The review, led by Maged T. Ghoche and Kalil G. Abdullah of Northwell Health, together with colleagues at the University of Pittsburgh and the University of Texas Southwestern Medical Center, systematically searched PubMed and Embase from inception through August 19, 2026. The team included in vivo preclinical experiments and clinical studies evaluating ketogenic diets or methionine restriction in glioma, extracting data on interventions, comparators, concurrent oncologic treatments, metabolic effects, safety, feasibility, and outcomes. Because most clinical studies were early-phase, single-arm feasibility investigations, the researchers allowed clinical studies with or without a non-diet comparator, then assessed risk of bias with design-appropriate tools and synthesized effects narratively and by direction of effect.

The search yielded 43 eligible reports: 14 preclinical and 18 clinical reports of ketogenic diet interventions, and 9 preclinical and only 2 clinical studies of methionine restriction. That imbalance itself tells a story. Ketogenic diets, which drastically cut carbohydrates and elevate fat intake to push the body into ketosis, have generated nearly two decades of clinical curiosity in neuro-oncology. Methionine restriction, by contrast, remains largely confined to the laboratory, despite a compelling mechanistic rationale rooted in the peculiar metabolic addictions of glioma cells.

The scientific logic behind both diets traces back to a century-old observation. In the 1920s, Otto Warburg described how cancer cells consume glucose voraciously even in the presence of oxygen, a phenomenon now known as the Warburg effect. Glioblastoma is a quintessential glycolytic tumor, and the healthy brain, uniquely dependent on glucose, can partially switch to ketone bodies when glucose is scarce. The therapeutic hypothesis is elegant: if ketone bodies can fuel normal neurons but not tumor cells that have lost key ketogenic enzymes, a ketogenic diet might create a metabolic mismatch that harms the tumor while sparing the brain. Studies cited in the review, including work by Tisdale and colleagues in the 1980s and Maurer and colleagues in 2011, documented loss of acetoacetate coenzyme A transferase activity and differential ketone body utilization in glioma cell lines compared with neurons, providing a cellular foundation for the strategy.

Preclinical results, however, proved strikingly heterogeneous. Several diet-only experiments in mice reduced tumor growth or prolonged survival, including Seyfried’s calorically restricted ketogenic diet work from 2007 and Abdelwahab’s 2012 demonstration that a ketogenic diet acted as an effective adjuvant to radiation therapy in malignant glioma. Other studies showed limited activity, and one informative trial by de Feyter and colleagues in 2016 found that a ketogenic diet increased ketone body transport and oxidation in RG2 and 9L gliomas without actually affecting tumor growth, a reminder that fuel delivery does not equal fuel dependence. The most dramatic preclinical effects emerged when diets were combined with other modalities: radiation, antiangiogenic therapy, glutamine targeting, or immune modulation. Mukherjee and colleagues in 2019 reported therapeutic benefit from combining a calorie-restricted ketogenic diet with glutamine targeting in late-stage experimental glioblastoma, and more recent work has shown that ketogenic diets can reshape the tumor immune microenvironment and even induce an inflammatory reactive astrocyte phenotype that slows glioma growth.

Methionine restriction operates through a different axis entirely. Many tumor cells, including glioma cells, display methionine dependence, an inability to proliferate when the essential amino acid methionine is scarce. Because methionine feeds one-carbon metabolism, polyamine synthesis, and epigenetic regulation through S-adenosylmethionine, withdrawing it can cripple tumor growth and sensitize cells to chemotherapy. The review highlights work showing that methionine restriction does not induce MGMT, the resistance enzyme that blunts temozolomide, and thereby greatly improves chemotherapy efficacy in orthotopic mouse models. Other studies have shown that restricting methionine and cysteine together sensitizes gliomas to ferroptosis, an iron-dependent form of cell death, and that methionine deprivation can alter the epigenome in H3K27M-mutant glioma. Yet the clinical translation has barely begun: only two small clinical trials, both combining methionine-free diets with chemotherapy, are included in the entire evidence base.

On the clinical side, the review’s findings are candid. Studies such as the ERGO pilot trial of a ketogenic diet in recurrent glioblastoma, the KEATING randomized feasibility study, the ERGO2 trial of calorie-restricted ketogenic diets and fasting alongside reirradiation, and a series of phase I safety trials in newly diagnosed patients consistently demonstrated that ketogenic diets are feasible and can induce ketosis, lower glucose, and produce measurable metabolic changes. Some reports described striking individual long-term survivors, but the review notes that such anecdotes emerged from small, uncontrolled, single-arm studies in which patients also received surgery, radiotherapy, chemotherapy, bevacizumab, corticosteroids, and other co-interventions. In nearly every case, the effects of the diet cannot be reliably separated from the effects of concurrent treatment. The authors conclude that clinical evidence primarily supports feasibility and metabolic activity, not independent efficacy.

The confounding problem is not trivial. Corticosteroids, routinely given to glioblastoma patients to control brain swelling, raise blood glucose and can directly counteract the metabolic goals of a ketogenic diet. Bevacizumab alters tumor vascularity. Tumor molecular subtype, particularly IDH mutation status, reshapes tumor metabolism in ways that may determine whether a glycolytic strategy can work at all. Without standardized reporting of the glucose-ketone index, careful documentation of treatment and steroid use, and longitudinal nutritional assessment, comparisons across studies become nearly impossible. The review calls explicitly for randomized trials with intention-to-treat analysis as the necessary next step before these diets can be recommended as effective glioma therapy.

For patients and families who have embraced ketogenic diets, sometimes with missionary zeal, the message is nuanced. The diets are not dangerous in most studied contexts, they genuinely change systemic and possibly intratumoral metabolism, and the biology underlying them is real. But biology is not yet medicine. The gap between a mouse whose glioma shrinks under a calorie-restricted ketogenic diet combined with glutamine targeting and a human patient whose survival depends on whether the diet independently matters remains wide. What this systematic review offers is a map of that gap: forty-three studies, a clear mechanistic rationale, consistent proof that the diets can be delivered safely, and an honest admission that no one has yet proven they work on their own. The next generation of rigorously controlled randomized trials will determine whether dietary metabolic therapy earns a place alongside surgery, radiation, and chemotherapy, or whether it remains one of oncology’s most seductive unfinished ideas.

Subject of Research: Metabolic modulation of glioblastoma through ketogenic diet and methionine restriction

Article Title: Metabolic modulation of glioblastoma by dietary intervention: a systematic review

Article References: Ghoche, M. T., Miki, K., Yuan, F., Mantica, M., Agnihotri, S., McBrayer, S. K., & Abdullah, K. G. (2026). Metabolic modulation of glioblastoma by dietary intervention: a systematic review. Journal of Neuro-Oncology, 179(3), Article 90. https://doi.org/10.1007/s11060-026-05779-x

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05779-x

Keywords: glioblastoma, ketogenic diet, methionine restriction, cancer metabolism, dietary intervention, metabolic therapy, systematic review, clinical trials, brain cancer, Warburg effect, glucose-ketone index, glioma

Autism spectrum disorder is a heterogeneous neurodevelopmental condition characterized by persistent difficulties in social communication and interaction, along with restricted, repetitive patterns of behavior or interests. Epidemiological surveillance across the globe has documented a steady rise in prevalence, intensifying the search for interventions that go beyond the current standards of care. Behavioral therapies remain the cornerstone of treatment, and medications can address certain comorbidities such as irritability or attention problems, but their efficacy for core symptoms is limited. That gap has prompted researchers to look toward metabolic approaches, including the ketogenic diet, a high-fat, low-carbohydrate regimen long used to treat refractory epilepsy.

The rationale for testing a ketogenic diet in autism rests on several converging lines of biological evidence. Ketone bodies, produced by the liver when carbohydrate intake is sharply restricted, serve as an alternative fuel for the brain and can alter neurotransmitter balance, mitochondrial function, and oxidative stress, all of which have been implicated in autism-related neurobiology. Animal models and small clinical studies have previously hinted at behavioral benefits, but large, well-controlled trials in young children have been scarce. The new study was designed to address that evidence gap with a prospective, randomized design and validated clinical outcome measures.

Between January 2024 and January 2025, the research team enrolled 62 young children diagnosed with autism spectrum disorder who were admitted to Hefei Maternal and Child Health Hospital. Participants were randomly assigned in equal numbers, 31 per group, to a treatment group or a control group. Both groups underwent an identical rehabilitation program consisting of hospital-based training on weekdays and home-based rehabilitation on weekends across a two-month period. The only difference between the arms was dietary: the treatment group additionally received a modified ketogenic diet built around ketogenic nutritional powder and a structured dietary plan.

The dietary protocol was deliberately engineered for real-world feasibility in families with young children. On training days, lunch was provided at the hospital, ensuring direct supervision of a substantial portion of daily intake, while breakfast, dinner, weekend dietary implementation, and home monitoring were managed by caregivers following the structured plan. The control group, by contrast, maintained a regular diet throughout the same rehabilitation program. This hybrid delivery model, combining clinical oversight with caregiver administration, reflects a pragmatic attempt to test whether a modified ketogenic approach can be implemented outside highly controlled metabolic wards, a persistent obstacle in pediatric dietary research.

Symptom change was assessed with two widely used clinical instruments: the Autism Behavior Checklist, known as the ABC, and the Childhood Autism Rating Scale, or CARS. Both scales quantify the severity and breadth of autism-related behaviors, with higher scores indicating more pronounced symptoms. After two months of intervention, both groups showed statistically significant reductions in ABC and CARS scores compared with their own baselines, indicating that the rehabilitation program itself conferred measurable benefit to all participants.

The decisive comparison, however, lay between the groups. When the researchers examined individual change scores, children receiving the modified ketogenic diet demonstrated significantly larger reductions on both measures. The between-group difference in ABC change scores reached statistical significance with a Mann-Whitney U statistic of 324.000 and a p-value of 0.027, while the CARS comparison yielded an even stronger signal, with U equal to 273.500 and a p-value of 0.003. In practical terms, adding the dietary intervention appeared to amplify the improvements attributable to rehabilitation training over the short term.

The authors conclude that in this cohort of young children with autism, the modified ketogenic diet combined with rehabilitation training was associated with greater short-term reductions in ABC and CARS scores than rehabilitation alone, and that the intervention demonstrated acceptable short-term feasibility. The use of ketogenic nutritional powder rather than a strictly classical ketogenic diet likely contributed to tolerability, since modified formulations can ease the burden of achieving and maintaining ketosis in children who are often selective eaters. Caregiver-mediated home implementation, supported by hospital-provided meals on training days, offers a template for how metabolic interventions might be scaled in clinical rehabilitation settings.

Nevertheless, the trial has boundaries that temper interpretation. The intervention lasted only two months, leaving open whether gains persist, plateau, or erode over longer follow-up. The sample, while adequate for detecting between-group differences, comprised children from a single hospital, raising questions about generalizability across demographic and clinical profiles. Blinding is notoriously difficult in dietary trials, as caregivers inevitably know what their children are eating, and outcome assessments may be influenced by expectations despite the use of standardized rating scales. The study was registered with the China Clinical Trial Registry under identifier ChiCTR2300075057 on 24 August 2023, and was funded by the Hefei Health Commission Research Fund, with no competing interests declared by the authors.

Even with those caveats, the results represent one of the more encouraging randomized evaluations of ketogenic dietary therapy in pediatric autism to date. They align with a broader shift in neuroscience toward viewing metabolic health as a modifiable lever in neurodevelopmental disorders, echoing the diet’s established success in epilepsy and emerging studies in conditions ranging from schizophrenia to Alzheimer’s disease. Larger, longer, and preferably multicenter trials will be needed to confirm the effect, identify which children benefit most, and define the optimal formulation and duration of treatment. For now, families and clinicians have reason for cautious optimism that a carefully supervised modified ketogenic diet may offer a meaningful complement to rehabilitation in early autism intervention.

Beyond the headline results, the trial offers a window into the practical physiology of ketogenic therapy in early childhood. When carbohydrate intake falls sufficiently low, hepatic metabolism shifts toward producing beta-hydroxybutyrate and acetoacetate, molecules that cross the blood-brain barrier and supply neurons with an energy substrate that yields more ATP per unit of oxygen than glucose. This metabolic flexibility matters in neurodevelopmental conditions, because a growing body of work links mitochondrial electron transport chain dysfunction and impaired cellular energy supply to subsets of autism spectrum disorder. By providing an efficient alternative fuel, ketogenic approaches may partially bypass compromised glucose utilization, which neuroimaging studies have occasionally documented in affected brain regions.

The clinical instruments used in the trial deserve closer attention for readers unfamiliar with pediatric assessment tools. The Autism Behavior Checklist originated in the 1980s as one component of the Autism Screening Instrument for Educational Planning and contains 57 items spanning sensory, relational, body and object use, language, and social domains, typically completed by caregivers or teachers. The Childhood Autism Rating Scale, developed by Eric Schopler and colleagues, contrasts a child’s behavior against age-typical norms across fifteen domains, with scores above a conventional threshold supporting an autism diagnosis. Because both instruments rely on observed behavior rather than biomarkers, their scores can fluctuate with context, rater training, and the child’s daily state, which is why randomized designs that balance such noise across arms carry particular weight.

Safety considerations loom large whenever a ketogenic regimen is proposed for young children, whose growth demands stable nutrition. Classical ketogenic diets, long deployed in epilepsy centers, carry recognized risks including constipation, dyslipidemia, kidney stones, and slow weight gain, and they typically require vitamin and mineral supplementation alongside careful monitoring of lipids, urine ketones, and growth trajectories. Modified variants relax the strict fat-to-carbohydrate ratios of the classical approach, trading some ketone elevation for improved palatability and adherence. In this study, the use of a powdered nutritional formulation with a structured plan suggests an attempt to standardize intake precisely because free-form ketogenic cooking in children, who frequently exhibit food selectivity, is difficult to control and document.

The shared improvement observed in both arms also merits interpretation. Intensive rehabilitation, with its predictable routines, structured social exposure, and engaged caregivers, is itself an active intervention, and the design wisely treated it as the comparator standard rather than an untreated baseline. Against that backdrop, the dietary supplement’s incremental effect on rating scale scores, while statistically robust by nonparametric testing, was measured over a window short enough that regression to the mean and heightened caregiver attention in the treatment group cannot be fully excluded. Objective correlates, such as sleep quality, gastrointestinal symptoms, attention measures, or biochemical markers of ketosis, would strengthen causal confidence in future work.

Looking forward, the field would benefit from trials that enroll diverse populations, stratify participants by metabolic or mitochondrial phenotype, and track outcomes well beyond the initial months of intervention. Dose-response relationships, the durability of ketosis achieved at home, and the identification of responder subgroups all represent open questions. If replicated at scale, a metabolically informed adjunct to behavioral rehabilitation could reshape early intervention pathways, but until then the present findings are best viewed as a promising signal that justifies expansion of controlled investigation rather than a basis for unsupervised dietary change outside clinical guidance.

Subject of Research: A randomized controlled trial evaluating a modified ketogenic diet as an adjunct therapy for children with autism spectrum disorder.

Article Title: Evaluating the therapeutic efficacy of a modified ketogenic diet in children with autism spectrum disorder: a randomized controlled trial

Article References: Liu, L., Zhao, Q., Zhou, J., Liu, Z., Wu, D., & He, K. (2026). Evaluating the therapeutic efficacy of a modified ketogenic diet in children with autism spectrum disorder: a randomized controlled trial. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07708-3

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07708-3

Keywords: autism spectrum disorder, ketogenic diet, randomized controlled trial, BMC Pediatrics, pediatric rehabilitation, ketosis, Autism Behavior Checklist, Childhood Autism Rating Scale, metabolic therapy, neurodevelopmental disorders, nutritional intervention, child health