Abstract
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that is mainly characterized by cognitive deficits. Although many studies have been devoted to developing disease-modifying therapies, there has been no effective therapy until now. However, dietary interventions may be a potential strategy to treat AD. The ketogenic diet (KD) is a high-fat and low-carbohydrate diet with adequate protein. KD increases the levels of ketone bodies, providing an alternative energy source when there is not sufficient energy supply because of impaired glucose metabolism. Accumulating preclinical and clinical studies have shown that a KD is beneficial to AD. The potential underlying mechanisms include improved mitochondrial function, optimization of gut microbiota composition, and reduced neuroinflammation and oxidative stress. The review provides an update on clinical and preclinical research on the effects of KD or medium-chain triglyceride supplementation on symptoms and pathophysiology in AD. We also detail the potential mechanisms of KD, involving amyloid and tau proteins, neuroinflammation, gut microbiota, oxidative stress, and brain metabolism. We aimed to determine the function of the KD in AD and outline important aspects of the mechanism, providing a reference for the implementation of the KD as a potential therapeutic strategy for AD.
Keywords
INTRODUCTION
Dementia describes many symptoms, including deficits in memory, thinking, and social abilities, that severely interfere with daily life [1, 2]. Alzheimer’s disease (AD) is the most common form of dementia. AD is an insidious and progressive disease associated with neurological disorders that induces problems with memory, thinking, and behavior [3]. AD is most common and prevailing among older adults [4, 5]. An observational study in 2018, published by Alzheimer’s Disease International, reported that the prevalence of dementia among 50 million people worldwide will triple by in 2050, and two-thirds live in low-income and middle-income countries [6]. Two large-sample, multilocation studies showed an increase in dementia prevalence from 5.14% in 2014 to 5.60% in 2019 [7]. The proportion of Chinese patients with dementia is approximately 25%, which is a great challenge for society and family members [5]. Clinical syndromes in AD present with by cognition and classical symptoms. Typical clinical symptoms of AD are characterized by deficits in memory in the early stages, followed by deficits in language, attention, executive ability, visuospatial ability, gnosis, emotion, etc. [1]. The deficits in cognitive function are rooted in the degeneration of the neural system, which is susceptible to neuropathological alterations located in the limbic system, subcortical structures, archicortex, and neocortex [8]. As a clinicopathological entity, AD is defined biologically based on a neuropathological profile [9]: cerebral plaques containing extracellular deposition of amyloid-β (Aβ) derived from amyloid-β protein precursor (AβPP) and the presence of intraneuronal neurofibrillary tangles (NFTs) and neuropil threads consisting of aggregated and hyperphosphorylated tau protein [10].
Since the late 1990 s and early 2000 s, the FDA-approved drugs, cholinesterase inhibitors (ChEIs) (rivastigmine, galantamine, and donepezil) and memantine have shown modest benefits for the symptoms of AD, but do not reverse the underlying disease progression [11]. ChEIs, which increase the level of acetylcholine and duration of its action in the central nervous system (CNS), are widely recommended for the treatment of mild to moderate AD patients. Memantine is a N-methyl-D-aspartate (NMDA) receptor antagonist and is registered for the treatment of moderate to severe AD [12]. Additionally, a few drugs targeting Aβ have failed to prove therapeutic effects in AD [13]. To date, there have only been a few anti-Aβ monoclonal antibodies that remove amyloid plaques from the brains of AD patients. On 7 June 2021, the FDA granted accelerated approval to the human recombinant anti-AβIgG1 monoclonal antibody aducanumab, which is the first drug aiming to both modify disease and marketing for the treatment of AD [13]. However, the extent of the beneficial clinical efficacy of these antibodies is under contentious debate. There is currently no effective therapy to modify AD progression. Current scientific evidence suggests that a healthy dietary style may reduce the risk of developing AD [14]. In addition, various dietary patterns, such as the ketogenic diet (KD) [15], caloric restriction [16], and Mediterranean diet [17], decrease the neuropathological hallmarks of AD.
One of the diets, the KD, comprising high-fat, ultralow-carbohydrate, and adequate protein diet, has a long history of use as a “cure” for epilepsy with a long history [18], and is recently considered to be a promising potential intervention for a wide spectrum of illnesses. Various adapted versions of KD are currently widely applied for the treatment of obesity [19], diabetes [20], and age-associated diseases [21], and it may serve as a potential therapy for neurodegenerative disorders, such as AD [15, 22], amyotrophic lateral sclerosis [23], and Parkinson’s disease [24]. Recently, increasing evidence has shown benefits of KD in both the pathophysiology and cognition of AD, revealing that KD may be an option for the treatment of AD [25, 26].
In this review, we provide a comprehensive review of KD and AD, including the physiological basis of KD, therapeutic effects, underlying mechanisms, and clinical efficacy. We also summarize the current experimental and clinical results supporting the therapeutic potential of KD in AD.
ALZHEIMER’S DISEASE
Pathophysiology and etiology of AD
The canonical neuropathology of AD involves extracellular Aβ plaques that are neuritic throughout the cerebral cortex and tau-containing intracellular NFTs that are located initially in the medial temporal lobe, followed by spreading to the isocortical regions of the temporal, parietal and frontal lobes in a stereotypical fashion [27, 28].
Aβ is generated from APP by the proteolytic functions of β-secretase enzyme (BACE1) and γ-secretase, known as the amyloidogenic pathway [29]. Following the release of secreted AβPPβ and the C99 fragment by BACE1 cleavage, the γ-secretase complex binds to the N-terminally cleaved C99 fragment and intramembranously cleaves and releases the C-terminal fragment and Aβ48 [30, 31]. Then the γ-secretase complex processes the remaining C-terminal end, sequentially producing the Aβ peptide, which is generally 27–43 amino acids in length. BACE1 and AβPP coexist in endosomes, in which intracellular Aβ is produced [30, 31]. Aβ is secreted into the extracellular interstitial fluid as a monomer, the production and secretion of which are affected by synaptic activity [31, 32]. The level of Aβ in the extracellular space is higher during wakefulness, and Aβ is effectively cleared through the glymphatic system during sleep [33]. Aβ (particularly Aβ42) has a high propensity for aggregation in a concentration-dependent manner. Aβ peptides aggregate into higher-order oligomers, protofibrils, and fibrils that lead to impaired synaptic activity, synapse and neuron loss, impaired cerebral capillary blood flow, and cognitive decline [34]. In addition, AβPP is also cleaved by α-secretase, producing AβPPsα and αCTF, which prevent Aβ formation [35]. The AβPP cleavage products exert normal physiological functions. For example, AβPPsα regulates synaptic transmission by binding to a subtype of gamma-aminobutyric acid (GABA) receptor [36].
Under physiological conditions, the tau protein encoded by the microtubule associated protein tau (MAPT) gene on chromosome 17 is mostly expressed in neurons in the brain [37]. Pathological patterns of tau phosphorylation occur, generating abnormally hyperphosphorylated tau protein following the development of NFTs. This increases cytosolic tau and is considered to facilitate aggregation and fibrillization [37]. Tau hyperphosphorylation impair synaptic function by inhibiting transportation of glutamate receptor or synaptic anchoring [38]. Misfolded tau seeds facilitate templated misfolding of native monomers in a prion action pattern, resulting in the seeding of pathological aggregates [37]. Tau aggregates subsequently spread trans-synaptically to distant, anatomically connected brain regions, representing a molecular mechanism for pathologic tau propagation from the entorhinal cortex to the neocortex in AD [39]. Tau pathology propagates in a stereotyped pattern across neuroanatomically connected networks throughout the AD brain [37]. Unlike Aβ, the development of tau pathology is closely associated with the progression of cognitive impairment [40]. With age, tau pathology aggregates in the medial temporal lobes [41]. Cognitive decline in AD is noted when tau-PET imaging combined with structural MRI finds tau distribution from the entorhinal cortex into the neocortex [42].
The etiology of AD remains intricate, and it is generally attributed to both genetic and environmental risk factors. AD pathology is primarily characterized by cerebral atrophy with Aβ plaques, dystrophic neurites, and NFTs within defined areas of the brain. Other characteristics of AD are regional hypometabolism [43], neuroinflammation [44], mitochondrial dysfunction [45], and oxidative stress [45] in the brain. This disorder in cerebral glucose metabolism occurs prior to the manifestation of pathology and symptoms of AD, and it continues with progression of the symptoms [43]. Increased levels of inflammatory markers were discovered in patients with AD, and AD risk genes involved in innate immune functions were identified, which suggests that neuroinflammation has a prominent effect on AD [44]. The appearance of Aβ aggregates and tau hyperphosphorylation are associated with mitochondrial malfunction in neurons [46, 47]. In an APP transgenic mouse, knockout of the mitochondrial antioxidant enzyme MnSOD obviously increased Aβ levels and plaques [48]. Studies have revealed that Aβ induced oxidative imbalance, leads to neuronal damage in AD, and elevated levels of byproducts of lipid peroxidation, protein oxidation, and DNA/RNA oxidation were discovered in AD [49]. Gene mutations in mitochondria and oxidative stress become risk factors for AD [45]. There are cellular events correlated with AD pathology, including dysfunction of calcium homeostasis and autophagy [50]. Loss of synaptic homeostasis, neuronal loss, and cortical circuitry system failures, which are the possible causes of cognitive impairment in AD, must be mentioned [50]. Additionally, the neurovascular systems manage blood-brain barrier (BBB) permeability and cerebral blood flow and maintains the primary composition of the neuronal “microenvironment” (neurovascular pathways to neurodegeneration in AD and other disorders). Abnormalities involving the impairment of BBB, brain arteries atherosclerosis, and brain hypoperfusion contributes to development of AD [51]. Genetic and genomic technologies have identified more than 20 genetic loci-implicated in AD risk [52]. Dominantly inherited mutations including APP, presenilin 1 (PSEN1) and presenilin 2 (PSEN2) are primarily genetic risk factors for AD, and the epsilon 4 (E4) variant of apolipoprotein E (APOE) also increases the risk of AD development [52]. There is a 3-fold increase in the risk of developing dementia in patients who are greater than 50 years of age following a herpes virus infection [53]. Oral infection of periodontal bacteria increases intraparenchymal Aβ levels [54], suggesting infection as a true risk factor for AD. In midlife, there are several potential metabolic risk factors, including obesity, diabetes mellitus, hypertension, and low high-density lipoprotein cholesterol [55]. In later life, depression, low physical activity, social isolation, air pollution, and smoking are risk factors for dementia [56]. Traumatic brain injury, hearing loss, and alcohol abuse are also correlated with an increased risk of later-life dementia [55].
Pharmacological treatment for AD
There are currently four drugs approved by the US Food and Drug Administration (FDA) that are beneficial for cognitive symptoms in AD. The drugs enhancing cognitive function include three ChEIs and the noncompetitive NMDA receptor antagonist memantine [37]. Acetylcholinesterase inhibitors are the premier drugs approved by the FDA for AD in the United States [57]. In the progression of AD pathology, aberrant cholinergic neurons in the nucleus basalis of Meynert and septal nuclei cause cholinergic deficiency [58], contributing to early attention deficits and memory disorders in AD [37]. ChEIs rescue the deficiency by inhibiting the brain acetylcholinesterase enzyme and increasing acetylcholine levels at the synaptic cleft. They (donepezil, galantamine, and rivastigmine) are currently approved for mild-to-moderate AD, playing a role in the delay of symptom progression [59]. The pharmacokinetic pattern and formulation of drugs are different, but overall efficacy is not. For example, donepezil has a much longer half-life than the others [37]. There are a few adverse effects of ChEIs, including nausea, vomiting, vivid dreams, increased frequency of bowel movements, loss of appetite, and insomnia [57]. Memantine may act to inhibit glutamate-regulated neurotoxicity, although the mechanism of its function remains unclear [60]. It is only considered for patients with moderate to severe dementia, but its effects are rather modest [59]. The adverse effects of memantine are minor, including headaches and constipation [60].
Despite enormous efforts devoted to developing disease-modifying therapies, there remains no effective therapy available today. To date, the amyloid cascade hypothesis indicates that Aβ accumulation is primarily AD pathology, thus lowering the abundance of parenchymal Aβ becomes the goal for slowing disease progression. However, clinical trials of Aβ-targeted agents have failed without any benefits for AD, and the use of anti-Aβ agents as therapies has been a point of intense debate [61]. In clinical trials, strategies of active and passive immunization and inhibiting secretase have been the primary ways of directing Aβ. Active vaccination strategies targeting Aβ peptides are still being explored: CAD106 and ABVac40 [62, 63]. Passive immunization is another strategy for clearing Aβ. Monoclonal antibodies remove soluble Aβ oligomers and insoluble plaques, slowing disease progression [64]. Monoclonal antibodies targeting Aβ include bapinezumab, Solanezumab, Aducanumab, Crenezumab, Gantenerumab, BAN2401, and Donanemab. Although the clinical results for Aβ-targeted monoclonal antibodies in AD patients are depressing, prevention using monoclonal antibodies at an earlier AD stage may show a better outcome for disease modification [37]. An alternative strategy for decreasing Aβ abundance is devoted to reducing production by reducing β- and γ-secretase activities. BACE1 and BACE2 inhibitors obviously decrease Aβ concentrations in cerebrospinal fluid (CSF), which is expected to be a disease-modifying therapy. Nevertheless, clinical trials exhibited deterioration in cognition and elevated liver enzymes [65].
There are other paths to develop disease-modifying therapies. Tau pathology is a critical target. Several strategies for diminishing tau pathogenicity in AD were applied in this clinical trial. Methylene blue diminished tau fibrillization and aggregation in mouse models of tauopathy [66], but its clinical trial failed [67]. A GSK-3 inhibitor, tideglusib, inhibits abnormal hyperphosphorylation of tau [68], but a phase II trial demonstrated no effect in mild-to-moderate AD [69]. There are many preclinical studies developing APOE-directed therapies in animal models, which will eventually be translated and investigated [70, 71]. Additionally, anti-inflammatory treatments for AD and administration of young-donor plasma are currently being tested in clinical and animal trials [37]. Promising pharmacological strategies for the treatment of AD have been in advanced stages of clinical trials, such as anti-Aβ and anti-tau strategies. To date, a disease-modifying therapy that is available in humans has not yet been identified. Key discoveries need to be made, ultimately leading to the development of disease-modifying therapies.
Non-pharmacological treatment for AD
Evidence primarily from clinical trials and observational studies indicates a potentially beneficial role of nonpharmacologic approaches in dementia. Cognitively stimulating activities such as reading and playing cognitive training games may slow cognition impairment in patients with dementia [72]. In a randomized controlled trial, multisensory stimulation therapy, offering visual, tactile, auditory, and olfactory stimulation, maintained the cognition of elderly people with severe dementia [73]. Randomized trials indicate that physical exercise, including aerobic and nonaerobic exercise, possibly improves cognitive and physical function in addition to keeping the cardiovascular system healthy [74]. A comprehensive training program, primarily including sleep hygiene education, elevated light exposure, and daily walking, helps improve sleep in elderly patients with AD who have sleep problems [75]. Mediterranean diet and the Mediterranean-DASH diet Intervention for Neurological Delay diet have potential effects on AD pathology and on maintaining cognitive function in AD [17]. There are a few factors involved in lifestyle changes, such as smoking and alcohol abstinence, weight loss, positive interactions, and proper personal hygiene, that contribute to cognitive improvement in individuals with AD [3, 57]. In addition, a multimodal intervention based on tailored combinations of pharmacological and nonpharmacological strategies showed positive effects on dementia[76, 77].
KETOGENIC DIET
What is ketogenic diet
In the early 1920 s, Wilder et al. found that a special high-fat and very low carbohydrate-based diet-controlled seizures in epilepsy patients, especially drug-resistant patients [18, 79]. These findings might first display the special function of the classical KD. Concurrently, Woodyatt et al. found increased levels of ketone bodies in healthy subjects following fasting or a low-carbohydrate and high-fat diet. The “classic” KD consists of a macronutrient ratio (4:1 or 3:1) of fat to both carbohydrates and protein by weight. The classic KD of 4:1 ratio supplies energy, 90% from fat, 2–4% from carbohydrate, and 6–8% from protein, and the 3:1 ratio supplies energy, 85–90% from fat, 2–5% from carbohydrate, and 8–12% from protein [80]. Use the following points for reference, the typical American diet derives energy, 35% from fat, 49% from carbohydrates, and 16% from protein [81]. KD generally contains eggs, meat, cream, fatty fish, butter, cheese, oils, non-starchy vegetables, avocado, and nuts and seeds. Strict restriction of carbohydrates normally results in a small quantity of intake of fiber-rich foods such as fruits, legumes, whole grains, and starchy vegetables, which is considered to be detrimental to health [81]. Normally, the oxidation of 100–120 g of glucose provides approximately 20% of the energy of basal metabolism, which is used for the functioning of the adult brain [82].
The carbohydrate restriction decreases insulin levels, triggering a shift from primary glucose metabolism to fatty acid metabolism, and the conversion yields ketone bodies (β-hydroxybutyrate (BHB), acetoacetate, and acetone) and supplies approximately 90% of the dietary energy in the classical KD [83]. The ketogenic diet is similar to the biochemical fashion of fasting [84], and the cells in the central nervous system utilize ketone bodies as the predominant fuel source [85]. Under normal physiological conditions, the ketone body concentration in the blood reaches the maximum value of 0.3 mM and 6 mM during prolonged fasting [86]. In diabetic ketoacidosis, the ketone bodies highly reach 25 mM [82], causing the death of the patient with additional problems [2].
Types of ketogenic diet
At present, there are several types of diets, including low-carbohydrate diets (LCDs; e.g., ketogenic diet [KD]), plant-forward diets, intermittent fasting, paleo-type diets, clean eating, traditional regional diets (e.g., the Mediterranean diet), and other specifically designed diets (e.g., dietary approaches to stop hypertension diet, Mayo Clinic diet), which diversify food patterns or fulfill specific purposes [87]. The KD comprises a high-fat component, very low carbohydrates, and adequate protein [78, 89]. Diverse types of KD include: classic KD, medium-chain triglyceride diet (MCTD), modified Atkins diet (MAD), and low glycemic index treatment (LGIT).
Classic KD was first reported by Wilder et al. for curing epilepsy in 1921 [88]. The Long-chain triglycerides (LCT) from standard foods is the most traditional category of KD, in which there is a 4:1 ratio of fat to carbohydrate plus protein. This ratio ranges from 3.5:1 to 3:1 for children, because of the need of a higher protein intake. Fat supplies about 80–90% of calories in classic KD [90, 91], widely applied in the clinical setting. Given the strict carbohydrate restriction, the LCT KD is unsavory, and difficult to maintain in patients [92]. The MCTD was devised in 1971 [92]. This diet tastes more delicious, is acceptable and ketogenic. Consequently, the MCTD obtains better compliance. In several studies [93–95], the treatment of MCTD achieves good results, comparable to the classic KD. There are also several clinical evidences, showing the similar efficacy between MCTD and classic KD [93, 96]. However, the MCTD could cause gastrointestinal side effects, which is the limitation of the MCTD. Besides, MCTD produces more ketones than the classic diet, which may result from its rich octanoic (C8) and decanoic (C10) fatty acids.
The MAD is devised based on the Atkins diet, prevalently used for weight loss [97–99]. Food ingredients of MAD are similar to the classic KD without the necessity for precise weight. The MAD has neither an accurate ketogenic ratio nor protein or calorie restrictions. Carbohydrate intake in the MAD is restrained from 10 to 15 g/day in the first month, and subsequently elevated to 20 g/day [100, 101]. The MAD is a type of more palatable, acceptable, and less restrictive diet, mainly for patients or physicians unwilling to administer the classic KD. Compared to the classical KD, the MAD does not need physicians to initiate the diet. In addition, the MAD may be similar to the efficiency of the classic KD [102]. In several studies, the MAD group displayed better tolerability and fewer side effects than the classic KD [102–104]. The LGIT is based on stable glucose levels induced by KD [105].
Effect of ketogenic diet in AD patients
Previous studies revealed the roles of KD in normal brain aging and AD. There are an increasing number of animal and clinical studies showing the potential of KD in treating AD, and KD is considered as a potential therapy for AD. Among one hundred fifty-three selected papers, we presented 18 animal studies and 29 human studies for the review. The inclusion criteria “KD intervention affects the cognitive function and/or pathology” is the main research topic of this article. In this section, we summarize the key preclinical and clinical studies involved in the role of KD or MCT in animal models and humans, which are detailed listed in Tables 1 and 2.
Preclinical studies on ketogenic diet treatment in normal brain aging and AD
WT, wild type; KD, ketogenic diet; MCT, medium-chain triglycerides; 5×FAD, 5×Familial Alzheimer’s Disease; BHB, β-hydroxybutyrate; AβPP, amyloid-β protein precursor; PS1, presenilin 1; HBME, 3-hydroxybutyrate methyl ester; Aβ, amyloid β; APOE, apolipoprotein E.
Clinical studies on ketogenic diet treatment in normal brain aging and AD
AD, Alzheimer’s disease; KD, ketogenic diet; MCT, medium-chain triglycerides; MCI, mild cognitive impairment; MCI, mild cognitive impairment; MMKD, modified Mediterranean-ketogenic diet; CSF, cerebrospinal fluid; SMC, subjective memory complaints.
PRECLINICAL
There are various animal models for exploring the roles of ketogenic interventions in AD, which are listed in the review. Such as: normal brain aging, 5×FAD mice (overexpressing human APP with the three Familial Alzheimer’s disease (FAD) mutations and human presenilin 1 (PS1) with two FAD mutations), APP/PS1 mice, 3xTg-AD (triple-transgenic mouse model expressing both Aβ plaques and NFTs), and ApoE (apolipoprotein E)-deficient mice. The models were applied to assess the effectiveness of KD or ketone supplementation on pathophysiological features of normal brain aging and AD. Ketogenic interventions, including traditional KD, MCT, and metabolite BHB, were used in animal experiments (Table 1).
Brain aging is an irreversible process associated with cognitive decline, which is considered as a main risk factor for AD [146]. Several studies found that KD had positive effects on cognition function in aged models [106–110]. Newman et al. reveal that a cyclic KD (feeding every other week) prevents memory decline and reduces midlife mortality as mice age, but does not improve maximum lifespan [108]. Dietary supplementation with MCT, similar to treatment of KD, produces high amounts of ketone bodies and has a lasting cognition-improving effect in aged dogs [109]. In addition to enhancement of cognition in aged animals, KD or MCT modulates protein expression of transporters for glucose, vesicular glutamate, and gamma-aminobutyric acid in the hippocampus of old wild type rats, ameliorating age-related abnormal energy metabolism, synaptic transmission, and network function [106]. Treatment of 8- and 10-carbon medium chain fatty acids play slight differential roles in synaptic stability, synaptic plasticity, and cognition in aged rats [107]. The beneficial results of ketogenic intervention indicate that the brain shifts towards using energy provided by the metabolism of ketone bodies.
In the AD transgenic models, animals treated with ketogenic intervention displayed improved cognitive function and motor ability, together with neuropathological and biochemical alterations. Several studies have confirmed that KD or MCT improves cognitive function in AD mouse models [26, 118]. Xu et al. described that long-term (four months) of a KD improved spatial memory and working memory in 5×FAD mice. A shorter period (2 months) of KD only had a weaker positive effect on working memory. However, KD initiated at a late stage of AD (age of 9 months) showed no role in cognitive improvement. The beneficial effect of KD on cognition relied on the starting time and the duration. The improvement in cognitive functions was associated with the rescue of synapses number in the hippocampus [26]. Supplementation of triheptanoin to KD for twelve weeks curbed the memory impairment in APP/PS1 mice (age of 6 months), but did not alter Aβ production and deposition [112]. AD mice treated with 3-hydroxybutyrate methyl ester showed significantly better performance in Morris water maze, displaying a pharmaceutical effect to enhancing the spatial learning and working memory [115]. In 3×Tg- AD mice model, four and seven months after diet initiation, ketone ester-fed mice exhibited not only significant improvements in cognitive performance, but also less anxiety [118]. Wu et al. also found that BHB prevented memory decline in 5×FAD mice by targeting diverse aspects of AD [147]. Moreover, KD fed for at least four weeks improved motor coordination in APP/PS1 mice [113, 114]. Pawlosky et al. reported that a dietary ketone ester normalized abnormal exploratory activity and avoidance-related behavior in 3×Tg-AD mice [117]. Several studies found that KD, MCT, and BHB lessen Aβ and tau deposition in various transgenic mouse models of AD [15, 148]. BHB protects against the multiple pathological events, such as attenuation of Aβ deposition and microglia hyperactivation, and improvement of mitochondrial respiratory function of neurons in 5×FAD mice [147]. BHB restrains the formation of Aβ plaques in 5xFAD mice by modulations of AβPP and neprilysin mediated by G-protein coupled receptor 109A (GPR109A) [147]. Exogenous BHB confers protection against AD pathology by inhibiting NLRP3 inflammasome activation [111]. Administration of BHB in an exogenous Aβ-induced rat AD model effectively reduced Aβ deposition and neuron apoptosis [116]. And BHB inhibits oxidative stress, shown by reduced intracellular reactive oxygen species (ROS) and Ca2 + levels, and increased superoxide dismutase and catalase activities [116]. Krishnan et al. explored the therapeutic outcome of BHB in an ApoE deficient mice model. Subcutaneous injection of BHB for eight weeks impeded lipid deposition in the choroid plexus and reduced amyloid plaques in brain in an ApoE deficient mice [148]. Ketone ester also protected against Aβ and hyperphosphorylated tau deposition in the subiculum, CA1 and CA3, and the amygdala [148]. In this study, van der Auwera et al. found that KD-fed female APP/V7171 mice showed a decrease in Aβ expression in the brain [15]. Four studies assessed the function of KD or BHB in inflammation through microglial hyperactivation, macroglia infiltration in choroid plexus, and NLRP3-inflammasome hyperactivation [26, 148]. Treatment of BHB in 5×FAD mice contributed to a decrease in the total number of microglia and primed, proinflammatory microglia in non-plaque-associated regions [111]. Four months of KD impaired microglial hyperactivation and decreased the levels of the pro-inflammatory cytokines interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) in brain of 5×FAD mice [26]. Mice fed with KD added with triheptanoin have shown reduced astroglial response around Aβ plaques and decreased pro-inflammatory cytokine interferon-γ level in astrocytes [112]. In addition, BHB attenuated CD68+ macroglia infiltration into the ChP in ApoE-deficient mice [148]. MCT notably supplied energy by ketone body metabolism, which markedly promoted mitochondrial function and maintained the intracellular redox state in APP mice [115]. In addition, Pawlosky et al. identified higher levels of TCA cycle or glycolytic intermediates and metabolites in addition to the energy biomarker, N-acetyl aspartate, in MCT-fed animals [149]. Based on the summary, we found a discrepancy on the role of ketogenic intervention in AD. Several factors could explain the discrepancy, such as types of diet and/or duration of diet; types of AD model and its diverse pathological processes. The duration of the treatment possibly plays a key role in the observed discrepancy. As per previous reports, the long term of an intervention improves cognition in 5×FAD mice while a shorter period only has a weaker benefit on cognition [26].
CLINICAL
Previous human studies explored that ketogenic intervention-induced ketosis had extensive neuroprotective effects, contributing to cognition improvement. In addition to KD, there is a ketogenic supplement, MCT, widely used for studying the function of ketosis in AD. Studies prove that MCT supplementation is beneficial for both healthy individuals and those suffering from MCI and AD. The commonly accepted viewpoints underlying MCT supplementation studies suggest that the benefits result from MCT-induced ketogenesis and that administration of MCT is safe [150]. In this section, we summarize the studies of KD, MCT supplementation, and ketogenic agents in patients with MCI and AD and healthy adults, discussing the intervention styles, administration duration, major outcomes, and nutritional state. The details of the human studies are presented in Table 2.
Within the current studies regarding ketogenic interventions, seven studies relied on a classic KD approach while twenty supplied MCT in human trials. Most reports aim at assessing cognitive outcomes, discussing whether ketogenic intervention has significant benefits in cognition, and other studies have pointed out the changes in brain metabolism, biomarkers in CSF, and daily function as well as quality of life. Regarding ketogenic interventions, we found seven studies regarding KD, among which two demonstrated that KD is sufficient to prevent cognitive deficit [131, 138]. Neth et al. reported that twenty participants with subjective memory complaints (SMC) (N = 11) and MCI (N = 9) completed tests including the Free and Cued Selective Reminding Test (FCSRT, reflecting verbal episodic memory) [151] and the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog)12 (testing general cognition that includes items related to executive function, attention, verbal abilities, and memory) [152], and cognitive test showed better performance in FCSRT, but no change in ADAS-Cog12 scores after KD diet [131]. Neth et al. still demonstrated that 6 weeks of KD markedly increased cerebral KB uptake (11C-acetoacetate PET, in subsample) and cerebral perfusion (18F-FDG PET imaging) in addition to improving the cerebrospinal fluid biomarker profile [131]. A randomized clinical trial using MAD for twelve weeks to induce ketogenesis showed an increase in Memory Composite Score (testing episodic memory) and enhanced vitality in patients with MCI and early AD [138]. Nagpal et al. and Neth et al. found that KD improved AD biomarkers in CSF [137]. One study of KD in mild AD showed no change in cognition in a randomized crossover trial [119]. However, Phillips et al. determined that a modified KD for 12 weeks improved daily function and quality of life in AD patients, and that the alteration in cardiovascular risk factors was acceptable [119, 137]. Three studies do not assess cognitive outcomes but other aspects [128, 139]. The results reported by Nagpal et al. reveal that specific gut microbial signatures are associated with improved AD biomarkers in CSF and that KD modulates gut microbiome-mediated short-chain fatty acids in fecal samples [137]. Brinkley et al. assigned twenty adults (mean age: 64.3) to a crossover trial, examining the influence of KD on adiposity measured by dual-energy x-ray absorptiometry and CSF biomarkers. And their results displayed favorable changes in CSF biomarkers, body fat distribution, and body composition [139]. Taylor et al. reported the food and micronutrient profile of the KD [128].
On the other hand, the majority of the studies (twenty out of twenty-nine) explored whether MCT supplementation affects neuropathological, biochemical, and cognitive function in healthy adults, MCI, and AD. In older patients with mild to moderate AD, five studies out of seven observed improved cognition in the MCT groups. In a randomized, double-blind, crossover study for 6 months followed by another 6 months of open-label extension in AD subjects (mean age: 72.6), twenty subjects provided with MCT supplementation at 42 g per day, or maximum tolerated dose, improved cognitive scores assessed with the Mini-Mental State Examination, Montreal Cognitive Assessment, and Cognigram® [122]. Juby et al. performed the longest duration of MCT supplementation in an AD study to date. In another double-blind, randomized, crossover study, 53 AD patients orally took MCT jelly (fat dose: 17.3 g per day) for a short period (4 weeks) per phase, but cognitive function measured by the ADAS-Cog (Chinese version) is still improved in mild to moderate AD patients with APOE4– /–. Besides, three studies with 15–20 patients with mild-to-moderate AD uniformly reveal the positive effect of MCT-based ketogenic formula on cognitive function [22, 129]. Two study reports no significant role in cognition [126], and the remaining studies do not evaluates cognitive function [126, 127], but brain energy metabolism [123]. Croteau et al. reported twenty participants from mild to moderate AD patients consumed MCT supplements (30 g/d) for two phase (four weeks per phases). The results using PET neuroimaging showed that increased ketone supply enhances total brain energy metabolism but not brain glucose utilization, which suggests plasma ketones achieved from MCT supplement compensates for the shortage of brain glucose metabolism in AD [123].
Six trials examined the influence of MCT interventions on subjects with MCI. Participants with MCI were randomly assigned to MCT (≥30 g per day) or placebo groups for at least 24 weeks [132, 135]. Rebelloa et al. and Fortiera et al. found that consumption of MCT increased ketone concentrations in serum and improved cognitive outcomes in MCI [132, 135]. Likewise, another three studies utilized MCT to treat MCI participants for 24 weeks, but improvement of cognition in MCI subjects by MCT was limited and change in cognitive scores was not significant [132, 140]. Nevertheless, the three studies reported that MCT improved white matter energy supply, brain functional connectivity, and axonal integrity [134, 140]. Besides, Myette et al. demonstrated the safety of MCT in treating individuals with MCI [136].
In the limited studies with healthy subjects, we found three out of five demonstrated that MCT supplements improve cognitive function in healthy elderly adults [141, 145]. In a pilot study, Yomogida et al. found that a single-dose of MCT improved cognitive performance in some tests, and the improvement is indirectly associated with an increased energy supplied by ketone bodies in the dorsolateral prefrontal cortex [141]. Elderly nursing home residents (mean age of 85.9) were randomly assigned to the MCT group (6 g per day for six weeks), in which both muscle and cognition function are enhanced in the frail elderly individuals [144]. Ota et al. also reported similar results in cognition [145]. In addition, the MCT meals play a positive role in visual attention and task switching positively correlated with an increase in the plasma β-hydroxybutyrate level in non-demented elderly [145]. Most studies aim to study the aged population or individuals suffering from cognitive dysfunction. It is necessary to clarify the function of MCT administration on cognition in healthy young adults. Interestingly, Ashton et al. just reported that supplementation of either 12 g/day or 18 g/day of MCT (the ratio of C8 and C10 is 30:70) for 3 weeks enhanced cognitive performance in healthy adults with a mean age of twenty, and supplying young participants with a 12 g/day dose of MCT possibly displayed better performance in cognition [143]. Very few studies provide subjects with MCT at doses lower than 20 g. One remaining study also demonstrates that administration of MCT (GSK2981710:30 g/day) for two weeks has no significant effects on cognitive function or memory-involved neuronal activity in healthy elderly individuals [142].
Weight loss is regarded as an adverse effect of KD because of insufficient energy intake. However, there are just three studies reporting weight loss in humans [119, 139], and seven studies founding stable weight in humans after KD or MCT [120, 135]. Most clinic reports do not assess the change in weight.
UNDERLYING MECHANISM OF KD IN AD
The effect of KD-evoked ketosis in AD
Normally, glycolysis supplies energy for human, and converts glucose to pyruvate. Nevertheless, under specific nutrient conditions, such as KD, fasting, dietary carbohydrate restriction, and prolonged physical exercise, the metabolism in humans triggers ketogenesis following the production of ketone bodies as the primary metabolic fuel. Generally, the increased levels of ketone in the blood are known as ketosis, indicated by reaching a ketone concentration of over 0.5–0.6 mmol/ L in blood [119]. Ketone bodies are metabolized in tissues, offering the main energy source. In the condition of ketosis, the levels of blood glucose are relatively stable.
Dietary triglycerides as a supplement provided by KD are broken down by lipases in the gastrointestinal tract, then hydrolyzed into medium-chain fatty acids, which are further absorbed directly in the intestine and transferred to the liver [20]. Most medium-chain fatty acids are rapidly broken into the ketone bodies, such as BHB, acetoacetate, and acetone [20]. Both fat and ketones are carried to the brain through the blood circulation system. Fatty acids and ketones are sufficient to cross the BBB, providing an energy source as a substitute for brain cells [20].
In brain tissue, ketone bodies are taken in by astrocytes and neurons through monocarboxylate transporters, which rely on the concentrations of ketone bodies in the blood circulation [25, 153]. The brain cells utilize the ketones to satisfy the energy demands under the condition of glucose deprivation [154]. The ketone body metabolism meets 80% of the energy demands in the brain, having neuroprotective effects, especially in neurological disorders. For the remaining 20% of the energy needs, glucose is still the unique source of energy [155, 156]. Ketosis or supplementations of BHB enhances the production and release of mitochondrial ATP, leading to an increase in extracellular purine nucleoside adenosine, and the process evokes neuroprotective effects [157, 158]. Previous studies found that BHB activates AMPK through HCAR2, leading to increased activity of SIRTs [159], and then evokes neuroprotective effects [149]. Therapeutic ketosis enhances the inhibitory effects of GABAergic neurons [84] and decreases the glutamate release as well as glutamate-induced neuronal excitatory toxicity [84, 160]. The ketone bodies in the blood play a positive role in AD. In 5×FAD mice model, exogenous BHB significantly improved cognitive function in addition to decreasing Aβ [147]. Two studies demonstrated that administration of BHB decreased Aβ and inflammation in the brain [111, 148]. Ketone bodies still decrease reactive oxygen forms and improve mitochondrial function in the hippocampus [116, 161]. Though many studies reveal a relationship between ketosis and improvement of cognition in AD patients, only two studies displayed distinctive results in patients suffering from mild and moderate AD (A Placebo-Controlled, Parallel-Group, Randomized Clinical Trial of AC-1204 in Mild-to-Moderate Alzheimer’s Disease; Randomized crossover trial of a modified ketogenic diet in Alzheimer’s disease). Henderson et al. reported that an increase in ketone bodies was not found in the participants supplied with the caprylic triglyceride formula (A Placebo-Controlled, Parallel-Group, Randomized Clinical Trial of AC-1204 in Mild-to-Moderate Alzheimer’s Disease). And Phillips et al. explored the influence of classical ketogenic low-carbohydrate diets-induced ketosis on mild AD patients, showing no positive effects on cognitive function (Randomized crossover trial of a modified ketogenic diet in Alzheimer’s disease). In clinical studies, most used MCT as a source of induction of ketosis (Table 2). Twelve out of fifteen studies suggest that supplementation of MCT improved cognitive function, and the remaining three studies used KD or AC-1204 (Table 2). Therefore, MCT supplementation may serve as a potential means for improving cognitive ability in AD patients, MCI patients, and healthy adults. In spite of the promising manifestation of ketosis in AD, more studies are still needed to determine the true effect of ketosison AD.
The effect of ketogenic diet on the deposition of amyloid and tau protein
The pivotal histopathological hallmarks in AD are both extracellular deposition of Aβ plaques and intracellular abnormal aggregation of tau protein. Preliminary studies reveal that ketogenic intervention may be beneficial for improvement in the pathology of AD. There is more evidence that KD may positively modulate Aβ in both transgenic or non-transgenic animal models or humans for the treatment of AD. In transgenic 5×FAD mice receiving 1.5 mmol/kg/d BHB for 4 weeks, BHB attenuated Aβ deposition in the brain, and enhanced mitochondrial respiratory function in hippocampal neurons to protect against Aβ toxicity [147]. Likewise, a 4-month KD fed to 5×FAD mice decreased the level of hippocampal Aβ expression compared to control [26]. Shippy et al. reported that 5×FAD mice drinking BHB-treated water significantly reduced the number of X04+ and 6E10+ plaques as well as the percentage of area covered by plaques in detail [111]. Additionally, they did not observe significant differences in plaque volume, plaque density, or plaque sphericity [111]. APP/V717I transgenic mice fed with KD for 6 weeks showed a 25% reduction in the soluble Aβ40 and Aβ42 levels of whole brain, and no difference in the ratio of Aβ42 to Aβ40 between the KD group and control group was found [15, 162]. A 1-month KD decreased the precursor for Aβ in the brain [113]. A reduced Aβ positive area in mouse brains was also observed after intragastric administration of a BHB derivative [115]. Supplementation with 21.5 g of ketone ester for eight months reduced Aβ and hyperphosphorylated tau deposition compared to a diet without ketone ester [162] or BHB impaired Aβ deposition, whereas KD had no effects on Aβ deposition [116, 163]. Besides, a modified Mediterranean-ketogenic diet in humans with either MCI or SMC increased CSF Aβ42 and decreased tau in CSF, suggesting that a 6-week ketogenic intervention-induced ketosis positively affects the CSF AD biomarker profile [131]. Interestingly, defects in mitochondrial function lead to diminished energy generated from the oxidation of glucose/pyruvate, and it can also modulate AβPP processing, producing neurotoxic Aβ deposition [164]. ApoE is closely related to AD. Krishnan et al. found that BHB also decreased amyloid plaques in the substantia nigra pars in ApoE deficient mice [148].
The effect of ketogenic diet on inflammation
Inflammation modulated by microglia and astrocytes in the brain is an essential factor affecting the neuropathology of AD, and the underlying neurotoxic mechanisms lead to neuronal loss and cognitive dysfunction [165, 166]. Resident glial cells in the brain and peripheral infiltrating immune comprehensively mediate the neuroinflammation through regulating inflammatory cytokines, chemokines, and small-molecule messengers. Generally, KD possibly regulates microglial activities and inflammatory response by special pathways. KD exerts its anti-inflammatory function via modulation of microglial activities [167], pro-apoptotic properties, increased concentrations of neuroprotective mediators, and molecular chaperones, preventing polypeptide aggregation into toxic molecules [168].
Hydrocarboxylic acid receptor 2 (HCA2), known as G-Coupled Protein Receptor 109A, is expressed abundantly on microglia [168], macrophages, and dendritic cells [169]. BHB, as a endogenous ligand, can directly bind to HCA2, and activates it with approximately 0.7 mmol/L [169, 170]. Activation of HCA2 treated with BHB further attenuates the production of both pro-inflammatory cytokines and enzymes via the NF-κB pathway in activated primary microglia, which are stimulated with lipopolysaccharide [171]. Another mechanism of KD-regulated neuroinflammation is to inhibit the NOD-like receptor 3 (NLRP3) inflammasome, which mediates the activities of caspase-1 and the release of proinflammatory cytokines [172–174]. KD increases the NAD (+)/NADH ratio [175, 176], which regulates transcription of pro-inflammatory cytokines in the brain [177]. KD stimulates AMP-activated protein kinase prior to inhibition of NF-kB activities, and transcription of pro-inflammatory cytokines [177]. Besides, accumulation of Aβ promotes the production of pro-inflammatory cytokines from microglia in early stages of AD pathology [178, 179]. Cellular KD treatment significantly reduced the Iba-1+ cells, cell body diameter, and increased the primary process length of microglia, attenuating microglial activation and microgliosis [26]. Meanwhile, an increase in IL-1β and TNF-α was reversed by KD treatment in hippocampus and cortex, and the results indicate that KD decreases the inflammation response in the brain [26]. In a word, impairing neuroinflammation could be one of the crucial modified effects of KDon AD.
Insulin resistance or hyperglycemia aggravates the pathological progression through a complex of mechanisms in AD [180]. Recent studies determined that KD improved insulin sensitivity [181–183]. Further studies found that BHB inhibits insulin glycation to prevent microglial apoptosis [184]. Therefore, KD may have anti-inflammatory effects by increasing insulin sensitivity.
The effect of ketogenic diet on brain metabolism
The phenomenon that glucose uptake is decreased in AD is observed [185]. It is commonly considered that the decrease in glucose demand results from neuronal death and reduced synaptic activity. On the other hand, the brains of patients with AD and MCI are capable of utilizing ketone bodies for energy metabolism [186]. Overall brain metabolism may be made up of a fuel alternate due to impaired glucose metabolism. Administration of ketone bodies is possibly considered as a preventive strategy by attenuating impaired brain metabolism [185].
The β-oxidation of fatty acids yields acetyl-CoA in mitochondrion [187, 188]. Acetyl-CoA molecules may be partially used to produce AcAc prior to being converted to acetone or BHB in the liver mitochondria [173, 187–189]. Ketone bodies, like BHB and AcAc, are transported in the bloodstream and are distributed in the brain, providing energy for brain cells in mitochondria [190]. In addition, ketone bodies are rapidly metabolized because of the bypass of the glycolytic pathway [86, 188]. In the brain, ketone bodies promote multiple complexes formation of the mitochondrial respiratory chain [191, 192], and ATP production [193]. Ketone bodies may also augment the number and bioenergetics of mitochondria through activating a special axis that regulate mitogenesis [194]. Another study suggest that KD upregulated hippocampal genes encoding mitochondrial and energy metabolism enzymes [195].
The effect of ketogenic diet on oxidative stress
Oxidative stress is considered as an important modulator in AD and aging [196]. It is an imbalance between the production of oxidants and antioxidants in organisms [197]. Dysfunction of the antioxidant system leads to oxidative stress, which is characterized by increased levels of reactive species. A disorder in oxidative phosphorylation may induce ROS, leading to mitochondrial dysfunction and inefficient production of ATP. To our knowledge, a lower concentration of ROS is essential for normal cellular function, but the higher concentration and longer exposure of ROS break the structure of cellular macromolecules, resulting in necrosis and cell apoptosis. A study found that excessive ROS is observed in early AD [49], contributing significantly to the pathogenesis and progression of AD [49, 198]. Biomarkers of oxidative stress are well related with Aβ levels in AD [199]. Nerve fiber tangles and neuritic plaques rich in fibrillar Aβ42 and Aβ40 contain diverse oxidized, and nitrated proteins (oxidative stress, dysfunctional glucose metabolism and AD). Interestingly, Aβ42 oligomers cause damage to synaptic membranes through oxidative pathways [200, 201], and contribute to synaptic dysfunction as well as cognitive deficit [202]. Oligomer-induced oxidative damage may be closely associated with synaptic dysfunction. Milder et al. reported that KD activated the hippocampal nuclear factor-E2 related factor2 (Nrf2) pathway via redox signaling in healthy rats and improved the mitochondrial redox state [203]. KD also activates Nrf2, enhancing the synthesis of antioxidants, such as manganese superoxide dismutase and glutathione, which protect from oxidative stress and mitochondrial damage with scavenging ROS [189, 204]. KD-induced Nrf2 activation is linked to decreased ROS in nervous tissue in some conditions [205, 206]. Ketogenic intervention may impede entry of small ROS particles through the mitochondrial membrane due to reduce the permeability [168, 207]. Besides, other studies suggest that metabolism of ketone bodies in mitochondria produces less ROS than glucose metabolism [190], possibly because of elevated levels of uncoupling protein in mitochondrial respiration [208, 209]. BHB also serves as an antioxidant to scavenge ROS, protecting against oxidative damage [210].
The effect of ketogenic diet on gut microbiota
Considering many studies show the crucial effects of the microbiota on neurodevelopment and behavior, focusing on revealing the two-way communication between gut bacteria and the central nervous system, the concept of the microbiome-gut-brain axis has been gradually accepted [211]. The gut microbiota affects the host central nervous system by diverse mechanisms involving metabolism, immune, and endocrine in the gut [212, 213].
The gut microbiota of an infant is considered to be largely implanted when they are exposed to the maternal microbiota during delivery [214]. There are various factors influencing the colonization of gut microbiota, including mode of maternal infection, delivery, breastfeeding, environment, healthy conditions of the host, stress, and diet [215, 216]. However, diet is perhaps the most crucial factor of the factors modulating the composition of the microbiota throughout life [217]. There is accumulating evidence supporting that KD improves the gut bacterial profile in various studies. Recently, a study suggests that KD-altered gut microbiota profiles mediate neuroinflammation in an epilepsy model [218]. Similarly, KD intervention modulated the composition of the gut microbiota, shifting toward probiotics, enhanced brain vascular function and improved metabolic profile (such as Turicibacter and Desulfovibrio) [219].
The gut microbiota in transgenic mouse models of AD have been shown to be altered with age [220]. Chronic intervention of transgenic mice using an antibiotic cocktail reduced microglia accumulation around Aβ plaques in the hippocampus, and insoluble Aβ plaques [221, 222]. A clinical study showed that subjects with MCI have a higher abundance of Akkermansia, Enterobacteriaceae, Christensenellaceae, Slackia, and Erysipelotriaceae, but a relatively lower abundance of Lachnobacterium and Bifidobacterium after treating with a modified Mediterranean-ketogenic diet (MMKD) [130]. Besides, the MMKD decreased acetate and lactate and increased propionate and butyrate in feces, protecting against AD pathology [130]. In previous animal and human studies, KD suppressed bifidobacterial growth mediated by the host ketone bodies in a manner different from a high-fat diet. The KD-mediated gut microbiota downregulates the number of pro-inflammatory Th17 cells [223]. The Th17-involved host immune response might account for the potential protective mechanism of KD in the microenvironment of the intestine [223]. Additionally, KD elevated the levels of beneficial short-chain fatty acids and reduced the production of γ-glutamyl amino acid via modulating the abundance of Lactobacillus and Akkermansia muciniphila [213, 224].
Limitation and adverse effects
The studies on the effects of the KD on treatment of AD provided promising results, and using KD lasted for at least several weeks. A key problem for treatment of the KD in patients is compliance with dietary recommendations [225]. A study also pointed out that KD-associated anorectic activity is found [161]. The reason may be decreased organoleptic attractiveness and appetite in KD. Besides, another problem using KD is its adverse effects following long-term treatment. For epilepsy patients using KD, the adverse effects are gastrointestinal reactions including constipation, lower appetite, vomiting and nausea, weight loss, and transient hyperlipidemia including elevated cholesterol, triglycerides, and low-density lipoprotein cholesterol in serum [85, 227]. Side effects were observed in patients with AD, including high nervousness, muscle cramps, fatigue, and constipation after the intervention lasted for six weeks [119]. Because of the low carbohydrate supply in KD, the level of dietary fiber is low. Especially, constipation may become worse in the premier period of KD treatment. Muscle pain, hypoglycemia, hypocalcemia, hypercalciuria, and hypertriglyceridemia belong to the early side effects of KD treatment [21, 227]. The consequences of dehydration are also extremely dangerous, especially for the elderly population [228]. Excessive ketone bodies were considered to be toxic, arising from knowledge of diabetic ketoacidosis [229]. Hyperketonemia arising from insulin deficiency may cause severe acidosis and even cause death in individuals [2]. The existing abnormalities in patients using KD may lead to reduced food consumption, finally causing energy and nutrient deficiencies. The negative results will continue to worsen the patient’s condition [230]. Implementation of the classic or modified KD may reduce body mass and increase the risk of malnutrition. The application of the KD to elderly men with AD may remain uncertain. Given the above problems and limitations in the treatment of a classic KD, it seems safer to use MCT supplements. Supplementation of MCT does not require drastic alteration in the composition of the AD patients’ diet. Medical strategies should be implemented to prevent diverse adverse effects arising from KD treatment, such as using medications that are appropriate for the adverseeffects.
CONCLUSION
Currently, neither effective strategies nor disease-modifying therapies are available for treatment of AD. Therefore, preventive or therapeutic interventions are crucial to delay the progression of AD. To our knowledge, KD is a traditional and beneficial treatment for epilepsy, and it has been suggested that KD has neuroprotective function in brain disorders. Subsequently, KD is recommended for the treatment of AD, observing whether cognitive function is improved efficiency among AD patients [124]. Multiple studies have shown that KD contributes to cognitive function through pleiotropic mechanisms involved in ketosis, pathophysiology, oxidative stress, neuroinflammation, and brain metabolism in AD.
Although the current findings point out that KD provides a promising approach for improving cognitive function in AD, preclinical studies identifying the mechanism of KD are still needed. Besides, performing large-scale clinical trials on AD, especially randomized control trials, is essential to evaluate the adverse effects in addition to safety, efficacy, and sustainability of KD. Various KDs need restrained carbohydrate intake, which puts more demand on caregivers. A standardized program with KD in AD patients may be more feasible, and it is easy to administer for the caregivers [231]. Physical state of patients should be paid attention, since KD may exacerbate the risk of pathological process in patients with malnutrition [232, 233]. It is important to develop a management system that ensures the patient safety [124]. In light of clinical reports, supplementation of MCT or BHB causing nutritional ketosis for individuals seems more reasonable and may be more feasible in future clinical application. Due to the limitations and adverse effects of KD, combined usage of multiple approaches may be more effective in slowing disease progression in patients with MCI or AD.
Although the current study suggests that a KD may be beneficial for AD and MCI as a potential treatment, further research into its mechanism and exploration of more effective molecular targets are needed in the future. Moreover, while the KD is widely popular and the market is promising, its safety and effectiveness need to be further studied in the future, and KD would be extended if the clinical benefits outweigh the risks. We are optimistic that the KD may be generalized to other neurodegenerative diseases due to the specific function, but the evidence remains to beaccumulated.
Footnotes
ACKNOWLEDGMENTS
The authors have no acknowledgments to report.
FUNDING
This work was supported by STI2030-Major Projects (2021ZD0202103), National Natural Science Foundation of China (81922024), Science, Technology and Innovation Commission of Shenzhen Municipality (RCJC20200714114556103 & ZDSYS20190902093601675).
CONFLICT OF INTEREST
The authors have no conflict of interest to report.