Young women partition fatty acids towards ketone body production rather
than VLDL-TAG synthesis, compared with young men
Kyriakoula Marinou1,2, Martin Adiels3, Leanne Hodson1, Keith N. Frayn1, Fredrik Karpe1
and Barbara A. Fielding1*
1Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, Churchill Hospital, Oxford OX3 7LJ, UK
2Department of Experimental Physiology, Athens University School of Medicine, Athens, Greece
3Wallenberg Laboratory, Sahlgrenska Center for Cardiovascular and Metabolic Research, Go ¨teborg University,
Go ¨teborg, Sweden
(Received 24 June 2010 – Revised 30 September 2010 – Accepted 5 October 2010 – First published online 21 January 2011)
Before the menopause, women are relatively protected against CVD compared with men. The reasons for this sex difference are not
completely understood, but hepatic fatty acid metabolism may play a role. The present study aimed to investigate the utilisation of
plasma NEFA by the liver and to determine whether they are partitioned differently into ketone bodies and VLDL-TAG in healthy,
lean young men and women. Volunteers were studied during a prolonged overnight fast (12–19h) using an intravenous infusion of
[U-13C]palmitate. After 12h fasting, the women had a more advantageous metabolic profile with lower plasma glucose (P,0·05) and
TAG (P,0·05) but higher plasma NEFA (P,0·05) concentrations. Plasma 3-hydroxybutyrate (3-OHB) concentrations rose more in
women than in men, and the transfer of13C from [U-13C]palmitate to plasma [13C]3-OHB reached a plateau 6–7h after the start of the infu-
sion in women but was still increasing at 6h in men. This implies a slower 3-OHB production rate and/or dilution by other precursor pools
in men. In women, the high isotopic enrichment of plasma 3-OHB suggested that systemic plasma fatty acids were the major source of
3-OHB production. However, in men, this was not observed during the course of the study (P,0·01). There were no sex differences
for the incorporation of
production rather than VLDL-TAG may contribute to their more advantageous metabolic profile compared with young men.
13C into VLDL1- or VLDL2-TAG. The ability of young women to partition fatty acids towards ketone body
Key words: 3-Hydroxybutyrate: Stable isotopes: Esterification
Premenopausal women are relatively protected against
CVD compared with men(1), but the reasons for this sex
difference are not completely understood. An under-
standing of the underlying mechanisms for this difference
may help us to prevent worsening of cardiovascular risk
factors in men and women in later life. There is evidence
for some sex differences in the plasma metabolic profile
in young adults. Short-term fasting reveals sex differences
in people who are young and lean; after fasting, women
have lower plasma glucose and higher plasma NEFA
concentrations than men, as summarised in Soeters
et al.(2). The ketone body 3-hydroxybutyrate (3-OHB) is
an important energy source for the brain and other tissues
during prolonged fasting. Under starving conditions, it is
considered to act as a central signal and energy-providing
substrate involved in the regulation of food intake(3). Also,
it seems that 3-OHB has an anti-lipolytic effect and is
a ligand for the nicotinic acid receptor. In this way,
its effect is similar to nicotinic acid(4). However, to the best
of our knowledge, sex comparisons of plasma 3-OHB
concentrations have rarely been reported in healthy, lean
young men and women. The difference in blood 3-OHB
concentrations has been reported to be greater in response
to fasting with significantly higher concentrations in
women than in men after 30h fasting, but plasma insulin
concentrations were similar(5). After an overnight fast,
plasma TAG concentrations in women were lower than
those in men, but this difference did not reach statistical
significance and did not change during a further 10h of
fasting(6). However, plasma VLDL-TAG concentrations were
lower in women(6)in response to the extended fasting.
The results from that study suggested that the liver in
women secretes fewer but TAG-richer VLDL particles
than the liver in men, and that clearance was faster in
women. There are two major classes of VLDL secreted by
the liver: VLDL1is larger and more TAG-rich than VLDL2.
*Corresponding author: Dr B. A. Fielding, email email@example.com
Abbreviations: 3-OHB, 3-hydroxybutyrate; MPE, mole percent excess; Ra, rate of appearance; TTR, tracer:tracee ratio.
British Journal of Nutrition (2011), 105, 857–865
q The Authors 2011
British Journal of Nutrition
The latter can either be secreted directly from the liver or be
formed by the peripheral hydrolysis of VLDL1. These two
classes of lipoproteins have different properties; in subjects
with type 2 diabetes, the secretion of VLDL1is associated
with liver fat content, hypertriacylglycerolaemia and
atherogenic risk(7). VLDL-TAG production represents the
export of fatty acids partitioned towards esterification and
secretion rather than storage or oxidation in the liver.
Therefore, the balance between the pathways of esterifica-
tion and oxidation in men and women may be important
for the development of CVD risk factors.
The aim of the present study was to investigate sex
differences in plasma metabolites and hepatic fatty acid
metabolic partitioning in the liver in response to an
extended overnight fast. This enabled us to stress meta-
bolic pathways in order to investigate sex differences in
young people before the onset of traditional cardiovascular
risk factors. Since systemic NEFA are the major substrate
for VLDL-TAG and 3-OHB production, we aimed to
follow the transfer of a stable isotope label from plasma
NEFA into the products of hepatic metabolism.
Materials and methods
Healthy, lean young men and women (n 12) were studied,
and matched for age and BMI with no significant difference
in waist circumference, but the women had a greater
percentage of body fat (Table 1). Subjects, recruited from
the wider Oxford community via advertisement, were
free from any disease, weight stable and not taking any
lipid-lowering medication or medication that would alter
lipid metabolism. The studies were performed during
the first week (follicular phase) of the menstrual cycle
and conducted according to the guidelines laid down in
the Declaration of Helsinki. All procedures involving
human subjects were approved by the Oxfordshire Clinical
Research Ethics Committee. Written informed consent was
obtained from all subjects.
Before the study day, subjects were asked to avoid food-
stuffs naturally enriched in13C for 48h and refrain from
strenuous exercise and alcohol for 24h. On the evening
before the study, subjects were required to eat a low-fat
meal. On the day of the study, subjects arrived at the
clinical research unit after an overnight fast. Fat mass
was measured by bioelectrical impedance analysis, and
fat-free mass was calculated as the difference between
total body weight and fat mass. A cannula was inserted
into an antecubital vein, and a baseline blood sample was
taken for background isotopic enrichment. At time 0, a
continuous intravenous infusion (0·04mmol/kg per min)
Cambridge Isotope Laboratories, Inc., Andover, MA, USA)
complexed to human albumin was started, to label the
plasma NEFA pool. This was continued until the end of
the experiment (at 7h).
Whole blood was collected into heparinised syringes
(Sarstedt, Leicester, UK), plasma was rapidly separated
by centrifugation at 48C, and plasma NEFA, VLDL-TAG
and 3-OHB concentrations were determined enzymatically,
as described previously(8,9). Plasma insulin concentrations
were measured by RIA(9).
In ten subjects, VLDL1 (Svedberg flotation rate (Sf)
60–400) and VLDL2(Sf20–60) were isolated by sequential
flotation using density gradient ultracentrifugation in an
SW40Ti swinging bucket rotor (Beckman Instruments,
Palo Alto, CA, USA) at 40000rpm at 158C for 4h for
Sf 60–400 lipoproteins and for a further 16h to float
Fatty acid analysis and isotopic enrichment
Fatty acid methyl esters were prepared from NEFA, VLDL1-
and VLDL2-TAG fractions and isotopic enrichment was
measured by GC and GC–MS, respectively, as described
previously(9). Specific fatty acid concentrations were deter-
mined by multiplying the proportion of the specific
fatty acid by the corresponding plasma concentration as
determined enzymatically for plasma NEFA, VLDL1- and
Isotopic enrichment from [U-13C]fatty acids appearing
in 3-OHB in deproteinised plasma was measured using
a modified method of Beylot et al.(10). Solutions of
[2,4-13C2]3-OHB (Cambridge Isotope Laboratories, Inc.)
and unenriched 3-OHB were prepared in 4% perchloric
acid and diluted in deionised water to make an enrichment
standard curve (100mmol/l). Plasma samples (0·5ml) were
Table 1. Volunteer characteristics and baseline overnight fasting (12h)
plasma metabolite concentrations
(Mean values with their standard errors, n 6)
Fat mass (% body wt)
Body wt (kg)
Mean values were significantly different from those of men: *P,0·05, ***P,0·001.
K. Marinou et al.858
British Journal of Nutrition
deproteinised with 1ml perchloric acid (70g/l). Duplicate
1ml aliquots of the supernatant were neutralised with
500ml neutralising reagent (0·5 M-KHCO3and K2CO3) on
ice. After centrifugation, 500ml aliquots of the deprotei-
nised plasma were pipetted into 10ml glass tubes and
acidified to pH 1 with 1 M-HCl. After prior addition of
50ml neutralising reagent, 500ml 3-OHB standards were
acidified in the same manner. The 3-OHB from samples
and standards was extracted into 4·5ml ethyl acetate–
diethyl ether (1:1, v/v) in 10ml glass tubes by mixing
by hand for 2min and then by rotary mixing for 1h.
The aqueous and organic phases were separated by
centrifugation, and the upper organic phase was evapor-
ated to dryness under N2 at room temperature. t-Butyl
dimethylsilyl derivatives of 3-OHB were prepared by the
addition of 20ml pyridine and 20ml N-(tert-butyldi-
methylsilyl)-N-methyltrifluoroacetamide þ 1% t-butyl dim-
ethylchlorosilane. After incubation for at least 45min at
room temperature, the contents of the tubes were trans-
ferred to GC vials and analysed by GC–MS. The GC was
equipped with a 30m capillary column with a 5% diphenyl–
diameter 0·32mm, film thickness 0·25mm; Thames Restek,
Saunderton, UK). The 5890 GC was coupled to a 5973N
MSD (Agilent Technologies, Stockport, UK). Ions with
mass-to-charge ratios (m/z) of 275 (M þ 0), 277 (M þ 2)
were determined by selected ion monitoring. We assume
that the latter corresponds with the formation of 3-OHB
from one labelled [13C2]acetyl-CoA molecule, derived
from [U-13C]palmitic acid via b-oxidation. Initially, we
selected ions of m/z 279 (M þ 4) but the GC–MS was not
sufficiently sensitive to consistently detect a peak.
The tracer:tracee ratio (TTR) of a baseline measurement
(before the administration of the stable isotope tracer)
was subtracted from each sample’s TTR to account for
natural abundance. The TTR for [U-13C]palmitate were
multiplied by the corresponding palmitate concentration
in plasma NEFA and VLDL1- and VLDL2-TAG to give
tracer concentrations. Likewise, the TTR for [13C]3-OHB
(M þ 2):(M þ 0) were multiplied by the corresponding
plasma 3-OHB concentrations to give tracer concen-
trations. Mole percent excess (MPE) was calculated from
the TTR using the formula MPE ¼ 1/(1 þ 1/TTR).
The proportion of systemic plasma NEFA contributing
to VLDL-TAG at 7h was calculated as described pre-
viously(11). The remainder are assumed to be fatty acids
from splanchnic sources, including those from visceral fat
lipolysis, liver fat and de novo lipogenesis.
Whole-body rate of appearance (Ra) of NEFA was calcu-
lated using the isotopic enrichment in the plasma after an
intravenous infusion of the [U-13C]palmitate(12)during
the last 5h of the infusion. The rate of disappearance of
NEFA was calculated as RaNEFA 2 dQ/dt, where dQ is
the change in the amount of tracee with time(12).
The classic mathematical modelling of 3-OHB pro-
duction could not be performed because the level of
enrichment in 3-OHB did not allow for calculation of
the precursor pool enrichment. However, we wished
to examine the extent to which the enrichment of [13C] in
plasma 3-OHB reflected the synthesis from systemic
NEFA. Therefore, we estimated the maximal enrichment
(MPE) in 3-OHB for (M þ 2) at 7h, assuming that the
precursor for 3-OHB production was only from the
plasma NEFA pool. Since we infused a palmitic acid
tracer, we assumed that the enrichment in the mitochon-
drial acetyl-CoA pool would be diluted by (non-selective)
contributions from other specific fatty acids. Therefore,
we accounted for the number of carbon atoms in each
fatty acid species and their relative proportions in the
NEFA pool for each participant. The plasma enrichment
of [U-13C]palmitate multiplied by the dilution factor gave
an estimation of the [13C]acetyl-CoA enrichment in the
mitochondria. The estimated maximal MPE was calculated
from the formula for the probability of picking exactly
one labelled and one unlabelled element, 2P(12P),
where P is the probability of an element to be labelled.
The Mþ 4 isotopomer for 3-OHB in women at plateau
enrichment was estimated from the calculated precursor
enrichment (MPE) and binomial expansion(13).
In order to determine an approximation of the relative
partitioning of systemic plasma NEFA into 3-OHB and
[13C]VLDL-TAG in the plasma, where [13C] is expressed
Data were analysed using SPSS for Windows version 16
(SPSS, Chertsey, UK). Statistical significance was set
at P,0·05. All data are presented as means with their
standard errors. Repeated-measures ANOVA, with time
and group as factors, was used to investigate the change
between men and women over time while fasting.
Comparisons between groups after the extended overnight
fast (7h) were made using a Mann–Whitney test.
Conventional terminology for stable isotope tracer tech-
niques is used in the present paper(12). The passage of a
molecule through the GC–MS results in the production
of a major ion of interest with mass M. An isotopomer is
a molecule that has the same chemical composition but a
different mass because it has an isotopic tracer incorpor-
ated somewhere in the molecule. Thus, the M isotopomer
for 3-OHB refers to an ion that has no13C tracer incorpor-
ated, and M þ 2 represents an ion that has a mass that
is 2 atomic mass units greater than the M isotopomer;
Hepatic fatty acid partitioning859
British Journal of Nutrition
this is assumed to be due to the substitution of two12C
atoms by two13C atoms.
Differences in plasma variables between men and women
after an overnight fast are given in Table 1. After 12h
fasting, the men had a less advantageous metabolic profile
in terms of higher plasma glucose (P,0·05) and TAG
(P,0·05) concentrations but lower plasma NEFA (P,0·05)
concentrations. In order to verify that the twelve subjects
whom we chose were representative of the wider commu-
nity in this respect, we retrieved data from 266 men and 405
women (aged 30–50 years) from the Oxford Biobank(14).
All individuals were selected to have a BMI of less than
25kg/m2. In exact accordance with the six men and six
women in the present study, the larger group of women
had higher plasma NEFA concentrations (553 (SE 13) v.
475 (SE 16)mmol/l; P,0·05); the same plasma 3-OHB con-
centrations, lower plasma TAG concentrations (870 (SE 20)
v. 1090 (SE 40)mmol/l; P,0·05); lower plasma glucose
concentrations (4·8 (SE 0·0) v. 5·2 (SE 0·0)mmol/l; P,0·05)
and similar plasma insulin concentrations. After extension
of the postabsorptive period during the 7h of the present
occurred (Fig. 1). Significant differences between men
and women for glucose, TAG and NEFA were maintained,
and the marginally higher values for 3-OHB found in
women after 12h fasting (P¼0·078) became significant
(P¼0·035). Plasma insulin concentrations decreased signifi-
cantly with time (P¼0·02) but remained similar between
men and women after prolonged fasting.
Concentrations of VLDL2-TAG but not VLDL1-TAG were
significantly higher in men than in women (Table 1). We
observed that VLDL1-TAG concentrations were higher
than VLDL2-TAG in both sexes combined (P¼0·028).
During prolonged fasting, the concentrations of VLDL1-
TAG decreased significantly (Fig. 1), but no further sex
differences were revealed. There was a small decrease in
VLDL2-TAG concentrations during the study, but this was
not significant (females P¼0·275, males P¼0·663).
The TTR of [U-13C]palmitate from the intravenous infu-
sion reached a plateau in the plasma NEFA pool by 1h
in men and women (Fig. 2). RaNEFA (mmol/min) was simi-
lar in men and women but lower in women when
expressed as per kg fat mass (Table 2). When expressed
according to fat-free mass, there was no difference
between men and women (Table 2).
The fatty acid tracer was rapidly incorporated into
plasma 3-OHB and VLDL-TAG. Whereas enrichment
reached a plateau in plasma 3-OHB at 6–7h after the
start of the infusion in women, it was still increasing at
6h in men (Fig. 2). Plasma concentrations of [13C]3-OHB
were correspondingly higher in women than in men
during the course of the experiment (P¼0·007 for the
effect of sex, P¼0·004 for the time £ sex interaction).
In women, the observed isotopic enrichment of the fatty
acid tracer in 3-OHB was close to that predicted, if systemic
plasma fatty acids were the sole source of 3-OHB pro-
duction (Table 2). In fact, the observed MPE reached
78% of the estimated maximal MPE, and thus only 22%
of the 3-OHB could not be accounted for by the systemic
NEFA sources. These values are in close agreement with
the levels reached for VLDL1- and VLDL2-TAG (Table 3)
in these lean women. However, in men, the observed
either an alternative source of fatty acids or a slower rate
of production. The incorporation of the [13C] tracer
into VLDL1- and VLDL2-TAG
different between men and women, in terms of either the
TTR (Fig. 2) or the [13C]VLDL-TAG concentration (data
The ratio of [13C]3-OHB:[13C]VLDL-TAG in the plasma
was significantly higher in women than in men (2·73
(SE 1·7) v. 0·66 (SE 0·30); P¼0·027 for VLDL1 and 3·24
(SE 1·6) v. 0·97 (SE 0·43); P¼0·014 for VLDL2).
As we have reported previously in the postprandial
period for healthy people for total VLDL-TAG(11), fatty
acids from systemic plasma NEFA contributed a greater
proportion of fatty acids to VLDL1-TAG and VLDL2-TAG
than those from splanchnic sources, in men and women
(Table 3). However, the proportion of systemic fatty acids
was lower in men than in women for VLDL2-TAG, with a
corresponding greater proportion from splanchnic sources.
In the present study, we found significant differences in
hepatic metabolism in healthy, lean young men and
women. Sex comparisons of plasma 3-OHB in very young
men and women have rarely been reported, but we are in
agreement with previous findings of higher concentra-
tions in young women after a 30h fast(5), and we found
that a shorter period of fasting (12–19h) revealed higher
concentrations in women. We found no difference in
plasma 3-OHB concentrations between men and women
without the provocation of a prolonged fast, in either
a young age group (n 12) or middle-aged individuals
(n 671). Since ketone bodies are only produced in the
liver, systemic concentrations, usually of 3-OHB, can be
taken to reflect hepatic ketone body production. Lower
3-OHB concentrations have been reported in subjects
with hyperlipidaemia(15,16), obesity(17)and insulin resist-
ance(11). Therefore, lower 3-OHB concentrations seem to
be associated with a less advantageous metabolic profile.
These differences could be due to differences in pro-
duction or clearance. The present findings suggest that in
young lean women, a higher plasma concentration of
3-OHB, compared with men, is due to a greater pro-
duction, although we did not measure production directly.
There were sex differences in the incorporation of
carbon units from [U-13C]palmitate, representing systemic
K. Marinou et al.860
British Journal of Nutrition
fatty acids, into 3-OHB. The higher 3-OHB TTR and attain-
ment of plateau enrichment in women suggests that new
3-OHB substantially contributes to total 3-OHB. Thus, it
could be hypothesised that women rapidly switch on
3-OHB production from systemic plasma NEFA, allowing
plasma concentrations of 3-OHB to increase rapidly and
the TTR to flatten out. Men have a slower 3-OHB switch,
and at 7h, the contribution of plasma NEFA is still increas-
ing as a source of 3-OHB. However, it should be noted
that the present findings are not necessarily representative
of n-3 PUFA, which may partition away from b-oxidation
The reason for a greater ability of women to turn on
3-OHB production is not clear. The production of ketone
bodies is dependent, to a large extent, on the supply of
plasma NEFA(19,20). Plasma NEFA concentrations that
increased experimentally lead to increased plasma concen-
trations of 3-OHB(19,20). Beylot et al.(21)found that the
0 60120180 240300360 420
060 120180 240300 360420
060 120180 240 300 360420
0 60 120180 240300360420
0 60120180 240300 360420
VLDL1- TAG (µmol/l)
VLDL2- TAG (µmol/l)
Fig. 1. Plasma concentrations of metabolites in response to continued overnight fasting and analysed by repeated-measures ANOVA. X, Men (n 6); W, women
(n 6). Values are means, with standard errors represented by vertical bars. (a) Glucose, there was a significant effect of time (P,0·05) and sex (P,0·05).
(b) TAG, there was a significant effect of time (P,0·001) and a tendency for an effect of sex (P¼0·087). (c) NEFA, there were significant effects of time (P,0·01)
and sex (P,0·01). (d) 3-Hydroxybutyrate (3-OHB), there were significant effects of time (P,0·01) and sex (P,0·05). (e) Plasma VLDL1concentrations (men (n 5),
women (n 5)), in response to continued overnight fasting. There was a significant effect of time (P,0·05). (f) Plasma VLDL2concentrations, there was a tendency
for an effect of sex (P¼0·066).
Hepatic fatty acid partitioning861
British Journal of Nutrition
increased in a linear fashion as the precursor concentration
increased. This is thought to be due to reduced malonyl-
CoA and reduced inhibition of carnitine palmitoyl transfer-
ase I, and could possibly account for the sex differences
that we observed. A greater percentage conversion of
plasma NEFA to 3-OHB may have occurred in the
women of the present study, whose mean plasma NEFA
concentration was almost twice that of men. Glucagon is
also an important stimulator of ketone body production,
conversion ofplasmaNEFA to3-OHB
but it is not clear if there is a sex difference. Glucagon
levels have been reported as similar in men and
women(5), but in an older study by Merimee & Fineberg(22),
plasma glucagon concentrations were significantly higher
in premenopausal women than in men after a prolonged
fast (72h). Insulin reduces 3-OHB production indirectly,
via a direct effect on adipose tissue lipolysis, affecting
the supply of NEFA. A direct effect on ketone body
production has been described in some(23)but not in all
VLDL1- TAG TTR
VLDL2- TAG TTR
Fig. 2. Plasma tracer:tracee ratio (TTR) resulting from an intravenous infusion of [U-13C]palmitate, starting at time 0. X, Men (n 6); W, women (n 6). (a) NEFA
(n 6), there were significant effects of time (P,0·05), time £ sex interaction (P,0·05) and a tendency for an effect of sex (P¼0·086). (b) 3-Hydroxybutyrate
(3-OHB, n 5), there was a significant effect of time (P,0·001) and sex (P,0·05). (c) VLDL1, there was a significant effect of time (P,0·001). (d) VLDL2, there
was a significant effect of time (P,0·001). Values are means, with standard errors represented by vertical bars.
Table 2. NEFA kinetics and 3-hydroxybutyrate (3-OHB)13C enrichment arising from the systemic NEFA pool at 420min
(Mean values with their standard errors, n 6)
RaNEFA (mmol/min per kg fat mass)*
RdNEFA (mmol/min per kg fat-free mass)
3-OHB estimated maximal MPE**
3-OHB observed MPE
3-OHB observed/maximal MPE (%)
Ra, rate of appearance; Rd, rate of disappearance; MPE, mole percent excess.
There was a significant time-by-sex effect: **P,0·005.
Mean values were significantly different from those of women: †P,0·05, ‡P,0·01. Effect of sex over a period of extended fasting (14–20h).
§ n 5.
K. Marinou et al. 862
British Journal of Nutrition
Although plasma NEFA concentrations were higher in
women, they had a similar delivery of fatty acids into
the systemic circulation than in men as determined by
RaNEFA. However, even though the women were lean,
they had a significantly higher fat mass than the men,
and when Ra NEFA was expressed in relation to fat
mass, the RaNEFA was significantly lower. This is in agree-
ment with the recent publication of Mittendorfer et al.(25)
who reported a down-regulation of RaNEFA per unit of fat
mass in obesity. So, even within lean individuals, we were
able to observe an effect of fat mass on RaNEFA. In contrast
to the present findings (no difference between men
and women), the study of Mittendorfer et al. found a
higher RaNEFA (mmol/min) in women, and reported similar
plasma NEFA concentrations. A possible explanation for the
be because we enrolled leaner, younger volunteers.
After 12h fasting, the men in the present study had
a more disadvantageous metabolic profile in terms of
higher plasma glucose and TAG concentrations but lower
plasma NEFA concentrations. Since the latter are generally
thought of as an adverse CVD risk factor, especially in
terms of obesity and diabetes(26), this was unexpected.
However, we have shown that a metabolic disadvantage
of a low plasma NEFA concentration is the reduced ability
of the liver to switch to ketone body production during
short-term fasting, and in the long term, this could possibly
lead to the development of fatty liver. Plasma NEFA is also
a substrate for VLDL-TAG synthesis. In the present study,
more than three quarters of VLDL-TAG were delivered
from systemic plasma NEFA, in men and women. However,
particularly in VLDL2, the contribution of systemic plasma
NEFA to VLDL-TAG was lower in men than in women. This
may have been due to a lower flux from systemic fatty
acids and/ora greater contribution from splanchnic sources.
The latterwould include denovo lipogenesis, fatty acidfrom
the lipolysis of visceral fat or from cytosolic storage pools.
We calculated an ‘estimated’ maximum13C enrichment
in 3-OHB that would be achieved if systemic fatty acids
were the sole precursor pool for 3-OHB production.
This was similar for men and women. For women,
the achieved value was 78% of the value expected,
implying that systemic fatty acids were the major precursor
pool for 3-OHB production. This is in line with the hypoth-
esis that cytosolic TAG fatty acids may not provide
substrate for ketone body production, suggesting compart-
mentalisation of the precursor pool of ketone bodies(27).
It also suggests that fatty acids from visceral fat did not
substantially contribute to ketone body production in these
lean young women. However, in men, the enrichment of
13C in 3-OHB was less than half the value expected. Consist-
ent with this would be dilution of the precursor pool by
non-systemic fatty acids, such as those from visceral fat.
Alternatively, a lower 3-OHB production rate in men
would mean that the maximum isotopic enrichment
had not been reached during the course of the experiment.
Parallel to these findings for 3-OHB isotopic enrichment,
the VLDL isotopic enrichment in women was close to that
of the plasma NEFA pool; thus over 85% of VLDL were
cursor fatty acids for both metabolic pathways over the dur-
ation of the present study were approximately 80% from
plasma NEFA, and implies that plasma NEFA were an
immediate and major source of fatty acids for hepatic fatty
acid metabolism, contributing equally to 3-OHB and VLDL-
TAG synthesis during a prolonged overnight fast. However,
since there were no sex differences in VLDL isotopic enrich-
ment, the implication is that women rapidly partition a
greater proportion of systemic fatty acids into ketone
bodies than men. In men, it cannot be ruled out that the
same sources are used for 3-OHB and VLDL-TAG, but since
the 3-OHB enrichment is still rising at the end of the experi-
ment, and the total 3-OHB pool is much lower than for
women, it is clear either that the time course is considerably
slower or that other sources of fatty acids are involved.
Ketone bodies may be infused intravenously to give
kinetic information on ketone body production(21,23,28),
but the use of a fatty acid stable isotope tracer has
not been utilised previously to investigate ketone body
metabolism in humans. The disadvantage of the latter
approach is that we were not able to do classic kinetic
modelling, because the lack of detectable M þ 4 tracer
Table 3. Contribution of splanchnic and systemic fatty acids to VLDL-TAG at 7h in men and women
(Mean values with their standard errors)
Splanchnic Systemic SplanchnicSystemic
Absolute contribution of different fatty acid sources to VLDL-TAG (mmol/l)
Relative contribution of different fatty acid sources to VLDL-TAG (%)
* Mean values were significantly different from those of men (P,0·05).
† Mean values were significantly different from those of splanchnic sources (P,0·05).
Hepatic fatty acid partitioning863
British Journal of Nutrition
labelling in 3-OHB did not allow for calculation of the
precursor pool enrichment. Quantitatively, the M þ 4 iso-
topomer would not contribute significantly to the total
appearance of the palmitate tracer in 3-OHB. We calculated
that in women, the enrichment (MPE) of the 3-OHB M þ 4
isotopomer was less than 0·5% of the M þ 2 isotopomer.
However, the results were informative. In particular, we
were able to show that in women, systemic plasma NEFA
was the main source of fatty acids for 3-OHB production,
andthat hepatic production
different kinetics in men and women. We calculated the
[13C]3-OHB:[13C]VLDL-TAG ratio in the plasma, and found
that mean values were greater than 2·5 in women.
This illustrates the quantitative importance of ketone
body production for NEFA turnover in women, in the
early fasting period. Beylot et al.(21)calculated that for
a plasma 3-OHB concentration of 464mmol/l, 13% of
plasma NEFA are converted to 3-OHB in healthy young
men. The present results would suggest an even higher
rate of conversion in women.
The hepatic partitioning of fatty acids into ketone bodies
or towards esterification is difficult to study in humans
but may be an important metabolic regulatory point that
warrants further study. We have shown that in young
women, plasma fatty acids tend to be readily converted
to ketone bodies after an extended overnight fast, possibly
partly because of a higher systemic pool than in men.
Moreover, there is evidence of preferential partitioning
to ketone bodies rather than VLDL-TAG, at least in early
starvation. From an evolutionary point of view, women
may therefore have been better adapted to cope with
periods without food. In modern times, the greater ability
of women to oxidise plasma NEFA into ketone bodies
may partly explain the fact that premenopausal women
have a better metabolic profile than men and may help
protect women against the accumulation of liver fat.
Fatty acid partitioning could potentially be manipulated
by lifestyle changes or pharmacological intervention.
This work was funded in part by the project ‘Hepatic and
adipose tissue and functions in the metabolic syndrome’
(http://www.hepadip.org/), which is supported by the
European Commission as an Integrated Project under the
Sixth Framework Programme (Contract LSHM-CT-2005-
018734). Funding is also gratefully acknowledged from
the European Atherosclerosis Society, Novo Nordisk and
the Sahlgrenska Center for Cardiovascular and Metabolic
Research funded by the Foundation for Strategic Research.
We thank Jane Cheeseman, Louise Dennis and Siobha ´n
McQuaid for expert help during the studies. We confirm no
conflict of interest for the work reported here. K. M. carried
out the studies, and performed laboratory and data
analyses. M. A. analysed the data and helped with the
analysis, data analysis and helped with the manuscript
preparation. B. A. F. developed stable isotope methods,
supervised laboratory work and took the lead on the
manuscript preparation. F. K. supervised the studies and
helped with the manuscript preparation. K. N. F., L. H.,
B. A. F. and F. K. designed the study.
1.Booth GL, Kapral MK, Fung K, et al. (2006) Relation between
age and cardiovascular disease in men and women with
diabetes compared with non-diabetic people: a population-
based retrospective cohort study. Lancet 368, 29–36.
Soeters MR, Sauerwein HP, Groener JE, et al. (2007) Gender-
related differences in the metabolic response to fasting. J Clin
Endocrinol Metab 92, 3646–3652.
Laeger T, Metges CC & Kuhla B (2010) Role of beta-hydroxy-
butyric acid in the central regulation of energy balance.
Appetite 54, 450–455.
Taggart AK, Kero J, Gan X, et al. (2005) (D)-beta-Hydroxybu-
tyrate inhibits adipocyte lipolysis via the nicotinic acid
receptor PUMA-G. J Biol Chem 280, 26649–26652.
Haymond MW, Karl IE, Clarke WL, et al. (1982) Differences
in circulating gluconeogenic substrates during short-term
fasting in men, women, and children. Metabolism 31, 33–42.
Magkos F, Patterson BW, Mohammed BS, et al. (2007)
Women produce fewer but triglyceride-richer very low-
density lipoproteins than men. J Clin Endocrinol Metab 92,
Adiels M, Taskinen MR, Packard C, et al. (2006) Overproduc-
tion of large VLDL particles is driven by increased liver fat
content in man. Diabetologia 49, 755–765.
Humphreys SM & Frayn KN (1988) Micro-method for prepar-
ing perchloric extracts of blood. Clin Chem 34, 1657.
Bickerton AS, Roberts R, Fielding BA, et al. (2007) Preferen-
tial uptake of dietary fatty acids in adipose tissue and muscle
in the postprandial period. Diabetes 56, 168–176.
Beylot M, Beaufrere B, Normand S, et al. (1986) Determi-
nation of human ketone body kinetics using stable-isotope
labelled tracers. Diabetologia 29, 90–96.
Hodson L, Bickerton AS, McQuaid SE, et al. (2007) The con-
tribution of splanchnic fat to VLDL-triglyceride is greater in
insulin resistant than insulin sensitive men and women:
studies in the postprandial state. Diabetes 56, 2433–2441.
Wolfe RR & Chinkes D (2005) Isotope Tracers in Metabolic
Research. Hoboken: John Wiley, Sons, Inc.
Hellerstein MK, Christiansen M, Kaempfer S, et al. (1991)
Measurement of de novo hepatic lipogenesis in humans
using stable isotopes. J Clin Invest 87, 1841–1852.
Tan GD, Neville MJ, Liverani E, et al. (2006) The in vivo
effects of the Pro12Ala PPARgamma2 polymorphism on
adipose tissue NEFA metabolism: the first use of the
Oxford Biobank. Diabetologia 49, 158–168.
Havel RJ, Kane JP, Balasse EO, et al. (1970) Splanchnic
metabolism of free fatty acids and production of triglycerides
of very low density lipoproteins in normotriglyceridemic and
hypertriglyceridemic humans. J Clin Invest 49, 2017–2035.
Evans K, Burdge GC, Wootton SA, et al. (2008) Tissue-
specific stable isotope measurements of postprandial lipid
Atherosclerosis 197, 164–170.
Vice E, Privette JD, Hickner RC, et al. (2005) Ketone body
metabolism in lean and obese women. Metabolism 54,
K. Marinou et al. 864
British Journal of Nutrition
18. Burdge GC & Wootton SA (2003) Conversion of alpha-linole-
nic acid to palmitic, palmitoleic, stearic and oleic acids in
men and women. Prostaglandins Leukot Essent Fatty Acids
Grey NJ, Karl I & Kipnis DM (1975) Physiologic mechanisms
in the development of starvation ketosis in man. Diabetes 24,
Johnston DG & Alberti KG (1982) Hormonal control of
ketone body metabolism in the normal and diabetic state.
Clin Endocrinol Metab 11, 329–361.
Beylot M, Picard S, Chambrier C, et al. (1991) Effect of phys-
iological concentrations of insulin and glucagon on the
relationship between nonesterified fatty acids availability
and ketone body production in humans. Metabolism 40,
Merimee TJ & Fineberg SE (1973) Homeostasis during
fasting. II. Hormone substrate differences between men
and women. J Clin Endocrinol Metab 37, 698–702.
23. Keller U, Gerber PP & Stauffacher W (1988) Fatty acid-
independent inhibition of hepatic ketone body production
by insulin in humans. Am J Physiol 254, E694–E699.
Miles JM, Haymond MW, Nissen SL, et al. (1983) Effects of
free fatty acid availability, glucagon excess, and insulin
deficiency on ketone body production in postabsorptive
man. J Clin Invest 71, 1554–1561.
Mittendorfer B, Magkos F, Fabbrini E, et al. (2009) Relation-
ship between body fat mass and free fatty acid kinetics in
men and women. Obesity (Silver Spring) 17, 1872–1877.
Wilding JP (2007) The importance of free fatty acids in the
development of type 2 diabetes. Diabet Med 24, 934–945.
Gibbons GF, Islam K & Pease RJ (2000) Mobilisation of
triacylglycerol stores. Biochim Biophys Acta 1483, 37–57.
Avogaro A, Nosadini R, Bier DM, et al. (1990) Ketone body
kinetics in vivo using simultaneous administration of acetoa-
cetate and 3-hydroxybutyrate labelled with stable isotopes.
Acta Diabetol Lat 27, 41–51.
Hepatic fatty acid partitioning865
British Journal of Nutrition