Multiple epidemiologic studies agree that the burden of pediatric mental health diagnoses is exponentially increasing (1). Commonly cited explanations are diverse, ranging from increased use of and dependence on social media, increased obesity rates, and decreased medical insurance coverage for mental health support. Yet, while additional funding and availability of psychopharmacological therapies and mental health practitioners clearly help, current evidence-based approaches are not enough. Classical medical therapies often have severe and lifelong adverse side effects, especially for the developing brain. Improvement to the desired level of independent functioning frequently demands polypharmacy, further amplifying the risk of side effects.

Use of a ketogenic diet as a medical therapy for pediatric epilepsy was first described by Dr. Russell Wilder in the Mayo Clinic Proceedings (2). Since that time, support for its efficacy at or above the level of standard antiseizure medications has solidified its keystone role (3). In conjunction with Dr. Wilder, a pediatrician Mynie Peterman noted parallel improvements in behavior and cognition on a ketogenic diet (4). However, these benefits are often seen as secondary, beneficial only as compared to the lack of an adverse outcome, or simply fortuitous. Our perspective is that this benefit needs to be brought front and center as a primary outcome, so that we can identify children with neurodevelopmental impairments, with or without seizures, who may require it most.

Within the subtext of epilepsy studies, there is evidence ranging from anecdotal to statistically-significant secondary outcomes that improvements in other domains besides epilepsy are present. This is highly relevant, as ~80% of children with epilepsy have comorbid behavioral or cognitive impairment (5) and the highest risk factor for autism spectrum disorder in children with epilepsy is intellectual disability (6). Given the cascade of paradigm-shifting studies in adult psychiatric disease, it stands to reason that children may also benefit from a ketogenic diet or nutritional ketosis (7) – and given the existing expertise in pediatric epilepsy centers, pediatric practitioners are well-equipped to initiate and study this powerful intervention. We acknowledge that evidence-based use of the ketogenic diet in pediatrics is currently limited to intractable epilepsy; however, we suggest that “off-label” indications are supported by anecdotal evidence while the academic community mobilizes to initiate objective, randomized controlled trials with standardized, quantitative outcome measures to improve the standard of care for multiple common neuropsychiatric conditions. Although review of proposed mechanisms of the ketogenic diet is beyond the scope of this paper, we suggest that the multifactorial, pleiotropic effects of the ketogenic diet in different clinical contexts converge upon and pave the way for a common therapeutic improvement for an increasingly broadening range of common neuropsychiatric conditions (Figure 1).

While a strict 4:1 classic ketogenic diet may be necessary for drug-resistant epilepsy (9), recent studies suggest a less-restrictive Modified Atkins Diet is often sufficient to yield demonstrable benefit (10, 11). If the goal of using a ketogenic diet is to improve a child’s mental health so that they more easily socialize and learn in parallel with other children in a regular school environment, a comparatively less restrictive protocol that can be initiated at home rather than during an inpatient hospital admission is ideal. The effective and therapeutic degree of ketosis in an individual is exquisitely patient-specific and does not correlate perfectly to the ketogenic diet ratio across individuals. An intervention with a higher degree of flexibility and reduced risk for side effects is more likely to be successful in the long-term.

There are several versions of a ketogenic diet available for use. Our version of the Modified Atkins Diet (MAD), also called Modified Ketogenic Diet (MKD), consists of 40–60 g total carbohydrate per day with 1–2 tablespoons fat added to each meal. Fat sources are often butter, mayo, and oil. If feasible for patients, MCT oil is the preferred fat source and is encouraged at each meal. MCT oil will bypass the usual fat metabolism to be used more immediately for conversion to ketones, thus expediting ketogenesis by bypassing digestive enzymes and the carnitine shuttle (12). In addition, use of MCT oil may allow for a comparatively higher carbohydrate and protein intake without compromising nutritional ketosis (3).

A challenge that’s imperative upon all pediatric practitioners is to identify inborn errors of metabolism that benefit from specific dietary therapies as early as possible. A subset of these (e.g., phenylketonuria) are readily identified on the Newborn Screen, a primarily biochemical screening test with modest state-to-state variability. However, other syndromes are not able to be screened for in a blood spot or do not emerge until later in infancy/childhood. For example, Glut1 deficiency syndrome is a genetic-metabolic condition that impairs transport of glucose across the blood brain barrier; it is diagnosed by identification of pathogenic variants in SLC2A1 and/or low absolute or relative glucose in the cerebrospinal fluid, in the proper clinical context. Most children do not develop symptoms until early infancy or later, and have broad phenotypic variability ranging from severe epilepsy to disruptive movement disorders and mild–moderate intellectual disability (13). Because a ketogenic diet improves not only seizure control but also cognitive development, motor function, and freedom from disabling hemiplegic migraines, it is critical that testing is made available early in life and with minimal invasiveness when possible (i.e., genetic or blood screening rather than diagnostic lumbar puncture). Similarly, other genetic/metabolic diagnoses where a ketogenic diet is considered a potential first-line therapy because glucose supply to the brain is impaired include glycogen storage disorders (type III/Forbes’ disease and type V/McArdle’s disease), phosphofructokinase deficiency, pyruvate dehydrogenase complex deficiency, and some mitochondrial disorders (14). In contrast, there are specific inborn errors of metabolism where implementation of a ketogenic diet is contraindicated and would be dangerous, including: carnitine deficiency (primary), carnitine palmitoyltransferase (CPT) I or II deficiency, carnitine translocase deficiency, beta-oxidation deficiencies including medium-chain acyl dehydrogenase deficiency (MCAD), long-chain acyl dehydrogenase deficiency (LCAD), short-chain acyl dehydrogenase deficiency (SCAD), long-chain 3-hydroxyacl-CoA deficiency, medium-chain 3-hydroxyacl-CoA deficiency, pyruvate carboxylase deficiency and porphyria (15). We bring these up to suggest that a recommendation to start a ketogenic diet as well as metabolic and genetic testing should be thoughtfully considered (see Table 1 for suggested testing) by a knowledgeable healthcare team, regardless of the indication (though this often is linked to the etiology).

Parents often wish to know potential adverse effects of a ketogenic diet in parallel with feasibility considerations. During the transition period onto a ketogenic diet, common issues such as dehydration, relative hypoglycemia, nausea, and decreased energy can be addressed with increased consumption of fluid-containing electrolytes, and quickly resolve. Further corrections for abnormal electrolytes, metabolic acidosis, vitamin deficiencies, and constipation are possible. Historically, negative pediatric health outcomes have included linear growth deceleration, increased gastroesophageal reflux and constipation, and increased risk of nephrocalcinosis, with some consideration for increased risk of osteoporosis and secondary carnitine deficiency (with decreased efficacy of a ketogenic diet and/or associated cardiomegaly), and hepatotoxicity in adulthood. In our experience, these side effects are rare when the ketogenic diet is initiated as a medical therapy and followed by an experienced dietitian in even the youngest patients (16). Using a more liberal ketogenic diet, such as the MAD/MKD that brings about a lower level of nutritional ketosis, will lower the risk of potential side effects (11). As many children take liquid medication with a high carbohydrate content, dose formulation must be considered and often adjusted to crushed tablets or opened capsules; parents must also be aware of antibiotics, antipyretics, and common supplements that may be suggested by physicians and available over the counter in sugary syrups. Overall, the highest risk is perhaps parental dissatisfaction or inability to adhere to a ketogenic diet, prompting discontinuation. For example, implementation of a ketogenic diet in children with intractable epilepsy has been demonstrated to increase parental stress (17). In an ideal situation, a multidisciplinary team is able to provide parental education and support to initiate and maintain the diet with the understanding that it is a major lifestyle change impacting the entire family; an unfortunate reality is that this may not be possible for many biopsychosocial reasons in all cases where it is medically indicated.

Given the number of children who have been treated with a ketogenic diet for epilepsy, what can we learn about other changes that they have experienced in their health in parallel? In two studies from Johns Hopkins, parental goals for starting a ketogenic diet cited improvement in cognition as the second most common reason for initiating the consult, second only to potential reduction in seizure frequency (18, 19). Remarkably, this goal was achieved in more than 50% of families. In a cohort of fifty children from the Netherlands, a positive cognitive and behavioral impact was noted in children with refractory epilepsy treated with a ketogenic diet as compared to a control group over a 4 month period, independent of any changes in seizure frequency (20). Parental questionnaires reflected lowered anxiety, mood-disturbed behavior, and higher productivity; in parallel, objective cognitive tests demonstrated improved receptive vocabulary and reaction time in the absence of any medication changes. If and when a ketogenic diet may be recommended for adolescent mood disorders is an active area of research with initial promise (21, 22). A key question moving forward is what are the optimal measures by which to quantify improved mood and cognition, as well as other non-seizure outcomes.

As we expand our practice to include a broader swathe of neuropsychiatric diagnoses, we continue to draw from our past experiences with children with intractable epilepsy. Below are case examples where other appreciable benefits of a ketogenic diet were clearly noted.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

TG: Writing – original draft, Writing – review & editing. RB: Writing – original draft, Writing – review & editing.