Brain & Development xxx (2017) xxx–xxx

www.elsevier.com/locate/braindev

Original article

Ketogenic diet using a Japanese ketogenic milk for patients with

epilepsy: A multi-institutional study
a,⇑ b b c
Tomohiro Kumada , Katsumi Imai , Yukitoshi Takahashi , Shin Nabatame , Hirokazu
d
Oguni
a
Department of Pediatrics, Shiga Medical Center for Children, Japan
b c
National Epilepsy Center, Shizuoka Institute of Epilepsy and Neurological Disorders, Japan
Department of Pediatrics, Osaka University Graduate School of Medicine, Japan
d
Department of Pediatrics, Tokyo Women’s Medical University, Japan
Received 20 June 2017; received in revised form 18 October 2017; accepted 12 November 2017

Abstract

Background: In Japan, Meiji 817-B (M817-B), a powdered ketogenic milk, has been available since the ketogenic diet was intro-duced
to infants and tube-fed children with medication-resistant epilepsy in the 1980s.
Methods: We retrospectively evaluated the efficacy, tolerability, and side e ffects of the ketogenic diet using M817-B as the main source
of daily food intake for patients with epilepsy by sending questionnaires to the members of a subcommittee of the Japan Epilepsy Society
that focuses on the proper use of M817-B.
Results: A total of 42 patients were enrolled. Age at the initiation of the diet therapy ranged from 3 to 244 months (median, 32.5
months). Thirty-four patients were fed via tube, and the remaining 8 were fed orally. About 93% of patients were able to continue the diet
for 1 month, 74% for 3 months, and 64% for 6 months. The median period of continuation was 16 months. One patient was able to
continue as long as 7 years. The ketogenic ratio was maintained at about 3.0. The seizure-free rate and responder (>50% seizure reduction)
rate were about 10% and 30–40%, respectively during the 12 months on the diet. Mean serum beta-hydroxybutyrate increased to almost 4
mM at 1 month and was maintained during the diet period. Side e ffects, which required discontinuation of the diet therapy, occurred in 11
of 42 patients and included hypertonia, weight loss, vomiting, hypoglycemia, metabolic acidosis, and hypokalemia.

Conclusion: M817-B could be used long-term with demonstrated efficacy in seizure reduction, although there are some side e ffects that
may require cessation of the diet therapy.
2017 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.

Keywords: Ketogenic diet; Ketogenic milk; Epilepsy; Meiji 817-B; Japan

1. Introduction the KD was introduced as early as the 1950s in Japan, only

a limited number of institutions have employed the KD for
The ketogenic diet (KD) has been used worldwide for the treatment of epilepsy [1]. Meiji 817-B (M817-B), a
the treatment of medication-resistant epilepsy. Although unique ketogenic milk, was developed in 1981 by the Meiji
Corporation (Tokyo) and is provided free of cost by the
⇑ company as a voluntary service to patients in Japan. It is a
Corresponding author at: Department of Pediatrics, Shiga Medical
Center for Children, 5-7-30, Moriyama, Moriyama City, Shiga Prefecture, powdered ketogenic milk and is designed to have a
Japan. ketogenic ratio of 3.0 when dissolved in water and is easy
E-mail address: tkumada@mccs.med.shiga-pref.jp (T. Kumada). to prepare. The recent

https://doi.org/10.1016/j.braindev.2017.11.003
0387-7604/ 2017 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
2 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx
worldwide ‘‘KD boom” originating from Johns Hop-kins
Hospital (Baltimore, Maryland, USA) has drawn the
attention of Japanese physicians and parents of Japanese
children with epilepsy. As a result, the demand for M817-B
has increased every year, even though the KD is still
considered to be a last resort for treatment of intractable
epilepsy in Japan [2].
We have not found any reports that have assessed the
usefulness of M817-B for epilepsy. Thus, we carried out
this retrospective study to clarify the efficacy, tolerabil-ity,
and side effects of M817-B in Japanese patients with
epilepsy.
2. Patients and methods
We investigated the efficacy, tolerability, and side
effects of M817-B in patients with epilepsy treated with the
formula by sending questionnaires to members of a
subcommittee of the Japan Epilepsy Society that focuses
on the proper use of ketogenic formulas.
2.1. Subjects
Patients who started the KD for the treatment of epi-
lepsy using M817-B as their main source of daily meals at
4 hospitals (Shiga Medical Center for Children, Shiga;
National Epilepsy Center, Shizuoka Institute of Epilepsy
and Neurological Disorders, Shizuoka; Osaka University
Graduate School of Medicine, Osaka; and Tokyo Women’s
Medical University, Tokyo) from Jan-uary 4, 2006 to
March 31, 2016 were enrolled. Those who had pyruvate
dehydrogenase complex deficiency or glucose transporter
type 1 deficiency were excluded. Those who took
ketogenic formulas other than M817-
B such as Ketonia , KetoCal , KetoVie , KetoVolve , and
KetoCuisine , or solid ketogenic meals orally dur-ing the
initial 3 months on the diet therapy were also excluded.
Patients’ profiles were investigated as follows: age at the
time of the investigation (October 31, 2016), sex,
underlying etiology of epilepsy, age at seizure onset,
classification of epileptic syndrome, seizure type, seizure
frequency, interictal electroencephalographic (EEG)
findings, and number of antiepileptic drugs (AEDs) used
previously. We also investigated age at the initiation of the
KD, nutritional routes, and concomitant AEDs.
2.2. Dietary specifics
M817-B was provided for the patients as basic meals.
When a ketogenic ratio higher than 3.0 was needed, oil or
fresh cream was allowed to be added, and when a
ketogenic ratio lower than 3.0 was needed, commercially
available regular milk was allowed to be added. There was
no common start-up protocol for the KD using M817-B
because this was a retrospective study. We investigated the
ketogenic ratio, nutritional routes, and
Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
food consistency, i.e., liquid, paste, or solid, at 1, 3, 6, 12,
24 months, and at the final visit.
2.3. Efficacy in seizure reduction
We assessed the efficacy of the diet therapy for reduc-
ing the seizure frequency of each patient and each sei-zure
type and for interictal EEG improvement at 1, 3, 6, 12, and
24 months on the diet therapy. Serum beta-hydroxybutyrate
was also checked at each visit. Efficacy in seizure
reduction was defined as ‘‘seizure-free” when seizures
disappeared, ‘‘excellent” when a >90% seizure reduction
was achieved, ‘‘good” when a >50% seizure reduction was
achieved, ‘‘poor” when a >50% seizure reduction was not
achieved, and ‘‘worse” when seizure frequency was
aggravated. Efficacy on interictal EEG findings was
defined as ‘‘excellent” when paroxysmal discharges
disappeared, ‘‘good” when >50% of paroxys-mal
discharges were reduced, ‘‘poor” when <50% of
paroxysmal discharges were reduced, and ‘‘worse” when
paroxysmal discharges increased. We also investigated the
correlation between serum beta-hydroxybutyrate and
efficacy in seizure reduction. EEG data were col-lected
between 10 and 14 months (12-month data) and between
22 and 26 months (24-month data). Modifica-tion of
concomitant AEDs was allowed during the study period.
2.4. Tolerability and side effects
We assessed the tolerability and side effects of the diet
therapy. We demonstrated the cumulative continuation of
the KD using Kaplan-Meier analysis. We researched the
reasons for discontinuation of the diet therapy. We also
investigated the side effects that required medical
intervention or discontinuation of the diet therapy. Because
this was a retrospective study, the need for and timing of
medical interventions to treat the side effects were left to
the discretion of the medical providers.
3. Results
3.1. Patient profiles
We analyzed 42 patients (25 male and 17 female
patients) from 4 hospitals (20 from Shiga, 10 from Tokyo,
8 from Shizuoka, and 4 from Osaka). The patient profiles,
including the underlying etiologies, epi-lepsy
classification, seizure phenotypes, interictal EEG findings,
number of previously used AEDs and con-comitant AEDs,
age at seizure onset, and KD initiation, are shown in Table
1. All but 2 patients had daily sei-zures. AEDs prescribed
concomitantly with diet therapy at initiation included
valproic acid (n = 23), phenobar-bital (n = 13), lamotrigine
(n = 12), topiramate
 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx 3

(n = 10), levetiracetam (n = 10), zonisamide (n = 8), nasoduodenal tube, 5 via gastrostomy tube, and 8 via oral
clonazepam (n = 6), clobazam (n = 5), phenytoin (n = 3), suckling. During the diet therapy, 3 patients’ nutri-tional
sulthiame (n = 2), nitrazepam (n = 2), and bromide (n = 2), routes were changed; 2 had a gastrostomy tube substituted
as well as ethosuximide, clorazepate, vigabatrin, for a nasogastric or nasoduodenal tube, respectively; and 1
carbamazepine, and vitamin B6 (n = 1, respectively). required a nasogastric tube instead of oral suckling. Three
About 90% of patients (37 of 42) started the KD before 6 patients added solid ketogenic foods to their M817-B
years of age. About half of the patients (23 of 42) started intake at 6, 12, and 24 months on the diet, respectively.
diet therapy within 2 years from seizure onset.
The mean ketogenic ratio (grams of fat per that of
3.2. Diet specifics carbohydrate and protein) was maintained at approxi-
mately 3.0 while on the diet therapy (2.8 ± 0.5 at 1 month,
Nutritional routes for M817-B intake at the initiation of n = 39; 2.9 ± 0.3 at 3 months, n = 31; 2.9 ± 0.6 at 6 months,
the diet were follows: 20 via nasogastric tube, 9 via n = 26; 2.9 ± 0.5 at 12 months, n = 23;
Table 1
Patient profiles.
Men/Women 25/17
Underlying etiologies Severe asphyxia 12
Chromosomal aberrations 7 Trisomy 13 (n = 2), trisomy 18 (n = 3), trisomy 21,
1p36.3 deletion
Single genetic disorders 5 SCN2A, SPTAN1, CDKL5, KCNQ2, MeCP2
duplication
Brain malformations 6 Heterotopia, hemimegalencephaly, lissencephaly,
FCD (n = 2), TSC
Congenital anomaly 1 Cardiofacial cutaneous syndrome
Neurodegenerative diseases 4 Leukodystrophy (n = 2), Gaucher disease, DRPLA
Unknown 7
Epilepsy classification Infantile spasms 26
Ohtahara syndrome 3
PME 1
Other generalized epilepsy 2
Focal epilepsy 10
Seizure phenotypes Epileptic spasms 31
Focal seizure 10
GTCs 6
Tonic seizure 4
Myoclonic seizure 2
Interictal EEG findings Hypsarrhythmia 24
Multifocal spikes 12
Focal spikes 3
Diffuse SWC 3
Suppression-burst 1
Number of previously used AEDs 0–12 (median, 6.5)
Number of concomitant AEDs at initiation 0–4 (median, 3)
Age at onset of seizures 0–84 mo (median, 3 mo)
Age at initiation of ketogenic diet 3–244 mo (median, 32.5 mo)
Duration from seizure onset to initiation of diet 0–241 mo (median, 22 mo)
Nutritional routes Nasogastric tube 20
Nasoduodenal tube 9
Gastrostomy tube 5
Oral suckling 8
Supplements L-carnitine 31
Multivitamins 23
Citrate 19
Calcium 15
Selenium 9
Biotin 3
Abbreviations: FCD, focal cortical dysplasia; TSC, tuberous sclerosis complex; DRPLA, dentatorubral-pallidoluysian atrophy; PME, progressive myoclonic
epilepsy; GTCs, generalized tonic-clonic seizures; EEG, electroencephalogram; SWC spike-and-wave complex; AEDs, antiepileptic drugs; mo, months.

Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
4 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx
and 2.9 ± 0.8 at 24 months, n = 17). Some supplements
such as L-carnitine, multivitamins, citrate, calcium, sele-
nium, and biotin were used during the KD as shown in
Table 1.
3.3. Efficacy in seizure reduction
As shown in Fig. 1, the seizure-free rate was 5/42 (12%)
at 1 month, 3/42 (7%) at 3 months, 4/42(10%) at 6 months,
3/42 (7%) at 12 months, and 1/42 (2%) at 24 months. The
responder (>50% seizure reduction) rate was 17/42 (40%)
at 1 month, 16/42 (38%) at 3 months, 17/42(40%) at 6
months, 14/42 (33%) at 12 months, and 10/42 (24%) at 24
months. As shown in Table 2, the rates of EEG
improvement (‘‘excellent” + ‘‘good”) were 9/42 (22%) at 1
month, 6/42 (14%) at 3 months, 8/42 (20%) at 6 months,
3/42 (7%) at 12 months, and 5/42 (12%) at 24 months.
Efficacy in decreasing seizure frequency was also
measured according to seizure type. The numbers of
patients with seizure types such as focal seizures (n =
10), secondarily or primarily generalized tonic-clonic sei-
zures (n = 6), tonic seizures (n = 4), and myoclonic sei-
zures (n = 2) were too small to evaluate the efficacy of
M817-B in specific types. Therefore, only the efficacy
in spasm reduction (n = 31) was investigated. Regarding
spasms, the seizure-free rate was 4/31 (13%) at 1 month,
2/31 (6%) at 3 months, 3/31(10%) at 6 months, 2/31
(6%) at 12 months, and 1/31 (3%) at 24 months. The
responder (>50% seizure reduction) rate was 13/31
(42%) at 1 month, 11/31 (35%) at 3 months, 11/31
(35%) at 6 months, 9/31 (29%) at 12 months, and 7/31
(23%) at 24 months. Specifically, for infantile spasms,
hypsarrhythmia diminished partially or completely in
Fig. 1. Efficacy in seizure reduction. The responder rate was calculated as the sum of the ‘‘seizure-free ” (blue), ‘‘excellent” (green), and ‘‘good” (pastel blue)
categories. Discontinued: those who discontinued the ketogenic diet. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)
Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
8/24 patients (33%) after 1 month on the diet (data not
shown).
Mean serum beta-hydroxybutyrate increased to almost 4
mM at 1 month and was maintained during the diet therapy
(3.8 ± 1.6 mM at 1 month, n = 38; 3.7 ± 1.6 mM at 3
months, n = 25; 4.0 ± 2.0 at 6 months, n = 26; 4.0 ± 1.9
mM at 12 months, n = 21; and 3.7 ± 2.4 mM at 24 months,
n = 15). As shown in Fig. 2, in which 38 patients at 1
month on the diet were included, there was no correlation
between ketogenic ratio and serum beta-hydroxybutyrate
because the levels ranged widely, from 0.6 to 7.4 mM (4.0
± 1.5 mM), at a ketogenic ratio of 3.0. Also, there was no
correlation between efficacy in seizure reduction and
serum beta-hydroxybutyrate at each visit on the diet (data
not shown).
3.4. Tolerability and side effects
Fig. 3 shows Kaplan-Meier estimates of cumulative
continuation rates.
The duration of KD therapy ranged from 0 to 84 months
(median, 16 months). The rates of discontinua-tion were
3/42 (7%) at 1 month, 11/42 (26%) at 3 months, 15/42
(36%) at 6 months, 19/42 (45%) at 12 months, and 25/42
(60%) at 24 months on the diet. Thirty-one patients had
discontinued the diet therapy by the end of the
investigation, and the reasons for dis-continuation were as
follows: 9 for ineffectiveness, including 1 patient with
infantile spasm who achieved significant seizure reduction
but did not develop motor milestones; 4 for seizure
recurrence; 1 for seizure aggra-vation; 11 due to side
effects; 3 because of death; 3 because of cure; and 2 for
other reasons. Three patients
 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx 5

Table 2
Efficacy in reduction of seizure activity on interictal EEG.
1 month 3 months 6 months 12 months 24 months
Number on diet (%) 39 (93) 31 (74) 27 (64) 23 (55) 17 (40)
Excellent 5 (12) 3 (7) 4 (10) 2 (5) 2 (5)
Good 4 (10) 3 (7) 4 (10) 1 (2) 3 (7)
Poor 13 (31) 9 (21) 11 (26) 6 (14) 5 (12)
Worse 1 (2) 1 (2) 0 (0) 1 (2) 0 (0)
Lost to EEG examination 16 (38) 15 (36) 8 (19) 13 (31) 7 (17)
Discontinued diet 3 (7) 11 (26) 15 (36) 19 (45) 25 (60)
N = 31.
Values are number (percentage).
Discontinued diet: those who discontinued the ketogenic diet.
Lost to EEG examination: those who continued the ketogenic diet but did not undergo EEG.

Fig. 2. Relationship between serum beta-hydroxybutyrate and ketogenic ratio. Serum beta-hydroxybutyrate ranged widely from 0.6 to 7.4 mM at a
ketogenic ratio of 3.0.

Fig. 3. Kaplan-Meier estimate of the cumulative continuation rate. The rates of continuation were 93% at 1 month, 55% at 12 months, and 40% at 24
months. The longest duration of maintenance of the ketogenic diet was 84 months.

Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
6 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx

died during the diet therapy because of acute bacterial
pneumonia (n = 1) or sudden unexpected death in epi-lepsy
(n = 2), which might not have been associated with the diet
therapy.
Side effects that required medical intervention or dis-
continuation of the diet therapy occurred in 21 of 42
patients. Eleven patients discontinued the diet therapy
because of the following side effects: 1 developed hypo-
kalemia without improvement with oral potassium; 4 lost
weight; 1 had hypoglycemia and metabolic acidosis; and 5,
all of whom were fed M817-B via tube, had vom-iting or
hypertonia (thought to be caused by nausea or abdominal
discomfort). The other 10 patients continued the diet
therapy with medical intervention for the fol-lowing side
effects: 3 developed metabolic acidosis, all of whom
required sodium bicarbonate; 11 had hyper-uricemia, all of
whom required allopurinol; 1 experi-enced bone fracture
and needed plaster cast fixation; 2 developed renal stones
and were treated with citric acid and a diuretic; 3
developed hyponatremia and used oral salt; 1 had
hypokalemia and used oral potassium; 3 experienced
alopecia due to biotin deficiency with improvement after
biotin supplementation; 1 developed diabetes mellitus with
hypertriglyceridemia and was treated with insulin; and 1
had severe constipation with-out improvement with
laxatives. We classified the side effects by timing of
symptoms and need for discontinu-ation of the diet, and
these are shown in Table 3. Over-all, 16/39 (41%) side
effects occurred within 1 month on the diet, and about half
of those were hyperuricemia. Side effects that required
discontinuation of the diet occurred at any time, even after
2 years.
4. Discussion
M817-B ketogenic powder has been provided to patients
with epilepsy, pyruvate dehydrogenase complex
deficiency, or glucose transporter type 1 deficiency free of
charge by the Meiji Corporation as a voluntary ser-vice in
Japan. Thus, M817-B has been used by many Japanese
patients for more than 20 years; however, there have been
no reports about the efficacy, tolerability, or side effects of
M817-B for patients with epilepsy, even though this
ketogenic formula has been mentioned in some articles
about KD written by Japanese researchers [1,3].
A comparison of M817-B and other ketogenic formu-las
used worldwide is shown in Table 4. The KD using a
ketogenic formula can be easily prepared and given to
infants orally or to older children and adolescents via tube
[4,5]. In this study, 34 patients were fed via tube, and the
remaining 8 were fed orally. About 93% of the patients
were able to continue the diet for 1 month, 74% for 3
months, and 64% for 6 months. One patient continued for 7
years. Only 5/42 (12%) patients discon-tinued the diet
because of vomiting or hypertonia (which we thought
reflected the patients’ hesitation and rejec-tion of M817-B)
perhaps because of stomach irritability caused by the
medium chain triglycerides (MCT) con-tained in M817-B.
Thus, the palatability of M817-B was very good compared
to solid ketogenic foods taken orally, and only around 30%
of patients were not able to continue because of
restrictiveness [4]. Our responder rate of about 30–40%
was less than that in a previous prospective report about
classical KD and a report about the KD using other
ketogenic formulas, in which, responder rates were over
50% [5,6]. This may have been due to the lower ketogenic
ratio in our study (about 3.0) than that in the previous
studies (about 4.0). Many of the responders achieved
seizure reduction within 1 month, which was similar to a
previous report [7]. M817-B improved epileptic spasms
and hypsarrhythmia in patients with infantile spasms
refractory to adreno-corticotropic hormone (ACTH)
therapy, which was also

Table 3
Side effects requiring medical intervention and discontinuation of the ketogenic diet.
Period 0–1 month 1–3 months 3–12 months 12–24 24–36 months 36+ months
months
Total number of side 16 8 4 3 5 3
effects
Side effects which needed Hypertonia Vomiting/ Vomiting Weight loss Weight loss (2)
hypertonia
Discontinuation of the Hypoglycemia/ Hypertonia (2) Weight loss Hypokalemia
diet acidosis
Side effects which needed Acidosis Hyperuricemia (2) Hyperuricemia Acidosis (2)
Diabetes mellitus
Medical intervention Hyperuricemia (9) Hyponatremia (2) Alopecia Alopecia Alopecia
Constipation Vomiting Bone fracture
Hyponatremia Renal stone
Hypokalemia
Renal stone
Total number of side effects includes the sum of all patients who required discontinuation of the diet as well as those who required medical intervention
during each period. Number in parentheses is that of patients who developed each side effect.

Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx 7

Table 4
Comparison of ketogenic formulas available worldwide.
Product M817-B KetoCal (3:1) KetoCal (4:1) KetoCal (4:1) Ketonia KetoVie KetoVie KetoVolve KetoCuisine
Unflavored Vanilla Vanilla/Unflavored Vanilla Chocolate Vanilla
Company Meiji Nutricia Nutricia Nutricia Namyang Cambrooke Cambrooke Metabolica Medica Nutrition
Form Powder Powder Powder Liquid Liquid Liquid Liquid Powder Powder
Ketogenic ratio 3 3 4 4 4 4 4 4 5
Composition/100 kcal
Net Carbohydrate 1.2 g 1.0 g 0.5 g 0.4 g 0.8 g 0.1 g 0.3 g 0.4 g 1.4 g
Protein 2.0 g 2.2 g 2.1 g 2.1 g 1.7 g 2.4 g 2.2 g 2.1 g 0.6 g
Fat 9.7 g 9.7 g 9.9 g 9.9 g 10.0 g 9.8 g 9.8 g 10.3 g 10.2 g
MCT% of fat (g) 55.2% 0% 0% 0% 0% 33.8% 31.2% 20.1% 0%
Other nutrients
L-carnitine – – 6.4 mg 5.5 mg 4.2 mg 13.9 mg 12.8 mg – Not found
Biotin – 2 lg 2.7 lg 2.7 lg 3 lg 2.1 lg 1.9 lg 1.7 lg Not found
Selenium – 2.7 lg 4.7 lg 4.7 lg – 6.3 lg 5.8 lg 4.3 lg Not found
Net carbohydrates, total carbohydrates minus total fiber; MCT, medium-chain triglycerides.

reported with the classical ketogenic diet [8]. The fre-
quency of side effects such as hyperuricemia, hypona-
tremia, metabolic acidosis, and renal stones, was almost the
same as those in previous reports about the classical
ketogenic diet, although hypokalemia and dia-betes
mellitus, which only a few patients experienced in this
study, were rare complications in previous reports about
KD [9–11]. A future task will be to determine whether or
not these complications may be unique side effects of
M817-B.
M817-B has a unique advantage over other ketogenic
formulas in that it includes more MCT than other keto-
genic formulas, as shown in Table 4. The MCT KD pro-
vides 40–55% of total calories from MCT. Therefore, the
KD using M817-B as the main source of daily meals
mimics the MCT KD [12]. Because MCT ingestion results
in much more ketone production than long chain
triglycerides (LCT) ingestion, M817-B can be expected to
produce abundant ketone bodies; however, we found no
experimental research involving human clinical stud-ies
comparing M817-B to other ketogenic formulas. In this
study, mean serum beta-hydroxybutyrate at a keto-genic
ratio of 3.0 was almost 4 mM and was sufficient as a
ketogenic formula. MCT are likely to induce gastroin-
testinal symptoms such as abdominal discomfort and
diarrhea; however, no patients experienced diarrhea in our
study. We speculate that MCT mixed with pow-dered milk
(M817-B) may be associated with less gas-trointestinal
side effects than MCT alone.
On the other hand, M817-B has some disadvantages
versus other ketogenic formulas. First, M817-B lacks some
trace elements and vitamins such as selenium, car-nitine,
and biotin. In this study, 3 patients experienced alopecia,
which was improved with biotin supplementa-tion.
Previously, we reported a reduction in serum sele-nium in
the use of M817-B without supplementation of selenium
[13]. Second, to raise the ketogenic ratio above 3.0,
additional labor is required to mix the M817-B with oil or
fresh cream. This may be the reason why many
patients were offered the ketogenic diet with a ketogenic
ratio of 3.0 in this study. Third, in the near future, it will be
difficult to steadily supply M817-B to patients with
epilepsy because the demand for M817-B has been
increasing every year, and M817-B production is cur-rently
dependent on the altruistic spirit of the Meiji Corporation.
In conclusion, M817-B could be used as a long-term
resource for patients on the KD, with proven efficacy in
seizure reduction, although some side effects are of par-
ticular concern and may be responsible for discontinuation.
Disclosure
None of the authors has any conflicts of interest to
disclose. We confirm that the design of this study was
approved by the ethical committees of all of the institu-
tions to which the authors belong.
Acknowledgement
M817-B is provided free of cost by the Meiji Corpo-
ration for patients in Japan. Because many patients in this
study had severe underlying etiologies of epilepsy, such as
severe asphyxia, chromosomal aberrations, brain
malformations, and neurodegenerative diseases, their
families may have experienced significant financial burden
providing for their daily nutrition in addition to their
epilepsy treatment. Therefore, they must be very grateful
for receiving M817-B free of charge.
References
[1] Oguni M, Oguni H, Ito S, Ito Y, Osawa M. Re-evaluation of
ketogenic diet therapy for childhood refractory epilepsy. No To
Hattatsu (in Japanese) 2009;41:339–42.
[2] Ito S, Oguni H. Ketogenic diet for intractable childhood epilepsy; as
an early option as well as a last resort. Brain Nerve (in Japanese)
2011;63:393–400.

Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
8 T. Kumada et al. / Brain & Development xxx (2017) xxx–xxx
[3] Hirano Y, Oguni H, Shiota M, Nishikawa A, Osawa M. Ketogenic
diet therapy can improve ACTH-resistant West syn- drome in Japan.
Brain Dev 2015;37:18–22.
[4] Kossoff EH, McGrogan JR, Freeman JM. Benefits of an all-liquid
ketogenic diet. Epilepsia 2004;45:1163.
[5] Hosain SA, La Vega-Talbott M, Solomon GE. Ketogenic diet in
pediatric epilepsy patients with gastrostomy feeding. Pediatr Neurol
2005;32:81–3.
[6] Vining EPG, Freeman JM, Ballaban-Gil K, Camfield CS, Camfield
PR, Holmes GR, et al. A multicenter study of the efficacy of the
ketogenic diet. Arch Neurol 1998;55:1433–7.
[7] Kossoff EH, Laux LC, Blackford R, Morrison PF, Pyzik PL, Hamdy
RM, et al. When do seizures usually improve with the ketogenic
diet? Epilepsia 2008;49:329–33.
[8] Hong AM, Turner Z, Hamdy RF, Kossoff EH. Infantile spasms
treated with the ketogenic diet: prospective single- center experience
in 104 consecutive infants. Epilepsia 2010;51:1403–7.
Please cite this article in press as: Kumada T et al. Ketogenic diet using a Japanese ketogenic milk for patients with epilepsy: A multi-institu-tional study.
Brain Dev (2017), https://doi.org/10.1016/j.braindev.2017.11.003
[9] Kang HC, Chung DE, Kim DW, Kim HD. Early- and late-onset
complications of the ketogenic diet for intractable epilepsy. Epilepsia
2004;45:1116–23.
[10] Kossoff EH, Zupec-Kania BA, Amark PE, Ballaban-Gil KR,
Christina Bergqvist AG, Blackford R, et al. Optimal clinical
management of children receiving the ketogenic diet: Recommen-
dations of the International Ketogenic Diet Study Group. Epilepsia
2009;50:304–17.
[11] Aguirre Castaneda RL, Mack KJ, Lteif A. Successful treatment of
type1 diabetes and seizures with combined ketogenic diet and
insulin. Pediatrics 2012;129:e511–4.
[12] Neal E. The medium-chain triglyceride ketogenic diet. In: Neal E,
editor. Dietary treatment of epilepsy. Practical implementation of
ketogenic therapy. Chichester: Wiley-Blackwell; 2012. p. 78–88.
[13] Hayashi A, Kumada T, Nozaki F, Hiejima I, Miyajima T, Fujii T.
Changes in serum levels of selenium, zinc and copper in patients on a
ketogenic diet using Ketonformula. No To Hattatsu (in Japanese)
2013;45:288–93.