• Ketogenic diet;
  • Clinical management;
  • Drug-resistant epilepsy


  1. Top of page
  2. Summary
  3. Consensus Recommendations
  4. Acknowledgment
  5. Disclosure
  6. References

Ketogenic diet is a nonpharmacologic treatment for childhood epilepsy not amenable to drugs. At the present time, two works based on national research, one in Germany and one in the United States provide international guidelines to ensure a correct management of the ketogenic diet. Our Italian collaborative study group was set up in order to formulate a consensus statement regarding the clinical management of the ketogenic diet, patient selection, pre–ketogenic diet, counseling, setting and enforcement of dietary induction of ketosis, follow-up management, and eventual discontinuation of the diet.

The ketogenic diet (KD) is a nonpharmacologic treatment for childhood epilepsy not amenable to drugs (Stafstrom & Rho, 2004; Freeman et al., 2007) that mimics the biochemical response to starvation, when ketone bodies, rather then glucose, become the main fuel for the brain energy demand. In the past decade, interest in ketogenic diet spread worldwide (Kossoff & McGrogan, 2005; Lord & Magrath, 2010; Neal & Cross, 2010). At present, two works based on national research, one in Germany (Klepper et al., 2004) and one in the United States (Kossoff et al., 2008b) provide international guidelines to ensure a correct management of the KD. Our Italian collaborative study group was set up to create a consensus statement regarding the clinical management of the KD. The aim of the study is to establish more standardized protocols and management recommendations for Italian groups of child neuropsychiatrist, clinical nutritionists, and dietitians. The main suggestions concern patient selection, pre-KD counseling, setting and enforcement of dietary induction of ketosis, follow-up management, and eventual KD discontinuation.

Consensus Recommendations

  1. Top of page
  2. Summary
  3. Consensus Recommendations
  4. Acknowledgment
  5. Disclosure
  6. References

Patient selection

The two main indications (Table 1) to the ketogenic diet are glucose transporter protein 1 (GLUT-1) deficiency syndrome (Klepper & Leiendecker, 2007; Pons et al., 2010) and pyruvate dehydrogenase deficiency (PDHD) (Wexler et al., 1997). In both conditions the enzymatic defect causes a disorder of brain energy metabolism. The KD is the treatment of choice because the brain can use ketones as an alternative energy source, bypassing the obstacle in physiologic metabolism. GLUT-1 deficiency syndrome is an inborn error of glucose transport across the blood–brain barrier, characterized by a variable combination of seizures, development delay, acquired microcephaly, spasticity, and a complex movement disorder (Leen et al., 2010). The KD should be considered as a first-line therapy for this disease (Wang et al., 2009; Veggiotti et al., 2010). PDHD is a severe mitochondrial disease where pyruvate cannot be metabolized into acetyl-CoA (Wexler et al., 1997; McWilliam et al., 2010; Barnerias et al., 2010). The KD is particularly useful for control of severe lactic acidosis in patients with this disorder.

Table 1.   Main indications for dietary therapy
Metabolic disorders
 Glucose transporter protein 1 (GLUT-1) deficiency
 Pyruvate dehydrogenase deficiency (PDHD)
 Mitochondrial respiratory chain complex disorders
 Severe myoclonic epilepsy of infancy (Dravet syndrome)
 Medically refractory epilepsy
 Severe intolerance to AEDs

Some cases of positive response to the diet are described in metabolic disorders such as phosphofructokinase deficiency (Swoboda et al., 1997), glycogenosis type V (Busch et al., 2005), and mitochondrial respiratory chain complex disorders (Kang et al., 2007a). In patients with medically refractory epilepsy or severe intolerance to antiepileptic drugs (AEDs), the KD is an important, potentially efficacious alternative. Some evidence indicates that patients with severe myoclonic epilepsy of infancy (or Dravet syndrome) (Caraballo et al., 2005, 2011) and myoclonic–astatic epilepsy may respond positively to the diet (Caraballo et al., 2006). Several retrospective studies reported that 60–75% of children with refractory seizures obtained a >50% decrease in their seizures (Kinsman et al., 1992; Freeman et al., 1998; Casey et al., 1999; Vining, 1999; Katyal et al., 2000; Coppola et al., 2002; Kossoff & McGrogan, 2005; Henderson et al., 2006; Freeman et al., 2007; Coppola et al., 2010; Lefevre & Aronson, 2010). The KD should be strongly considered in a child who failed two to three anticonvulsant therapies, regardless of age, and particularly in those with symptomatic generalized epilepsies.

The KD was shown to be particularly effective in infantile spasm (Hong et al., 2010), even before trying any anticonvulsant or steroid (Kossoff et al., 2008a,b; Kossoff, 2010). Furthermore, children receiving either concurrent vagus nerve stimulation or zonisamide may show preferential benefit when the KD is started (Kossoff et al., 2007b; Morrison et al., 2009). On the contrary children with partial seizures tend to have a worse response to the KD (Stainman et al., 2007). Clinical trials are currently ongoing on patients with Alzheimer disease, amyotrophic lateral sclerosis, migraine, and brain tumors (Van der Auwera et al., 2005; Zhao et al., 2006; Seyfried et al., 2008, 2009).

KD is contraindicated in several specific inborn errors of metabolism that could lead to a severe metabolic crisis. These metabolic diseases should, therefore, be ruled out before starting the diet. Patients with a disorder of fat metabolism might develop a severe worsening of the disease under KD. Therefore, before initiating the KD, children must be screened for disorders of fatty acid transport and oxidation (Kossoff et al., 2008a,b). An inborn metabolic error at any point along the pathway of fatty acid transportation by carnitine can lead to a devastating catabolic crisis (i.e., coma, death) during fasting or the KD. Deficiency of pyruvate carboxylase, a mitochondrial enzyme that catalyzes the conversion of pyruvate to oxaloacetate, can impair tricarboxylic acid cycle function and energy production in patients on the KD. The diet can exacerbate acute intermittent porphyria.

Pre–ketogenic diet counseling

Several important prerequisites assessing the patient eligibility for the KD ensure both safety and maximization of chances of success (Table 2). Pediatric, neurologic, and nutritional consults before the introduction of KD are recommended. An evaluation of epilepsy is needed to identify seizure type(s), frequency, and etiology; the clinician should also review all past and current AEDs. A wakefulness and a sleep electroencephalography (EEG) (24 h EEG if needed) and brain magnetic resonance imaging (MRI) will be useful to identify those patients who are susceptible to surgical treatment. A cognitive or developmental assessment is needed to evaluate the neuropsychological outcome. As part of diagnostic work-up in progressive epileptic encephalopathies, a full serum and urine metabolic evaluation (urine organic acids, serum amino acids, ammonium, lactic acid, serum acylcarnitine profile) should be performed if no clear etiology for the child’s epilepsy was identified.

Table 2.   Pre-KD evaluation
  1. aIf these last two tests are not available, the use of predictive equations of basal metabolic rate and wrist x-ray could be performed

Neurologic evaluation
 Seizure type
 Seizure frequency
 AEDs and other medication review
 EEG/Holter EEG
 Cognitive/development assessment
 Full serum and urine metabolic evaluation
Pediatric evaluation
 ECG if history of heart disease
 Abdomen ultrasound
 Laboratory analysis
Nutritional evaluation
 Baseline weight, height, and ideal weight for stature
 Body mass index (BMI)
 Skinfold thickness measurement
 Dietary history
 Bioelectrical impedance analysis
 Indirect calorimetrya
 Dual energy x-ray absorptiometry (DEXA)a

Heart and abdominal ultrasound is important to rule out metabolic disorders unsuited to the diet and to point out any complication (presence of kidney stones, dyslipidemia, liver disease, gastroesophageal reflux, and cardiomyopathy). Laboratory evaluation before starting the KD should include complete blood count with platelets, serum, liver, and kidney tests (including albumin, prealbumin, ammonium, AST, ALT, ALP, γGT, total and direct bilirubin, blood urea, nitrogen, and creatinine), a fasting lipid profile (triglycerides, total cholesterol, high-density lipoproteins, low-density lipoproteins, and atherogenic index), blood sugar level, electrolytes (paying particular attention to serum bicarbonate, total protein, calcium, zinc, selenium, magnesium, and phosphate), blood gas analysis, parathyroid hormone and vitamin D, osteocalcin (if osteopenia), urinalysis and 24 h urine calcium and creatinine, and antiepileptic drug levels (if applicable).

Dietetic and nutritional evaluation consists of measurement of baseline weight, height, body mass index (BMI) and ideal weight for stature, skinfold thickness measurement, bioelectrical impedance analysis, indirect calorimetry (basal metabolism evaluation), and dual energy x-ray absorptiometry (DEXA; if the latter two tests are unavailable, the use of predictive equations of basal metabolic rate and wrist x-ray should be performed). A dietary history (7-day food record, food preferences, allergies, aversions, and intolerances) is needed.

All these investigations are strongly recommended in first-time patient evaluation; during the patient follow-up, tests should be chosen based on clinical course.

Before starting the diet, it is also crucial to discuss psychosocial issues. The physician should ensure that parents or caregivers understand their involvement in administering the KD to their child, specifically the importance of strict adherence to the diet, avoidance of carbohydrates, need for multivitamin and mineral supplementation, and awareness of potential adverse effects (Kossoff et al., 2008a,b).

Patients’ families are often worried about how much time is needed to obtain therapeutic success with the KD; the clinician is advised to discuss this choice with the child’s parents and to indicate a minimum of 3 months.

Setting and enforcement of dietary induction of ketosis

The traditional method of initiating KD involves a period of fasting, with no carbohydrate-containing fluids and, periodically monitoring serum glucose (Freeman et al., 2006). Duration of fasting varies from 12–48 h, depending on the time required to achieving adequate ketosis; after reaching a level of beta-hydroxybutyrate in the blood >2 mm, food may be progressively administered. The caloric contribution of meals is increased daily in one-third caloric intervals to reach full calories meals; an adequate level of beta-hydroxybutyrate is considered to be 2–5 mm, 7 days after the diet. Fasting may result in hypoglycemia, acidosis, nausea, vomiting, dehydration, and lethargy; therefore, instituting the traditional KD protocol in the hospital is indicated to prevent any possible complication.

Nowadays it’s evident that fasting may be appropriate when it’s necessary to obtain a quicker response to the diet (Freeman & Vining, 1999; Kossoff et al., 2008b), but it is not necessary for long-term efficacy. Retrospective (Kim et al., 2004) and prospective studies indicate that gradual initiation protocols offer the same seizure control at 3 months compared to traditional KD protocol, together with significant lower frequency and severity of initiation related side effects (Bergqvist et al., 2005).

The KD can also be started in outpatients, in selected cases, following a standardized procedure to screen neurologic, general pediatric, metabolic, and nutritional conditions before administration. Potential advantages of gradually starting the diet outside the hospital include reduced stress for the child, lower risk of hypoglycemia and dehydration, reduced number of laboratory analysis, and reduced costs.

The optimal way to administer the KD depends on the type of feeding. Children who have normal oral feeding can get the diet in the food that is prepared following specific dietary indications, whereas enterally (including gastrostomy and jejunostomy) fed children can use a formula-based KD, which is generally simpler for dietitians to calculate; the formula is also often used as add on in children who are orally fed.

Classic KD is calculated in grams of fat to grams of proteins plus carbohydrates. The most common ratio is 4 g of fat to 1 g of protein plus carbohydrate (described as ‘‘4:1”). This means that 90% of the energy comes from fats and 10% from proteins and carbohydrates combined. Sometimes it is necessary to provide the KD at a lower ratio to increase intake of proteins or carbohydrates (often “3:1”). There is some evidence that a 4:1 ratio, when used at the start, may be more advantageous for the first 3 months (Seo et al., 2007). During the first month, calories are typically restricted to 75% of the daily recommendations for age, to promote ketosis. But this reduction is not necessary; underweight children should not follow this protocol and all children should increase calories gradually over time to the regular recommendations for age.

Similarly, fluid restriction to 90% is based on consolidated habits rather than on scientific evidence. We do not suggest fluid restriction for children on KD; the fluid administration (noncaloric fluids) is individualized (1–10 kg need 80 ml/kg; 10–20 kg need 800 ml + 40 ml/kg; >20 kg need 1,200 ml + 20 ml/kg) and increased contextually to the child’s level of activity or adjusted for climate; in infants fluids should be increased to 100 ml per kilogram of body weight.

The KD may also be easily administered to enterally fed children. Prescription of a formula-based KD is generally needed in the case of coma, low caloric intake by mouth due to conditions such as respiratory and gastrointestinal problems, severe neurologic impairment, failure to thrive, poor compliance with traditional KD, and for children younger than 12 months of age. The transition from an enteral diet to KD should be gradual (Table 3). The formula-based KD is easily administered and is used also in mouth-fed children to adjust caloric intake and substitute regular food in special circumstances (on a journey, etc.). To prepare a formula-based KD, in Europe a commercial product is currently available. KetoCal (for North America: Nutricia, Rockville, MD, U.S.A. and for Europe: SHS International, Liverpool, U.K.) is a milk protein-based, powdered formula that, added to water, provides either a 3:1 or 4:1 KD.

Table 3.   Transition from enteral diet in Ketocal (4 days)
StageDaysEnteral diet (%)Ketocal (%)
11–2Energy 75Energy 25
21–2Energy 50Energy 50
31–2Energy 25Energy 75
4 Energy 0Energy 100

In the past few years, modified dietary approaches have been developed for the treatment of epilepsy, including the modified Atkins diet (Kossoff et al., 2006, 2007b; Kang et al., 2007b), the MCT diet (Liu, 2008; Neal et al., 2008) and the low-glycemic diet (Pfeifer et al., 2008). Unlike the classic KD, the modified Atkins diet is started without hospitalization and does not require precise weighing of food ingredients and portions. The daily carbohydrate consumption in the modified Atkins diet is 10 or 20 g (Kossoff et al., 2007b). However, there are no limitations in protein, fluids, and calories. This diet is easier to administer in adolescents and adults than in children.

Because of limited food choice, the classic KD is also deficient in minerals and vitamins, and needs to be supplemented with sugar-free products. Inadequate calcium intake and limited sun exposure can impair bone mineralization in children at risk of osteopenia and osteoporosis due to long-term antiepileptic drug (AED) therapy. Therefore, both vitamin D and calcium should be supplemented (Bergqvist et al., 2007). Additional supplementation (zinc, selenium, magnesium, phosphorus, and so on) using standard multivitamin products is suggested. Carnitine oral supplementation (50 mg/kg/die) is needed if laboratory levels are low (Coppola et al., 2006) or children exhibit symptoms indicating hypocarnitinemia such as generalized weakness, excessive fatigue, and decreased muscle strength.

Follow-up management

Follow-up of children receiving KD includes regular neurologic, nutritional, and pediatric evaluations (Table 4). The child should be examined by the neurologist initially at the 8th and 15th day after hospital discharge and EEG performed (if the child is is younger than 1 year of age or an epileptic encephalopathy is present) or at least after 1, 3, 6, 9, and 12 months in the first year of treatment. Cognitive and developmental assessment should be performed 6 and 12 months after introducing the KD. Electrocardiography (ECG), heart ultrasound and an abdominal echo should be performed every 6 months.

Table 4.   Follow-up KD management
Neurologic assessment
 Neurologic evaluation (at 1–3–6–12 months)
 Electroencephalography (at 1–3–6–12 months)
 Review efficacy of the diet
 Cognitive/development evaluation (at 6–12 months)
Pediatric assessment
 Electrocardiography (every 6 months)
 Abdominal echo (every 6 months)
 Laboratory evaluation (at 1–3–6–12 months)
 Complete blood count with plates
 Serum liver and kidney tests
 Blood sugar level
 Blood gas analysis
 Laboratory evaluation (at 3–6–12 months)
 Fasting lipid profile
 Parathormone and vitamin D
 Osteocalcin (if osteopenia)
 Urinalysis and 24 h urine calcium and creatinine (only if previously altered)
 Anticonvulsant drug levels
Nutritional assessment
 Assess compliance to therapy
 Height and body mass index (BMI)
 Skinfold thickness measurements
 Bioelectrical impedance analysis
 Indirect calorimetry (each 3 months)
 Dual energy x-ray absorptiometry or wrist x-ray (every 6–12 months)
 Review appropriateness of diet prescription (calories, protein, and fluid)
 Review vitamin and mineral supplementation

The laboratory evaluation at 1, 3, 6, and 12 months after introduction of the KD includes complete blood count with platelets, serum, liver and kidney tests (including albumin, prealbumin, ammonium, AST, ALT, ALP, Gamma-GT, total and direct bilirubin, blood urea, nitrogen, and creatinine), glycemia, electrolytes (especially serum bicarbonate, total protein, calcium, zinc, selenium, magnesium, and phosphate) and blood gas analysis.

At 3, 6, and 12 months, the following evaluations should be performed: fasting lipid profile (triglycerides, total cholesterol, HDL, LDL), parathyroid hormone and vitamin D, osteocalcin (if osteopenia), urinalysis, 24 h urine calcium and creatinine (only if previously altered), and AED levels (if applicable).

The dietetic and nutritional evaluation consists of a review of diet and compliance, the measurement of weight, height, BMI, skinfold thickness, and bioelectrical impedance analysis at each visit; indirect calorimetry (basal metabolism evaluation) every 3 months, DEXA, or wrist x-ray every 6–12 months.

At home, routine urine ketosis evaluation should be performed by parents twice per day (morning and evening) at least in the first 3 months, and then once a week; the optimal level of ketones in the urine is 8–16 mm evaluated by keto-diastix (Bayer Health Care, Milano, Italy). Serum β-hydroxybutyrate (BHB) should be measured every 12 h until its stabilization, and then regular controls should be performed at 1, 3, 6, 12, and 24 months or in case of symptoms referable to hyperketosis or worsening of seizures. It seems that BHB is better correlated to seizure reduction than are ketones in the urine. Therefore, it is preferable to use BHB to monitor the KD even if BHB is measured less frequently than urinary ketones (Van Delft et al., 2010). Serum glycemia should be evaluated every 12 h (every 6 h in the first 48 h if the child is aged <12 months); optimal level of serum glycemia is 2.5–5 mm.

Glycemic values <2.5 mm and/or ketones values >6.5 mm, if symptomatic (sweating, palor, and tremor), require immediate oral administration of glucose, maltodextrin, or fruit juice; in case of impaired alertness, administration of intravenous glucose is essential.

Growth parameters such as weight and height should be regularly controlled during the first year of the KD, in order to monitor appropriate weight gain for age and length; infants younger than 2 years of age should be monitored more frequently (almost weekly) to prevent growth disturbance (Vining et al., 2002). If a child is overly hungry or refuses his meal, calorie contribution should be adjusted accordingly.

As in all medical therapies, side effects can occur during the diet and neurologists, clinical nutritionists, and dietitians need to be alerted (Ballaban-Gil et al., 1998; Wheless, 2001). However, the risk of serious adverse events is low. Metabolic abnormalities include hypoproteinemia that causes muscular mass loss or alteration in bone metabolism; hyperlipidemia or hypercholesterolemia, reported in 14–59% of children on KD (Chesney et al., 1999; Kwiterovich et al., 2003; Kang et al., 2004) that can cause atherosclerosis; hypocalcemia (2%) that can worsen a preexisting osteopenia or osteoporosis; potassium or selenium deficit that can cause cardiac abnormalities; hyperuricemia (2–26%), hypomagnesemia (5%), decreased amino acid levels, and acidosis (2–5%) (Schwartz et al.,1989; Chesney et al., 1999; Kang et al., 2004). Metabolic abnormalities can be prevented by careful monitoring.

Gastrointestinal symptoms including vomiting, constipation, diarrhea, and abdominal pain occur in 12–50% of children on the KD (Kang et al., 2004); renal calculi occur in 3–7% (Furth et al., 2000; Kossoff et al., 2002; Sampath et al., 2007) making regular monitoring necessary (Choi et al., 2010). Potassium citrate supplementation is warranted to reduce the incidence of kidney stones (McNally et al., 2009).

Long-term complications in children treated with KD for >2 years have been reported (Groesbeck et al., 2006). In the small population studies, the higher risk was referred to bone fractures, kidney stones, and decreased growth. KD discontinuation depends on the patient’s response to the diet; even if apparently ineffective, the diet must be maintained for at least 3 months (Freeman et al., 2006) before considering discontinuation. In children with >50% seizure response, the KD is often discontinued after approximately 2 years; however, in children in whom seizure control is nearly complete and side effects are low, the diet can be prolonged up to 6–12 years (Groesbeck et al., 2006). A longer diet duration is needed for children with GLUT-1 and PDHD deficiency syndromes. No information is available about the maximum duration of the KD, but seizure-free patients on the diet for >10 years have been reported. A recent study revaluated the long-term outcomes in safety and efficacy of KD after its discontinuation, providing reassuring results with regard to seizures and adverse effects (Patel et al., 2010).

In assessing the effectiveness of the KD, rate of seizure reduction should not be the only parameter (Beniczky et al., 2010), duration of improved seizure control and the possible global improvement of patient well being are also important.


  1. Top of page
  2. Summary
  3. Consensus Recommendations
  4. Acknowledgment
  5. Disclosure
  6. References

This work was obtained from a Consensus established in Verona on November 2009.


  1. Top of page
  2. Summary
  3. Consensus Recommendations
  4. Acknowledgment
  5. Disclosure
  6. References

Drs. Veggiotti, Tagliabue, Coppola, and Guerrini and Dalla Bernardina have received support for research and/or consulting from Nutricia. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.


  1. Top of page
  2. Summary
  3. Consensus Recommendations
  4. Acknowledgment
  5. Disclosure
  6. References