Study participant
This double blind parallel randomized controlled trial was conducted according to the CONSORT guidelines. Participants included 76 KOA women recruited from the Khatam Al-Anbia Super Specialty clinic at the Department of Rheumatology, Yazd, Iran. To be eligible for inclusion, women had to be over 45 years old, have a body mass index (BMI) in the range of 25-35 kg/m2, and have diagnosed KOA according to the clinical classification of KOA (18). The exclusion criteria were: having former or planned knee-joint replacement, being under pharmacologic treatment for obesity, having any history or active presence of other rheumatic diseases, using any nonsteroidal anti-inflammatory drug (NSAID), consuming ≤ 20 percent of the supplements, consuming multivitamin, minerals, or other nutritional supplements, and having severe liver, kidney, or heart diseases.
Randomization and intervention
This study was a double blind randomized controlled trial. Patients, who met the study criteria, were assigned to L-carnitine group (LCG) or placebo group (PLG) through randomization lists made by a computerized random-number generator and simple randomization process in a ratio of 1:1. The L-carnitine group received 1 g/d L-carnitine daily for 12 weeks. The placebo group received a placebo according to the same regimen and the same duration. The placebo pills contained inactive ingredients with no therapeutic activity and had an identical appearance. These tablets were produced by Karen Pharmaceutical & Nutrilife Co., Yazd, Iran. As a double-blind study, the bottles were labeled A and B respectively for the placebo and drug by the factory but neither the patients nor the research team were aware of the codes. Every month, patients received a bottle of tablet containing 30 tablets. Compliance with the medication was monitored by the research personnel using pill counts and patients’ self-reporting. Participants who did not consume more than 30 percent of their supplements were eliminated from the analysis. All participants followed a low- calorie diet. A registered dietitian estimated the energy expenditure for each patient through Harris-Benedict formula using the individual activity factor [19]. The recommended composition of the dietary regimen was 50% to 60% carbohydrates, 15% to 20% proteins, and less than 30% total fat. Initially, a dietitian completed the 3-day food recall for all participants and visited patients every month to check their compliance with the diet according to the patients’ feedback and 24-hour food recall. At the baseline, physical activity during the past week was assessed using the long version of International Physical Activity Questionnaire (IPAQ). Patients were also asked not to change their level of activity during the study.
Outcome Measurements
The following variables were assessed at the baseline and 12 weeks after the start of treatment: primary outcome included WOMAC, CRP and MDA. Secondary outcome was LDL-C, total cholesterol (TC), high density lipoprotein-cholesterol (HDL-C), triglycerides (TG), BMI, fat mass, free fat mass, as well as waist and hip circumference.
In order to conduct the laboratory tests, 5 mL of venous blood samples was obtained after the patients had fasted for 8 hours overnight. Immediately after the centrifugation (3000g, 10 min), serum samples were produced from the collected blood samples. They were then frozen at -20 ºC, stored at -70 ºC, and measured at the same time. The total TC, HDL-C, and TG were later measured using Pars Ammon kit (Iran). LDL-C was then calculated using Friedewald’s equation (20). Serum CRP and serum MDA concentrations were measured through enzyme-linked immunosorbent assay kits and thiobarbituric acid reactive substances Zellbio kit (Germany), respectively.
The patients’ weights were also recorded on a portable digital scale (Omeron BF511, Japan) to the nearest 0.1 kg. Participants were in light clothes and stood on the scale without help. Fat mass and free fat mass were measured with this scale. Further, height was measured in standing position using an audiometer fixed on a straight wall to the nearest 0.1 cm. Waist circumference (WC) was recorded to the nearest 1 centimeter using non-stretch plastic tape placed midway between iliac crest and lowest rib while participants were in standing position. Hip circumference was also measured over the largest part of buttocks with the accuracy of 1 cm. BMI was also calculated as weight (kg) divided by height squared (m2). To assess the clinical symptoms Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) were used. Patients filled out the Persian version of WOMAC index [21] that consists of 24 questions (pain, stiffness, and physical functional). Items should be answered on a Likert scale: none (0), mild (1), moderate (2), severe (3), or extreme (4). For physical activity, the Persian version of IPAQ was applied [22]. The continuous score shows the weekly energy expenditure expressed in MET-min/week (metabolic equivalent-minutes). Individuals were classified into three categories of ‘inactive’, ‘moderately active’, and ‘highly active’ using the categorical classification.
Sample size
Power calculations were conducted based on pain scores of 72 women with KOA who participated in a previous trial [23]. On the basis of an assumed 10% dropout rate, we estimated that a total sample size of 76 patients (38 patients per group) would provide 80% of the power to detect a 2.6 pain score difference between the PLG and LCG.
Statistical analysis
Statistical analyses were carried out using SPSS (version16). The normal distribution of variables was tested by Kolmogorov-Smirnov test. Differences in patients’ anthropometrics, WOMAC scores, and hematological measurement data between the PLG and LCG were analyzed by Student’s t-test or the Mann–Whitney rank sum test for parametric and non-parametric continuous variables, respectively. The paired t-test or Wilcoxon signed rank test was used to analyze the data within each group before (baseline) and after the intervention (week 12). Analysis of covariance (ANCOVA) was used to identify the differences between the two groups after adjusting for change weight. Results were considered statistically significant at P < 0.05. Normal data were indicated by means ± standard deviations (SD) and the non-normal scores were presented in median and inter quartile range (IQR). The dietary information was analyzed with N4 software (Nutritionist: version 4.0; Tinuviel Software, Warrington, United Kingdom).
Efficacy and tolerability assessment
For the safety, all participants were interviewed every month for any signs of L-carnitine toxicity or other adverse problems considering the diet, including serious illnesses or hospitalizations.