Effects of n26 PUFAs compared with SFAs on liver fat, lipoproteins,
and inflammation in abdominal obesity: a randomized controlled trial1–4
Helena Bjermo, David Iggman, Joel Kullberg, Ingrid Dahlman, Lars Johansson, Lena Persson, Johan Berglund, Kari Pulkki,
Samar Basu, Matti Uusitupa, Mats Rudling, Peter Arner, Tommy Cederholm, Ha ˚kan Ahlstro ¨m, and Ulf Rise ´rus
Background: Replacing SFAs with vegetable PUFAs has cardio-
metabolic benefits, but the effects on liver fat are unknown. In-
creased dietary n26 PUFAs have, however, also been proposed to
promote inflammation—a yet unproven theory.
Objective: We investigated the effects of PUFAs on liver fat, sys-
temic inflammation, and metabolic disorders.
Design: We randomly assigned 67 abdominally obese subjects
(15% had type 2 diabetes) to a 10-wk isocaloric diet high in vege-
table n26 PUFA (PUFA diet) or SFA mainly from butter (SFA diet),
without altering the macronutrient intake. Liver fat was assessed by
MRI and magnetic resonance proton (1H) spectroscopy (MRS).
Proprotein convertase subtilisin/kexin type-9 (PCSK9, a hepatic
LDL-receptor regulator), inflammation, and adipose tissue expres-
sion of inflammatory and lipogenic genes were determined.
Results: A total of 61 subjects completed the study. Body weight
modestly increased but was not different between groups. Liver fat
was lower during the PUFA diet than during the SFA diet [between-
group difference in relative change from baseline; 16% (MRI; P ,
0.001), 34% (MRS; P = 0.02)]. PCSK9 (P = 0.001), TNF receptor-2
(P , 0.01), and IL-1 receptor antagonist (P = 0.02) concentrations were
lower during the PUFA diet, whereas insulin (P = 0.06) tended to be
higher during the SFA diet. In compliant subjects (defined as change in
serum linoleic acid), insulin, total/HDL-cholesterol ratio, LDL choles-
terol, and triglycerides were lower during the PUFA diet than during the
SFA diet (P , 0.05). Adipose tissue gene expression was unchanged.
Conclusions: Compared with SFA intake, n26 PUFAs reduce liver
fat and modestly improve metabolic status, without weight loss. A
high n26 PUFA intake does not cause any signs of inflammation or
oxidative stress. Downregulation of PCSK9 could be a novel mecha-
nism behind the cholesterol-lowering effects of PUFAs. This trial was
registered at clinicaltrials.gov as NCT01038102.
Am J Clin Nutr
Nonalcoholic fatty liver disease (NAFLD)5affects 25% of the
adult population (1) and is strongly associated with metabolic
disorders and type 2 diabetes, even independently of abdominal
obesity (2, 3). Reduction of liver fat is thus an interesting target
for preventing and treating obesity-related metabolic diseases.
Besides physical activity and weight loss (4), there is no ef-
fective and safe treatment to reduce liver fat. Interestingly, di-
etary fatty acids may influence the accumulation of hepatic (5)
and abdominal (6, 7) fat. A diet high in vegetable n26 PUFAs
decreases abdominal fat content and peripheral insulin re-
sistance compared with a diet high in SFAs (6). A high fat intake
may promote liver fat accumulation (8), but observational data
also suggest that dietary fat composition could play a role (9,
10). SFAs have been positively related to liver fat (9, 10),
whereas the essential n26 PUFA linoleic acid (18:2n26) has
been inversely related to plasma alanine aminotransferase (ALT)
1From the Departments of Public Health and Caring Sciences (HB, DI,
SB, TC, and UR) and Radiology (JK, LJ, JB, and HA), Uppsala University,
Uppsala, Sweden; the Center for Clinical Research Dalarna, Falun, Sweden
(DI); the Department of Medicine, Karolinska Institutet at Karolinska Uni-
versity Hospital, Huddinge, Stockholm, Sweden (ID and PA); AstraZeneca
R&D, Mo ¨lndal, Sweden (LJ); the Department of Endocrinology, Metabolism
& Diabetes and Center for Biosciences, Department of Medicine, Karolinska
Institutet at Karolinska University Hospital, Huddinge, Stockholm, Sweden
(LP and MR); the Department of Clinical Chemistry, University of Eastern
Finland and Eastern Finland Laboratory Centre, Kuopio, Finland (KP); the
Laboratory of Biochemistry, Molecular Biology and Nutrition, Universite
d’Auvergne, Clermont-Ferrand, France (SB); the Institute of Public Health
and Clinical Nutrition, Clinical Nutrition, University of Eastern Finland,
Kuopio, Finland (MU); and the Research Unit, Kuopio University Hospital,
Kuopio, Finland (MU).
2Presented orally and as a poster at the 29th International Symposium on
Diabetes and Nutrition of the Nutrition Study Group, Rome, Italy, 1 July 2011.
3Supported by a grant from the Swedish Council for Working Life and
4Address correspondence and reprint requests to U Rise ´rus, Clinical Nu-
trition and Metabolism, Department of Public Health and Caring Sciences,
Uppsala University, Uppsala Science Park, 75185 Uppsala, Sweden. E-mail:
5Abbreviations used: ALT, alanine aminotransferase; CAD, coronary ar-
tery disease; Ct, comparative threshold cycle; IL-1RA, IL-1 receptor antag-
onist; MRS, magnetic resonance proton (1H) spectroscopy; NAFLD,
nonalcoholic fatty liver disease; OGTT, oral-glucose-tolerance test; PCSK9,
proprotein convertase subtilisin/kexin type-9; SCD-1, stearoyl-CoA desatur-
ase 1; TNF-R2, TNF receptor-2.
Received November 10, 2011. Accepted for publication February 22, 2012.
Am J Clin Nutr doi: 10.3945/ajcn.111.030114. Printed in USA. ? 2012 American Society for Nutrition
1 of 10
AJCN. First published ahead of print April 4, 2012 as doi: 10.3945/ajcn.111.030114.
Copyright (C) 2012 by the American Society for Nutrition
Linoleic acid is the main dietary n26 PUFA and is abundant in
plant oils, seeds, and nuts. Solid evidence indicates that replacing
SFAs with PUFAs reduces coronary artery disease (CAD) events
(11, 12) and possibly prevents diabetes (13). The mechanism for
the CAD risk reduction involves the LDL-cholesterol lowering of
n26 PUFAs (14), but other potential metabolic effects need
further study. The mechanism behind SFA-induced increases in
serum cholesterol may involve both lowered LDL-receptor
numbers and PGC-1b induction (15, 16) but the LDL-cholesterol
lowering of PUFAs is unclear. Proprotein convertase subtilisin/
kexin type 9 (PCSK9) has been identified as a key regulator of
plasma LDL cholesterol by modulation of the degradation of LDL
receptors (17, 18). PCSK9 inhibition is a drug target to reduce
serum cholesterol, particularly in combination with statins, be-
cause statins increase PCSK9 (19). Subjects with a certain loss of
function variants of the PCSK9 gene have 15–28% lower serum
cholesterol and a 47–88% lower CAD risk (18). The effect of
PUFA or SFA on circulating PCSK9 concentrations is to our
In earlier literature it has been argued that n26 PUFAs may
promote inflammation, mainly by increasing circulating linoleic
acid and consequently arachidonic acid and proinflammatory
metabolites (20)—a theoretical reasoning that remains to be
established in controlled studies. Whereas most lipid research
has focused on n23 PUFA, few controlled studies have in-
vestigated the effects of n26 in humans. On the basis of animal
(5) and observational (9, 10) data, we hypothesized that dietary
fat modification could influence liver fat, even in the absence of
weight loss or caloric restriction. In the randomized HEPFAT
trial, we examined the effects of n26 PUFA and SFA on liver fat
content, serum PCSK9, blood lipid concentrations, glucose
metabolism, lipid peroxidation, and inflammation in abdomi-
nally obese subjects.
SUBJECTS AND METHODS
The HEPFAT trial was a randomized, 10-wk, parallel-group
April 2010. Subjects were recruited by advertisements in local
newspapers, supermarkets, and primary health care centers. In-
clusion criteria assessed by screening were age 30–65 y, sagittal
abdominal diameter .25 cm, or waist circumference .88 cm
(women) or .102 cm (men). Exclusion criteria were diagnosed
liver disease, type 1 diabetes, history of a serious cardiovascular
event, BMI (in kg/m2) .40, excessive alcohol intake, and in-
ternal metal or electronic device. After being screened, 67 in-
dividuals were eligible (Figure 1). All participants gave written
informed consent before entering the study. The study was ap-
proved by the regional ethical committee in Uppsala.
Participants were randomly assigned to either a PUFA diet or
an SFA diet. The randomization was stratified according to sex
and performed in blocks of 4 (allocation ratio 1:1). The par-
ticipants were instructed (unblinded) to change the quality of
their dietary fat without altering their intakes of total fat and the
type and amount of carbohydrates and protein. The participants
were encouraged not to change their physical activity or their
fish and alcohol intakes during the study. Some key food items
were provided: the PUFA group received foods rich in n26
FIGURE 1. Flow diagram.*The diseases were known before enrollment (heart valve disorder and chronic obstructive pulmonary disease). MRI, magnetic
resonance imaging; MRS, magnetic resonance proton (1H) spectroscopy.
2 of 10
BJERMO ET AL
linoleic acid, ie, scones (baked-on sunflower oil), margarine,
sunflower oil, and sunflower seeds, and the SFA group received
scones (baked-on butter) and butter. On the basis of weight
and sex, participants consuming the PUFA diet were instructed
to consume the given food items corresponding to ;15% of
energy as linoleic acid. To avoid weight loss during the in-
tervention, all participants were urged to weigh themselves
The primary outcomewas change in liver fat content measured
by MRI and by magnetic resonance proton (1H) spectroscopy
(MRS). Secondary outcomes were abdominal and total body fat,
serum lipid and PCSK9 concentrations, insulin sensitivity, sys-
temic inflammation, oxidative stress (lipid peroxidation), and
adipose tissue gene expression.
The examinations were performed at baseline (visit 1), after 5
wk (visit 2), and after 10 wk (visit 3). All visits took place in the
morning after an overnight fast. Visits 1 and 3 started with an MR
investigation, followed by anthropometric and blood pressure
measurements, blood sampling, an oral-glucose-tolerance test
collected at visits 1 and 3. Visit 2 included anthropometric
measurements and blood sampling.
Subcutaneous and visceral adipose tissue volumes were
assessed by MRI, and liver fat content was assessed by both MRI
and MRS. MR measurements were performed by using a 1.5T
clinical scanner (Gyroscan NT; Philips Health Care). Collection
and analyses of the MR data were performed at one center under
blinded conditions and is described in detail elsewhere (see
supplemental material under “Supplemental data” in the online
issue). Total body fat mass was determined by air-displacement
plethysmography (BOD POD; LmiTech).
Body weight was measured while the subjects were in un-
derwear to the nearest 0.1 kg. Blood was drawn from an ante-
350 mL water was consumed, and blood samples were collected
at 30, 60, 90, and 120 min.
Biopsy samples were collected from the subcutaneous adipose
tissue fat pad below and lateral to the umbilicus by needle as-
piration under local anesthesia (1% lidocaine), washed with
physiologic saline, frozen on dry ice covered with ethanol, and
stored at 270?C. Dietary intake was assessed from 3-d weighed
food records before randomization and between visits 2 and 3.
The Dietist XP software package (version 3.0, 2007) was used to
calculate dietary intake.
Plasma concentrations of glucose, triglycerides, cholesterol,
LDL cholesterol, HDL cholesterol, apolipoproteins, ALT,
c-glutamyltransferase, C-reactive protein, and serum insulin
concentrations were measured by routine laboratory methods at
Uppsala University Hospital. AUCs for insulin and glucose
concentrations during the OGTT were calculated according to
the trapezoid rule. Serum fatty acid composition was measured
in cholesterol esters by gas chromatography as previously de-
scribed (21), with the following divergences: use of a 30-m glass
capillary column coated with Thermo TR-FAME (Thermo
Electron Corporation); an Agilent Technologies system con-
sisting of model GLC 6890N, autosampler 7683, and Agilent
ChemStation; and a programmed temperature of 150?C to 260?C.
The stearoyl-coA desaturase-1 (SCD-1) index was estimated by
serum 16:1/16:0. ELISAwas used to determine serum fibroblast
growth factor 21 (R&D Systems), PCSK9 (CycLex), proinsulin
(Mercodia), and fetuin-A (Mercodia) as well as plasma con-
centrations of high-molecular-weight adiponectin (Millipore),
IL-1b, IL-6, IL-10, IL-1 receptor antagonist (IL-1RA), and
soluble TNF receptor-2 (TNF-R2; all R&D Systems). Serum
unesterified lathosterol was determined by isotope dilution mass
spectrometry and corrected for total cholesterol concentration
Urinary 8-iso-prostaglandin F2aand 15-keto-13,14-dihydro-
prostaglandin F2awere assessed by radioimmunoassay as indicators
of oxidative stress and lipid peroxidation, ie, free radical–induced
and enzymatic induced oxidation of arachidonic acid, respectively
(23, 24). Data were adjusted for urinary creatinine; 15-keto-
dihydro-PGF2awas analyzed only in compliant participants
(definition described below).
Adipose tissue mRNA expression of target genes was assessed
in compliant subjects by quantitative real-time polymerase chain
reaction (iCycler IQ; Bio-Rad Laboratories) by using a com-
parative threshold cycle (Ct) method. Ct values were normalized
to the reference gene LDL receptor–related protein 10 or 18S,
according to the following formula:
2DCt?target gene=2DCt?reference gene¼ arbitrary units
Variables are presented as means 6 SDs or medians (quartiles
1–3). In power analyses, 31 individuals per group would be
needed to observe a 5% difference in liver fat with an SD for
liver fat of 7% (8) (a = 0.05, b = 0.20). Non-normally distributed
variables were logarithmically transformed, and nonparametric
tests were used if normality was not attained. Statistical analyses
were based on per-protocol instead of intention-to-treat princi-
ples because we believe the latter analyses are not always op-
timal in dietary interventions with a limited number of subjects.
Also, in contrast with drug therapy with clinical endpoints, the
primary aim was academic, ie, to investigate the role of dietary
fat quality on liver fat content rather than to evaluate a clinical
treatment and its adverse events in patients. We a priori defined
compliance using serum fatty acid biomarkers and conducted
post hoc analyses, because the change in dietary SFAs and
PUFAs will be dependent on the baseline (ie, habitual) intake of
the individual. Compliance was defined according to changes in
serum linoleic acid, ie, 18:2n26 .0.0% during the PUFA diet (n
= 27) and 18:2n26 ,0.0% during the SFA diet (n = 19). The
rationale for using 18:2n26 is that changes in serum cholesterol
esters are strongly associated with changes in dietary intake
(25), ie, if no change in serum 18:2n26 can be observed in an
individual, then little actual change from baseline in fatty acid
intake has taken place. Hemolytic samples were excluded
in sensitive analyses (fasting insulin, n = 3; AUCinsulin, n = 7;
c-glutamyltransferase, n = 1; cytokines, n = 1; adiponectin, n =
1). Baseline values and changes during the intervention [meas-
urement(visit3)2 measurement(visit1)] were compared between
DIETARY FATTY ACIDS AND LIVER FAT
3 of 10
the groups by 2-sided t test or Wilcoxon’s rank-sum test.
Baseline adjustments were made by ANCOVA or the residual
method. P values represent the differences in change between
the groups when adjusted for baseline. The covariates in sec-
ondary analyses were as follows: baseline value for the variable
tested, weight change, change in energy intake, and change in
total fat intake. The primary statistical analyses were performed
blinded. P ? 0.05 was considered statistically significant. The
software package STATA version 11 (STATA Corporation) was
Of the 67included subjects, 61 completed the study (Figure 1).
in Table 1. The results were similar (all significant differences
remained) regardless of whether intention-to-treat or per-protocol
analyses were used. None of the variables differed between the
groups at baseline (P . 0.05), and 66% of both groups were
women. Subjects in the PUFA diet were aged 57 (51–63) y and
in the SFA diet were aged 56 (50–64) y, and 15% of the subjects
had diabetes (previous diagnosis, fasting glucose ?7.0 mmol/L,
or glucose ?11.0 mmol/L 2 h after an OGTT). Use of antihy-
pertensive drugs and lipid-lowering drugs was 31% and 16%,
respectively. Two subjects in the PUFA-diet group reported
loose stool shortly after study initiation; no other adverse events
Energy intakes were 2052 6 514 kcal (PUFA diet) and 1945
6 493 kcal (SFA diet) at baseline and increased by 138 6 451
kcal (PUFA diet) and 225 6 509 kcal (SFA diet) during the
intervention, but did not differ significantly between the groups
(P = 0.59). Changes in nutrient composition are presented in
Figure 2A. During the PUFA diet, the n26/n23 ratio increased
from 4 (3–5) to 14 (10–19). The intervention had considerable
effects on serum fatty acid composition (Figure 2B; see sup-
plemental material under “Supplemental data” in the online is-
sue). The proportion of linoleic acid increased from baseline by
11% in the PUFA group. Those n26 PUFAs that are mainly
endogenously synthesized (18:3n26, 20:3n26, and 20:4n26)
did not differ between the diets (P . 0.36). The SCD-1 index
was lower during the PUFA diet than during the SFA diet (P ,
Baseline characteristics and changes in body composition and metabolic factors during the intervention1
Body compositionPUFA diet SFA dietPUFA dietSFA dietP4
Waist circumference (cm)
Visceral AT (L)
Subcutaneous AT (L)
Visceral AT/subcutaneous AT
Total fat mass (%)
Lipids and cholesterol
Serum PCSK9 (lg/L)
Fasting plasma triglycerides (mmol/L)
Plasma cholesterol (mmol/L)
Plasma HDL cholesterol (mmol/L)
Plasma LDL cholesterol (mmol/L)
Plasma apolipoprotein A-I (g/L)
Plasma apolipoprotein B (g/L)
Serum lathosterol/cholesterol (mg/mol)
Fasting plasma glucose (mmol/L)
Fasting serum insulin (pmol/L)
Other biochemical markers
Plasma HMW adiponectin (mg/L)
Plasma ALT (lkat/L)
Serum FGF21 (ng/L)
Serum fetuin-A (lg/mL)
85.7 6 10.65
30.3 6 3.7
5.1 6 1.7
10.0 6 3.3
42.9 6 7.5
90.8 6 14.4
31.3 6 3.9
6.2 6 2.8
10.8 6 3.5
43.2 6 6.6
0.4 6 1.4
0.1 6 0.5
21 (22 to 1)
0.0 6 0.3
0.3 6 0.5
20.01 (20.02 to 0.01)
20.1 6 1.4
0.8 6 1.6
0.3 6 0.6
1 (21 to 2)
0.1 6 0.3
0.1 6 0.5
0.00 (20.01 to 0.01)
0.6 6 1.4
273 6 70
1.51 6 0.70
5.5 6 0.9
3.4 6 0.9
1.4 6 0.2
1.0 6 0.2
327 6 86
278 6 54
1.44 6 0.66
5.5 6 1.0
3.3 6 0.7
1.5 6 0.3
1.0 6 0.2
350 6 99
236 6 69
20.07 6 0.53
20.2 6 0.4
0.0 (20.2 to 0.1)
20.1 6 0.3
0.0 6 0.1
0.0 6 0.1
244 6 82
15 6 78
0.06 6 0.39
0.0 6 0.5
0.0 (20.1 to 0.1)
0.0 6 0.4
0.1 6 0.2
0.0 6 0.1
34 6 83
0.0 (20.2 to 0.2)
2.2 (212 to 15)
0.1 (20.5 to 0.6)
0.6 (21.0 to 1.7)
0.0 (20.4 to 0.2)
0.7 (20.2 to 4.3)
240.4 6 38.6
251.0 6 28.6
20.02 (20.36 to 0.36)
20.04 (20.15 to 0.09)
216.0 (248.5 to 27.5)
20.2 6 22.1
20.03 (20.74 to 0.68)
20.01 (20.04 to 0.10)
223.2 (238.1 to 23.4)
2.2 6 23.5
1n = 32 (PUFA diet) and n = 29 (SFA diet). ALT, alanine aminotransferase; AT, adipose tissue; FGF21, fibroblast growth factor 21; HMW, high-
molecular-weight; MRS, magnetic resonance proton (1H) spectroscopy; PCSK9, proprotein convertase subtilisin/kexin type 9.
2The variables did not differ significantly between the groups at baseline.
3Change denotes measure at follow-up 2 measure at baseline.
4P values were derived by ANCOVA or the residual method and were adjusted for baseline values.
5Mean ± SD (all such values).
6Median (quartiles 1–3) (all such values).
4 of 10
BJERMO ET AL
Liver and body fat
Body weight modestly increased during both diets without be-
tween-groups difference (Table 1). Liver fat values at baseline were
and SFA diet [7.5%, 5.5–15.7 (MRI); 3.2%, 1.3–7.7 (MRS)]. Liver
significantly lower during the PUFA diet than during the SFA diet.
Changes during the trial were 20.5% (22.3 to 0.2; MRI) and
20.9% (21.7 to 0.0; MRS) with the PUFA diet and 0.7% (20.2 to
2.1; MRI) and 0.3% (20.6 to 1.8; MRS) for the SFA diet (Figure
3). The results were unaltered after adjustment for weight change
and total fat intake or after the exclusion of subjects with liver fat
?1.0% (MRS) at baseline (n = 12), ie, when the assessment is less
sensitive. Change in liver fat was inversely related to change in
serum linoleic acid concentrations and was positively associated
with change in serum SFAs (Figure 4). No differences were ob-
served for subcutaneous or visceral adipose tissue. However, a small
but significant difference in the visceral-to-subcutaneous adipose
tissue ratio was found (Table 1). The changes in percentage total
body fat were 20.1 6 1.4% (PUFA diet) and 0.6 6 1.4% (SFA
diet) (P = 0.09).
PCSK9 and blood lipids
Serum PCSK9 and the lathosterol/cholesterol ratio were lower
during the PUFA diet than during the SFA diet. Plasma cholesterol
slightly decreased during the PUFA diet (P = 0.01), whereas no
significant effects on other blood lipids were found unless com-
pliance with the diets was taken into account (Table 1).
Fasting insulin concentrations tended to be higher during the
SFA diet than during the PUFA diet (P = 0.06) and was significant
in compliant subjects (Figure 5). Fasting glucose, HOMA,
AUCglucose, AUCinsulin, and proinsulin did not differ between the
diet groups (Table 1).
Liver enzymes and adiponectin
No differences between the diets were observed for ALT, fi-
broblast growth factor 21, or c-glutamyltransferase concentrations
(Table 1). Change in ALT was related to change in liver fat [r =
0.33, P = 0.01 (MRI) and r = 0.29, P = 0.03 (MRS)], which in-
dicated a stronger correlation in subjects with MRS liver fat ?1.0%
(r = 0.59, P , 0.001; n = 43). All 3 liver markers correlated with
MRS liver fat at baseline (P , 0.04). Plasma high-molecular-
weight adiponectin was unchanged by the intervention (Table 1).
Inflammation and oxidative stress
Plasma IL-1RA and TNF-R2 were lower during the PUFA diet
than during the SFA diet (Table 2). No effects were observed for
the other markers of inflammation or oxidative stress.
FIGURE 2. Dietary intake and serum fatty acid composition at baseline and changes during the intervention. n = 30 (PUFA diet) and n = 29 (SFA diet).
Data are presented as medians (quartiles 1–3) in A and as means (6SDs) in B. A 2-sided t test was used in the statistical analyses. Change denotes measure at
follow-up 2 measure at baseline. E%, percentage of energy.
DIETARY FATTY ACIDS AND LIVER FAT
5 of 10
Results similar to those from the per-protocol analyses were
observed for most variables in the compliance analyses (data not
(P = 0.04) between groups [PUFA diet: 20.7 (214 to 7.2) pmol/
L; SFA diet: 14 (20.7 to 23) pmol/L], whereas other measures
of glucose metabolism were nonsignificant (P ? 0.08). More-
over, triglycerides, cholesterol, LDL cholesterol, and the total
cholesterol/HDL-cholesterol ratio were significantly lower dur-
ing the PUFA diet than during the SFA diet (Figure 5).
Adipose tissue gene expression
No effects on adipose tissue mRNA expression were observed
for the investigated preselected target genes involved in lipid
FIGURE 4. Linear regressions showing relations between changes in serum fatty acid composition and changes in logarithmized liver fat measured by
magnetic resonance imaging. n = 29 (PUFA diet) and n = 28 (SFA diet). Linear regression and Pearson’s correlation were used for the statistical analyses.
Fatty acids were measured in serum cholesterol esters. Change denotes measure at follow-up 2 measure at baseline.
FIGURE 3. Changes in absolute and relative liver fat content during the intervention. The boxes represent the 25th, 50th, and 75th percentiles and reflect
absolute changes (percentage units) in liver fat. The lower lines indicate the 10th percentile and the upper lines the 90th percentile. Higher and lower values are
not presented in the figure. Numbers within boxes refer to relative changes from baseline. P ? 0.03 for the difference between groups. P values were derived
from ANCOVA and adjusted for baseline values. Compliant participants were defined according to changes in serum linoleic acid, ie, 18:2n26 .0.0% during
the PUFA diet [n = 26 (MRI), n = 24 (MRS)] and 18:2n26 ,0.0% during the SFA diet [n = 18 (MRI and MRS)]. MRI, magnetic resonance imaging; MRS,
magnetic resonance proton (1H) spectroscopy.
6 of 10
BJERMO ET AL
metabolism and inflammation (see supplemental material under
“Supplemental data” in the online issue).
This study provides severalnew insightsconcerning therole of
n26 PUFAs in liver fat accumulation, lipoprotein metabolism,
and inflammation. First, in the absence of weight loss, a diet
high in PUFAs reduced liver fat compared with an SFA-rich diet.
Second, this effect was accompanied by a moderate improve-
ment in blood lipids and fasting insulin, which was more evident
in compliant subjects. Third, PUFAs decreased serum PCSK9
concentrations—a mechanism potentially explaining the PUFA-
induced lowering of plasma LDL cholesterol. Fourth, PUFA had
no adverse effects on oxidative stress or inflammation; instead,
n26 PUFA may act as an anti-inflammatory, as evidenced by the
reduced TNF-R2 and IL-1RA concentrations.
The reduction in liver fat during the PUFA diet as compared
FIGURE 5. Changes in blood lipids and insulin during the intervention in compliant participants. Compliant participants were defined according to changes
in serum linoleic acid, ie, 18:2n26 .0.0% during the PUFA diet (n = 27) and 18:2n26 ,0.0% during the SFA diet (n = 19). P values were derived by
ANCOVA and adjusted for baseline values. TG, Total-C, LDL-C, and insulin are skewed and presented as medians (quartiles 1–3). HDL-C, total-C/HDL-C
ratio, and PCSK9 are presented as means 6 SDs. Change denotes measure at follow-up 2 measure at baseline. C, cholesterol; PCSK9, protein convertase
subtilisin/kexin type 9; TG, triglycerides.
Baseline characteristics and changes in markers of inflammation and oxidative stress during the intervention1
PUFA diet SFA dietPUFA dietSFA dietP4
Plasma C-reactive protein (mg/L)
Plasma IL-1b (ng/L)
Plasma IL-6 (ng/L)
Plasma IL-10 (ng/L)
Plasma IL-1RA (ng/L)
Plasma TNF-R2 (ng/L)
Urinary 8-iso-PGF2a(nmol/mmol creatinine)
0.56 6 0.206
0.60 6 0.18
20.2 (20.6 to 0.5)
0.00 (20.36 to 0.45)
0.00 (20.02 to 0.21)
27.0 (252.8 to 19.2)
2108 (2204 to 0)
0.01 6 0.18
0.00 (20.03 to 0.02)
0 (20.5 to 0.7)
0.00 (20.08 to 0.03)
0.05 (20.30 to 0.45)
23.0 (223.7 to 101.8)
94 (222 to 261)
20.04 6 0.18
20.01 (20.04 to 0.02)
1n = 32 (PUFA diet) and n = 29 (SFA diet). IL-1RA, IL-1 receptor antagonist; 8-iso-PGF2a, TNF-R2, TNF receptor-2; 8-iso-prostaglandin 2a; 15-keto-
dihydro-PGF2a, 15-keto-13,14-dihydro-prostaglandin 2a.
2The variables did not differ between the groups at baseline.
3Change denotes measure at follow-up 2 measure at baseline.
4P values were derived by ANCOVA or the residual method and were adjusted for baseline values.
5Median (quartiles 1–3) (all such values).
6Mean ± SD (all such values).
7The analysis was performed only in compliant participants. Compliance was defined according to changes in serum linoleic acid, ie, 18:2n26 .0.0%
during the PUFA diet (n = 27) and 18:2n26, 0.0% during the SFA diet (n = 19).
DIETARY FATTY ACIDS AND LIVER FAT
7 of 10
in most participants. Possibly, the effect might have been even
larger in subjects with fatty livers (26). The magnitude of the
reduction is comparable with the effects of a low-fat or hypo-
caloric diet (8), exercise (26), and moderate weight loss (4).
According to the food records, slight increases in total fat intake
were observed with both diets, especially with the SFA diet.
Still, effects on liver fat content were diverse during the diets,
and adjustment for changes in both total fat and energy intake
did not influence the results. Another possible confounder is di-
etary cholesterol, although there is yet no human data to suggest
a direct effect of cholesterol on liver fat content (27).
Clinical trials examining the effects of dietary fats on body fat
content are sparse (27). In one study, n23 PUFA decreased liver
fat but, in contrast with our study, that effect may have been
caused by weight loss (28). Diets rich in n26 PUFAs were
previously shown to reduce subcutaneous adipose tissue (6) and
trunk adipose tissue (7), whereas we observed only a small but
significant decrease in the visceral-to-subcutaneous adipose
tissue ratio. It has been proposed that n26 PUFAs may promote
adiposity (29)—a view that is apparently not supported by the
current or other controlled studies (6, 7).
Circulating PCSK9 and lathosterol, the latter a serum marker
for cholesterol synthesis (22), were both reduced during PUFA
feeding, which indicates a novel pathway to explain the plasma
LDL-cholesterol lowering effect of PUFAs. This finding is of
high interest because PCSK9 is a novel drug target and lowering
PCSK9 reduces plasma LDL cholesterol (18). Gene activities of
both PCSK9 and HMG CoA reductase are regulated by sterol
regulatory element-binding protein-2 and, thus, the hepatic
cholesterol concentration (30, 31). Therefore, the finding of
reduced concentrations of circulating PCSK9 and lathosterol are
in linewith thenotionthat PUFAincreases hepatic cholesterol, as
has also been shown in rats (32), resulting in decreased activity of
sterol regulatory element-binding protein-2 and subsequently
reduced serum concentrations of PCSK9 and lathosterol. In
compliant subjects, the PUFA diet caused an overall improve-
ment in the blood lipid profile, in line with several previous trials
in which PUFA was substituted for SFA (14, 33–35). However,
the overall effects on LDL cholesterol, and the total/HDL-
cholesterol ratio in particular were less pronounced as compared
with previous data (14). As judged from the current reported
dietary linoleic acid intake and, to some extent, changes in serum
linoleic acid, somewhat greater effects would havebeen expected
(14, 25). Because we did not provide meals to the subjects, com-
pliance was probably lower than that in shorter strictly controlled
feeding trials (14, 25, 35).
diet as compared with the PUFA diet, which indicated impaired
insulin sensitivity. This agrees with a recent 16-wk intervention in
which n26 PUFA improved insulin sensitivity (36). Furthermore,
in a 5-wk crossover study, an n26 PUFA diet improved insulin
sensitivity (euglycemic clamp) compared with an SFA diet (6).
Note that the current relatively high intake of total fat (;39% of
energy) may counteract an insulin-sensitizing effect of un-
saturated fats (37). On the other hand, recent epidemiologic data
showed that a diet rich in total fat (;40% of energy) was asso-
ciated with a decreased risk of type 2 diabetes as long as the fat
was predominantly plant-based and rich in n26 PUFAs (38).
Whether a PUFA-induced reduction of liver fat could mediate
such an association warrants further investigation.
Dietary n26 PUFA or a high n26/n23 ratio has been sug-
gested to increase inflammation and lipid peroxidation through
its conversion to arachidonic acid (20). We found no support for
such a hypothesis. Despite the marked increase in linoleic acid
intake (14% of energy) and the 3.5-fold increase in the dietary
n26/n23 ratio, serum arachidonic acid concentrations were not
elevated. Moreover, neither systemic proinflammatory effects
nor signs of free radical–mediated or cyclooxygenase-2–mediated
lipid peroxidation were observed. In contrast, IL-1RA and TNF-
R2 decreased during the PUFA diet compared with the SFA diet,
possibly suggesting antiinflammatory effects of PUFA and/or
proinflammatory properties of SFA (20). Notably, these markers
are elevated in individuals long before the onset of type 2 di-
abetes (39, 40). An annual increase in IL-1RA concentrations
of 11 ng/L was observed during the 6 y before diabetes onset
(39), which suggests that the current changes in IL-1RA (PUFA
diet: 27.0 ng/L; SFA diet: +23.0 ng/L) could be clinically rel-
evant. Insufficient statistical power may have been a factor in the
lack of effect on other inflammation markers or gene expression
in adipose tissue. Possible chance findings due to multiple testing
also need to be considered.
The mechanism behind a PUFA-induced reduction in liver fat
is unclear. However, long-chain PUFAs are preferentially
b-oxidized compared with long-chain SFAs (41). Furthermore,
PUFA in contrast with SFA inhibits de novo hepatic fatty acid
synthesis and lipogenic gene expression (41–43). The reduction
in liver fat was strongly associated with changes in the SCD-1
index, in accordance with observational data (10, 44). The SCD-1
index is associated with hepatic SCD-1 expression in humans (45)
and with SCD-1 activity in animals (46, 47); however, this index
should still be regarded as a marker of SCD-1 activity from which
we cannot draw conclusions on enzyme activity (48). This
finding may, however, imply that the SCD-1 index may be a
useful serum marker of liver fat content. A decreased serum
SCD-1 index during the PUFA diet was not reflected by reduced
adipose tissue SCD-1 expression, which indicates that it may
mainly reflect hepatic SCD-1 activity, in accordance with pre-
vious data (45–47).
This study was limited by a nonblinded design, which was not
feasible because of the diets the subjects were advised to eat. The
MR analyses and primary statistical analyses were, however,
blinded. Nonstandardized diets may increase bias but may also
underestimate the effects and reflect realistic changes that can be
achieved in clinical practice. Strengths of the study included the
assessment of liver fat by both MRI and MRS. Although MRS is
considered the gold standard, MRI analysis includes the whole
liver and may better represent total liver fat. A limitation was that
the imaging and spectroscopy methods used did not include full
characterization of all lipid resonances of the liver spectra and
therefore did not allow for a more detailed analysis of changes in
liver lipid saturation, for example. We therefore cannot com-
pletely exclude the possibility that our liver fat measures were
biased by changes in lipid saturation levels. The significant
correlation between changes in liver fat and changes in plasma
similarly to morphologic and histologic data in rodents that
showed a reduction in liver fat content after PUFA feeding (49,
50). The low dropout rate strengthens the data, and the inclusion
of free-living subjects with diabetes, hypertension, and dyslipi-
demia increases the generalizability of the data. Also, this study
8 of 10
BJERMO ET AL
was supported by governmental funding without involvement of
any food industry.
In conclusion, compared with SFAs, dietary n26 PUFAs re-
duce liver fat in overweight individuals in the absence of weight
loss. This difference was observed even though both diets were
rather high in total fat. These results have potential implications
for public health, considering the high prevalence of NAFLD;
however, the findings need to be confirmed in other populations
and in NAFLD patients. A reduced plasma PCSK9 concentra-
tion could be a novel mechanism behind the plasma total and
LDL-cholesterol-lowering effects of PUFAs.
We thank Gunilla Arvidsson and Anders Lundberg at the Radiology unit,
UppsalaUniversity for the MRassessment;SivTengbladat ClinicalNutrition
and Metabolism, Uppsala University; Lillemor Ka ¨llstro ¨m at the Oxidative
Stress and Inflammation Unit, Uppsala University; Eva Sjo ¨lin at Karolinska
University Hospital, Huddinge, Karolinska Institutet, for excellent laboratory
work; Maritta Siloaho, University of Eastern Finland, for supervising the
high-molecular-weight adiponectin analyses; Arvo Ha ¨nni at Uppsala Aca-
demic Hopsital; and Good Food Practice in Uppsala for contributing to
the data collection.
had primary responsibility for the conception and design of the study, includ-
ingthe finalcontent;HBandUR:designedtheresearch; HB,DI, TC,andUR:
conducted the research; DI: had medical responsibility in the study; HB, JK,
ID, LJ, LP, MR, KP, JB, SB, MU, and HA; analyzed and interpreted the data;
PA, TC, and HA: provided essential materials and interpreted the data; HB:
performedthestatistical analyses;andHB, DI, andUR: wrotethe manuscript.
All authors reviewed and edited the manuscript. No potential conflicts of in-
terest relevant to this article were reported.
1. Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohepatitis:
summary of an AASLD Single Topic Conference. Hepatology 2003;
2. Yki-Ja ¨rvinen H. Liver fat in the pathogenesis of insulin resistance and
type 2 diabetes. Dig Dis 2010;28:203–9.
3. Fabbrini E, Magkos F, Mohammed BS, Pietka T, Abumrad NA, Patterson
BW, Okunade A, Klein S. Intrahepatic fat, not visceral fat, is linked with
metabolic complications of obesity. Proc Natl Acad Sci USA 2009;106:
4. Lazo M, Solga SF, Horska A, Bonekamp S, Diehl AM, Brancati FL,
Wagenknecht LE, Pi-Sunyer FX, Kahn SE, Clark JM. Effect of a
12-month intensive lifestyle intervention on hepatic steatosis in adults
with type 2 diabetes. Diabetes Care 2010;33:2156–63.
5. Buettner R, Parhofer KG, Woenckhaus M, Wrede CE, Kunz-Schughart
LA, Scholmerich J, Bollheimer LC. Defining high-fat-diet rat models:
metabolic and molecular effects of different fat types. J Mol Endo-
6. Summers LK, Fielding BA, Bradshaw HA, Ilic V, Beysen C, Clark ML,
Moore NR, Frayn KN. Substituting dietary saturated fat with poly-
unsaturated fat changes abdominal fat distribution and improves insulin
sensitivity. Diabetologia 2002;45:369–77.
7. Norris LE, Collene AL, Asp ML, Hsu JC, Liu LF, Richardson JR, Li D,
Bell D, Osei K, Jackson RD, et al. Comparison of dietary conjugated
linoleic acid with safflower oil on body composition in obese post-
menopausal women with type 2 diabetes mellitus. Am J Clin Nutr
8. Westerbacka J, Lammi K, Hakkinen AM, Rissanen A, Salminen I, Aro
A, Yki-Jarvinen H. Dietary fat content modifies liver fat in overweight
nondiabetic subjects. J Clin Endocrinol Metab 2005;90:2804–9.
9. Tiikkainen M, Bergholm R, Vehkavaara S, Rissanen A, Hakkinen AM,
Tamminen M, Teramo K, Yki-Jarvinen H. Effects of identical weight
loss on body composition and features of insulin resistance in obese
women with high and low liver fat content. Diabetes 2003;52:701–7.
10. Petersson H, Arnlov J, Zethelius B, Riserus U. Serum fatty acid com-
position and insulin resistance are independently associated with liver fat
markers in elderly men. Diabetes Res Clin Pract 2010;87:379–84.
11. Mozaffarian D, Micha R, Wallace S. Effects on coronary heart disease
of increasing polyunsaturated fat in place of saturated fat: a systematic
review and meta-analysis of randomized controlled trials. PLoS Med
12. Astrup A, Dyerberg J, Elwood P, Hermansen K, Hu FB, Jakobsen MU,
Kok FJ, Krauss RM, Lecerf JM, LeGrand P, et al. The role of reducing
intakes of saturated fat in the prevention of cardiovascular disease: where
does the evidence stand in 2010? Am J Clin Nutr 2011;93:684–8.
13. Rise ´rus U, Willett WC, Hu FB. Dietary fats and prevention of type 2
diabetes. Prog Lipid Res 2009;48:44–51.
14. Mensink RP, Katan MB. Effect of dietary fatty acids on serum lipids
and lipoproteins. A meta-analysis of 27 trials. Arterioscler Thromb
15. Mustad VA, Etherton TD, Cooper AD, Mastro AM, Pearson TA,
Jonnalagadda SS, Kris-Etherton PM. Reducing saturated fat intake is
associated with increased levels of LDL receptors on mononuclear
cells in healthy men and women. J Lipid Res 1997;38:459–68.
16. Lin J, Yang R, Tarr PT, Wu PH, Handschin C, Li S, Yang W, Pei L,
Uldry M, Tontonoz P, et al. Hyperlipidemic effects of dietary saturated
fats mediated through PGC-1beta coactivation of SREBP. Cell 2005;
17. Abifadel M, Varret M, Rabes JP, Allard D, Ouguerram K, Devillers M,
Cruaud C, Benjannet S, Wickham L, Erlich D, et al. Mutations in
PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet
18. Cohen JC, Boerwinkle E, Mosley TH Jr, Hobbs HH. Sequence varia-
tions in PCSK9, low LDL, and protection against coronary heart dis-
ease. N Engl J Med 2006;354:1264–72.
19. Persson L, Cao G, Stahle L, Sjoberg BG, Troutt JS, Konrad RJ, Galman
C, Wallen H, Eriksson M, Hafstrom I, et al. Circulating proprotein
convertase subtilisin kexin type 9 has a diurnal rhythm synchronous
with cholesterol synthesis and is reduced by fasting in humans. Arte-
rioscler Thromb Vasc Biol 2010;30:2666–72.
20. Fritsche KL. Too much linoleic acid promotes inflammation-doesn’t it?
Prostaglandins Leukot Essent Fatty Acids 2008;79:173–5.
21. Boberg M, Croon LB, Gustafsson IB, Vessby B. Platelet fatty acid
composition in relation to fatty acid composition in plasma and to
serum lipoprotein lipids in healthy subjects with special reference to
the linoleic acid pathway. Clin Sci (Lond) 1985;68:581–7.
22. Kempen HJ, Glatz JF, Gevers Leuven JA, van der Voort HA, Katan
MB. Serum lathosterol concentration is an indicator of whole-body
cholesterol synthesis in humans. J Lipid Res 1988;29:1149–55.
23. Basu S. Radioimmunoassay of 8-iso-prostaglandin F2alpha: an index
for oxidative injury via free radical catalysed lipid peroxidation.
Prostaglandins Leukot Essent Fatty Acids 1998;58:319–25.
24. Basu S. Radioimmunoassay of 15-keto-13,14-dihydro-prostaglandin
F2alpha: an index for inflammation via cyclooxygenase catalysed lipid
peroxidation. Prostaglandins Leukot Essent Fatty Acids 1998;58:347–52.
25. Zock PL, Mensink RP, Harryvan J, de Vries JH, Katan MB. Fatty acids
in serum cholesteryl esters as quantitative biomarkers of dietary intake
in humans. Am J Epidemiol 1997;145:1114–22.
26. Kantartzis K, Thamer C, Peter A, Machann J, Schick F, Schraml C,
Konigsrainer A, Konigsrainer I, Krober S, Niess A, et al. High car-
diorespiratory fitness is an independent predictor of the reduction in
liver fat during a lifestyle intervention in non-alcoholic fatty liver
disease. Gut 2009;58:1281–8.
27. Yki-Ja ¨rvinen H. Nutritional modulation of nonalcoholic fatty liver
disease and insulin resistance: human data. Curr Opin Clin Nutr Metab
28. Spadaro L, Magliocco O, Spampinato D, Piro S, Oliveri C, Alagona C,
Papa G, Rabuazzo AM, Purrello F. Effects of n23 polyunsaturated
fatty acids in subjects with nonalcoholic fatty liver disease. Dig Liver
29. Ailhaud G, Massiera F, Weill P, Legrand P, Alessandri JM, Guesnet P.
Temporal changes in dietary fats: role of n26 polyunsaturated fatty
acids in excessive adipose tissue development and relationship to
obesity. Prog Lipid Res 2006;45:203–36.
30. Horton JD, Shah NA, Warrington JA, Anderson NN, Park SW, Brown
MS, Goldstein JL. Combined analysis of oligonucleotide microarray
data from transgenic and knockout mice identifies direct SREBP target
genes. Proc Natl Acad Sci USA 2003;100:12027–32.
31. Maxwell KN, Soccio RE, Duncan EM, Sehayek E, Breslow JL. Novel
putative SREBP and LXR target genes identified by microarray anal-
ysis in liver of cholesterol-fed mice. J Lipid Res 2003;44:2109–19.
DIETARY FATTY ACIDS AND LIVER FAT
9 of 10
32. Gustavsson C, Parini P, Ostojic J, Cheung L, Hu J, Zadjali F, Tahir F, Download full-text
Brismar K, Norstedt G, Tollet-Egnell P. Cocoa butter and safflower oil
elicit different effects on hepatic gene expression and lipid metabolism
in rats. Lipids 2009;44:1011–27.
33. Sacks FM, Katan M. Randomized clinical trials on the effects of di-
etary fat and carbohydrate on plasma lipoproteins and cardiovascular
disease. Am J Med 2002;113(Suppl 9B):13S–24S.
34. Heine RJ, Mulder C, Popp-Snijders C, van der Meer J, van der Veen
EA. Linoleic-acid-enriched diet: long-term effects on serum lipopro-
tein and apolipoprotein concentrations and insulin sensitivity in non-
insulin-dependent diabetic patients. Am J Clin Nutr 1989;49:448–56.
35. Iggman D, Gustafsson IB, Berglund L, Vessby B, Marckmann P, Riserus
U. Replacing dairy fat with rapeseed oil causes rapid improvement of
hyperlipidaemia: a randomized controlled study. J Intern Med 2011;270:
36. Asp ML, Collene AL, Norris LE, Cole RM, Stout MB, Tang SY, Hsu
JC, Belury MA. Time-dependent effects of safflower oil to improve
glycemia, inflammation and blood lipids in obese, post-menopausal
women with type 2 diabetes: a randomized, double-masked, crossover
study. Clin Nutr 2011;30:443–9.
37. Vessby B, Uusitupa M, Hermansen K, Riccardi G, Rivellese AA,
Tapsell LC, Nalsen C, Berglund L, Louheranta A, Rasmussen BM,
et al. Substituting dietary saturated for monounsaturated fat impairs
insulin sensitivity in healthy men and women: the KANWU Study.
38. Halton TL, Liu S, Manson JE, Hu FB. Low-carbohydrate-diet score
and risk of type 2 diabetes in women. Am J Clin Nutr 2008;87:339–46.
39. Carstensen M, Herder C, Kivimaki M, Jokela M, Roden M, Shipley
MJ, Witte DR, Brunner EJ, Tabak AG. Accelerated increase in serum
interleukin-1 receptor antagonist starts 6 years before diagnosis of type
2 diabetes: Whitehall II prospective cohort study. Diabetes 2010;59:
40. Hu FB, Meigs JB, Li TY, Rifai N, Manson JE. Inflammatory markers and
risk of developing type 2 diabetes in women. Diabetes 2004;53:693–700.
41. Moussavi N, Gavino V, Receveur O. Could the quality of dietary fat,
and not just its quantity, be related to risk of obesity? Obesity (Silver
42. Girard J, Perdereau D, Foufelle F, Prip-Buus C, Ferre P. Regulation of
lipogenic enzyme gene expression by nutrients and hormones. FASEB
43. Wang Y, Jones Voy B, Urs S, Kim S, Soltani-Bejnood M, Quigley N,
Heo YR, Standridge M, Andersen B, Dhar M, et al. The human fatty
acid synthase gene and de novo lipogenesis are coordinately regulated
in human adipose tissue. J Nutr 2004;134:1032–8.
44. Kotronen A, Seppanen-Laakso T, Westerbacka J, Kiviluoto T, Arola J,
Ruskeepaa AL, Oresic M, Yki-Jarvinen H. Hepatic stearoyl-CoA de-
saturase (SCD)-1 activity and diacylglycerol but not ceramide con-
centrations are increased in the nonalcoholic human fatty liver.
45. Peter A, Cegan A, Wagner S, Lehmann R, Stefan N, Konigsrainer A,
Konigsrainer I, Haring HU, Schleicher E. Hepatic lipid composition
and stearoyl-coenzyme A desaturase 1 mRNA expression can be es-
timated from plasma VLDL fatty acid ratios. Clin Chem 2009;55:
46. Attie AD, Krauss RM, Gray-Keller MP, Brownlie A, Miyazaki M,
Kastelein JJ, Lusis AJ, Stalenhoef AF, Stoehr JP, Hayden MR, et al.
Relationship between stearoyl-CoA desaturase activity and plasma
triglycerides in human and mouse hypertriglyceridemia. J Lipid Res
47. Gutie ´rrez-Jua ´rez R, Pocai A, Mulas C, Ono H, Bhanot S, Monia BP,
Rossetti L. Critical role of stearoyl-CoA desaturase-1 (SCD1) in the
onset of diet-induced hepatic insulin resistance. J Clin Invest 2006;116:
48. Bjermo H, Riserus U. Role of hepatic desaturases in obesity-related
metabolic disorders. Curr Opin Clin Nutr Metab Care 2010;13:703–8.
49. Martin PG, Guillou H, Lasserre F, Dejean S, Lan A, Pascussi JM,
Sancristobal M, Legrand P, Besse P, Pineau T. Novel aspects of
PPARalpha-mediated regulation of lipid and xenobiotic metabolism
revealed through a nutrigenomic study. Hepatology 2007;45:767–77.
50. Cho YY, Kwon EY, Kim HJ, Jeon SM, Lee KT, Choi MS. Differential
effect of corn oil-based low trans structured fat on the plasma and
hepatic lipid profile in an atherogenic mouse model: comparison to
hydrogenated trans fat. Lipids Health Dis 2011;10:15.
10 of 10
BJERMO ET AL