Ketogenic Diet and Neurodegenerative Diseases
Executive Summary
The ketogenic diet (KD) and ketogenic interventions (e.g., MCT-supplemented diets, modified KD protocols, and strategies aimed at increasing β-hydroxybutyrate HB) are described in the literature as potentially beneficial in several neurodegenerative diseases, including Alzheimer's disease (AD), Parkinson's disease (PD), and amyotrophic lateral sclerosis (ALS), but the clinical evidence base is still limited.[1–3]
The most consistent clinical signal in AD concerns improvements in daily functioning and quality of life after 12 weeks of a modified KD in a randomized crossover study (ADCS-ADL ; ; QOL-AD ; ), with no statistically significant improvement in ACE-III.[4]
In PD, clinical data indicate that ketogenic interventions may bring greater benefits in non-motor domains and quality of life than in hard motor outcomes: in an RCT with 47 patients (KD vs. low-fat diet), both groups improved MDS-UPDRS scores, but KD was associated with greater improvement in non-motor symptoms, and a review covering a total of 152 patients in 6 studies indicates a small-to-moderate effect on quality of life, particularly in non-motor areas (fatigue, sleep), with inconsistent results and marginal/inconsistent motor benefits.[1, 5]
Mechanistically, KD/ketone bodies (KBs) are linked to a metabolic shift towards fatty acid oxidation and ketone production, improved mitochondrial function, reduced oxidative stress (e.g., by decreasing ROS generated by complex I), activation of antioxidant pathways (Nrf2), inhibition of neuroinflammatory axes (NF-κB, NLRP3, IL-1β), and signaling and epigenetic phenomena (e.g., HDAC inhibition), as well as modifications of the gut-brain axis through effects on the microbiome.[1, 6]
The biggest limitations of current evidence are small sample sizes, short intervention durations, frequent lack of randomization and control groups, and heterogeneity of protocols and ketosis criteria, resulting in a strong need for large, long-term, prospective, randomized, blinded studies.[3, 7, 8]
Introduction
The ketogenic diet is described as a \"biochemical model of starvation\" that promotes the utilization of ketone bodies as the dominant fuel instead of glucose for the central nervous system.[6]
In clinical practice and research on neurodegeneration, several approaches are encountered that share the goal of achieving ketosis, understood as blood ketone concentrations of around .[9]
Variants used in the cited studies include a modified ketogenic diet (e.g., in a 12-week randomized crossover study in AD) and strategies based on MCT as a ketogenic agent, as a review of two studies with cognitive improvement in AD highlighted that both protocols utilized MCT.[9, 10]
In the clinical literature on early cognitive impairment, the Modified Atkins Diet (MAD) also appears as a ketogenic intervention tested in a 12-week RCT in individuals with MCI of early AD etiology.[11]
In the context of \"exogenous ketones\", it is worth emphasizing that in preclinical models, exogenous β-OHB (exogenous HB) is mentioned as one of the interventions alongside KD and MCT, although such data are preclinical and do not determine clinical efficacy in humans.[6]
Neuroprotective Mechanisms
From a bioenergetic perspective, KD causes a metabolic shift towards ketone production and fatty acid oxidation, which in the cited synthesis is linked to improved mitochondrial function, anti-inflammatory capacity, endogenous antioxidant effect, anti-apoptotic activity, and improved energy supply for the brain.[1]
At the mitochondrial level, it has been described that HB and acetoacetate reduce ROS production by complex I of the respiratory chain, and acetoacetate increases the survival of hippocampal cell lines by reducing ROS, which constitutes one of the mechanistic justifications for neuroprotection.[1]
From the perspective of the antioxidant response, KD has been linked to the activation of the Nrf2 pathway and the attenuation of oxidative stress.[1]
In the area of neuroinflammation, cited works emphasized that KD-induced increases in HB can inhibit inflammation by blocking IL-1β expression and affecting the NLRP3 inflammasome, which controls caspase-1 activation and release.[1]
Along the same mechanistic axis, a direct effect of KD and HB on microglia associated with neuroinflammation was also indicated, with microglia polarizing towards M2-like phenotypes, supporting regeneration and neuroprotection.[1]
Additionally, another mechanistic view indicates that KD inhibits inflammatory processes by inhibiting NF-κB activation.[6]
From the perspective of the innate immune response, it was also emphasized that ketones can inhibit the NLRP3 sensor, which controls caspase-1 activation and the release of pro-inflammatory cytokines (IL-1β, IL-18) by limiting the efflux of from cells.[6]
In the dimension of signaling and epigenetics, it was indicated that one of the mechanisms of KD is the inhibition of histone deacetylases (HDACs), which participate in chromatin structure modification and genetic information accessibility.[6]
At the same time, the cited mechanistic work described that KD can lead to PPAR-α activation mediated by fatty acids, which has been linked to inhibition of glycolysis and fatty acid metabolism in specific regulatory contexts.[6]
The gut-brain axis theme is present in syntheses where ketogenic diets have been shown to affect the abundance and diversity of the gut microbiome and microbial-derived molecules involved in CNS homeostasis and neuroprotection.[1]
In Parkinson's disease, a "bypass" mechanism for complex I was additionally proposed: ketones are said to provide alternative fuel to compromised neurons while improving mitochondrial function and increasing ATP production.[12]
In the same context, it was described that at the mitochondrial level, ketone bodies can attenuate the intrinsic apoptosis cascade by reducing ROS, inhibiting mitochondrial permeability pore opening, reducing cytochrome c release, and subsequent caspase activation.[12]
Alzheimer's Disease and MCI
In AD/MCI, the rationalization for ketogenic interventions includes, among other things, the fact that AD pathogenesis is associated with impaired glucose metabolism (alongside Aβ accumulation and tau pathology), while ketone metabolism in the brain remains normal in AD and can potentially compensate for cerebral insulin resistance and glucose metabolism deficits.[4, 13]
Clinical Evidence
The most detailed clinical data in the provided material come from a randomized crossover study in clinically confirmed Alzheimer's disease, in which, compared to a usual diet, patients on KD increased their mean within-person ADCS-ADL scores by points () and QOL-AD by points (), while ACE-III increased non-significantly (; ).[4]
In the same study, sustained physiological ketosis was achieved, and the 12-week mean HB value was , which is consistent with the concept that efficacy (if present) may depend on achieving ketosis.[4, 9]
In the study, 26 patients were randomized, of whom 21 (81%) completed the ketogenic diet, and only one withdrawal was attributed to the diet, which supports the feasibility of the intervention in the selected population and with appropriate support.[4]
In another 3-month single-arm study in patients with mild/very mild AD, the intervention was well-tolerated, no serious adverse events were observed, and in the group achieving sustained or intermittent ketosis (n=10), ADAS-Cog scores after 3 months significantly exceeded baseline values, while after a one-month \"washout\", they returned to near-baseline levels.[14]
In MCI of early AD etiology, MAD was tested in a controlled study where the mean change in Memory Composite Score was 1.37 points greater in the MAD group than in control (95% CI from −0.87 to 4.90), and the effect size was estimated as moderate (Cohen’s D=0.57; 95% CI −0.67 to 1.33), with concomitant problems with adhesion (only two participants met adhesion criteria).[11]
Clinical reviews in the field of AD/MCI highlight both signals of cognitive improvement in small groups (e.g., 6 weeks KD in MCI and 12 weeks in a single-arm AD study) and the fact that improvement is not uniform, with some studies observing no cognitive improvement in individuals with mild and mild-to-moderate AD.[1, 9, 10]
One synthesis emphasizes that ketosis, defined as ketones in the blood, was achieved in two studies where cognitive improvement was observed after KD, and both these protocols used MCT as a ketogenic agent, suggesting a practical role for variant selection and ketogenesis support.[9]
AD-Specific Mechanisms
In the mechanistic literature, KD is described as a starvation model that promotes the use of KBs as the dominant fuel for the CNS, forming the basis of the \"alternative fuel\" hypothesis in AD.[6]
In this view, KBs are thought to reduce glycolytic ATP production and increase ATP generation via mitochondrial oxidation, which has been linked to beneficial metabolic changes (ketosis, higher serum lipid levels, and lower glycemia) and protection against neuronal loss through apoptosis and necrosis.[6]
In the inflammatory and immunometabolic components, it was indicated that KD can inhibit NF-κB and that ketones can inhibit the NLRP3 inflammasome, limiting the release of pro-inflammatory cytokines such as IL-1β and IL-18.[6]
In epigenetic mechanisms, the possibility of HDAC inhibition by KD was emphasized, which is one of the potential targets for long-term changes in gene expression and plasticity.[6]
Regarding Aβ pathology, data cited from animal models indicate that in rodents treated with KD, exogenous β-OHB, and MCT, a reduction in brain Aβ levels, protection against Aβ toxicity, and improved mitochondrial function were observed, and in a transgenic AD model, a 25% decrease in soluble Aβ deposits was described after 40 days of KD.[6]
Practice and Safety in AD and MCI
In available data, achieving ketosis and adhesion appear as critical implementation limitations: in a single-arm study, five participants did not achieve sustained ketosis and withdrew, and these individuals tended to have more advanced dementia (CDR 2).[14]
In the RCT with MAD in MCI, despite some participants completing the study, only two participants in the MAD arm met adhesion criteria, indicating that ketogenic protocols may require intensive support and monitoring to achieve the intended metabolic effect.[11]
In terms of tolerability in the 3-month study, the intervention was well-tolerated, and no serious adverse events were reported.[14]
At the same time, an analysis of nutritional quality during KD showed compliance with RDA/AI recommendations for some micronutrients (e.g., choline, vitamins A, C, and K) and non-compliance for others (including calcium, folate, magnesium, potassium, thiamine, vitamins D and E), as well as significantly lower fiber intake (P=0.025), which justifies the need for planning supplementation and diet quality in research and practice.[15]
Limitations of evidence are highlighted by reviews indicating that the few human studies were often pre-post designs without randomization and control groups, and that large RCTs with long-term patient observation are needed.[7, 8]
Parkinson's Disease
In PD, ketogenic interventions are described as a potential supplementary strategy capable of affecting many aspects of pathology, although review authors draw attention to the limitations of clinical evidence and the need for cautious interpretation.[16]
Clinical Evidence
In an RCT involving 47 PD patients, a low-fat diet and KD were compared, and MDS-UPDRS scores in both groups significantly decreased, with the KD group showing greater improvement in non-motor symptoms.[1]
In a classic, uncontrolled 28-day study, PD patients experienced an average 43% reduction in UPDRS score after exposure to KD, which represents an early signal of potential symptomatic efficacy but is also burdened by the limitations of a design without a control group.[17]
In a randomized study aimed at the feasibility of short-term KD supplemented with MCT, 15 out of 16 individuals completed the protocol, acceptability was rated as 2.3/3 on average, and ketosis (BHB >0.5 mM) was achieved in 94% of participants by week 3.[18]
In the same study, after adjustment for baseline TUG, no significant difference between groups was found on day 7 of hospitalization (KD 8.4 s, SD 9.1 s; ), and the lack of a significant effect between groups on mobility in TUG or UPDRS-3 (day 7 or week 3) led to the study being terminated due to \"futility\".[18]
A review integrating 6 studies (total of 152 patients) indicated that KD was associated with a small-to-moderate effect size in improving quality of life, especially in non-motor domains (fatigue and sleep quality), however, results were inconsistent, and motor benefits were assessed as marginal or inconsistent.[5]
The same review indicates that most studies did not show significant changes in bradykinesia, tremor, or rigidity, with the exception of one report of improved motor functions related to voice (VHI-10).[5]
In a single-arm 12-week study after KD intervention in PD patients (27 enrolled; 16 completed), a significant decrease in total MDS-UPDRS III motor score () and significant improvements in non-motor domains (MDS-UPDRS I ; NMSS ), including constipation, daytime sleepiness, anxiety, depression, and REM sleep behavior disorder, were observed.[19]
In the same study, cognitive function (MMSE ; MoCA-B ) also improved, which is consistent with the hypothesis that non-motor domains may be particularly sensitive to metabolic interventions, although the single-arm design limits causal inference.[19]
Descriptive data also include a case study of a PD patient (stage I) who followed a traditional KD (70% fat, 25% protein, 5% carbohydrates) for 24 weeks, and the authors reported improvements in biomarkers (e.g., HbA1c, CRP, triglycerides, fasting insulin), improved HDL, and improvements in the \"mentation and behavior\" domain of UPDRS and in depressive symptoms.[20]
PD-Specific Mechanisms
In preclinical models, it was indicated that HB can protect against dopaminergic neuron death and alleviate PD symptoms in mice, which provides a mechanistic rationale for clinical studies but is not proof of efficacy in humans.[21]
From a neuroinflammatory perspective, it was described that HB exhibits strong anti-inflammatory properties, reducing pro-inflammatory cytokines and microglia activation by inhibiting pathways such as NF-κB and the NLRP3 inflammasome, and KD was identified as an intervention that can reduce inflammation and oxidative stress through HB's action.[5, 12]
Additionally, a receptor mechanism was indicated: HB is thought to attenuate microglia activation by stimulating HCAR2, and HB binding to HCAR2 on macrophages and microglia is believed to inhibit NFκB-mediated neuroinflammation, which is considered a critical pathological feature in PD.[12, 21]
In an MPTP-induced mouse model, KD was linked to a decrease in IL-1β, IL-6, and TNF-α, reduced microglia activation in the substantia nigra, and improved dopaminergic transmission and motor functions, which mechanistically aligns with clinical interest in inflammatory domains.[12]
In the bioenergetic and redox dimension, a review on HB emphasized that HB metabolism can modify redox pair ratios (NAD+/NADH and Q/QH2), which potentially reduces ROS production and enhances antioxidant defense.[21]
Gut-Brain Axis in PD
In a 12-week KD study in PD patients, no significant changes in diversity or of the microbiota were observed, but at the same time, a significant increase in Enterococcus and Synergistota and a decrease in Alloprevotella after the intervention were described.[19]
The authors noted that the microbiological shift co-occurred with clinical improvement, which was interpreted as suggesting a role for the gut-brain axis involving anti-inflammatory pathways and dopaminergic regulation.[19]
Practice and Limitations in PD
Reviews emphasized that KD studies in PD are often limited by small populations and short intervention durations, which reduces statistical power and hinders the assessment of long-term effects.[16]
At the same time, synthetic data indicating inconsistency of results between studies and marginal/inconsistent motor benefits underscore the importance of selecting endpoints (non-motor vs. motor) and study design quality (randomization, control, length of observation) in future trials.[5]
ALS
Regarding ALS, the cited syntheses indicate that KD has been described as potentially beneficial in several neurodegenerative diseases, including ALS, however, the limited availability of clinical evidence in neurodegenerative diseases as a group is simultaneously emphasized.[1, 3]
In this situation, a key methodological conclusion is the recommendation for large, long-term, randomized, double-blind controlled trials that could determine whether KD affects the development, progression, and symptoms in neurodegenerative diseases, which also applies to ALS.[3]
MS
Clinical Evidence
In the provided material, the authors explicitly point to the lack of human studies on the use of ketones/KD in neurodegenerative disorders and the absence of human data on KD in MS, thus characterizing the premises for KD use in MS as largely theoretical.[22]
The same source notes that despite the dominance of immunological therapies in MS, there is currently no definitive therapy for progressive forms (primary and secondary progressive), which provides a clinical context in which supportive strategies for the neurodegenerative component are sought.[22]
MS-Specific Mechanisms
In the proposed pathophysiological model, mitochondrial dysfunction can lead to reduced ATP availability, which is consistent with hypotheses about the energetic basis of axonal damage.[22]
According to in vitro and animal data, KD is believed to increase ATP production, promote mitochondrial biogenesis, bypass dysfunctional bioenergetic steps, increase antioxidant levels, and reduce oxidative damage.[22]
In the inflammatory components, it was indicated that the anti-inflammatory effect of KD may be partially explained by HB's inhibition of the NLRP3 inflammasome, in a manner independent of starvation-induced mechanisms (such as AMPK, autophagy, or glycolysis inhibition).[22]
Consequently, it was proposed that since increased ATP and improved mitochondrial function correlate with axon survival, KD may offer a therapeutic benefit for the neurodegenerative component of MS, with the caveat of a lack of human clinical evidence.[22]
Safety
From available clinical data on AD, ketogenic interventions appear to be well-tolerated in the short term: in a 3-month study, no serious adverse events were observed.[14]
In an AD crossover RCT, it was noted that out of 26 randomized patients, 21 (81%) completed the ketogenic diet, and only one withdrawal was attributed to KD, which supports the thesis of acceptability in a clinical study setting.[4]
In PD, short-term KD with MCT was feasible in >90% of participants and was associated with willingness to continue (acceptability 2.3/3), while lacking a significant effect on mobility in TUG/UPDRS-3 in that specific protocol.[18]
Regarding nutritional safety, a significant point is the risk of micronutrient deficiencies and decreased fiber intake during KD, as demonstrated in a dietary quality analysis (non-compliance with recommendations for calcium, magnesium, potassium, vitamins D and E, among others, and significantly lower fiber intake).[15]
Limitations of Evidence
Reviews covering neurodegenerative diseases indicate that clinical data are still limited and heterogeneous, and part of the potential therapeutic value may relate to earlier disease stages and patients with more favorable metabolic or genetic profiles.[2]
" }At the same time, it is noted that clinical evidence is scarce, and most existing studies are small in number, often uncontrolled, and limited to short-term KD effects.[3]
In the area of AD/MCI, it is emphasized that the few human studies often have a before-and-after design without a control group and without randomization, which limits causal inference.[7]
For PD, limitations include small populations and short intervention durations, which hinders the assessment of long-term effects and leads to inconsistency in results across studies, particularly regarding motor outcomes.[5, 16]
For MS, the premises are directly referred to as theoretical, as there is a lack of human study data, making it impossible to formulate clinical recommendations regarding efficacy.[22]
Research Directions
Syntheses concerning neurodegenerative diseases unequivocally recommend large, long-term, prospective, randomized, double-blind controlled studies to determine whether KD can mitigate or treat the development, progression, and symptoms of neurodegenerative diseases.[3]
In the area of AD/MCI, the need for large RCTs with long-term observation is emphasized, stemming from the limitations of previous designs and inconsistencies in cognitive outcomes.[8, 9]
For PD, research directions include clarifying whether ketogenic interventions primarily affect non-motor domains (fatigue, sleep, autonomic symptoms, and cognition) and what their impact is on quality of life compared to other dietary patterns, which is consistent with review findings of small-to-moderate effects on QoL and marginal motor effects.[5]
In mechanistic studies, a rational direction is the integration of axes: mitochondrial bioenergetics (ATP/ROS), neuroinflammation (NF-κB, NLRP3, IL-1β), signaling (HCAR2), and potential microbiota mediators, as these elements are repeatedly indicated as KD/ketone targets.[1, 21]
Practical Conclusions for Clinicians
Ketogenic interventions can only be considered as a potential adjunctive treatment, as reviews emphasize a limited and heterogeneous clinical basis and the need for large RCTs before conclusions about their impact on the progression of neurodegenerative diseases can be drawn.[2, 3]
In AD, the most clinically justified hypothesis, based on available data, is the possibility of short-term improvement in daily functioning and quality of life with maintained ketosis, although improvement in global cognitive tests may be modest or inconsistent.[4, 9]
In MCI and AD, practical implementation should consider that adherence and achieving ketosis are common barriers (percentage of non-achievers of ketosis and dropouts in a single-arm study, and only two meeting adherence criteria in the MAD arm), implying the need for monitoring (e.g., BHB measurements) and dietary support.[9, 11, 14]
In PD, patients should be realistically informed that while some studies suggest improvement in non-motor domains and quality of life, motor outcomes in reviews are often marginal or inconsistent, and a randomized feasibility study found no significant effect on TUG/UPDRS-3 and was stopped due to “futility”.[5, 18]
For each of the discussed disease entities, KD planning should include an assessment of nutritional quality and the risk of deficiencies (e.g., calcium, magnesium, potassium, vitamins D and E, and fiber), as such deviations have been shown in the analysis of micronutrient intake in KD.[15]
In MS, due to the lack of human clinical data, KD cannot be recommended as an intervention with confirmed efficacy, and all decisions should consider that the premises are described as theoretical.[22]
The table below synthesizes in which areas the strongest clinical signals appear in the provided data and what the key limitations are.
| Disease | Best Available Clinical Signal | Key Evidence Limitations |
|---|---|---|
| AD | Improvement in ADCS-ADL and QOL-AD after 12 weeks of KD in a crossover RCT, with no significant improvement in ACE-III.[4] | Small samples and heterogeneity; frequent non-randomized designs in other studies; need for large RCTs.[7, 8] |
| MCI | MAD: moderate effect (Cohen’s D=0.57), but without certain significance and with low adherence.[11] | Achieving ketosis and adherence as barriers; ambiguity of cognitive results between studies.[9, 11] |
| PD | Greater improvement in non-motor symptoms vs. low-fat diet in an RCT (n=47) and a small-to-moderate effect on QoL in a review (152 patients), with inconsistent results.[1, 5] | Often small samples and short interventions; motor benefits marginal/inconsistent; feasibility RCT without effect on TUG/UPDRS-3.[5, 16, 18] |
| ALS | Potential benefit indicated in syntheses, but without sufficient clinical basis in the provided data.[1, 3] | Need for large, long-term RCTs to assess impact on progression and symptoms.[3] |
| SM | Lack of human study data for KD in MS.[22] | Theoretical premises; conclusions based on mechanisms and preclinical studies.[22] |
Summary
The collected data indicate that ketogenic interventions in neurodegeneration have a strong mechanistic rationale encompassing mitochondrial bioenergetics, oxidative stress, neuroinflammation (NF-κB, NLRP3, IL-1β), BHB signaling (including HCAR2), epigenetics (HDAC), and potential gut mediators.[1, 6, 21]
Clinically, the strongest and most quantifiable signals in the provided material relate to short-term improvements in functioning and quality of life in AD in crossover RCTs, and improvements in non-motor domains/quality of life in some PD studies, alongside inconsistencies in motor outcomes and methodological limitations.[1, 4, 5]
Further progress in this field primarily requires large, long-term randomized studies with clear ketosis criteria and protocol standardization, as current data are scarce, heterogeneous, and often short-term and uncontrolled.[3]
