In the present cross-sectional study, the adult patients (age>18 years) were selected randomly from the East-Azerbaijan celiac disease registry database. In this registry, the patients were registerd if they have a positive serology tests confirmed by compatible duodenal histological findings.
The sample size for the present study was calculated based on the Bentler recommendation  with a minimum of ten observations per estimated parameters. So, we needed a sample size of 90 participants at least. In the present study, a total of 170 participants was recruited.
The Ethics Committee of Tabriz University of medical sciences (IR.TBZMED.REC.1399.904) approved the study and all participants gave written informed consent.
The information regarding demographic characteristics including age, education level, employment status, marital status, were obtained using a questionnaire. Having comorbidities was defined as the diagnosis of diabetes mellitus, cancers, chronic kidney diseases, chronic liver diseases, autoimmune diseases, inflammatory diseases, and psychological disorders.
Adherence to diet was assessed using the celiac disease adherence test (CDAT) questionnaire. This questionnaire assesses the level of adherence to gluten-free diet using seven questions on the five Likert scale type and the total score ranged between seven and 35. We considered a CDAT score<13 as good adherence, 13-17 as a moderate adherence, and >17 as a poor adherence to GFD . This questionnaire was previously translated to Persian, and its validity was confirmed in the previous study .
The severity of gastrointestinal symptoms was assessed by the Persian version of the gastrointestinal symptom rating scale (GSRS) questionnaire. This questionnaire includes fifteen questions on a seven-point Likert scale and higher scores indicate more severe symptoms. The questionnaire assesses five domains including diarrhea, constipation, abdominal pain, reflux, and indigestion.
The HRQOL was assessed using the SF-36 questionnaire. This questionnaire is a 36-item questionnaire that assesses physical (Physical Component Summary (PCS)) and psychological (Mental Component Summary (MCS)) health with a higher score indicating better health. This questionnaire was previously translated to Persian, and its validity was confirmed .
All analysis was performed using STATA-16. The normality of data distribution was assessed using the Kolmogorov-Smirnov test. The continuous variables were presented as mean±SD and the categorical variable were presented as frequency (%). An independent t-test and chi-square were used to compare the continuous variables and nominal and categorical variables between males and females respectively. Association between sociodemographic factors, disease-related factors, and quality of life was assessed using structure equation modeling (SEM). As can be seen in the conceptual model (Figure 1), adherence to diet and GSRS score were considered as mediators. The standardized regression weights was used to assess total, direct and indirect effects of variables on the HRQoL. The model fit was assessed to determine the “goodness of fit” between the hypothesized model and the data by use of several methods including the ratio of chi-square to the degree of freedom, root-mean-squared error of approximation (RMSEA), comparative fit index (CFI); and standardized root mean squared residual (SRMR). The acceptable values were Chi-square/DF <5, RMSEA<0.08, CFI>0.9, SRMR <0.08. A p-value less than 0.05 was considered significant.