Abstract
Alzheimer’s disease (AD) is an irreversible neurodegenerative disease characterized by brain function disorder and chronic cognitive function impairment. The onset of AD is complex and is mostly attributed to interactions between genetic factors and environmental factors. Lifestyle, dietary habits, and food consumption are likely to play indispensable functions in aged-related neurodegenerative diseases in elderly people. An increasing number of epidemiological studies have linked dietary fatty acid factors to AD, raising the point of view that fatty acid metabolism plays an important role in AD initiation and progression as well as in other central nervous system disorders. In this paper, we review the effects of the consumption of various dietary fatty acids on AD onset and progression and discuss the detrimental and beneficial effects of some typical fatty acids derived from dietary patterns on the pathology of AD. We outline these recent advances, and we recommend that healthy dietary lifestyles may contribute to preventing the occurrence and decreasing the pathology of AD.
INTRODUCTION
There are over 50 million people living with de-mentia around the world, and this number is expected to reach 150 million in 2050 [1]. Epidemiological studies have indicated that Alzheimer’s disease (AD) is the main form of dementia, accounting for 50∼56% of all clinical cases [2]. AD is a kind of degenerative disease of the central nervous system (CNS) in elderly people. AD patients perform irreversible brain disorders and suffer from chronic cognitive function impairment. The criteria of progressive decline for AD include at least two cognitive aspects: impaired memory, along with language changes, visuospatial or executive dysfunction [3]. This cognitive decline is bound to affect daily life; ultimately, patients may become bedridden, mute, incontinent, and unresponsive [4]. With the growth in the aging population, AD has become a serious global problem. However, to date, there is still a lack of disease-modifying treatments [5, 6].
Emerging evidence indicates that dietary factors could affect the risk or pathological process of AD [7–11]. Among the components of food, fatty acids play indispensable roles in human nutrient demand and healthy regulation. Functions of fatty acids in the CNS include energy source, cell membrane composition, axonal growth and function regulation, and neuroinflammation response [7, 13]. The brain is the second-most lipid-rich organ in the human body, and it is the organ with the highest diversity of lipid composition in the human body [14, 15]. Because of these unique properties, lipid homeostasis is one of the key points of AD and other CNS disorders [16].
In this review, several typical dietary fatty acids, including ω-3 and ω-6 polyunsaturated fatty acids, monounsaturated fatty acids, medium-chain fatty acids, saturated fatty acids, trans and conjugated fatty acids in foods, are described and summarized; then, the correlation between these fatty acids and AD is analyzed. Furthermore, the possible molecular mechanisms of these fatty acids in AD onset and pathology are discussed and proposed. We are looking forward to these recent study advances that may provide a certain theoretical basis for the prevention and treatment of AD.
PATHOPHYSIOLOGY OF AD
The main pathological features of AD include the extracellular deposition of amyloid-β protein (Aβ) into neuritic plaques and the formation of neurofibrillary tangles (NFTs) associated with aggregated hyperphosphorylation of tau protein [17, 18]. Evidence associated with amyloid protein suggests that an imbalance in the production and clearance of Aβ is the main cause of AD pathogenesis [19]. Aβ is produced by sequential cleavage of amyloid-β protein precursor (AβPP) through β-secretase enzyme 1 (BACE1) and γ-secretase enzyme [20]. After secretion, extracellular Aβ can aggregate to form oligomers and neuritic plaques. The pathological deposition of Aβ induces synaptic toxicity, mitochondrial dysregulation [17, 21], inflammatory responses, and the activation of glial cells [22] as well as the facilitation of tau pathology [23, 24]. Although the accumulation of Aβ in the brain begins 15∼20 years earlier than clinical AD symptoms occur [25], Aβ plaques may create a unique environment that allows for rapid seeded induction of tau pathology from endogenous tau [23]. Tau aggregation is highly associated with AD onset and neurodegenerative processes as well as synaptic dysfunction and synapse loss [6]. Tau is recognized as a microtubule-associated protein in neurons, taking part in microtubule aggregation and assembly. Tau participates in multiple molecular pathways, including regulating neuronal cell signaling and synaptic plasticity [26]. Tau protein is phosphorylated at 85 different residues [26]. Under physiological conditions, each tau protein carries 2-3 phosphate groups, participating in microtubule formation along with tubulin [27]. However, under AD pathological conditions, the hyperphosphorylation of tau protein decreases the binding of tau to microtubules, resulting in axonal degeneration, the loss of synaptic neurons, and neuronal function damage [17]. The presence and pathology of tau seem to be predictors of cognitive impairment in AD, whereas Aβ seems to be a predictor of further tau-associated cognitive impairment [28]. Above all, both Aβ and tau pathology are clearly critical in AD pathogenesis, yet the exact mechanistic link between the two AD lesions is still unclear [5, 23].
Although most AD occurs on an apparently sporadic basis, a small portion of AD (<5%), in which symptoms appear before the age of 65 years, is commonly referred to as early-onset AD (EOAD). EOAD is associated with mutations in genes involved in Aβ metabolism, such as APP, presenilin 1 (PSEN1) and presenilin 2 (PSEN2). Typically, most AD occurs at age 65 and older and is called late-onset AD (LOAD), which is likely to be driven by a complex interplay between genetic factors and environmental factors [29]. The apolipoprotein E (APOE) gene is the most influential genetic factor in sporadic AD: the APOE ɛ4 allele is associated with increased risk for AD [29, 30], whereas the ɛ2 allele is associated with decreased risk [31]. Recent advances in genetic risk factors for LOAD and their role in AD pathogenesis have been unveiled and implicate inflammatory, cholesterol metabolism and endosomal-vesicle recycling pathways [32].
CLASSIFICATION AND COMPOSITION OF DIETARY FATTY ACIDS
Generally, natural dietary fatty acids can be classified as short-chain fatty acids (SCFAs; 1–6 carbons), medium-chain fatty acids (MCFAs; 7–12 carbons), and long-chain fatty acids (LCFAs;>12 carbons) according to the number of carbon atoms [33] and as saturated fatty acids (SFAs; no double bonds), monounsaturated fatty acids (MUFAs; one double bond), and polyunsaturated fatty acids (PUFAs; more than one double bond) according to the number of unsaturated double bonds. Most food-derived fatty acids are LCFAs, whereas dairy products and some plant oils (e.g., palm kernel oil and coconut oil) contain large portions of dietary MCFAs [34]. MUFAs are mainly found in olive oil, rapeseed oil, nuts, and seeds [35]. Compared to animal fat, plant and fish oil contain a higher proportion of PUFAs [36]. Briefly, the main dietary sources of fatty acids are listed in Table 1.
Common dietary fatty acids
1C, number of carbon atoms; D, number of double bonds; ω-X, position of the first double bond counting from terminal methyl carbon.
Two kinds of fatty acids are found in the human body: essential and unessential fatty acids. The difference depends on whether the human body can endogenously synthesize them. For example, the human body has the ability to synthesize SFAs and most MUFAs de novo from glucose and amino acids, whereas in most cases, the diet has already supplied abundant fatty acids for the body’s needs [37]. However, ω-3 and ω-6 PUFAs are essential fatty acids for the human body: due to the lack of enzymes to insert a double bond before the ninth carbon chain position counted from the methyl end, the human body cannot synthesize α-linolenic acid (ALA; 18:3, ω-3) and linoleic acid (LA; 18:2, ω-6), which are precursors of the ω-3 and ω-6 PUFA families, respectively [38]. Although the human body contains enzymes to synthesize ω-3 and ω-6 PUFAs from ALA and LA, most of these PUFAs are obtained from foods [12]. Plant, algae, and marine fishes are natural sources of ω-3 and ω-6 PUFAs [36, 37]. LA is abundant in plant oils, such as in soybean, corn, safflower, and sunflower oils; ALA is rare, but can be found in soybean, canola and flax seed, corn, safflower, and olive oils; and eicosapentaenoic acid (EPA; 20:5, ω-3) and docosahexaenoic acid (DHA; 22:6, ω-3) are abundantly found in fatty fish, such as salmon, trout, and tuna [39].
Except for a very small proportion of free fatty acids, dietary fatty acids commonly exist in triglycerides, which are composed of 1 glycerol molecule esterified to 3 fatty acid molecules. In humans and mammals, adipose tissue contains the largest proportion of lipid mass in the body, where excess fatty acids are synthesized to triglycerides with glycerol and stored in lipid droplets [40]. Milk fat is mainly composed of triglycerides (∼98%) and stored in a kind of double-membrane structure named milk lipid globules [41]. Plant oils come from oilseeds, legumes, nuts, or the flesh of some fruits [42]. Despite the variety of plant oil compositions, one kind of fatty acid usually predominates over the others [43].
FATTY ACID METABOLISM IN THE BRAIN
The brain contains the second highest proportion of lipid mass (after only adipose tissue) and the highest diversity of lipid composition in the human body [14, 40]. Brain lipids are largely located in myelin sheaths, arborized neurons, and astrocytes. In contrast to adipose tissue, where fatty acids are stored as energy, in the brain, fatty acids are essential constituents of various membrane systems for neural cell crosstalk [15]. For instance, phospholipids are the major component of biomembranes (approximately 32∼70 mg/g wet weight of human brain) [44]. A disturbance in brain lipid homeostasis has been linked to brain metabolism disorders and neurodegenerative diseases, including AD [16, 45]. The origin of brain fatty acids includes both in situ synthesis and uptake from blood.
Dietary fatty acid uptake and transport
Dietary fat is absorbed and hydrolyzed by lipases within the intestinal lumen. The released LCFAs are activated by acyl coenzyme A (acyl-CoA) synthetase into acly-CoA, and then re-esterified to form triglyceride molecules by acyl-CoA transferase in small intestinal mucosa, which are packaged into chylomicrons with apolipoprotein, phospholipid, and cholesterol, and delivered primarily to muscle and adipose tissue, with the remaining chylomicrons being transported to the liver [46]. SCFA and MCFA esters are preferentially digested in the intestine and taken up by enterocytes; they are not incorporated into chylomicrons [33, 47].
Because the presence of the blood-brain barrier (BBB) restricts the passage of many compounds into the brain, fatty acids must first move across luminal and abluminal membranes of brain endothelial cells and then arrive at the plasma membrane of neural cells [48]. The uptake of LCFAs by cells occurs through passive diffusion referred to as a “flip-flop” mechanism independent of proteins [49], or protein-mediated transport [50]. For example, nonesterified (NE)-DHA or lysophosphatidylcholine (LysoPtdCho)-DHA are released from plasma lipoproteins via the hydrolysis of lipoproteins at the brain endothelium and then enter endothelial cells through a “flip-flop” mechanism, or facilitated transport mediated by a complex transmembrane protein. The process includes fatty acid-binding protein binding and activation by fatty acid transport protein and acyl-CoA synthetase long-chain to form DHA-CoA in the brain, which participates in various metabolic pathways [48]. In contrast to LCFAs, SCFAs and MCFAs possess higher cell membrane permeability coefficients and do not depend on proteins for binding, transport, or transmembrane translocation [33].
De novo fatty acid synthesis and triglyceride formation
The liver is the most active tissue of de novo fatty acid synthesis and triglyceride formation. De novo fatty acid synthesis starts from the synthesis of acetyl coenzyme A (acetyl-CoA) and (malonyl-CoA) precursors for palmitic acid (16:0), followed by their elongation and desaturation to form more complex fatty acid types. Due to the lack of enzymes to insert a double bond before the ninth position in the carbon chain, counted from the methyl end, the human body cannot synthesize ω-3 and ω-6 PUFAs [38]. Fatty acid synthesis is regulated by factors such as hormones and metabolites. Under normal conditions, these endogenous triglycerides are assembled into lipoprotein (very-low-density lipoprotein) and transported to extrahepatic tissues [51].
The brain has been proven to synthesize several fatty acids in situ. In the developing brain, SFAs and MUFAs, such as 16:0, 18:0, and 18:1, come from de novo biosynthesis, whereas brain essential PUFAs, such as ω-3 and ω-6, are dependent on exogenous uptake into the CNS [52, 53]. However, evidence has demonstrated that peripheral LCFAs, such as palmitic acid (16:0), can be taken up by the brain, thus regulating central palmitate levels [54, 55]; brain uptake of 16:0 has even increased by 80% in the hippocampus of people with metabolic syndrome [55].
Lipoprotein metabolism
Plasma lipoproteins are the main form of plasma lipid vesicles and are complexes composed of a lipid and apolipoproteins. The structure of lipoproteins contains a lipid core (mainly triacylglycerols and cholesterol esters) surrounded by a monolayer of polar lipids (mainly phospholipids and unesterified cholesterol) and apolipoproteins [56]. Apolipoproteins serve as receptor ligands as well as enzyme cofactors and mainly include apolipoprotein A, B, C, D, and E. According to diameter and density, lipoproteins are mainly classified as chylomicrons (diameter > 100 nm), very-low-density lipoproteins (VLDLs; 40–100 nm), intermediate-density lipoproteins (IDLs; 25–35 nm), low-density lipoproteins (LDLs; diameter 20–24 nm), and high-density lipoproteins (HDLs; diameter 8–12 nm) [56].
Both chylomicrons and VLDLs are mainly responsible for triglyceride transport, and they are produced by the intestine and liver [51]. Chylomicrons contain apoB48 and exchangeable proteins (such as apoE and apoC), whereas VLDLs are composed of apoB and multiple exchangeable proteins (mainly apoE and apoC). IDLs are transient metabolic products of VLDLs. LDLs are the plasma intermediate of VLDLs, which are the main transporters of endogenous cholesterol. The major protein of LDL is apoB. LDL can be taken up and degraded by many tissues, especially the liver. HDLs are synthesized by the liver and also result from the metabolism of VLDLs and chylomicrons. The major protein of HDL is apoA-I [57]. HDLs participate in reverse cholesterol transport, which remove excess cholesterol from peripheral cells and then deliver it to the liver by binding to HDL receptors; cholesterol is ultimately excreted as bile acid [58].
In contrast to plasma, lipoproteins in the cerebrospinal fluid (CSF) are composed solely of particles resembling the size and density of HDL, and apoE is a major apolipoprotein that is predominantly expressed by astrocytes and microglia [59]. Lipoproteins in the CNS are secreted largely by glial cells, and as we mentioned before, the APOE gene is the most influential genetic factor in sporadic AD: the APOE ɛ4 allele is associated with an increased risk for AD [29, 30].
β-Oxidation and ketogenesis
Most tissues are capable of catabolizing fatty acids by mitochondrial β-oxidation. LCFAs are first catalyzed by acyl-CoA synthetase to acyl-CoA and then transported to mitochondria by carnitine acyl transferase I and carnitine acylcarnitine translocation. In contrast, SFAs and MCFAs entering the mitochondria do not depend on the carnitine transport system and undergo preferential oxidation [47]. The production of β-oxidation includes FADH2 and FADH for oxidative phosphorylation and acetyl-CoA for the Krebs cycle.
The metabolic activities of the brain, such as synaptic transmission, depend on high energy consumption. The main energy source of the brain is glucose uptake from the blood. It has been reported that the brain utilizes approximately 25% of the total glucose required by the entire body [60]. Although glucose is the primary energy source of the brain, approximately 20% of the total energy expenditures of the adult brain rely on fatty acid oxidation [46]. Despite the abundance of fatty acid content and high energy consumption, the efficiency of brain fatty acid oxidation is considered low compared with that in peripheral tissues [47]. Fatty acid oxidation occurs almost exclusively in astrocytes [46]; however, recent evidence suggests that neurons are able to utilize fatty acids as a source of energy as well [61]. That is, brain fatty acid oxidation seems not to be limited to gliocytes only.
Ketone bodies, including acetoacetic acid, β-hydroxybutyrate, and acetone, are produced from acetyl-CoA. In vivo β-hydroxybutyrate can be formed reversibly from acetoacetate. Ketone bodies act as the primary alternative fuel of the brain when plasma glucose is inadequate. In prolonged starvation, ketone bodies are able to provide up to 80% of brain’s energy demands [62]. Ketogenesis occurs in liver mitochondria and catabolizes the mitochondria of extrahepatic tissues to form acetyl-CoA, thus participating in the Krebs cycle. Recent research has indicated that MCFAs can be utilized to generate ketone bodies and activate shuttle systems by astrocytes, thus providing energy to neighboring neurons [63].
THE LINK BETWEEN FATTY ACIDS AND AD
Ω-3 and ω-6 PUFAs
The ω-3 PUFAs mainly include DHA (22:6, ω-3), EPA (20:5, ω-3), ALA (18:3, ω-3), and docosapentaenoic acid (DPA; 22:5, ω-3). DHA is the most abundant long-chain PUFA in the human brain, constituting over 10% of the total fatty acids in human adult brains and approximately 30% of the total fatty acid content in brain membranes [15]. Inadequate supplementation with DHA leads to poor neural development, neurotransmitter metabolism, and altered learning and visual function [39]. DHA and EPA are abundantly found in fatty fishes, whereas ALA can be found in plants and vegetable oils [39]. DPA is an intermediate metabolite between EPA and DHA, which is elongated by EPA in vivo [64]. The ω-6 PUFAs mainly include linoleic acid (LA; 18:2, ω-6), γ-linolenic acid (GLA; 18:3, ω-6), and arachidonic acid (AA; 20:4, ω-6). Except for ω-3 PUFAs, ω-6 PUFAs are essential components for brain membrane and neurological function maintenance as well [65]. The level of AA is almost comparable to that of DHA, whose content is apparently higher than any other ω-6 PUFAs in the brain [66]. For most people, an adequate amount of ω-6 PUFAs can be acquired from foods, but this is not the case for ω-3 PUFAs.
An increasing number of studies have indicated that the balance of ω-3/ω-6 PUFA intake plays an important role in brain health. The balance between ω-6 and ω-3 PUFA metabolites in the brain is close to 1:1, which meets the demands of brain development and structural integrity [67]. Higher brain DHA accretion probably indicates higher dietary DHA intake; likewise, higher intake of LA compromises DHA accretion [68]. With breastfeeding during the first year of life, higher levels of PUFAs and a higher ω-3/ω-6 PUFA ratio in colostrum are positively correlated with infant mental development [69]. However, modern diets contain larger amounts of ω-6 PUFAs derived from animal fat and plant oils; for example, the ratio of ω-6/ω-3 PUFAs reaches 10∼20:1 in a Western diet, compared with the ratio of 1:1 in the diet of our ancestors [67, 71]. Hence, an imbalance in ω-3/ω-6 PUFA intake is commonly recognized as one of the risk factors for chronic inflammatory diseases.
Clinical studies
Previous epidemiological studies have pointed out the positive link between higher dietary ω-3 PUFA intake and decreased dementia risk in older people [72–76]. Dietary fish consumption or DHA supplementation may have beneficial effects on healthy older people, reducing the risk of dementia and AD [77], whereas reduced dietary ω-3 PUFA intake or lower brain ω-3 PUFA levels are associated with AD occurrence [78]. Nooyens et al. assessed lipid consumption in 2,612 healthy middle-aged and older people (43–70 years) with a 5-year follow-up in the Doetinchem Cohort Study, and they found that a higher dietary ω-3 PUFA intake, especially ALA, was associated with a slower decline in global cognitive function and memory with aging at middle age [79]. The advantages in this study included the prospective design, relatively younger participants and longer follow-up. From these clinical studies, it is found that higher dietary ω-3 PUFA consumption may decrease the risk of cognitive decline or delay the occurrence of AD.
For AD therapy, the ω-3 PUFA intake seems not as effective as reducing the risk of AD occurrence. In a systematic review of Burckhardt et al., three trials, including 632 people with mild to moderate AD, were investigated. These interventions included 1.7 g/d DHA and 0.6 g/d EPA for 6 months (OmegAD trial; NCT00211159) [80], 0.9 to 1.1 g/d DHA for 18 months (ADCS-NIA trial) [81], and 675 mg/d DHA and 975 mg/d EPA for 12 months [82]. The results indicated that there were no effects of ω-3 PUFA supplementation in people with mild to moderate AD [83]. Another study, the Multidomain Alzheimer Preventive Trial (MAPT; NCT00672685) [84], was also conducted. During this 36-month phase III, multicenter, randomized, placebo-controlled trial, older adults with subjective memory complaints but without clinical dementia received ω-3 PUFAs (daily dose of 0.80 g/d DHA and 0.225 g/d EPA) [84]; however, the results indicated ω-3 PUFAs had no significant effects on cognitive decline in these older people [85]. Arellanes et al. assumed that the benefits of ω-3 PUFAs in AD therapy are restricted by brain bioavailability, that is, a low dose (1 g/d or less) DHA supplementation may not provide an adequate brain level to fully evaluate the efficacy cognitive outcomes, and a higher does DHA supplementation (2 g/d or more) is required to ensure adequate brain delivery [86].
Most of all, with consideration of eating habits and food preferences, higher dietary ω-3 PUFA intake seems to reduce the occurrence of AD in aging. However, as a therapeutic regimen, ω-3 PUFA supplementation for dementia requires further studies, perhaps considered by the perspectives of increasing the ω-3 PUFA bioavailability of the brain, such as raising the supplementation dose of ω-3 PUFAs.
Preclinical studies
Accumulating in vitro and in vivo studies verified that higher ω-3 PUFA intake or a higher ω-3/ω-6 PUFA ratio exerts positive neuroprotective effects in AD models, and the potential molecular mechanisms are linked to some critical processes in AD pathology, including Aβ aggregation, tau hyperphosphorylation, neuroinflammation, cell death, and neurotoxicity. We cannot deny that these advances may provide, at the very least, some constructive insights into AD prevention and treatment.
The effects of ω-3 and ω-6 PUFAs may target Aβ processing, fibrillogenesis, and downstream neurotoxicity. A cross-sectional observational study indicated that seafood consumption, at least one meal per week, is related to fewer neuritic plaques and neurofibrillary tangles in brain autopsies, and this phenomenon is mainly observed in APOE ɛ4 allele carriers [87]. An in vitro study found that PUFAs, including DHA, directly interact with both Aβ40 and Aβ42, acting as inhibitors of Aβ fibrillogenesis [88]. Exposure to a ω-3 PUFA-enriched diet reduced Aβ-related AD pathologies after central Aβ administration in rats [89]. In AD transgenic Tg2576 mice, dietary DHA supplementation decreased Aβ accumulation and its potential downstream neurotoxicity [90]. In contrast, a ω-6 PUFA-enriched diet enhanced plasmatic Aβ1–42 levels in Wistar rats [91]. Furthermore, in the study of AD patients, a kind of DHA-derived 10,17S-docosatriene known as neuroprotectin D1 (NPD1) was demonstrated to attenuate Aβ secretion and Aβ-induced neurotoxicity [92].
The fatty acid type and proportion may modulate lipid composition and functionality through lipid raft domains, thus influencing Aβ deposition, which is the critical procedure of AD pathology. Lipid rafts are lipid-derived membrane microdomains that function in signal transduction, protein processing and membrane turnover [93]. These microdomains contain high contents of SFAs, sphingolipids, and cholesterol, and low contents of PUFAs [94]. Compared with healthy brains, the lipid rafts from AD brains displayed aberrant lipid profiles and low levels of ω-3 PUFAs and monoenoic acids (mainly DHA and OA), along with reduced unsaturation and peroxidability indexes [94]. DHA was proven to shift AβPP toward nonamyloidogenic processing with multiple effects, including decreased β- and γ-secretase activity and the promotion of cholesterol shifts from raft to nonraft domains, thus effectively reducing Aβ release [95].
On the other hand, ω-3 PUFAs may modulate tau and synaptic plasticity as well. Dietary DHA was confirmed to reduce tau accumulation in 3×Tg-AD mice [96]. Additionally, ω-3 PUFAs improved memory and learning ability in AD mouse models of tau, reducing the formation of NFTs and Aβ plaques and preventing damage to hippocampal neurons [97]. Furthermore, adequate levels of ω-3 PUFAs may benefit the integrity of brain neurons and enhance synaptic plasticity [98], although the details of how ω-3 PUFAs influence tau-induced synaptic loss are unclear.
Moreover, because ω-3 PUFAs have the potential to regulate inflammation by modulating conditioning membrane composition and function, eicosanoid production, and gene expression [99], ω-3 PUFAs may protect the brain against AD pathology by ameliorating neuroinflammation. Increasing evidence indicates that not only Aβ and tau pathologies but also neuroinflammation have a significant influence on AD pathogenesis [22, 100]. Neuroinflammation is a complex immune response that includes molecular and cellular alterations, peripheral immune cell recruitment, intracellular signal transduction, and inflammatory mediator release in the brain [22]. Eicosanoids are bioactive oxygenated metabolites, such as prostaglandins, thromboxanes, and leukotrienes, derived from PUFAs. Generally, but not all, eicosanoids derived from ω-6 PUFAs, such as AA (20:4, ω-6), are proinflammatory, whereas eicosanoids derived from ω-3 PUFAs, such as EPA (20:5, ω-3), are anti-inflammatory [101]. Dietary intake of DHA may provide protective effects to counter the damage of high levels of AA-derived proinflammatory cytokines through cyclooxygenase and lipoxygenase enzymes, thus improving the neuroinflammatory condition [102]. In a mouse model, supplementation with ω-3 PUFA-enriched fish oil decreased neuroinflammatory gene expression induced by Aβ, indicating that ω-3 PUFAs may participate in the anti-neuroinflammatory pathways in Aβ-mediated AD pathology [103].
In summary, the beneficial effects of ω-3 PUFAs against AD are probably attributed to at least several mechanisms, including decreasing Aβ accumulation, modulating tau and synaptic plasticity and improving neuroinflammation. Likewise, this neuroprotective effect is complex, multiple and interactive. Although more powerful clinical data are needed on ω-3 PUFAs in the treatment of AD, we should not ignore the potential mechanisms of ω-3 PUFAs in AD pathology, which provide insights and possible targets in AD therapy and drug development.
MUFAs and the Mediterranean diet
Dietary MUFAs mainly include palmitoleic acid (PA; 16:1 ω-7) and oleic acid (OA; 18:1 ω-9). The natural sources of PA include salmon, cod liver oil, nuts, and olive oil [104], whereas OA is the predominant fatty acid type in olive oil and is mainly found in olive oil, rapeseed oil, nuts, and seeds [35] (Table 1). Similar to PUFAs, the existence of double bonds provides MUFAs with unique configurations and biological functions distinguished from saturated fatty acids.
Olive oil is a high-quality dietary fat source and is recognized as an essential part and hallmark of the Mediterranean diet (MeDi) [105], a dietary pattern with high consumption of nonrefined cereals and products, fresh fruits, vegetables, nuts, extra virgin olive oil; moderate consumption of fish; low to moderate consumption of dairy products; low consumption of red meat; and regular but moderate intake of red wine [106, 107]. Olive oil contains 70–80% of MUFAs among the total lipid composition [108]. Furthermore, the minor components (comprising only 1-2%) in olive oil, such as phenols and other antioxidant components, also contribute to the health benefits of olive oil [109]. The extra virgin olive oil (EVOO), which is obtained from the first physical separation of the olive paste and is rich in phenolic compound [110], is considered as the highest grade of olive oil for consumption.
Clinical studies
Previous epidemiological studies have provided numerous pieces of evidence that a higher ratio of MUFA intake or an MUFA-enriched MeDi is associated with lower cognitive decline [11, 111–115]; by contrast, patients with AD showed lower MeDi adherence than healthy elderly individuals [116]. This point of view was supported by a recent 3-year follow-up study: in 70 clinically and cognitively healthy participants over 30∼60 years of age, lower MeDi adherence was associated with progressive AD biomarker abnormalities, whereas higher MeDi adherence was estimated to provide 1.5∼3.5 years of protection against AD onset [117]. In a multicenter, randomized, primary prevention trial (PREDIMED), 522 participants aged approximately 75 years received an MeDi intervention plus extra virgin olive oil and nuts, demonstrating an improvement in cognition after 6.5 years the of nutritional intervention when compared with the low-fat control diet [118]. These above data suggested the healthy benefit of the MeDi in cognition improvement; furthermore, the following one-year study among 180 elderly participants based on the MeDi seemed to clarify the role of MUFAs in improving cognition. In this study, the vegetable oils were replaced by extra virgin olive oil at a dose of 20∼30 g per day, aiming at increasing the daily intake of MUFAs; the results indicated that, compared with the control group, which received the MeDi invention alone, the group that received the MeDi intervention with extra virgin olive oil had a higher short-term improvement in cognitive function scores [119].
Preclinical studies
Similar to PUFAs, the beneficial effects of MUFAs in AD may occur through their effects on Aβ metabolism. As previously mentioned, the lipid rafts from AD brains displayed aberrant lipid profiles and low levels of ω-3 PUFAs and monoenoic acids (mainly DHA and OA) and reduced unsaturation and peroxidability indexes [94]; that is, a low level of OA might shift AβPP toward amyloidogenic processing, inducing Aβ release and AD pathology. Through aggregation kinetic experiments, transmission electron microscopy and molecular docking studies, both OA and DHA were identified as effective inhibitors of Aβ40 and Aβ42 fibrillogenesis [88]. In wild-type C57BL/6J mice, after a 9-month administration of low-fat chow (4% fat w/w) or either an SFA-, MUFA-, or PUFA-enriched diet (23% fat w/w), an SFA-enriched diet significantly increased the enterocytic abundance and plasma concentration of Aβ, while groups that received MUFA- and PUFA-enriched diets maintained comparable levels as those that received the control diet [120]. Additionally, an in vivo study in AD transgenic mice unveiled that a high-protein, low-fat, cholesterol-free, and OA-enriched diet decreased the levels of BACE and reduced the number of cerebral amyloid plaques [121]. Furthermore, due to their anti-inflammatory potential [122, 123], MUFAs may contribute to ameliorating neuroinflammation in the AD brain.
It is noteworthy that most of the clinical studies were carried out with a MeDi intervention; that is to say, MUFAs, although acting as the main fatty acid source in the MeDi, may not be sufficient as an independent beneficial dietary factor for AD treatment or prevention. Given the complexity of the MeDi pattern, MUFAs—more importantly, the dietary pattern, low-level SFA intake, and other bioactive compounds, such as polyphenols and vitamins—may coordinate to improve cognitive status and reduce AD onset and pathological processes.
MCFAs and ketone bodies
Increasing evidence suggests that MCFAs and ketone bodies may help to alleviate neurodegeneration and AD pathology [8, 125]. Ketogenic diet (KD) is a typical approach of nutritional ketosis with low-carbohydrate, adequate-protein and high-fat, mimicking the metabolic profile of fasting by reducing blood glucose concentration and increasing blood ketone concentration [7]. Typically, there are two forms of a KD: the classic KD and the alternative medium-chain triglycerides (MCTs) KD. In the classic KD, the ratio of fat: carbohydrate plus protein is usually 3:1 or 4:1, and the fat composition is long-chain triglycerides (LCTs; 16–20 carbons), providing up to 90% dietary energy [126, 127]. In the MCT KD, MCTs provide 30–60% of energy [127]. MCTs can be predominantly oxidized and utilized for ketogenesis in organs [33], and produce more ketone bodies per kilocalorie of energy than LCTs in the classic KD; thus MCT-induced ketogesis allows a larger proportion of carbohydrate and protein [126]. Besides that, dietary supplementation with moderate MCTs (e.g., 20–70 g/d, containing 8:0 and 10:0) is a more convenient method of nutritional ketosis compared with severe carbohydrate restriction dietary supplementation [128].
Clinical studies
Given that a KD is widely applied in the treatment of epilepsy [129, 130], a series of clinical trials were conducted in AD treatment. A previous randomized, placebo-controlled trial containing 6 mild cognitive impairment (MCI) patients was conducted for 24 weeks, and the results indicated that MCTs oil intake (56 g/d) both increased serum ketone concentrations and improved memory [131]. Recently, a two-phase crossover study in which MCTs (17.3 g/d) or placebo (19.7 g canola oil daily) were administered for 30 days per phase was conducted in 53 mild-to-moderate AD patients, and the results indicated the positive effects of MCTs on cognitive ability in patients without the APOE ɛ4 allele [132]. Another randomized study in MCI patients (30 g/d MCTs) for 6 months suggested that cognitive outcomes, such as episodic memory, language, executive function, and processing speed, were improved compared with baseline [133]. Furthermore, in a study of 20 elderly patients with mild to moderate AD, a 12-week ketogenic formula (Ketonformula®, containing 50 g/d MCTs) significantly improved the results of immediate and delayed logical memory tests at 8 weeks and significantly improved the results of the digit-symbol coding test and immediate logical memory test at 12 weeks compared to the baseline [134]. A pilot study in 44 patients following a coconut oil (40 mL/d)-enriched MeDi invention for 21 days indicated positive effects on cognition, especially in women with a mild-moderate cognitive dysfunction [135]. In contrast, another 3-month clinical trial of Axona® (5–20 g/d 10:0 triglycerides) claimed that Axona did not improve cognitive function in mild-to-moderate AD patients, even in those without the APOE ɛ4 allele, and might be effective for some patients with relatively mild cognitive dysfunction [136]. Overall, these studies are mainly concentrated in mild to moderate AD patients, and most of the clinical data support the effect of an MCFA intervention on cognitive improvement.
Preclinical studies
The possible neuroprotective mechanisms of MCFAs have been widely investigated in both in vitro studies and in vivo studies. The potential role of MCFAs in ameliorating AD is probably attributed to improvements in brain energy metabolism and its antioxidative and anti-neuroinflammatory effects. As we previously mentioned, Aβ deposition induces synaptic toxicity, mitochondrial dysregulation [17, 21], inflammatory responses, and the activation of glial cells [22]; therefore, the beneficial effects of MCFAs on AD patients might not be through directly interaction with Aβ or tau but probably be through improvements the cerebral metabolic environment, thus alleviating the aggregation of AD pathology, such as Aβ aggregation or tau hyperphosphorylation.
Due to cerebral energy metabolism, one of the critical effects of MCFAs is energy supply when glucose is inadequate or the brain glucose pathway is impaired. Both MCFAs and ketone bodies can cross the BBB, serving as alternative energy sources for astrocytes and neurons [137]. Positron emission tomography (PET) and magnetic resonance imaging (MRI) data pointed out that lower brain glucose energy metabolism is a specific symptom in MCI and AD patients [138, 139]; additionally, the decline in brain glucose metabolism is already present before the onset of clinical cognitive decline in people at risk of AD, including APOE ɛ4 carriers and those who have a maternal family history of AD [138]. All these studies have linked AD onset and pathogenesis to insufficient energy supply in the brain, which might explain the effect of a KD on cognitive improvement in AD patients to some extent. Furthermore, MCFA-derived ketone bodies were shown to compensate for the brain glucose deficit in mild-to-moderate AD patients without affecting brain glucose utilization [140], supporting the view that a ketogenic diet may provide a potential therapeutic approach to overcome energy-deficit-induced brain dysfunction in AD.
It is worth mentioning that the benefits of an MCFA-derived KD are mainly attributed to the effects of ketone metabolites. In addition to ameliorating brain energy metabolism, ketone bodies also play important roles in antioxidation, anti-inflammation, and the inhibition of microglial activation [141–143], and these neuroprotective effects were verified in several AD animal models and in vitro studies. In 5×FAD mice, β-hydroxybutyrate significantly improved AD-like pathological events and attenuated Aβ deposition and microglial activation, and in mouse HT22 hippocampal neuronal cells, β-hydroxybutyrate protected mitochondrial function against Aβ toxicity [144]. Similarly, in PDGFB-APPSwInd mouse models, a solution mixture of both β-hydroxybutyrate and acetoacetate improved memory and reduced brain oxidative stress and Aβ burden; furthermore, ketone bodies were verified to reduce Aβ42 entry into neurons and helped to rescue mitochondrial function by restoring Complex I activity [145]. Moreover, β-hydroxybutyrate activates hydroxy-carboxylic acid receptor 2 (HCA2), thus redirecting monocytes and/or macrophages to a neuroprotective role [146]. Additionally, β-hydroxybutyrate participates as an endogenous histone deacetylase inhibitor [147], regulating gene expression through chromatin modifications, such as brain-derived neurotrophic factor, a trophic factor associated with cognitive improvement, both after exercise [148] and under adequate glucose supply [149].
In contrast to ketone bodies, MCFAs such as 10:0 possess unique neuroprotective functions. In addition to ketone bodies, these MCFAs are able to permeate the BBB, acting as bioactive molecules in the CNS [7]. As a peroxisome proliferator activated receptor gamma activator, 10:0 improved mitochondrial citrate synthase and complex I activity in neuronal cells [150]. In addition, 10:0 was proven to directly and selectively inhibit the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor in animal models [151]. Because Aβ-induced AMPA receptor subunit internalization has been proposed to be linked with memory loss in AD [152], whether this inhibitory effect is beneficial to AD requires full investigation.
SFAs, trans FAs, and conjugated FAs
A higher intake of saturated, trans, and conjugated fats is proposed to be a risk factor for several diseases, such as cardiovascular disease and type 2 diabetes [153]. It is recommended that saturated fats should be limited to less than 10% of the total energy intake, and trans fats should be limited to less than 1% or as low as possible [154–157]; the World Health Organization (WHO) plans to eliminate industrially produced trans FAs from the global food supply. SFAs are widely found in various foods, mainly animal products, including dairy products (cow milk and butter), meat, salmon, egg yolks, and animal grease, and several plant products, such as coconut, chocolate and cocoa butter, and palm kernel oils [153].
Among unsaturated fatty acids, one double bond may exhibit a cis (carbon chain extends from the same side) or trans (carbon chain extends from the opposite side) geometric configuration. Dietary unsaturated fatty acids generally form a cis configuration, whereas trans fatty acids (trans FAs) mainly consist of ruminant products, industrial partially hydrogenated vegetable oils and some fried foods [158, 159]. Conjugated fatty acids refer to the positional and geometric isomers of unsaturated fatty acids containing one or more non-methylene-interrupted double bonds in either cis or trans conformation [160]. Ruminal biohydrogenation produces natural trans and conjugated FAs of ruminant products, such as trans-11 18:1 (vaccenic acid) and cis-9, trans-11 conjugated linoleic acid (cis-9, trans-11 CLA; rumenic acid) [161], whereas partially hydrogenated vegetable oils contain trans-9 18:1 (elaidic acid) and trans-10 18:1 [159].
Clinical studies
In their meta-analysis of cohort studies up to 2017, Ruan et al concluded that there is significant evidence of a positive association between higher saturated fat intake and AD and dementia, but this was not the case for total, monounsaturated, polyunsaturated fat intake [162]. Notably, dietary high SFA intake at midlife increased the risk of cognitive impairment and AD later in life [163], especially in APOE ɛ4 carriers [164]. Combined with the beneficial effect that higher dietary PUFA intake is associated with slower decline with aging at middle age [79], it is inferred that the effect of dietary factors in AD pathogenesis is long term, and precautions should be taken at younger age.
On the other hand, several epidemiological studies indicated that diets with high trans-unsaturated fat or low non-hydrogenated unsaturated fats content may be associated with cognitive decline among the older people and people with type 2 diabetes [165, 166], whereas others seem not find the association between higher intake of trans fats and cognitive decline [113, 114]. However, these studies did not distinguish the different types of trans fatty acids. As mentioned earlier, the major geometric configurations of natural and industrial trans fatty acids are different. Through systematic review and meta-analysis, the intake of trans fats was verified to be associated with cardiovascular disease and type 2 diabetes, and the link is probably due to higher levels of intake of industrial trans fat rather than ruminant trans fat [153].
Preclinical studies
The functions of SFAs in increasing AD risk may include regulating the gene expression of Aβ production and subsequent neuroinflammation. For example, palmitate acid (16:0) acts as an activator of BACE1 expression, the enzyme participating in AβPP cleavage and Aβ genesis. In a study of the mouse hippocampus and Neuro-2a neuroblastoma cells, palmitate acid (16:0) was proven to increase the binding of sterol-regulatory element binding protein 1 to the BACE1 promoter region, thus inducing subsequent Aβ genesis [167]. Additionally, palmitate acid-induced BACE1 expression and Aβ genesis are mediated by the C/EBP homologous protein (CHOP)-NF-κB signaling pathway [168]. However, another in vitro study in SH-SY5Y human neuroblastoma cells indicated that SFAs with longer chains (20:0 and 26:0), rather than 16:0, have more potential to promote Aβ production and induce oxidative stress [169]. This difference may be due to the cell types, treatment methods, and in vitro systems. For the in vivo study, as we previously mentioned, SFA-enriched, but not MUFA- and PUFA-enriched diets, significantly increased the enterocytic abundance and plasma concentration of Aβ in mice [120]. Furthermore, SFAs are supposed to increase the production of the cytokinesIL-1, IL-6, and TNF-α, inducing neuroinflammation in glial cells, thus causing neuronal death [170, 171].
Nevertheless, few reports have unveiled the detailed molecular mechanisms between trans fatty acids and AD. An in vitro study on neurons indicated that the accumulation of trans FAs (trans-9 18:1) on the cell membrane, rather than cis FAs (OA; 18:1 ω-9), increased amyloidogenic processing of AβPP, resulting in increased Aβ peptide production, oligomerization, and aggregation [172]. In studies of other major human diseases, different from the positive correlation between industrial trans FAs and heart disease risk [173], two key trans fatty acids in ruminants, vaccenic acid (trans-11 18:1) and rumenic acid (cis-9, trans-11 CLA), showed anti-carcinogenic and anti-atherogenic properties [161]. For AD, we consider it more objective and accurate to discuss the effect of natural and industrial trans fatty acids on AD patients, respectively, and relevant studies are required.
PERSPECTIVE
The pathogenesis of AD is not completely unveiled, and disease-modifying treatments remain elusive. Therefore, the prevention AD onset and the delay of pathological processes seem to be possible approaches, and the association between dietary factors and cognitive decline or AD has been largely investigated. Fatty acids are one of the most essential nutrient components, participating in energy supply, cell composition, and metabolism pathways in the human brain. Healthy dietary habits and appropriate fatty acid intake are beneficial for brain lipid homeostasis and neurodegenerative disease prevention.
In this review, several fatty acids, including ω-3 and ω-6 PUFAs, MUFAs, MCFAs, SFAs, trans FAs, and conjugated FAs, were described, and relevant dietary consumptions were discussed. First, it was found that the intake of ω-3 PUFAs and MUFAs are likely to have a positive association with brain health and decrease the risk of dementia in older people. Evidence for a role of ω-3 PUFAs in AD prevention is increasing, although the results from these clinical studies in AD treatment are not significant. A MUFA-enriched MeDi is recommended for AD prevention, although MUFAs do not seem to be the only factor in improving cognitive decline considering the complexity of this dietary pattern. The possible targets of these dietary fatty acids to modulate AD include a decrease in Aβ accumulation and antioxidative stress and an improvement in neuroinflammation. Given the cerebral energy metabolism in AD patients, a KD and MCFAs provided novel insights into methods to alleviate cognitive decline, showing potential for AD prevention and treatment, and this effect may contribute to improvements in mitochondrial function and energy metabolism, as well as antioxidative and anti-inflammatory properties. It is worth mentioning that these benefits are mainly attributed to the effects of fatty acid metabolites, such as β-hydroxybutyrate. Furthermore, high dietary SFA intake is proposed to increase the risk of AD and dementia, but there is insufficient evidence to support a correlation between trans FAs intake and cognitive decline. We speculate that excessive intake of trans FAs is likely to be associated with a high-fat diet or the so-called “Western diet”, and it is difficult to analyze the effects of trans FAs alone. Further clinical studies and investigations are still required.
However, some limitations of fatty acid approaches in the treatment of AD must be noted. First, dietary intervention in the early stage of AD, such as in the stage of subjective memory complaints or MCI, seems to have a better effect on cognitive improvement. Second, in some cases, one single fatty acid or one kind of fatty acid series is not sufficient to act as the only dietary approach to AD prevention; that is, it is inappropriate to exaggerate the effect of fatty acid factors. Third, AD pathology is a long-term, dynamic, and irreversible neurodegenerative process. As mentioned before, changes in cerebral molecular mechanisms, such as the accumulation of Aβ, occur far earlier than clinical AD symptoms; thus, a long-term healthy lifestyle and dietary habits from young to middle age may be more important for the prevention or delay of AD onset.
In conclusion, a healthy lifestyle and dietary structure and higher ω-3 PUFA and MUFA intake may play an important role in the prevention of AD onset. Additionally, the minimization of SFA and trans FA intake is recommended. A MCFA-derived ketogenic diet provided novel insights into methods to improve AD symptoms, although further studies are required. Additionally, we should pay more attention to the potential molecular mechanisms of these fatty acids in AD pathology, seeking more insights and targets in AD therapy and drug development.
Footnotes
ACKNOWLEDGMENTS
This article was jointly supported by National Natural Science Foundation of China (Program No. 31900807, 81971330, 81873740), the Special Scientific Research Plan of Shaanxi Provincial Education Department (19JK0766), the Science Research Program of Xi’an Medical University (Program No. 2018DOC11). We also thank the support of the Youth Innovation Team of Shaanxi Universities.
