INTRODUCTION

Gastroesophageal reflux disease (GERD) is “a condition that develops when the reflux of stomach contents causes troublesome symptoms and/or complications” (Vakil et al., 2006). True GERD with symptoms sufficiently pronounced to affect quality of life (QOL), seems to have an international variation in its prevalence. In fact, the prevalence in the Western world ranges between 10 and 20% (Dent et al., 2005; Zagari et al., 2008) which is higher than what is observed in Asia (5-7%) (Dent et al., 2005; Li et al., 2008; Ma et al., 2009; Wang et al. 2009).

Typical symptoms of GERD include heartburn and acid regurgitation, yet an extensive list of atypical and extra-esophageal manifestations has been identified during the past 20 years. These symptoms have a detrimental impact on the lifestyle of patients, compromising their physical, social and emotional well-being (Yamamichi et al., 2012). Although, GERD is not related to higher death rates, it presents with high morbidity rates (Chen et al., 2012) and its impact on the QOL of patients has been clearly demonstrated (Chen et al., 2012; Yamamichi et al., 2012).

Since, symptoms and QOL improvements are the main goals of treatment, assessment of quality of life through a questionnaire-based evaluation is an important asset in disease evaluation (Mouli and Ahuja, 2011). The REFLUX-QUAL short-form (RQS) is a quality of life measurement instrument specific to GERD which proved to be reliable, valid and sensitive to within and between-subject differences. It consists of 8 items covering the main domains of QOL (Amouretti et al., 2005). This score has not been validated for use in the Lebanese population.

MATERIALS AND METHODS

Procedure: A convenience sample was randomly selected across the five districts or "Mouhafazat" of Lebanon between March and May 2012. Being older than 18 years old and accepting to participate in the study were the only inclusion criteria, whilst no specific exclusion criteria were defined. After an oral informed consent, the questionnaire in Arabic was administered by a face-to face interview for participants in the streets.

Questionnaire: The questionnaire included sections about socio-demographic characteristics and lifestyle habits. Moreover, the psychological state was measured by the Beirut Distress Scale-22 (BDS-22) which is validated in Lebanon: A total score of 18 or less reflects a good mental state, otherwise the subject will be considered as stressed and at higher risk of psychological impairment (Barbour et al., 2012).

Participants were also asked to report their history of diseases and questions regarding the use of medications known to have a possible effect on GERD, followed by questions regarding the typical and atypical symptoms of GERD (Yuksel and Vaezi, 2012), their frequency and the beginning of their onset. Furthermore, the subjects having manifested any of these symptoms had to complete questions about their visit to the physician, if it occurred, the medication that was prescribed to them or self-medication and the lifestyle changes, if any.

Diagnosis of GERD: The majority of patients with GERD suffers from minor and intermittent symptoms and does not consult a specialist (Poynard, 1999). Therefore, a presumptive diagnosis of GERD can only be established in the presence of typical and pathognomonic symptoms of heartburn and regurgitation; however, a systematic review found the sensitivity of these two symptoms for the presence of erosive esophagitis to be 30-76% and their specificity from 62-96% (Moayyedi et al., 2006). The symptoms of heartburn and regurgitation remain the most reliable for making a presumptive diagnosis based on the patient’s history alone, (Moraes-Filho et al., 2010). Thus GERD prevalence was determined based on the presence of heartburn and/or regurgitation at least once per week and/or physician-made diagnosis of GERD. Subjects were then divided into GERD positive and GERD negative groups.

Impact of symptoms on patients’ QOL: The impact of symptoms on patients’ QOL was measured by the Reflux-Qual Short form (RQS) questionnaire (Amouretti et al., 2005) which was administered in Arabic local language, the translation process was done according to the WHO guidelines.

First, a forward translation into Arabic was done by a bilingual Lebanese researcher, instructions were given to him in the approach to translating, emphasizing conceptual and cultural equivalent of a word or phrase rather than literal translations, as well as the need to use natural and acceptable language for the broadest audience.

Second, the translated questionnaire was reviewed by a panel of experts which included the original translator, experts in health, as well as experts with experience in instrument development and translation to identify and resolve any inadequate expressions or concepts of the translation, as well as any discrepancies.

Third, an independent translator with no knowledge of the questionnaire back translated the questions into English, using the same instructions as in the initial translation. Discrepancies were then resolved by consensus between the researchers and the translator.

Fourth, the questionnaire was pilot tested on 20 subjects by an experienced interviewer; all questions were found to be clear by these individuals and no further changes were made to the initial questions.

The RQS score ranges from 0 (lowest QOL level) to 100 (highest QOL level) and is calculated by Amouretti et al. (2005) as follows:

RQS score = Mean value of the 8 items×25

Statistical analysis: Statistical analysis was made using the Statistical Package for the Social Sciences (SPSS), version 17.0. A p-value of 0.05 was considered significant. Weighting was performed according to the numbers published by the Lebanese Central Administration of Statistics in 2007 (CAS, 2008), taking into account gender and age to adjust the representativity of the sample.

Scale structure of RQS score: Internal structure of the RQS was assessed by a principal component factor analysis (PCA) using a Varimax rotation. The Kaiser-Meyer-Olkin value (KMO) index was used to determine the suitability of the data for dimensionality analysis (Cerny and Kaiser, 1977). Bartlett’s test of sphericity was also provided to ensure the model adequacy. The number of factors to retain was decided by eigen values higher than one. Cronbach’s reliability coefficient α was used for the internal consistency of the scale. Individual items were assessed by examination of their impact on the coefficient.

Validity of RQS score: In order to verify the apparent or construct validity of the RQS score, two hypotheses were generated. The first was based on the belief that the QOL should be lower in the patients suffering from symptoms. Evidence for construct validity was examined using Student’s t test to compare RQS scores of patients in the presence or absence of the different symptoms associated with GERD. In the second hypothesis, a dose-response relationship was assumed. Therefore, construct validity was examined using ANOVA to determine the relationship between the QOL and the frequency of the most common symptoms. The relationship between RQS scores and the number of symptoms manifested by the patient was also evaluated using Pearson’s correlation.

A univariate analysis was performed with the RQS score as a dependent variable and background factors as well as characteristics of GERD as explanatory variables. Student’s t-test was used for these analyses. As patients were categorized in GERD positive and GERD negative groups, a bivariate analysis was also carried out to compare characteristics’ differences between the two groups using the Chi square test.

Multivariate analysis: A forward multiple linear regression model was next applied for predictive background factors selected from the bivariate analyses (p≤0.2), with the RQS score as the dependent variable. The predictive variables which were entered were: Sex, age, Body Mass Index (BMI), marital status, education level, work status, salary, exercising, alcohol drinking, smoking, having any disease, GERD, cardiovascular diseases, diseases of the digestive system, cancer, joint pain, osteoporosis, asthma, diabetes, chronic inflammation, liver diseases, renal diseases, tuberculosis, hyperlipidemia, psychological state, number of symptoms related to GERD. The linearity of the relationship, the normality of residuals and the noncolinearity of variables were ensured so that the model could be accepted. This method served to find significant predictors of quality of life among patients with GERD.

Moreover, the potential risk factors of GERD were derived from a forward stepwise likelihood ratio logistic regression. Unadjusted (crude) and next multivariable-adjusted Odds Ratios (ORas) and their respective 95% Confidence Intervals (CIs) were calculated after ensuring the adequacy of the models using the Hosmer-Lemeshow test. Variables which were entered in the model were those who showed a significant effect in the bivariate analysis (p≤0.2). They consisted of socio demographic factors, lifestyle and dietary habits of participants, history or conditions of any diseases and drugs or supplements used by participants.

RESULTS

Nine hundred and seventy subjects were enrolled during the study period. The mean age was 33.55±16.01 years and 41.4% were males.

Factor analysis and reliability: First, the examination of the items correlation matrix revealed the presence of many correlation coefficients of 0.30 and higher. In addition, The Kaiser-Meyer-Olkin value (KMO) was 0.939, exceeding the recommended value of 0.60 (Table 1). The Bartlett’s test of sphericity also reached statistical significance (p<0.001) which also supported the appropriateness of dimensionality analyses of the correlation matrix.

Secondly, the rotated principal component factor analysis (PCA) revealed the presence of one component with eigenvalues exceeding one (λ = 6.52) which explained 81.5% of the total variance.

Inspection of the screeplot revealed a clear break after the first component as well. Catell’s screeplot test and the eigenvalues criteria suggested that only one component should be retained (Cattell, 1966). The components matrix showed that approximately all items loaded on the first component, with factor or components loadings ranging from 0.879-0.929.

Internal consistency, degree to which items that make up this scale represent a coherent set that measures the same underlying construct, was evaluated using Cronbach’s coefficient alpha. The results showed Cronbach’s coefficient alpha of approximately 0.967 (Table 1). The removal of any item from the scale was accompanied by a decrease in the value of the model’s Cronbach alfa.

Predictors of lower RQS: In the multivariate linear regression model estimating the predictors of low RQS scores, a higher number of symptoms (β = -0.364) had the strongest association with lower RQS scores, followed by the presence of any disease of the digestive system (β = -0.186) and by the presence of GERD (β = -0.176). Having a worse psychological state (β = -0.105), a higher BMI (β = -0.091), as well as suffering from any heart or artery disease (β = -0.088) or renal disease (β = -0. 087) were also associated with low RQS scores (Table 2).

Face validity of RQS: The evaluation of the symptoms’ effects on the RQS score was done in two parts. First, the effect of the symptoms presence on the score was evaluated. All symptoms lowered the RQS score significantly (p≤0.05). Moreover, average scores ranged from 95.45 in subjects with no symptoms to 73.77 in those manifesting at least one symptom (p<0.001, Fig. 1).

The use of digestive drugs (PPIs, H2RAs and others) shows a high correlation with RQS scores, all use associated with poorer quality of life (Fig. 2).

Scores on the RQS demonstrated a strong significant relationship with the frequency of the most common symptoms, showing reduced scores as the symptom became more frequent (p≤0.05) (Fig. 3).

A strong negative relationship appeared between the RQS score and the number of symptoms manifested (r = -0.513; p<0.001); average scores ranged from 95.45 in patients having no symptoms related to GERD to 51 in those having nine symptoms or more.

Comparison between GERD positive and negative participants: According to the method of diagnosis adopted, prevalence of GERD in the sample was 23.5% before adjustment and 24.8% for adjusted age and gender data. There was a significant relation between GERD prevalence and the age groups, indicating higher rates in older age groups even in weighted age-adjusted data (p<0.001). Reflux prevalence was not significantly different between the sexes for the weighted data (p = 0.534).

The comparison of characteristics’ differences between GERD positive and GERD negative groups are presented in Table 3: Men (p = 0.035), being married (p<0.001), working as a state employee (p<0.001) and non-smokers (p<0.001) had higher GERD rates. Older age was also associated with a higher GERD prevalence (p<0.001). Moreover, those having the greatest body mass index (BMI≥35) had the highest GERD rates (p = 0.004). However, there was no significant difference between GERD positive and negative patients regarding the region of residence (p>0.05).

GERD associates in bivariate analysis: Some factors were associated with GERD in bivariate analysis but were not retained in the multivariate regression model (Table 4). Smoking (OR = 1.61, 95% CI = 1.21-2.16) and drinking coffee or tea (OR = 2.18, 95% CI = 1.39-3.43) seemed to increase the risk of GERD, whereas exercising had a protective role (OR = 0.65, 95% CI = 0.50-0.85). Some foods could also be added: Vinegar (OR = 1.43, 95% CI = 1.08-1.89), honey (OR = 1.66, 95% CI = 1.27-2.17), chilli pepper (OR = 1.67, 95% CI = 1.07-2.61) and vegetables (OR = 1.63, 95% CI = 1.01-2.63). Many diseases or conditions were also correlated to GERD in the bivariate analysis.

First, having any disease at all had a significant impact on the occurrence of GERD (OR = 3.29, 95% CI = 2.43-4.44). Cancer (OR = 5.31, 95% CI = 1.65-17.07), renal diseases (OR = 3.42, 95% CI = 1.80-6.50), chronic inflammation (OR = 2.69, 95% CI = 0.99-7.30), osteoporosis (OR = 2.14, 95% CI = 1.13-4.04), cardiovascular diseases (OR = 1.95, 95% CI = 1.20-3.16), a worse psychological state (OR = 1.78, 95% CI = 1.35-2.36), headache (OR = 1.76, 95% CI = 1.10-2.81), joint pain (OR = 1.63, 95% CI = 1.24-2.15) and allergies (OR = 1.56, 95% CI = 1.05-2.31) seemed to be factors favoring GERD.

Moreover, many drugs and supplements appeared to have significant effects in the bivariate analysis (Table 4). Hydrocortisone had the most important impact (OR = 17.59, 95% CI = 5.26-58.88), followed by fosamax^® (alendronate sodium) (OR = 6.27, 95% CI = 1.69-23.34). Calcium supplements (OR = 1.67, 95% CI = 1.22-2.29), mebeverine (OR = 2.09, 95% CI = 1.03-4.24) and non steroidal anti-inflammatory agents, such as diclofenac (OR = 1.84, 95% CI = 1.39-2.44), diclofenac with vitamins B1, B6 and B12 (OR = 3.26, 95% CI = 1.89-5.64), aceclofenac (OR = 1.58, 95% CI = 1.06-2.37) and acetylsalicylic acid (OR = 1.79, 95% CI = 1.34-2.38), also appeared to be risk factors to GERD. On the other hand, paracetamol seemed to have a protective effect against GERD (OR = 0.46, 95% CI = 0.28-0.76).

GERD associates in multivariate analysis: In the multivariate analysis, older age and higher education had an adjusted OR of 3.10, 95% CI: 2.15-4.46 and 2.53, 95% CI = 1.74-3.68 for developing GERD, respectively (Table 4). Eating thyme (ORa = 1.73, 95% CI = 1.28-2.32), citrus (ORa = 1.34, 95% CI = 1.03-1.75), chocolate (ORa = 1.29, 95% CI = 0.99-1.69), garlic or onions (ORa = 1.51, 95% CI = 1.11-2.04) and cucumber (ORa = 1.60, 95% CI = 1.21-2.12) were associated with the occurrence of GERD, as well as drinking carbonated beverages (ORa = 1.62, 95% CI = 1.30-2.03). Drinking milk (ORa = 0.72, 95% CI = 0.57-0.91) appeared to have a protective effect on GERD occurrence. The number of meals per day had also a significant effect, showing that a higher number of meals decreased the risk of GERD (ORa of 0.60, 95% CI = 0.41-0.88).

Compared with those without, subjects with asthma have an adjusted OR of 8.35, 95% CI = 2.17-32.22 to report GERD symptom. Subjects with any disease related to the digestive system were more likely to have GERD, with an adjusted OR of 4.97, 95% CI = 2.67-9.25. Self-reported diabetes mellitus was also strongly associated with the development of GERD symptoms (ORa = 2.74, 95% CI = 1.38-5.43). Similarly, those who had a family member suffering from GERD tended to be more prone to get it themselves, with ORa of 3.73, 95% CI = 1.99-7.02 (Table 4).

DifenB12^® (betamethasone, diclofenac, vitamin B₁₂) was strongly associated with the presence of GERD (ORa = 69.186, 95% CI = 3.41-1405.31). Similarly, Profinal^® and Advil^® (both ibuprofen) were predictors of reflux (ORa = 2.76, 95% CI = 1.66-4.61 and ORa = 2.12, 95% CI = 1.31-3.44, respectively).

Also, iron supplements seemed to be associated with a rise in the risk of GERD (ORa: 1.62, 95% CI = 0.98-2.68) (Table 4).

DISCUSSION

The RQS scale was internally consistent and all item-total correlations were sufficiently high to justify inclusion in one, aggregated score for this population. Factor analysis demonstrated a single factor that explained most of the variation in the data. The scale was sensitive to the frequency of symptoms as well as their multiplicity. These data were consistent with previous studies of this instrument (Amouretti et al., 2005; Mouli and Ahuja, 2011). In fact, Amouretti et al. found the RQS to have a strong discriminative power based on the severity and frequency of reflux episodes (Amouretti et al., 2002). Other studies showed that the quality of life was all the more impaired as the symptoms were more frequent (Des Varannes et al., 2006).

Based on these criteria, we ended up with a one-dimensional 8-item RQS with items forming a coherent set that explained 81.5% of the variance in GERD patients’ quality of life. The results of the psychometric analyses support the validity of the RQS for this sample (Wiklund, 2004; Tabachnick and Fidell, 2007) and the scale reliability assessment supported the use of the RQS as a reliable tool for the assessment of the quality of life of GERD patients in Lebanon.

As predictors of low RQS scores, accounting for a low quality of life, a higher number of symptoms had the strongest association with lower RQS scores, followed by the presence of any disease of the digestive system and by the sole presence of GERD. In most cases, the symptoms of GERD impede patients’ lives in physical, social and mental well-being, lowering the QOL (Wiklund and Talley, 2003; Jo et al. 2013). The effects of GERD on patient health-related quality of life have been highlighted in several studies (Revicki et al., 1998; Wahlqvist, 2001; Kulig et al., 2003; El-Dika et al., 2005; Ronkainen et al. 2006). They all revealed a significant reduction in health-related quality of life compared with the general population. The ProGERD study gives strong support for this association (Kulig et al., 2003). Having a higher BMI worsens the RQS scores of GERD patients. In fact, another study supports this finding (Ponce et al. 2009). The results also showed that GERD patients having a worse psychological state had lower RQS scores than the other patients. A cohort study supports this finding by concluding that QOL in GERD patients may be more related to psychological factors rather than to symptom severity (Oh et al., 2009).

According to the method of diagnosis adopted in the study, prevalence of GERD in the sample population could be estimated to be 24.8% for adjusted age and gender data. This prevalence is close to that observed in other epidemiological studies defining GERD as at least weekly symptoms of heartburn and/or acid regurgitation, with no reference to symptom severity. In fact, the reported prevalence of pathological GERD was 10-20% in Western countries (Dent et al., 2005; Zagari et al., 2008).

The risk factors belonged to 4 distinct categories: Socio-demographic, lifestyle and alimentation, history of diseases and use of medication. Male gender, being married and working as a state employee first appeared to have higher GERD rates. However, after adjusting for age and gender, the previous factors showed no significant differences. In fact, several studies concluded that there was no association between gender and GERD (Kotzan et al., 2001). A systematic review showed that the prevalence of GERD may increase with age (Becher and Dent, 2011). In this study, older age remained associated with a higher GERD prevalence even after the adjustment.

Moreover, many lifestyle-related factors showed strong correlation with GERD symptoms. The extremely obese (BMI≥35) had the highest GERD rates. In fact, it is now widespread that obesity may increase the risk of GERD (Hampel et al., 2005; Corley et al., 2007), through mechanical alterations at the esophagogastric junction (Sise and Friedenberg, 2008). A higher education appeared to be related to GERD as well. Higher education was also significantly associated with GERD in previous studies (Jiang et al., 2010), as well as higher income (Sonnenberg, 2011), whereas opposite results were found in other studies (Diaz-Rubio et al., 2004; Jansson et al., 2007).

Dietary habits had significant correlation with GERD occurrence. Consuming thyme, citrus, chocolate, garlic or onions and cucumber were associated with the occurrence of GERD, as well as drinking carbonated beverages. However, drinking milk appeared to have a protective effect. Some works show that chocolate (Richter and Castell, 1981; Murphy and Castell, 1988), acid juices (McArthur et al., 1982), carbonated beverages (Feldman and Barnett, 1995; Csendes and Burdiles, 2007) and onions (Terry et al., 2000) are refluxogenic agents (Csendes and Burdiles, 2007). Chocolate, onions, citrus fruit and carbonated beverages are associated with temporary symptoms of reflux, most likely through a relaxation of the Lower Esophageal Sphincter (LES) (Murphy and Castell, 1988; Shukla et al., 2012). Milk’s inverse association may be explained by the transient buffering effect, although this effect is followed by a significant gastric acid secretion (Martin-de-Argila and Martinez-Jimenez, 2013).

Alcohol drinking, tea drinking and smoking were not associated with any change in the risk of GERD. In fact, the relationships between alcohol, caffeine or smoking and GERD are controversial since some studies have observed a positive relationship (Matsuki et al., 2013), whereas some other reports have pointed to a lack of association between them (Dent et al., 2005; Mouli and Ahuja, 2011; Chen et al., 2012; Pandeya et al., 2012). Finally, the number of meals per day had also a significant effect, showing that a higher number of meals decreased the risk of GERD. In fact, patients suffering from GERD are advised to have small frequent meals (Havemann et al. 2007).

Subjects with asthma reported more GERD symptoms. This has been shown in other studies as well (Thompson and Heaton, 1982). Subjects with diseases related to the digestive system were more likely to have GERD. In fact, a history of Irritable Bowel Syndrome (IBS), abdominal pain or peptic ulcer disease was associated with an increased risk of GERD diagnosis (Thompson and Heaton, 1982; Lee et al., 2009). Moreover, a systematic review showed that several co-morbidities identified with GERD were of gastrointestinal origin (Dent et al., 2005). Self-reported diabetes mellitus was also strongly associated with the development of the symptoms. Observations demonstrated the association between GERD and metabolic risk factors including diabetes mellitus (Chiba et al., 2012). Similarly, those who had a family member suffering from GERD tended to be more prone to get it themselves. GERD in first-degree relatives has already been correlated with the risk of GERD (Kotzan et al., 2001; Nasseri-Moghaddam et al., 2008; Toghanian et al. 2011); there have even been speculations about a genetic component regarding the disease, implied by some authors (Dent et al., 2005).

For medications, DifenB12^® (betamethasone, diclofenac, vitamin B12) was strongly associated with the presence of GERD. Similarly, Profinal^® and Advil^® (both ibuprofen) were predictors of reflux with close ORa values. Non Steroidal Anti-Inflammatory Drugs and corticosteroids are known for their gastrointestinal adverse effects; therefore, their intake has been classified as a risk factor for GERD and was found to have an effect in many studies (Yamamichi et al., 2012; Nasseri-Moghaddam et al., 2008). Also, iron supplements seemed to be associated with a rise in the risk of GERD.

Moreover, all symptoms lowered the RQS score significantly while average scores ranged from 95.45 in subjects with no symptoms to 73.77 in those manifesting at least one symptom. Many patients on digestive drug medications were still suffering from GERD symptoms and having low RQS scores. This could be due to the fact that the study was cross-sectional and could not measure the temporal effects but this was also noticed in another study where the QOL impairment remained positively correlated with the GERD symptoms despite optimized therapy (Stalhammar et al., 2012; Des Varannes et al., 2013). In fact, Yamamichi et al., 2012 that digestive medicine could not relieve the GERD symptoms completely.

Scores on the RQS demonstrated a strong significant relationship with the frequency of the most common symptoms, showing reduced scores as the symptom became more frequent. Several studies showed a positive correlation between the quality of life and the severity and frequency of GERD symptoms (Ronkainen et al. 2006; Wahlqvist et al., 2008; Stalhammar et al., 2012; Des Varannes et al., 2013). Average scores ranged from 95.45 in patients having no symptoms related to GERD to 51 in those having nine symptoms or more with a clear descending mean as the number of symptoms rose.

Several limitations of this study are worth citing: First, given the cross-sectional nature of the study, one cannot verify the temporality of the associations between dietary habits and GERD, since there is a risk of reversed causality: GERD positive patients would avoid foods that trigger the symptoms, whereas they would consume those that soothe them. The convenience of the sample could introduce some selection bias; however, the low refusal rate and the weighting performed for sample adjustment could overcome this problem.

Another limitation of this study is the fact that data were self-reported; therefore, the possibility of information bias cannot be fully refuted. Also, some recall periods were relatively long (one month for the RQS questions and up to six months for the symptom severity/frequency); these data may therefore be subject to recall bias. Moreover, symptoms were assessed only by using patient questionnaires and not by a physician. But, in the absence of a gold standard for diagnosing GERD, patient questionnaires remain the common outcome in clinical or epidemiological studies, especially since many people with GERD symptoms do not consult a physician, particularly if they have mild disease rather than severe and/or at least weekly symptoms of GERD (Rey et al., 2004; Bretagne et al., 2006).

Although, both GERD diagnosed by physicians and the presence of typical symptoms for diagnosis were taken into account, atypical symptoms were not considered part of GERD, although patients with atypical manifestations may not have concomitant complaints of heartburn or regurgitation. Due to that, many patients may not be appropriately diagnosed (Vaezi, 2005). However, diagnostic tests having low specificity and establishing a cause-and-effect association between GERD and atypical symptoms being difficult (Vaezi, 2005), the diagnostic method was valid. Furthermore, since the study was cross-sectional, responsiveness of the measure to change was not assessed. That is, how well the RQS will capture differences in health care processes is not known. The issue of reverse causality remains unresolved as well from such cross-sectional designs. Despite these limitations, findings from the current study are in accordance with population-based studies from other countries.

Nevertheless, some strong points of this study deserve to be cited: The relationship between symptom load and the RQS instrument provides indirect support for the construct validity of the scale in the studied population (face validity), as the severity and frequency of GERD symptoms appear to have a predictive value for the extent of QOL deterioration. The questionnaire was administered orally in order to minimize effects of low literacy. It included not only the major and most common GERD symptoms, heartburn and acid regurgitation but also atypical symptoms and extra-esophageal manifestations of GERD. The data obtained on many variables, including body mass index, diet, smoking status and alcohol intake and other possible confounding factors are all worth noting.

CONCLUSION

The RQS items coherence in assessing the underlying construct of GERD’s impact on quality of life was demonstrated by high indices of internal consistency and reliability. Furthermore, the frequency and the multitude of GERD symptoms appeared to lower quality of life which confirms face validity of the scale. Therefore, the RQS questionnaire is a simple, 8-items questionnaire that can be used by doctors in Lebanon to assess the effect of GERD on the quality of life of their patients and to appropriately treat patients’ symptoms. The study also revealed an association between GERD and some socio-demographic factors, lifestyle and dietary habits, concomitant diseases and the use of certain drugs.

Lastly, additional studies that examine divergent and convergent characteristics of the RQS with other relevant psychometrically valid and reliable health measures, as well as studies that compare the RQS with other disease-specific scales measuring the effects of GERD on HRQOL, are needed in order to provide evidence of the strength of the RQS and to calculate the specificity and sensitivity of the scale in Lebanon.

ACKNOWLEDGMENT

This study was funded by the Lebanese University, Doctoral School of Sciences and Technology.