Materials and Methods
The participants of the study were recruited between June 2016 and June 2017 in Navarra, Spain. In total, 228 subjects were evaluated to determine the presence of liver steatosis by abdominal ultrasonography. A total of 98 men and women with overweight or obesity (Body Mass Index (BMI) ≥27.5 kg/m2 to <40 kg/m2) between 40–80 years old fulfilled the selection criteria and started the trial (Figure 1). Exclusion criteria included the presence of known hepatic disease other than NAFLD, excessive alcohol consumption (>21 units of alcohol per week for men and >14 per women), weight loss ≥3 kg in the last 3 months, endocrine disorders (hyperthyroidism or uncontrolled hypothyroidism), pharmacological treatment with immunosuppressants, cytotoxic agents, corticosteroids or other drugs that could potentially cause liver steatosis or alteration in hepatic tests, severe psychiatric disorders, active autoimmune disease or requiring pharmacological treatment, the use of weight modifiers, and the lack of autonomy or inability to follow the diet, as well as the difficulties in following the scheduled visits. In case of habitual use of medication (eg antidiabetic or antihypertensive drugs), participants were encouraged to continue to take the pharmacological treatment prescribed by his/her physician, and the changes were registered at all the follow-up evaluations. The trial was approved by the Research Ethics Committee of the University of Navarra (ref. 54/2015) and registered at www.clinicaltrials.gov (FLiO: Fatty Liver in Obesity study; NCT03183193). Each subject gave written informed consent prior to enrolment in the study. All the participants were blinded to the group of allocation and the professionals in charge of hepatic assessment and other determinations were blinded to the dietary treatment, clinical information and laboratory data. All the procedures were performed in accordance with the Declaration of Helsinki and the study was conducted following the CONSORT 2010 guidelines.
Flowchart of the participants in the FLiO study. American Heart Association (AHA). Fatty Liver in OBESITY (FLiO). ‡One participant did not attend to 6-mo follow-up visit but did attend to 12-mo follow-up visit
Overview of the Study Protocol
This randomized controlled trial was designed to compare the effects of two weight loss dietary strategies with different nutritional features on liver status, anthropometric measurements, body composition and biochemical markers in overweight or obese subjects with ultrasonography proven liver steatosis. The participants were randomly assigned to the American Heart Association (AHA) group or the Fatty Liver in Obesity (FLiO) group (Figure 1). The intervention had a total duration of two years with a complete evaluation of the participants at baseline and after 6, 12 and 24 months. The assessment of the participants at each point of the study included anthropometric measurements, body composition by dual-energy X-Ray absorptiometry (DXA), biochemical determinations, and evaluation of the liver by ultrasonography, acoustic radiation force impulse (ARFI) elastography, transient elastography (TE), and Magnetic Resonance Imaging (MRI). Qualified radiologist and hepatologist performed the assessment of the liver status at the University of Navarra Clinic (CUN). In addition, fasting blood samples were properly collected, processed and stored at −80°C for further analyses.
Two different diets were prescribed and compared according to the group of allocation. Both diets applied an energy restriction of 30% of the total energy requirements of each participant with the objective to achieve a loss of at least 3%-5% of the initial body weight, in accordance with the recommendations of the American Association for the Study of Liver Diseases (AASLD). The energy prescription for each participant was estimated using the equation of the Institute of Medicine to calculate the resting metabolic rate, as elsewhere described. One diet was based on the guidelines of the AHA which propose 3–5 meals/day with a conventionally balanced distribution of macronutrients in relation to the total caloric value: 55% from carbohydrates, 15% from proteins and 30% from lipids with a healthy fatty acid profile. On the other hand, the FLiO diet was designed with a higher meal frequency (7 meals/d). The macronutrient distribution according to the total energy value was: 40%-45% carbohydrates (preferring those with low glycemic index), 25% proteins (predominantly from vegetable sources), and 30%-35% from lipids (favouring extra virgin olive oil and omega-3 polyunsaturated fatty acids to the detriment of saturated and trans fats). The FLiO diet proposed a high adherence to the MedDiet, involving an increased quantity of natural antioxidants based on previous studies of this research group. Moreover, a 7 days menu plan was provided to the participants in both groups.
Anthropometry, Body Composition and Biochemical Evaluation
Anthropometric and body composition evaluation was carried out at the Metabolic Unit of the University of Navarra. Body weight, height and waist circumference were determined as elsewhere described. Blood pressure (Intelli Sense. M6, OMRON Healthcare) and DXA body composition (Lunar iDXA, encore 14.5) were assessed following the instructions of the manufacturers. Blood samples for biochemical determinations were properly collected after overnight fasting of 8–10 hours. The samples were processed at the Laboratory of Biochemistry of the University of Navarra Clinic (CUN). Blood glucose, insulin, total cholesterol, high-density lipoprotein cholesterol (HDL-c), triglycerides, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and gamma-glutamyltransferase (GGT) concentrations were determined on an autoanalyzer with specific commercial kits and following the instructions of the company (Cobas 8000, Roche Diagnostics). Leptin and adiponectin concentrations were measured in a Triturus autoanalyzer (Grifols) using specific ELISA kits (Demeditec). BMI was calculated as the body weight divided by the squared height (kg/m2). The Homeostatic Model Assessment of Insulin Resistance (HOMA-IR), was computed as HOMA-IR = (insulin (μU/mL) × glucose (mmol/L))/22.5. The low-density lipoprotein cholesterol (LDL-c) levels were estimated using the Friedewald formula. The Triglycerides/Glucose index (TyG) was calculated as ln[triglycerides (mg/dL) × glucose (mg/dL)/2] and the Atherogenic index of plasma (AIP) as log [triglycerides/HDL-c)].
Assessment of Liver Status
The presence of hepatic steatosis was determined by means of ultrasonography (Siemens ACUSON S2000 and S3000) in accordance with previously described methodology. ARFI elastography and TE were performed to assess liver stiffness as a surrogate marker of hepatic fibrosis. ARFI was carried out along with the ultrasonography. Briefly, 10 valid determinations were performed on each participant and the median value of liver stiffness was computed. TE was carried out through FibroScan® (Echosens) as previously described. Finally, MRI (Siemens Aera 1.5T) was used to determine the fat content of the liver by Dixon technique as elsewhere described. The whole liver evaluation was carried out under fasting conditions at the University of Navarra Clinic.
The stages of fibrosis of the participants, as well as the number of participants with advanced fibrosis, were determined by means of TE according to cut-off points previously proposed.[23,24] Hepatic Steatosis Index (HSI) was calculated using the following formula: HSI = 8 × (ALT/AST ratio) ± BMI (±2, if female; ±2, if diabetes mellitus). The Fatty Liver Index (FLI) was computed using serum triglycerides, BMI, waist circumference and GGT concentrations using the formula elsewhere described.
Lifestyle Assessment: Diet and Physical Activity
Information regarding the diet and physical activity of the participants was collected at each timepoint of the study (baseline, 6, 12 and 24 months). The dietary intake was registered with a semi-quantitative food frequency questionnaire (FFQ), which consisted of 137-item and was previously validated in Spain for energy and nutrient intake. Each item in the FFQ included a typical portion size and the frequencies of consumption were registered in nine categories that ranged from "Never or almost never" to "≥6 times/d". The composition of the food items was derived from accepted Spanish food composition tables as previously described.[17,28] The adherence to the MedDiet was evaluated using a 17-point screening questionnaire. The final score ranged from 0 to 17 and a higher punctuation indicated a better adherence to the MedDiet. Physical activity was estimated using the validated Spanish version of the Minnesota Leisure-Time Physical Activity Questionnaire.[31,32]
Randomization and Statistical Analyses
The primary outcome of the study was the change in body weight, given that the current recommendations of the AASLD to ameliorate NAFLD features are focused on weight loss. The sample size was calculated to achieve a significant outcome with a 95% confidence interval (α = 0.05) and a statistical power of 80% (β = 0.8). A total of 36 participants per study group was estimated with this approach, but 50 subjects were included in each arm of the study considering the estimated dropout rate of 35%-40% according to the experience of the research group. The group assignment of the participants was performed using random numbers generated by a computer. The randomization and allocation were done by a physician not involved in the intervention, adherence and follow-up of the subjects throughout the study and were concealed from the investigators and the participants. Two individuals from the AHA group were excluded due to important alterations in the initial assessment of biochemical parameters, which required medical management. Therefore, the study started with 48 participants in the AHA group and 50 participants in the FLiO group. The differences in primary and secondary outcomes between the effects of the two dietary interventions during the entire trial (baseline, 6, 12 and 24 months), as well as the repeated measures overtime within each dietary group, were analysed using linear mixed model approach considering all available measurements. Categorical variables were compared using the Chi-squared test at different timepoints of the study. All p-values presented are two-tailed and a P < .05 was considered statistically significant. Statistical analyses were carried out using the software Stata version 12.0 (StataCorp). Graphs were generated using GraphPad Prism 6 (Graph-Pad Software).
Liver International. 2021;41(7):1532-1544. © 2021 Blackwell Publishing