Introduction

Worldwide, 50 million people are living with Alzheimer’s disease (AD) and related dementias, and this figure is projected to triple by 2050 unless preventive measures are developed [1]. Addressing modifiable risk factors is considered to be the most promising strategy to prevent AD [1]. In this regard, increasing evidence suggests that dietary interventions have potential to protect against cognitive decline during ageing [2,3,4].

The mechanisms governing dietary influences on cognition are not clear, but inflammatory pathways are likely to be involved [5, 6]. Inflammation has been strongly implicated in the pathogenesis of AD [7••, 8]. Microglial macrophages in the brain become chronically activated during ageing [9], and particularly under pathological conditions [10], to promote sustained production of pro-inflammatory cytokines including interleukin-1β, interleukin-6 (IL-6) and tumour necrosis factor-α (TNF-α) [8, 11]. Production of these molecules can perpetuate a cycle of neuroinflammatory processes including amyloidosis, neuronal death [10, 12], cortical thinning [13, 14], reduced brain volume [13], cerebral vascular disease-related events such as microbleeds and infarcts [15, 16] and neurodegeneration [7••]. Genetic studies provide strong evidence for a causative role of neuroinflammation in AD, with several mutations in microglial genes, for example, TREM2 and CD33, being independently associated with elevated AD risk [17]. Furthermore, greater numbers of activated microglial cells and cytokines have been confirmed in brain tissue from individuals with AD [18].

Diet is known to modulate the immune system [19••], and several nutrients and bioactive components can influence neuroinflammatory processes in animals. For example, polyphenols, unsaturated fats and antioxidant vitamins inhibit oxidative stress and neuroinflammation [20, 21••, 22], while saturated fat promotes inflammation, particularly in the hypothalamus [23]. However, it is not clear whether diet-induced effects on neurocognition are mediated directly by neuroinflammatory processes and/or via other immune mechanisms in vivo. An increasing body of evidence suggests that peripheral inflammation and alterations to the gut microbiome can amplify neuroinflammation and accelerate neurodegeneration [7••, 24, 25] and these external factors can also be influenced by diet [26].

Mechanistic studies in the brain have primarily focused on single nutrients. However, the synergistic effects of nutrients and foods when consumed together as a usual dietary pattern are likely to exert greater effects on inflammatory processes and neurodegeneration during ageing [25, 27]. There has been much interest in examining the Mediterranean diet (MD) and dietary approaches to stop hypertension (DASH) dietary patterns as potential strategies for dementia prevention due to their proven anti-inflammatory [28,29,30] and cardioprotective effects [31, 32]. The MD and DASH have shown promising associations with slower rate of cognitive decline [33, 34] and reduced AD risk [35,36,37,38] although results have not been consistent [39, 40] and evidence for a causal relationship is limited due to the small number of diet intervention studies conducted. Interestingly, more recent neuroimaging studies have shown protective effects of MD on neuronal structures and early morphological changes linked to neurodegeneration and AD [41,42,43,44,45,46,47,48].

This review aims to summarise available evidence to evaluate the role of MD and DASH dietary patterns for cognitive health and AD risk during ageing and discuss underlying inflammatory mechanisms of diet on cognitive function in humans and animal models.

Dietary Patterns and Cognitive Health During Ageing

The Mediterranean Diet

The MD describes the traditional dietary pattern consumed in Mediterranean countries and is characterised by high intake of fruits, vegetables, wholegrains, nuts and legumes; moderate intake of fish, poultry and alcohol (particularly red wine, with meals) and low intake of red and processed meats with olive oil used as the main fat source [49, 50]. Adherence to the MD has demonstrated clear benefits for both primary [32] and secondary [51] CVD prevention. This may also be the case for cognitive function. A recent pooled analysis of available prospective data in older adults reported a beneficial association between adherence to the MD and cognitive function, especially for domains of global cognition and episodic memory [52].

Observational studies that have evaluated the MD for cognitive health are summarised in Table 1. In summary, greater concordance to the MD has been associated with better cognitive performance [53, 54, 57, 58, 65], slower rate of cognitive decline [61, 66,67,68] and reduced risk of cognitive impairment [53, 55, 56, 61] and AD [36]. However, the variable and sometimes disparate study findings [27, 59, 60, 62,63,64] are likely due to differences between studies in the populations investigated and methods used to assess diet and cognition. Most epidemiologic studies have examined MD adherence using a score system derived from population-specific median reported intake thresholds for each individual food component in the MD score and this approach limits the comparability of findings with other populations [53]. Furthermore, there is heterogeneity between studies with regards to the cognitive outcomes measured, follow-up time and population characteristics including baseline cognitive status.

Table 1 An overview of studies associated with anti-inflammatory dietary patterns
Full size table

It is possible that MD contributes to neuronal integrity across the life-course. Emerging evidence from observational studies report a link between greater MD adherence and more favourable brain structures and functions that protect against neurodegeneration including increased cortical thickness [45, 48], greater brain volumes [42], slower rate of hippocampal atrophy and improved structural connectivity [43, 44, 46] as well as less amyloid (Aβ) accumulation at both midlife and older age [41, 47].

The effect of MD on neurocognition has been evaluated in a limited number of randomised controlled trials (RCTs). The PREDIMED study demonstrated a modest beneficial effect of increased MD adherence over 4–6 years on cognitive function in cognitively healthy adults at high CVD risk [2, 4, 71] particularly in domains of global cognition, memory and executive function [4]. In contrast, adopting a MD over 6 months had no effect on cognitive function in healthy older Australian adults [72]. The recent NU-AGE trial also reported no benefit of a Mediterranean-style diet on cognitive function after 1 year in older European adults; however, participants with greatest MD adherence demonstrated improved global cognition and episodic memory compared to those with low adherence [73].

The Dietary Approaches to Stop Hypertension Diet

The DASH diet is an accepted non-pharmacological treatment for hypertension [31] and, like the MD, recommends a high intake of fruits, vegetables, nuts and wholegrain products. However, in contrast to a MD, DASH places greater emphasis on low fat dairy foods, low dietary sodium and does not recommend alcohol. In older adults, higher DASH scores have been associated with better cognitive function [69] and slower cognitive decline [33] but, again, findings have not been consistent [70]. There is limited evidence from intervention studies for an effect of DASH on neurocognition. The ENCORE study showed improved cognitive function in response to a calorie restricted DASH diet among overweight adults with hypertension [3], while preliminary 6-month data from the ENLIGHTEN study in cognitively impaired adults demonstrated no benefit of DASH diet alone on cognition, but improvement in executive function among those consuming DASH combined with aerobic exercise [74].

Collectively, the available evidence suggests that MD and DASH patterns may provide protection against neurodegeneration during ageing, with more consistent associations in favour of the MD, probably owing to a greater number of studies that have examined MD and cognition, relative to DASH or indeed other dietary patterns. Further prospective studies in diverse populations are recommended to determine relations between dietary patterns and clinically relevant measures of cognitive decline as well as incident AD. While few dietary intervention studies have been conducted, beneficial effects of MD on cognition are shown in trials of longer duration (12 months or more) [2, 4, 60] and in participants who adhere more closely to the intervention diet [63]. Further research is required to determine the duration of dietary intervention needed for optimal effect on cognition and to identify cognitive end-point measures that can respond to subtle diet-induced changes in the ageing brain. Furthermore, intervention strategies would benefit from learning how best to support behaviour change towards a healthy dietary pattern in different population groups.

The MD and DASH dietary patterns are typically low in saturated fat, sugar and high in antioxidants, fibres and polyphenols, which have potential independent effects on brain health. It is likely that the myriad of bioactive compounds and nutrients consumed at synergistic levels within these dietary patterns exert the potential to reduce neuroinflammatory processes involved in neurodegeneration [75].

Inflammatory Mechanisms of Dietary Action on Cognitive Function

Effects of Diet on Neuroinflammation

Nutrients have important physiological roles for normal brain functioning and are transported into the brain via the blood–brain barrier (BBB) or from the choroid plexus transport locus of the blood-cerebrospinal fluid barrier by distinct mechanisms, such as facilitated diffusion and active transport [76]. The effect of whole dietary patterns on neuroinflammation is yet to be discerned; however, animal studies have consistently reported anti-neuroinflammatory effects of several nutrients typically eaten in high amounts in MD and DASH diets. Antioxidants from fruit and vegetables suppress neuroinflammatory processes and neuronal apoptosis by inhibiting free radicals and cytokine production in activated microglia cells [22, 75, 77] and plant-derived flavonoids are thought to play a role in preventing neuroinflammation by downregulating transcription factor activity, e.g. NF-KB and inhibiting pro-inflammatory cell signalling pathways [78, 79]. Long chain omega-3 fatty acids from fish attenuate the expression of pro-inflammatory cytokines in microglia and help to resolve inflammation in the brain [20, 80].

While these animal studies provide important insights into mechanistic actions of diet on neuroinflammation, it is not clear how well the findings translate to humans, due to potential differences in nutrient bioavailability and metabolism. Indeed, single nutrient supplementation has shown no cognitive benefit among older adults in the few trials that have been conducted so far [81•, 82•], although positive signals have been observed for individuals with low baseline nutrient status [83] and for fish oil supplementation in early cognitive impairment [84], suggesting that population subgroups may derive cognitive benefit from nutrient supplementation.

Indirect Inflammatory Actions of Diet on Neuroinflammation

Systemic Inflammation

Increased peripheral inflammatory markers have been associated with neurodegeneration [85] and are suggested to increase neuroinflammation via neuronal and hormonal pathways, as previously reviewed [7••, 19••]. The MD and DASH diet have demonstrated anti-inflammatory effects in humans. In a recent meta-analysis of six RCTs, DASH significantly decreased serum (high-sensitivity C-reactive protein (hs-CRP) concentration (mean difference − 1.01: 95% CI − 1.64, − 0.38; I2 = 67.7%) compared to usual diet which tended to be of greater magnitude in trials with longer duration [30]. The MD has been shown to decrease a broader range of inflammatory biomarkers. The PREDIMED study reported reductions in cytokines (IL-1, IL-6, IL-8, IL-12p70, CRP, TNF-α) and chemokines (MCP-1 and macrophage inflammatory proteins (MIP-1β)) in response to a 3-month MD supplemented with either olive oil or nuts, and the anti-inflammatory effects were observed up to 5 years of intervention [28, 29]. Adoption of a MD also significantly reduced hs-CRP, IL-6, IL-7, and IL-18 in patients with the metabolic syndrome [86]. Furthermore, key components of the MD, such as oily fish and omega 3 fatty acids, as well as bioactive polyphenols found in fruits and vegetables, red wine and olive oil have also been shown to reduce pro-inflammatory markers [78, 87, 88••]. Hence, the ability of dietary patterns, particularly the MD, to decrease systemic inflammation may help to attenuate neuroinflammation.

Emerging observational data lend support to systemic inflammation as a driver between diet and neurocognition (Table 2). High pro-inflammatory dietary scores have been linked with poor cognitive performance [91, 94] and cognitive impairment [92, 93]. Moreover, a proinflammatory dietary pattern, based on circulating IL-6 and characterised by higher intake of red meat, processed meat, peas and legumes, and fried food, and lower intake of whole grains has been associated with accelerated cognitive decline [89], while an inflammatory nutrient pattern derived from IL-6 and CRP levels and characterised by low intake of calcium, antioxidant vitamins, omega-3 and high intake of cholesterol, has been inversely associated with brain volume and cognitive function [90]. Further longitudinal studies are needed to evaluate whether dietary patterns based on systemic inflammatory biomarkers are related to incident AD.

Table 2 An overview of studies associated with pro-inflammatory dietary patterns
Full size table

The Gut Microbiome

The human gut microbiome represents the collective genomes of 10–100 trillion microorganisms harboured in the gastrointestinal tract and is considered important for healthy immune function [95]. The ageing process and other environmental factors can result in alterations to the microbiome composition (dysbiosis) and contribute to the development of chronic low-grade inflammation [96]. Dysbiosis stimulates excretion of endotoxins, e.g. lipopolysaccharides (LPSs) and microbial amyloids to promote permeability of the gut wall and increase peripheral circulation of proinflammatory cytokines [97]. Dybiosis has been implicated in the pathogenesis of AD by initiating and prolonging neuroinflammatory processes. Gut-derived bacteria and toxins can compromise the integrity of the BBB and contribute to early neuroinflammatory changes and AD by priming microglia and impairing amyloid clearance [24, 25]. Moreover, circulating LPSs and microbial amyloid activate innate resistance receptors, e.g. toll-like receptor (TLR) and receptor for advanced glycation end-products (RAGE) to amplify pro-inflammatory signalling and promote chronic neuroinflammation and progress neurodegeneration, particularly in brain regions sensitive to AD such as the hippocampus [25, 98].

Almost 60% of variation in the gut microbiome is attributable to diet [99]; therefore, modulating the gut microbiome through dietary means could be an effective approach to reduce inflammation associated with AD. However, only few studies have evaluated dietary patterns and gut microbiota. Preliminary data has shown positive associations between the MD and increased number of beneficial microbiota species, e.g. Bacteroidetes and their short chain fatty acid metabolites [100, 101] that have anti-inflammatory effects [26]. Further research is needed to understand the complex relationships between the gut microbiome and cognitive health and whether diet-induced effects on cognitive function are mediated by alterations in gut microbiota.

Conclusion

Given the lack of effective treatments and projected increased prevalence of AD, there is considerable interest in understanding the contribution of neuroinflammation to the pathogenesis of AD in order to develop effective preventive strategies for cognitive decline.

Compelling evidence shows that nutrients and other bioactive dietary compounds influence neuroinflammatory processes leading to neurodegeneration in animals and that nutrients can act synergistically to exert greater biological effects. It is plausible that diets rich in anti-inflammatory components attenuate neuroinflammation via several immune pathways within the brain and indirectly from the gut microbiome and systemic circulation. However, data from human studies are lacking and the exact inflammatory mechanisms linking diet to cognitive function remain elusive.

Growing evidence supports a protective effect of anti-inflammatory dietary patterns, especially the MD, against cognitive decline in older persons but causal associations between diet and AD remain uncertain. The evidence base has been recently strengthened by small intervention studies showing improvements in cognitive function in response to MD and calorie-controlled DASH. However, adequately powered intervention studies with larger sample sizes and longer durations are required to examine the effect of dietary modification on clinically relevant cognitive outcomes. Neuroimaging studies have shown direct beneficial associations of MD on preclinical changes associated with AD; therefore, brain biomarkers should be considered as end-points in future intervention studies to investigate diet-associated neurological change from the earliest through to latest stages of cognitive decline. Furthermore, incorporation of PET (positron emission tomography) could allow measurement of in vivo microglia activation [102] and advance current understanding of the mechanistic effects of dietary modification on neuroinflammation in the ageing brain.

While most research attention has been given to MD and DASH, there is still much to learn about the ideal combination of foods and nutrients for optimal cognitive health during ageing. More recently, the MIND (Mediterranean-DASH diet intervention for neurodegeneration delay) dietary pattern that incorporates foods based on evidence in the diet-dementia field has shown to be more predictive of cognitive decline than the MD [103] and has also been associated with reduced AD risk [104]. The effect of MIND on cognitive decline is currently being evaluated in several US trials (e.g. clinicaltrials.gov Reference NCT02817074, NCT03688126) and will generate valuable data to determine whether improving diet quality is an effective strategy to improve brain health in older adults.