bs_bs_banner

Hepatology Research 2017; 47: 622–631 doi: 10.1111/hepr.12781

Original Article

Effect of resistance training on liver fat and visceral adiposity

in adults with obesity: A randomized controlled trial
Shelley E. Keating,1,2 Daniel A. Hackett,1 Helen M. Parker,1 Kimberley L. Way,1
Helen T. O’Connor,1 Amanda Sainsbury,3 Michael K. Baker,4 Vivienne H. Chuter,5
Ian D. Caterson,3 Jacob George6 and Nathan A. Johnson1,3
1
Discipline of Exercise and Sport Science, 3Boden Institute of Obesity, Nutrition, Exercise and Eating Disorders, Sydney
Medical School and 6Storr Liver Centre, Westmead Millennium Institute and Westmead Hospital, University of Sydney,
and 4School of Exercise Science, Australian Catholic University, Sydney, New South Wales, and 2Centre for Research
on Exercise, Physical Activity and Health, School of Human Movement and Nutrition Sciences, The University of
Queensland, Brisbane, Queensland and 5School of Health Sciences, University of Newcastle, Newcastle, New South
Wales, Australia

Aim: Regular aerobic exercise reduces visceral adipose tissue
(VAT) and liver fat, however, not all individuals are able to adopt
and adhere to such programs. Progressive resistance training
(PRT) may be an alternative therapy, but there is limited available
evidence. We examined the efficacy of PRT as per current
exercise guidelines, compared with sham exercise placebo on
liver fat and VAT.
Methods: Twenty-nine inactive and overweight/obese (body
mass index ≥25 kg/m2) adults (age 29–59) were randomized to
receive 8 weeks of PRT (n = 15, 10 exercises per session, 8–12
repetitions, 2–3 sets per exercise at 80–85% of one-repetition
maximum, 3 days per week) or a sham exercise placebo control
(CON) (n = 14). Change in liver fat, VAT, and abdominal s.c.
adipose tissue (SAT) were assessed by magnetic resonance
spectroscopy and imaging).
Results: There were no significant group by time interactions
for change in liver fat in PRT versus CON groups ( 0.07 ± 0.31%
vs. 0.55 ± 0.77%, respectively, P = 0.19), VAT ( 175 ± 85 cm3 vs.
10 ± 64 cm3, respectively, P = 0.11), or abdominal SAT ( 436
± 245 cm3 vs. 127.29 ± 182 cm3, respectively, P = 0.10) despite a
significant increase in muscle volume (55 ± 78 cm3 vs. 0.04
± 8 cm3, respectively, P = 0.03).
Conclusion: Traditional PRT is not effective for reducing liver
fat in overweight/obese adults compared with placebo control.
Although PRT has known metabolic benefits, an adequate
volume of aerobic exercise should be promoted if liver fat is
the therapeutic target.
Key words: exercise, liver fat, obesity, resistance training,
visceral adiposity

INTRODUCTION increased visceral adipose tissue (VAT) and intrahepatic

T HE BENEFIT OF exercise training for obesity-related
chronic disease management1 and all-cause mortality
reduction2 is well established. Obesity-related cardiovas-
cular and metabolic disease is particularly linked to
Correspondence: Dr Nathan Johnson, Discipline of Exercise and Sport
Science, University of Sydney, Lidcombe NSW 2141, Australia. Email:
nathan.johnson@sydney.edu.au
Conflict of interest: N.A. Johnson has received honoraria from Merck
Sharp & Dohme. I.D. Caterson has undertaken clinical trials funded
by the National Health and Medical Research Council, NovoNordisk,
Pfizer, BMS, and SFI and receives honoraria as a member of steering
committees of international trials. He has given talks for
NovoNordisk, Servier Laboratories, Ache, and Pfizer in the last 3 years.
The other authors have no conflict of interest.
Received 12 May 2016; revision 23 July 2016; accepted 26 July 2016.
lipid (IHL).3,4 Pharmacological agents do not reduce fat
specifically from these regions or are unsuitable for long-
term management. Therefore, establishing the efficacy of
lifestyle therapies, including exercise, is increasingly
important to inform guidelines for the management of
obesity and related comorbidities.
The utility of aerobic exercise training for VAT and IHL
reduction is well recognized, with benefits observed with
as little as 4 weeks of training and in the absence of clini-
cally significant weight loss.5–7 However, not all individ-
uals who would benefit from aerobic training are able to
adopt and adhere to such a program. Barriers in patients
with non-alcoholic fatty liver disease (NAFLD), for exam-
ple, include the presence of comorbid cardiovascular
disease8 and fear of falling.9 Progressive resistance training

© 2016 The Japan Society of Hepatology 622

Hepatology Research 2017; 47: 622–631
(PRT), which involves bursts of strenuous weight
lifting/movement has been suggested as an alternative
therapy.10 However, evidence for the efficacy of PRT on
VAT is conflicting6,7,11 and there is a paucity of data for
change in liver fat, with benefit observed in some10,12–15
but not other6,16–18 studies. These inconsistencies may be
due to variations in study design, with few studies prescrib-
ing the PRT dose recommended in current exercise
guidelines.19
The “dose” of PRT can be varied by manipulation of var-
iables including repetitions (the number of times a weight
is lifted), sets (number of repetitions performed), rest
(time between sets), and intensity (typically expressed as
a percentage of maximal lift [one repetition maximum,
1-RM]). Current guidelines recommend PRT on 2–3 days
per week involving 2–3 sets of 8–12 repetitions
progressing to an intensity of ≥80% 1-RM. To date, there
have been no studies examining the efficacy of supervised
PRT as per current guidelines on IHL and VAT using a
randomized placebo-controlled design and accurate
assessment of fat depots. This study sought to compare
8 weeks of supervised traditional PRT with a placebo
exercise control on IHL, VAT, and abdominal s.c. adipose
tissue (SAT) in previously inactive adults who are over-
weight or obese. We hypothesized that PRT would signifi-
cantly reduce IHL, VAT, and abdominal SAT compared
with control.
PARTICIPANTS AND METHODS
Trial design
A
ban
MULTICENTER, RANDOMIZED, placebo-controlled
trial was undertaken at two centers within a large ur-
university. The study was registered
(ACTRN12614000723684) and approved by the
University’s Human Research Ethics Committee and
conformed to the ethical guidelines of the 1975 Declara-
tion of Helsinki. Medical clearance was gained, and written
informed consent obtained, for each participant prior to
baseline assessments. Participants were required to abstain
from alcohol, vigorous exercise, and over-the-counter
medications for 24 h prior to baseline and post-
intervention assessments. Randomization to 8 weeks of
PRT or a placebo control (CON) was undertaken after
baseline assessment with an equally distributed pre-
generated list of permuted blocks with group allocation
concealed (by S.K.). An 8-week intervention was selected
because numerous studies have reported significant change
in hepatic steatosis with exercise interventions of this dura-
tion,5 including with circuit-based resistance exercise.13
Resistance training and liver fat 623
Participants
Included participants were inactive (undertaking struc-
tured exercise <3 days/week, and/or <150 min/week of
moderate-intensity exercise), overweight or obese (body
mass index [BMI] ≥25 kg/m2) adults (age, 29–59 years).
Participants were recruited through noticeboards,
electronic bulletins, and clinical databases between August
2011 and February 2015. Participants were excluded if
they reported a high alcohol intake (>20 g/day), treatment
with lipid-lowering or insulin-sensitizing agents, recent
changes in medication dose or body weight (previous
3 months), or had secondary causes of steatohepatitis,
viral hepatitis, alcoholic liver disease, diabetes, or any
non-controlled medical disorder.
Interventions
Progressive resistance training was supervised by an
accredited exercise physiologist with blood pressure
monitored throughout each session. In accordance with
recommendations,19 participants undertook PRT on
3 days/week, for 30–60 min (including 5 min warm-up
and cool-down at ~60% maximum heart rate on a cycle
ergometer) for 8 weeks. Exercises were: seated leg press,
chest press, lateral pull-down, calf raises, lunges, bicep
curls, triceps press-down, seated row, shoulder press, and
abdominal crunches. Volume and intensity were progres-
sively increased during weeks 1–3 so that by week 4, three
sets of 8–12 repetitions at 80–85% 1-RM was completed
with 60–120 s rest between each. Intensity was monitored
using the OMNI scale for rating of perceived exertion20
and increased progressively with strength gains through-
out the 8 weeks. Assessment of 1-RM occurred in the first
session of week 1 for the purpose of exercise prescription,
after participants had been educated and familiarized with
the correct technique.
Participants in CON undertook stretching, light fitball,
and self-massage exercises on 3 days/week as a home-
based intervention with sessions logged for participation
and compliance. Supervised sessions were provided once
or twice per fortnight and involved instruction on new
exercises. During the home-based sessions, participants
were instructed to warm-up and cool-down with 5 min
of slow walking.
Primary outcomes
Intrahepatic lipid, visceral, and abdominal SAT
M AGNETIC RESONANCE IMAGING (MRI) and pro-
ton magnetic resonance spectroscopy (1H-MRS)
were used to measure abdominal visceral and s.c. fat
volume and intrahepatic lipid concentration, respectively,
© 2016 The Japan Society of Hepatology
624 S. E Keating et al.
using a 1.5 Tesla Achieva whole-body system (Philips
Medical Systems, Best, the Netherlands). Measures were
acquired with the patient supine, by technicians blinded
to group allocation and the purpose of the study. Axial
T1-weighted fast field echo images were acquired during
suspended end-expiration (TR = 11 ms, TE = 4.5 ms, flip
angle = 40°) from diaphragm to pelvis (slice thickness,
10 mm; inter-slice gap, 10 mm) for determination of
abdominal fat volumes.
Intrahepatic lipid was quantified by 1H-MRS as previ-
ously outlined.21 Volumes of interest (3.0 × 2.0 × 2.0 cm
voxel) were centered within the right lobe of the liver
and spectra were acquired during respiratory gating (end
expiration) using a torso coil (flex M multichannel). A
point-resolved spectroscopy sequence (TR = 5000 ms,
TE = 34 ms, 32 measurements, 1024 sample points) was
used after fully automated high-order shimming on the
volume of interest. Excitation water suppression was used
to suppress the water signal during data acquisition.
Unsuppressed water spectra were acquired in vivo for use
as the internal standard.
Both MRI and 1H-MRS processing was carried out by an
experimenter blinded to treatment allocation. Cross-
sectional areas for VAT and abdominal SAT were analyzed
using automated software (Hippo Fat version 2.11, Pisa, It-
aly)22 with manual editing as necessary as previously de-
scribed.23 Liver spectra data were analyzed by magnetic
resonance user interface software (jMRUI version 4.0, EU
Project, http://www.jmrui.eu/) using a five resonance
model.21 Hepatic water signal amplitudes were measured
from the non-water suppressed spectrum using Hankel
Lanczos squares singular values decomposition.
Secondary outcomes
Muscle and intermuscular fat volume
T HE AREA OF the erector spinae, psoas, and paraspinal
muscle groups were quantified in six slices (from MRI)
from L4/L5 toward the left lobe of the liver using the
“region growing” function with thresholds adjusted manu-
ally. Partial muscle and intermuscular fat volumes were
calculated by summation of areas with adjustment for slice
thickness and interslice gap. Analysis was undertaken by a
trained experimenter blinded to treatment allocation
(Slice-O-matic version 5.0; Tomovision, Montréal, Canada).
Anthropometrics and blood pressure
Height was measured by stadiometer (SECA model 220
Telescopic Height Rod, Hamburg, Germany) and body
weight measured using an electronic digital platform scale
(Tanita BC-418 Body Composition Analyzer; Tanita
© 2016 The Japan Society of Hepatology
Hepatology Research 2017; 47: 622–631
Corporation, Tokyo, Japan). Waist circumference was
measured in triplicate against the skin in the horizontal
plane at the midpoint between the inferior margin of the
lowest rib and the iliac crest during deep expiration. Blood
pressure was assessed bilaterally using a manual cuff after
10–15 min of quiet sitting, with repeated measures if
>10 mmHg difference was observed and the highest
reading recorded.
Blood sampling and analysis
Venous blood was collected after an overnight fast (>10 h)
into serum separation (8.5 mL) and ethylenediaminetetra-
acetic acid (4 mL) tubes. Blood in the serum separation
tube was stored at 4°C prior to the analyses of serum
glucose, insulin and lipids, alanine aminotransferase
(ALT), aspartate aminotransferase, and high sensitivity
C-reactive protein (hs-CRP). Following centrifugation at
room temperature at 4000 g, plasma from the ethylenedi-
aminetetraacetic acid sample was extracted and stored at
20°C prior to analysis of serum free fatty acids (FFA).
All analyses were carried out on the same day as collection
by a commercially accredited laboratory. The homeostasis
model assessment for insulin resistance (HOMA-IR) was
calculated as glucose (mmol/L)*insulin (mU/L)/22.5.
Cardiorespiratory fitness/work capacity
Cardiorespiratory fitness/work capacity was assessed by
graded maximal exercise test on an electronically braked
cycle ergometer (Corival; Lode, Groningen, the
Netherlands) as previously described.23 The test was per-
formed to volitional fatigue, under the supervision of the
study physician with heart rate, blood pressure, 12-lead
electrocardiogram, and rating of perceived exertion24
obtained at each stage. Peak work capacity was measured25
and peak oxygen consumption (VO2peak) estimated.26
Habitual physical activity and dietary control
Participants were instructed to maintain their usual
diet/eating behavior and habitual activity levels for the
duration of the intervention. Three-day diet diaries (1
weekend day, 2 week days) were obtained during week 1
and week 8 of the study and analyzed by a dietitian
blinded to group allocation. The average daily intake of
macronutrients and total energy were quantified
(FoodWorks 2009; Xyris Software, version 6.0.6502, Bris-
bane, Australia). A tri-axial accelerometer was worn on
the left upper arm (SenseWear; BodyMedia Inc., Pitts-
burgh, PA, USA) to quantify mean time spent in sedentary
behavior and physical activity, steps per day, and total
daily energy expenditure as previously described.23 Data
Hepatology Research 2017; 47: 622–631
were omitted from analysis if the accelerometer was not
worn for at least 85% of a 24-h period.
Sample size
Controlled exercise intervention studies with detailed
assessment of body fat composition, with gold standard
methodology, typically involve 11–30 partici-
pants.10,13,27–29 A previous study recruited 11 subjects in
each group and reported a mean 13% reduction in IHL
(P < 0.01) with 8 weeks of circuit-based resistance
exercise.13 Therefore, we aimed to recruit a total of 30
participants (n = 15 each group).
Statistical methods
Primary and secondary outcome measures were explored
between PRT and CON with the group × time interaction
for the absolute change score of all outcome measures by
ANCOVA using the baseline value as the covariate (SPSS ver-
sion 22.0; IMB Corp., Armonk, NY, USA). Exercise and
control intervention compliance was determined as
Figure 1 CONSORT diagram of study process to assess the effect of resistance training on liver fat and visceral adiposity in adults with
obesity. CON, placebo control group; F, female; M, male; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy;
PRT, progressive resistance training.
Resistance training and liver fat 625
(number of sessions attended / total number of sessions
available) × 100. An intention-to-treat analysis was used
with group mean change scores imputed for dropouts.30
Effect size was calculated using Hedges’ g corrected for bias
with 95% confidence intervals. Relationships between
change in IHL, VAT, and abdominal SAT with change in
other variables, and potential confounders (diet or non-
exercise physical activity time) during treatment were
assessed by Pearson’s correlation coefficient using data from
all study participants. Statistical significance was accepted at
P < 0.05. Values are reported as means ± standard error.
RESULTS
T WENTY-NINE ELIGIBLE PARTICIPANTS (4 men and
25 women) undertook initial assessment and ran-
domization (Fig. 1), with a mean age of 42.9 ± 1.6 years
and mean BMI of 31.5 ± 0.8 kg/m2. All participants in
CON (n = 14) completed the intervention and assessment
of the primary outcomes; however, one participant did
© 2016 The Japan Society of Hepatology
626 S. E Keating et al. Hepatology Research 2017; 47: 622–631

not complete secondary outcome measures. Three of the

15 participants in PRT did not complete the intervention
or final assessment and one completed the full interven-
tion and secondary outcomes but not the final MRI (Fig. 1).
Baseline characteristics were not statistically different
between PRT and CON (P > 0.05 for all, Table 1). Compli-
ance with the study protocol was 93% ± 3.1% with PRT
and 87% ± 2.9% with CON. No adverse events were
reported during exercise testing or training.

Intrahepatic lipid, visceral, and abdominal SAT

There was no significant effect of PRT versus CON on IHL
(P = 0.19), which reduced by 0.07 ± 0.31% in PRT and
increased by 0.55 ± 0.77% in CON (Fig. 2). There was no
significant group × time interaction for VAT (P = 0.11) with
a non-significant reduction of 175.00 ± 85.16 cm3
observed in PRT and increase of 9.95 ± 64.13 cm3 in
Figure 2 Effect of 8 weeks of progressive resistance training
CON (Fig. 3a). There was no significant effect of PRT com- (closed triangles) or control (open triangles) on change in
pared with CON for reduction of SAT (P = 0.10) with a intrahepatic lipid in overweight or obese adults. Triangles show
non-significant reduction of 435.91 ± 245.31 cm3 in means (standard error).
PRT and increase of 127.29 ± 182.31 cm3 in CON (Fig. 3b).
These data are shown in Table 2. Subanalysis of partici-
± 7.52 cm3. There was no significant effect of PRT
pants who met the criteria for NAFLD at baseline
compared to CON for change in intermuscular fat partial
(>5.5% IHL, n = 5 in CON and n = 4 in PRT) found no
volume with a 6.95 ± 7.32 cm3 decrease in PRT and
effect of PRT on IHL, which increased by 0.40 ± 1.17%
a 16.03 ± 14.16 cm3 reduction in CON (Table 2).
and 1.00 ± 0.48% in PRT and CON, respectively
(P = 0.64). There were also no significant interactions for Anthropometrics and blood pressure
change in VAT or SAT (P > 0.05 for both).
There was no significant group × time effect for change in
Muscle and intermuscular fat volume body weight (P = 0.72), systolic blood pressure (P = 0.70),
There was a significant group × time interaction for change or DBP (P = 0.63). A significant group × time interaction
in partial muscle volume (P = 0.03) which increased in was observed for change in waist circumference, with a
PRT by 54.87 ± 20.05 cm3 and reduced in CON by 0.04 1.36 ± 0.37 cm reduction in PRT and a 0.14 ± 0.56 cm
increase in CON (P = 0.04) (Table 2).

Blood lipids and biochemistry

A significant group × time effect was observed for total
Table 1 Baseline characteristics of participants in a randomized
controlled trial assessing the effect of resistance training on liver
cholesterol (P = 0.049) which reduced in CON by 0.49
fat and visceral adiposity in adults with obesityParticipant ± 0.22 mmol/L but not PRT (increase of 0.15
± 0.12 mmol/L). There was a trend toward change in aspar-
Characteristics PRT (n = 15) CON (n = 14) tate aminotransferase with a 2.2 ± 1.47 U/L reduction in
Demographics CON and a slight increase (0.82 ± 1.52 U/L) in PRT
Age, years 45.4 (1.9) 44.2 (2.8)
(P = 0.055). There was no significant group × time interac-
Sex, n, male / female 2/13 2/12
tion for change in ALT, triglycerides, high density lipopro-
BMI, kg/m2 32.2 (1.2) 30.8 (1.0)
Waist circumference, cm 97.6 (3.3) 92.8 (2.2)
tein, low density lipoprotein, hs-CRP, fasting glucose,
Intrahepatic lipid, % 3.3 (0.7) 5.1 (1.4) insulin, or FFA (P > 0.05 for all, Table 2).
Energy intake, kJ/day 9487 (693)† 9198 (745)‡
Work capacity and cardiorespiratory fitness
Presented as mean (standard error).BMI, body mass index; CON, pla-
cebo control; PRT, progressive resistance training.
There was a significant group × time interaction for change
†n = 11; in peak work capacity with PRT improving by 10.18
‡n = 12. ± 3.09 W and 0.347 ± 2.36 W in CON (P = 0.01). A

© 2016 The Japan Society of Hepatology

Hepatology Research 2017; 47: 622–631
Figure 3 Effect of 8 weeks of progressive resistance training
(closed triangles) or control (open triangles) on change in visceral
adipose tissue (a) and s.c. adipose tissue in overweight or obese
adults. Triangles show means (standard error).
significant group × time effect was observed for change in
peak oxygen consumption which improved by 1.45
± 0.35 mL/kg/min in PRT and by 0.11 ± 0.41 mL/kg/min
in CON (P = 0.03) (Table 2).
Habitual diet and activity
Diet diary data were unavailable for 4/15 (26%) partici-
pants in PRT and 2/14 (14%) participants in CON due
to non-compliance with food records. Physical activity
data were unavailable for 3/15 (20%) participants in PRT
and 1/14 (7%) in CON due to not wearing the monitor.
There was no group × time interaction for change in total
energy expenditure, which reduced by 165 ± 156 kJ in
PRT and by 405 ± 456 kJ in CON (P = 0.90), or for change
Resistance training and liver fat 627
in number of steps per day, reducing by 487 ± 419 steps
and by 1658 ± 908 steps in PRT and CON respectively
(P = 0.41). No significant group × time effects were
observed in time spent in sedentary behavior, which
increased by 0.5 ± 8.8 min and by 1.6 ± 25 min in PRT
and CON, respectively (P = 0.38), or in time spent in
moderate activity, which reduced by 2.7 ± 5.8 min and by
15.2 ± 14.4 min in PRT and CON, respectively (P = 0.72).
There was no significant group × time interaction for
total energy intake, which reduced by 396 ± 452 kJ in
PRT and by 516 ± 388 kJ in CON (P = 0.77) for change in
macronutrients. Protein % reduced by 0.6 ± 1.8% in PRT
and increased by 1.4 ± 0.8% in CON (P = 0.20). Fat %
reduced by 4.6 ± 5.6% in PRT and increased by 4.5
± 1.4% in CON (P = 0.86). Carbohydrate % reduced by
1.8 ± 2.2% and by 4.0 ± 1.7% in PRT and CON, respec-
tively (P = 0.30).
Correlations between IHL, VAT, abdominal SAT
and other cardio-metabolic risk variables
There was no significant association between change in
IHL and change in any other variable (P > 0.05). Change
in VAT was positively associated with change in SAT
(r = 0.44, P = 0.02), weight (r = 0.38, P = 0.04), and waist
circumference (r = 0.51, P = 0.005). There was a moderate
association between change in SAT and change in weight
(r = 0.52, P = 0.004), waist circumference (r = 0.41,
P = 0.03), systolic blood pressure (r = 0.46, P = 0.01), hs-
CRP (r = 0.64, P = 0.00), and FFA (r = 0.43, P = 0.02).
There was no significant correlation between change in
IHL, VAT, or SAT and change in energy intake, macronutri-
ent composition, habitual activity energy expenditure,
sedentary behavior, moderate activity, or daily step count
(P > 0.05) except for a moderate association observed
between change in VAT and change in % protein
(r = 0.50, P = 0.02).
DISCUSSION
HIS IS THE first study to compare the effects of super-
T vised PRT with a placebo sham exercise control on
IHL, VAT, and abdominal SAT using accurate, quantifiable
magnetic resonance techniques. Using a randomized
controlled design and a resistance exercise program that
complied with current recommendations, we found no
evidence to suggest that PRT is effective for reducing liver
fat, VAT, or abdominal SAT in previously inactive adults
with obesity. These data are important as it is of consider-
able clinical interest whether resistance training reduces
intrahepatic lipid and visceral adiposity.31
© 2016 The Japan Society of Hepatology
628 S. E Keating et al. Hepatology Research 2017; 47: 622–631

Table 2 Outcome measures in a randomized controlled trial assessing the effect of resistance training on liver fat and visceral adiposity
in adults with obesity

PRT CON P-value

(group × time)

Baseline Post Baseline Post ES (95% CI)

Weight, kg 86.5 (4.9) 86.6 (4.3) 85.5 (3.6) 85.7 (3.8) –0.12 ( 0.85, 0.61) 0.72
BMI, kg/m2 32.2 (1.2) 32.2 (1.2) 30.8 (1.0) 31.3 (1.2) –0.41 ( 1.15, 0.32) 0.21
Waist circumference, 97.6 (3.3) 96.3 (3.2) 92.8 (2.2) 92.9 (2.4) 0.82 ( 1.58, 0.06) 0.04†
cm
Intrahepatic lipid, % 3.3 (0.7) 3.2 (0.8) 5.1 (1.4) 5.6 (1.5) –0.56 ( 1.34, 0.15) 0.19
Subcutaneous adipose 11 874 (914) 11 439 (825) 11 251 (871) 11 378 (890) –0.66 ( 1.41, 0.09) 0.10
tissue, cm3
Visceral adipose tissue, 2447 (360) 2272 (345) 2334 (343) 2344 (345) 0.62 ( 1.37, 0.13) 0.11
cm3
Partial muscle volume, 714 (28.4) 757 (39.4) 758 (39.9) 758 (41.7) 0.90 (0.14, 1.67) 0.03††
cm3
Partial intermuscular 122 (13.0) 115.0 (9.8) 126 (17.5) 142 (19.3) 0.21 ( 0.52, 0.94) 0.61
adipose tissue volume,
cm3
Biochemistry
AST, U/L 20.6 (1.7) 21.4 (1.7) 20.4 (2.3) 18.1 (1.2) 0.52 ( 0.22, 1.26) 0.06
ALT, U/L 19.7 (3.5) 21.0 (3.4) 26.2 (5.1) 19.9 (3.1) 0.67 ( 0.08, 1.42) 0.16
Fasting glucose, 4.3 (0.2) 4.2 (0.2) 4.0 (0.1) 4.0 (1.2) 0.50 ( 0.24, 1.24) 0.15
mmol/L
Insulin, mU/L 6.9 (0.6) 7.3 (0.6) 8.9 (1.4) 8.1 (1.0) 0.53 ( 0.21, 1.27) 0.46
HOMA-IR 1.4 (0.2) 1.4 (0.2) 1.6 (0.3) 1.3 (0.3) 0.55 ( 0.19, 1.29) 0.18
hs-CRP, mg/L 5.1 (1.4) 4.6 (1.3) 4.5 (1.8) 4.3 (1.6) -0.20 ( 0.93, 0.53) 0.64
Total cholesterol, 5.3 (0.2) 5.4 (0.2) 5.9 (0.3) 5.4 (0.2) 0.94 (0.17, 1.71) 0.049††
mmol/L
HDL, mmol/L 1.5 (0.1) 1.7 (0.1) 1.8 (0.3) 1.5 (0.1) -0.08 ( 0.81, 0.65) 0.09
LDL, mmol/L 3.3 (0.2) 3.2 (0.2) 3.6 (0.3) 3.3 (0.2) -0.44 ( 0.30, 1.18) 0.38
Triglycerides, 1.1 (0.1) 1.2 (0.1) 1.3 (0.2) 1.4 (0.2) 0.21 ( 0.52, 0.94) 0.42
mmol/L
Free fatty acids, 446 (51) 605 (78) 323 (38) 416 (43) 0.33 ( 0.41, 1.06) 0.30
μmol/L
Blood pressure
SBP, mmHg 121.3 (3.2) 116.8 (2.2) 120.1 (4.0) 117.4 (2.3) -0.15 ( 0.88, 0.58) 0.70
DBP, mmHg 76.6 (1.8) 76.0 (1.9) 77.6 (1.8) 75.7 (1.2) 0.22 ( 0.51, 0.96) 0.63
Fitness
VO2peak, mL/kg/min‡ 23.6 (0.9) 25.1 (1.1) 21.5 (1.5) 21.6 (1.5) 0.90 (0.15, 1.67) 0.01††
Wpeak 138.8 (6.9) 149.0 (7.3) 121.9 (12.0) 122.2 (11.3) 0.91 (0.14, 1.66) 0.03††

‡Estimated from peak work capacity (Wpeak).

†Significant group × time interaction (P < 0.05).
Data are shown as mean (standard error). ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CI, confidence
interval; CON, placebo control; DBP, diastolic blood pressure; ES, Hedges’ bias corrected effect size; HDL, high density lipoprotein; HOMA-IR,
homeostasis model assessment of insulin resistance; hs-CRP, high sensitivity C-reactive protein; LDL, low density lipoprotein cholesterol; PRT,
progressive resistance training; SBP, systolic blood pressure; VO2peak: peak aerobic capacity.

Unlike the general consensus that aerobic exercise is not all,6,16,17,32 previous reports that have examined the
beneficial for reducing liver fat, there is less agreement in hepatic benefits of resistance training. We observed no
current reports regarding the efficacy of PRT on liver fat effect of PRT on liver fat in adults with obesity using
reduction. Our finding is in contrast to some,10,12–15 but current recommendations. Although not all participants

© 2016 The Japan Society of Hepatology

Hepatology Research 2017; 47: 622–631
had liver steatosis at baseline, which could possibly dilute
the effects of PRT on IHL in this study, previous studies are
nevertheless conflicting regarding the importance of base-
line steatosis versus other factors, such as the exercise dose,
in eliciting liver fat reduction. For instance, an apparent
benefit of PRT on liver fat has previously been reported
in both NAFLD10,12,15 and non-NAFLD14 cohorts, and
our observation of no significant effect has been similarly
observed in both NAFLD16,32 and non-NAFLD6,17 cohorts.
Indeed, the variations in outcomes may relate to differ-
ences in resistance training interventions. For example,
Hallsworth et al. observed a 13% relative reduction (2%
absolute reduction) in IHL, but not VAT or abdominal
SAT, with an 8-week circuit-based resistance training inter-
vention (which had an additional aerobic component) in
patients with NAFLD and elevated ALT.13 However, in
contrast, 6 months of circuit training was not effective in
reducing biopsy-quantified steatosis in adults with
NAFLD.16 Other studies have used atypical exercises with
weighted belts,12 or have lacked supervision of the training
sessions which makes adherence to the prescribed inten-
sity difficult to determine.15 Clearly, more data is needed
to confirm the efficacy of PRT on IHL. Importantly, to date,
most studies have not included a placebo/control
group;6,10,12–14,16–18 our study is the first to include both
sham exercise placebo control and to use accurate liver
lipid quantification techniques (1H-MRS).
The prescription of PRT at the dose used here is known
to enhance insulin sensitivity (assessed by euglycemic
hyperinsulinemic clamp) and increase lean body mass.19
As expected, we observed a small but significant increase
in muscle mass in the PRT group compared with CON.
However, the magnitude of increase, coupled with the
small, non-significant reduction in VAT and SAT in PRT,
meant that there was no net change in BMI. Furthermore,
we did not observe any changes in static measures of insu-
lin sensitivity inferred by HOMA-IR. This finding may
reflect the poor sensitivity of HOMA-IR to detect change
in peripheral insulin sensitivity, and is consistent with sim-
ilar studies.33 Given that hepatic fat accumulation and
visceral adiposity are strongly associated with insulin resis-
tance,34 the inclusion of PRT in exercise interventions for
patients with abdominal obesity and fatty liver seems
logical. Furthermore, there is an increased risk of liver fat
accumulation in sarcopenic groups.35
Only two studies in adults6,10 and three studies in
children and adolescents14,17,32 have examined the role
of traditional PRT on liver fat using imaging assessment
techniques in adults. Bacchi and colleagues directly
compared (n = 30) 4 months of regular aerobic exercise
(60 min at 60–70% heart rate reserve on 3 days/week)
Resistance training and liver fat 629
with PRT (nine exercises, three sets of 10 repetitions at
70–80% 1-RM) and observed a significant and compara-
ble mean reduction in apparent liver fat of 33% and 26%
in adults with type 2 diabetes and NAFLD, respectively.10
However, whether PRT was better than placebo was not
answered because of lack of a control group. Furthermore,
liver fat was determined using MRI, which lacks the extent
of evidence base for reliability and validity compared with
spectroscopy, which is considered the gold standard for
non-invasive quantification of liver steatosis.36 In contrast
to the findings by Bacchi et al., Slentz and coworkers
compared (n = 107) aerobic exercise (~40 min at 75%
VO2peak on 3 days/week), progressive resistance training
(eight exercises, three sets of 8–12 repetitions at the equiv-
alent of ~70–85% 1-RM, on 3 days/week), and a com-
bined aerobic exercise and resistance training group
(participants performed both interventions) directly.6 Sig-
nificant reduction in liver fat score (computed tomogra-
phy) was observed with aerobic training alone but,
consistent with our findings, not with PRT alone.6 How-
ever, this study also did not include a control group and
quantification of steatosis by computed tomography is less
accurate than 1H-MRS.36 The only other study using a ran-
domized controlled design compared 12 weeks of
stretching on 3 days/week (n = 31) with PRT (eight exer-
cises, three sets of 8–12 repetitions with 1–2 min between
sets; n = 33) and observed a 12% relative reduction in liver
fat. However, this was assessed using an ultrasound-based
score (hepatorenal index), which is suboptimal for the
quantification of steatosis,36 and the majority of gym ses-
sions were unsupervised.15 Similar inconsistencies in re-
sults with PRT have been observed in adolescent
populations. Lee and colleagues (2012) observed a non-
significant mean 2% reduction in IHL with PRT (8–12 rep-
etitions to fatigue with 1–2 min rest between sets, 3 days
per week for 12 weeks; n = 16) in adolescent boys with obe-
sity, which was comparable with the aerobic exercise pro-
gram (n = 15).14 However, no benefit of PRT was
observed in adolescent girls with abdominal obesity with
only aerobic exercise training producing significant reduc-
tions in IHL, VAT, and insulin sensitivity.17 Similarly, there
were no significant reductions in IHL or VAT in obese
youth with NAFLD (mean baseline IHL 9.2%) with
12 weeks of PRT based on current guidelines.32
Limitations of the present study include the mostly
female (86%) and Caucasian (76%) cohort, which may
limit the generalizability of our results. Notably, only
31% of participants had NAFLD as diagnosed by >5.5%
IHL on baseline 1H-MRS, but this characterizes most stud-
ies in the field. Although our subanalysis in participants
with NAFLD (n = 9) indicated no reduction in IHL with
© 2016 The Japan Society of Hepatology
630 S. E Keating et al.
PRT, further cohort studies with NAFLD-only subjects are
required to ascertain the real effects of PRT on IHL in
NAFLD. Moreover, while changes in IHL and VAT have
been observed in as little as 4 weeks of aerobic exercise
training27 and with 8 weeks of circuit training,13 PRT at
the dose used in this study may take longer to yield signif-
icant benefits. We were also unable to determine the sever-
ity of NAFLD and the effect of this dose of PRT on fibrosis
due to the ethical considerations regarding the use of liver
biopsy in research studies and the lack of other validated,
accurate surrogates for fibrosis. Our sample size was rela-
tively low; however, it is the largest using both a control
arm and gold standard assessment methods and compara-
ble to other studies examining resistance training for liver
fat reduction which have observed an effect.10,12,13 Larger
studies with longer duration of supervised PRT are there-
fore warranted.
In conclusion, this study showed that PRT at the level
promoted in current guidelines did not significantly
reduce IHL, VAT, or abdominal SAT in previously inactive
adults with obesity. Exercise guidelines for health and
fitness promote the use of both regular aerobic and resis-
tance training. Based on findings from this study, we
recommend that a sufficient volume of aerobic exercise
should be principally promoted for the reduction of liver
fat and VAT.
ACKNOWLEDGMENTS
T HIS RESEARCH WAS partly supported by funding
from the Diabetes Australia Research Trust (Establish-
ment Grant: N.A. Johnson). Jacob George is supported
by the Robert W. Storr bequest to the Sydney Medical
Foundation and by grants from the National Health and
Medical Research Council. Amanda Sainsbury is
supported by fellowships from the University of
Sydney/Sydney Medical School and the National Health
and Medical Research Council.
REFERENCES
1 Pedersen BK, Saltin B. Evidence for prescribing exercise as
therapy in chronic disease. Scand J Med Sci Sports 2006; 16:
3–63.
2 Kodama S, Saito K, Tanaka S et al. Cardiorespiratory fitness as
a quantitative predictor of all-cause mortality and cardiovas-
cular events in healthy men and women: a meta-analysis.
JAMA 2009; 301: 2024–35.
3 Hocking S, Samocha-Bonet D, Milner K-L, Greenfield JR,
Chisholm DJ. Adiposity and insulin resistance in humans:
the role of the different tissue and cellular lipid depots. Endocr
Rev 2013; 34: 463–500.
© 2016 The Japan Society of Hepatology
Hepatology Research 2017; 47: 622–631
4 Tchernof A, Després J. Pathophysiology of human visceral
obesity: an update. Physiol Rev 2013; 93: 359–404.
5 Keating S, Hackett D, George J, Johnson N. Exercise and non-
alcoholic fatty liver disease: a systematic review and meta-
analysis. J Hepatol 2012; 57: 157–66.
6 Slentz CA, Bateman LA, Willis LH et al. Effects of aerobic vs.
resistance training on visceral and liver fat stores, liver
enzymes, and insulin resistance by HOMA in overweight
adults from STRRIDE AT/RT. Am J Physiol Endocrinol Metab
2011; 301: E1033–E1039.
7 Ismail I, Keating SE, Baker MK, Johnson NA. A systematic
review and meta-analysis of the effect of aerobic vs. resistance
exercise training on visceral fat. Obes Rev 2012; 13: 68–91.
8 Ballestri S, Lonardo A, Bonapace S, Byrne CD, Loria P, Targher
G. Risk of cardiovascular, cardiac and arrhythmic complica-
tions in patients with non-alcoholic fatty liver disease. World
J Gastroenterol 2014; 20: 1724–45.
9 Frith J, Day CP, Robinson L, Elliott C, Jones DE, Newton JL.
Potential strategies to improve uptake of exercise interven-
tions in non-alcoholic fatty liver disease. J Hepatol 2010; 52:
112–6.
10 Bacchi E, Negri C, Targher G et al. Both resistance training and
aerobic training reduce hepatic fat content in type 2 diabetic
subjects with nonalcoholic fatty liver disease (the RAED2 ran-
domized trial). Hepatology 2013; 58: 1287–95.
11 Strasser B, Arvandi M, Siebert U. Resistance training, visceral
obesity and inflammatory response: a review of the evidence.
Obes Rev 2012; 13: 578–91.
12 Damor K, Mittal K, Bhalla AS et al. Effect of progressive resis-
tance exercise training on hepatic fat in Asian Indians with
non-alcoholic fatty liver disease. Br J Med Med Res 2014; 4:
114–24.
13 Hallsworth K, Fattakhova G, Hollingsworth KG et al. Resis-
tance exercise reduces liver fat and its mediators in non-
alcoholic fatty liver disease independent of weight loss. Gut
2011; 60: 1278–83.
14 Lee S, Bacha F, Hannon T, Kuk JL, Boesch C, Arslanian S. Ef-
fects of aerobic versus resistance exercise without caloric
restriction on abdominal fat, intrahepatic lipid, and insulin
sensitivity in obese adolescent boys: a randomized, controlled
trial. Diabetes 2012; 61: 2787–95.
15 Zelber-Sagi S, Buch A, Yeshua H et al. Effect of resistance train-
ing on non-alcoholic fatty-liver disease a randomized-clinical
trial. World J Gastroenterol 2014; 20: 4382–92.
16 Hickman IJ, Byrne NM, Croci I et al. A pilot randomised study
of the metabolic and histological effects of exercise in non-
alcoholic steatohepatitis. J Diabetes Metab 2013; 4: 300.
17 Lee S, Deldin AR, White D et al. Aerobic exercise but not resis-
tance exercise reduces intrahepatic lipid content and visceral
fat and improves insulin sensitivity in obese adolescent girls:
a randomized controlled trial. Am J Physiol Endocrinol Metab
2013; 305: E1222–E1229.
18 van der Heijden G-J, Wang ZJ, Chu ZD et al. A 12-week aerobic
exercise program reduces hepatic fat accumulation and insulin
resistance in obese, Hispanic adolescents. (Published erratum
appears in Obesity 2010; 18: 1062). Obesity 2010; 18: 384–90.
Hepatology Research 2017; 47: 622–631
19 Garber CE, Blissmer B, Deschenes MR et al. Quantity and
quality of exercise for developing and maintaining cardiore-
spiratory, musculoskeletal, and neuromotor fitness in
apparently healthy adults: guidance for prescribing exercise.
Med Sci Sports Exerc 2011; 43: 1334–59.
20 Naclerio F, Rodríguez-Romo G, Barriopedro-Moro MI,
Jiménez A, Alvar BA, Triplett NT. Control of resistance training
intensity by the OMNI perceived exertion scale. J Strength
Cond Res 2011; 25: 1879–88.
21 Johnson NA, Walton DW, Sachinwalla T et al. Noninvasive
assessment of hepatic lipid composition: advancing under-
standing and management of fatty liver disorders. Hepatology
2008; 47: 1513–23.
22 Positano V, Gastaldelli A, Sironi AM, Santarelli MF, Lombardi
M, Landini L. An accurate and robust method for unsuper-
vised assessment of abdominal fat by MRI. J Magn Reson
Imaging 2004; 20: 684–9.
23 Keating SE, Hackett DA, Parker HM et al. Effect of aerobic ex-
ercise training dose on liver fat and visceral adiposity. J Hepatol
2015; 63: 174–82.
24 Borg G. Psychophysical bases of perceived exertion. Med Sci
Sports Exerc 1982; 14: 377–81.
25 Kuipers H, Verstappen FTJ, Keizer HA, Geurten P,
van Kranenburg G. Variability of aerobic performance in the
laboratory and its physiologic correlates. Int J Sports Med
1985; 6: 197–201.
26 Hawley J, Noakes T. Peak power output predicts maximal
oxygen uptake and performance time in trained cyclists. Eur
J Appl Physiol Occup Physiol 1992; 65: 79–83.
27 Johnson N, Sachinwalla T, Walton D et al. Aerobic exercise
training reduces hepatic and visceral lipids in obese individ-
uals without weight loss. Hepatology 2009; 50: 1105–12.
Resistance training and liver fat 631
28 Pugh CJ, Spring VS, Kemp GJ et al. Exercise training
reverses endothelial dysfunction in nonalcoholic fatty liver
disease. Am J Physiol Heart Circ Physiol 2014; 307:
H1298–H1306.
29 Sullivan S, Kirk EP, Mittendorfer B, Patterson BW, Klein S.
Randomized trial of exercise effect on intrahepatic triglyceride
content and lipid kinetics in nonalcoholic fatty liver disease.
Hepatology 2012; 55: 1738–45.
30 Armijo-Olivo S, Warren S, Magee D. Intention to treat analy-
sis, compliance, drop-outs and how to deal with missing
data in clinical research: a review. Phys Ther Rev 2009; 14:
36–49.
31 Browning JD, Szczepaniak LS, Dobbins R et al. Prevalence of
hepatic steatosis in an urban population in the United
States: impact of ethnicity. Hepatology 2004; 40: 1387–95.
32 Van Der Heijden GJ, Wang ZJ, Chu Z et al. Strength exercise
improves muscle mass and hepatic insulin sensitivity in obese
youth. Med Sci Sports Exerc 2010; 42: 1973–80.
33 Klimcakova E, Polak J, Moro C et al. Dynamic strength train-
ing improves insulin sensitivity without altering plasma
levels and gene expression of adipokines in subcutaneous ad-
ipose tissue in obese men. J Clin Endocrinol Metab 2006; 91:
5107–12.
34 Snel M, Jonker JT, Schoones J et al. Ectopic fat and insulin re-
sistance: pathophysiology and effect of diet and lifestyle
interventions. Int J Endocrinol 2012; 2012: 983814.
35 Hong HC, Hwang SY, Choi HY et al. Relationship between
sarcopenia and nonalcoholic fatty liver disease: the Korean
Sarcopenic Obesity Study. Hepatology 2014; 59: 1772–8.
36 Machado MV, Cortez-Pinto H. Non-invasive diagnosis of
non-alcoholic fatty liver disease. A critical appraisal. J Hepatol
2013; 58: 1007–19.
© 2016 The Japan Society of Hepatology