Table 1.
Literature review of ketogenic interventions for MCI, AD, PD, and aging.
| Study | Condition | Methods | Phase | Primary outcome measure(s) | Positive outcomes | Negative outcomes | Apoε4 | Protein biomarkers | Ketone blood levels | Lipid levels | Glucose levels |
AAN level of evidence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Krikorian et al. (34) | 23 older adults with MCI | Randomized to KD vs. high-carb diet for 6 wk | 2 | Cognition | Improved verbal memory (p = 0.01) | N/A | N/A | N/A | Positively correlated with memory improvement | N/A | Fasting glucose levels ↓ | Class II |
| Fortier et al.* (35) | 52 subjects with MCI | Blindly randomized to 30 g kMCT daily vs. placebo for 6 mo | 2 | Cognition, brain AcAc and FDG PET, MRI | Brain ketone metabolism increased by 230% for kMCT group (p < 0.001). Improvements in cognitive tests positively a/w ↑ brain ketone metabolism | Adverse effects primarily GI upset | No significant effect of APOε4 status on ketosis or cognitive outcomes, though not adequately powered to assess this | N/A | Mean [ß-HB] significantly increased to 0.543 mmol/L (p = 0.001) | N/A | Brain glucose uptake unchanged | Class II |
| Neth et al. (19) | 20 subjects with SMC or MCI | Randomized crossover: MMKD vs. LF-AHAD for 6 wk (with 6 wk washout period prior to crossover) | 2 | Cognition, brain AcAc and FDG PET, MRI | Improved metabolic indices. Increased cerebral perfusion among subjects with MCI (no change in perfusion among subjects with SMC). Increased cerebral ketone body uptake (assessed via PET tracer). | Memory improved w/ both diets (though practice effects may have played a role), with relatively greater effect size for MMKD. Not all cognitive outcome parameters reached significance | Higher prevalence of APOε4 among MCI Pts (44% vs 20% for SMC); lower ketone levels in MCI group despite similar dietary compliance; MCI group also had higher fasting insulin and TG levels at baseline | ↑ CSF Aβ42 ↓ CSF Tau ↓ CSF Neurogranin |
Significantly increased (mean [ß-HB] of 0.7 mmol/L) w/ ↑ cerebral uptake of ketones | ↑ LDL ↓ VLDL ↓ TG (p < 0.05) |
Significant reduction in HbA1c | Class II |
| Fortier et al.* (36) | 122 subjects with MCI | Blindly randomized to 30 g kMCT daily vs. placebo for 6 mo | 2 | Cognition, brain AcAc and FDG PET, MRI | Significant Improvements in cognitive tests compared to placebo. Some cognitive outcomes also positively correlated with ↑ brain ketone metabolism | Adverse effects primarily GI upset | No significant effect of APOε4 status on ketosis or cognitive outcomes, though not adequately powered to assess this | Measured plasma amyloid beta (Aβ), but did not see any differences between groups or after the kMCT or placebo | Mean [ß-HB] increased significantly to >0.4 mmol/L Immediately after administration of kMCT (p < 0.0001) | Total cholesterol increased in treatment group, but stayed within clinical reference range | Blood glucose increased in treatment group, but stayed within clinical reference range | Class II |
| Roy et al.* (37) | 37 subjects with MCI | Blindly randomized to 30 g kMCT daily vs. placebo for 6 mo | 2 | Processing speed, brain AcAc PET | Ketone uptake was increased in kMCT group 2.5-3.2-fold in white matter areas of interest. Improvements in processing speed, which were positively associated with white matter uptake | N/A | N/A | N/A | Mean [ß-HB] increased significantly to 0.572 mmol/L 2 h after administration of kMCT (p < 0.0001) | N/A | kMCT → nearly 3-fold increase in white matter ketone uptake, but there was no significant effect on glucose uptake | Class II |
| Myette-Côté et al.* (38) |
39 subjects with MCI | Blindly randomized to 30 g kMCT daily vs. placebo for 6 mo | 1 | Cardio-metabolic and inflammatory biomarkers | MCT supplementation was assessed for its effect on cardiometabolic and inflammatory markers, revealing a reassuring safety profile | Increased IL8 observed in MCT group (unclear clinical significance) w/ no significant effects on other inflammatory markers | N/A | Measured plasma amyloid beta (Aβ), but did not see any differences between groups or after the kMCT or placebo | kMCT group showed significant increase in total ketones post-intervention (+0.416 mmol/L) | Cholesterol and TG levels all remained within normal range throughout intervention | No significant change in HbA1c | Class II |
| Roy et al.* (39) |
32 subjects with MCI | Blindly randomized to 30 g kMCT daily vs. placebo for 6 mo | 2 | Brain MRI connectivity, brain AcAc and FDG PET | kMCT supplementation a/w increased functional connectivity in the dorsal attention network (DAN), which correlated w/ improvement in cognitive tests. Ketone uptake in DAN cortical areas significantly increased in kMCT group, which was directly a/w improvements in DAN functional connectivity | N/A | N/A | N/A | N/A | N/A | After the intervention, mean DAN white matter glucose uptake increased by 8% in the kMCT group (p = 0.039), with no change in the placebo group | Class II |
| Henderson et al. (24) | 152 subjects with mild-to-moderate AD | Randomized, double-blinded, placebo-controlled trial of an oral ketogenic compound (AC-1202); taken daily for 90 d | 2 | Cognition and safety | Improvement in cognitive scores compared to placebo (p < 0.05) | GI adverse effects most common; 24.4% of subjects in experimental group reported diarrhea | Cognitive benefits only observed among APOε4- participants, which correlated with ↑ serum ß-HB levels; no correlation b/n cognitive scores and serum ß-HB levels among APOε4+ subjects | N/A | Post-dose ß-HB levels significantly elevated to mean of 0.36-0.39 mmol/L for experimental group as compared to placebo, though pre-dose ß-HB levels did not significantly differ at any time point | N/A | N/A | Class II |
| Taylor et al. (25) | 15 subjects with AD | Single arm: KD + MCT for 3 mo, followed by 1 mo washout (normal diet resumed) | 1 | Cognition | Improved cognitive scores during diet among completers (p = 0.02); scores reverted after washout | High dropout rate among those w/ moderate AD, citing caregiver burden GI upset, primarily MCT-related |
N/A | N/A | Serum ß-HB levels significantly increased during diet (peaked in first month at average of 0.52 mmol/L) | ↑ HDL, LDL, and total cholesterol (though did not reach statistical significance) | Remained stable | Class III |
| Croteau et al. (26) | 20 subjects with mild-to-moderate AD | Consumed two different kMCT supplements: C8C10 30 g daily for 1 mo followed by 1 mo washout then C8 30 g daily for 1 mo | 2 | Brain AcAc and FDG PET | Brain ketone metabolism doubled for both supplements. Slope of relationship b/n plasma ketones and brain ketone uptake same as in healthy young adults | N/A | N/A | N/A | Mean [ß-HB] significantly increased to 0.57 mmol/L post-C8; mean [ß-HB] significantly increased to 0.46 mmol/L post-C8C10 | ↑ TG No significant difference in HDL, LDL, and total cholesterol |
Total brain metabolism ↑ increased 2/2 ketone metabolism w/ brain glucose metabolism unchanged No significant change in HbA1c | Class III |
| Torosyan et al. (27) | 16 subjects with mild-to-moderate AD | Small double-blinded, placebo-controlled RCT. 14 subjects assigned to caprylidene (ketone precursor); 2 assigned to placebo. Intervention taken for 45 d | 2 | Cerebral blood flow (via O-water PET) | Daily ingestion of caprylidene a/w increased regional cerebral blood flow | N/A | Increased regional cerebral blood flow observed only among APOε4- subjects. No significant effect for APOε4+ subjects | N/A | N/A | N/A | N/A | Class III |
| de la Rubia Ortí et al. (28) | 44 subjects with AD | Randomized to Mediterranean diet enriched with coconut oil vs. isocaloric control diet for 21 d | 2 | Cognition | Improved cognitive function (more pronounced in women w/ less severe disease) | N/A | N/A | N/A | N/A | N/A | N/A | Class II |
| Ota et al. (29) | 20 subjects with mild-to-moderate AD | SS1: Double-blinded RCT comparing MCT formula to placebo w/ cognitive testing and ketone blood level measurement 2 h after single administration SS2: 3-mo open-label study of daily MCT formula ingestion, measuring cognitive function over time |
2 | Cognition | Significant improvements in cognitive scores for open-label trial of daily supplementation w/ MCT formula (p < 0.01), though potentially confounded by training effects in addition to lack of blinding, with subjects serving as own controls | No significant difference in cognitive scores following single administration of supplement. No significant longitudinal increase in endogenous baseline ketone blood level (measured in setting of withholding daily MCT formula at 4, 8, and 12 wks) |
Did not assess. Authors acknowledged this may have been a confounding factor | N/A | MCT formula significantly increased [ß-HB] to mean of 0.4709 mmol/L when measured 2 h after administration (p < 0.001) | Only measured at baseline | Only measured at baseline | Class III |
| Xu et al. (30) | 53 subjects with mild-to-moderate AD | Double-blinded, randomized, placebo-controlled crossover study comparing 17.3 g/d of MCTs to canola oil for 30 d | 2 | Cognition | Significant improvements in cognitive scores compared to placebo | Changes in ADL scores did not significantly differ between MCT and placebo groups | Statistically significant cognitive benefits observed only among APOε4- Pts, though subtle signal suggesting intervention may stabilize cognitive decline among APOε4+ Pts | N/A | After 30 d of MCT intervention, [ß-HB] was 129% higher at baseline in subjects fasting for more than 12 hr (measured prior to MCT dosing): attained mean fasting [ß-HB] of 0.09015 mmol/L (p < 0.01 compared to placebo) | ↑ total cholesterol ↑ HDL (p < 0.01) |
N/A | Class I |
| Henderson et al. (31) | 413 subjects with mild-to-moderate probable AD | Randomized, double-blinded, placebo-controlled trial of an oral ketogenic compound (AC-1204); taken daily for 26 wk | 3 | Cognition and safety | N/A | Failed to improve cognitive or functional capacity. GI upset most common adverse effect |
Primary analysis focused on APOε4- Pts, but did not demonstrate significant benefit | N/A | Mean post-dose [ß-HB] ranging from 0.109-0.272 mmol/L | N/A | N/A | Class II design; however, did not reach endpoint |
| Phillips et al. (32) | 26 subjects with AD | Randomized crossover: KD vs. control diet for 3 mo (with 1 mo washout prior to crossover) | 2 | Quality of life, ADL capacity, and cognition | Improved quality of life (p = 0.02) and ADL capacity (p < 0.01) | Cognitive scores also increased, but not significant (p = 0.24) | Authors concluded that carrier status may be a/w ↓ cognitive benefit of KD, though acknowledged potential confounding factors | N/A | Mean serum ß-HB level of 0.95 mmol/L during diet (significantly increased) | ↑ HDL, LDL, and total cholesterol (p < 0.05) | Significant reduction in HbA1c | Class II |
| Juby et al. (33) | 20 subjects with mild-to-moderate AD | 8-month randomized, double-blinded, placebo-controlled, crossover study comparing 42 g/d of MCT oil to olive oil (phase 1), followed by open-label extension of MCT oil by all subjects for 7 months (phase 2) | 2 | Cognition and safety | Overall, at the conclusion of the extended 15-month protocol (crossover phase followed by open-label extension of MCT for all participants), 80% of subjects demonstrated either cognitive improvement or slowed rates of cognitive decline (thus considered responders to stabilization effects) | During RCT phase, cognitive outcomes for experimental arm did not reach statistical significance. Adverse events were primarily GI-related |
No measurable effect of APOε4 status on response to MCT oil | N/A | Average fasting (pre-intervention) baseline ß-HB level was 0.19 mmol/L, and at study completion was 0.22 mmol/L (no significant change) | No significant change in total cholesterol, LDL, and TG. Weak association noted for lower LDL and total cholesterol in those consuming higher doses of MCT oil |
No impact on HbA1c and fasting insulin | Class III |
| VanIttalie et al. (2) | 7 subjects with PD | Open-label feasibility study of strict KD for 28 days | 1 | UPDRS scores and lipid levels | 5/7 subjects completed the study, though 2/5 completers had occasional lapses in adherence to KD → all 5 completers showed improved total UPDRS scores and improved motor subscores | One subject was unable to prepare diet; another withdrew for unrelated reasons. Potential confounding factors (lack of control group) make it difficult to interpret results beyond confirmation of feasibility for future research |
N/A | N/A | Notably high levels of serum ß-HB (mean of 6.6 mmol/L among three most adherent subjects) | Serum cholesterol levels were not significantly different for 4/5 subjects; for one subject who had high cholesterol at baseline, total cholesterol increased 30% after 28 days | N/A | Class III |
| Phillips et al. (41) | 47 subjects with PD | Randomized to less stringent KD (as compared to the diet used by VanIttalie et al.) vs. LFD for 8 wk | 2 | UPDRS scores | 41% improvement in UPDRS I scores (non-motor) (p < 0.01) | Intermittent exacerbation of PD tremor and/or rigidity in KD group | N/A | N/A | Mean serum [ß-HB] significantly increased to 1.15 mmol/L | KD significantly increased HDL, LDL, and total cholesterol (p < 0.001) | Significant reduction in mean weekly bedtime glucose levels (p = 0.001) | Class II |
| Krikorian et al. (42) | 14 subjects with MCI in the setting of PD | Randomized to KD vs. control diet for 8wk | 2 | Cognition, motor symptoms (via UPDRS-III) | Improvement in memory | No improvement in motor function | N/A | N/A | Mean serum [ß-HB] significantly increased to 0.31 mmol/L | N/A | Fasting insulin significantly declined, whereas fasting glucose did not | Class III |
| Koyuncu et al. (43) | 74 subjects with PD who reported a voice disorder related to their disease | Randomly assigned to KD vs. regular diet for 3 months | 2 | Vocal quality | KD significantly improved voice quality | GI upset in 5 subjects for KD group, compared to 2 in control group | N/A | N/A | N/A | N/A | N/A | Class II |
| Norwitz et al. (40) | 14 subjects with Hoehn and Yahr stage 1-2 PD | Randomized, placebo-controlled, crossover study comparing acute effect of ketone ester drink vs. isocaloric, taste-matched placebo drink on exercise endurance | 1 | Exercise endurance | Participants sustained exercise for 24% longer after consuming the ketone ester drink compared to the isocaloric control drink | N/A | N/A | N/A | Mean serum [ß-HB] increased significantly to 3.5 mmol/L within 30 min of ketone ester consumption | N/A | Less significant rise in glucose associated with formula that contained ketone ester | Class I (Exemplary trial design, though this study did not assess direct effects on clinical symptoms) |
| Freemantle et al. (60) | 32 healthy subjects stratified by age group | A ketogenic meal was consumed with various metabolic markers tracked to assess potential interaction effects between age and ketone metabolism | 1 | Blood ß-HB, glucose, and lipid levels | Elderly people in relatively good health have similar capacity to produce ketones and oxidize ß-HB as compared to middle-aged or young adults | N/A | APOε4 carriers had significantly elevated cholesterol levels, but no significant differences in other metabolites | N/A | Mean serum ß-HB level increased similarly for all three groups | No significant acute change in lipid levels | Among elderly subjects, glucose was oxidized more rapidly as compared to healthy middle-aged adults | Class III |
| Abe et al. (61) | 64 elderly nursing home residents with frailty | Randomly assigned to receive 6 g/d MCTs, 6 g/d MCTs plus Vitamin D3 and L-leucine, or 6 g/d LCTs for 3 mo | 2 | Cognition | Significant improvement in MMSE for subjects receiving MCTs compared to LCTs (p < 0.001), though only partially blinded (examiners aware of subject group) | N/A | N/A | N/A | N/A | N/A | N/A | Class III |
| Mujica-Parodi et al. (62) | 42 adults were assessed via fMRI for network stability (marker of brain aging) | Two experiments conducted involving: (1) comparison of standard diet to overnight fasting and ketogenic diet conditions (KD was sustained for 1 wk prior to fMRI) (2) exogenous ketone ester drink was also compared to calorie-matched glucose drink, with network stability measured as primary outcome |
2 | fMRI network stability | Networks were destabilized by glucose and stabilized by ketones, regardless of whether ketosis was achieved via ketogenic diet or exogenous ketone ester. MRS showed ketones reaching peak concentrations in brain at ~30 min post-dose. Network stability validated as a marker of aging; destabilization effects emerged at 47 years, with most rapid degeneration occurring at 60 years | N/A | N/A | N/A | Ketone ester raised [ß-HB] to mean of 3.5 mmol/L (measured 50 min after administration) | N/A | Ketone ester consumption associated with acute reduction in blood glucose (mean reduction of 18 mg/dL, as measured 50 min after consumption of ketone ester) | Class II |
*
Several publications were derived from the multi-phase BENEFIC trial dataset.
a/w, associated with; AAN, American Academy of Neurology; AD, Alzheimer's disease; ADL, Activities of Daily Living; APOε4, Apolipoprotein ε4; ß-HB, beta-hydroxybutyrate; CSF, cerebrospinal fluid; FA, first author; fMRI, functional Magnetic Resonance Imaging; GI, gastrointestinal; HDL, high-density lipoprotein; IL, interleukin; KD, ketogenic diet; kMCT, ketogenic medium-chain triglycerides; LCT, long-chain triglyceride; LDL, low-density lipoprotein; LF-AHA, low-fat American Heart Association diet; LFD, low-fat diet; MCI, mild cognitive impairment; MCT, medium-chain triglyceride; MMKD, modified Mediterranean ketogenic diet; MMSE, Mini-Mental Status Exam; mo, month; MRS, Magnetic Resonance Spectroscopy; N/A, not applicable; PD, Parkinson's disease; PMID, PubMed reference number; SMC, subjective memory complaints; SS, Sub-study; TG, triglycerides; UPDSR, Unified Parkinson's Disease Rating Scale; VLDL, very-low-density lipoprotein; 2/2, secondary to.