function through vasodilation, reduction in arterial blood pressure, and anti-inflammatory effects. Clinical studies to date have o ffered valuable insights into the therapeutic ability of PPs in treating cognitive decline and neurodegenerative disorders. However, although these findings are encouraging, there is still a need for more robust and rigorous randomized con- trolled trials in larger cohorts to replicate and solidify the therapeutic potential of PPs, as well as pave the way for more targeted interventions to combat cognitive neurodegeneration. Additionally, investigating sex- based differences in PP intervention throughout cohorts of future studies may provide further insight into the unique needs of both men and women, enhancing the effectiveness and precision of therapeutic approaches for cognitive decline and neurodegenerative disorders. The gut microbiome The gut microbiota refers to the microbial groups associated with the human gut, while the term micro- biome refers to the catalog of these microbes and their genes. Until recently, the potential bene fits of PP intake on the brain have largely been related to their antioxi- dant e ffects observed in in vitro systems [156 –159]. Both in vitro assays with human gut microbiota and in vivo preclinical and clinical studies have demonstrated the ways in which PPs can modulate gut microbes to promote the growth of benefi cial bacteria while inhibit- ing the growth of pathogenic bacteria [ 160].
Some bac- teria that have been identi fied in the metabolization of polyphenols are Eubacterium ramulus, Bifi dobacterium spp., and Flavonifractor plautii [161]. Based on inter- individual variations in the abundance of microbial taxa able to metabolize PPs, di fferent metabotypes have been identi fied [162–164]. For example, Eubacter- ium is associated with the metabolization of flavonoids, while Lactobacillus and Bifidobacterium release hydro- xycinnamic acids within the colon [ 161]. Ma et al. con- ducted a meta-analysis and systematic review revealing that health-promoting Lactobacillus and Bifidobacter- ium abundance signi ficantly increased by 220% and 56%, respectively, and Clostridium pathogenic species were decreased in abundance following polyphenol sup- plementation [165]. It is also proposed that the phytoac- tive metabolites resulting from the consumption of flavonoids, following their breakdown within the gut microbiome, engage with cellular and molecular targets (such as signaling pathways). This interaction is believed to enhance neuronal connectivity and facilitate improved blood circulation in the brain, including vas- cular and peripheral flow, particularly when in conjunc- tion with exercise [166 ].
Current evidence suggests a role of the gut microbiome in central nervous system (CNS) disorders, with a par- ticular focus on the brain-gut microbiome (BGM) sys- tem playing a distinct function in neurodegenerative processes [167–171]. The realization that the microbiota could be targeted for modulating brain function across the lifespan has been a complete paradigm shift in neuro- science. Various factors that in fluence the composition of the microbiota also modulate brain function and behavior [172]. A human study carried out on 63 healthy adults aged 67–83 years found a correlation between cog- nitive performance (spatial working memory and paired- associate learning measured by (ERP and CANTAB) and alpha diversity of gut microbiota. Furthermore, the research postulated that diminished cognitive perform- ance (i.e. as an indicator of cognitive impairment) was associated with reduced diversity in the gut microbiome [173]. In brain aging, there is a growing appreciation of the role of the microbiome, although this is still largely restricted to animal studies [ 174–176]. It is proposed that many of the bene ficial effects of nutrition on brain health may be mediated by microbially generated metab- olites [48,172,176]. For example, phenyl-γ-valerolactones produced exclusively by the gut microbiota from flavan- 3-ol-rich foods exert anti- β-amyloid oligomer activity NUTRITIONAL NEUROSCIENCE 1067 and reduce memory deterioration, as well as neuroi- nflammation, in vivo [55]. Likewise, xanthohumol, quer- cetin, and phlorotannin extract reduced anxiety and depressive symptoms in a rat model. In this study, the phenolic-derived gut microbial metabolite, xanthohu- mol, profoundly altered microbial diversity and compo- sition, and improved brain-derived neurotrophic factor plasma levels [177]. Similarly, a preclinical study showed that amyloid-beta load was reduced by urolithin A, a gut metabolite produced from the metabolism of ellagic acid. This was shown to improve spatial memory, exploratory behavior, and fear response in mice [ 178]. Additionally, S-equol, a phenolic-derived gut metabolite of soy iso fla- vone daidzein, has shown therapeutic evidence in precli- nical and some small-scale clinical studies of targeting arterial sti ffness, tau phosphorylation, reducing amy- loid-beta induced neurotoxicity, as well as having antiox- idant and anti-in flammatory e ffects [as reviewed in [179]].
In summary, the intricate interplay between PPs and the gut microbiome is a promising avenue in the prevention and treatment of cognitive decline and neurodegenera- tive disorders. With a greater understanding of the BGM system, the microbiome-modulating potential of PPs is a compelling strategy to potentially mitigate the onset of neurodegenerative disorders.
Conclusions Largely plant-based diets such as the traditional Medi- terranean diet, and the related DASH and MIND diets have been associated with a wide range of brain and behavioral, cardiovascular, and metabolism-related health benefits based on a large body of epidemiological evidence (reviewed in [13 ]). However, future well-con- trolled and strategically designed intervention studies are essential to unequivocally identify a causal role for specific dietary components and these benefi ts for PP. While many factors may be involved, the main mediators are likely to be dietary fiber and PPs and their interactions with the gut microbiota and the gut- associated immune system. PPs and fiber are both large molecules that require microbial metabolic trans- formation into smaller molecules for absorption in the distal, small, and large intestines. The benefi cial e ffects of di fferent members of the flavonoid family of PPs on cognitive function have been demonstrated in a number of human largely epidemiological studies and some have identified the brain changes underlying these benefi ts isms responsible for flavonoid-associated bene fits in brain function, as well as cognitive and behavioral par- ameters, have been studied mainly in animal models. These include indirect brain e ffects involving vascular, endothelial, and immune activation in the brain, many of which may be mediated by microbiome-related immune signals, such as infl ammatory cytokines, anti- inflammatory SCFAs, lipopolysaccharides, and microbe-associated molecular patterns. In addition, a number of neurotransmitters and hormones a ffecting the brain are either generated by or under partial con- trol of certain gut microbes. Examples of microbe- derived neuroactive molecules include tryptophan metabolites, kynurenine, indoles, and certain secondary BAs. However, the number and amount of these neu- roactive and hormonal microbial metabolites reaching the brain show signi ficant inter-individual variability. Furthermore, limitations remain as the great majority of existing studies fail to measure the great effect that confounding factors may have on PP concen- trations. When conducting global intervention studies, it is essential to account for the diverse food compo- sitions and the e ffect that parameters such as cooking, pesticides, food ripening, climate, etc. have on the over- all results of the study. Thus, knowing the systemic con- centration of biomarkers, such as the flavan-3-ol metabolite phenyl- γ- valerolactone or hippuric acid is currently the only reliable way to minimize these con- founders [181 ]. Additionally, human studies directly measuring the strains of bacteria present, such as Bifido- bacteria, Lactobacillus, Eubacterium, and Akkermansia, should be conducted to investigate their modulatory effects on the impact of dietary PPs on cognitive benefits. Now that scientists appreciate the essential role the gut microbiota plays in transforming ingested PPs into absorbable metabolites, future high-quality controlled clinical trials that take inter-individual di ffer- ences in gut microbiota composition into account should help unravel how this class of botanicals may play signi ficant roles in brain health.
Disclosure statement No potential conflict of interest was reported by the author(s). Funding The authors were funded in part by Science Foundation Ire- land, APC Microbiome Ireland [grant number SFI/12/RC/ 2273] and National Institutes of Health grant R01 DK064539. Data availability statement All data used in this manuscript is publicly available and refer- enced throughout.
R. P. Ross http://orcid.org/0000-0003-4876-8839 1068 F. C. ROSS ET AL.