Mediterranean Dietary Pattern and Inflammation in adults with Non-Alcoholic Fatty Liver Disease

Journal Pre-proof

Adherence to a Mediterranean diet may improve serum adiponectin

in adults with non-alcoholic fatty liver disease: The MEDINA
randomized controlled trial

Anjana Reddy , Paul Della Gatta , Shaun Mason ,

Amanda J. Nicoll , Marno Ryan , Catherine Itsiopoulos ,
Gavin Abbott , Nathan A. Johnson , Siddharth Sood ,
Stuart K. Roberts , Elena S. George , Audrey C. Tierney

PII: S0271-5317(23)00090-8
DOI: https://doi.org/10.1016/j.nutres.2023.09.005
Reference: NTR 8470

To appear in: Nutrition Research

Received date: 1 March 2023

Revised date: 6 September 2023
Accepted date: 7 September 2023

Please cite this article as: Anjana Reddy , Paul Della Gatta , Shaun Mason , Amanda J. Nicoll ,
Marno Ryan , Catherine Itsiopoulos , Gavin Abbott , Nathan A. Johnson , Siddharth Sood ,
Stuart K. Roberts , Elena S. George , Audrey C. Tierney , Adherence to a Mediterranean diet may
improve serum adiponectin in adults with non-alcoholic fatty liver disease: The MEDINA randomized
controlled trial, Nutrition Research (2023), doi: https://doi.org/10.1016/j.nutres.2023.09.005

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2023 Published by Elsevier Inc.

Highlights:

• The Mediterranean diet improved serum adiponectin following a 12-week intervention

• Adiponectin and visceral fat were reduced after a Mediterranean diet, without weight loss

• The low-fat diet did not elicit improvements in inflammatory markers

• Management of Non-Alcoholic Fatty Liver Disease (NAFLD) should focus on diet

composition and its functional components

Adherence to a Mediterranean diet may improve serum adiponectin in adults
with non-alcoholic fatty liver disease: The MEDINA randomized controlled trial

Anjana Reddy1,2, Paul Della Gatta3, Shaun Mason3, Amanda J. Nicoll4, Marno Ryan5,

Catherine Itsiopoulos1,6 Gavin Abbott2, Nathan A. Johnson7, Siddharth Sood8,9, Stuart K.

Roberts10,11, Elena S. George1,3*a, Audrey C. Tierney1,12a

1. School of Allied Health, Human Services and Sport, La Trobe University, Australia

2. Monash Centre for Health Research and Implementation, School of Public Health and

Preventative Medicine, Monash University, Melbourne, VIC 3004, Australia

3. Institute for Physical Activity and Nutrition, School of Exercise and Nutrition Sciences, Deakin

University, Geelong, Australia

4. Department of Gastroenterology, Eastern Health, Box Hill, Australia

5. Department of Gastroenterology and Hepatology, St Vincent’s Hospital, Fitzroy, Australia

6. School of Health and Biomedical Sciences, RMIT University, Melbourne, Australia

7. The Boden Collaboration for Obesity, Nutrition, Exercise and Eating Disorders, The University of

Sydney, Sydney, New South Wales, Australia

8. Department of Gastroenterology, Melbourne Health, Melbourne, Australia

9. Department of Medicine, University of Melbourne, Parkville, Australia

10. Department of Gastroenterology, Alfred Health, Prahran, Australia

11. Central Clinical School, Monash University, Clayton, Australia

12. School of Allied Health, Health Implementation Science and Technology Research Cluster,

Health Research Institute, University of Limerick, Ireland

*Corresponding author: Elena S George, Institute for Physical Activity and Nutrition

(IPAN), School of Exercise and Nutrition Sciences, Deakin University, 3220 Geelong,

Australia. Email: elena.george@deakin.edu.au

aThese two authors share co-senior authorship

Australian New Zealand Clinical Trials Registry (ANZCTR): ACTRN12615001010583

Running title: Mediterranean Dietary Pattern and Inflammation in adults with Non-

Alcoholic Fatty Liver Disease

Number of figures: 1

Number of tables: 3

Supplementary Files: 1

Manuscript word count (Abstract through Conclusions): 4,709

List of Abbreviations:

APD Accredited Practicing Dietitian

BIA Bioelectrical Impendence Analysis
BMI Body Mass Index
HOMA-IR Homeostasis Model Assessment of Insulin Resistance
1
H-MRS proton magnetic resonance spectroscopy
hs-CRP high sensitive C-Reactive Protein
IL-6 Interleukin-6
LFD Low Fat Diet
MEDINA Mediterranean Dietary Intervention for patients with Non-alcoholic fatty liver
disease study
MedDiet Mediterranean diet
MetS the Metabolic Syndrome
n-3 Omega-3 Fatty Acids
NAFLD Non-Alcoholic Fatty Liver Disease
NASH Non-Alcoholic Steatohepatitis
PREDIMED Prevención con Dieta Mediterránea
PUFA Polyunsaturated Fatty Acids
RCT Randomised Controlled Trial
T2DM Type 2 Diabetes Mellitus
TNF-α Tumor Necrosis Factor-Alpha
ABSTRACT

Non-alcoholic fatty liver disease (NAFLD) affects approximately thirty percent of adults

worldwide, with chronic low-grade inflammation a key pathophysiological feature of

progression. The Mediterranean diet (MedDiet) is recognised for improving metabolic and

hepatic outcomes in people with diabetes and NAFLD, in part, via anti-inflammatory

properties. The aim of this study was to determine the effect of an ad libitum MedDiet versus

low-fat diet (LFD) on inflammatory markers in adults with NAFLD. It was hypothesised that

the MedDiet, and it’s individual components, would improve inflammation. This multicentre,

randomized controlled trial, randomized participants to a MedDiet or LFD intervention for 12-

weeks. Primary outcomes included change from baseline to 12-weeks for serum hs-CRP, IL-

6, TNF-α, adiponectin, leptin and resistin. Forty-two participants (60% female, 52.3 ±

12.6years, BMI 32.2 ± 6.2kg/m²) were randomized to the MedDiet (n=19) or low-fat diet

(n=23). At 12-weeks, the LFD showed a greater decrease in leptin compared to the MedDiet (-

1.20 ± 3.9ng/mL vs 0.64 ± 3.5ng/mL, p =0.010). Adiponectin significantly improved within

the MedDiet (13.7 ± 9.2µg/mL to 17.0 ± 12.5µg/mL, p =0.016), but not within the LFD group.

No statistically significant changes were observed for other inflammatory markers following

the MedDiet or LFD. Adherence to the MedDiet significantly improved for all participants

regardless of diet allocation, although greater in the MedDiet group. Adiponectin significantly

improved following a Mediterranean diet intervention, in the absence of weight-loss. The low-

fat diet did not elicit improvements in inflammatory markers. High-quality clinical trials

appropriately powered to inflammatory markers are required in this population.

Keywords: Non-alcoholic fatty liver disease; diet intervention; Mediterranean diet;

inflammation; adiponectin

1
1. Introduction

Non-alcoholic fatty liver disease (NAFLD) is the most common liver disease worldwide,

affecting up to 30% of adults [1]. Prevalence has quickly risen in parallel with rates of obesity

and type 2 diabetes mellitus (T2DM); which are also risk factors for cardiovascular disease

(CVD) and CVD-related mortality [2]. NAFLD occurs due to excess hepatic triglyceride

accumulation, portal vein inflammation and tissue injury, which can progress to non-alcoholic

steatohepatitis (NASH), with or without fibrosis [3]. The pathogenesis of NAFLD involves

multiple metabolic ‘hits’ including genetic and environmental factors [4], and affects several

extra-hepatic organs and regulatory pathways [5]. A combination of these factors, including an

unfavourable diet and lifestyle, leads to intrahepatic fat accumulation which impairs hepatocyte

insulin signalling, augments metabolic derangement, and worsens peripheral insulin resistance

and other downstream events [6]. Inflammatory cytokines and adipokines play a key role in

the progression from steatosis to NASH, while IR amplifies de novo lipogenesis, together

sustaining a chronic low-grade inflammatory and insulin resistant state [7].

Several pathways lead to an increase in cytokine and adipokine production and

secretion in the liver. The up-regulation of interleukin-6 (IL-6) and tumour necrosis factor

alpha (TNF-α) are responsible for an increase in the production of acute phase proteins such as

fibrinogen and C-reactive protein (CRP) [8, 9], a well-known marker of inflammation,

positively associated with the risk of CVD [9]. Adipocyte derived markers such as adiponectin,

leptin and resistin, are predominately released from visceral adipose tissue (VAT) [10]. Poor

diet quality promotes an inflammatory milieu by increasing the production and release of pro-

inflammatory cytokines (IL-6, TNF-α, CRP) and adipokines (leptin, resistin), while lowering

concentrations of the anti-inflammatory adipokine adiponectin [11].

2
 Lifestyle interventions centred around diet and weight loss are primary management

strategies for NAFLD due to established benefits on hepatic outcomes [12]. A growing number

of clinical trials have investigated inflammatory markers as primary endpoints for prevention

and management of NAFLD, considering that systemic inflammation is a main

pathophysiological feature of disease progression [14]. Significant reductions in hs-CRP, IL-

6, TNF-α, adiponectin and leptin have been observed following a hypocaloric diet, with weight-

loss identified as a key driver of improvements [13]. However, calorie restriction is difficult to

maintain in metabolically diseased partcipants and long-term adherence and maintenance of

weight-loss are often not assessed [14]. Recently, a systematic review and meta-analysis found

that an isocaloric diet complemented with an anti-oxidant, anti-inflammatory, anti-lipidemic

and/or insulin sensitising supplements produced significant improvments in hs-CRP, IL-6 and

TNF-α for patients with NAFLD [14]. A diet known to provide similar benefits, the

Mediterranean diet, is also reported to improve CRP, IL-6, TNF-α and adiponectin in

individuals diagnosed with NAFLD [14], however existing studies have been conducted in

Mediterranean regions and include another feature to the dietary intervention such as an energy

deficit [15-17] or supplementation [18, 19]. Few pragmatic clinical trials have assessed whether

the principles of the Mediterranean diet lower inflammatory markers in a free-living, non-

Mediterranean population with NAFLD, where weight-loss is not the primary outcome and

adherence to the Mediterranean diet is not common. [20]. This lack of evidence significantly

effects translation and implementation of this dietary pattern as a mangagement strategy in

Australia.

To address this gap in the literature, we conducted a randomized controlled trial (RCT) of

a Mediterranean Dietary Intervention for Adults with Non Alcoholic Fatty Liver Disease

(MEDINA), and a detailed trial protocol has been published [21]. Primary findings from the

MEDINA study investigating hepatic steatosis, insulin resistance, body composition and

3
dietary intake have been published elsewhere [22]. To better understand the role of diet as a

mediator of inflammatory processes in a sample of multi-ethnic Australian participants who

significantly increased their level of adherence to a Mediterranean-style dietary patern, we

conducted a secondary analysis of the RCT. We aimed to assess the effects of the 12-week ad

libitum Mediterranean diet (MedDiet) against a standard low-fat diet (LFD) on serum markers

of inflammation in adults with NAFLD, as part of a secondary analysis of our RCT. We also

aimed to determine whether adherence to the MedDiet was associated with a favourable

cardiometabolic and inflammatory profile, and whether changes in MedDiet adherence,

irrespective of diet allocation, were associated with an improvement in markers of

inflammation. It was hypothesised that a MedDiet would be beneficial in improving the

inflammatory profiles in patients with NAFLD, and that greater adherence to the MedDiet

pattern would be associated with greater improvements in inflammatory markers.

2. Methods

2.1. Study Design

This is a secondary analysis of a 12-week multicentre, randomized dietary intervention trial in

adults with NAFLD, with detailed study methods previously published [21] and hepatic and

metabolic outcomes, body composition and dietary intake published elsewhere [22]. The study

was registered with the Australian New Zealand Clinical Trials Registry (ANZCTR) [Trial ID:

ACTRN12615001010583] and designed in accordance with and adherence to the Consolidated

Standards of Reporting Trials (CONSORT) statement [23]. Data presented for this study were

collected at 0-weeks (baseline) and 12-weeks (end-intervention).

2.2. Recruitment and Eligibility Criteria

Adults (>18 years) who were diagnosed with NAFLD as determined by routine ultrasound or

biopsy were recruited to the study. Recruitment sites, dates, processes, and inclusion criteria

4
are detailed in George et al., (2022) [24] and the full exclusion criteria are detailed elsewhere

[21]. Individuals with a diagnosis of insulin dependent diabetes mellitus or those taking

gliclazides were not eligible to participate in this study. Written informed consent was obtained

from elibile particpants recruited to the study. Human research ethics committee approval was

obtained for La Trobe University and the Alfred Health (76/14), Melbourne Health

(HREC/15/MH/283) and Eastern Health (LR31/2015).

2.3. Randomization, Appointment Schedule and Blinding

Participants were randomized to either one of two dietary intervention arms by a researcher

using a 1:1 computer-generated randomization, stratified by sex and presence of diabetes. An

Accredited Practicing Dietitian (APD) was allocated to the LFD group and a separate APD

allocated to the MedDiet group to ensure that there was no cross-over of advice given to either

diet group. Over the 12-week intervention period, each participant attended three face-to-face

dietary consultation appointments (weeks 0, 6, and 12) and received three phone-call follow

up sessions (weeks 2, 4 and 9) with their respective dietitian. Considering that both study

groups involved the active delivery and uptake of dietary interventions, blinding of the study

condition was not possible for both lead researchers (AR, EG), study dietitians and participants.

Outcome assessors were blinded for specific measures (pathology, MR-S and Fibroscan) and

the laboratory staff who analysed inflammatory markers were blinded.

2.4. Dietary Intervention

Details of the diet intervention are described briefly here, and comprehensive details regarding

the dietary development and prescription have been published elsewhere [24]. Participants in

the LFD followed dietary recommendations based on the Australian Dietary Guidelines and

the Heart Foundation [26, 27]. Advice focussed on portion sizes, low-fat options and cooking

methods [26]. The MedDiet intervention was based on a traditional MedDiet as previously

5
published in the MEDINA Study protocol paper [21] and described elsewhere [28]. The

macronutrient composition, reflective of a traditional Cretan Diet but modified for this

Australian cohort, comprised approximately 44% fat (>50% monounsaturated fatty acids), 36%

carbohydrates, 17-20% protein and up to 5% alcohol. Food group recommendations were given

to participants in both diet groups, including daily intake of fruit, vegetables and wholegrains,

and limited intake of processed foods. However, the MedDiet also promoted key dietary

components, such as consumption of extra virgin olive oil, legumes and nuts, fish and

fermented dairy, and a lower consumption of red meat. Both diets were delivered using an ad-

libitum approach.

General physical activity (PA) recommendations were in line with the National Health and

Medical Research Council for participants in both diet intervention groups [29]. No tailored

PA advice was provided to participants.

Participants in the LFD group received three AUD $20 Supermarket gift voucher at the

baseline, mid- and end-intervention appointments to put towards purchasing recommended

food items. Participants in the MedDiet group received a food hamper at the baseline, mid-

intervention and end-intervention appointments containing examples of staple MedDiet foods.

2.5. Demographic Data

During the first face-to-face (0-week) appointment, a trained researcher collected demographic

information from each participant using a self-reported questionnaire. Information regarding

the diagnosis of co-morbidities or pre-baseline results were extracted from patient medical

histories. Medication and supplement use were recorded at baseline, and at each subsequent

face-to-face appointment. Participants were asked to keep medication and supplement intake

consistent over the 12-week intervention period. The National Cholesterol Education Program

6
(NCEP), Adult Treatment Panel III (ATP III) criteria [30] were used to classify participants as

having the MetS.

2.6. Anthropometric, Haemodynamic and Body Composition Measures

Anthropometric parameters (body weight, height, waist and hip circumferences), systolic and

diastolic blood pressure were measured at 0- and 12-weeks according to standard methods.

BMI was calculated as body weight divided by height squared (kg/m2); World Health

Organization reference ranges were used [31]. Body composition was also assessed at 0- and

12-weeks using Seca© Bioelectrical Impedance Analysis (BIA) scales. Parameters reported in

this study are fat mass in kilograms (kg) or percent (%) and visceral fat in litres (L).

2.7. Primary Outcome Measures – Inflammatory Markers

The primary outcome of the overall MEDINA trial was a reduction in homeostasis model

assessment of insulin resistance (HOMA-IR) [21] and results for this outcome are published

elsewhere [22]. The primary outcomes investigated in this study were inflammatory cytokines;

high sensitivity (hs)-CRP, TNF-α, and IL-6, and adipokines; adiponectin, leptin and resistin,

collected and measured at baseline and week 12. Data for hs-CRP has been presented in our

previous study and is re-presented here due to inclusion as both a metabolic outcome and an

inflammatory outcome, and to ensure the full inflammatory profile that was collected is

presented here. Blood samples were collected using SST vacutainers and after centrifugation

sera was separated and frozen at -80°C. Cytokine and adipokine markers were later analysed

using multiplex immunoassay kits (Millipore Corp., Billerica, MD, USA) which

simultaneously measured serum concentrations of cytokines (interleukin-6 (IL-6) and tumor

necrosis factor alpha (TNF-α), cat num: HSTCMAG-28SK) and metabolic hormones

(Adiponectin, Resistin and Leptin, cat num: HMHEMAG-34K) as previously described [32].

7
The assay was performed according to the manufacturer’s instructions and all samples were

run in duplicate.

2.8. Secondary Outcome Measures

Fasting blood samples were also used to measure blood glucose and insulin, and to calculate

Homeostatic model assessment-Insulin Resistance (HOMA-IR) [33]. Proton magnetic

resonance spectroscopy (1H-MRS) was used to measure intrahepatic lipid (IHL) concentration

as detailed elsewhere [21]. Transient elastography (TE) FibroscanTM was used to measure liver

stiffness (in Kilo Pasqual’s (kPa)) to estimate the likelihood offibrosis.

2.9. MEDAS Score

Adherence to the MedDiet was assessed using a validated 14-point checklist containing key

foods, beverages and practices characteristic of a traditional Mediterranean diet [35]. One point

is allocated for each checklist item and higher overall score indicates greater adherence. A

separate 9-item checklist was used to measure for adherence to the LFD [34]. Diet adherence

checklists were completed at baseline and week 12, and reviewed by an APD to minimise

errors. MedDiet adherence scores were calculated retrospectively for participants in the LFD

group to allow for additional pooled analyses. Results for the MEDAS score was presented in

our previous publication [22] and was included in this analysis due to its direct applicability to

the study aims.

2.10. Statistical Analyses

Outcome data available for participants at each timepoint were used in all analyses. Any

participants who had a measurement of >10 mg/L for hs-CRP (indicating likely acute

inflammation) were excluded from CRP analysis [35]. Statistical analyses were conducted

using Stata/SE 16.1 (StataCorp®, TX) and statistical significance was set at p<0.05.

8
For between-group analyses of the LFD and MedDiet groups, restricted maximum likelihood

linear mixed models, with observations (0- and 12-weeks) nested within individuals, were

fitted for each outcome. Models included a term to estimate group differences at 12-weeks and

fixed effects of time [36]. Baseline concentration of the outcome(s) were adjusted via

specification of an unstructured residual covariance matrix, and models were adjusted for

baseline body weight. Initial testing of models was conducted using both raw outcome

variables and their log-transformed counterparts, and plots of residual normality and

heterogeneity of variance were inspected to determine if log-transformations were warranted

to meet parametric assumptions. Consequently, between-group analyses included hs-CRP,

TNF-α and resistin in their raw form and log-transformed variables for IL-6, adiponectin, and

leptin. Sensitivity of between-group findings to influential observations was examined by

inspection of DFBETA scores (values which indicate how much estimates differ if an

observation is excluded) for each model. Where potentially influential observations were

identified, the model was refitted with them excluded and results were compared to the primary

analyses.

For within dietary intervention group analyses between the baseline and end-intervention

timepoints, paired samples t-tests were used for parametric data and Wilcoxon Signed Rank

Tests for non-parametric data. To assess the impact of adherence to the MedDiet, MEDAS

scores for the combined cohort were divided into two categories labelled “higher adherence”

and “lower adherence” derived from above and below mean values. Mean ± SD values of key

variables were assessed between levels of adherence. The two groups were compared using

independent samples t-tests for parametric data and Mann-Whitney U Tests for non-parametric

data. Finally, fixed effects models, with data from both treatment groups pooled, were fitted to

examine associations between concurrent changes in MedDiet scores and inflammatory

9
markers from 0- to 12-weeks [37]. Sensitivity analyses for influential observations as indicated

by DFBETA scores were performed for these fixed effects models.

2.11.Power Calculation

The overarching MEDINA study was powered based on IHL summary data published in the

Study Protocol paper [21]. The sample size was 17 per group (power of 80%, α<0.05),

adjusting for 20% dropout, 17/0.8 = 21 participants per group. The total sample size of 42

participants (without dropouts) was achieved.

Inflammatory outcomes were a secondary outcome for the overarching study and given the

absence of literature in the area this analysis was considered novel and warranted. As such, an

a priori sample size calculation, powered to see a change in each inflammatory cytokine and

adipokine, was conducted prior to analysis. This calculation was based on the results from

previous studies [13, 38-40] and used the statistical software program G*Power 3.0.10. The

estimated sample size required to detect meaningful differences between groups with 80%

power (α=0.05) varied 120-350 participants, depending on the sensitivity of each inflammatory

marker. Secondary power calculations are available in Supplementary Figure 1.

3. Results

3.1. Participant Characteristics at Baseline

Table 1 presents the baseline characteristics and inflammatory profile of the 42 participants

recruited to the MEDINA Study. Of these, 18 participants (43%) had diagnosed T2DM, and

23 (55%) met the NCEP ATP III [30] criteria for the MetS. Thirty-five participants (83%)

reported prescription medication use at baseline, and 24 participants (57%) reported

supplement intake. Medication and supplement intake remained consistent over the 12-week

intervention. There were a higher proportion of females and participants were on average obese

10
with abdominal obesity, based on World Health Organization classicifation of BMI and waist

circumference [31, 41].

Participants were randomized to either the LFD (n=23) or MedDiet (n=19) intervention groups.

A total of three participants withdrew from the study (7% attrition rate), due to personal

circumstances. The participant study flow diagram is shown in Figure 1; three participants at

baseline and six participants at end intervention had a measurement >10 mg/L for hs-CRP and

were excluded from hs-CRP analysis [35].

3.2. Inflammatory Markers and Anthropometry

Differences in inflammatory markers at 12-weeks, while adjusting for baseline concentrations,

were assessed and compared between groups (Table 2). Adjusted for baseline leptin

concentration and body weight, leptin was significantly higher at 12-weeks in participants in

the MedDiet group when compared to LFD group (B=0.27 [95% CI: 0.07, 0.48], p=0.010).

Findings from sensitivity analyses were consistent with those from the primary analyses

(Supplementary Table 1).

Both dietary intervention groups had an increased concentration of serum adiponectin over

time, however this change was only statistically significant within the MedDiet group (13.7 ±

9.2 µg/mL to 17.0 ± 12.5 µg/mL, p =0.016). There were no significant changes in hs-CRP,

TNF-α or IL-6 in either group (p >0.05).

The effect of the dietary interventions on anthropometry and body composition are detailed in

our previous paper [24]. Briefly, weight, BMI and waist circumference were not significantly

different from pre- to post intervention within each diet group. Visceral fat was reduced

significantly in both groups; LFD ([log scale] -76%, p = <.0005), MedDiet (−61%, p = <.0005)

although there were no between-group differences [24].

3.3. Dietary Adherence

11
From baseline to end intervention, dietary adherence significantly improved in the MedDiet

group by an average 2.7 units (out of a maximum possible score of 14) (p <0.001). Adherence

to the LFD for the LFD group improved by 1.4 units (out of a maximum possible score of 9)

(p=0.035). Given the overlap in dietary prescription MedDiet adherence was also assessed in

the LFD group, whose participants increased their MEDAS scores from baseline to end-

intervention by 1.4 units (p <0.001).

Given that adherence to the MedDiet pattern significantly improved in both diet groups

between the baseline and week 12 timepoints, data for the entire cohort was pooled to evaluate

whether the overall improvement in adherence to a MedDiet was associated with an

improvement in inflammatory outcomes.

3.4. Baseline Adherence to the MedDiet Pattern

When assessed at baseline, the mean MEDAS score for the pooled cohort was 5.7 ± 2.0,

indicating low adherence (Table 3). Only one participant met the acceptable adherence

criterion (9/14) [42]. This mean score at baseline was used to classify participants into low

and high adherence to MedDiet; 0-5, and 6-14, respectively. TNF-α was significantly lower for

participants with higher MedDiet adherence (p=0.05). All other cytokines were lower in

participants with higher MedDiet adherence scores, albeit not significantly. Participants with

higher MedDiet adherence had significantly lower body weight (-8.3 kg, p=0.021), though

there was not a significant difference for BMI.

3.5. Baseline to 12-week Change in Adherence to the MedDiet

In the total cohort (n=39 completers), MedDiet adherence scores increased significantly from

baseline to end-intervention (5.64 ± 2.02 to 7.67 ± 2.23, p =0.0005). Individual components of

the MedDiet that significantly increased from baseline to end-intervention for the pooled/entire

cohort are repored in Supplememtary Figure 2.

12
 3.6. Association between change in Adherence to the Med Diet and Inflammatory

Outcomes

Fixed effects models, using the combined cohort, did not show any statistically significant

associations between concurrent changes in adherence to MedDiet and inflammatory markers

over time (baseline to end intervention) (Supplementary Table 2). These results were not

found to be sensitive to influential observations. Changes in inflammatory makers and MedDiet

adherence for each participant are presented in Supplementary Figure 3.

4. Discussion

The aim of this study was to determine the effects of an ad libitum Mediterranean versus low-

fat diet on inflammation in individuals with NAFLD. The results showed the participants in

the MedDiet group had significantly improved concentrations of circulating adiponectin. In our

previous publication the MedDiet group did not experience significant weight loss, therefore

these improvements in adiponectin are in the absence of weight change [24]. Despite

significantly improved adherence to a Mediterranean-style diet, there was no significant effect

of the MedDiet or LFD on markers hs-CRP, TNF-α, IL-6, leptin or resistin. This study

identified that participants who were more adherent to the MedDiet pattern prior to any dietary

intervention had significantly lower body weight and serum TNF-α than those with lower

adherence. There was no clear correlation between unit increases in MedDiet adherence

(MEDAS Score) and change in inflammatory markers from baseline to end-intervention, likely

due to the small sample size.

To our knowledge, two other RCTs have observed significant improvements in

adiponectin in individuals with NAFLD following a Mediterranean-type die, both of which did

not administer traditional Cretan diets. One study supplemented olive oil enriched with omega-

3 (n-3) polyunsaturated fatty acid (PUFA) alongside the MedDiet for one year [43] and the

13
other applied an energy restriction of 30% total energy to the MedDiet prescription for two

years [16]. These studies highlight the benefits of long-term adherence to the MedDiet

combined with the capacity for increasing active key components of the MedDiet which may

directly influence molecular events that alter adipokine synthesis and interrelated lipid and

glucose metabolism [44]. Monounsaturated and polyunsatured fatty acids are ligands of

peroxisome proliferation activated receptor γ (PPARγ) which stimulate the secretion and

expression of adiponectin.[45] Long-chain polyunsaturated fatty acids – docosahexaenoic acid

(DHA) and eicosapentaenoic acid (EPA) – are important structural compoments which increase

adiponectin concentration by mediating mechanisms of the activated PPARγ receptors [46].

The increase in adiponectin mediates fatty acid oxidation which can reduce triglyceride storage

in skeletal muscle whilst increasing high-density lipoprotein via activation of PPARα and

reduced catabolism of apolipoprotein A-I [47]. In agreement with this, a recent meta-analysis

found that n-3 PUFA supplementation, especially DHA, has a favorable effect in the treating

NAFLD by modulating lipid metabolism, enhancing fatty acid oxidation, and decreasing de

novo lipogenesis [48].

Despite satisfying adherence with the MedDiet program and adequate diet quality in

both MedDiet and LFD groups, the Mediterranean diet had no impact on the concentrations of

hs-CRP, TNF-α, IL-6, leptin or resistin. This result may reflect the true lack of effect of the

Mediterranean diet on the selected risk markers or may indicate that dietary changes made by

participants in this sample were not extreme enough to produce a dramatic shift in markers. Of

the few RCTs that have tested the effect of a Mediterranean-type diet on cytokine markers in

NAFLD, two interventions applied energy restriction resulting in a significant decrease in hs-

CRP at 12-weeks [17] and 6-months [15]. Another study observed a significant reduction in

TNF-α following a 12-week MedDiet plus bergamot polyphenolic fraction (BPF) and Cynara

Cardunculus (CyC) extract, but not the MedDiet alone [18]. Nutraceuticals are increasingly

14
tested alongside diet to investigate the added effect of biologically active compounds outcomes

of NAFLD [49]. Cellular models for BPF and CyC show dietary polyphenols in the extract

counteract reactive oxygen species production and associated oxidative damage by cintrilling

mitochondrial memberane potential and oxidative phosphorylation. [18] Dietary polyphenols

may also indirectly up-regulate endogenous antioxidant defences. [18]

The MedDiet patten is characterised as a high polyphenol diet as key componenets

including extra-virgin olive oil, nuts, red wine, legumes, vegetables, fruits, and whole-grain

cereals contain phenolic compounds [50]. The analysis of individual diet items (food groups)

on the MEDAS checklist (supplementary material figure 2) shows a significant increase for

the aforementioned components, though adherence did not increase above ~70% (of

participants adhering) for any of the items. Additionally, important components of the MedDiet

are diversity of vegetables consumed (particularly tomatoes and dark leafy greens), onion,

herbs and spices [51]. These vegetables and condiments enhance concentrations of vitamins,

especially vitamins C and A; minerals, especially electrolytes; and phytochemicals, especially

antioxidants and carotenoids; in the dietary pattern [52]. Isolated effects of herbs and spices

can result in significant improvements in markers of inflammation and oxidative stress via

dietary polyphenol content [53]. Single molecules do not work efficiently in NAFLD, the

synergic effect of all these polyphenols would represent a novel approach in reducing

inflammation. In the present study, dietary advice, diet records and the MEDAS score may not

have accurately captured these components. Moreover, the bioavailability of compounds in

whole foods may be altered by the physical property or cooking practice and these interactions

are difficult to study in humans [52]. Although adherence data was cross-checked against food

records and dietary intake data, these figures may be subject to performance or social

desirability bias, as is a common flaw of most self-reported scales [54]. Perhaps a greater true

15
increase in adherence was necessary to produce a reduction in the selected cytokines and

adipokines.

A main feature of our study was the ad libitum, non-restrictive approach to intervening

with diet to increase acceptability of the MedDiet in a multi-ethnic Australian population.

Although target macronutrient distribution was identified for both dietary approaches, this was

not delivered to participans in a prescriptive or controlled way. In NAFLD, clinical trials that

have produced the greatest change in inflammation [55], hepatic steatosis and insulin resisance

[56] have been restrictive and/or controlled-feeding trials. These studies are limited by the short

(6- or 12-week) intervention period and no long-term follow up. These results may be

indicative that an alternate approach is needed. For example, a controlled, higher intensity

intervention period may first be administered to produce a significant change in clinical

outcomes, followed by a longer term ad libitum stage during which dietary coaching focusses

on sustaining changes by implementing techniques linked to the Health Belief Model to centre

the intervention on the partcipants’ challenges and behaviour change.

Limitations of this study include an underpowered sample size. Furthermore, dietary

data collection always holds the potential for under-reporting or over-reporting in self-reported

questionnaires (MEDAS checklist) due to increased dietary awareness. Efforts to minimise this

in this study were employed, whereby data was thoroughly checked by an APD during each

appointment. This study contained two experimental groups, with the LFD was intended to

represent “standard outpatient care” in Australia, the frequency and length of visits with the

APD in this study did not reflect a true control or standard care patient group as the

interventions were matched to reflect contact with a clinician and study intensity. The MedDiet

intervention contained approximately 44% fat, which may be considered a limitation in the

treatment of NAFLD. Outpatient tertiary hospital access to a clinical dietitian for this

16
participant group would more realistically be approximately 30-minute consultation every

three months. Future research should consider direct imaging methodologies such as CT or

MRI to verify the current findings [24] which inferred abdominal adiposity including visceral

fat via indirect BIA and waist circumference methods. Another wider limitation for the

implication of these findings is that at present, reference ranges for inflammatory markers are

not available in patients with NAFLD and large population based data sets which measure

cytokine and adipokine concentrations in Mediterranean countries are not reliable for

comparison.

In this study, participants with NAFLD in the MedDiet group experienced significant

improvements in serum adiponectin concentrations in the absence of weight loss, whereas these

improvements were not significant in the LFD group. Despite an improvement in diet quality

and adherence to a Mediterranean-style diet, hs-CRP, TNF-α, IL-6, leptin and resistin did not

change. The small sample size and non-restrictive nature of our intervention may be an

explanation for the non-sigificant results observed. Future studies should be adequately

powered with a longer duration of intervention to determine significant changes in

inflammatory markers. Particular focus on diet composition and functional components which

produce physiological benefits for individuals with NAFLD will assist the provision of

effective intervention strategies.

17
Acknowledgment: none

Sources of support: This project was funded by a La Trobe University Research Focus Area

‘Understanding Disease’ grant. The analysis of inflammatory cytokine and adipokine markers

was funded by an Australasian Society of Parenteral and Enteral Nutrition small project grant.

This work was supported by an Australian Government Research Training Program

Scholarship (AR, ESG). Food was donated for study participant hampers from Boundary Bend,

Almond Board of Australia, Jalna, HJ Heinz, Simplot and Carmen’s, these companies had no

input into the trial design or delivery.

Author Contributions: ESG and ACT conceptualised the study. ESG designed the

intervention, ESG and AR collected and analysed data. AR, PDG and SM analysed the primary

outcomes. AR drafted the manuscript. All authors provided intellectual input into the

manuscript and approved the final version.

Author Declarations: All authors have no perceived conflict of interest to declare. The authors

have noting to disclose.

18
REFERENCES

[1] Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M. Global epidemiology of
nonalcoholic fatty liver disease—Meta‐analytic assessment of prevalence, incidence, and outcomes.
Hepatology. 2016;64:73-84. doi: 10.1002/hep.28431.
[2] Vernon G, Baranova A, Younossi ZJAp, therapeutics. Systematic review: the epidemiology and natural
history of non‐alcoholic fatty liver disease and non‐alcoholic steatohepatitis in adults. Aliment Pharmacol
Ther. 2011;34:274-85. doi: 10.1111/j.1365-2036.2011.04724.x
[3] Neuschwander‐Tetri BA, Caldwell SHJH. Nonalcoholic steatohepatitis: summary of an AASLD Single
Topic Conference. Hepatology. 2003;37:1202-19. doi: 10.1053/jhep.2003.50193.
[4] Lonardo A, Nascimbeni F, Maurantonio M, Marrazzo A, Rinaldi L, Adinolfi LEJWjog. Nonalcoholic
fatty liver disease: Evolving paradigms. World J Gastroenterol. 2017;23:6571. doi:
10.3748/wjg.v23.i36.6571.
[5] Byrne CD, Targher G. NAFLD: a multisystem disease. J Hepatol. 2015;62:S47-S64. doi:
10.1016/j.jhep.2014.12.012.
[6] Buzzetti E, Pinzani M, Tsochatzis EA. The multiple-hit pathogenesis of non-alcoholic fatty liver
disease (NAFLD). Metabolism. 2016;65:1038-48. doi: 10.1016/j.metabol.2015.12.012
[7] Tilg H, Moschen AR. Insulin resistance, inflammation, and non-alcoholic fatty liver disease. Trends
Endocrinol. Metab. 2008;19:371-9. doi: 10.1016/j.tem.2008.08.005.
[8] Ajmal MR, Yaccha M, Malik MA, Rabbani M, Ahmad I, Isalm N, et al. Prevalence of nonalcoholic
fatty liver disease (NAFLD) in patients of cardiovascular diseases and its association with hs-CRP and
TNF-α. IHJ. 2014;66:574-9. doi: 10.1016/j.ihj.2014.08.006
[9] Richard C, Couture P, Desroches S, Lamarche B. Effect of the Mediterranean diet with and without
weight loss on markers of inflammation in men with metabolic syndrome. Obesity. 2013;21:51-7. doi:
10.1002/oby.20239.
[10] Milić S, Lulić D, Štimac D. Non-alcoholic fatty liver disease and obesity: biochemical, metabolic and
clinical presentations. World J Gastroenterol. 2014;20:9330. doi: 10.3748/wjg.v20.i28.9330.
[11] Ahluwalia N, Andreeva V, Kesse-Guyot E, Hercberg SJD, metabolism. Dietary patterns,
inflammation and the metabolic syndrome. Diabetes Metab J. 2013;39:99-110. doi:
10.1016/j.diabet.2012.08.007.
[12] Liver EAftSot, Diabetes EAftSo. EASL-EASD-EASO Clinical practice guidelines for the
management of non-alcoholic fatty liver disease. 2016;9:65-90. doi: 10.1016/j.jhep.2015.11.004.
[13] Reddy AJ, George ES, Roberts SK, Tierney AC. Effect of dietary intervention, with or without co-
interventions, on inflammatory markers in patients with nonalcoholic fatty liver disease: a systematic
literature review. Nutr. Rev. 2019;77:765-86. doi: 10.1093/nutrit/nuz029.
[14] Hall R, George ES, Tierney AC, Reddy AJ. Effect of dietary intervention, with or without co-
interventions, on inflammatory markers in patients with nonalcoholic fatty liver disease: a systematic
literature review and meta-analysis. Adv in Nutr. 2023;14;475-499. doi: 10.1016/j.advnut.2023.01.001.

19
[15] Marin-Alejandre BA, Abete I, Cantero I, Monreal JI, Elorz M, Herrero JI, et al. The metabolic and
hepatic impact of two personalized dietary strategies in subjects with obesity and nonalcoholic fatty liver
disease: the Fatty Liver in Obesity (FLiO) randomized controlled trial. Nutrients.
2019;11:22.10.3390/nu11102543. doi: 10.3390/nu11102543
[16] Marin-Alejandre BA, Cantero I, Perez-Diaz-Del-Campo N, Monreal JI, Elorz M, Herrero JI, et al.
Effects of two personalized dietary strategies during a 2-year intervention in subjects with nonalcoholic
fatty liver disease: a randomized trial. Liver Int. 2021;41:1532-44.10.1111/liv.14818. doi:
10.1111/liv.14818
[17] Ristic-Medic D, Kovacic M, Takic M, Arsic A, Petrovic S, Paunovic M, etdal. Calorie-restricted
Mediterranean and low-fat diets affect fatty acid status in individuals with nonalcoholic fatty liver disease.
Nutrients. 2020;13:15.10.3390/nu13010015. doi: 10.3390/nu13010015.
[18] Ferro Y, Montalcini T, Mazza E, Foti D, Angotti E, Gliozzi M, et al. Randomized clinical trial:
bergamot citrus and wild cardoon reduce liver steatosis and body weight in non-diabetic individuals aged
over 50 years. Front Endocrinol. 2020;11:494. doi: 10.3389/fendo.2020.00494
[19] Sofi F, Giangrandi I, Cesari F, Corsani I, Abbate R, Gensini GF, et al. Effects of a 1-year dietary
intervention with n-3 polyunsaturated fatty acid-enriched olive oil on non-alcoholic fatty liver disease
patients: a preliminary study. Int J Food Sci Nutr. 2010;61:792-802. doi: 10.3109/09637486.2010.487480.
[20] Martínez-González MÁ, Hershey MS, Zazpe I, Trichopoulou A. Transferability of the Mediterranean
diet to non-Mediterranean countries. What is and what is not the Mediterranean diet. Nutrients.
2017;9:1226. doi: 10.3390/nu9111226.
[21] Papamiltiadous ES, Roberts SK, Nicoll AJ, Ryan MC, Itsiopoulos C, Salim A, et al. A randomised
controlled trial of a Mediterranean dietary intervention for adults with non alcoholic fatty liver disease
(MEDINA): study protocol. BMC Gastroenterol. 2016;16:14. doi: 10.1186/s12876-016-0426-3.
[22] George ES, Reddy A, Nicoll AJ, Ryan MC, Itsiopoulos C, Abbott G, et al. Impact of a Mediterranean
diet on hepatic and metabolic outcomes in non-alcoholic fatty liver disease: The MEDINA randomised
controlled trial. Liver Int. 2022;42:1308-22. doi: 10.1111/liv.15264.
[23] Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting
parallel group randomized trials. Ann Intern Med. 2010;152:726-32. doi: 10.4103/0976-500X.72352.
[24] George ES, Reddy A, Nicoll AJ, Ryan MC, Itsiopoulos C, Abbott G, et al. Impact of a Mediterranean
diet on hepatic and metabolic outcomes in non‐alcoholic fatty liver disease: The MEDINA randomised
controlled trial. Liver Int. 2022; 42:1308-1322. doi: 10.1111/liv.15264.
[25] Papamiltiadous ES, Roberts SK, Nicoll AJ, Ryan MC, Itsiopoulos C, Salim A, et al. A randomised
controlled trial of a Mediterranean dietary intervention for adults with non alcoholic fatty liver disease
(MEDINA): study protocol. BMC Gastroenterol. 2016;16:14. doi: 10.1186/s12876-016-0426-3.
[26] National Health and Medical Research Council. Australian Dietary Guidelines. Canberra, Australia;
2013.

20
[27] Chew DP, Scott IA, Cullen L, French JK, Briffa TG, Tideman PA, et al. National Heart Foundation of
Australia and Cardiac Society of Australia and New Zealand: Australian clinical guidelines for the
management of acute coronary syndromes 2016. MJA. 2016;205:128-33. doi: 10.1016/j.hlc.2016.06.789.
[28] George E, Kucianski T, Mayr H, Moschonis G, Tierney A, Itsiopoulos C. A Mediterranean diet model
in Australia: strategies for translating the traditional Mediterranean diet into a multicultural setting.
Nutrients. 2018;10:465. doi: 10.3390/nu10040465.
[29] Brown WJ, Bauman AE, Bull F, Burton NW. Development of evidence-based physical activity
recommendations for adults (18-64 years). Report prepared for the Australian Government Department of
Health, August 2012.
[30] Thomas GN, Ho S-Y, Janus ED, Lam KS, Hedley AJ, Lam TH, et al. The US national cholesterol
education programme adult treatment panel III (NCEP ATP III) prevalence of the metabolic syndrome in a
Chinese population. Diabetes Res Clin Pract. 2005;67:251-7. doi: 10.1016/j.diabres.2004.07.022.
[31] World Health Organization. Physical status: The use of and interpretation of anthropometry, Report of
a WHO Expert Committee. 1995.
[32] Gabel L, Ridgers ND, Della Gatta P, Arundell L, Cerin E, Robinson S, et al. Associations of sedentary
time patterns and TV viewing time with inflammatory and endothelial function biomarkers in children.
Pediatr Obes. 2016;11:194-201. doi: 10.1111/ijpo.12045.
[33] Keskin M, Kurtoglu S, Kendirci M, Atabek ME, Yazici C. Homeostasis model assessment is more
reliable than the fasting glucose/insulin ratio and quantitative insulin sensitivity check index for assessing
insulin resistance among obese children and adolescents. Pediatrics. 2005;115:500-503. doi:
10.1542/peds.2004-1921.
[34] Estruch R, Ros E, Salas-Salvadó J, Covas M-I, Corella D, Arós F, et al. Primary prevention of
cardiovascular disease with a Mediterranean diet supplemented with extra-virgin olive oil or nuts. N Engl J
Med 2018; 378:34. doi: 10.1056/NEJMoa1800389.
[35] Pearson TA, Mensah GA, Alexander RW, Anderson JL, Cannon RO, Criqui M, et al. Markers of
inflammation and cardiovascular disease: application to clinical and public health practice: A statement for
healthcare professionals from the Centers for Disease Control and Prevention and the American Heart
Association. Circ J. 2003;107:499-511. doi: 10.1161/01.cir.0000052939.59093.45.
[36] Liang K-Y, Zeger SL. Longitudinal data analysis of continuous and discrete responses for pre-post
designs. Sankhyā: The Indian Journal of Statistics, Series B. 2000:134-48.
[37] Gunasekara FI, Richardson K, Carter K, Blakely T. Fixed effects analysis of repeated measures data.
Int J Epidemiol. 2014;43:264-9. doi: 130.194.114.109.
[38] Sofi F, Giangrandi I, Cesari F, Corsani I, Abbate R, Gensini GF, et al. Effects of a 1-year dietary
intervention with n-3 polyunsaturated fatty acid-enriched olive oil on non-alcoholic fatty liver disease
patients: a preliminary study. Int J Food Sci Nutr. 2010;61:792-802. doi: 10.3109/09637486.2010.487480.
[39] Markova M, Pivovarova O, Hornemann S, Sucher S, Frahnow T, Wegner K, et al. Isocaloric diets
high in animal or plant protein reduce liver fat and inflammation in individuals with type 2 diabetes.
Gastroenterology. 2017;152:571-85. doi: 10.1053/j.gastro.2016.10.007.

21
[40] Spadaro L, Magliocco O, Spampinato D, Piro S, Oliveri C, Alagona C, et al. Effects of n-3
polyunsaturated fatty acids in subjects with nonalcoholic fatty liver disease. Dig Liver Dis. 2008;40:194-9.
doi: 10.1016/j.dld.2007.10.003.
[41] Organization WH. Obesity: preventing and managing the global epidemic. 2000.
[42] Martínez-González MA, García-Arellano A, Toledo E, Salas-Salvado J, Buil-Cosiales P, Corella D, et
al. A 14-item Mediterranean diet assessment tool and obesity indexes among high-risk subjects: the
PREDIMED trial. PloS one. 2012;7:43134. doi: 10.1371/journal.pone.0043134.
[43] Sofi F, Giangrandi I, Cesari F, Corsani I, Abbate R, Gensini GF, et al. Effects of a 1-year dietary
intervention with n-3 polyunsaturated fatty acid-enriched olive oil on non-alcoholic fatty liver disease
patients: a preliminary study. Int J Food Sci Nutr. 2010;61:792-802. doi: 10.3109/09637486.2010.487480.
[44] Cassidy A, Skidmore P, Rimm EB, Welch A, Fairweather-Tait S, Skinner J, et al. Plasma adiponectin
concentrations are associated with body composition and plant-based dietary factors in female twins. J
Nutr. 2008;139:353-358. doi: 10.3945/jn.108.098681.
[45] Izadi V, Azadbakht L. Specific dietary patterns and concentrations of adiponectin. J Res Med Sci
2015;20:178-84.
[46] Yamauchi T, Kamon J, Minokoshi Ya, Ito Y, Waki H, Uchida S, et al. Adiponectin stimulates glucose
utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat Med. 2002;8:1288-95.
doi: 10.1038/nm788.
[47] Verges B, Petit JM, Duvillard L, Dautin G, Florentin E, Galland F, et al. Adiponectin is an important
determinant of apoA-I catabolism. Arterioscler Thromb Vasc Biol. 2006;26:1364-9. doi:
10.1161/01.ATV.0000219611.50066.bd.
[48] Guo XF, Yang B, Tang J, Li D. Fatty acid and non-alcoholic fatty liver disease: meta-analyses of
case-control and randomized controlled trials. Clin Nutr. 2018;37:113-22. doi: 10.1016/j.clnu.2017.01.003.
[49] Simon J, Casado-Andres M, Goikoetxea-Usandizaga N, Serrano-Macia M, Martinez-Chantar ML.
Nutraceutical properties of polyphenols against liver diseases. Nutrients. 2020;12:15. doi:
10.3390/nu12113517.
[50] Castro-Barquero S, Lamuela-Raventós RM, Doménech M, Estruch R. Relationship between
Mediterranean dietary polyphenol intake and obesity. Nutrients. 2018;10:1523. doi: 10.3390/nu10101523.
[51] George ES, Kucianski T, Mayr HL, Moschonis G, Tierney AC, Itsiopoulos C. A Mediterranean diet
model in Australia: Strategies for translating the traditional Mediterranean diet into a multicultural detting.
Nutrients. 2018;10:465. doi: 10.3390/nu10040465.
[52] Slavin JL, Lloyd B. Health benefits of fruits and vegetables. Adv Nutr. 2012;3:506-516. doi:
10.3945/an.112.002154.
[53] Farzaei MH, Zobeiri M, Parvizi F, El-Senduny FF, Marmouzi I, Coy-Barrera E, et al. Curcumin in
liver diseases: A systematic review of the cellular mechanisms of oxidative stress and clinical perspective.
Nutrients. 2018;10:855. doi: 10.3390/nu10070855.
[54] Rosenman R, Tennekoon V, Hill LG. Measuring bias in self-reported data. Int J Behav Healthc Res.
2011;2:320-32. doi: 10.1504/IJBHR.2011.043414.

22
[55] Medrano M, Cadenas-Sanchez C, Alvarez-Bueno C, Cavero-Redondo I, Ruiz JR, Ortega FB, et al.
Evidence-based exercise recommendations to reduce hepatic fat content in youth- a systematic review and
meta-analysis. Prog Cardiovasc Dis. 2018;61:222-231. doi: 10.1016/j.pcad.2018.01.013.
[56] Ryan MC, Itsiopoulos C, Thodis T, Ward G, Trost N, Hofferberth S, et al. The Mediterranean diet
improves hepatic steatosis and insulin sensitivity in individuals with non-alcoholic fatty liver disease. J
Hepatol. 2013;59:138-143. doi: 10.1016/j.jhep.2013.02.012

Figure 1. Flow diagram of study participants with Non-Alcoholic Fatty Liver Disease (NAFLD) and

primary outcomes collected at each timepoint through the study.

23
Table 1. Baseline characteristics and inflammatory profile of participants with Non-Alcoholic Fatty
Liver Disease (NAFLD) enrolled in this randomized controlled trial (RCT) (n=42)

Characteristics Mean ± SD
Age (years) 52.3 ± 12.6
Sex, females (n, %) 25, 60%
Body Mass Index (kg/m2) 32.2 ± 6.3
Waist Circumference (cm) 103.5 ± 21.3
Systolic Blood Pressure (mmHg) 126.5 ± 16.3
Diastolic Blood Pressure (mmHg) 83.1 ± 8.6
Inflammatory markers
hs-CRP (0-3 mg/L) 2.4 ± 4.2
TNF-α (pg/ml) 4.4 ± 3.0
IL-6 (pg/ml) 3.4 ± 11.8
Adiponectin (µg/mL) 11.3 ± 13.9
Resistin (ng/mL) 36.3 ± 15.9
Leptin (ng/ml) 12.5 ± 16.3
Abbreviations: hs-CRP, high sensitivity C-reactive protein; TNF-α, tumour necrosis factor-alpha; IL-6,
interleukin-6.

24
Table 2. Descriptive statistics (mean ± standard deviation) of participants with Non-Alcoholic Fatty Liver Disease (NAFLD), within groups, for inflammatory
markers in the two study groups (Low-Fat Diet versus MedDiet) at baseline and end-intervention timepoints.

LFD Group MedDiet Group Estimated Group Differencea

End End
Baseline LFD group Baseline MedDiet group
Intervention Intervention B (95% CI) p-value
(n=22) p-value* (n=19) p-value*
(n=19) (n=18)
Inflammatory Markers

hs-CRP (mg/L) 3.8 ± 2.7 3.5 ± 2.6 0.33 2.6 ± 2.8 2.2 ± 1.9 0.81 0.20 (-0.81, 1.21) 0.69
TNF-α (pg/mL) 6.1 ± 6.4 4.5 ± 2.0 0.23 3.8 ± 2.0 4.1 ± 1.9 0.51 0.10 (-0.93, 1.14) 0.84
IL-6 (pg/mL) 7.9 ± 9.5 11.4 ± 14.3 0.68 15.0 ± 24.3 16.6 ± 25.6 0.076 -0.01 (-0.44, 0.42) 0.96
Adiponectin (µg/mL) 17.3 ± 14.4 19.5 ± 21.0 0.82 13.7 ± 9.2 17.0 ± 12.6 0.016 0.20 (-0.26, 0.67) 0.38
Resistin (ng/mL) 41.6 ± 22.6 39.0 ± 20.9 0.50 37.6 ± 13.1 39.9 ± 15.3 0.99 2.77 (-4.24, 9.79) 0.43
Leptin (ng/mL) 18.7 ± 13.6 16.7 ± 13.1 0.10 13.5 ± 9.9 13.7 ± 10.2 0.47 0.27 (0.07, 0.48) 0.010
MEDAS Dietary
5.0 ± 1.8 6.4 ± 1.6 <0.001 6.5 ± 2.0 9.2 ± 1.9 <0.001
Adherence Score (0-14)
LFD Dietary
5.4 ± 2.0 6.4 ± 2.3 0.035
Adherence Score (0-9)
aGroup differences for each outcome were estimated using restricted maximum-likelihood linear mixed models using outcome data from all available time points.
The models were adjusted for baseline body weight. *Within-group changes from baseline to post-intervention. Significance p <0.05. Abbreviations: MedDiet;
Mediterranean Diet; LFD, Low-Fat Diet; hs-CRP, high sensitivity C-reactive protein; TNF-α, tumour necrosis factor-alpha; IL-6, interleukin-6.

25
Table 3. Baseline characteristics of the Non-Alcoholic Fatty Liver Disease (NAFLD) cohort according
to Med Diet Score (n=42)

Characteristics Med Diet Score at Baseline

Lower Adherence Higher Adherence
(0 – 5) (6 – 14) p-value
(n=20) (n=21)
MEDAS Score 4.0 ± 1.12 7.3 ± 1.1 0.001
Demographics
Age (years) 50.5 ± 16.2 54.6 ± 8.1 0.316
Sex, females (n, %) 13, 65% 11, 52% 0.425
Diabetes n (%) 10 (50%) 7 (33%) 0.291
Inflammatory Markers
hs-CRP (0-3 mg/L) 3.6 ± 2.5 2.7 ± 2.4 0.252

TNF-α (pg/ml) 6.6 ± 6.7 3.5 ± 1.7 0.054

IL-6 (pg/ml) 13.5 ± 21.1 9.3 ± 15.6 0.480

Adiponectin (µg/mL) 15.6 ± 12.6 15.9 ± 12.4 0.931

Resistin (ng/mL) 40.9 ± 23.5 37.4 ± 13.2 0.552

Leptin (ng/ml) 18.8 ± 11.4 12.4 ± 10.2 0.07

Anthropometry and Body Composition
Weight (kg) 91.5 ± 23.7 83.2 ± 16.5 0.021
BMI (kg/m2) 32.8 ± 6.1 30.7 ± 4.5 0.228
WC (cm) 109.1 ± 15.8 102.3 ± 12.3 0.131
Fat Mass (%) 40.9 ± 6.1 38.5 ± 8.4 0.300
Visceral Fat (L) 3.9 ± 2.3 3.8 ± 1.5 0.935
Liver Outcomes
IHL (%) 12.4 ± 10.6 10.5 ± 8.6 0.568
LSM (kPa) 13.0 ± 15.0 7.4 ± 4.2 0.114
Insulin Profile
Glucose (mmol/L) 6.5 ± 1.8 6.1 ± 1.9 0.550
Insulin (mIU/L) 20.3 ± 13.4 16.4 ± 8.3 0.273
HOMA-IR 6.5 ± 6.0 4.6 ± 3.1 0.223

All data presented as mean ± SD. Abbreviations: MedDiet; Mediterranean Diet; LFD, Low-Fat Diet; hs-
CRP, high sensitivity C-reactive protein; TNF-α, tumor necrosis factor-alpha; IL-6, interleukin-6; WC, waist
circumference; WHR, waist-to-hip ratio; IHL, intrahepatic lipids; LSM, liver stiffness measure; HOMA-IR,
homeostatic model assessment of insulin resistance; BMI, body mass index.

26
Graphical Abstract

In a 12-week intervention of individuals with non-alcoholic fatty liver disease, participants who

were randomized to the Mediterranean diet group significantly improved levels of serum

adiponectin. A greater absolute reduction in leptin levels was observed in the low-fat diet

group, when compared to the Mediterranean diet group. No statistically significant changes

were observed for other inflammatory markers following the Mediterranean or low-fat diet

intervention.

Abbreviations: kJ, kilojoules; %E, percentage of total energy intake; CHO, carbohydrate; vs, versus;

hs-CRP, high-sensitivity C-reactive protein; IL-6, interleukin-6; TNF-α, tumor necrosis factor-alpha.

27