Embed Size (px)
Citation preview
K
A
K
I
t
FUs
7
0d
Applied nutritional investigation
Breath acetone predicts plasma ketone bodies in children with epilepsyon a ketogenic diet
athy Musa-Veloso, Ph.D.a, Sergei S. Likhodii, Ph.D.b, Exequiel Raramac, Stephanie Benoitd,Yeou-mei Christiana Liu, M.S., R.D., C.H.E.S.e, Dominic Chartrandd,
Rosalind Curtis, M.D., F.R.C.P.(C)e, Lionel Carmant, M.D.d, Anne Lortie, M.D.d,Felix J. E. Comeauc, and Stephen C. Cunnane, Ph.D.a,*
a Department of Nutritional Sciences, University of Toronto, Toronto, Ontario, Canadab Department of Pharmacology, University of Toronto, Toronto, Ontario, Canada
c Alcohol Countermeasure Systems, Mississauga, Ontario, Canadad Department of Neurology, Hôpital Ste-Justine, Montréal, Quebec, Canada
e Bloorview MacMillan Children’s Centre, Toronto, Ontario, Canada
Manuscript received November 18, 2004; accepted April 19, 2005.
bstract Objective: The high-fat ketogenic diet has long been used to treat refractory childhood seizures, butwhether there is a relation between the degree of ketosis and effectiveness of seizure control remainsunclear. Frequent measurements of plasma ketones are difficult in children so the goal was todetermine the utility of breath acetone as a marker of systemic ketosis and seizure control in childrengiven the ketogenic diet because of seizures refractory to medication.Methods: In experiment I, breath acetone and plasma ketones were assessed every 2 h during an8-h test day in seven children. In experiment II, a preliminary assessment of the possible relationbetween breath acetone and seizure frequency was made over 14 d in five children and oneadolescent on the ketogenic diet.Results: Breath acetone was positively and curvilinearly related to plasma acetone (r2 � 0.99, P� 0.0001), plasma acetoacetate (r2 � 0.89, P � 0.0001), and plasma �-hydroxybutyrate (r2 � 0.94,P � 0.0001). No significant relation was found between breath acetone and seizure frequency orchange in seizure frequency.Conclusions: Breath acetone is indicative of systemic ketosis while on the ketogenic diet. How-ever, owing to the wide range of seizure types and plasma acetone, more subjects will be needed todetermine whether there is a clear link between breath acetone and seizure frequency or decreasedseizure frequency while on the high-fat ketogenic diet. © 2006 Elsevier Inc. All rights reserved.
Nutrition 22 (2006) 1–8www.elsevier.com/locate/nut
eywords: Acetoacetate; �-Hydroxybutyrate; Breath acetone analyzer; Gas chromatography; Isopropanol
akSbfa[sob
ntroduction
The high-fat ketogenic diet has been used to treat refrac-ory childhood seizures for over 80 y but its mechanism of
The Bloorview MacMillan Children’s Hospital Foundation, Dairyarmers of Canada, NSERC, Stanley Thomas Johnson Foundation, and theniversity of Toronto Awards Division are thanked for their financial
upport.* Corresponding author. Tel.: 819-821-1170, ext. 2670; fax: 819-829-
141.
fE-mail address: [email protected] (S.C. Cunnane).899-9007/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.oi:10.1016/j.nut.2005.04.008
ction remains unknown. Whether there is a threshold ofetosis for effective seizure control on this diet is unclear.ome clinical studies have demonstrated a positive relationetween ketosis and seizure control [1–3], but others haveound no significant association [4–6]. Therefore, the situ-tion remains similar to that expressed by Bridge and Iob7] more than 70 y ago: “Frequently, no improvement re-ults in spite of severe ketosis, and at times, good results arebtained without the formation of ketone bodies � it haseen impossible to establish any constant correlation.”
Despite the unclear role of ketosis in seizure control,
requent monitoring of ketones (acetoacetate [AcAc], �-
htTbqnpbkmsi
rc[llcmsbmhamssn
wkreafr
M
cSafw
E
D
darwhca
B
bUltf1Maopm
app
TP
Sn
12345567
2 K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
ydroxybutyrate [�-HBA], or acetone) is an integral aspect ofhe ketogenic diet regimen in the home and the clinical setting [8].he most widely used ketone test in the home is the nitroprusside-ased urinary dipstick. This non-invasive test provides a semi-uantitative measurement of urinary AcAc concentration but doesot reliably represent blood ketone values [3,9–12]. In contrast,lasma ketone analysis is invasive, particularly because frequentlood sampling would be required to assess the relation betweenetosis and seizure control. Therefore, a less invasive measure-ent of systemic ketosis would be useful to determine whether
eizure control on the high-fat ketogenic diet depends on achiev-ng a certain degree of ketosis.
We recently demonstrated that breath acetone is aeliable measurement of plasma ketone levels in chroni-ally ketotic rats [13] and in mildly ketotic adult humans14]. We also showed that children with refractory epi-epsy on the high-fat ketogenic diet have breath acetoneevels that are about 125 times higher than that in healthyhildren or children with epilepsy that is controlled byedication [15]. Breath acetone is most commonly mea-
ured by gas chromatography, which is a laboratory-ased method. In collaboration with Alcohol Counter-easure Systems Co. (Mississauga, ON, Canada), we
ave developed and tested a prototype hand-held breathcetone analyzer (BAA) intended for rapid, home-basedeasurement of breath acetone. The model used in this
tudy provided acetone values in less than 1 min and wasensitive to breath acetone in the range of 500 to 9000mol/L (unpublished data).
The objective of experiment I therefore was to determinehether breath acetone could be used to predict plasmaetone level (AcAc, �-HBA, or acetone) in children whoseefractory seizures were treated using the ketogenic diet. Inxperiment II, the objective was to evaluate whether there issignificant relation between breath acetone and seizure
requency or change in seizure frequency in children withefractory seizures on a ketogenic diet.
able 1atient information (experiment I)
ubjectumber
Sex Age (y) Weight (kg) (percentile)
Before KD During KD
M 9 28.6 (66) 26.5 (21)M 19 38.3 (�5) 40.5 (�5)F 11 24.4 (92) 35.1 (29)M 10 30.0 (48) 33.6 (50)
a† F 16 67.6 (88) 50.2 (23)b† F 16 67.6 (88) 51.8 (30)
M 11 30.0 (10) 27.2 (�5)‡ F 12 33.3 (21) 36.6 (16)
F, female; KD, ketogenic diet; M, male; MCT, medium-chain triacylgl* The KD ratio refers to the ratio, in grams, of fat to (carbohydrate plu† Subject number 5 was studied twice, once when the ketogenic diet ra
‡ Breath samples but not blood samples were collected from subject number 7aterials and methods
Both experiments were approved by the ethical reviewommittees of the University of Toronto, the Hospital forick Children, Bloorview MacMillan Children’s Hospital,nd Hôpital Ste-Justine. All guardians and subjects wereully informed of the experimental procedures before givingritten consent and assent.
xperiment I: breath acetone and plasma ketones
esignThe characteristics of the seven subjects, their ketogenic
iet formulations, and length of time on the ketogenic dietre presented in Table 1. Each subject arrived at the labo-atory between 6:30 and 9:00 AM. Blood and breath samplesere collected at fasting and every 2 h for 8 consecutiveours (five samples in total) during a single day. Subjectsonsumed their home-prepared ketogenic meals and drinksnd took their medications at the usual times.
lood sampling and analysisFive minutes before blood collection, an electric heating
lanket was placed on the hand to stimulate vasodilatation.sing disposable lancet pens, 375 �L of blood was col-
ected from fingertips into 75-�L heparinized capillaryubes. To separate the plasma, blood samples were centri-uged in 2.5-mL heparinized Microfuge tubes at 2500g for0 min at 4°C. An enzymatic assay kit (Sigma, St. Louis,O, USA) was used to analyze plasma �-HBA. The AcAc
ssay developed by Harano et al. [16] was used, but insteadf using 50 �L of plasma for the AcAc analysis, 25 �L oflasma was diluted in 25 �L of water to avoid reaching theaximum detection limit of the assay [14].Plasma acetone was analyzed as the dinitrophenylhydr-
zine derivative by high-performance liquid chromatogra-hy using protein free plasma extracts. To remove plasmaroteins, 50 �L of plasma was mixed with 100 �L of
Height (cm) (percentile) KD ratio* MCT oil (g/d)
Before KD During KD
124.5 (18) 125.0 (�5) 4:1 0159.0 (�5) 160.0 (�5) 4:1 18117.0 (83) 145.0 (36) 3.5:1 0137.0 (57) 138.4 (37) 40% MCT 93162.5 (60) 163.0 (54) 4:1 0162.5 (60) 163.0 (54) 3.5:1 0144.0 (38) 144.0 (38) 4:1 0148.5 (50) 153.2 (43) 3:1 0
n).4:1 and once when the ketogenic diet ratio was reduced to 3.5:1.
ycerols proteitio was
.
ap22vTc(psicolvpw
B
mtatmt(B
E
kasaBttptd
cuoevast(
wtradltw
S
taItitdos(rytcgddttt
TP
Sn
123456
m
3K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
cetonitrile and centrifuged at 2500g for 10 min at 4°C. Torepare sufficient reagent for the derivation of 50 samples,.5 mg of 2,4-dinitrophenylhydrazine were dissolved in a.5-mL mixture of concentrated HCl, distilled water (4:6ol/vol), and incubated in a water bath for 20 min at 60°C.o derivatize the acetone, 50 �L of plasma supernatant wasombined with 50 �L of 2,4-dinitrophenylhydrazineSigma). Acetone standards were prepared by using controllasma spiked with known amounts of acetone. Derivatizedamples were loaded onto an autosampler and automaticallynjected into a C18 Symmetry column (5-�m particles, 25m � 4.2 mm inner diameter; Waters, Milford, MA, USA)f a high-performance liquid chromatographic system (Agi-ent 1100 Series, Agilent, Palo Alto, CA, USA). The ultra-iolet absorbance detector was set to 365 nm. The mobilehase was 63:37 (vol/vol) acetonitrile:water. The flow rateas 1.0 mL/min.
reath sampling and analysisBreath acetone was analyzed in all subjects by gas chro-
atography [14,15]. The coefficients of variation acrossriplicate breath and calibrator samples were 6.7 � 0.7%nd 2.2 � 0.5%, respectively. In four subjects, breath ace-one was also analyzed using the BAA prototype. The re-aining three subjects were incapable of using the BAA or
heir breath acetone values were below its sensitivity limit�500 nmol/L). The coefficient of variation across triplicateAA measures averaged 7.2 � 0.7%.
xperiment II: breath acetone and seizure frequency
The physical characteristics of the six subjects, includingetogenic diet type, length of time on the ketogenic diet, andnticonvulsant medications, are listed in Table 2. Someubjects in experiment I were also in experiment II. Breathcetone was the only ketone measured in experiment II.reath samples were obtained in triplicate, one to three
imes daily from each subject for 2 wk. Caregivers main-ained a log of daily seizure frequency throughout the studyeriod. Subject numbers 2, 3, and 6 were physically unableo use the BAA so they were visited at home once or twiceaily, and breath acetone was collected for analysis by gas
able 2atient information (experiment II)
ubjectumber
Sex Age (y) Time onKD (mo)
KD ratio MCT oil (g/d)
M 9 10 4:1 0M 19 22 4:1 18M 13 31 3:1 15M 10 8 40% MCT 93F 11 64 3.5:1 0M 8 35 3.25:1 0
AEDs, antiepileptic drugs; CBZ, carbamazepine (Tegretol); CLB, cloba
ale; MCT, medium-chain triacylglycerol; PB, phenobarbital; TPM, topiramate (hromatography. Subject numbers 1, 4, and 5 were able tose the BAA so their breath acetone was measured by BAAnly. Weekly home visits were conducted to calibrate andnsure optimal functioning of the BAA. During the homeisits, breath samples were also collected for subsequentnalysis by gas chromatography to determine whether anyubjects were exhaling isopropanol, a metabolite of acetonehat can yield a false breath acetone value on the BAAMusa-Veloso, unpublished data).
Throughout the 2-wk study, subject number 4, who wasithout seizure, had breath acetone levels that were below
he sensitivity limitations of the BAA. To investigate theeason for these unexpectedly low values, this subjectgreed to participate in three additional 8-h tests on separateays, in each of which triplicate breath samples were col-ected hourly and analyzed by gas chromatography. Duringhese 3 d of the study, medications and ketogenic mealsere consumed as usual.
tatistical analyses
All values are expressed as mean � standard error ofhe mean. A repeated measures one-way analysis of vari-nce was performed using SAS statistical software (SASnstitute, Cary, NC, USA) to determine whether any ofhe measurements were significantly different from fast-ng values (P � 0.05). Tukey’s test was then conductedo determine where significant differences existed. Toetermine the relation between breath acetone and eachf the plasma ketones (experiment I), non-linear regres-ion analyses were conducted using GraphPad PrismGraphPad, San Diego, CA, USA). To determine theelations between plasma ketones, linear regression anal-ses were conducted using GraphPad Prism. To evaluatehe relation between breath acetone measured by gashromatography and that measured by BAA, linear re-ression analysis was conducted. The statistical methodescribed by Bland and Altman [17] was employed toetermine the agreement between breath acetone as de-ermined by gas chromatography versus that measured byhe BAA. To evaluate possible within-subject and be-ween-subject relations between breath acetone and sei-
t (kg) (%) Height (cm) (%) AEDs before KD AEDs during KD
1) 125.0 (�5) None None5) 160.0 (�5) CLB, CBZ CLB, CBZ5) 149.0 (13) CLB None
0) 138.4 (37) LTG, TPM LTG, TPM9) 145.0 (36) VPA VPA, CLB, PB5) 115.0 (�5) VPA, CLB CLB
risium); F, female; KD, ketogenic diet; LTG, lamotrigine (Lamictal); M,
Weigh
26.5 (240.5 (�34.0 (�33.6 (535.1 (220.7 (�
zam (F
Topamax); VPA, valproic acid (Depakene Epival)
zrbs
R
E
pl04�8
(H
Fitsi�
Fao�00�0nssL
4 K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
ure control (experiment II), a repeated measures linearegression analysis was conducted between mean dailyreath acetone and daily seizure frequency using SAStatistical software.
esults
xperiment I: breath acetone and plasma ketones
Compared with morning fasting values (0 h), all threelasma ketones were modestly but significantly increased 8 hater in the midafternoon. Plasma AcAc increased from 1.7 �.4 to 2.1 � 0.3 mmol/L, �-HBA increased from 3.5 � 0.7 to.4 � 0.6 mmol/L, and acetone increased from 3.5 � 1.0 to 4.5
1.0 mmol/L (all P � 0.05; Fig. 1A). Mean values over theh were �1.9 � 0.2 mmol/L (AcAc), 3.6 � 0.3 mmol/L
ig. 1. Plasma ketones (A) and breath acetone (B) during a single 8-h dayn children on a ketogenic diet. Points represent mean � standard error ofhe mean (n � 7 subjects/point), except for plasma acetone (n � 4ubjects/point). Plasma �-HBA, AcAc, and acetone were significantlyncreased from fasting at 8 h (P � 0.05). AcAc, acetoacetate; �-HBA,
�-hydroxybutyrate.
�-HBA), and 4.0 � 0.8 mmol/L (acetone), respectively.ence, plasma acetone was about 12% higher than plasma
ig. 2. Relation between breath acetone and plasma ketones. (A) Breathcetone and plasma acetoacetate (r2 � 0.89, P � 0.0001, n � 7). Equationf the line: Y � AXB � C, where A � 3.4884, B � 0.0803, and C �4.6065. (B) Breath acetone and plasma �-hydroxybutyrate (r2 � 0.94, P �
.0001, n � 7). Equation of the line: Y � AXB � C, where A � 5.1365, B �
.0864, and C � �6.5048. (C) Breath acetone and plasma acetone (r2 � 0.99, P0.0001, n � 4). Equation of the line: Y � AXB � C, where A � 2.2177, B �
.1900, and C � �5.6040. Each symbol represents a single subject (subjectumber 1, solid square; subject number 2, X; subject number 3, open triangle;ubject number 4, open square; subject number 5a, open circle; subject number 5b,olid circle; subject number 6, solid triangle; subject number 7, open diamond).arger symbols represent the mean of five time points in the 8-h test day.
-HBA and about twice as high as plasma AcAc. Averaged
oc
ankapo0�ATF
iBmgto2
E
dTtsabqr
D
hpviosea
taplp�on
FpAplblr2snrepresent the mean of five time points. AcAc, acetoacetate.
5K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
ver the day, the ratio of plasma �-HBA to AcAc did nothange significantly from 2.0 � 0.1 mmol/L.
Mean breath acetone remained stable throughout the day,veraging 5063 � 933 nmol/L (Fig. 1B). Although isopropa-ol is occasionally found on the breath in subjects on theetogenic diet, none was detected in these subjects. Breathcetone was positively and curvilinearly related to all threelasma ketones (P � 0.0001; Fig. 2). A linear relation was alsobserved between plasma acetone and plasma AcAc (r2 �.98, P � 0.011; Fig. 3A), between plasma acetone and plasma-HBA (r2 � 0.99, P � 0.0073; Fig. 3B), and between plasmacAc and plasma �-HBA (r2 � 0.92, P � 0.0006; Fig. 3C).he equations for these curves are presented in the legend toig. 3.
Breath acetone measured by gas chromatography was pos-tively and linearly related to measurements made using theAA (r2 � 0.99, P � 0.0002; Fig. 4A). All BAA measure-ents were found to be within 10% of the values obtained by
as chromatography (Fig. 4B). During the additional 3 d ofesting of subject number 4, breath acetone increased through-ut the day with day-long averages of 124 � 20, 182 � 49, and15 � 43 nmol/L, respectively (Fig. 5).
xperiment II: breath acetone and seizure frequency
Mean breath acetone and number of seizures before anduring treatment with the ketogenic diet are presented inable 3. Breath acetone values varied widely (means of 174
o 14 342 nmol/L), representing an 82-fold range over theix subjects. No significant relation between breath acetonend seizure frequency was detected. The possibility thatreath acetone might be related to change in seizure fre-uency was also evaluated, but no statistically significantelation was found.
iscussion
Experiment I of this study shows that, in children on aigh fat ketogenic diet, breath acetone accurately predictslasma acetone, AcAc, and �-HBA (Fig. 2). This was pre-iously shown in rats on the high-fat ketogenic diet [13] andn adults in short-term mild ketosis [14] but had not previ-usly been established in children maintaining higher keto-is induced by the high-fat ketogenic diet. To our knowl-dge, significant positive relations between AcAc, �-HBA,nd acetone in plasma have not been described before.
Other studies have shown a positive linear relation be-ween breath acetone up to 25 000 nmol/L and plasmacetone up to 15 mmol/L [18–21]. In the present study,lasma acetone up to 9 mmol/L was positively but curvi-inearly related to breath acetone up to 18 000 nmol/L. Aositive, linear relation between breath acetone and blood-HBA has been reported in five patients during 10 to 36 df fasting [22]. At breath acetone level higher than 7500
mol/L, the relation became curvilinear as plasma �-HBAig. 3. Relation among the three plasma ketones. (A) Plasma acetone andlasma AcAc (r2 � 0.98, P � 0.0109, n � 4). Equation of the line: Y �X � B, where A � 0.2704 and B � 0.6038. (B) Plasma acetone andlasma �-hydroxybutyrate (r2 � 0.99, P � 0.0073, n � 4). Equation of theine: Y � AX � B, where A � 0.4558 and B � 1.3712. (C) �-Hydroxy-utyrate and plasma AcAc (r2 � 0.92, P � 0.0006, n � 7). Equation of theine: Y � AX � B, where A � 1.5236 and B � 0.6734. Each symbolepresents a single subject (subject number 1, solid square; subject number, X; subject number 3, open triangle; subject number 4, open square;ubject number 5a, open circle; subject number 5b, solid circle; subjectumber 6, solid triangle; subject number 7, open diamond). Larger symbols
bp
wtaqtaaAttcbf
tb
vlpiwboabsdkbitbsstsb
cmtqoia3
FtpBarcntsnp
Fdtsw(
6 K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
egan to plateau above 3 mmol/L. In the present study,lasma �-HBA began to plateau at 4 to 5 mmol/L.
Compared with plasma �-HBA and plasma AcAc, theider range in plasma and breath acetone suggests a poten-
ially important physiologic role of plasma acetone. �-HBAnd AcAc dissociate in blood to become anions. Conse-uently, these ions affect blood pH, bicarbonate concentra-ions, and arterial gases [18]. Although AcAc is an acid,cetone is neutral and carbonic acid is a weak bufferingcid, so the formation of acetone and carbon dioxide fromcAc counteracts acidosis. Unlike �-HBA or AcAc, ace-
one is strongly lipophilic and hydrophilic. Thus, althoughhe distribution of �-HBA and AcAc is limited to the waterompartment [23], acetone is distributed throughout theody including fat and membranes [24]. Acetone is alsoreely diffusible into the brain but �-HBA and AcAc require
ig. 4. (A) Relation between breath acetone as measured by gas chroma-ography versus that measured by a hand-held breath acetone analyzerrototype (r2 � 0.99, P � 0.0002, n � 7). Equation of the line: Y � AX �, where A � 0.97 and B � 338.4. (B) Relation between mean breathcetone and the difference between the two methods. Mean breath acetoneefers to the average of the two methods. Dashed lines represent 95%onfidence intervals [17]. Each symbol represents a single subject (subjectumber 1, solid square; subject number 2, X; subject number 3, openriangle; subject number 4, open square; subject number 5a, open circle;ubject number 5b, solid circle; subject number 6, solid triangle; subjectumber 7, open diamond). Larger symbols represent the mean of all time
eoints for each subject.
he monocarboxylic acid transporter to cross the blood-brainarrier.
Recent reports have suggested that acetone has anticon-ulsant properties [25,26]. Some studies examining the re-ation between ketosis and seizure control have reported aositive association [1–3], but others have found no signif-cant association [4–6]. In the present study, no significantithin-subject or between-subject associations were foundetween breath acetone and seizure frequency. Although wenly had six subjects in experiment II, the multiple breathcetone sampling per day over 14 d led to more than 250reath acetone measurements in this experiment. These re-ults suggest that good seizure control was not statisticallyependent on achieving a particular breath acetone level oretone threshold in plasma. The lack of association betweenreath acetone and seizure control therefore seems to be duen large part to the wide cross-subject range in breath ace-one (120 to 14 000 nmol/L). Another possibility is thatreath acetone may be more clearly related to a change ineizure frequency rather than to actual seizure frequency pere. We found no such a relation but may not have been ableo adequately assess this possibility because four of the sixubjects were without seizure during the 2-wk period whenreath acetone was collected.
High breath acetone concentration in some subjectsould indicate slower acetone catabolism. Acetone can beetabolized to lactate, pyruvate, and acetyl coenzyme A via
he propanediol or methyl-glyoxal pathways and subse-uently recycled into amino acids, fatty acids, cholesterol,r glucose [27–29]. Carbon from acetone is incorporatednto cholesterol and other lipids, glycogen, and severalmino acids [19,24,30–32]. In one tracer study, less than0% of a dose of [14C] acetone given to fasted humans was
ig. 5. Breath acetone in subject number 4 who was studied on 3 differentays. Each point represents the mean � standard error of the mean ofriplicate samples analyzed by gas chromatography. Throughout the 2-wktudy, this subject was without seizure but had breath acetone readings thatere consistently below the detection limits of the breath acetone analyzer
500 nmol/L).
liminated as acetone in urine, breath, and sweat, whereas
6atdfkmtwbbt
S
tddeicew
A
R
[
[
[
[
[
[
[
[
[
[
TB
Sn
123456
sence.
7K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
0% was recycled into other products such as CO2, glucose,nd amino acids [24]. Up to11% of plasma glucose produc-ion was calculated to potentially be derived from acetoneuring prolonged fasting. Incorporation of radioactivityrom acetone into glucose has been observed in rats [31],etotic cows [33,34], fasted humans [24], and diabetic hu-ans [19]. Thus, subjects with unusually high breath ace-
one may actually be converting more into glucose, whichould impair seizure control. Conversely, our subject num-er 4 who was without seizure despite very low fastingreath acetone may reflect more efficient excretion of ace-one.
ummary
Breath acetone is a good, non-invasive marker of sys-emic ketosis but shows that ketosis varies widely in chil-ren given similar formulations of the high-fat ketogeniciet. Owing to this variability in breath acetone and mark-dly differing seizure frequencies in different forms of ep-lepsy, establishing a relation between ketosis and seizureontrol will require frequent measurements of both param-ters over an extended period, preferably from the point athich the diet is introduced.
cknowledgments
Mary Ann Ryan is thanked for technical assistance.
eferences
[1] Huttenlocher PR. Ketonemia and seizures: metabolic and anticonvul-sant effects of two ketogenic diets in childhood epilepsy. Pediat Res1976;10:536–40.
[2] Whedon B, Sirven JI, O’Dwyer J, Sperling MR. Therapeutic serumbeta-hydroxybutyrate levels in adult ketogenic diet patients. Epilepsia
able 3reath acetone and seizure frequency in children on a KD*
ubjectumber
Breath acetone(nmol/L)
Seizure frequency before KD
1389 � 161‡ 10–20 absence/d14342 � 351 30–60 atonic/mo, 4–5 absence/d
670 � 65 48 generalized/y174 � 37§ 1–15 absence/d
6428 � 86 150–200 multiple type/d�
† 2639 � 68 100 multiple type/d�
KD, ketogenic diet* Same subjects as presented in Table 2.† Studied for 12 instead of 14 d.‡ Mean � standard error of the mean.§ Value was determined by gas chromatography only.� Multiple type refers to atonic drop, myoclonic, generalized, and/or ab
1999;40(suppl 7):221.
[3] Gilbert DL, Pyzik PL, Freeman JM. The ketogenic diet: seizurecontrol correlates better with serum �-hydroxybutyrate than withurine ketones. J Child Neurol 2000;15:787–90.
[4] Dekaban AS. Plasma lipids in epileptic children treated with the highfat diet. Arch Neurol 1966;15:177–84.
[5] Schwartz RM, Boyes S, Aynsley-Green A. Metabolic effects of threeketogenic diets in the treatment of severe epilepsy. Dev Med ChildNeurol 1989;31:152–60.
[6] Fraser DD, Whiting S, Andrew RD, Macdonald EA, Musa-Veloso K,Cunnane SC. Elevated polyunsaturated fatty acids in blood serumobtained from children on the ketogenic diet. Neurology 2003;60:1026–9.
[7] Bridge EM, Iob LV. The mechanism of the ketogenic diet in epilepsy.Johns Hopkins Med J 1931;48:373–89.
[8] Freeman JM, Freeman JB, Kelly MT. The ketogenic diet: a treatmentfor epilepsy. 3rd ed. New York: Demos Publications; 2000.
[9] Freund G. The calorie deficiency hypothesis of ketogenesis tested inman. Metabolism 1965;14:985–90.
10] Livingston S. Dietary management of epilepsy. In: Comprehensivemanagement of epilepsy in infancy, childhood, and adolescence.Springfield, IL: CC Thomas; 1972, p. 378–405.
11] Schwartz RM, Eaton J, Bower BD, Aynsley-Green A. Ketogenicdiets in the treatment of epilepsy: short-term clinical effects. Dev MedChild Neurol 1989;31:145–51.
12] Laffel L. Ketone bodies: a review of physiology, pathophysiologyand application of monitoring to diabetes. Diabetes Metab Res Rev1999;15:412–26.
13] Likhodii SS, Musa K, Cunnane SC. Breath acetone as a measure ofsystemic ketosis assessed in a rat model of the ketogenic diet. ClinChem 2000;48:115–20.
14] Musa-Veloso K, Likhodii SS, Cunnane SC. Breath acetone is areliable indicator of ketosis in adult volunteers consuming ketogenicmeals. Am J Clin Nutr 2002;76:65–70.
15] Musa-Veloso K, Rarama E, Comeau F, Curtis R, Cunnane SC. Epi-lepsy and the ketogenic diet: Assessment of ketosis in children usingbreath acetone. Pediatr Res 2002;52:443–8.
16] Harano Y, Ohtsuki M, Ida M. Direct automated assay method forserum or urine levels of ketone bodies. Clin Chim Acta 1985;151:177–83.
17] Bland JM, Artman DO. Statistical methods for assessing agreementbetween two methods of clinical measurement. Lancet 1986;i:307–10.
18] Sulway MJ, Malins JM. Acetone in diabetic ketoacidosis. Lancet1970;2:736–40.
19] Owen OE, Trapp VE, Skutches CL, Mozzoli MA, Hoeldtke RD,Boden G, Reichard GA Jr. Acetone metabolism during diabetic ke-
Seizure frequency during KD Decrease in seizurefrequency (%)
0 absence/d 1001–2 atonic/mo, 2–3 absence/d 95 (atonic) 45 (absence)2 generalized/y 960 absence/d 1001–2 absence/d 992–3 absence/y, 2–3 myoclonic jerks/y �99
toacidosis. Diabetes 1982;31:242–8.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
8 K. Musa-Veloso et al. / Nutrition 22 (2006) 1–8
20] Levey S, Balchum OJ, Medrano V, Jung R. Studies of metabolicproducts in expired air. II. Acetone. J Lab Clin Med 1964;63:574–84.
21] Brechner VL, Bethune RWM. Determination of acetone concentra-tion in arterial blood by vapor phase chromatography of alveolar gas.Diabetes 1965;14:663–5.
22] Tassopoulos CN, Barnett D, Fraser TR. Breath acetone and bloodsugar measurements in diabetes. Lancet 1969;i:1282–6.
23] Owen OE, Caprio S, Reichard GA Jr. Mozzoli MA, Boden G, OwenRS. Ketosis of starvation: a revisit and new perspectives. Clin Endo-crinol Metab 1983;12:359–79.
24] Reichard GA Jr. Haff AC, Skutches CL, Paul P, Holroyde CP, OwenOE. Plasma acetone metabolism in the fasting human. J Clin Invest1979;63:619–26.
25] Likhodii SS, Burnham WM. Ketogenic diet: does acetone stop sei-zures? Med Sci Monit 2002;8:HY19–24.
26] Rho JM, Anderson GD, Donevan SD, White HS. Acetoacetate, acetone,and dibenzylamine (a contaminant in L-(�)-�-hydroxybutyrate) exhibitdirect anticonvulsant actions in vivo. Epilepsia 2002;43:358–61.
27] Argiles JM. Has acetone a role in the conversion of fat to carbohy-
drate in mammals? Trends Biochem Sci 1986;11:61–3.28] Kalapos MP. Possible physiological roles of acetone metabolism inhumans. Med Hypotheses 1999;53:236–42.
29] Kalapos MP. On the mammalian acetone metabolism: from chem-istry to chemical implications. Biochem Biophys Acta 2003;1621:122–39.
30] Price TD, Rittenberg D. The metabolism of acetone. I. Grossaspects of catabolism and excretion. J Biol Chem 1950;185:449 –59.
31] Sakami W, Rudney H. The metabolism of acetone and acetoace-tate in the mammalian organism. Brookhaven Symp Biol 1952;5:176 –91.
32] Mourkides GA, Hobbs DC, Koeppe RE. The metabolism of acetone-2-C14 by intact rats. J Biol Chem 1959;234:27–30.
33] Luick JR, Black AL, Simesen MG, Kametaka M, Kronfeld DS.Acetone metabolism in normal and ketotic cows. J Dairy Sci 1967;50:544–9.
34] Black AL, Luick JR, Lee SL, Knox K. Glucogenic pathway foracetone metabolism in the lactating cow. Am J Physiol 1972;222:
1575–80.
