Nutrition, Metabolism & Cardiovascular Diseases (2015) 25, 734e741

Available online at www.sciencedirect.com

Nutrition, Metabolism & Cardiovascular Diseases

journal homepage: www.elsevier.com/locate/nmcd

A double-blind, placebo-controlled randomized trial to evaluate the

efficacy of docosahexaenoic acid supplementation on hepatic fat
and associated cardiovascular risk factors in overweight children
with nonalcoholic fatty liver disease
L. Pacifico a, E. Bonci b, M. Di Martino c, P. Versacci a, G. Andreoli a, L.M. Silvestri a,
C. Chiesa d,*
a
Department of Pediatrics and Child Neuropsychiatry, Sapienza University of Rome, 00161 Rome, Italy
b
Department of Experimental Medicine, Sapienza University of Rome, 00161 Rome, Italy
c
Department of Radiological Sciences, Sapienza University of Rome, 00161 Rome, Italy
d
Institute of Translational Pharmacology, National Research Council, Via del Fosso del Cavaliere, 100, 00133 Rome, Italy

Received 5 February 2015; received in revised form 3 April 2015; accepted 16 April 2015
Available online 25 April 2015

KEYWORDS Abstract Background and Aims: Very little information is available on whether docosahexaenoic
Docosahexaenoic acid (DHA) supplementation has a beneficial effect on liver fat and cardiovascular disease (CVD)
acid; risk factors in children with nonalcoholic fatty liver disease (NAFLD). In a double-blind, placebo-
NAFLD; controlled randomized trial we investigated whether 6-month treatment with DHA improves
Liver fat; hepatic fat and other fat depots, and their associated CVD risk factors in children with biopsy-
Visceral adipose proven NAFLD.
tissue; Methods and Results: Of 58 randomized children, 51 (25 DHA, 26 placebo) completed the study.
Epicardial adipose The main outcome was the change in hepatic fat fraction as estimated by magnetic resonance
tissue; imaging. Secondary outcomes were changes in visceral adipose tissue (VAT), epicardial adipose
tissue (EAT), and left ventricular (LV) function, as well as alanine aminotransferase (ALT), triglyc-
Cardiac function;
erides, body mass index-standard deviation score (BMI-SDS), and insulin sensitivity. At 6
Children
months, the liver fat was reduced by 53.4% (95% CI, 33.4e73.4) in the DHA group, as compared
with 22.6% (6.2e39.0) in the placebo group (P Z 0.040 for the comparison between the two
groups). Likewise, in the DHA group VAT and EAT were reduced by 7.8% (0e18.3) and 14.2% (0
e28.2%), as compared with 2.2% (0e8.1) and 1.7% (0e6.8%) in the placebo group, respectively
(P Z 0.01 for both comparisons). There were no significant between-group changes for LV func-
tion as well as BMI-SDS and ALT, while fasting insulin and triglycerides significantly decreased in
the DHA-treated children (P Z 0.028 and P Z 0.041, respectively).
Conclusions: DHA supplementation decreases liver and visceral fat, and ameliorates metabolic
abnormalities in children with NAFLD.
ª 2015 Elsevier B.V. All rights reserved.

* Corresponding author. Tel.: þ39 (0)6 49979215; fax: þ39 (0)6 49979216.
E-mail address: claudio.chiesa@ift.cnr.it (C. Chiesa).

http://dx.doi.org/10.1016/j.numecd.2015.04.003
0939-4753/ª 2015 Elsevier B.V. All rights reserved.
Dietary patterns and lipoprotein subclasses
Introduction
Nonalcoholic fatty liver disease (NAFLD) encompasses a
range of liver histology severity and outcomes in the
absence of chronic alcohol use. The mildest form is sim-
ple steatosis in which triglycerides accumulate within
hepatocytes. A more advanced form of NAFLD, nonalco-
holic steatohepatitis (NASH), includes inflammation and
liver cell injury, progressive to cryptogenic cirrhosis.
NAFLD has become the most common cause of chronic
liver disease in children and adolescents [1]. The pres-
ence of NAFLD also has important implications regarding
cardiovascular health. As described in adults, children
and adolescents with fatty liver display metabolic ab-
normalities that are risk factors for cardiovascular dis-
ease(CVD), such as insulin resistance, glucose intolerance,
and dyslipidemia [1,2]. Several studies, including pedi-
atric subjects, have reported independent associations
between NAFLD and markers of subclinical atheroscle-
rosis, such as impaired flow-mediated vasodilation,
increased carotid artery intima-media thickness and
arterial stiffness, after adjusting for CVD risk factors and
metabolic syndrome (MetS) [3,4]. Thus NAFLD has
emerged as the hepatic component of MetS and a strong
risk factor for the development of atherosclerotic disease,
even at a very early age [5,6]. More recent work has
identified NAFLD as a risk factor also for myocardial in-
sulin resistance, altered cardiac energy metabolism,
abnormal left ventricular (LV) structure, and impaired
diastolic function [7e9].
Currently, only weight loss and increased physical ac-
tivity decrease hepatic fat, and to date there are no
evidenced-based guidelines, and no approved pharma-
cologic therapy for the treatment of NAFLD in children [1].
Growing evidence suggests that n-3 long-chain poly-
unsaturated fatty acids (LC-PUFAs) may have a beneficial
role on many cardio-metabolic risk factors, including
NAFLD [10,11]. In adults, randomized and non-
randomized clinical trials have shown that the supple-
mentation of n-3 LC-PUFAs, including both eicosa-
pentaenoic acid (EPA) and docosahexaenoic acid (DHA),
leads to amelioration of liver fatness [12,13]. In children,
Nobili et al. have recently reported that DHA supple-
mentation improves liver steatosis and is able to reduce
the levels of serum alanine aminotransferase (ALT) and
triglycerides, and to improve insulin sensitivity [14,15].
However, very little information is available on whether
DHA treatment has a significant benefit on cardiovascular
health in children with NAFLD. In a double-blind, parallel-
group, placebo-controlled randomized trial we investi-
gated in children and adolescents with NAFLD, the impact
of 6-month n-3 LC-PUFAs supplementation on hepatic fat
and other fat depots [i.e. abdominal visceral adipose tissue
(VAT), abdominal subcutaneous adipose tissue (SAT), and
epicardial adipose tissue (EAT)], and their associated CVD
risk factors including LV dysfunction. We focused only on
the effects of DHA supplementation, which has been
suggested to have greater benefit for CVD risk factors than
EPA [16].
735
Methods
The study protocol was approved by the local Ethics
Committee (Policlinico Umberto I Hospital, Rome, Italy)
(2485/24.05.2012) and written consent was obtained from
the next of kin, caretakers, or guardians on behalf of the
children enrolled in this study, in accordance with prin-
ciples of the Helsinki Declaration.
Study population
Patients were eligible for the study if they had at enroll-
ment: 1) age <18 years; 2) body mass index (BMI) > 85th
percentile according to age- and gender-specific percen-
tiles of BMI [17]; 3) persistently elevated aminotransferase
levels; 4) magnetic resonance imaging (MRI)-diagnosed
NAFLD [hepatic fat fraction (HFF)  5%] [18]; and 5) liver
biopsy consistent with NAFLD. Secondary causes of stea-
tosis including hepatic virus infections (hepatitis AeE and
G, cytomegalovirus, and EpsteineBarr virus), autoimmune
hepatitis, metabolic liver disease, a-1-antitrypsin defi-
ciency, cystic fibrosis, Wilson’s disease, hemochromatosis,
and celiac disease were excluded after appropriate tests.
Other exclusion criteria were smoking, and history of type
1 or type 2 diabetes, renal disease, total parenteral nutri-
tion, alcohol intake, use of hepatotoxic medications, and
previous use of n-3 LC-PUFAs.
Study design
The study was a double-blind, parallel-group, placebo-
controlled randomized trial performed at the Hepatology
outpatient Clinic of the Department of Pediatrics, Sapienza
University of Rome, Italy, between May 2012 and
September 2014. Participants were randomly assigned to
DHA supplementation [250 mg/day (39% DHA algae oil;
Dietetic Metabolic Food e DMF, Limbiate, MB, Italy)] or
placebo (290 mg linoleic acid supplied with germ oil; IBSA,
Lodi, Italy). A randomization list (in a 1:1 ratio to treat-
ment with DHA or placebo) was generated by an inde-
pendent statistician who was blinded to participants’
clinical data and did not perform the final analysis. DHA
and placebo pills were of similar appearance and taste and
provided about 7 kcal of energy (DMF, Limbiate, MB, Italy).
Pills were stored at the hospital pharmacy and dispensed
at the baseline visit and every month thereafter. All par-
ticipants and research staff were blind to the group
assignment. Compliance to treatment was encouraged by
weekly phone calls and text messages and monitored by
pill count and direct interview at every monthly visit.
Blood DHA concentrations before and after treatment were
used as an objective measure of compliance. Adverse
events were defined as those injuries related to or caused
by the treatments under study. At each visit, parents were
specifically asked about adverse events, and the first
author checked for any association between the adverse
events and morbidity. A balanced low-calorie diet was
736
prescribed to all patients with a recommendation to
engage in a moderate daily exercise program (60 min/day
at least 5 days a week), and to reduce sedentary activities.
Specifically, diet was hypocaloric (25e30 calories/kg/day),
consisting of carbohydrate (50%e60%); protein (15%e20%);
and fat (23%e30%), with a composition of two in third
unsaturated and one in third saturated; the u6/u3 ratio
was approximately 4:1 as recommended by the Italian
Recommended Dietary Allowances [18].
Abdominal MRI
In this study, we used a previously described and validated
MRI technique that has been demonstrated to correlate
well with histology-determined hepatic steatosis [18].
Measurement of HFF was performed by drawing three
different regions of interest (ROIs) (diameter 1e2 cm2; 2 in
the right hepatic lobe and 1 in the left hepatic lobe) at
three different sections of the liver (above, at the level of,
and below the porta hepatis). These ROIs were carefully
drawn in order to avoid vascular structures, motion arte-
facts, and partial volume effects. ROIs were placed at
anatomically matched locations on paired images by using
a co-registration tool available on the picture archiving
and communication system workstation.
The three-point chemical-shiftefatewater separation
method (fat-only dataset) was used to measure VAT and
SAT [9]. MRI results were interpreted by an experienced
radiologist who was blinded to the clinical, laboratory and/
or histologic findings.
Liver biopsy
Percutaneous needle liver biopsy was performed as pre-
viously described [19]. The main histologic features of
NAFLD were scored according to the scoring system
developed by the NASH Clinical Research Network (CNR)
[20]. Features of steatosis, lobular inflammation, and he-
patocyte ballooning were combined to obtain the NAFLD
activity score. As recommended by a recent NASH CRN
article [21], a microscopic diagnosis that is based on
overall injury pattern (i.e. steatosis, hepatocellular
ballooning, and inflammation) as well as the presence of
additional lesions such as zonality of lesions, portal
inflammation, and fibrosis, was assigned to each case.
Accordingly, biopsies were subdivided into not-NASH and
definite-NASH subcategories.
Clinical and laboratory data
All participants underwent physical examination including
weight measurement, standing height, BMI, waist
circumference (WC), determination of pubertal status, and
systolic and diastolic blood pressure, as reported in detail
previously [6]. The degree of obesity was quantified using
Cole’s least mean-square method, which normalizes the
skewed distribution of BMI and expresses BMI as standard
deviation score (SDS) [17].
L. Pacifico et al.
Blood samples were taken from each subject, after an
overnight fast, for estimation of glucose, insulin, total
cholesterol, high-density lipoprotein cholesterol (HDL-C),
triglycerides, ALT, high-sensitivity C reactive protein
(HSCRP), and fatty acids. An oral glucose tolerance test
(OGTT) was performed for all obese children with the
administration of glucose at 1.75 g per kg of body weight
(maximum dose 75 g); blood samples were obtained at
0 min and every 30 min thereafter for 120 min for deter-
mination of serum glucose and insulin. Estimates of insulin
sensitivity were calculated using the homeostasis model
assessment of insulin resistance (HOMA-IR), defined by
fasting insulin and fasting glucose and whole-body insulin
sensitivity index (WBISI), based on mean values of insulin
and glucose obtained from OGTT and the corresponding
fasting values. All analyses were conducted by COBAS 6000
(Roche Diagnostics). While insulin concentrations were
measured on cobas e 601 module (Electro-
chemiluminescence Technology, Roche Diagnostics), the
remaining analytes on cobas e 501 clinical chemistry
module (Photometric Technology). Blood fatty acids,
including DHA, were analyzed in a drop of whole blood
absorbed on a strip and transmethylated for gas-
chromatography [22].
Echocardiographic parameters
Echocardiography was performed with a commercially
available echocardiographic system (SONOS 5500, Phillips,
Andover, Massachusetts, USA). As described in detail
elsewhere [9], LV function was evaluated using the
following pulse-wave Doppler echocardiographic param-
eters: early (E) and late (A) mitral velocity, and decelera-
tion time. Tissue Doppler imaging echocardiography of the
septal and lateral mitral annulus was used to measure the
early (e0 ) and late (a0 ) annular diastolic and systolic (s0 )
tissue velocities, and the mean values of septal and lateral
annulus measurements were used for analysis. The ratios
E/e0 and e0 /a0 were also calculated. The Doppler-derived
index that combines systolic and diastolic myocardial
performance (Tei index), defined as the sum of iso-
volumetric relaxation time and isovolumetric contraction
time divided by ejection time was used for quantification
of the global LV function [9]. EAT thickness was measured
during end-systole at the point on the free wall of the right
ventricle along the midline of the ultrasound beam, as
closely as possible perpendicular to the aortic annulus,
used as an anatomic landmark [9].
Statistical analysis
Statistical analyses were performed using the SPSS pack-
age (version 22.0), SPSS Inc., Chicago, IL, USA. The primary
endpoint used for the sample size determination was the
change in liver fat %. A sample size of 44 subjects (22 per
group) would provide 80% power to detect a 20% change in
liver fat with an estimated value for sigma of 3.5, and an
alpha of 0.05 (two tailed test). Secondary outcomes were
the changes in VAT, EAT and LV function, as well as in
Dietary patterns and lipoprotein subclasses
insulin sensitivity, ALT, triglycerides and BMI-SDS. Data are
reported as means and standard deviations for normally
distributed variables, or as median and interquartile range
for non-normally distributed variables. Baseline differ-
ences between the two study groups were evaluated by t-
test or ManneWhitney U-test, as appropriate. Proportions
were compared by the chi square test. Changes from
baseline to 6 months were compared by paired t-test or
Wilcoxon’s rank sum test within each group, and by t-test
or ManneWhitney U-test between groups. A P < 0.05 was
considered significant.
Results
Patients characteristics
We evaluated 118 children with suspected NAFLD at our
outpatient Clinic. Forty-four did not meet the inclusion
criteria: 7 had viral infections, 3 celiac disease, 4 predia-
betes; in 6 children liver enzymes were not elevated, in 5
subjects MRI images were not of good quality while in 5
MRI HFF was <5%; and 14 did not give consent to liver
biopsy. Sixteen patients did not give consent to randomi-
zation. Thus, a total of 58 children were randomized to
treatment with DHA (n Z 29) or placebo (n Z 29) for 6
months (Fig. 1). The study completion rates were high
(>85%) for both randomized groups. Of the 29 patients
randomized to each group, four patients from the DHA and
three patients from the placebo groups were excluded
from the efficacy evaluable set. The principal reasons were
Figure 1 Algorithm of the study.
737
low compliance and lack of MRI at follow-up. There were
no adverse events in either group. The final analysis was
performed with a total of fifty-one participants who
received DHA (n Z 25) or placebo (n Z 26). The two
groups were similar with respect to baseline clinical,
metabolic, and echocardiographic characteristics (Tables
1A and 1B). The only exception was HDL-C, which by
chance was “significantly” lower in the DHA group. Ran-
domized groups also did not differ in baseline histological
features of NAFLD (Table 2). There was evidence to suggest
high compliance with DHA supplementation. Blood DHA
significantly increased in the DHA-supplemented group as
compared to baseline or placebo treatment (Table 1A and
Supplemental Table 1).
Effects of DHA on fatty liver (main outcome) and other fat
depots
As shown in Table 1B, HFF significantly decreased from
baseline to 6 months in the DHA-treated patients (from
14.0% to 6.5%; P Z 0.036) while in the placebo-treated
subjects HFF did not change significantly (from 15.5% to
12.0%; P Z 0.39). The liver fat was reduced by 53.4% (95%
CI, 33.4e73.4) in the DHA group, as compared with 22.6%
(6.2e39.0) in the placebo group (P Z 0.040 for the com-
parison between the two groups). Likewise, VAT and EAT
significantly decreased from baseline to 6 months in the
DHA-treated group, whereas they remained unchanged in
the placebo group (Table 1B). In the DHA group, VAT and
EAT were reduced by 7.8% (0e18.3) and 14.2% (0e28.2%),
as compared with 2.2% (0e8.1) and 1.7% (0e6.8%) in the
738 L. Pacifico et al.

Table 1A Baseline clinical and biochemical characteristics of the subjects and outcome at 6 months.

Placebo DHA
Variables Baseline End of study P value* Baseline End of study P value* P valueþ
Age, years 10.8 (2.8) 11.0 (2.6) 0.74
Male gender 16 (61.5) 14 (56.0) 0.57
Weight, kg 63 (20) 64 (19) 0.69 66 (18) 63 (17) 0.22 0.041
Height, cm 149 (13) 153 (11) 0.01 149 (13) 151 (14) 0.25 0.049
BMI 27.5 (5.5) 27.2 (5.4) 0.71 28.9 (4.3) 27.3 (4.1) 0.024 0.16
BMI-SDS 2.07 (0.35) 1.87 (0.51) 0.039 2.15 (0.32) 1.97 (0.42) 0.005 0.62
Waist circumference, cm 91 (12) 92 (14) 0.68 94 (12) 91(11) 0.056 0.06
Systolic blood pressure, mmHg 115 (12) 114 (10) 0.54 115 (14) 115 (13) 0.69 0.44
Diastolic blood pressure, mmHg 66 (10) 66 (9) 0.57 68 (10) 66 (7) 0.14 0.48
ALT, U/l 56 (19) 45 (22) 0.27 57 (20) 27 (14) 0.004 0.06
Total cholesterol, mg/dl 150 (43) 147 (30) 0.46 152 (33) 158 (40) 0.54 0.34
HDL-C, mg/dl 47 (9) 50 (11) 0.22 41 (10) 43 (9) 0.41 0.69
Triglycerides, mg/dl 85 (55e126) 75 (48e109) 0.56 92 (58e143) 75 (49e118) 0.045 0.041
Glucose, mg/dl 84 (7) 84 (6) 0.94 83 (7) 82 (6) 0.88 0.88
Insulin, mU/ml 18 (12e21) 15 (8e21) 0.28 20 (13e24) 11 (10e15) 0.005 0.028
HOMA-IR 3.27 (2.0e4.4) 2.97 (1.7e4.4) 0.19 4.0 (2.4e4.9) 2.50 (2.0e3.1) 0.028 0.32
WBISI 4.8 (2.8) 4.2 (2.5) 0.79 4.6 (2.9) 4.9 (3.7e6.6) 0.45 0.44
HSCRP, mg/l 2000 (999e4000) 1450 (475e3650) 0.22 2700 (990e5100) 2800 (800e5100) 0.95 0.73
DHA, %{ 2.08 (0.65) 2.23 (0.53) 0.87 2.02 (0.82) 3.29 (0.74) <0.0001 <0.0001
Results are expressed as n (%), mean (SD), or median (IQR).
*P values indicate comparison within groups; þP values indicate comparison of the changes of each variable between the 2 groups; {% of total
fatty acids.
DHA, docosahexaenoic acid; BMI, body mass index; BMI-SDS, BMI-standard deviation score; ALT, alanine aminotransferase; HDL-C, high-density
lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; WBISI, whole-body insulin sensitivity index; HSCRP,
high-sensitivity C reactive protein.

placebo group, respectively (P Z 0.01 for both compari-
sons). No statistically significant within- and between-
group changes in SAT were detected.
Effects of DHA on anthropometric, biochemical, and
echocardiographic variables
No significant between-group changes in SDS of BMI were
found (Table 1A). There was a trend toward a statistically
significant decrease in WC in the DHA-supplemented
group when compared with placebo treatment
(P Z 0.06). Serum triglycerides and fasting insulin signif-
icantly decreased in the DHA group with respect to pla-
cebo group (P Z 0.041 and P Z 0.028, respectively). The
effects of DHA treatment on HOMA-IR, WBISI, and HSCRP
were not significant when compared to placebo treatment,
while ALT changes reached borderline significance
(P Z 0.06).
Between-group comparisons showed that there were
no significant changes for parameters of systolic(s0 ) and
diastolic (e0 , e0 /a0 ratio, E/e0 ratio) LV function, and the Tei
index (Table 1B). Nonetheless, compared to placebo, in the

Table 1B Baseline liver fat content, abdominal adipose tissue, epicardial fat and left ventricular function of the subjects, and outcome at 6
months.

Placebo DHA
Variables Baseline End of study P value* Baseline End of study P value* P valueþ
HFF, % 15.5 (8.0e23.0) 12.0 (5.0e25.0) 0.39 14.0 (8.0e19.0) 6.5 (3.2e15.5) 0.036 0.04
VAT, cm2 445 (335e689) 435 (315e659) 0.40 450 (337e679) 415 (295e559) 0.031 0.01
SAT, cm2 1830 (1650e2460) 1800 (1630e2350) 0.54 1820 (1630e2475) 1750 (1620e2100) 0.45 0.77
Epicardial fat, mm 11.5 (8.6e12.9) 11.3 (9.8e14.0) 0.30 12.1 (10.3e13.5) 10.3 (8.6e12.1) 0.029 0.01
s0 velocity, cm/s 7.96 (6.34e8.97) 7.63 (6.98e8.56) 0.27 7.81(6.15e8.79) 8.20 (6.67e9.97) 0.26 0.87
e0 velocity, cm/s 14.7 (12.3e18.0) 15.0 (13.1e18.0) 0.89 14.6 (12.4e15.6) 15.1 (12.3e15.4) 0.60 0.79
a0 velocity, cm/s 6.10 (5.11e7.32) 6.02 (5.01e7.35) 0.75 6.07 (5.15e7.30) 6.04 (5.12e7.31) 0.65 0.82
e0 /a0 2.60 (2.30e2.85) 2.82 (2.10e3.05) 0.68 2.56 (2.31e2.99) 2.93 (2.60e3.19) 0.55 0.75
E/e0 ratio 7.34 (5.73e7.93) 6.94 (5.31e7.45) 0.22 7.36 (6.26e8.31) 6.76 (5.59e8.04) 0.48 0.79
Tei index 0.38 (0.35e0.39) 0.40 (0.38e0.43) 0.10 0.39 (0.34e0.44) 0.36 (0.34.0.40) 0.60 0.053
Results are expressed as mean (SD), or median (IQR).
*P values indicate comparison within groups; þP values indicate comparison of the changes of each variable between the 2 groups.
DHA, docosahexaenoic acid; HFF, hepatic fat fraction; VAT, visceral adipose tissue; SAT, subcutaneous adipose tissue; s0 , systolic tissue velocity; e0 ,
early diastolic tissue velocity; a0 , late diastolic tissue velocity; e0 /a0 , early diastolic tissue velocity/late diastolic tissue velocity; E/e0 , early mitral
velocity/early diastolic tissue velocity; Tei index, myocardial performance index.
The bold values indicates HFF % is the primary endpoint.
Dietary patterns and lipoprotein subclasses 739

Table 2 Histopathological features of the children randomized to

small amounts of fatty liver infiltration [24]. This confusing
placebo or DHA. evidence led the authors of a recent systematic review and
meta-analysis to conclude that “well designed randomized
Histopathological features Placebo group DHA group
controlled trials which quantify the magnitude of effect of
(n Z 26) (n Z 25)
omega-3 PUFA supplementation on liver fat are needed”
Steatosis, n (%)
[12]. In consequence, we selected MRI to measure liver fat
1 7 (26.9) 7 (28.0)
2 9 (34.6) 8 (32.0) (the primary outcome) because it is noninvasive, uses no
3 10 (38.5) 10 (40.0) ionizing radiation, provides objective and quantitative es-
Lobular inflammation, n (%) timates of fat content throughout the entire liver, is more
0 2 (7.7) 1 (4.0) objective than ultrasound, is more practical for general use
1 10 (38.5) 12 (48.0)
and provides greater spatial coverage than MR spectros-
2 10 (38.5) 8 (32.0)
3 4 (15.4) 4 (16.0) copy (which requires a skilled operator to correctly
Ballooning, n (%) perform the examination, process the data, and interpret
0 4 (15.4) 3 (12.0) the results) [25].
1 17 (65.4) 16 (64.0) In this longitudinal study, we have shown that DHA
2 5 (19.2) 6 (24.0)
significantly decreases the MRI-determined liver fat con-
Fibrosis, n (%)
0 5 (19.2) 6 (24.0) tent independently of BMI-SDS changes. Moreover, this
1 10 (38.5) 10 (40.0) study highlights the beneficial effects of 6-month DHA on
2 9 (34.6) 7 (28.0) other fat depots. An important result is the positive effect
3 2 (7.7) 2 (8.0) on the reduction of abdominal visceral fat in the DHA
NAS score, mean (SD) 4.6 (0.5) 4.4 (0.6)
group, with a significant decrease in VAT and a trend to-
NASH, n (%) 17 (65.4) 16 (64.0)
ward a statistically significant decrease in WC. The
DHA, docosahexaenoic acid; NAS, nonalcoholic fatty liver disease reduction in central body fat is of importance, as this
activity score; SD, standard deviation; NASH, nonalcoholic steato-
hepatitis.
better demonstrates abdominal obesity than BMI [26].
Excessive visceral adipose tissue increases the portal free
fatty acids load flowing to the liver, and this provides an
impetus for hepatic fat accumulation [27]. In the DHA-
DHA group there was a trend toward an improvement in
treated children, there was also a significant decrease in
the Tei index that reflects the combined systolic and dia-
EAT thickness. As recently noted, among other visceral fat
stolic function.
depots, epicardial fat is the highest source of free fatty
acids [28,29]. Certainly epicardial fat and fatty liver share
Discussion similar biochemical properties with the intra-abdominal
visceral fat. Cardiac and hepatic fat are associated with
In this randomized double-blind, placebo-controlled trial insulin resistance and lipotoxicity [30]. Our data show that
we evaluated the efficacy of DHA for 6 months on a subjects with NAFLD had significant reduction in insulin
quantitative measurement of liver fat and a range of and triglyceride levels, a reversal of the metabolic milieu
cardio-metabolic risk factors in pediatric NAFLD. permissive of fat depots.
In the last decade, there has been growing interest in a The effects of DHA supplementation in our study
potential role for omega-3 FA treatment in subjects with confirm the results of previous studies in children. Nobili
NAFLD, and several biological mechanisms have been et al. reported short-term (6 months) [14] and long-term
suggested, proposing for a benefit of this treatment [10,11]. (up to 24 months) [15] effects of DHA, after 6, 12, 18, and
The beneficial effects of n-3 FA in NAFLD may be related to 24 months of treatment with different concentrations
their action in regulating hepatic lipid metabolism, adi- (DHA 250 mg/day and 500 mg/day) combined with diet
pose tissue function, and inflammation [10,11]. To date and exercise. In these studies, algae DHA supplementation
there have been a few randomized trials investigating the improved liver steatosis (as detected by ultrasound) and
effects of omega-3 LC PUFA in adults and children with was able to improve insulin sensitivity and to reduce the
NAFLD [12e15,23]. Notably, the limitations of these studies levels of serum triglycerides at 6 months. ALT decreased in
were the nonblinding of participants and investigators, the the DHA-treated groups from month 12 onwards.
lack of a placebo control group, the lack of testing for There is evidence that the fatty acid milieu predicts
adherence to the omega-3 intervention, and the semi- structural and functional changes in the heart that occur
quantitative or non-specific measures of NAFLD severity, with obesity. Dietary supplementation with n-3 PUFA at-
including ultrasound and ALT. Most of these trials have tenuates pathologic cardiac remodeling in response to
used ultrasound because of its availability, low cost, and pressure overload [31], myocardial ischemia [32], and
minimal risk to the patients [24]. It measures the fat treatment with rosiglitazone [33]. Accordingly we sought
content of the liver indirectly by assessing the liver texture to investigate the contribution of DHA supplementation to
and echogenicity. However, ultrasound is both machine- early cardiac outcomes such as LV dysfunction. In the
and operator-dependent and lacks the ability to accurately present study, similar echocardiographic findings of
quantify the liver fat content. In addition, its sensitivity is myocardial function were observed in the DHA- and
reduced in morbidly obese subjects and in those with placebo-treated children. It is possible that this may be
740
due to the relatively short follow-up. We suggest that
treatment with DHA needs to be tested for longer than 6
months if improvement in cardiac function is secondary to
improvements in liver and other visceral fat depots.
The main strengths of our study are the parallel-group,
double-blind, randomized controlled trial design; the
careful clinical characterization of the participants with
their risk factors; objective evidence of good compliance;
objective, validated outcomes such as MRI-determined
hepatic and visceral fat; and the meticulous conduct of
the echocardiographic procedures. Despite these
strengths, our study has limitations. First, the small sample
size may limit the interpretation of the results, but could
provide indications for further research. Second, DHA
supplements were given for only six months. However, 6-
month duration of treatment is frequently applied in
clinical trials with dietetic supplements and has been
considered sufficient to achieve therapeutic results and
maintain satisfactory compliance [12,34]. Third, although
all children who were recruited had a baseline diagnostic
liver biopsy, a further biopsy as a research test for moni-
toring NAFLD was neither feasible nor ethical at the end of
the study. Finally, a potential confounding factor was the
use as placebo of germ oil, high in omega 6 fatty acids.
In conclusion, DHA supplementation in children and
adolescents with NAFLD decreases liver and visceral fat,
and ameliorates metabolic abnormalities. Based on our
results and on those of previous studies, we suggest that
management of children with NAFLD might consider
incorporating DHA as part of a multifaceted approach.
However, larger clinical trials of longer duration are
needed to determine the long-term benefits of omega-3
supplementation for cardiovascular risk factors in these
patients as well as to establish optimal doses for study
outcomes.
Acknowledgments
Declaration of personal interests: None.
Declaration of funding interests: This study was funded
by the Sapienza University of Rome (Progetti di Ricerca
Universitaria 2011e2012).
Appendix B. Supplementary material
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.numecd.2015.04.003.
References
[1] Mencin AA, Lavine JE. Advances in pediatric nonalcoholic fatty
liver disease. Pediatr Clin North Am 2011;58:1375e92.
[2] D’Adamo E, Cali AM, Weiss R, Santoro N, Pierpont B, Northrup V,
et al. Central role of fatty liver in the pathogenesis of insulin
resistance in obese adolescents. Diabetes Care 2010;33:1817e22.
[3] Pacifico L, Chiesa C, Anania C, De Merulis A, Osborn JF,
Romaggioli S, et al. Nonalcoholic fatty liver disease and the heart
in children and adolescents. World J Gastroenterol 2014;20:
9055e71.
L. Pacifico et al.
[4] Targher G, Day CP, Bonora E. Risk of cardiovascular disease in
patients with nonalcoholic fatty liver disease. N Engl J Med 2010;
363:1341e50.
[5] Schwimmer JB, Pardee PE, Lavine JE, Blumkin AK, Cook S. Car-
diovascular risk factors and the metabolic syndrome in pediatric
nonalcoholic fatty liver disease. Circulation 2008;118:277e83.
[6] Pacifico L, Anania C, Martino F, Cantisani V, Pascone R,
Marcantonio A, et al. Functional and morphological vascular
changes in pediatric nonalcoholic fatty liver disease. Hepatology
2010;52:1643e51.
[7] Perseghin G, Lattuada G, De Cobelli F, Esposito A, Belloni E, Ntali G,
et al. Increased mediastinal fat and impaired left ventricular en-
ergy metabolism in young men with newly found fatty liver.
Hepatology 2008;47:51e8.
[8] Goland S, Shimoni S, Zornitzki T, Knobler H, Azoulai O, Lutaty G,
et al. Cardiac abnormalities as a new manifestation of nonalco-
holic fatty liver disease: echocardiographic and tissue Doppler
imaging assessment. J Clin Gastroenterol 2006;40:949e55.
[9] Pacifico L, Di Martino M, De Merulis A, Bezzi M, Osborn JF,
Catalano C, et al. Left ventricular dysfunction in obese children and
adolescents with nonalcoholic fatty liver disease. Hepatology
2014;59:461e70.
[10] Lorente-Cebrián S, Costa AG, Navas-Carretero S, Zabala M,
Martínez JA, Moreno-Aliaga MJ. Role of omega-3 fatty acids in
obesity, metabolic syndrome, and cardiovascular diseases: a re-
view of the evidence. J Physiol Biochem 2013;69:633e51.
[11] Pacifico L, Giansanti S, Gallozzi A, Chiesa C. Long chain omega-3
polyunsaturated fatty acids in pediatric metabolic syndrome.
Mini Rev Med Chem 2014;14:791e804.
[12] Parker HM, Johnson NA, Burdon CA, Cohn JS, O’Connor HT,
George J. Omega-3 supplementation and non-alcoholic fatty liver
disease: a systematic review and meta-analysis. J Hepatol 2012;
56:944e51.
[13] Scorletti E, Bhatia L, McCormick KG, Clough GF, Nash K, Hodson L,
et al., on behalf of the WELCOME Study Investigators. Effects of
purified eicosapentaenoic and docosahexaenoic acids in non-
alcoholic fatty liver disease: results from the *WELCOME study.
Hepatology 2014;60:1211e21.
[14] Nobili V, Bedogni G, Alisi A, Pietrobattista A, Risé P, Galli C, et al.
Docosahexaenoic acid supplementation decreases liver fat content
in children with non-alcoholic fatty liver disease: double-blind
randomized controlled clinical trial. Arch Dis Child 2011;96:
350e3.
[15] Nobili V, Alisi A, Della Corte C, Risé P, Galli C, Agostoni C, et al.
Docosahexaenoic acid for the treatment of fatty liver: randomized
controlled trial in children. Nutr Metab Cardiovasc Dis 2013;23:
1066e70.
[16] Cottin SC, Sanders TA, Hall WL. The differential effects of EPA and
DHA on cardiovascular risk factors. Proc Nutr Soc 2011;70:215e31.
[17] Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard
definition for child overweight and obesity worldwide: interna-
tional survey. BMJ 2000;320:1240e3.
[18] Pacifico L, Arca M, Anania C, Cantisani V, Di Martino M, Chiesa C.
Arterial function and structure after a 1-year lifestyle intervention
in children with nonalcoholic fatty liver disease. Nutr Metab Car-
diovasc Dis 2013;23:1010e6.
[19] Pacifico L, Di Martino M, Catalano C, Panebianco V, Bezzi M,
Anania C, et al. T1-weighted dual-echo MRI for fat quantification
in pediatric nonalcoholic fatty liver disease. World J Gastroenterol
2011;17:3012e9.
[20] Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ,
Cummings OW, et al. Design and validation of a histological
scoring system for nonalcoholic fatty liver disease. Hepatology
2005;41:1313e21.
[21] Brunt EM, Kleiner DE, Wilson LA, Belt P, Neuschwander-Tetri BA.
NASH Clinical Research Network (CRN). Nonalcoholic fatty liver
disease (NAFLD) activity score and the histopathologic diagnosis
in NAFLD: distinct clinicopathologic meanings. Hepatology 2011;
53:810e20.
[22] Marangoni F, Colombo C, Galli C. A method for the direct evalu-
ation of the fatty acid status in a drop of blood from a fingertip in
humans: applicability to nutritional and epidemiological studies.
Anal Biochem 2004;326:267e72.
[23] Nobili V, Carpino G, Alisi A, De Vito R, Franchitto A, Alpini G, et al.
Role of docosahexaenoic acid treatment in improving liver
Dietary patterns and lipoprotein subclasses
histology in pediatric nonalcoholic fatty liver disease. PLoS One
2014;9:e88005.
[24] Mishra P, Younossi ZM. Abdominal ultrasound for diagnosis of
nonalcoholic fatty liver disease (NAFLD). Am J Gastroenterol 2007;
102:2716e7.
[25] Le TA, Chen J, Changchien C, Peterson MR, Kono Y, Patton H, et al.,
for San Diego Integrated NAFLD Research Consortium (SINC). Ef-
fect of colesevelam on liver fat quantified by magnetic resonance
in nonalcoholic steatohepatitis: a randomized controlled trial.
Hepatology 2012;56:922e32.
[26] Daniels SR, Khoury PR, Morrison JA. Utility of different measures
of body fat distribution in children and adolescents. Am J Epi-
demiol 2000;152:1179e84.
[27] Nielsen S, Guo Z, Johnson CM, Hensrud DD, Jensen MD. Splanchnic
lipolysis in human obesity. J Clin Invest 2004;113:1582e8.
[28] Iacobellis G, Bianco AC. Epicardial adipose tissue: emerging
physiological, pathophysiological and clinical features. Trends
Endocrinol Metab 2011;22:450e7.
[29] Iacobellis G, Barbarini G, Letizia C, Barbaro G. Epicardial fat
thickness and nonalcoholic fatty liver disease in obese subjects.
Obesity (Silver Spring) 2014;22:332e6.
741
[30] Kankaanpää M, Lehto HR, Pärkkä JP, Komu M, Viljanen A,
Ferrannini E, et al. Myocardial triglyceride content and epicardial
fat mass in human obesity: relationship to left ventricular function
and serum free fatty acid levels. J Clin Endocrinol Metab 2006;91:
4689e95.
[31] Duda MK, O’Shea KM, Tintinu A, Xu W, Khairallah RJ, Barrows BR,
et al. Fish oil, but not flaxseed oil, decreases inflammation and
prevents pressure overload induced cardiac dysfunction. Car-
diovasc Res 2008;81:319e27.
[32] Fang Y, Favre J, Vercauteren M, Laillet B, Remy-Jouet I, Skiba M,
et al. Reduced cardiac remodelling and prevention of glutathione
deficiency after omega-3 supplementation in chronic heart failure.
Fundam Clin Pharmacol 2011;25:323e32.
[33] Halade GV, Williams PJ, Lindsey ML, Fernandes G. Fish oil de-
creases inflammation and reduces cardiac remodeling in rosigli-
tazone treated aging mice. Pharmacol Res 2011; 63: 300e7.
[34] Musso G, Gambino R, Cassader M, Pagano G. A meta-analysis of
randomized trials for the treatment of nonalcoholic fatty liver
disease. Hepatology 2010;52:79e104.