Risk Factors Associated with Oral Manifestations and Oral

BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on October 1, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from

Risk factors, associated oral manifestations, and oral health impact of gastro-oesophageal reflux disease: A multicentre, cross-sectional study in Pakistan

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-021458

Article Type: Research

Date Submitted by the Author: 04-Mar-2018

Complete List of Authors: Warsi, Ibrahim; Dow University of Health Sciences, General Dentistry Ahmed, Javeria; Dow University of Health Sciences, General Dentistry Younus, Anjum; Dow University of Health Sciences, Department of Community Dentistry Rasheed, Abdur; Dow University of Health Sciences, Department of Research and Biostatistics Hashmi, Rimsha; Rehmat Memorial Dental Hospital, Abbottabad, Pakistan, Department of Operative Dentistry Akhtar, Tayyab; Rawalpindi Medical College, Gastroenterology and Liver Centre, Holy Family Hospital

Gastroenterology < INTERNAL MEDICINE, PUBLIC HEALTH, PREVENTIVE Keywords: MEDICINE, ORAL MEDICINE, Endoscopy < GASTROENTEROLOGY,

EPIDEMIOLOGY http://bmjopen.bmj.com/

on October 1, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 17 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 Risk factors, associated oral manifestations, and oral health impact of gastro- 2 3 oesophageal reflux disease: A multicentre, cross-sectional study in Pakistan 4 5 6 1 1 2 3 4 5 7 Ibrahim Warsi , Javeria Ahmed , Anjum Younus , Abdur Rasheed , Rimsha Hashmi , Tayyab Akhtar 8

9 1 10 Dr. IshratulEbad Khan Institute of Oral Health Sciences, Dow University of Health Sciences, Karachi, Pakistan 2 11 Department of Community Dentistry, Dr. IshratulEbad Khan Institute of Oral Health Sciences, Dow University of Health 12 Sciences, Karachi, Pakistan. 13 3Department of Research, Dow University of Health Sciences, Karachi, Pakistan 14 4Department of Operative Dentistry, Rehmat Memorial Dental Hospital, Abbottabad, Pakistan 15 5 16 Gastroenterology and Liver Centre,For Holy Familypeer Hospital, review RawalpindiIslamabad, only Pakistan 17 18 Ibrahim Warsi 19 House officer, Dr. IshratulEbad Khan Institute of Oral Health Sciences, Dow University of Health Sciences, Karachi, Pakistan 20 21 22 Javeria Ahmed 23 House officer, Dr. IshratulEbad Khan Institute of Oral Health Sciences, Dow University of Health Sciences, Karachi, Pakistan 24 25 Anjum Younus 26 27 Postgraduate trainee MDS, Department of Community Dentistry, Dr. IshratulEbad Khan Institute of Oral Health Sciences, Dow 28 University of Health Sciences, Karachi, Pakistan 29 30 31 Abdur Rasheed 32 Deputy director research & biostatistician, Department of Research, Dow University of Health Sciences, Karachi, Pakistan

33 http://bmjopen.bmj.com/ 34 Rimsha Hashmi 35 Postgraduate resident FCPS Operative dentistry, Rehmat Memorial Dental Hospital, Abbottabad, Pakistan 36 37 38 Tayyab Akhtar 39 Senior Registrar, Gastroenterology and Liver Center, Holy Family Hospital, RawalpindiIslamabad, Pakistan 40

41 on October 1, 2021 by guest. Protected copyright. 42 43 Correspondence: Dr. Ibrahim Warsi 44 178, Street 18, Sector F10/2, Islamabad, Pakistan 44000 45 Tel: 00923328262980 46 47 Email: [email protected] 48 49 WORD COUNT: 50 51 MAIN TEXT: 3174 52 ABSTRACT: 298 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open Page 2 of 17 ABSTRACT BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 Objective : Gastrooesophageal reflux disease (GORD) is a relatively common disorder and manifests as extra 3 4 oesophageal symptoms, such as dental erosions (DE), cough, laryngitis, asthma, and oral soft/hard tissue pathologies. 5 6 This study aimed (i) to identify oral soft and hard tissue changes in patients with GORD and (ii) to evaluate these oral 7 8 changes as indices for assessing GORD and its severity. 9 Setting: This crosssectional study was conducted at four major tertiary care government hospitals in two of the major 10 11 cities of Pakistan. 12 13 Participants: In total, 187 of 700 patients who underwent oesophago–gastro–duodenoscopy (OGD) and having 14 GORD were included in the study. The GORD patients were divided according to the presence of dental erosions (DE) 15 16 into group A (with DE, chronic/severeFor peer GORD) andreview group B (without only DE, mild GORD). Patients who were 17 18 unconscious and extremely limited mouth opening were excluded. 19 20 Primary & secondary outcome measures: The impact of oral changes on the oral healthrelated quality of life was 21 assessed using the Pakistani (Urdu) version of the validated Oral Health Impact Profile14 (OHIP14) instrument. 22 23 Abnormal conditions and lesions of the oral mucosa were recorded. 24 25 Results: Oral submucosal fibrosis (66.3%), ulceration (59.4%), and xerostomia (47.6%) were significantly more 26 27 common in group A (p < 0.05). The prevalence of GORD was 26.71%, within which the prevalence of DE was 35.3%. 28 Dietary pattern, nausea/vomiting, loss of appetite, and oesophagitis showed a statistical significant association with 29 30 DE. All subscales of OHIP14 were positively correlated in patients with GORD and DE (p < 0.05). 31 32 Conclusion: Patients with GORD and DE presented with more severe oral symptoms than did those with GORD and

33 no DE. We recommend a controlled diet and timely dental checkups to assess the severity of both systemic and oral http://bmjopen.bmj.com/ 34 35 disease. 36 37 Keywords: GORD, dental erosions, oral manifestations, extraoesophageal symptoms, oral health, OHIP14 38

39 40

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open Page 4 of 17 INTRODUCTION BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 Gastrooesophageal reflux disease (GORD) is a common disorder, affecting approximately 10%–20% of the general 4 population.[1] The Montreal consensus,[2] classifies GORD as an entity manifesting as oesophageal and extra 5 6 oesophageal syndromes. The oesophageal symptomatology includes regurgitation or burning retrosternal chest pain, 7 8 reflux oesophagitis, strictures, Barrett’s oesophagus, and adenocarcinoma, while extraoesophageal symptomatology 9 includes reflux, cough, laryngitis, asthma, and dental erosions (DE). DE are a multifactorial phenomenon whereby 10 11 the protective buffering capacity of the oral cavity is compromised by reduced secretion of saliva or high volumes of 12 13 harmful gastric reflux.[3] DE appears to be the most prevalent injury caused by GORD.[4] The prevalence of DE in 14 15 the general population is not accurately known; however, depending on the underlying cause, a prevalence of 2%–77% 16 was reported.[5] Tooth involvementFor seems peer to be universal, review but the most only commonly observed damage occurs on the 17 18 palatal surfaces of the posterior teeth, with a reported prevalence of up to 42%.[6] Studies have highlighted the 19 20 injurious effects of gastroduodenal contents in oral softtissue pathology as well as on DE.[3] These include effects on 21 the oesophageal epithelium, soft palate, and oral mucosa, and manifest as burning mouth syndrome, aphthoidlesions, 22 23 hoarseness of voice, erythema of the soft palate and uvula, glossitis, epithelial atrophy, and xerostomia.[7, 8, 9] 24 25 Epithelial atrophy and xerostomia further aggravate GORDinduced injury to the epithelium in the oral cavity and 26 27 oesophagus.[10] 28 29 The severity of DE is directly proportional to the amount of time gastric acid, in contact with the enamel, so the 30 31 frequency and duration of the reflux problem can be assessed by the amount of enamel loss and vice versa.[11, 12] We 32 hypothesized a strong association between DE and chronic severe GORD and that an absence of DE would indicate a 33 http://bmjopen.bmj.com/ 34 lesssevere form of GORD of shorter duration. The aims of this study were as follows: to assess the frequency of 35 36 GORD with and without DE, and oral symptoms in patients undergoing oesophago–gastro–duodenoscopy (OGD); to 37 38 determine the association between severity of GORD and the likelihood of orodental manifestations, and to determine 39 40 the effect of GORD on the oralhealth related quality of life (HRQoL).

41 on October 1, 2021 by guest. Protected copyright. 42 43 MATERIALS AND METHODS 44 45 46 47 The study was conducted at four major tertiary care government hospitals in two of the major metropolitan cities of 48 Pakistan, namely the twin city of RawalpindiIslamabad (Pakistan Institute of Medical Sciences, PIMS and Holy 49 50 Family Hospital) and Karachi (Jinnah Hospital, and Civil Hospital). In total, 187 of 700 patients who underwent OGD 51 52 along with a comprehensive gastrointestinal (GI) examination over a fourmonth period were diagnosed to have 53 54 GORD, confirmed medically by treatment with a proton pump inhibitor, and were included in the study. A mandatory 55 dental examination was included to evaluate the orodental effects of GORD. Patients who were unconscious and 56 57 those with extremely limited mouth opening (precluding an oral examination) were excluded. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Assessment of Demographic and Clinical Characteristics Page 5 of 17 BMJ Open A questionnaire was used to collect data on sociodemographics, past and current medical history (comorbidities),

drug history (including use of nonsteroidal antiinflammatory drugs [NSAIDs]), and risk factors for GORD (dietary BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 pattern, addiction profile, consumption of tea, eating habits, weight changes, and other GI disorders). 3 4 5 6 Assessment of Health-related quality-of-life (HRQoL) 7 The Oral Health Impact Profile (OHIP14) is an instrument designed to assess oralHRQoL and includes seven 8 9 subscales. The Pakistani (Urdu) validated version of this instrument,[13] was used to assess factors related to GORD 10 11 and their effect on the oral cavity in terms of speech production, mastication, psychosocial wellbeing, and life in 12 13 general, and to estimate the impact of GORD on oral and systemic health. 14 15 16 Gastrointestinal (GI) examinationFor peer review only 17 18 All patients underwent a comprehensive GI examination with baseline laboratory investigations and were on 19 appropriate medical therapy.[12] OGD was performed to confirm the diagnosis of GORD. 20 21 22 23 Dental examination 24 25 The World Health Organization (WHO) criteria for oral and dental examination,[19] was adapted to examine patients 26 for decayed teeth by using the decayed, missed, filled, and trauma (DMFT) index value, mouth opening status, and 27 28 presence of oral submucosal fibrosis (OSF) by using the OSFstaging index. Abnormal conditions and lesions of the 29 30 oral mucosa were recorded. Oral examination was conducted by single calibrated dentist (IW). 31 32

33 Statistical analysis http://bmjopen.bmj.com/ 34 35 The data were analysed using SPSS Version 21 (IBM Corp., Armonk, NY, USA). Categorical variables are reported as 36 37 frequencies and percentages. Associations between GI symptoms and oral manifestations of GORD were determined 38 using the chisquared test. The Spearman rank correlation was used to test for associations of the mean values, on the 39 40 OHIP14 subscales with GORD and DE. Bivariate logistic regression was used to obtain crude odds ratios (ORs)

41 on October 1, 2021 by guest. Protected copyright. 42 assessing the likelihood of GORD and DE according to patient characteristics. Variables with p ≤ 0.2 were entered 43 44 into multivariable logistic regression analysis. The results are reported with the adjusted OR and 95% confidence 45 interval (CI). 46 47 48 49 Patient involvement 50 51 This study was noninterventional, and the participants played no active role in the research other than answering a 52 questionnaire. Any invasive procedures (such as, OGD and lab investigations) performed were related to routine GI 53 54 investigations ordered by gastroenterologists in concerned hospital departments and therefore were outside our study 55 56 protocol. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 17 Ethical approval

The study protocol was approved by the ethics committees of each participating hospital before participant enrolment. BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 Written informed consent was obtained from all study participants after explaining the study protocol. 3 4 5 6 7 RESULTS 8 9 10 11 Most study participants were males (58.3%) and in the adult age group (79.7%), as shown in Table 1. The study 12 sample was ethnically and racially diverse; most participants were Punjabis (54.5%), followed by Urduspeaking 13 14 Mohajirs (16.6%), Pathans (15%), and Sindhis (8%). Most were urban residents (63.6%) of low (62%) to intermediate 15 16 (35.3%) socioeconomic status.For The addiction peer profile review included cigarette only smoking (26.7%) and alcohol (2.7%), oral 17 tobacco (30.5%), and tea (72.7%) consumption. A sizeable proportion experienced generalized body pain (44%), for 18 19 which they reported seeking selfmedication, most often NSAIDs (46%). The most common comorbidities were 20 21 diabetes (34.8%) and hypertension (23%). 22 23 24 Slightly more than half (56.7%) of the study population had GORD alone, and the remaining had additional upper GI 25 26 conditions, including gastritis (19.8%), portal gastropathy (11.2%), peptic ulcer disease (10.2%), GORDrelated 27 28 oesophagitis (4.8%), and hiatal hernia (4.3%). 29 30 31 Table 1. Demographic variables and risk factors in patients with GORD 32

33 http://bmjopen.bmj.com/ 34 S.NO. DEMOGRAPHIC VARIABLES N (%) 35 1 Gender Male 109 (58.3%) 36 Female 78 (41.7%) 37 2 Age Adolescents (<19 years) 07 (3.7%) 38 Adults (2060 years) 149 (79.7%) 39 Old (>60 years) 31 (16.6%) 40 3 Race/Ethnicity Sindhi 15 (8%) Balochi 05 (2.7%)

41 on October 1, 2021 by guest. Protected copyright. 42 Punjabi 102 (54.5%) 43 Pathan 28 (15%)

44 Kashmiri 06 (3.2%) Urdu Speaking Mohajirs 31 (16.6%) 45 4 Location of residential Urban 119 (63.6%) 46 Background (Domicile) Rural 68 (36.4%) 47 5 Socioeconomic status (SES) Low 116 (62%) 48 Moderate 66 (35.3%) 49 High 05 (2.7%) 50 6 Smoking Yes 50 (26.7%) 51 No 137 (73.3%) 52 7 Alcohol Yes 05 (2.7%) 53 No 182 (97.3%) 54 8 Tea consumption Yes 136 (72.7%) 55 No 51 (27.3%) 56 9 Oral Tobacco consumption Yes 57 (30.5%) 57 No 130 (69.5%) 58 10 NSAIDS consumption Yes 86 (46%) 59 No 101 (54%) 60 11 History of Body For peer reviewYes only - http://bmjopen.bmj.com/site/about/guidelines.xhtml83 (44%) Ache/General Body Pain No 104 (56%) Page 7 of 17 BMJ Open 12 Weight Loss (from <5Kg to No 45 (24.1%) >20Kg) 15 Kg 72 (38.5%) BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 510 Kg 13 (7%) 2 1015 Kg 36 (19.3%) 1520 Kg 15 (8%) 3 >20Kg 06 (3.2%) 4 13 Hypertension Yes 65 (34.8%) 5 No 122 (65.2%) 6 14 Diabetes Mellitus Yes 43 (23%) 7 No 144 (77%) 8 16 Dental Erosions Group A: GORD with DENTAL EROSIONS (DE) 66 (35.3%) 9 Group B: GORD without DE 121 (64.7%) 10 17. GORD (alone) and (in GORD Alone 106 (56.7%) 11 Combination with other GI GORD with additional UpperGI conditions 81 (43.3%) 12 illness) i. Gastritis 16 (19.8%) 13 ii. Portal Gastropathy 09 (11.2%) 14 iii. Peptic Ulcer Disease (PUD) 08 (10.2%) 15 iv. GERD Esophagitis 04 (4.8%) 16 For peerv. Hiatal review Hernia only04 (4.3%) 17 vi. Others 40 (49.7%) 18 GORD, gastrooesophageal reflux disease 19 20 DE is a known comorbidity of chronic GORD; therefore, we divided the patients according to the presence of DE into 21 22 two groups, namely group A (with DE, chronic/severe GORD; n = 66, 35.3%) and group B (without DE, mild GORD; 23 24 n = 121, 64.7%). The oral clinical and subjective findings were compared between the two groups. Table 2 shows the 25 26 GI signs and symptoms that were found in both groups A and B, the most common of which was 27 heartburn/regurgitation (95%), followed by nausea/vomiting (82%) and dysphagia (42%). A statistically significant 28 29 difference in the frequency of nausea/vomiting (p = 0.006) and early satiety/loss of appetite (p = 0.042) was found 30 31 between patients who had GORD with and without DE. 32

33 http://bmjopen.bmj.com/ 34 Table 2. Gastrointestinal signs and symptoms in patients with GORD 35 36 37 GI SYMPTOMS Group A Group B ALL P-value 38 (n=66) (n=121) (n=187) 39 Heartburn/Regurgitation Yes 63 115 178 (95.2%) 0.900 40 No 03 06 09 (4.8%) Nausea/Vomiting Yes 48 107 155 (82.9%) 0.006* 41 on October 1, 2021 by guest. Protected copyright. 42 No 18 14 32 (17.1%) 43 Abdominal Pain/discomfort Yes 49 92 141 (75.4%) 0.786 44 No 17 29 46 (24.5%) 45 Abdominal distension Yes 40 62 102 (54.5%) 0.219 46 No 26 59 85 (45.4%) 47 Early Satiety/ Loss of Appetite Yes 44 62 106 (56.7%) 0.042* No 22 59 81 (43.3%) 48 Dysphagia Yes 32 48 80 (42.8%) 0.244 49 No 34 73 107 (57.2%) 50 *p<0.05 (chisquared test) was considered to be statistically significant. GORD, gastrooesophageal reflux disease. Group A, chronic/severe 51 GORD with DE. Group B, mild GORD without DE. 52 53 54 The oral manifestations associated with GORD in groups A and B are shown in Table 3. In group A, the most 55 significant association was with oral submucosal fibrosis, followed in decreasing order by ulceration, xerostomia, 56 57 gingivitis, atrophic glossitis, candidiasis, and angular cheilitis. Overall, patients with GORD and DE had significantly 58 59 more oral manifestations than did those without DE. There was a trend of worse dentition status, as indicated by a 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml higher mean DMFT index value in group A than in group B, suggesting that teeth with DE were more vulnerable to BMJ Open Page 8 of 17 decay and tooth loss. Patients with a worse dietary pattern were more likely to develop DE with GORD than did those

with a good dietary pattern (OR 6.034, 95% CI 1.889–19.268; Table 4). In addition, nausea/vomiting (OR 0.349, 95% BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 CI 0.160–0.759) and early satiety/loss of appetite (OR 1.903, 95% CI 1.020–3.551) increased the likelihood of 3 4 developing DE by 0.349fold and 1.903fold, respectively. Patients with peptic ulcer disease were more likely to 5 6 develop DE (OR 2.825, 95% CI 1.075–7.423), as were those with gastritis (OR 2.013; CI 0.970–4.179) and 7 oesophagitis (OR 2.398, 95% CI 0.621–9.255). 8 9 10 11 12 Table 3. Oral manifestations in patients with GORD 13 14 15 ORAL MANIFESTATIONS Group A Group B ALL P-value 16 For peer(n=66) review (n=121) only (n=187) 17 Xerostomia Yes 47 42 89 (47.6%) <0.01* 18 No 19 79 98 (52.4%) 19 Ulceration Yes 60 51 111 (59.4%) <0.01* 20 No 6 70 76 (40.6%) 21 Gingivitis Yes 47 26 73 (39.0%) <0.01* 22 No 19 95 114 (61.0%) 23 Candidiasis Yes 20 20 40 (21.4%) 0.028* 24 No 46 101 147 (78.6%) 25 Angular Cheilitis Yes 24 12 36 (19.3%) <0.01* 26 No 42 109 151 (80.7%) 27 Atrophic Glossitis Yes 30 27 57 (30.5%) 0.001* 28 No 36 94 130 (69.5%) 29 Leukoplakia Yes 2 6 08 (4.3%) 0.533 30 No 64 115 179 (95.7%) 31 Oral Submucosal Fibrosis Yes 53 71 124 (66.3%) 0.003* 32 No 13 50 63 (33.7%) DMFT status <3 (acceptable) 24 64 88 (47.06%) 0.069 33 http://bmjopen.bmj.com/ 34 410 (average) 24 37 61 (32.62%) 35 >10 (poor) 18 20 38 (20.32%) 36 *p<0.05 (chisquared test) was considered significant. GORD, gastrooesophageal reflux disease. Group A, chronic/severe GORD with DE. 37 Group B, mild GORD without DE. DMFT (decayed, missed, filled, trauma), pertaining to tooth decay and tooth loss. 38 39 40 The results of the multivariable logistic regression are also presented in Table 4, with adjustments made for all

41 on October 1, 2021 by guest. Protected copyright. 42 independent variables. Some variables that had a statistically significant association with DE in bivariate analysis (i.e., 43 44 peptic ulcer disease, gastritis, and hiatal hernia) were no longer significant in multivariable analysis. Portal gastropathy 45 46 had a pvalue of >0.2; therefore, was excluded from multivariable analysis. However, dietary pattern, nausea/vomiting, 47 loss of appetite, and oesophagitis showed a statistical significant association with DE. Further, all these variables 48 49 (except for nausea/vomiting) were associated with an increased likelihood of developing DE in multivariable logistic 50 51 regression analysis. 52 53 54 55 Table 4. Factors associated with GORD and dental erosions: results of univariate and multivariable analysis. 56 57 CHARACTERISTICS GORD with Dental Erosions 58 C.OR (95% CI) p-value A.OR (95% CI) p-value 59 1. Dietary Pattern <0.01* <0.001* 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Good 1 1 Average 1.096 (0.3253.697) 1.271(0.3454.684) Page 9 of 17 BMJ Open Worse 6.034 (1.88919.268) 6.437(1.81822.791) 2. GI symptom: Nausea/vomiting 0.008* 0.005* BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 No 1 1 2 Yes 0.349 (0.1600.759) 0.277 (0.1120.683) 0.043* 0.058* 3 3. GI symptom: Early satiety/loss of appetite No 1 1 4 Yes 1.903 (1.0203.551) 2.018 (0.9754.177) 5 4. GI disorder: Esophagitis 0.20 0.028* 6 No 1 1 7 Yes 2.398 (0.6219.255) 5.911 (1.21228.831) 8 5. GI disorder: Peptic ulcer disease (PUD) 0.035* 0.125 9 No 1 1 10 Yes 2.825 (1.0757.423) 2.468 (0.7797.821) 11 6. GI disorder: Hiatal hernia 0.20 0.341 12 No 1 1 13 Yes 0.251 (0.032.082) 0.337 (0.0363.150) 14 7. GI disorder: Portal gastropathy 0.214 15 No 1 16 ForYes peer1.786 review(0.7154.458) only 17 8. GI disorder: Gastritis 0.06 0.218 18 No 1 1 19 Yes 2.013 (0.9704.179) 1.718 (0.7264.066) 20*p < 0.05 was considered statistically significant. A.OR, odds ratios adjusted for all independent variables; 95% CI, confidence interval; C.OR, 21crude/unadjusted odds ratios. 22 23 24 Table 5 shows the OHIP14 subscale scores for groups A and B. The mean values for all subscales were relatively 25 26 higher in group A, suggesting a higher risk of ‘poor oralHRQoL’ than that in group B. The coexistence of GORD and 27 28 DE had a notable negative impact on oralHRQoL in terms of physical disability (rs=0.303), psychological discomfort 29 (rs=0.280), psychological disability (rs=0.254), and functional limitation (rs=0.204). All subscales were positively 30 31 correlated in patients with GORD and DE (p < 0.05). 32

33 http://bmjopen.bmj.com/ 34 Table 5. Psychometric properties of OHIP14 correlated with GORD with and without dental erosions 35 36 37 OHIP-14 Sub-scales GORD + DENTAL EROSIONS 38 39 Characteristics With DE (group A) Without DE (group B) Statistics p-value* 40 N= 66 N= 121 Spearman %= 35.3 %= 64.7 Ranks (rs) 41 on October 1, 2021 by guest. Protected copyright. 42 MEAN subscale (S.D) MEAN subscale (S.D) correlation 43 1. Functional Limitation 3.5303 (2.19944) 2.5702 (1.92712) 0.204 0.005 44 2. Physical Pain 1.1818 (1.26380) 0.7273 (0.96609) 0.167 0.023 45 3. Psychological Discomfort 5.1667 (1.85293) 4.0744 (1.98396) 0.280 0.001 46 4. Physical Disability 6.3939 (2.74495) 4.4463 (2.81351) 0.303 0.001 47 5. Psychological Disability 4.5909 (1.74512) 3.5868 (1.94795) 0.254 0.001 48 6. Social Disability 7.1061 (2.83456) 6.1405 (3.42127) 0.148 0.043 49 7. Handicap 1.6818 (1.22987) 1.2149 (1.31153) 0.193 0.008 50 *Correlation is significant at p<0.05 level (2tailed). 51 52 53 54 DISCUSSION 55 56 57 GORD with DE and its associated risk factors has been the subject of much research interest in recent years. The 58 59 present study focused on patients who had GORD with or without DE and recorded the prevalence, distribution, and 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 17 severity of this clinical entity in a relatively large sample. The influences of known risk indicators and other possible

determinants that have been less well studied in the past were also investigated. BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 4 GORD has multiple systemic effects, particularly in the GI tract (Table 2). There is considerable medical literature on 5 6 GORD and its effects on oral hard tissue (i.e. DE), as well as its periodontal effects;[7] however, data on the 7 relationship between acidic oral mucosal lesions and oral health are scarce.[7, 10, 13, 14] Our finding of a prevalence 8 9 of DE in patients with GORD of 35.3% fits well within the reported range of 5.00%–58.41%.[34] 10 11 12 13 Our results indicate that xerostomia, ulceration, gingivitis, and OSF are the most frequent oral symptoms associated 14 with GORD. This is in accordance with previous studies.[4, 8, 25] There have been some studies of the prevalence of 15 16 xerostomia in patients with GORD,For butpeer with controversial review results. Salivaonly is considered to play a leading role in 17 18 protecting the oesophageal mucosa against gastric reflux, and its qualitative (e.g., deficiencies of salivary growth 19 factor and cytokines) and quantitative (e.g., hyposalivation) abnormalities have been linked to the pathogenesis of 20 21 GORD,[24, 2628] compromised dental health, and DE.[29] The pathogenesis of GORD appears to be connected with 22 23 decreased production of saliva, and our findings suggest that xerostomia should be included in the extraoesophageal 24 25 symptomatology of GORD.[33] 26 27 28 Approximately 60% of participants had ulcerative lesions on the soft and hard palate mucosa, buccal mucosa, uvula, 29 30 and tongue, which could not be characterised clinically as any other disease; these lesions are typically recognised as 31 32 ‘soft tissue aphthoidlesions’ related to GORD,[4, 8, 25] and are caused directly by the corrosive effects of refluxed

33 acid. http://bmjopen.bmj.com/ 34 35 36 37 A recent study in a rat model of the soft palate by Habesoglu et al,[6] identified epithelial and microscopic alterations 38 in response to acid injury. Interest in this phenomenon dates to the late 1980s, when a study by Järvinen et al, 39 40 demonstrated a marked prevalence of oral mucosal changes in the presence of GORD, including burning mouth,

41 on October 1, 2021 by guest. Protected copyright. 42 aphthoid lesions, erythema of the soft palate and uvula, glossitis, and epithelial atrophy.[8] However, these findings 43 44 were called into question by the studies of Ranjitkar et al,[10] and Petruzzi et al,[14] who found that mucosal changes 45 are quite common and not an entity specific for GORD. Deppe et al,[7] also studied the effects of GORD on the oral 46 47 mucosa and found positive and negative signs for erythema and ulceration, respectively, but could not find a 48 49 statistically significant difference between their erosive and nonerosive reflux groups. Similarly, a study by Meurman 50 51 et al,[5] found no mucosal changes that could be linked to GORD. Following on from the findings of Rajalalitha et 52 al,[15] we hypothesized that the chronicacid injury caused by GORD would be a source of persistent irritation and 53 54 inflammation affecting the oral mucosa, ultimately resulting in the infiltration of inflammatory markers. These 55 56 markers include increased populations of interleukins, cytokines, tumour necrosis factoralpha, interferonalpha, and 57 growth factors such as transforming growth factorbeta, that are produced at the site of inflammation. This mechanism 58 59 of inflammation is clearly exaggerated in situations of chronic and constant irritation such as GORD and supra 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml oesophageal reflux disease.[1517, 21, 34] This finding is accompanied by epithelial atrophy and fibroelastic changes Page 11 of 17 BMJ Open in the lamina propria.[15, 30] This mechanism is consistent with the finding of significant oral submucosal fibrosis in

our sample, which can lead to metaplastic changes in the mucosa and can also cause restricted mouth opening, BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 resulting in trismus, eating difficulties, impaired speech, and generalised impairment of oral health.[1517] Our study 3 4 findings clearly demonstrate a positive association of chronic GORD with OSF, which is a precancerous lesion.[30] 5 6 OSF has never been reported in association with GORD; however, because of its significant impact on quality of life, 7 it should be considered and addressed in future research.[12, 1418, 2029, 3335] 8 9 10 11 With chronic GORDmediated acid exposure, the salivary gland epithelium sustains severe damage (i.e. epithelial 12 13 metaplasia), resulting in xerostomia.[24, 34] This association may further aggravate acidmediated orooesophageal 14 epithelium injury, resulting in GORDrelated oesophagitis.[9, 24, 31]. Other conditions, including oral ulceration, 15 16 associated infections (candidiasisFor and angular peer cheilitis), review and gustatory dysfunction, only are also common.[31] 17 18 19 Correa et al noted that patients with chronic GORD had reduced salivary buffering capacity and concluded that this 20 was the predominant cause of tooth erosion.[20] The same study also reported a lower prevalence of dental caries, 21 22 which was attributed to the low prevalence of cariogenic bacteria (Lactobacilli and Streptococci) in the saliva of 23 24 patients with chronic GORD; this may be a possible association with GORDinduced DE that needs confirmation in 25 future studies. However, our findings with regard to dental caries are in contrast with those of Correa et al in that the 26 27 mean DMFT index values were higher in the group with GORD and DE (DMFT > 4 in 63.6%) than in the group with 28 29 GORD and no DE (DMFT > 4 in 47.1%). We postulate that the additive effects of direct acid injury, low salivary 30 31 buffering capacity, and increased opportunistic bacterial populations cause this marked increase in tooth decay and 32 loss (as indicated by increased DMFTindex values in our study) in the highrisk population with GORD and DE.[18]

33 http://bmjopen.bmj.com/ 34 However, it can also be attributed to poor oral hygiene status, in that 27% of our study population were smokers and 35 36 30% chewed tobacco. An acidic/unfavourable dietary pattern was common, as was excessive consumption of 37 38 traditional beverages [i.e. sugary chai: tea with milk (73%)], which would also have contributed to a low oral pH and 39 provided a favourable environment for opportunistic bacteria causing tooth decay and tooth loss.[39, 40] 40

41 on October 1, 2021 by guest. Protected copyright. 42 There is a wealth of literature identifying DE as being comorbid with GORD, and a review by Ranjitkar et al, reported 43 44 the median prevalence for DE in patients with GORD to be 24%.[18] Accordingly, the Montreal consensus postulated 45 that the prevalence of DE is directly related to GORD, particularly when noted on the lingual and palatal tooth 46 47 surfaces.[2] Our study findings are consistent with the concept that DE has a significant association with GORD and 48 49 may serve as a marker of disease severity. As in a study by Meurman et al,[5] our statistical comparison of patients 50 who had GORD with and without DE demonstrated that, unlike subjective oral symptoms, oral manifestations were 51 52 significantly more common in the group with GORD and DE. 53 54 55 56 Considering the findings of Meurman et al,[5] we used the validated OHIP14 instrument,[13] instead of subjective 57 oral symptoms to assess the impact of severity of GORD (with DE) on oral health from a psychological and general 58 59 wellbeing perspective. Until now, this tool has not been tested in such a patient population. Using the OHIP14, we 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml found that GORD with DE was significantly correlated (p < 0.05) with the OHIP14 subscale psychometric BMJ Open Page 12 of 17 characteristics of physical disability, psychological discomfort, psychological disability, functional limitation,

handicap, physical pain, and social disability (in decreasing order of frequency). This result is consistent with our BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 finding that GI conditions and their oral manifestations have a considerable adverse impact on oral health. Notably, in 3 4 our study, 43.3% of the study population had GORD alone, and the rest had additional upper GI conditions, including 5 6 gastritis, portal gastropathy, peptic ulcer disease, GORDrelated oesophagitis, and hiatal hernia, which are believed to 7 be factors that initiate and/or lead to progression of GORD.[3638] Further, a good proportion of patients self 8 9 medicated with NSAIDs for pain relief, which would have aggravated their upperGI illness further.[36] 10 11 12 This study has some limitations. First, the staging of DE was not recorded. Second, only one investigation was used to 13 diagnose GORD, and no measures of disease severity, such as manometry, 24hour pH monitoring, or biopsies, were 14 15 included.[20] Third, other oral manifestations of GORD that might have a causal relationship with DE, such as 16 For peer review only 17 periodontal disease and gingival disease, were not recorded in detail. Moreover, we did not control for the effect of 18 19 treatment for GORD. Patients on treatment for GORD are more likely to have controlled disease, so chronic symptoms 20 of GORD, including DE and other oral manifestations, would be relatively less prevalent in this group. Further, an 21 22 aggravating effect of proton pump inhibitor therapy on xerostomia cannot be ruled out, and most patients in our study 23 24 were receiving these agents. Fourth, we could not rule out xerostomia being the result of an interplay of various 25 comorbidities, such as diabetes or hepatic disease, or a side effect (albeit minor) of medication, including 26 27 antihistamines, proton pump inhibitors, calcium channel blockers, and betablockers.[5, 24, 31] Further trials and 28 29 clinical studies are needed to rule out these potentially confounding factors. Future studies in humans should include 30 31 biopsies of the oral epithelium to correlate clinical and histological findings. Our present findings suggest that the 32 Montreal consensus recommendations should now be expanded to include xerostomia, OSF, aphthoidulcerative

33 http://bmjopen.bmj.com/ 34 lesions, and candidiasis in the extraoesophageal symptomatology of GORD.[2, 35] 35 36 37 Dental health is widely neglected in Pakistan, where dentalvisits are restricted to a downstream approach.[32] Further, 38 general practitioners and gastroenterologists are often the main health care providers for patients with GORD, but 39 40 while addressing their main GI concerns, the oral manifestations of these systemic conditions are often overlooked.

41 on October 1, 2021 by guest. Protected copyright. 42 This study highlights the need for dental referral in patients with upper GI disorders (in this case GORD), which can 43 44 have a marked effect on both systemic and oral health. 45 46 47 48 49 CONCLUSION 50 51 52 We found a positive correlation of GORD with DE and oral conditions, such as xerostomia and mucosal ulceration, in 53 our study population. Patients with DE had more severe oral symptoms and more compromised oral health than did 54 55 those without DE. Further, acidic dietary patterns increase the likelihood of a more aggressive form of the disease. We 56 57 recommend timely dental visits for evaluation of oral health, to assess the severity of DE and prevent its progression, 58 as well as dietary control. We also urge gastroenterologists to perform routine oral examinations and make referrals to 59 60 a dentist to prevent exacerbationFor peer of review the oral only conditions - http://bmjopen.bmj.com/site/about/guidelines.xhtml caused by GORD. Page 13 of 17 BMJ Open Acknowledgements BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 The authors acknowledge the participating hospitals and respective gastroenterologists for supervising this research 2 3 and providing ethical and institutional/departmental permission to conduct this study, namely: Dr. Tayyab Saeed 4 5 Akhtar (Holy Family Hospital), Dr. Farzana Memon (CHK), Dr. Saeed Qureshi (CHK), Dr. Mashood (PIMS), Dr. 6 7 Waseem Said (PIMS), Dr. Waqar (JPMC). The study would not have been possible without the efforts of the 8 following clinicians who were trained and calibrated as the data collection team at four participating hospitals: Dr. 9 10 Sumbul Zakir, Dr. Iqra Raja, Dr. Rimsha Hashmi, Dr. Javeria Ahmed, Dr. Emmad Naeem Khan and Dr. Ibrahim 11 12 Warsi. We would also like to thank Editage for English language editing. 13 14 Declaration of interests 15 16 For peer review only 17 All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: no 18 19 support from any organization for the submitted work; no financial relationships with any organizations that might 20 have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to 21 22 have influenced the submitted work. 23 24 25 Funding 26 27 None 28 29 30 Data sharing statement 31 32 No additional data are available

33 http://bmjopen.bmj.com/ 34 35 Author contributions 36 37 IW, study conception, data collection/analysis, manuscript writing and editing; JA, data collection and data 38 39 entry/analysis; AY, study design, manuscript editing; AR, statistical analysis and proofreading; RH, data collection 40 and data entry; TA, supervision of gastroenterology research (clinical supervisor) and proofreading. 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 License to Publisher 45 46 The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, a 47 48 worldwide license (http://www.bmj.com/sites/default/files/BMJ%20Author%20Licence%20March%202013.doc) to 49 the Publishers and its licensees in perpetuity in all forms. 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 17 REFERENCES BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 1. Patel A, Amaechi BT, Brady C. Prevention and control of dental erosion: gastroesophageal reflux disease management. 4 Dental Erosion and Its Clinical Management. Springer International Publishing 2015:203224. 5 2. Vakil N, Van Zanten SV, Kahrilas P, Dent J, Jones R. The Montreal definition and classification of gastroesophageal reflux 6 disease: a global evidencebased consensus. The American journal of gastroenterology 2006;101(8):190020. 7 3. Pace F, Pallotta S, Tonini M, Vakil N, Bianchi Porro G. Systematic review: gastro‐oesophageal reflux disease and dental 8 lesions. Alimentary pharmacology & therapeutics 2008;27(12):117986. 9 4. Lazarchik DA, Filler SJ. Effects of gastroesophageal reflux on the oral cavity. The American journal of medicine 10 1997;103(5):10713. 11 5. Meurman JH, Toskala J, Nuutinen P, Klemetti E. Oral and dental manifestations in gastroesophageal reflux disease. Oral 12 surgery, oral medicine, oral pathology 1994;78(5):58389. 13 6. Habesoglu TE, Habesoglu M, Sürmeli M, Deveci I, Toros SZ, Gunes P, et al. Histological changes of rat soft palate with 14 exposure to experimental laryngopharyngeal reflux. Auris Nasus Larynx 2010;37(6):73036. 15 7. Deppe H, Mücke T, Wagenpfeil S, Kesting M, Rozej A, Bajbouj M, et al. Erosive esophageal reflux vs. non erosive 16 esophageal reflux: oral findingsFor in 71 patients. peer BMC oralreview health 2015;15(1):84. only 17 8. Jarvinen V, Meurman JH, Hyvarinen H, Rytomaa I, Murtomaa H. Dental erosion and upper gastrointestinal disorders. Oral 18 19 Surg Oral Med Oral Pathol 1988;65:298–303. 20 9. Friesen LR, Bohaty B, Onikul R, Walker MP, Abraham C, Williams KB, et al. Is histologic esophagitis associated with 21 dental erosion: a crosssectional observational study. BMC oral health 2017;17(1):116. 22 10. Ranjitkar S, Smales RJ, Kaidonis JA. Oral manifestations of gastroesophageal reflux disease. Journal of gastroenterology and 23 hepatology 2012;27(1):2127. 24 11. Daley TD, Armstrong JE. Oral manifestations of gastrointestinal diseases. Canadian Journal of Gastroenterology and 25 Hepatology 2007;21(4):24144. 26 12. Picos AM, Poenar S, Opris A, Chira A, Bud M, Berar et al. Prevalence of dental erosions in GERD: a pilot study. Clujul 27 Medical 2013;86(4):344–346. 28 13. Batra M, Aggarwal VP, Shah AF, Gupta M. Validation of Hindi version of oral health impact profile14 for adults. J Indian 29 Assoc Public Health Dent 2015;13:46974 30 14. Petruzzi M, Lucchese A, Campus G, Crincoli V, Lauritano D, Baldoni E. Oral stigmatic lesions of gastroesophageal reflux 31 disease (GERD). Rev Med Chil 2012;140(7):91518. 32 15. Rajalalitha P, Vali S. Molecular pathogenesis of oral submucous fibrosis–a collagen metabolic disorder. Journal of oral

33 pathology & medicine 2005;34(6):32128. http://bmjopen.bmj.com/ 34 16. Pindborg JJ, Sirsat SM. Oral submucous fibrosis. Oral Surgery, Oral Medicine, Oral Pathology. 1966;22(6):76479. 35 17. Hogaboam CM, Steinhauser ML, Chensue SW, Kunkel SL. Novel roles for chemokines and fibroblasts in interstitial fibrosis. 36 Kidney international 1998;54(6):215259. 37 38 18. Ranjitkar S, Kaidonis JA, Smales RJ. Gastroesophageal reflux disease and tooth erosion. International journal of dentistry 39 2011;2012. 40 19. Kramer IR, Pindborg JJ, Bezroukov V, Infirri JS. Guide to epidemiology and diagnosis of oral mucosal diseases and conditions. World Health Organization. Community dentistry and oral epidemiology 1980;8(1):126. 41 on October 1, 2021 by guest. Protected copyright. 42 20. Corrêa MC, Lerco MM, Cunha MD, Henry MA. Salivary parameters and teeth erosions in patients with gastroesophageal 43 reflux disease. Arquivos de gastroenterologia 2012;49(3):21418. 44 21. Scott DR, Simon RA. Supraesophageal reflux disease: A review of the literature. Allergy and asthma proceedings. OceanSide 45 Publications, Inc 2014;35(2):104110. 46 22. IsmailBeigi F, Horton PF, Pope CE. Histological consequences of gastroesophageal reflux in man. Gastroenterology 47 1970;58(2):16374. 48 23. Silva MA, Damante JH, Stipp AC, Tolentino MM, Carlotto PR, Fleury RN. Gastroesophageal reflux disease: new oral 49 findings. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology 2001;91(3):30110. 50 24. Korsten MA, Rosman AS, Fishbein S, Shlein RD, Goldberg HE, Biener A. Chronic xerostomia increases esophageal acid 51 exposure and is associated with esophageal injury. The American journal of medicine 1991;90(6):70106. 52 25. Meurman JH, Rantonen P. Salivary flow rate, buffering capacity, and yeast counts in 187 consecutive adult patients from 53 Kuopio, Finland. European Journal of Oral Sciences 1994;102(4):22934. 54 26. Kongara K, Varilek G, Soffer EE. Salivary growth factors and cytokines are not deficient in patients with gastroesophageal 55 reflux disease or Barrett's esophagus. Digestive diseases and sciences 2001;46(3):60609. 56 27. Rourk RM, Namiot Z, Edmunds MC, Sarosiek J, Yu Z, McCallum RW. Diminished luminal release of esophageal epidermal 57 58 growth factor in patients with reflux esophagitis. American Journal of Gastroenterology 1994;89(8). 59 28. Kao CH, Ho YJ, ChangLai SP, Liao KK. Evidence for decreased salivary function in patients with reflux esophagitis. 60 Digestion 1999;60(3):19195.For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 29. Jarvinen VK, Rytomaa II, Heinonen OP. Risk factors in dental erosion. Journal of dental research 1991;70(6):94247. Page 15 of 17 BMJ Open 30. Khan S, Chatra L, Prashanth SK, Veena KM, Rao PK. Pathogenesis of oral submucous fibrosis. Journal of cancer research and therapeutics 2012;8(2):199. BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 31. Sreebny LM, Valdini A. Xerostomia: a neglected symptom. Archives of internal medicine 1987;147(7):13337. 2 32. Harchandani N. Oral health challenges in Pakistan and approaches to these problems. Pakistan Oral & Dental Journal 3 2012;32(3). 4 33. Di Fede O, Di Liberto C, Occhipinti G, Vigneri S, Lo Russo L, Fedele S, et al. Oral manifestations in patients with gastro‐ 5 oesophageal reflux disease: a single‐center case–control study. Journal of oral pathology & medicine 2008;37(6):33640. 6 34. RoeschRamos L, RoeschDietlen F, RemesTroche JM, RomeroSierra G, MataTovar Cde J, AzamarJacome AA, et al. 7 Dental erosion, an extraesophageal manifestation of gastroesophageal reflux disease. The experience of a center for digestive 8 physiology in Southeastern Mexico. Rev Esp Enferm Dig 2014;106(2):927. 9 35. Pauwels, ANS. Dental erosions and other extraoesophageal symptoms of gastrooesophageal reflux disease: Evidence, 10 treatment response and areas of uncertainty. United European Gastroenterology Journal 2015;3(2):166170. 11 36. Pruthi S, Nirupama M, Chakraborti S. Evaluation of gastric biopsies in chronic gastritis: Grading of inflammation by Visual 12 Analogue Scale. Medical Journal of Dr. DY Patil University 2014;7(4):463. 13 37. AlSaadi AM, AlKhayat JQ, Muhammad IM, Anwar SA. The role of Helicobacter pylori in esophagitis and peptic ulcer 14 15 disease in Iraq. Saudi Med J 2004;25(9):121622. 16 38. Schechter RB, Lemme EM,For Coelho HS.peer Gastroesophageal review reflux in cirrhoticonly patients with esophageal varices without 17 endoscopic treatment. Arquivos de gastroenterologia 2007;44(2):14550. 18 39. Stephan RM. Intraoral hydrogenion concentrations associated with dental caries activity. Journal of dental research 19 1944;23(4):25766. 20 40. Lussi A, Jaeggi T, Zero D. The role of diet in the aetiology of dental erosion. Caries research 2004;38(1):3444. 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

7 cross-sectional studies 8 GORD-study STROBE checklist of items, that should be included in reports of . 9 Items ticked in ‘blue’ () are included in manuscript that has been submitted, while items not ticked are not 10 applicable for our study. 11 Item 12 No Recommendation 13 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the 14 15 abstract 16 For peer (b) Provide in reviewthe abstract an informative only and balanced summary of what was 17 done and what was found 18 19 Introduction 20 Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 21 Objectives 3 State specific objectives, including any prespecified hypotheses 22 Methods 23 24 Study design 4 Present key elements of study design early in the paper 25 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, 26 exposure, follow-up, and data collection 27 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of 28 participants 29 30 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect 31 modifiers. Give diagnostic criteria, if applicable 32 Data sources/ 8* For each variable of interest, give sources of data and details of methods of

33 measurement assessment (measurement). Describe comparability of assessment methods if there is http://bmjopen.bmj.com/ 34 more than one group 35 Bias 9 Describe any efforts to address potential sources of bias 36 37 Study size 10 Explain how the study size was arrived at 38 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, 39 describe which groupings were chosen and why 40 Statistical methods 12 (a) Describe all statistical methods, including those used to control for

41 on October 1, 2021 by guest. Protected copyright. confounding 42 43 (b) Describe any methods used to examine subgroups and interactions 44 (c) Explain how missing data were addressed 45 (d) If applicable, describe analytical methods taking account of sampling strategy 46 (e) Describe any sensitivity analyses 47 48 Results 49 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially 50 eligible, examined for eligibility, confirmed eligible, included in the study, 51 completing follow-up, and analysed 52 53 (b) Give reasons for non-participation at each stage 54 (c) Consider use of a flow diagram 55 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and 56 information on exposures and potential confounders 57 (b) Indicate number of participants with missing data for each variable of interest 58 59 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 17 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 Outcome data 15* Report numbers of outcome events or summary measures 3 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates 4 and their precision (eg, 95% confidence interval). Make clear which confounders 5 were adjusted for and why they were included 6 b 7 ( ) Report category boundaries when continuous variables were categorized 8 (c) If relevant, consider translating estimates of relative risk into absolute risk for a 9 meaningful time period 10 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 11 sensitivity analyses 12 13 Discussion 14 Key results 18 Summarise key results with reference to study objectives 15 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or 16 For imprecision.peer Discuss review both direction and magnitudeonly of any potential bias 17 18 Interpretation 20 Give a cautious overall interpretation of results considering objectives, 19 limitations, multiplicity of analyses, results from similar studies, and other relevant 20 evidence 21 Generalisability 21 Discuss the generalisability (external validity) of the study results 22 23 Other information 24 Funding 22 Give the source of funding and the role of the funders for the present study and, if 25 applicable, for the original study on which the present article is based 26 27 28 I declare that the checklist maintained in reference to the submitted original research article, is correct to the best of my 29 30 knowledge. For any concerns, you can direct them to me directly. 31 32 Thank you for your consideration. I look forward to hearing from you.

33 http://bmjopen.bmj.com/ 34 Sincerely, 35 36 37 Dr. Ibrahim Warsi 38 Dow University of Health Sciences, Karachi, Pakistan (74200) 39 178, Street 18, Sector F-10/2, Islamabad, Pakistan (44000) 40 Mobile # 0092-3328262980

41 on October 1, 2021 by guest. Protected copyright. Email: [email protected] 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 2 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from

Risk factors associated with oral manifestations, and oral health impact of gastro-oesophageal reflux disease: A multicentre, cross-sectional study in Pakistan ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-021458.R1

Article Type: Research

Date Submitted by the 16-Oct-2018 Author:

Complete List of Authors: Warsi, Ibrahim; Harvard Medical School, Masters in medical science & Clinical Investigation Ahmed, Javeria; Jinnah Sindh Medical University, Oral and Maxillofacial surgery (Jinnah postgraduate medical center, JPMC) Younus, Anjum; Dow University of Health Sciences, Department of Community Dentistry Rasheed, Abdur; Dow University of Health Sciences, Department of Research and Biostatistics Akhtar, Tayyab; Rawalpindi Medical College, Gastroenterology and Liver Centre, Holy Family Hospital Ain, Qurrat; Shalamar Hospital, Internal Medicine Khurshid, Zohaib; King Faisal University, Department of Prosthodontics and Implantology http://bmjopen.bmj.com/

Primary Subject Dentistry and oral medicine Heading:

Gastroenterology and hepatology, General practice / Family practice, Secondary Subject Heading: Epidemiology, Health services research, Public health

Gastroenterology < INTERNAL MEDICINE, PUBLIC HEALTH, PREVENTIVE Keywords: MEDICINE, ORAL MEDICINE, Endoscopy < GASTROENTEROLOGY,

EPIDEMIOLOGY on October 1, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 Risk factors associated with oral manifestations, and oral health impact of gastro- 2 3 oesophageal reflux disease: A multicentre, cross-sectional study in Pakistan 4 5 Ibrahim Warsi1,7, Javeria Ahmed1,8, Anjum Younus1,2, Abdur Rasheed3, Tayyab Akhtar4, Qurrat-ul-Ain5, Zohaib 6 Khurshid6 7 8 1Dr. Ishrat-ul-Ebad Khan Institute of Oral Health Sciences, Dow University of Health Sciences, Karachi, Pakistan 9 2Department of Community Dentistry, Dow Dental College, Dow University of Health Sciences, Karachi, Pakistan. 10 3Department of Research, Dow University of Health Sciences, Karachi, Pakistan 11 4 12 Gastroenterology and Liver Centre, Holy Family Hospital, Rawalpindi-Islamabad, Pakistan 5 13 Department of Internal Medicine, Shalamar Hospital, Lahore. 14 6Department of Prosthodontics and Implantology, College of Dentistry, King Faisal University, Saudi Arabia. 15 7Masters in Medical Science and Clinical Investigation Program, Harvard Medical School, Harvard University, Boston, MA, USA 16 8Department of Oral & MaxillofacialFor Surgery, peerJinnah Postgraduate review Medical Center (JPMC),only Jinnah Hospital, Karachi, Pakistan 17 18 19 Ibrahim Warsi (IW) 20 Postgraduate student Masters (medical sciences and clinical investigation), Harvard Medical School, Boston, MA, USA 21 22 23 Javeria Ahmed (JA) 24 Post-graduate trainee (Oral & Maxillofacial Surgery), Jinnah Postgraduate Medical Center (JPMC), Jinnah Hospital, Karachi, Pakistan 25 26 Anjum Younus (AY) 27 Postgraduate trainee MDS, Department of Community Dentistry, Dr. Ishrat-ul-Ebad Khan Institute of Oral Health Sciences, Dow University 28 of Health Sciences, Karachi, Pakistan 29 30 Abdur Rasheed (AR) 31 Deputy director of research & biostatistician, Department of Research, Dow University of Health Sciences, Karachi, Pakistan 32

33 http://bmjopen.bmj.com/ 34 Tayyab Akhtar (TA) 35 Senior Registrar, Gastroenterology and Liver Center, Holy Family Hospital, Rawalpindi-Islamabad, Pakistan 36 37 Qurrat-ul-Ain (QA) 38 Internal Medicine Resident, Shalamar Hospital, Lahore, Pakistan 39 40 Zohaib Khurshid (ZK)

41 Lecturer and Course Coordinator, Department of Prosthodontics and Implantology, College of Dentistry, King Faisal University, Al-Ahsa on October 1, 2021 by guest. Protected copyright. 42 31982, Saudi Arabia 43 44 45 46 Corresponding Author: Dr. Ibrahim Warsi 47 614-A, Shawmut Avenue, Unit-4, Boston, MA (USA), 02118 48 Tel: (+1)832-458-5039 49 Email: [email protected] 50 51 52 WORD COUNT: 53 MAIN TEXT: 3174 ~ 3876~4182 54 ABSTRACT: 298 ~ 304 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 21 BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 ABSTRACT 2 3 4 Objective: Gastro-oesophageal reflux disease (GORD) is a relatively common disorder and manifests with extra- 5 6 oesophageal symptoms, such as dental erosions (DE), cough, laryngitis, asthma, and oral soft/hard tissue pathologies. 7 8 This study aimed (i) to identify oral soft and hard-tissue changes in patients with GORD and (ii) to evaluate these oral 9 10 changes as indices for assessing GORD and its severity. 11 Setting: This cross-sectional study was conducted at four major tertiary-care government hospitals, in two metropolitan 12 13 cities of Pakistan. 14 15 Participants: In total, 187 of 700 patients who underwent oesophago–gastro–duodenoscopy (OGD) and having GORD 16 were included in the study. TheFor GORD peer patients were review divided according only to the presence of dental erosions (DE) into 17 18 group A (with DE, chronic/severe GORD) and group B (without DE, mild GORD). Patients who were unconscious 19 20 and extremely limited mouth opening were excluded. 21 22 Primary & secondary outcome measures: Abnormal conditions and lesions of the oral mucosa were recorded. The 23 impact of oral hard & soft-tissue changes on the oral health-related quality-of-life (OHRQoL) was assessed using the 24 25 Pakistani (Urdu) version of the validated Oral Health Impact Profile-14 (OHIP-14) instrument. 26 27 Results: Oral submucosal fibrosis (66.3%), ulceration (59.4%), and xerostomia (47.6%) were significantly more 28 29 common in group A (p<0.05). The prevalence of GORD was 26.7%, within which the prevalence of DE was 35.3%. 30 Unhealthy dietary pattern, nausea/vomiting, oesophagitis, xerostomia, ulceration, gingivitis, and angular cheilitis 31 32 showed a statistically significant association with DE. All subscales of OHIP-14 were positively correlated (p<0.05) in

33 http://bmjopen.bmj.com/ 34 patients with GORD and DE, with notable impact on psychological discomfort (rs=0.297), physical disability 35 (rs=0.288), psychological disability (rs=0.271), and functional limitation (rs=0.200). 36 37 Conclusion: Patients with GORD and DE presented with more severe oral symptoms than did those with GORD and 38 39 no DE. We recommend a healthy diet and timely dental check-ups to assess the severity of both systemic and oral 40 disease.

41 on October 1, 2021 by guest. Protected copyright. 42 Keywords: GORD, dental erosions, oral manifestations, oral hard and soft-tissue changes, extra-oesophageal 43 44 symptoms, oral health, OHIP-14, OHRQoL 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 4 5 6 7 STRENGTH & LIMITATIONS OF THIS STUDY: 8 9 10 11 o Patients with GORD and DE presented with more severe oral symptoms than did those with GORD and no DE. 12 o This study is the first of its kind to bridge the gap created by either lacking evidence or controversial literature on 13 14 the effect of GORD on oral soft tissues. Moreover, this cross-sectional study lays the foundation for future 15 16 researchers to identify a temporalFor relationship peer in thisreview specific domain, only through randomized clinical trials. 17 A novel element of this study is also the use of the validated OHIP-14 instrument (for assessment of the Oral health- 18 o 19 related quality of life) in this (GORD) population for the first time. While Chronic-GORD reflects as oral lesions 20 21 and abnormal mucosal conditions, it also poorly affects oral health and quality-of-life; having a significant impact 22 23 on psychological discomfort, physical disability, psychological disability, and functional limitation. 24 o The significant highlight of this study is that the ‘Montreal consensus for extra-oesophageal symptomatology of 25 26 GORD,’ should now be expanded to include xerostomia, oral submucosal fibrosis, aphthoid-ulcerative lesions, and 27 28 candidiasis. Furthermore, this study highlights the need for dental referral in patients with upper-GI disorders (in 29 30 this case GORD), which can have a marked effect on both oral and systemic health. 31 o This study has a few limitations of note. Patients on the treatment of GORD have controlled disease thus it would 32

33 under-estimate the associations found (between GORD & oral manifestations). Also, some oral side-effects of http://bmjopen.bmj.com/ 34 35 prescription could not be ruled out (such as the effect of proton-pump inhibitors on xerostomia). 36 37 38 39 40

41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 21 INTRODUCTION BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 Gastro-oesophageal reflux disease (GORD) is a common disorder, affecting approximately 10%–20% of the general 4 population.[1] The Montreal consensus,[2] classifies GORD as an entity manifesting as oesophageal and extra- 5 6 oesophageal syndromes. The oesophageal symptomatology includes regurgitation or burning retrosternal chest pain, 7 8 reflux oesophagitis, strictures, Barrett’s oesophagus, and adenocarcinoma; while extra-oesophageal symptomatology 9 comprises reflux, cough, laryngitis, asthma, and dental erosions (DE). DE is a multifactorial phenomenon whereby 10 11 the protective buffering capacity of the oral cavity is compromised by reduced secretion of saliva or high volumes of 12 13 harmful gastric reflux.[3] DE appears to be the most common injury caused by GORD.[4] The global prevalence of 14 15 DE in the general population is not accurately known; it varies depending on the underlying cause, a prevalence of 2%– 16 77% was reported.[5] Tooth Forinvolvement peer seems to bereview universal, but theonly most commonly observed damage occurs on 17 18 the palatal surfaces of the posterior teeth, with a reported prevalence of up to 42%.[6] Studies have highlighted the 19 20 detrimental effects of gastro-duodenal contents in oral soft-tissue pathology as well as on DE.[3] These include effects 21 22 on the oesophageal epithelium, soft palate, and oral mucosa, and manifest as burning mouth syndrome, aphthoid- 23 lesions, hoarseness of voice, erythema of the soft palate and uvula, glossitis, epithelial atrophy, and xerostomia.[7, 8, 24 25 9] Epithelial atrophy and xerostomia further aggravate the GORD-induced injury to the epithelium in the oral cavity 26 27 and oesophagus.[10] 28 29 The severity of DE is directly proportional to the amount of time gastric acid, is in contact with the enamel, so the 30 31 frequency and duration of the reflux problem can be assessed by the amount of enamel loss and vice versa.[11, 12] We 32 hypothesized a strong association between DE and chronic severe GORD and that an absence of DE would indicate a 33 http://bmjopen.bmj.com/ 34 less-severe form of GORD of shorter duration. The aims of this study were as follows: to assess the frequency of GORD 35 36 with and without DE, and oral symptoms in patients undergoing oesophago–gastro–duodenoscopy (OGD); to determine 37 38 the association between severity of GORD and the likelihood of oro-dental manifestations, and to determine the effect 39 40 of GORD on the oral-health-related quality of life (OHRQoL).

41 on October 1, 2021 by guest. Protected copyright. 42 43 MATERIALS AND METHODS 44 45 46 47 This cross-sectional study was conducted at four major tertiary care government hospitals in two of the major 48 metropolitan cities of Pakistan, namely the twin city of Rawalpindi-Islamabad (Pakistan Institute of Medical Sciences, 49 50 PIMS, and Holy Family Hospital) and Karachi (Jinnah Hospital, and Civil Hospital). In total, 187 of 700 patients who 51 52 underwent OGD along with a comprehensive gastrointestinal (GI) examination over a four-month period were 53 54 diagnosed to have GORD,[13] confirmed medically by treatment with a proton pump inhibitor, were included in the 55 study (100% participation rate). A mandatory dental examination was included to evaluate the oro-dental effects of 56 57 GORD. Patients who were unconscious and those with minimal mouth opening (precluding an oral exam) were 58 59 excluded. 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 21 BMJ Open Assessment of Demographic and Clinical Characteristics BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 A questionnaire was used to collect data on socio-demographics, past and current medical history (comorbidities), drug 2 history (including use of non-steroidal anti-inflammatory drugs [NSAIDs]), and risk factors for GORD (dietary pattern, 3 4 addiction profile, consumption of tea, eating habits, weight changes, and other GI disorders). 5 6 7 Assessment of Health-related quality-of-life (HRQoL) 8 9 The Oral Health Impact Profile (OHIP-14) is an instrument designed to assess oral-HRQoL (OHRQoL) and includes 10 11 seven subscales. The Pakistani (Urdu) validated the version of this instrument,[13] was used to evaluate factors related 12 13 to GORD and their effect on the oral cavity regarding speech production, mastication, psychosocial well-being, and 14 life in general, and to estimate the impact of GORD on oral and systemic health. 15 16 For peer review only 17 18 Gastrointestinal (GI) examination 19 All study patients underwent a comprehensive GI examination with baseline laboratory investigations and were on 20 21 appropriate medical therapy.[12, 13] In this study, GORD was diagnosed on the basis of the clinical picture and 22 23 endoscopic (OGD) findings by a gastroenterologist. It manifests in OGD findings as Oesophageal erosions, erythema 24 25 and strictures, lax esophageal sphincters, Oesophagitis, Barrett's oesophagus, Oesophageal adenocarcinoma, and Hiatal 26 hernia.[14] Moreover, the accurate diagnosis of Oesophagitis was also crucial in this study, as one of the predicting 27 28 risk factors for dental erosions. Compelling evidence from ESR lab biomarkers, patients’ clinical symptoms (painful 29 30 swallowing & dysphagia), and signs of erythema, inflammation, and ulceration on OGD were markers for definite 31 32 diagnosis of Oesophagitis.[15]

33 http://bmjopen.bmj.com/ 34 35 Dental examination 36 37 The World Health Organization (WHO) criteria for oral and dental examination,[16] was adapted to examine patients 38 for decayed teeth by using the decayed, missed filled teeth (DMFT) index value, mouth opening status, and presence 39 40 of oral submucosal fibrosis (OSF) by using the OSF-staging index. DMFT was recorded as a simple count of decayed

41 on October 1, 2021 by guest. Protected copyright. 42 and loss of teeth. Abnormal conditions and lesions of the oral mucosa (xerostomia, ulceration, candidiasis, gingivitis, 43 44 angular cheilitis, atrophic glossitis, and leukoplakia) were recorded as per WHO guidelines of the screening 45 protocol.[13, 16, 17] Dental Erosion (DE) was diagnosed on the basis of patient’s complaint of tooth hypersensitivity, 46 47 together with clinical sign of tooth discoloration, dentin exposure, evident loss of tooth occlusal surfaces and decrease 48 49 in tooth height (measured from cementoenamel-junction to crest of enamel).[16 , 18, 19] Oral examination was 50 51 conducted by a single calibrated dentist (IW). 52 53 54 Statistical analysis 55 56 Data analysis was performed using SPSS Version 21 (IBM Corp., Armonk, NY, USA). Categorical variables are 57 reported as frequencies and percentages. Associations between GI symptoms and oral manifestations of GORD were 58 59 determined using the chi-squared test and logistic regression analysis. The Spearman rank correlation and Mann- 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Whitney were used to test for associations and difference of the mean values between GORD groups and OHIP-14 BMJ Open Page 6 of 21 subscales, respectively. A binary outcome variable was created (GORD without DE, 0; and GORD with DE, 1) against BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 which tables (2-5) are computed in this manuscript. Bivariate logistic regression (table 5) was used to obtain crude odds 2 ratios (ORs), for assessing the ‘likelihood of GORD and DE’ with ‘independent predictor variables’ (extracted from 3 4 table 1, 2 and 4). Variables with p≤0.2 were entered into a multivariable logistic regression model for analysis. The 5 6 results are reported with the adjusted odds ratio (AOR), 95% confidence interval (CI) and p-value. In addition, 7 cumulative logistic regression model (extracting significant variables from table 3 & 5) statistics are reported in text 8 9 only, as AOR, 95% CI and p-value; while it has been tested for external validation based on its ability to discriminate 10 11 between GORD with & without DE, using area-under-ROC (AUC) statistics. AUC value range from 0.5 (no 12 13 discrimination) to 1 (perfect discrimination).[20] 14 15 16 Patient & Public involvementFor peer review only 17 18 No patients were involved in the development of this study. However, pilot patients played a crucial role in testing the 19 responsiveness of the questionnaire and identifying gaps in the study. Study variables like xerostomia, oral submucosal 20 21 fibrosis and DMFT were added to the clinical dental section after observing a trend of repetitive observations within a 22 23 specific population. Moreover, translated OHIP-14 questionnaire was tested, and feedback was obtained on pilot 24 25 patients, this process helped in making various amendments concerning language and comprehension of the OHIP- 26 scale items.[13] Within GI-section, it was observed from the patient’s history, being prone to self-medication due to 27 28 the history of chronic back pain. Therefore, these variables were added to the study. All the patients in this study were 29 30 assigned a unique identifier code, and their results cannot be traced back to the original patients to maintain patient 31 32 privacy and confidentiality. Results of this study will be published in the medical and dental journal, and the study

33 centers will be circulated with published manuscripts. Furthermore, patients did not have an active role in the initiation http://bmjopen.bmj.com/ 34 35 of this study or writing the manuscript. 36 37 38 Ethical approval 39 40 The Dow ethical review committee approved the study protocol (IRB department of Dow University of Health

41 on October 1, 2021 by guest. Protected copyright. 42 Sciences). While departmental permissions were sought from participating all four hospitals (CHK, JPMC, PIMS & 43 44 Holy Family), before participant enrolment. Written informed consent was obtained from all study participants after 45 explaining the study protocol. 46 47 48 49 50 RESULTS 51 52 53 54 Most study participants were males (58.3%) and in the adult age group (79.7%), and within the age range of 41-60 55 years as shown in Table 1. The study sample was ethnically and racially diverse; most participants were Punjabi’s 56 57 (54.5%), followed by Urdu-speaking Mohajir’s (16.6%), Pathan’s (15%), and Sindhi’s (8%). Of them, most were urban 58 59 residents (63.6%) of intermediate (35.3%) to low (62%) socioeconomic status. The addiction profile predominantly 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml comprised of oral tobacco (30.5%) & cigarette smoking (26.7%) while minimal alcohol consumption was also reported Page 7 of 21 BMJ Open (2.7%). The primarily consumed beverage is chai-tea (72.7%). A sizeable proportion experienced generalized body BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 pain (44%), for which they reported seeking self-medication, most often NSAIDs (46%). The most common 2 comorbidities were diabetes (34.8%), hypertension (23%), hepatitis C (34.2%), and CLD (18.7%). 3 4 5 6 Slightly more than half (56.7%) of the study population had GORD alone, while the remaining had additional upper 7 GI conditions, including gastritis (19.8%), portal gastropathy (11.2%), peptic ulcer disease (10.2%), GORD-related 8 9 oesophagitis (4.8%), and hiatal hernia (4.3%). 10 11 12 13 Table 1. Demographic variables and baseline characteristics in patients with GORD

14 a 15 S.No. DEMOGRAPHIC VARIABLES N (%) 16 1 Gender For peerMale review only109 (58.3%) 17 Female 78 (41.7%) 18 2 Age Category <19years 09 (4.8%) 19 20-30 years 34 (18.2%) 20 31-40 years 26 (13.9%) 21 41-50 years 48 (25.7%) 22 51-60 years 42 (22.5%) 23 61-70 years 19 (10.2%) 24 71-80 years 09 (4.8%) 25 3 Race/Ethnicity Sindhi 15 (8.0%) 26 Balochi 05 (2.7%) 27 Punjabi 102 (54.5%) 28 Pathan 28 (15%) 29 30 Kashmiri 06 (3.2%) 31 Urdu Speaking Mohajirs 31 (16.6%) 32 4 Domicile Urban 119 (63.6%)

33 Rural 68 (36.4%) http://bmjopen.bmj.com/ 34 5 Socio-economic status (SES) b Low 116 (62%) 35 Moderate 66 (35.3%) 36 High 05 (2.7%) 37 6 Dietary Pattern c Healthy diet 23 (12.3%) 38 Satisfactory diet 80 (42.8%) 39 40 Unsatisfactory diet 84 (44.9%)

41 7 Addiction Smoking 50 (26.7%) on October 1, 2021 by guest. Protected copyright. 42 Oral Tobacco 57 (30.5%) 43 Alcohol 05 (2.7%) 44 8 Tea consumption Yes 136 (72.7%) 45 No 51 (27.3%) 46 47 9 NSAIDS consumption Yes 86 (46.0%) 48 No 101 (54.0%) 49 10 History of Body Ache/General Yes 83 (44.0%) 50 Body Pain No 104 (56.0%) 51 11 Weight Loss (from <5Kg to No 45 (24.1%) 52 >20Kg) 1-5 Kg 72 (38.5%) 53 5-10 Kg 13 (7.0%) 54 10-15 Kg 36 (19.3%) 55 15-20 Kg 15 (8.0%) 56 >20Kg 06 (3.2%) 57 12 Co-morbidities Hypertension 65 (34.8%) 58 Diabetes Mellitus 43 (23%) 59 Hepatitis B 08 (4.3%) 60 For peer review onlyHepatitis - http://bmjopen.bmj.com/site/about/guidelines.xhtml C 64 (34.2%) Chronic Liver Disease (CLD) 35 (18.7%) BMJ Open Page 8 of 21 13 Dental Erosions Group A: GORD with DE 66 (35.3%)

Group B: GORD without DE 121 (64.7%) BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 14. GORD (alone) and (in GORD Alone 106 (56.7%) 2 Combination with other GI GORD with additional Upper-GI 81 (43.3%) 3 illness) conditions 4 i. Gastritis 16 (19.8%) 5 ii. Portal Gastropathy 09 (11.2%) 6 iii. Peptic Ulcer Disease (PUD) 08 (10.2%) 7 iv. GERD Esophagitis 04 (4.8%) 8 v. Hiatal Hernia 04 (4.3%) 9 vi. Others 40 (49.7%) 10 GORD, gastro-oesophageal reflux disease. GI, Gastro-intestinal. DE, Dental Erosions. aItems in bold highlight notable 11 prevalence in each category. bSES (categorized on the basis of monthly family income, number of family dependents and 12 educational status). cDietary pattern (Healthy diet refers to a balanced diet; Satisfactory diet: being intermediary between 13 healthy and unhealthy diet, explicitly relating to occasionally eating junk food, while diet relatively lacks in natural abrasives like fruits/vegetables, cereals, leafy vegetables, and fibers; Unhealthy diet refers to consuming high fatty/oily 14 meals, frequent consumption of junk food/fast food, and soft drinks, while completely devoid of natural abrasives).[13] 15 16 For peer review only 17 DE is a known comorbidity of chronic GORD; therefore, we divided the patients based on presence of DE into two 18 groups (binary variable), namely group A (with DE, chronic/severe GORD; na = 66, 35.3%) and group B (without DE, 19 20 mild GORD; nb = 121, 64.7%). The oral clinical and subjective findings were compared between the two groups. Table 21 22 2 shows the GI signs and symptoms that were found in both groups A and B, the most common of which was 23 heartburn/regurgitation (95%), followed by nausea/vomiting (82%) and dysphagia (42%). A statistically significant 24 25 difference in the frequency of nausea/vomiting (OR 0.349, 95% CI 0.160-0.759; p=0.006), early satiety/loss of appetite 26 27 (OR 1.903, 95% CI 1.020-3.551; p=0.042) and melena (OR 1.957, 95% CI 1.065-3.593; p=0.029) was found between 28 29 patients who had GORD with and without DE. 30 31 32 Table 2. Gastrointestinal signs and symptoms in patients with GORD

33 http://bmjopen.bmj.com/ 34 GI SYMPTOMS Group A Group B ALL OR* Confidence P- 35 (n =66) (n =121) (N=187) Interval value 36 a b 37 38 Heartburn/Regurgitation Yes 63 115 178 (95.2%) 1.096 (0.265, 4.531) 0.900 No 03 06 09 (4.8%) 39 Nausea/Vomiting Yes 48 107 155 (82.9%) 0.349 (0.160, 0.759) 0.006** 40 No 18 14 32 (17.1%)

41 on October 1, 2021 by guest. Protected copyright. Abdominal Pain/discomfort Yes 49 92 141 (75.4%) 0.909 (0.455, 1.815) 0.786 42 No 17 29 46 (24.5%) 43 Abdominal distension Yes 40 62 102 (54.5%) 1.464 (0.796, 2.692) 0.219 44 No 26 59 85 (45.4%) 45 Early Satiety/ Loss of Appetite Yes 44 62 106 (56.7%) 1.903 (1.020, 3.551) 0.042** 46 No 22 59 81 (43.3%) 47 Dysphagia Yes 32 48 80 (42.8%) 1.431 (0.782, 2.620) 0.244 48 No 34 73 107 (57.2%) 49 Hematemesis Yes 27 44 71 (38.0%) 1.212 (0.655, 2.240) 0.540 50 No 39 77 116 (62.0%) 51 Melena Yes 36 46 82 (43.9%) 1.957 (1.065, 3.593) 0.029** 52 No 30 75 105 (56.1%) 53 Group A, chronic/severe GORD with DE. Group B, mild GORD without DE. *OR, Odds ratios (of GI symptoms with GERD+DE as reference category). 54 **p<0.05 (chi-squared test of proportion) was statistically significant. 55 56 The oral manifestations associated with GORD in groups A and B are shown in Table 3. Overall, patients with GORD 57 58 and DE had significantly more oral manifestations than did those without DE. Also, there was a trend of worse dentition 59 status (poor DMFT>10), as indicated by a higher mean DMFT index value in group A than in group B, suggesting that 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml teeth with DE were more vulnerable to decay and tooth loss. Page 9 of 21 BMJ Open The coexistence of GERD & DE is most likely with (in decreasing order of frequency) aphthoid ulcerative lesions (OR BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 13.7, p<0.01), followed by gingivitis (OR 9.0, p<0.01), angular cheilitis (OR 5.2, p<0.01), xerostomia (OR 4.7, p<0.01), 2 atrophic glossitis (OR 2.9, p=0.001), oral submucosal fibrosis (OR 2.9, P=0.003), and oral candidiasis (OR 2.2, 3 4 p=0.028). Oral manifestations with set p-value<0.05 were included in the multivariable logistic regression model, based 5 6 on it, leukoplakia and DMFT were excluded. This multivariable logistic regression model was created by backward- 7 selection (likelihood ratio) method. Since candidiasis has a biological plausibility with GERD & DE, therefore it was 8 9 retained in the model. Interestingly, the multivariable logistic regression-model represented a good AUROC value of 10 11 0.855, and revealed the highest correlation for GORD & DE, was obtained with ulceration (AOR 6.705, CI 2.481- 12 13 18.122, p<0.01), and gingivitis (AOR 4.768 CI 2.123-10.706, p<0.01); followed by angular cheilitis (AOR 2.526, CI 14 1.038-6.148, p=0.041), and xerostomia (AOR 2.372, CI 1.047-5.373, p=0.038). However, candidiasis did not retain its 15 16 significance in the multivariableFor analysis. peer review only 17 18 19 Table 3. Oral manifestations in patients with GORD 20 21 ORAL MANIFESTATIONS Group Group ALL C. OR’s P- A. OR’s P- 22 A B (n=187) value value 23 (n=66) (n=121) * ** 24 25 1. Xerostomia Yes 47 42 89 (47.6%) 4.653 (2.426 , 8.923) <0.01 2.372 (1.047, 5.373) 0.038 26 * ** 27 No 19 79 98 (52.4%) 1 1 28 2. Ulceration Yes 60 51 111 (59.4%) 13.725 (5.506, 34.218) <0.01 6.705 (2.481, 18.122) <0.01 29 * ** 30 No 6 70 76 (40.6%) 1 1 3. Gingivitis Yes 47 26 73 (39.0%) 9.038 (4.547, 17.968) <0.01 4.768 (2.123, 10.706) <0.01 31 * ** 32 No 19 95 114 (61.0%) 1 1

33 http://bmjopen.bmj.com/ 4. Candidiasis Yes 20 20 40 (21.4%) 2.196 (1.078, 4.471) 0.028 0.706 (0.267, 1.868) 0.483 34 * 35 No 46 101 147 (78.6%) 1 1 36 5. Angular Yes 24 12 36 (19.3%) 5.190 (2.382, 11.312) <0.01 2.526 (1.038, 6.148) 0.041 37 Cheilitis * ** 38 No 42 109 151 (80.7%) 1 1 39 6. Atrophic Yes 30 27 57 (30.5%) 2.901 (1.520, 5.538) 0.001 ** - 40 Glossitis *

41 No 36 94 130 (69.5%) 1 on October 1, 2021 by guest. Protected copyright. 42 7. Leukoplakia Yes 2 6 08 (4.3%) 0.599 (0.117, 3.055) 0.533 - - 43 No 64 115 179 (95.7%) 1 44 8. Oral Yes 53 71 124 (66.3%) 2.871 (1.417, 5.818) 0.003 ** - 45 Submucosal * 46 Fibrosis (OSF) No 13 50 63 (33.7%) 1 47 9. DMFT status <3 (acceptable) 24 64 88 (47.06%) - 0.069 - - 48 4-10 (average) 24 37 61 (32.62%) 49 >10 (poor) 18 20 38 (20.32%) 50 Items in bold represent major prevalence (of note) in their respective field. *p<0.05 (chi-squared test) was considered statistically significant. Group A, 51 chronic/severe GORD with DE. Group B, mild GORD without DE. DMFT (decayed, missed, filled teeth), pertaining to tooth decay and tooth loss. C.OR’s, Crude Odds Ratio (chi-square/linear regression). A.OR’s, Adjusted Odds Ratio (values from multivariate logistic regression analysis). **Odds ratios were 52 computed for oral manifestations against a reference variable, ‘GORD with DE;’ using backward selection (likelihood ratio) method, variables were entered 53 into the multivariable analysis. The adjustment was made for atrophic glossitis, oral submucosal fibrosis, leukoplakia, and DMFT status. Area under ROC 54 (AUROC) value: 0.855, was used as an assessment of multivariable logistic regression model’s discriminative ability (between the presence & absence of 55 DE). 56 57 58 59 Table 4 shows the seven-subscales (OHIP-14) scores for groups A and B. The mean values for all subscales were For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 60 relatively higher in group A, suggesting a higher risk of ‘poor oral-HRQoL,’ than that in group B. Spearman ranks BMJ Open Page 10 of 21 correlation (rs) and Mann-Witney U statistical tests, are used for correlations, and to compare means, in both groups A BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 & B respectively. Mann-Whitney U test reveals that a statistically significant difference has been observed for all OHIP 2 subscales (except social disability) when compared between the two groups (A & B). The coexistence of GORD and 3 4 DE had a positive correlation with OHIP subscales (p<0.05), with notable impact on psychological discomfort 5 6 (rs=0.297), physical disability (rs=0.288), psychological disability (rs=0.271), and functional limitation (rs=0.200). 7 8 Table 4. Psychometric properties of OHIP-14, correlated with GORD, with and without dental erosions. 9 10 11 OHIP-14 Sub-scales Mean subscale (S.D) Statistical Tests Characteristics 12 With DE (group A) Without DE (group B) Spearman Ranks: rs Mann-Whitney U test 13 na=66 nb=121 correlation value (p-value)* 14 Functional Limitation 3.530 (2.199) 2.5702 (1.92712) 0.200 * 3065.0 (0.006) * 15 Physical Pain 1.182 (1.264) 0.7273 (0.96609) 0.167 * 3260.0 (0.023) * Psychological Discomfort 5.167 (1.853) 4.0744 (1.98396) 0.297 * 2625.0 (<0.001) * 16 For peer review only 17 Physical Disability 6.394 (2.745) 4.4463 (2.81351) 0.288 * 2672.0 (<0.001) * Psychological Disability 4.591 (1.745) 3.5868 (1.94795) 0.271 * 2754.5 (<0.001) * 18 Social Disability 7.106 (2.835) 6.1405 (3.42127) 0.142 3332.5 (0.053) 19 Handicap 1.682 (1.230) 1.2149 (1.31153) 0.193 * 3099.0 (0.008) * 20 *Correlation is significant at p<0.05 level (2-tailed). rs, ranks correlation coefficient. Bold highlight represents important values in each statistical test results. 21 22 23 24 Table 5 represents the bivariate relationship between significant variables (extracted from table one, two, and four), & 25 26 dichotomous/binary outcome variable (GORD with DE, as reference category), were included in the logistic regression 27 model. Univariate analysis revealed patients with an unhealthy dietary pattern were more likely to develop DE with 28 29 GORD, than did those with a healthy dietary pattern (OR 6.034, 95% CI 1.889-19.268). Also, early satiety/loss of 30 31 appetite (OR 1.903, 95% CI 1.020–3.551), and nausea/vomiting (OR 0.349, 95% CI 0.160–0.759) increased the 32 likelihood of developing DE by 1.903-fold, and 0.349-fold respectively. Moreover, melena was observed as a 33 http://bmjopen.bmj.com/ 34 significant manifestation within chronic GORD population (OR 1.957, 95% CI 1.065-3.593). Patients with peptic ulcer 35 36 disease were more likely to develop chronic GORD with DE’s (OR 2.825, 95% CI 1.075–7.423). Likewise, OHIP- 37 38 subscales such as functional limitation, physical pain, physical disability, psychological discomfort, psychological 39 40 disability are sensitive to the severity of GORD and existence of Dental Erosions.

41 on October 1, 2021 by guest. Protected copyright. 42 The results of the multivariable logistic regression model are also presented in Table 5, with adjustments made for all 43 44 independent variables. Some variables that had a statistically significant association with DE in univariate analysis (i.e., 45 46 loss of appetite, melena, peptic ulcer disease, OHIP subscales namely functional limitation, physical pain, 47 psychological discomfort, & psychological disability) were no longer significant in multivariable analysis. Portal 48 49 gastropathy had a p-value of >0.2; therefore, was excluded from multivariable analysis. However, unhealthy dietary 50 51 pattern (AOR 2.307, 95% CI 0.403-13.207; Table 5), nausea/vomiting (AOR 0.130, 95% CI 0.033-0.512), oesophagitis 52 (AOR 12.427, 95% CI 1.658-93.143) and physical disability (advanced stage) (AOR 0.564, 95% CI 0.020-16.163) 53 54 retained its significant association with DE. Therefore, these variables are concluded to have statistically correlated 55 56 with GORD and DE’s, in multivariable logistic regression analysis. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Table 5. Risk factors associated with GORD and dental erosions: results of the univariate and multivariable analysis. Page 11 of 21 BMJ Open

CHARACTERISTICS GORD with Dental Erosions BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 C. OR (95% CI) p-value A. OR (95% CI) p-value 2 1. Dietary Pattern <0.01* <0.001* 3 4 Healthy 1 1 5 Satisfactory 1.096 (0.325, 3.697) 0.234 (0.041, 1.320) 6 Unhealthy 6.034 (1.889, 19.268) 2.307 (0.403, 13.207) 7 2. GI symptom: Nausea/vomiting 0.008* 0.004* 8 No 1 1 Yes 0.349 (0.160, 0.759) 0.130 (0.033, 0.512) 9 3. GI symptom: Early satiety/loss of appetite 0.043* 0.262 10 No 1 1 11 Yes 1.903 (1.020, 3.551) 1.778 (0.650, 4.867) 12 4. GI symptom: Melena 0.030* 0.421 13 No 1 1 14 Yes 1.957 (1.065, 3.593) 1.509 (0.554, 4.111) 15 5. GI disorder: Oesophagitis 0.20 0.014* 16 No For peer1 review only 1 17 Yes 2.398 (0.621, 9.255) 12.427 (1.658, 93.143) 18 6. GI disorder: Peptic ulcer disease (PUD) 0.035* 0.302 19 No 1 1 20 Yes 2.825 (1.075, 7.423) 2.143 (0.505, 9.095) 21 7. GI disorder: Hiatal hernia 0.20 0.201 22 No 1 1 23 Yes 0.251 (0.03, 2.082) 0.159 (0.010, 2.655) 24 8. GI disorder: Portal gastropathy 0.214 - 25 No 1 - 26 Yes 1.786 (0.715, 4.458) - 27 9. GI disorder: Gastritis 0.06 0.065 28 No 1 1 29 Yes 2.013 (0.970, 4.179) 3.388 (0.926, 12.398) 30 10. OHIP Subscale 1: Functional Limitation 0.047* 0.277 31 1= Null 1 1 32 2= Mild 1.379 (0.506, 3.753) 0.511 (0.106, 2.464)

33 3= Moderate 1.495 (0.581, 3.843) 0.537 (0.132, 2.193) http://bmjopen.bmj.com/ 34 4= Severe 3.646 (1.204, 11.044) 0.607 (0.086, 4.313) 35 5= Advanced 5.469 (1.265, 23.640) 6.149 (0.491, 76.934) 36 11. OHIP Subscale_2: Physical Pain 0.023* 0.118 37 1= Null 1 1 38 2= Mild 0.484 (0.168, 1.391) 0.202 (0.034, 1.186) 39 3= Moderate 1.935 (0.929, 4.032) 2.706 (0.633, 11.574) 40 4= Severe 3.116 (0.917, 10.591) 4.271 (0.473, 38.585)

41 on October 1, 2021 by guest. Protected copyright. 5= Advanced 6.677 (0.668, 66.769) 2.391 (0.104, 55.127) 42 12. OHIP Subscale_3: Physical Disability 0.002* 0.024* 43 44 1= Null 1 1 45 2= Mild 0.346 (0.057, 2.095) 0.042 (0.003, 0.674) 46 3= Moderate 3.033 (0.936, 9.822) 0.861 (0.101, 7.331) 47 4= Severe 3.414 (1.011, 11.527) 0.125 (0.010, 1.613) 5= Advanced 15.750 (2.335, 106.227) 0.564 (0.020, 16.163) 48 13. OHIP Subscale_4: Psychological Discomfort 0.002* 0.358 49 50 1= Null 1 1 51 2= Mild 2.857 (0.282, 28.960) 2.105 (0.073, 60.804) 52 3= Moderate 2.264 (0.262, 19.563) 3.080 (0.144, 65.939) 53 4= Severe 6.303 (0.741, 53.650) 3.303 (0.145, 75.217) 54 5= Advanced 11.692 (1.302, 105.028) 10.356 (0.457, 234.689) 55 14. OHIP Subscale_5: Handicap 0.057 0.171 56 1= Null 1 1 57 2= Mild 1.981 (0.795, 4.935) 1.884 (0.415, 8.563) 58 3= Moderate 2.869 (1.303, 6.315) 2.257 (0.569, 8.948) 59 4= Severe 3.900 (1.282, 11.860) 9.210 (1.348, 62.911) 60 5= Advanced For peer1.891 review (0.600, only 5.964)- http://bmjopen.bmj.com/site/about/guidelines.xhtml0.756 (0.126, 4.529) 15. OHIP Subscale_6: Psychological Disability 0.006* 0.207 BMJ Open Page 12 of 21 1= Null 1 1

2= Mild 3.000 (0.520, 17.316) 13.130 (0.913, 188.820) BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 3= Moderate 2.763 (0.582, 13.122) 6.336 (0.638, 62.938) 2 4= Severe 8.077 (1.682, 38.784) 13.513 (1.273, 143.448) 3 5= Advanced 8.437 (1.457, 48.851) 9.434 (0.899, 98.955) 4 All study variables expressing p-value<0.2 in univariate analysis, were included in multivariable analysis, and are presented here in this table. *p- 5 value<0.05 was considered statistically significant (bold highlight). 95% CI, confidence interval; C. OR, crude/unadjusted odds ratios (obtained from 6 univariate analysis); A. OR, odds ratios adjusted for all independent variables (obtained from multi-variable analysis, by step-wise selection method). 7 8 9 Conclusively, using backward selection-LR (likelihood ratio) test method, all (seventeen) significant (from univariate 10 11 analysis) variables (of table 3 & 5), were entered into the multivariable logistic regression model. Amongst which, 12 13 unhealthy dietary pattern (AOR 0.474, 95% CI 0.091-2.452; p=0.003), nausea/vomiting (AOR 0.256, 95% CI 0.067- 14 15 0.980; p=0.047), xerostomia (AOR 3.005, 95% CI 1.026-8.805; p=0.045), ulceration (AOR 6.609, 95% CI 2.007- 16 21.765; p=0.002), gingivitis (AORFor 7.516, peer 95% CI 2.517-22.443;review p<0.001), only and angular cheilitis (AOR 4.028, 95% CI 17 18 1.237-13.115; p=0.021) were significant in relation to GORD & DE. To assess discriminative ability (between GORD 19 20 with and without DE), using external validation of the model; area under the receiver operating curve (AU-ROC) 21 statistics were employed. It showed excellent discrimination (AUROC=0.922) of the multivariable regression model. 22 23 Henceforth, proving these variables as major risk factors for ‘GORD with Dental Erosions.’ 24 25 26 27 Table 6: Risk factors for GORD with DE: results of the cumulative multi-variable logistic regression model. 28 29 30 RISK FACTORS (for GORD with C. OR’s P- A. OR’s P- 31 DE’s) value* value** 32 1. Unhealthy Dietary pattern Yes 6.034 (1.889, 19.268) <0.01 0.474 (0.091, 2.452) 0.003 No 1 1 33 http://bmjopen.bmj.com/ 34 2. Nausea/ Vomiting Yes 0.349 (0.160, 0.759) 0.008 0.256 (0.067, 0.980) 0.047 35 No 1 1 36 3. Xerostomia Yes 4.653 (2.426, 8.923) <0.01 3.005 (1.026, 8.805) 0.045 No 1 1 37 4. Ulceration Yes 13.725 (5.506, 34.218) <0.01 6.609 (2.007, 21.765) 0.002 38 No 1 1 39 5. Gingivitis Yes 9.038 (4.547, 17.968) <0.01 7.516 (2.517, 22.443) <0.01 40 No 1 1 41 6. Angular Cheilitis Yes 5.190 (2.382, 11.312) <0.01 4.028 (1.237, 13.115) 0.021 on October 1, 2021 by guest. Protected copyright. 42 No 1 1 43 C.OR represent a univariate analysis, while A.OR represent multi-variable analysis (using backward-elimination method). 44 *All significant variables with p-value<0.05, obtained through univariate analysis (from table 3 and 5), against dichotomous outcome variable, 45 ‘GORD with DE.’ **Using backward selection (likelihood ratio), all seventeen variables were entered in multivariable logistic regression model; 46 only six of seventeen significant variables (from table 3 & 5), retained significance within the final regression model, which are shown here in 47 this table. This model was tested for discrimination (external validity), with an AUROC value of 0.922. 48 49 50 51 52 DISCUSSION 53 54 55 56 GORD with DE and its associated risk factors has been the subject of much research interest in recent years. The present 57 study focused on patients who had GORD with or without DE and recorded the prevalence, distribution, and severity 58 59 of this clinical entity in a relatively large sample. The influences of known risk indicators and other possible 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml determinants that have been less well studied in the past were also investigated. Page 13 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 GORD has multiple systemic effects, particularly in the GI tract (Table 2). There is considerable medical literature on 2 GORD and its impact on the oral hard tissue (i.e., DE), as well as its periodontal effects;[7] however, data on the 3 4 relationship between acidic oral mucosal lesions and oral health are scarce or controversial.[7-13, 19, 21, 22] Our 5 6 finding of a prevalence of DE in patients with GORD of 35.3%, fits well within the regionally (South Asian population) 7 reported prevalence of DE (ranges between 5.0 to 58.4 percent);[22-25] while global prevalence ranges from 2% to 8 9 77%, this wide range varies with age, gender, ethnicity and underlying risk-factors of different populations.[5] The 10 11 influence of major risk factors (such as obesity, metabolic syndrome, strong addiction pattern, and alcohol 12 13 consumption) for development of GORD & DE, in this specific population, was minimal or negligible; which supports 14 the fact that, the reported prevalence of DE in this study was relatively low (unlike the western population).[14, 26, 27] 15 16 Moreover, this study’s populationFor is predominantly peer reviewhospital-based patients, only seeking tertiary care for chronic ailments. 17 18 As evident around 43% of subjects have other GI pathology in addition to GORD, while chronic illness like Hepatitis 19 C (34%) and chronic liver disease (19%) contribute as major comorbidities; Furthermore, around 75% of this population 20 21 had experienced weight loss of varying degrees, which eliminates obesity (or weight gain) as a risk factor. 22 23 24 25 Our results indicate that xerostomia, ulceration, gingivitis, and angular cheilitis have the highest correlation with GORD 26 and DE. Some of our findings are in accordance with few previous studies,[4, 8, 22, 28] since there have been lacking 27 28 studies assessing the prevalence of xerostomia and oral lesions in patients with GORD. Therefore literature reveals 29 30 mixed results. Saliva is considered to play a leading role in protecting the oesophageal mucosa against gastric reflux, 31 32 and its qualitative (e.g., deficiencies of salivary growth factor and cytokines) and quantitative (e.g., hyposalivation)

33 abnormalities have been linked to the pathogenesis of GORD,[29-32] compromised dental health,[33] and DE.[34] http://bmjopen.bmj.com/ 34 35 The pathogenesis of GORD appears to be connected with the decreased production of saliva, and our findings suggest 36 37 that xerostomia should be included in the extra-oesophageal symptomatology of GORD.[35] 38 39 40 Approximately 60% of participants had ulcerative lesions on the soft and hard palate mucosa, buccal mucosa, uvula,

41 on October 1, 2021 by guest. Protected copyright. 42 and tongue, which could not be characterized clinically as any other disease; these lesions are typically recognized as 43 44 ‘soft tissue aphthoid-lesions’ related to GORD,[4, 8, 22, 28] and are caused directly by the corrosive effects of refluxed 45 acid. 46 47 48 49 A recent study in a rat model of the soft palate by Habesoglu et al.,[6] identified epithelial and microscopic alterations 50 51 in response to acid injury. Interest in this phenomenon dates to the late 1980s, when a study by Järvinen et al., 52 demonstrated a marked prevalence of oral mucosal changes in the presence of GORD, including burning mouth, 53 54 aphthoid lesions, erythema of the soft palate and uvula, glossitis, and epithelial atrophy.[8] However, these findings 55 56 were called into question by the studies of Ranjitkar et al.,[10] and Petruzzi et al.,[21] who found that mucosal changes 57 are quite common and not an entity-specific for GORD. Deppe et al.,[7] also studied the effects of GORD on the oral 58 59 mucosa and found positive and negative signs for erythema and ulceration, respectively, but could not find a statistically 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml significant difference between their erosive and non-erosive reflux groups. Similarly, a study by Meurman et al.,[5] BMJ Open Page 14 of 21 found no mucosal changes that could be linked to GORD. Following on from the findings of Rajalalitha et al.,[36] we BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 hypothesized that the chronic-acid injury caused by GORD would be a source of persistent irritation and inflammation 2 affecting the oral mucosa, ultimately resulting in the infiltration of inflammatory markers. These markers include 3 4 increased populations of interleukins, cytokines, tumor necrosis factor-alpha, interferon-alpha, and growth factors such 5 6 as transforming growth factor-beta, which are produced at the site of inflammation. This mechanism of inflammation 7 is exaggerated in situations of chronic and constant irritation such as GORD and supra-oesophageal reflux disease.[23, 8 9 36-39] This finding is accompanied by epithelial atrophy and fibro-elastic changes in the lamina propria.[36, 40] This 10 11 mechanism is consistent with the finding of significant oral submucosal fibrosis in our sample, which can lead to 12 13 metaplastic changes in the mucosa and can also cause restricted mouth opening, resulting in trismus, eating difficulties, 14 impaired speech, and generalized impairment of oral health.[36-38] Our study findings demonstrate a positive 15 16 association of chronic GORDFor with OSF, peer which is a precancerousreview lesion.[40] only OSF has never been reported in association 17 18 with GORD, partly because these oral pathologies were not studied in-depth amongst GORD population;[12, 17-19, 19 21-23, 28-32, 34-39, 41-43] however, because of its significant impact on quality of life, it should be considered and 20 21 addressed in future researches. 22 23 24 25 With chronic GORD-mediated acid exposure, the salivary gland epithelium sustains severe damage (i.e., epithelial 26 metaplasia), resulting in xerostomia.[17, 23] This association may further aggravate acid-mediated oro-oesophageal 27 28 epithelium injury, resulting in GORD-related oesophagitis.[9, 29, 44] Other conditions, including oral ulceration, 29 30 associated infections (candidiasis and angular cheilitis), and gustatory dysfunction, are also common.[44] 31 32 Correa et al. noted that patients with chronic GORD had reduced salivary buffering capacity and concluded that this

33 http://bmjopen.bmj.com/ 34 was the predominant cause of tooth erosion.[41] The same study also reported a lower prevalence of dental caries, 35 36 which was attributed to the low prevalence of cariogenic bacteria (Lactobacilli and Streptococci) in the saliva of patients 37 38 with chronic GORD; this may be a possible association with GORD-induced DE that needs confirmation in future 39 studies. However, our findings with regard to dental caries are contrary to Correa et al.’s, in sense that the mean DMFT 40

41 index values were higher in the group with GORD and DE (DMFT > 4 in 63.6%) than in the group with GORD and on October 1, 2021 by guest. Protected copyright. 42 43 no DE (DMFT > 4 in 47.1%). We postulate that the additive effects of direct acid injury, low salivary buffering capacity, 44 and increased opportunistic bacterial populations cause this marked increase in tooth decay and loss (as indicated by 45 46 increased DMFT-index values in our study) in the high-risk population with GORD and DE.[18, 33] However, it can 47 48 also be attributed to poor oral hygiene status, primarily because 27% of our study population were smokers, and 30% 49 50 chewed tobacco.[45] An acidic/unhealthy dietary pattern was common in GORD population,[19] as was excessive 51 consumption of traditional beverages [i.e. sugary chai: tea with milk (73%)], which would also have contributed to a 52 53 low oral pH and provided a favourable environment for opportunistic bacteria causing tooth decay and tooth loss.[46, 54 55 47] 56 57 There is a wealth of literature identifying DE as being comorbid with GORD, and a review by Ranjitkar et al., reported 58 59 the median prevalence for DE in patients with GORD to be 24%.[18] Accordingly, the Montreal consensus postulated 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml that the prevalence of DE is directly related to GORD, particularly when noted on the lingual and palatal tooth Page 15 of 21 BMJ Open surfaces.[2] Our study findings are consistent with the concept that DE has a significant association with GORD and BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 may serve as a marker of disease severity. As in a study by Meurman et al.,[5] our statistical comparison of patients 2 who had GORD with and without DE demonstrated that unlike subjective oral symptoms, oral manifestations were 3 4 significantly more common in the group with GORD and DE. 5 6 7 Considering the findings of Meurman et al.,[5] we used the validated OHIP-14 instrument,[13] instead of subjective 8 9 oral symptoms to assess the impact of severity of GORD (with DE) on oral health from a psychological and general 10 11 well-being perspective. Until now, this tool has not been tested in GORD patient population. Using the OHIP-14, we 12 13 found that GORD with DE was significantly correlated (p<0.05) with the OHIP-14 subscale psychometric 14 characteristics namely: psychological discomfort, physical disability, psychological disability, functional limitation, 15 16 handicap, and physical pain. ForThis result peer is consistent review with our finding thatonly GI conditions and their oral manifestations 17 18 have a considerable adverse impact on oral health. Notably, in our study, 43.3% of the study population had GORD 19 alone, and the rest had additional upper GI conditions, including gastritis, portal gastropathy, peptic ulcer disease, 20 21 GORD-related oesophagitis, and hiatal hernia, which are believed to be factors that initiate and/or lead to progression 22 23 of GORD.[48-50] Further, a good proportion of patients self-medicated with NSAIDs for pain relief, which would have 24 25 further aggravated their upper-GI illness.[48] 26 27 This study has some limitations. First, the study design being cross-sectional, limits the ability of the study results to 28 29 correlation extent only. Since this study is the first of its kind to report risk factors extensively, future prospective cohort 30 31 studies are needed to establish causality. Second, the staging of DE was not recorded. Third, only one investigation 32 was used to diagnose GORD, and no measures of disease severity, such as manometry, 24-hour pH monitoring, or

33 http://bmjopen.bmj.com/ 34 biopsies, were included.[41] Fourth, other oral manifestations of GORD that might have a causal relationship with DE, 35 36 such as periodontal disease and gingival disease, were not recorded in detail. Moreover, we did not control for the effect 37 38 of ongoing treatment for GORD patients. Thus it would have slightly under-estimated the associations found (between 39 GORD & oral manifestations), but it would not be a serious threat to the associations found. Also, some oral side- 40

41 effects of prescription could not be ruled out (such as the effect of proton-pump inhibitors on xerostomia). Fifth, we on October 1, 2021 by guest. Protected copyright. 42 43 could not rule out xerostomia being the result of an interplay of various comorbidities, such as diabetes or hepatic 44 disease, or a side effect (albeit minor) of medication, including antihistamines, proton pump inhibitors, calcium channel 45 46 blockers, and beta-blockers.[5, 29, 44] Further trials and clinical studies are needed to rule out these potentially 47 48 confounding factors. Future studies in humans should include biopsies of the oral epithelium to correlate clinical and 49 50 histological findings. Our present findings suggest that the Montreal consensus recommendations should now be 51 expanded to include xerostomia, OSF, aphthoid-ulcerative lesions, and candidiasis in the extra-oesophageal 52 53 symptomatology of GORD.[2, 43] 54 55 56 57 Dental health is widely neglected in Pakistan, where dental-visits are restricted to a downstream approach (interventions 58 at micro-level).[45, 51, 52] Further, general practitioners and gastroenterologists are often the primary health care 59 60 providers for patients withFor peer GORD, review but only while - http://bmjopen.bmj.com/site/about/guidelines.xhtml addressing their main GI concerns, the oral manifestations of these BMJ Open Page 16 of 21 systemic conditions are often overlooked. This study highlights the need for dental referral in patients with upper GI BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 disorders (in this case GORD), which can have a marked effect on both systemic and oral health. 2 3 4 5 CONCLUSION 6 7 8 We found a positive correlation between GORD with DE and oral conditions, such as xerostomia, aphthoid mucosal 9 ulceration, gingivitis, and angular cheilitis in our study population. Patients with DE had more severe oral symptoms 10 11 and more compromised oral health than did those without DE. This study reinforces DE as a strong comorbidity of 12 13 ‘Chronic GORD.’ Further, unhealthy dietary patterns and frequent nausea/vomiting increases the likelihood of 14 15 developing chronic GORD & DE, and thereby the development of oral soft-tissue lesions, with increasing severity. We 16 recommend timely dental visitsFor for evaluation peer of oral review health, to assess only the degree of oral manifestations, and prevent 17 18 its progression. We also urge gastroenterologists to perform routine oral examinations and make referrals to a dentist 19 20 to avoid worsening of the oral conditions caused by GORD. 21 22 Acknowledgments 23 24 25 The authors acknowledge the participating hospitals and respective gastroenterologists for supervising this research 26 27 and providing ethical and institutional/departmental permission to conduct this study, namely: Drs. Tayyab Saeed 28 Akhtar (Holy Family Hospital), Farzana Memon (CHK), Saeed Qureshi (CHK), Mashood (PIMS), Waseem Said 29 30 (PIMS), Waqar (JPMC), Masroor (JPMC). The study would not have been possible without the efforts of the following 31 32 clinicians who were trained and calibrated as the data collection team at four study centers: Drs. Emmad Naeem Khan,

33 Sumbul Zakir, Iqra Raja, Rimsha Hashmi, Javeria Ahmed, Maryam Nazeer, Ovais Khan Muhammad, Qurratul-Ain, http://bmjopen.bmj.com/ 34 35 Hafsa Mahida, Yumna Shaheen Ali, Sajida Jhamro, Saad Aziz, Prem Shankar, Sham Kumar, Sanjay Kumar, and 36 37 Orangzeb Khatri. We would also like to thank Editage for English language editing. Also, thanks for the kind timely 38 39 support from http://www.phsrg.com for upgrading of the manuscript. This research would not have been possible 40 without the valuable support and guidance of Professor Dr. Ambrina Qureshi (Chairperson community dentistry

41 on October 1, 2021 by guest. Protected copyright. 42 department, DUHS). 43 44 45 Declaration of interests 46 47 All authors have completed the ICMJE uniform disclosure form at "http://www.icmje.org/coi_disclosure.pdf" and 48 49 declare no support from any organization for the submitted work; no financial relationships with any organizations that 50 51 might have an interest in the submitted work in the previous three years; no other relationships or activities that could 52 appear to have influenced the submitted work. 53 54 55 Funding 56 57 None 58 59 60 Data sharing statement For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 21 BMJ Open No additional data are available BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 Author contributions 2 3 4 IW, study conception, data collection/analysis, manuscript writing and editing; JA, data collection and data 5 6 entry/analysis; AY, study design/analysis, manuscript editing; AR, statistical analysis and proofreading; TA, 7 supervision of gastroenterology research (clinical supervisor) and proofreading; QA, data analysis, re-editing & peer 8 9 review (internal medicine expert); ZK, statistical interpretation and peer-review (dentistry expert). 10 11 12 13 License to Publisher 14 The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, a 15 16 worldwide license to the PublishersFor and peer its licensees review in perpetuity in all only forms. 17 18 19 20 REFERENCES 21 22 1. Patel A, Amaechi BT, Brady C. Prevention and control of dental erosion: gastroesophageal reflux disease 23 management. Dental Erosion and Its Clinical Management. Springer International Publishing 2015:203-224. 24 2. Vakil N, Van Zanten SV, Kahrilas P, et al. The Montreal definition and classification of gastroesophageal reflux 25 26 disease: a global evidence-based consensus. The American journal of gastroenterology 2006;101(8):1900-20. 27 3. Pace F, Pallotta S, Tonini M, et al. Systematic review: gastro‐oesophageal reflux disease and dental lesions. 28 Alimentary pharmacology & therapeutics 2008;27(12):1179-86. 29 4. Lazarchik DA, Filler SJ. Effects of gastroesophageal reflux on the oral cavity. The American journal of medicine 30 1997;103(5):107-13. 31 5. Meurman JH, Toskala J, Nuutinen P, et al. Oral and dental manifestations in gastroesophageal reflux disease. 32 Oral Surg Oral Med Oral Pathol 1994;78(5):583-89.

33 http://bmjopen.bmj.com/ 34 6. Habesoglu TE, Habesoglu M, Sürmeli M, et al. Histological changes of rat soft-palate with exposure to 35 experimental laryngopharyngeal reflux. Auris Nasus Larynx 2010;37(6):730-36. 36 7. Deppe H, Mücke T, Wagenpfeil S, Kesting M, et al. Erosive esophageal reflux vs. non erosive esophageal 37 reflux: oral findings in 71 patients. BMC oral health 2015;15(1):84. 38 8. Jarvinen V, Meurman JH, Hyvarinen H, et al. Dental erosion and upper gastrointestinal disorders. Oral Surg 39 Oral Med Oral Pathol 1988;65:298–303. 40 9. Friesen LR, Bohaty B, Onikul R, et al. Is histologic esophagitis associated with dental erosion: a cross-sectional

41 on October 1, 2021 by guest. Protected copyright. 42 observational study. BMC oral health 2017;17(1):116. 43 10. Ranjitkar S, Smales RJ, Kaidonis JA. Oral manifestations of gastroesophageal reflux disease. Journal of 44 gastroenterology and hepatology 2012;27(1):21-27. 45 11. Daley TD, Armstrong JE. Oral manifestations of gastrointestinal diseases. Canadian Journal of 46 Gastroenterology and Hepatology 2007;21(4):241-44. 47 12. Picos AM, Poenar S, Opris A, et al. Prevalence of dental erosions in GERD: a pilot study. Clujul Medical 48 49 2013;86(4):344–346. 50 13. Warsi I, Younus A, Rasheed A, et al. Oral health-related quality of life in patients with upper gastrointestinal 51 and hepatic disorders in Pakistan: validation of the Oral Health Impact Profile-14 in the Urdu language. BDJ 52 Open 2018;4:1-7. 53 14. Barron RP, Carmichael RP, Marcon MA, et al. Dental erosion in gastroesophageal reflux disease. Journal 54 Canadian Dental Association 2003;69(2):84-9. 55 15. Tripathi M, Streutker CJ, Marginean EC. Relevance of histology in the diagnosis of reflux esophagitis. Annals 56 57 of the New York Academy of Sciences 2018;1420(1). 58 16. World Health Organization. In Oral Health Surveys: Basic Methods 5th edn. Petersen PE, Baez RJ, (World 59 Health Organization, TX, USA, 2013). 60 17. Silva MA, DamanteFor peerJH, Stippreview AC,only -et http://bmjopen.bmj.com/site/about/guidelines.xhtml al. Gastroesophageal reflux disease: new oral findings. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;91(3):301-10. BMJ Open Page 18 of 21 18. Ranjitkar S, Kaidonis JA, Smales RJ. Gastroesophageal reflux disease and tooth erosion. International journal

of dentistry 2011;2012. BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 19. Milani DC, Venturini AP, Callegari-Jacques SM, et al. Gastro-oesophageal reflux disease and dental erosions 2 in adults: Influence of acidified food intake and impact on quality of life. European journal of gastroenterology 3 & hepatology 2016;28(7):797-801. 4 5 20. Roelen et al. External validation of two prediction models identifying employees at risk of high sickness 6 absence: cohort study with 1-year follow-up. BMC Public Health 2013;13:105. 7 21. Petruzzi M, Lucchese A, Campus G, et al. Oral stigmatic lesions of gastroesophageal reflux disease (GERD). 8 Rev Med Chil 2012;140(7):915-18. 9 22. Vinesh E, Masthan KMK, Kumar MS, et al. A Clinicopathologic Study of Oral Changes in Gastroesophageal 10 Reflux Disease, Gastritis, and Ulcerative Colitis. J Contemp Dent Pract 2016;17(11):943-47. 11 23. Roesch-Ramos L, Roesch-Dietlen F, Remes-Troche JM, et al. Dental erosion, an extraesophageal manifestation 12 13 of gastroesophageal reflux disease. The experience of a center for digestive physiology in Southeastern Mexico. 14 Rev Esp Enferm Dig 2014;106(2):92-7. 15 24. Wang H-Y, Leena KB, Plymoth A, et al. Prevalence of gastroesophageal reflux disease and its risk factors in a 16 community-based populationFor in peer southern India. review BMC Gastroenterology only 2016;16:36. 17 25. Javadzadeh F, Rafeey M. 715 Dental Erosion and Gastroesophageal Reflux Disease (GERD). Children Archives 18 of Disease in Childhood 2012;97:A206. 19 26. Ness-Jensen E, Lagergren J. Baillière's Best Practice & Research: Clinical Gastroenterology 2017;31(5). 20 21 27. Mantonanaki M, Koletsi-Kounari H, Mamai-Homata E, et al. Dental erosion prevalence and associated risk 22 indicators among preschool children in Athens, Greece. Clin Oral Invest 2013;17(2):585-93. 23 28. Meurman JH, Rantonen P. Salivary flow rate, buffering capacity, and yeast counts in 187 consecutive adult 24 patients from Kuopio, Finland. European Journal of Oral Sciences 1994;102(4):229-34. 25 29. Korsten MA, Rosman AS, Fishbein S, et al. Chronic xerostomia increases esophageal acid exposure and is 26 associated with esophageal injury. The American journal of medicine 1991;90(6):701-06. 27 28 30. Kongara K, Varilek G, Soffer EE. Salivary growth factors and cytokines are not deficient in patients with 29 gastroesophageal reflux disease or Barrett's esophagus. Digestive diseases and sciences 2001;46(3):606-09. 30 31. Rourk RM, Namiot Z, Edmunds MC, et al. Diminished luminal release of esophageal epidermal growth factor 31 in patients with reflux esophagitis. American Journal of Gastroenterology 1994;89(8). 32 32. Kao CH, Ho YJ, ChangLai SP, et al. Evidence for decreased salivary function in patients with reflux esophagitis.

33 Digestion 1999;60(3):191-95. http://bmjopen.bmj.com/ 34 33. Ahmad MS, Bhayat A, Zafar MS, et al. The Impact of Hyposalivation on Quality of Life (QoL) and Oral Health 35 36 in the Aging Population of Al Madinah Al Munawarrah. Int. J. Environ. Res. Public Health 2017;14:445. 37 34. Jarvinen VK, Rytomaa II, Heinonen OP. Risk factors in dental erosion. Journal of dental research 38 1991;70(6):942-47. 39 35. Di Fede O, Di Liberto C, Occhipinti G, et al. Oral manifestations in patients with gastro‐oesophageal reflux 40 disease: a single‐center case-control study. Journal of oral pathology & medicine 2008;37(6):336-40.

41 36. Rajalalitha P, Vali S. Molecular pathogenesis of oral submucous fibrosis–a collagen metabolic disorder. Journal on October 1, 2021 by guest. Protected copyright. 42 of oral pathology & medicine 2005;34(6):321-28. 43 44 37. Pindborg JJ, Sirsat SM. Oral submucous fibrosis. Oral Surg Oral Med Oral Pathol 1966;22(6):764-79. 45 38. Hogaboam CM, Steinhauser ML, Chensue SW, et al. Novel roles for chemokines and fibroblasts in interstitial 46 fibrosis. Kidney international 1998;54(6):2152-59. 47 39. Scott DR, Simon RA. Supraesophageal reflux disease: A review of the literature. Allergy and asthma 48 proceedings. OceanSide Publications, Inc 2014;35(2):104-110. 49 40. Khan S, Chatra L, Prashanth SK, et al. Pathogenesis of oral submucous fibrosis. Journal of cancer research 50 51 and therapeutics 2012;8(2):199. 52 41. Corrêa MC, Lerco MM, Cunha MD, et al. Salivary parameters and teeth erosions in patients with 53 gastroesophageal reflux disease. Arquivos de gastroenterologia 2012;49(3):214-18. 54 42. Ismail-Beigi F, Horton PF, Pope CE. Histological consequences of gastroesophageal reflux in man. 55 Gastroenterology 1970;58(2):163-74. 56 43. Pauwels, ANS. Dental erosions and other extra-oesophageal symptoms of gastro-oesophageal reflux disease: 57 Evidence, treatment response, and areas of uncertainty. United European Gastroenterology Journal 58 59 2015;3(2):166-170. 60 44. Sreebny LM, ValdiniFor peer A. Xerostomia:review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml a neglected symptom. Archives of internal medicine 1987;147(7):1333- 7. Page 19 of 21 BMJ Open 45. Niaz MO, Naseem M, Siddiqui SN, et al. An outline of the oral health challenges in “Pakistani” population and

a discussion of approaches to these challenges. JPDA 2013;22(03):00. BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 46. Stephan RM. Intra-oral hydrogen-ion concentrations associated with dental caries activity. Journal of dental 2 research 1944;23(4):257-66. 3 47. Lussi A, Jaeggi T, Zero D. The role of diet in the aetiology of dental erosion. Caries research 2004;38(1):34- 4 5 44. 6 48. Pruthi S, Nirupama M, Chakraborti S. Evaluation of gastric biopsies in chronic gastritis: Grading of 7 inflammation by Visual Analogue Scale. Medical Journal of Dr. DY Patil University 2014;7(4):463. 8 49. Al-Saadi AM, Al-Khayat JQ, Muhammad IM, et al. The role of Helicobacter pylori in esophagitis and peptic 9 ulcer disease in Iraq. Saudi Med J 2004;25(9):1216-22. 10 50. Schechter RB, Lemme EM, Coelho HS. Gastroesophageal reflux in cirrhotic patients with esophageal varices 11 without endoscopic treatment. Arquivos de gastroenterologia 2007;44(2):145-50. 12 13 51. Harchandani N. Oral health challenges in Pakistan and approaches to these problems. Pakistan Oral & Dental 14 Journal 2012;32(3). 15 52. AMA. Australian Medical Association Position Statement: Social Determinants of Health and the Prevention 16 of Health Inequities. For peer review only 17 ww.ama.com.au/system/tdf/documents/AMA_Position_Statement_on_the_Social_Determinants_of_Health_a 18 nd_the_Prevention_of_Health_Inequities_2007_0.pdf?file=1&type=node&id=40625(2007). Accessed August 19 2018. 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on October 1, 2021 by guest. Protected copyright. confounding 42 43 (b) Describe any methods used to examine subgroups and interactions 44 (c) Explain how missing data were addressed 45 (d) If applicable, describe analytical methods taking account of sampling strategy 46 (e) Describe any sensitivity analyses 47 48 Results 49 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially 50 eligible, examined for eligibility, confirmed eligible, included in the study, 51 completing follow-up, and analysed 52 53 (b) Give reasons for non-participation at each stage 54 (c) Consider use of a flow diagram 55 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and 56 information on exposures and potential confounders 57 (b) Indicate number of participants with missing data for each variable of interest 58 59 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 Outcome data 15* Report numbers of outcome events or summary measures 3 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates 4 and their precision (eg, 95% confidence interval). Make clear which confounders 5 were adjusted for and why they were included 6 b 7 ( ) Report category boundaries when continuous variables were categorized 8 (c) If relevant, consider translating estimates of relative risk into absolute risk for a 9 meaningful time period 10 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 11 sensitivity analyses 12 13 Discussion 14 Key results 18 Summarise key results with reference to study objectives 15 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or 16 For imprecision.peer Discuss review both direction and magnitudeonly of any potential bias 17 18 Interpretation 20 Give a cautious overall interpretation of results considering objectives, 19 limitations, multiplicity of analyses, results from similar studies, and other relevant 20 evidence 21 Generalisability 21 Discuss the generalisability (external validity) of the study results 22 23 Other information 24 Funding 22 Give the source of funding and the role of the funders for the present study and, if 25 applicable, for the original study on which the present article is based 26 27 28 I declare that the checklist maintained in reference to the submitted original research article, is correct to the best of my 29 30 knowledge. For any concerns, you can direct them to me directly. 31 32 Thank you for your consideration. I look forward to hearing from you.

Risk factors associated with oral manifestations, and oral health impact of gastroesophageal reflux disease: a multicentre, cross-sectional study in Pakistan ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-021458.R2

Article Type: Research

Date Submitted by the 10-Jan-2019 Author:

Primary Subject Dentistry and oral medicine Heading:

Gastroenterology and hepatology, General practice / Family practice, Secondary Subject Heading: Epidemiology, Health services research, Public health

Gastroenterology < INTERNAL MEDICINE, PUBLIC HEALTH, PREVENTIVE Keywords: MEDICINE, ORAL MEDICINE, Endoscopy < GASTROENTEROLOGY,

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 Risk factors associated with oral manifestations, and oral health impact of 2 3 gastroesophageal reflux disease: a multicentre, cross-sectional study in Pakistan 4 5 6 Ibrahim Warsi1,7, Javeria Ahmed1,8, Anjum Younus1,2, Abdur Rasheed3, Tayyab Akhtar4, Qurrat-ul-Ain5, Zohaib 7 6 8 Khurshid 9 1 10 Dr. Ishrat-ul-Ebad Khan Institute of Oral Health Sciences, Dow University of Health Sciences, Karachi, Pakistan 2 11 Department of Community Dentistry, Dow Dental College, Dow University of Health Sciences, Karachi, Pakistan. 12 3Department of Research, Dow University of Health Sciences, Karachi, Pakistan 13 4Gastroenterology and Liver Centre, Holy Family Hospital, Rawalpindi-Islamabad, Pakistan 14 5Department of Internal Medicine, Shalamar Hospital, Lahore. 15 6Department of Prosthodontics and Implantology, College of Dentistry, King Faisal University, Saudi Arabia. 16 7Department of Postgraduate MedicalFor Education, peer Harvard Medical review School, Harvard University,only Boston, MA, USA 17 8Department of Oral & Maxillofacial Surgery, Jinnah Postgraduate Medical Center (JPMC), Jinnah Hospital, Karachi, Pakistan 18 19 20 21 22 Ibrahim Warsi (IW) 23 Postgraduate student Masters (medical sciences and clinical investigation), Harvard Medical School, Boston, MA, USA 24 25 Javeria Ahmed (JA) 26 Post-graduate trainee (Oral & Maxillofacial Surgery), Jinnah Postgraduate Medical Center (JPMC), Jinnah Hospital, Karachi, Pakistan 27 28 Anjum Younus (AY) 29 Postgraduate trainee MDS, Department of Community Dentistry, Dr. Ishrat-ul-Ebad Khan Institute of Oral Health Sciences, Dow University 30 of Health Sciences, Karachi, Pakistan 31 32 Abdur Rasheed (AR) 33 http://bmjopen.bmj.com/ 34 Deputy director of research & biostatistician, Department of Research, Dow University of Health Sciences, Karachi, Pakistan 35 36 Tayyab Akhtar (TA) 37 Senior Registrar, Gastroenterology and Liver Center, Holy Family Hospital, Rawalpindi-Islamabad, Pakistan 38 39 Qurrat-ul-Ain (QA) 40 Internal Medicine Resident, Shalamar Hospital, Lahore, Pakistan

41 on October 1, 2021 by guest. Protected copyright. 42 43 Zohaib Khurshid (ZK) 44 Lecturer and Course Coordinator, Department of Prosthodontics and Implantology, College of Dentistry, King Faisal University, Al-Ahsa 45 31982, Saudi Arabia 46 47 48 Corresponding Author: Dr. Ibrahim Warsi 49 614-A, Shawmut Avenue, Unit-4, Boston, MA (USA), 02118 50 Tel: (+1)832-458-5039 51 Primary Email: [email protected] 52 53 Institutional Email: [email protected] 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 ABSTRACT 4 5 6 Objective: Gastro-oesophageal reflux disease (GORD) is a relatively common disorder and manifests with extra- 7 8 oesophageal symptoms, such as dental erosions (DE), cough, laryngitis, asthma, and oral soft/hard tissue pathologies. 9 10 This study aimed (i) to identify oral soft and hard-tissue changes in patients with GORD and (ii) to evaluate these oral 11 12 changes as indices for assessing GORD and its severity. 13 Setting: This cross-sectional study was conducted at four major tertiary-care government hospitals, in two metropolitan 14 15 cities of Pakistan. 16 For peer review only 17 Participants: In total, 187 of 700 patients who underwent oesophago–gastro–duodenoscopy (OGD) and having GORD 18 were included in the study. The GORD patients were divided according to the presence of dental erosions (DE) into 19 20 group A (with DE, chronic/severe GORD) and group B (without DE, mild GORD). Patients who were unconscious 21 22 and had extremely limited mouth opening were excluded. 23 24 Primary & secondary outcome measures: Abnormal conditions and lesions of the oral mucosa were recorded. The 25 impact of oral hard & soft-tissue changes on the oral health-related quality-of-life (OHRQoL) was assessed using the 26 27 Pakistani (Urdu) version of the validated Oral Health Impact Profile-14 (OHIP-14) instrument. 28 29 Results: Oral submucosal fibrosis (66.3%), ulceration (59.4%), and xerostomia (47.6%) were significantly more 30 31 common in group A (p<0.05). The prevalence of GORD was 26.7%, within which the prevalence of DE was 35.3%. 32 Unhealthy dietary pattern, nausea/vomiting, oesophagitis, xerostomia, ulceration, gingivitis, and angular cheilitis

33 http://bmjopen.bmj.com/ 34 showed a statistically significant association with chronic GORD and DE. All subscales of OHIP-14 were positively 35 36 correlated (p<0.05) in patients with GORD and DE, with notable impact on psychological discomfort (rs=0.30), 37 physical disability (rs=0.29), psychological disability (rs=0.27), and functional limitation (rs=0.20). 38 39 Conclusion: Patients with GORD and DE presented with more severe oral manifestations than did those with GORD 40

41 and no DE. We recommend timely dental check-ups to assess the severity of both systemic and oral disease. on October 1, 2021 by guest. Protected copyright. 42 43 Keywords: GORD, GERD, dental erosions, oral manifestations, oral hard and soft-tissue lesions, extra-oesophageal 44 symptoms, oral health, OHIP-14, OHRQoL 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 4 5 STRENGTH & LIMITATIONS OF THIS STUDY: 6 7 8 9 o This study is the first of its kind to bridge the gap created by either lacking evidence or controversial literature on 10 the effect of gastroesophageal reflux disorder (GORD) on oral soft tissues. 11 12 o This study assessed oral-health related quality of life using OHIP-14 instrument, the results of which highlight how 13 14 oral health and systemic health are inter-linked from a psychological and general well-being perspective. 15 o General practitioners and gastroenterologists are often the primary health care providers for patients with GORD, 16 For peer review only 17 but while addressing their main gastroenterological concerns, the oral manifestations of these systemic conditions 18 19 are often overlooked. This study highlights the need for dental referral in patients with upper GI disorders (in this 20 21 case GORD), which can have a marked effect on both systemic and oral health. 22 o Cross-sectional design of this study limits the ability to rule out confounders and establish causal-inference 23 24 relationships. Hence the study results are limited to correlation extent only. We recommend future researchers to 25 26 assess the reported risk factors in a prospective study design, in order to understand the evolution of oral morbidities 27 28 linked with the pathogenesis of GORD. 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 22 INTRODUCTION BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 Gastro-oesophageal reflux disease (GORD) is a common disorder, affecting approximately 10%–20% of the general 4 5 population.[1] The Montreal consensus classifies GORD as an entity manifesting as oesophageal and extra- 6 7 oesophageal symptoms.[2] The oesophageal symptomatology includes regurgitation or burning retrosternal chest 8 9 pain, reflux oesophagitis, strictures, Barrett’s oesophagus, and adenocarcinoma, and the extra-oesophageal 10 symptomatology comprises reflux, cough, laryngitis, asthma, and dental erosions (DE).[2] DE is a multifactorial 11 12 phenomenon whereby the protective buffering capacity of the oral cavity is compromised by reduced secretion of saliva 13 14 or high volumes of harmful gastric reflux.[3] DE appears to be the most common injury caused by GORD.[4] The 15 global prevalence of DE in the general population is not accurately known; it varies depending on the underlying cause, 16 For peer review only 17 and prevalence rates in the range of 2%–77% have been reported.[5] Tooth involvement seems to be universal, but the 18 19 most commonly observed damage occurs on the palatal surfaces of the posterior teeth, with a reported prevalence of 20 21 up to 42%.[6] Studies have highlighted the detrimental effects of gastroduodenal contents on oral soft tissue pathology 22 as well as their propensity to cause DE.[3,7] These include effects on the oesophageal epithelium, soft palate, and oral 23 24 mucosa, and manifest as burning mouth syndrome, aphthoid lesions, hoarseness of voice, erythema of the soft palate 25 26 and uvula, glossitis, epithelial atrophy, and xerostomia.[6–11] Epithelial atrophy and xerostomia further aggravate the 27 28 GORD-induced injury to the epithelium in the oral cavity and oesophagus.[12,13] 29 30 The severity of DE is directly proportional to the amount of time that gastric acid is in contact with the enamel, so the 31 32 frequency and duration of the reflux problem can be assessed by the amount of enamel loss and vice versa.[11,14,15]

33 http://bmjopen.bmj.com/ 34 We hypothesized a strong association between DE and chronic severe GORD, and that an absence of DE would indicate 35 a less severe form of GORD that is of shorter duration. The aims of this study were as follows: to assess the frequency 36 37 of GORD with and without DE and oral symptoms in patients undergoing oesophagogastroduodenoscopy (OGD); to 38 39 determine the association between the severity of GORD and the likelihood of orodental manifestations; and to 40 determine the effect of GORD on oral health-related quality of life (OHRQoL).

41 on October 1, 2021 by guest. Protected copyright. 42 43 44 MATERIALS AND METHODS 45 46 47 48 This cross-sectional study was conducted at four major tertiary care government hospitals in two of the major 49 50 metropolitan cities of Pakistan, namely the twin city of Rawalpindi-Islamabad (Pakistan Institute of Medical Sciences, 51 52 PIMS, and Holy Family Hospital) and Karachi (Jinnah Hospital, and Civil Hospital). In total, 187 of 700 patients who 53 54 underwent OGD along with a comprehensive gastrointestinal (GI) examination over four months were diagnosed to 55 have GORD,[16] and confirmed medically by treatment with a proton pump inhibitor were included in the study (100% 56 57 participation rate). A mandatory dental examination was included to evaluate the oro-dental effects of GORD. Patients 58 59 who were unconscious and those with minimal mouth opening (precluding an oral examination) were excluded. 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Assessment of demographic and clinical characteristics Page 5 of 22 BMJ Open A questionnaire was used to collect data on sociodemographics, past and current medical history (comorbidities), drug BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 history (including use of non-steroidal anti-inflammatory drugs [NSAIDs]), and risk factors for GORD (dietary pattern, 2 addiction profile, consumption of tea, eating habits, weight changes, and other GI disorders). 3 4 5 6 Assessment of oral health-related quality of life 7 The Oral Health Impact Profile (OHIP-14) is an instrument designed to assess oral health-related quality of life 8 9 (OHRQoL) and includes seven subscales. The Pakistani (Urdu) validated version of this instrument,[16] was used to 10 11 evaluate factors related to GORD and their effect on the oral cavity concerning speech production, mastication, 12 13 psychosocial well-being, and life in general, and to estimate the impact of GORD on oral and systemic health. 14 15 16 Gastrointestinal examinationFor peer review only 17 18 All study patients underwent a comprehensive gastro-intestinal (GI) examination with baseline laboratory 19 investigations, and were on their routine medical prescriptions.[15,16] In this study, GORD was diagnosed on the basis 20 21 of the clinical picture and endoscopic (OGD) findings by a gastroenterologist. The OGD findings in patients with 22 23 GORD include oesophageal erosions, erythema and strictures, a lax esophageal sphincter, oesophagitis, Barrett’s 24 25 oesophagus, oesophageal adenocarcinoma, and hiatal hernia.[17] Accurate diagnosis of oesophagitis was crucial in this 26 study because it is a risk factor for DE.[10] Compelling evidence from plasma biomarkers (increased ESR), clinical 27 28 symptoms (painful swallowing and dysphagia), signs of erythema, inflammation, and ulceration on OGD are markers 29 30 for a definitive diagnosis of oesophagitis.[18] 31 32

33 Dental examination http://bmjopen.bmj.com/ 34 35 The World Health Organization (WHO) criteria for an oral and dental examination was adapted to examine patients for 36 37 decayed teeth by using the decayed, missed filled teeth (DMFT) index value, mouth opening status, and presence of 38 oral submucosal fibrosis (OSF) using the OSF staging index.[19] DMFT was recorded as a simple count of decayed 39 40 and lost teeth. Abnormal conditions and lesions of the oral mucosa (xerostomia, ulceration, candidiasis, gingivitis,

41 on October 1, 2021 by guest. Protected copyright. 42 angular cheilitis, atrophic glossitis, and leukoplakia) were recorded as per the WHO screening protocol.[16,19,20] DE 43 44 was diagnosed when a patient complained of tooth hypersensitivity, with accompanying clinical signs of tooth 45 discoloration, dentin exposure, loss of occlusal surfaces, and a decrease in tooth height (measured from the 46 47 cementoenamel junction to the crest of the enamel).[19,21,22] All oral examinations were performed by the same 48 49 dentist (IW). 50 51 52 Statistical analysis 53 54 Data analysis was performed using SPSS Version 21 (IBM Corp., Armonk, NY, USA). Categorical variables are 55 56 reported as frequencies and percentages. Associations between GI symptoms and oral manifestations of GORD were 57 determined using the chi-squared test and logistic regression analysis. The Spearman rank correlation and Mann- 58 59 Whitney were used to test for associations and difference of the mean values between GORD groups and OHIP-14 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml subscales, respectively. A binary outcome variable (GORD without DE, 0; and GORD with DE, 1) was created to BMJ Open Page 6 of 22 compute results.This paper presents multiple regression models to report risk factors related to the outcome variable BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 (GORD with DE). The results are reported as crude odds ratios (COR), adjusted odds ratio (AOR), with 95% confidence 2 interval (CI), and p-values. Bivariate logistic regression was used to obtain COR’s for assessing the ‘likelihood of 3 4 GORD and DE’ against ‘independent predictor variables,’ extracted from tables’ one, two, and four. Variables with 5 6 p≤0.2 were entered in the multivariable logistic regression model to obtain AOR’s. Model’s performance was evaluated 7 using measures of discrimination. It addresses the extent to which the model predicts a higher probability of the outcome 8 9 (GORD with DE), with certain set of predictors in a model; using the area under the receiver operating characteristic 10 11 curve (AU-ROC or AUC) or c-statistics. AUC value range from 0.5 (no discrimination) to 1 (perfect 12 13 discrimination).[23] 14 15 16 Patient and public involvementFor peer review only 17 18 No patients were involved in the development of this study. However, a pilot group of patients played a crucial role in 19 testing the responsiveness of the questionnaire and identifying gaps in the study. Additional study variables, including 20 21 xerostomia, oral submucosal fibrosis, and DMFT, were added to the clinical dental section after observing consistency 22 23 of these characteristics in the specified population. Moreover, the translated OHIP-14 questionnaire was tested, and 24 25 feedback was obtained from the pilot patients; this process helped in making various amendments concerning language 26 and comprehension of the OHIP scale items.[16] Within the GI section, it was observed from the patient’s history, that 27 28 they were prone to self-medication (particularly NSAIDs) due to the history of chronic back pain. Therefore, these 29 30 variables were also assessed in this study. All the patients were assigned a unique identifier code to maintain patient 31 32 privacy and confidentiality. The patients did not play any role in the initiation of this study or drafting this manuscript.

33 http://bmjopen.bmj.com/ 34 35 36 Ethical approval 37 38 The Dow ethical review committee approved the study protocol (IRB department of Dow University of Health 39 Sciences). While departmental permissions were sought from all four participating hospitals (CHK, JPMC, PIMS & 40

41 Holy Family), before participant enrolment, written informed consent was obtained from all study participants after on October 1, 2021 by guest. Protected copyright. 42 43 explaining the study protocol. 44 45 46 47 RESULTS 48 49 50 51 Most study participants were males (58.3%), around 80% were in the adult age group (41–60 years), as shown in Table 52 1. The study sample was ethnically and racially diverse; most participants were Punjabis (54.5%), followed by Urdu- 53 54 speaking Mohajirs (16.6%), Pathans (15%), and Sindhis (8%). Most were urban residents (63.6%) of intermediate 55 56 (35.3%) to low (62%) socioeconomic status. The addiction profile predominantly consisted of oral tobacco (30.5%) 57 58 and cigarette smoking (26.7%), with a low rate of alcohol consumption (2.7%). The most commonly consumed 59 beverage was chai-tea (72.7%). A sizeable proportion experienced generalized body pain (44%), and reported self- 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml medication, particularly with NSAIDs (46%). The most common comorbidities were diabetes (34.8%), hypertension Page 7 of 22 BMJ Open (23%), hepatitis C (34.2%), and chronic liver disease (CLD) (18.7%). Slightly more than half (56.7%) of the study BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 population had GORD alone, while the remaining had additional upper GI conditions, including gastritis (19.8%), 2 portal gastropathy (11.2%), peptic ulcer disease (10.2%), GORD-related oesophagitis (4.8%), and hiatal hernia (4.3%). 3 4 5 6 Table 1. Demographic variables and baseline characteristics in patients with GORD 7 a 8 S.No. DEMOGRAPHIC VARIABLES N (%) 9 1 Gender Male 109 (58.3%) 10 Female 78 (41.7%) 11 2 Age Category <19years 09 (4.8%) 12 20-30 years 34 (18.2%) 31-40 years 26 (13.9%) 13 41-50 years 48 (25.7%) 14 51-60 years 42 (22.5%) 15 61-70 years 19 (10.2%) 16 For peer71-80 years review only 09 (4.8%) 17 3 Race/Ethnicity Sindhi 15 (8.0%) 18 Balochi 05 (2.7%) 19 Punjabi 102 (54.5%) 20 Pathan 28 (15%) 21 Kashmiri 06 (3.2%) 22 Urdu Speaking Mohajirs 31 (16.6%) 23 4 Domicile Urban 119 (63.6%) 24 Rural 68 (36.4%) 25 5 Socio-economic status (SES) b Low 116 (62%) 26 Moderate 66 (35.3%) 27 High 05 (2.7%) 28 6 Dietary Pattern c Healthy diet 23 (12.3%) 29 Satisfactory diet 80 (42.8%) 30 Unhealthy diet 84 (44.9%) 31 7 Addiction Smoking 50 (26.7%) 32 Oral Tobacco 57 (30.5%) Alcohol 05 (2.7%)

33 http://bmjopen.bmj.com/ 34 8 Tea consumption Yes 136 (72.7%) 35 No 51 (27.3%) 9 NSAIDS consumption Yes 86 (46.0%) 36 No 101 (54.0%) 37 10 History of Body Ache/General Body Yes 83 (44.0%) 38 Pain No 104 (56.0%) 39 11 Weight Loss (from <5Kg to >20Kg) No 45 (24.1%) 40 1-5 Kg 72 (38.5%)

41 5-10 Kg 13 (7.0%) on October 1, 2021 by guest. Protected copyright. 42 10-15 Kg 36 (19.3%) 43 15-20 Kg 15 (8.0%) 44 >20Kg 06 (3.2%) 45 12 Co-morbidities Hypertension 65 (34.8%) 46 Diabetes Mellitus 43 (23%) 47 Hepatitis B 08 (4.3%) 48 Hepatitis C 64 (34.2%) 49 Chronic Liver Disease (CLD) 35 (18.7%) 50 13 Dental Erosions Group A: GORD with DE 66 (35.3%) 51 Group B: GORD without DE 121 (64.7%) 52 14. GORD (alone) and (in combination GORD Alone 106 (56.7%) 53 with other GI illness) GORD with additional Upper-GI conditions 81 (43.3%) 54 i. Gastritis 16 (19.8%) 55 ii. Portal Gastropathy 09 (11.2%) 56 iii. Peptic Ulcer Disease (PUD) 08 (10.2%) 57 iv. GERD Esophagitis 04 (4.8%) 58 v. Hiatal Hernia 04 (4.3%) vi. Others 40 (49.7%) 59 GORD, gastro-oesophageal refluxFor disease. peer GI,review Gastro-intestinal. only - http://bmjopen.bmj.com/site/about/guidelines.xhtml DE, Dental Erosions. aItems in bold highlight notable prevalence in each 60 category. bSES (categorized on the basis of monthly family income, number of family dependents and educational status). cDietary pattern (Healthy diet refers to a balanced diet; Satisfactory diet: being intermediary between healthy and unhealthy diet, explicitly relating to BMJ Open Page 8 of 22 occasionally eating junk food, while diet relatively lacks in natural abrasives like fruits/vegetables, cereals, leafy vegetables, and fibers; Unhealthy diet refers to consuming high fatty/oily meals, frequent consumption of junk food/fast food, and soft drinks, while completely devoid of natural abrasives).[16] BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 3 DE is a known comorbidity of chronic GORD; hence, we divided the patients based on the presence of DE into two 4 groups (binary variable), namely group A (with DE, chronic/severe GORD; na = 66, 35.3%) and group B (without DE, 5 6 mild GORD; nb = 121, 64.7%). The oral and systemic findings were compared between the two groups. Table 2 shows 7 8 the GI signs and symptoms that were found in both study groups; the most common of which was 9 heartburn/regurgitation (95%), followed by nausea/vomiting (82%) and dysphagia (42%). A statistically significant 10 11 difference in the frequency of nausea/vomiting (OR 0.35, 95% CI 0.16–0.76, p=0.006), early satiety/loss of appetite 12 13 (OR 1.90, CI 1.02–3.55, p=0.042), and melena (OR 1.96, CI 1.07–3.59, p=0.029) was found between patients who had 14 15 GORD with DE than did those without DE. 16 For peer review only 17 18 19 Table 2. Gastrointestinal signs and symptoms in patients with GORD 20 21 GI SYMPTOMS Group A Group B ALL OR* Confidence P-value 22 (na=66) (nb=121) (N=187) Interval 23 24 Heartburn/Regurgitation Yes 63 115 178 (95.2%) 1.096 (0.265, 4.531) 0.900 No 03 06 09 (4.8%) 25 Nausea/Vomiting Yes 48 107 155 (82.9%) 0.349 (0.160, 0.759) 0.006** 26 No 18 14 32 (17.1%) 27 Abdominal Pain/discomfort Yes 49 92 141 (75.4%) 0.909 (0.455, 1.815) 0.786 28 No 17 29 46 (24.5%) 29 Abdominal distension Yes 40 62 102 (54.5%) 1.464 (0.796, 2.692) 0.219 30 No 26 59 85 (45.4%) 31 Early Satiety/ Loss of Appetite Yes 44 62 106 (56.7%) 1.903 (1.020, 3.551) 0.042** 32 No 22 59 81 (43.3%)

33 Dysphagia Yes 32 48 80 (42.8%) 1.431 (0.782, 2.620) 0.244 http://bmjopen.bmj.com/ 34 No 34 73 107 (57.2%) 35 Hematemesis Yes 27 44 71 (38.0%) 1.212 (0.655, 2.240) 0.540 36 No 39 77 116 (62.0%) 37 Melena Yes 36 46 82 (43.9%) 1.957 (1.065, 3.593) 0.029** 38 No 30 75 105 (56.1%) 39 Group A, chronic/severe GORD with DE. Group B, mild GORD without DE. *OR, Odds ratios (of GI symptoms against GERD+DE as reference category). 40 **p<0.05 (chi-squared test of proportion) was statistically significant.

41 on October 1, 2021 by guest. Protected copyright. 42 43 The oral manifestations associated with GORD in groups A and B are shown in Table 3. Overall, patients with GORD 44 45 and DE had significantly more oral manifestations than did those without DE. There was also a trend of worse dentition 46 status (poor DMFT >10), as indicated by a higher mean DMFT index value in group A than in group B, suggesting that 47 48 teeth with DE were more vulnerable to decay and tooth loss. 49 50 51 The coexistence of GERD and DE was found most likely (in decreasing order of frequency) with aphthoid ulcerative 52 53 lesions (OR 13.7, p<0.01), followed by gingivitis (OR 9.0, p<0.01), angular cheilitis (OR 5.2, p<0.01), xerostomia (OR 54 4.7, p<0.01), atrophic glossitis (OR 2.9, p=0.001), oral submucosal fibrosis (OR 2.9, p=0.003), and oral candidiasis 55 56 (OR 2.2, p=0.028). Oral manifestations with a p-value <0.05 were included in the multivariable logistic regression 57 58 model, hence leukoplakia and DMFT were excluded. This multivariable logistic regression model was created by the 59 backward selection (likelihood ratio) method. Given that candidiasis has a biological association with GERD and DE, 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml it was retained in the model. Interestingly, the multivariable logistic regression model had a good area under the Page 9 of 22 BMJ Open receiver-operating characteristic curve (AUROC), with a value of 0.855. The highest correlations for GORD with DE, BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 was found with ulceration (AOR 6.71, 95% CI 2.48–18.12, p<0.01) and gingivitis (AOR 4.76, CI 2.12–10.71, p<0.01), 2 followed by angular cheilitis (AOR 2.53, CI 1.03–6.15, p=0.041) and xerostomia (AOR 2.37, CI 1.05–5.37, p=0.038). 3 4 However, candidiasis did not retain its significance in the multivariable analysis. 5 6 Table 3. Oral manifestations in patients with GORD 7 8 9 ORAL Group Group ALL C. OR (95% CI) P- A. OR (95% CI) P- 10 MANIFESTATIONS A B (n=187) value* value** 11 (n=66) (n=121) 12 1. Xerostomia Yes 47 42 89 (47.6%) 4.653 (2.426 , 8.923) <0.01* 2.372 (1.047, 5.373) 0.038** 13 No 19 79 98 (52.4%) 1 1 14 2. Ulceration Yes 60 51 111 (59.4%) 13.725 (5.506, 34.218) <0.01* 6.705 (2.481, 18.122) <0.01** 15 No 6 70 76 (40.6%) 1 1 16 3. Gingivitis Yes For47 peer26 73 (39.0%) review9.038 (4.547, only 17.968) <0.01* 4.768 (2.123, 10.706) <0.01** 17 No 19 95 114 (61.0%) 1 1 4. Candidiasis Yes 20 20 40 (21.4%) 2.196 (1.078, 4.471) 0.028* 0.706 (0.267, 1.868) 0.483 18 No 46 101 147 (78.6%) 1 1 19 5. Angular Cheilitis Yes 24 12 36 (19.3%) 5.190 (2.382, 11.312) <0.01* 2.526 (1.038, 6.148) 0.041** 20 No 42 109 151 (80.7%) 1 21 6. Atrophic Glossitis Yes 30 27 57 (30.5%) 2.901 (1.520, 5.538) 0.001* - - 22 No 36 94 130 (69.5%) 1 23 7. Leukoplakia Yes 2 6 08 (4.3%) 0.599 (0.117, 3.055) 0.533 - - 24 No 64 115 179 (95.7%) 1 25 8. Oral Submucosal Yes 53 71 124 (66.3%) 2.871 (1.417, 5.818) 0.003* - - 26 Fibrosis (OSF) No 13 50 63 (33.7%) 1 27 9. DMFT status <3 24 64 88 (47.06%) - 0.069 - - 28 4-10 24 37 61 (32.62%) 29 >10 18 20 38 (20.32%) 30 Items in bold represent major findings (of note) in their respective field. *p<0.05 (chi-squared test) was considered statistically significant. Group A, 31 chronic/severe GORD with DE. Group B, mild GORD without DE. DMFT (decayed, missed, filled teeth), pertaining to tooth decay and tooth loss. DMFT<3 32 (acceptable), DMFT 4-10 (average), DMFT>10 (poor).[16] 95% CI, confidence interval; C.OR, Crude Odds Ratio (chi-square/linear regression); A.OR, Adjusted Odds Ratio (values from multivariate logistic regression analysis). **Odds ratios were computed for oral manifestations against a reference variable,

33 http://bmjopen.bmj.com/ ‘GORD with DE.’ Using backward selection (likelihood ratio) method, variables were entered in the multivariable analysis. The adjustment was made for 34 atrophic glossitis, oral submucosal fibrosis, leukoplakia, and DMFT status. Area under ROC (AUROC) value: 0.855, was used as an assessment of the 35 multivariable logistic regression model’s discriminative ability (to predict between the presence & absence of DE). 36 37 38 39 Table 4 shows the seven-subscale (OHIP-14) scores for groups A and B. The mean values for all subscales were 40 relatively higher in group A, suggesting a higher risk of ‘poor OHRQoL,’ than that in group B. Spearman ranks

41 on October 1, 2021 by guest. Protected copyright. 42 correlation (rs) and Mann-Witney U statistical tests, are used for correlations, and to compare means, in both groups A 43 44 & B respectively. Mann-Whitney U test reveals that a statistically significant difference has been observed for all OHIP 45 46 subscales (except social disability) when compared between the two groups (A & B). The coexistence of GORD and 47 DE had a positive correlation with OHIP subscales (p<0.05), with notable impact on psychological discomfort 48 49 (rs=0.30), physical disability (rs=0.29), psychological disability (rs=0.27), and functional limitation (rs=0.20). 50 51 52 Table 4. Psychometric properties of OHIP-14, correlated with GORD, with and without dental erosions. 53 54 OHIP-14 Sub-scales Mean subscale (S.D) Statistical Tests 55 Characteristics With DE (group A) Without DE (group B) Spearman Ranks: rs Mann-Whitney U test 56 na=66 nb=121 correlation value (p-value)* 57 Functional Limitation 3.530 (2.199) 2.5702 (1.92712) 0.200 * 3065.0 (0.006) * 58 Physical Pain 1.182 (1.264) 0.7273 (0.96609) 0.167 * 3260.0 (0.023) * 59 Psychological Discomfort 5.167 (1.853) 4.0744 (1.98396) 0.297 * 2625.0 (<0.001) * 60 Physical Disability For peer 6.394review (2.745) only - http://bmjopen.bmj.com/site/about/guidelines.xhtml4.4463 (2.81351) 0.288 * 2672.0 (<0.001) * Psychological Disability 4.591 (1.745) 3.5868 (1.94795) 0.271 * 2754.5 (<0.001) * BMJ Open Page 10 of 22 Social Disability 7.106 (2.835) 6.1405 (3.42127) 0.142 3332.5 (0.053) Handicap 1.682 (1.230) 1.2149 (1.31153) 0.193 * 3099.0 (0.008) * BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 *Correlation is significant at p<0.05 level (2-tailed). rs, ranks correlation coefficient. Bold highlight represents important values in each statistical test results. 2 3 4 Table 5 shows the bivariate relationship between the significant study variables (extracted from tables 1, 2, and 4), and 5 6 the dichotomous/binary outcome variable (i.e., GORD with DE, as the reference category), computed in a logistic 7 8 regression model. Univariate analysis revealed that patients with an unhealthy diet were more likely to develop GORD 9 with DE than those with a healthy diet (OR 6.03, 95% CI 1.89–9.27). Early satiety/loss of appetite (OR 1.90, CI 1.020– 10 11 3.56) and nausea/vomiting (OR 0.35, CI 0.16–0.76) increased the likelihood of developing DE by 1.90-fold and 0.35- 12 13 fold, respectively. Moreover, melena was found to be a common manifestation in this population with chronic GORD 14 15 (OR 1.96, CI 1.07–3.59). Patients with peptic ulcer disease (PUD) were more likely to develop chronic GORD with 16 DE (OR 2.83, CI 1.08–7.42).For Similarly, peer the OHIP reviewsubscale scores for only functional limitation, physical pain, physical 17 18 disability, psychological discomfort, and psychological disability were sensitive to the severity of GORD and the 19 20 presence of DE. 21 22 The results of the multivariable logistic regression model are also presented in Table 5, with adjustments made for all 23 24 independent variables. Some variables that had a statistically significant association with DE in the univariate analysis 25 26 (i.e., loss of appetite, melena, peptic ulcer disease, and the OHIP subscales: functional limitation, physical pain, 27 psychological discomfort, and psychological disability) were no longer significant in multivariable analysis. Portal 28 29 gastropathy had a p-value of >0.2 therefore excluded from multivariable analysis. However, an unhealthy diet (AOR 30 31 2.31, 95% CI 0.40–13.21), nausea/vomiting (AOR 0.13, CI 0.03–0.51), oesophagitis (AOR 12.43, CI 1.66–93.14), and 32 advanced physical disability (AOR 0.56, CI 0.02–16.16) retained their significant association with GORD and DE, in

33 http://bmjopen.bmj.com/ 34 the multivariable analysis. 35 36 37 38 Table 5. Risk factors associated with GORD and dental erosions: results of the univariate and multivariable analysis. 39 40 CHARACTERISTICS GORD with Dental Erosions 41 C. OR (95% CI) p-value A. OR (95% CI) p-value on October 1, 2021 by guest. Protected copyright. 42 1. Dietary Pattern <0.01* <0.001* 43 44 Healthy 1 1 45 Satisfactory 1.096 (0.325, 3.697) 0.234 (0.041, 1.320) 46 Unhealthy 6.034 (1.889, 19.268) 2.307 (0.403, 13.207) 47 2. GI symptom: Nausea/vomiting 0.008* 0.004* 48 No 1 1 49 Yes 0.349 (0.160, 0.759) 0.130 (0.033, 0.512) 50 3. GI symptom: Early satiety/loss of appetite 0.043* 0.262 51 No 1 1 52 Yes 1.903 (1.020, 3.551) 1.778 (0.650, 4.867) 53 4. GI symptom: Melena 0.030* 0.421 54 No 1 1 55 Yes 1.957 (1.065, 3.593) 1.509 (0.554, 4.111) 56 5. GI disorder: Oesophagitis 0.20 0.014* 57 No 1 1 Yes 2.398 (0.621, 9.255) 12.427 (1.658, 93.143) 58 6. GI disorder: Peptic ulcer disease (PUD) 0.035* 0.302 59 No 1 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Yes 2.825 (1.075, 7.423) 2.143 (0.505, 9.095) 7. GI disorder: Hiatal hernia 0.20 0.201 Page 11 of 22 BMJ Open No 1 1 Yes 0.251 (0.03, 2.082) 0.159 (0.010, 2.655) BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 8. GI disorder: Portal gastropathy 0.214 - No 1 - 2 Yes 1.786 (0.715, 4.458) - 3 9. GI disorder: Gastritis 0.06 0.065 4 No 1 1 5 Yes 2.013 (0.970, 4.179) 3.388 (0.926, 12.398) 6 7 10. OHIP Subscale 1: Functional Limitation 0.047* 0.277 8 1= Null 1 1 9 2= Mild 1.379 (0.506, 3.753) 0.511 (0.106, 2.464) 10 3= Moderate 1.495 (0.581, 3.843) 0.537 (0.132, 2.193) 11 4= Severe 3.646 (1.204, 11.044) 0.607 (0.086, 4.313) 12 5= Advanced 5.469 (1.265, 23.640) 6.149 (0.491, 76.934) 13 11. OHIP Subscale_2: Physical Pain 0.023* 0.118 14 1= Null 1 1 15 2= Mild 0.484 (0.168, 1.391) 0.202 (0.034, 1.186) 16 3= Moderate For1.935 (0.929, peer 4.032) review only2.706 (0.633, 11.574) 17 4= Severe 3.116 (0.917, 10.591) 4.271 (0.473, 38.585) 18 5= Advanced 6.677 (0.668, 66.769) 2.391 (0.104, 55.127) 19 12. OHIP Subscale_3: Physical Disability 0.002* 0.024* 20 1= Null 1 1 21 2= Mild 0.346 (0.057, 2.095) 0.042 (0.003, 0.674) 22 3= Moderate 3.033 (0.936, 9.822) 0.861 (0.101, 7.331) 23 4= Severe 3.414 (1.011, 11.527) 0.125 (0.010, 1.613) 24 5= Advanced 15.750 (2.335, 106.227) 0.564 (0.020, 16.163) 25 13. OHIP Subscale_4: Psychological Discomfort 0.002* 0.358 26 27 1= Null 1 1 2= Mild 2.857 (0.282, 28.960) 2.105 (0.073, 60.804) 28 3= Moderate 2.264 (0.262, 19.563) 3.080 (0.144, 65.939) 29 30 4= Severe 6.303 (0.741, 53.650) 3.303 (0.145, 75.217) 31 5= Advanced 11.692 (1.302, 105.028) 10.356 (0.457, 234.689) 32 14. OHIP Subscale_5: Handicap 0.057 0.171 1= Null 1 1

33 http://bmjopen.bmj.com/ 34 2= Mild 1.981 (0.795, 4.935) 1.884 (0.415, 8.563) 35 3= Moderate 2.869 (1.303, 6.315) 2.257 (0.569, 8.948) 36 4= Severe 3.900 (1.282, 11.860) 9.210 (1.348, 62.911) 37 5= Advanced 1.891 (0.600, 5.964) 0.756 (0.126, 4.529) 38 15. OHIP Subscale_6: Psychological Disability 0.006* 0.207 39 1= Null 1 1 40 2= Mild 3.000 (0.520, 17.316) 13.130 (0.913, 188.820)

41 3= Moderate 2.763 (0.582, 13.122) 6.336 (0.638, 62.938) on October 1, 2021 by guest. Protected copyright. 42 4= Severe 8.077 (1.682, 38.784) 13.513 (1.273, 143.448) 43 5= Advanced 8.437 (1.457, 48.851) 9.434 (0.899, 98.955) 44 All study variables expressing p-value<0.2 in univariate analysis, were included in the multivariable analysis, and are presented here in this table. *p- 45 value<0.05 was considered statistically significant (bold highlight). 95% CI, confidence interval; C. OR, crude/unadjusted odds ratios (obtained from 46 univariate analysis); A. OR, adjusted odds ratios for all independent variables (obtained from multi-variable analysis, by step-wise selection method). 47 48 49 Using the backward selection (likelihood ratio) method, all seventeen variables that were significant in univariate 50 51 analysis (of table 3 and 5) were entered in the cumulative multivariable logistic regression model (as shown in table 6). 52 In this model, gingivitis (adjusted OR 7.516, 95% CI 2.517–22.443; p<0.001), ulceration (adjusted OR 6.609, 95% CI 53 54 2.007–21.765; p=0.002), angular cheilitis (adjusted OR 4.028, 95% CI 1.237–13.115; p=0.021), xerostomia (adjusted 55 56 OR 3.005, 95% CI 1.026–8.805; p=0.045), unhealthy diet (adjusted OR 0.474, 95% CI 0.091–2.452; p=0.003), and 57 nausea/vomiting (adjusted OR 0.256, 95% CI 0.067–0.980; p=0.047), were significant in relation to GORD and DE. 58 59 The AUROC was used to assess the ability of the multivariable regression model to discriminate between GORD with 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 22 DE and GORD without DE. The AUROC was 0.922, indicating excellent discrimination and confirming that these BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 variables presented in the model, are major risk factors for GORD with DE. 2 3 4 Table 6: Risk factors for GORD with DE: results of the cumulative multi-variable logistic regression model. 5 6 RISK FACTORS (for GORD with DE’s) C. OR (95% CI) P-value* A. OR (95% CI) P-value** 7 8 1. Unhealthy Dietary pattern Yes 6.034 (1.889, 19.268) <0.01 0.474 (0.091, 2.452) 0.003 9 No 1 1 10 2. Nausea/ Vomiting Yes 0.349 (0.160, 0.759) 0.008 0.256 (0.067, 0.980) 0.047 11 No 1 1 12 3. Xerostomia Yes 4.653 (2.426, 8.923) <0.01 3.005 (1.026, 8.805) 0.045 13 No 1 1 14 4. Ulceration Yes 13.725 (5.506, 34.218) <0.01 6.609 (2.007, 21.765) 0.002 15 No 1 1 16 5. Gingivitis ForYes peer9.038 (4.547, review 17.968) <0.01 only7.516 (2.517, 22.443) <0.01 17 No 1 1 18 6. Angular Cheilitis Yes 5.190 (2.382, 11.312) <0.01 4.028 (1.237, 13.115) 0.021 19 No 1 1 20 95% CI, confidence interval; C.OR (crude odds ratio) represent a univariate analysis, while A.OR (adjusted odds ratio) represent multi-variable analysis 21 (using backward-elimination method). *All significant variables with p-value<0.05, obtained through univariate analysis (from table 3 and 5), are 22 computed against dichotomous outcome variable, ‘GORD with DE.’ **Using backward selection (likelihood ratio), all seventeen variables were entered 23 in multivariable logistic regression model; only six of seventeen significant variables (from table 3 & 5), retained significance within the final regression 24 model, which are shown here in this table. This model was tested for discrimination, with an AUROC value of 0.922. 25 26 27 28 29 DISCUSSION 30 31 32 GORD with DE and its associated risk factors has been the subject of much research interest in recent years. The present

33 http://bmjopen.bmj.com/ 34 study focused on patients who had GORD with or without DE and recorded the prevalence, distribution, severity, and 35 36 risk factors of this clinical entity in a relatively large sample. The influences of known risk indicators and other possible 37 determinants that have been less well studied in the past were also investigated.[7,11] 38 39 40

41 GORD has multiple systemic effects, particularly in the GI tract (Table 2). There is considerable medical literature on on October 1, 2021 by guest. Protected copyright. 42 43 GORD and its impact on the oral hard tissue (i.e., DE), as well as its periodontal effects;[8] however, data on the 44 relationship between GORD-mediated acidic oral mucosal lesions and oral health are either scarce or controversial.[7- 45 46 12,14–16,22-25] Our finding of a prevalence of DE in patients with GORD of 35.3%, fits well within the regionally 47 48 (South Asian population) reported prevalence of DE (ranges between 5.0 to 58.4 percent);[24,26–28] while global 49 50 prevalence ranges from 2% to 77%, this wide range differ with age, gender, ethnicity and population-based underlying 51 risk factors.[5] The influence of notable risk factors (such as obesity, metabolic syndrome, strong addiction pattern, 52 53 and alcohol consumption) for development of GORD & DE, in this specific population, was negligible; which supports 54 55 the fact that the reported prevalence of DE in this study was relatively low (unlike the western population).[17,29,30] 56 Moreover, this study’s population was predominantly hospital-based patients, seeking tertiary care for chronic ailments. 57 58 As evident around 43% of subjects have other GI pathologies in addition to GORD, while chronic illness like Hepatitis 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 22 BMJ Open C (34%) and chronic liver disease (19%) contribute as eminent comorbidities; Furthermore, around 75% of this BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 population had experienced weight loss of varying degrees, which eliminates obesity (or weight gain) as a risk factor. 2 3 4 Our results indicate that xerostomia, ulceration, gingivitis, and angular cheilitis have the highest correlation with GORD 5 6 and DE. Some of our findings are in accordance with the few previous studies.[7,11] Since there have been lacking 7 studies assessing the prevalence of xerostomia and oral lesions in patients with GORD, therefore literature reveals 8 9 mixed results.[4,9,24,31] Saliva is considered to play a leading role in protecting the oesophageal mucosa against 10 11 gastric reflux, and its qualitative (e.g., deficiencies of salivary growth factor and cytokines) and quantitative (e.g., 12 13 hyposalivation) abnormalities have been linked to the pathogenesis of GORD,[13,32–34] compromised dental 14 health,[35] and DE.[36] The pathogenesis of GORD appears to be connected with the decreased production of saliva, 15 16 and our findings suggest For that xerostomia peer should review be included inonly the extra-oesophageal symptomatology of 17 18 GORD.[7,11,37] 19 20 21 Approximately 60% of participants had ulcerative lesions on the soft and hard palate mucosa, buccal mucosa, uvula, 22 23 and tongue, which could not be characterized clinically as any other disease; these lesions are typically recognized as 24 25 ‘soft tissue aphthoid-lesions’ related to GORD,[4,7,9,24,31] and are caused directly by the corrosive effects of refluxed 26 acid. 27 28 29 30 A recent study in a rat model of the soft palate by Habesoglu et al.,[6] identified epithelial and microscopic alterations 31 32 in response to acid injury. Interest in this phenomenon dates to the late 1980s, when a study by Järvinen et al.,

33 demonstrated a marked prevalence of oral mucosal changes in the presence of GORD, including a burning mouth, http://bmjopen.bmj.com/ 34 35 aphthoid lesions, erythema of the soft palate and uvula, glossitis, and epithelial atrophy.[9] However, these findings 36 37 were called into question by the studies of Ranjitkar et al.,[12] and Petruzzi et al.,[25] who found that mucosal changes 38 are quite common and not an entity-specific for GORD. Similarly Deppe et al.,[8] also studied the effects of GORD on 39 40 the oral mucosa and found positive and negative signs for erythema and ulceration, respectively, but could not find a

41 on October 1, 2021 by guest. Protected copyright. 42 statistically significant difference between their erosive and non-erosive reflux groups. In addition, a study by Meurman 43 44 et al.,[5] found no mucosal changes, that could be linked to GORD. On the contrary, a recent retrospective analysis by 45 Watanabe et al identified significant association of GORD with dysphagia, xerostomia, oral ulceration, gingivitis, and 46 47 oral inflammation on buccal mucosa and tongue.[7] Considering the findings of Watanabe and Rajalalitha et al.,[7,38] 48 49 we hypothesized that the chronic-acid injury caused by GORD would be a source of persistent irritation and 50 51 inflammation affecting the oral mucosa, resulting in the infiltration of inflammatory markers. These markers include 52 increased populations of interleukins, cytokines, tumor necrosis factor-alpha, interferon-alpha, and growth factors such 53 54 as transforming growth factor-beta, which are produced at the site of inflammation. This mechanism of inflammation 55 56 is exaggerated in situations of chronic and constant irritation such as GORD and supra-oesophageal reflux 57 disease.[26,38–41] While this finding is accompanied by epithelial atrophy and fibro-elastic changes in the lamina 58 59 propria.[38,42] Also, this mechanism is consistent with the significant finding of oral submucosal fibrosis (OSF) in our 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml study population, which can lead to metaplastic changes in the oral mucosa; and can also cause restricted mouth BMJ Open Page 14 of 22 opening, resulting in trismus, eating difficulties, impaired speech, and generalized impairment of oral health.[7,16,38– BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 40] Our study findings demonstrate a positive association of chronic GORD with OSF, which is an oral precancerous 2 lesion.[42] OSF has never been reported in association with GORD, partly because these oral pathologies were not 3 4 studied in-depth amongst GORD population.[3-15,20-22,24-28,43,44] However, because of its significant impact on 5 6 quality of life, it should be considered and addressed in future researches. 7 8 9 With chronic GORD-mediated acid exposure, the salivary gland epithelium sustains severe damage (i.e., epithelial 10 11 metaplasia), resulting in xerostomia.[7,11,20,26] This association may further aggravate acid-mediated oro- 12 13 oesophageal epithelium injury, resulting in GORD-related oesophagitis.[10,13,45] Other conditions, including oral 14 ulceration, associated infections (candidiasis and angular cheilitis), and gustatory dysfunction, are also common.[7,45] 15 16 For peer review only 17 Correa et al. noted that patients with chronic GORD had reduced salivary buffering capacity and concluded that this 18 19 was the predominant cause of tooth erosion.[46] The same study also reported a lower prevalence of dental caries, 20 which was attributed to the low prevalence of cariogenic bacteria (Lactobacilli and Streptococci) in the saliva of patients 21 22 with chronic GORD. This may have a possible association with GORD mediated DE that needs confirmation in future 23 24 studies. However, our findings with regard to dental caries are contrary to Correa et al.’s, in sense that the mean DMFT 25 26 index values were higher in the group with GORD and DE (DMFT > 4 in 63.6%) than in the group with GORD and 27 no DE (DMFT > 4 in 47.1%). We postulate that the additive effects of direct acid injury, low salivary buffering capacity, 28 29 and increased opportunistic bacterial populations caused a marked increase in tooth decay and loss (indicated by 30 31 increased DMFT-index values in our study), as evident in the high-risk population with chronic GORD and 32 DE’s.[7,21,35] However, it can also be attributed to poor oral hygiene status, primarily because 27% of our study

33 http://bmjopen.bmj.com/ 34 population were smokers, and 30% chewed tobacco.[47] An acidic/unhealthy dietary pattern was common in GORD 35 36 population,[22] as was excessive consumption of traditional beverages (i.e., sweet chai/ milk tea - 73%), which would 37 38 also have contributed to a low oral pH and provided a favorable environment for opportunistic bacteria causing tooth 39 decay and tooth loss.[7,48,49] 40

41 on October 1, 2021 by guest. Protected copyright. 42 There is a wealth of literature identifying DE as being comorbid with GORD,[2] and a review by Ranjitkar et al., 43 44 reported the median prevalence for DE in patients with GORD to be 24%.[21] Accordingly, the Montreal consensus 45 postulated that the prevalence of DE is directly related to GORD, particularly when noted on the lingual and palatal 46 47 tooth surfaces.[2] Our study findings are consistent with the concept that DE has a significant association with GORD 48 49 and may serve as a marker of disease severity. As in a study by Meurman et al.,[5] our study’s statistical comparison 50 of GORD patients with and without DE, demonstrated that unlike subjective oral symptoms, oral manifestations were 51 52 significantly more common in the group with GORD and DE. 53 54 55 56 Considering the findings of Meurman et al.,[5] we used the validated OHIP-14 instrument,[16] instead of subjective 57 oral symptoms, and perceived oral health to assess the impact of severity of GORD (with DE) on oral health from a 58 59 psychological and general well-being perspective. Until now, this tool has not been tested amongst this population.[22] 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Using the OHIP-14, we found that GORD with DE, was significantly correlated (p<0.05) with the psychometric Page 15 of 22 BMJ Open characteristics (of OHIP-14) namely: psychological discomfort, physical disability, psychological disability, functional BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 limitation, handicap, and physical pain. This result is consistent with our finding that GI conditions and their oral 2 manifestations have a considerable adverse impact on oral health.[16] Notably, in this study, 43.3% of the study 3 4 population had GORD alone, and the rest had additional upper GI conditions, including gastritis, portal gastropathy, 5 6 peptic ulcer disease, GORD-related oesophagitis, and hiatal hernia, which are believed to be factors that initiate and/or 7 lead to progression of GORD.[16,50–52] However, a significant proportion of patients were habitual towards unhealthy 8 9 dietary pattern (45%), experienced frequent nausea/ vomiting (83%), and also self-medicated themselves with NSAIDs 10 11 for pain relief (46%), which are risk factors that further aggravate upper-GI illness (particularly GORD).[22,27,50] 12 13 This study has some limitations. First, the study design being cross-sectional, limits the ability of the study results to 14 15 correlation extent only. Since this study is the first of its kind to extensively report risk factors, future prospective cohort 16 For peer review only 17 studies are needed to establish causality. Second, the staging of DE was not recorded. Third, only one investigation 18 19 was used to diagnose GORD, and no measures of disease severity, such as manometry, 24-hour pH monitoring, or 20 biopsies, were included.[46] Fourth, other oral manifestations of GORD that might have a causal relationship with DE, 21 22 such as periodontal and gingival diseases, were not recorded in detail.[7] Fifth, we did not control for the effect of 23 24 ongoing treatment for GORD patients. Thus it would have slightly under-estimated associations reported between 25 26 GORD & oral manifestations, but it would pose a serious threat to the associations found. Also, some oral side-effects 27 of prescription could not be ruled out (such as the effect of proton-pump inhibitors on xerostomia). Sixth, we could not 28 29 rule out the effect of other risk factors linked to xerostomia, such as: diabetes, hepatic disease, or a side effect (albeit 30 31 minor) of medication, including antihistamines, proton pump inhibitors, calcium channel blockers, and beta- 32 blockers.[5,13,45] Further trials and clinical studies are needed to rule out these potential confounding risk factors.

33 http://bmjopen.bmj.com/ 34 Future studies in humans should include biopsies of the oral epithelium to correlate clinical and histological findings. 35 36 Our present findings (from table 3) suggest that the Montreal consensus recommendations should now be expanded to 37 38 include aphthoid-ulcerative lesions, xerostomia, gingivitis, atrophic glossitis, angular cheilitis, oral submucosal 39 fibrosis, and candidiasis in the extra-oesophageal symptomatology of GORD.[2,44] 40

41 on October 1, 2021 by guest. Protected copyright. 42 43 Dental health is widely neglected in Pakistan, where dental-visits are restricted to a downstream approach (interventions 44 at micro-level).[47,53,54] Further, general practitioners and gastroenterologists are often the primary health care 45 46 providers for patients with GORD, but while addressing their main GI concerns, the oral manifestations of these 47 48 systemic conditions are often overlooked. This study highlights the need for dental referral in patients with upper GI 49 50 disorders (in this case GORD), which can have a marked effect on both systemic and oral health. 51 52 53 54 CONCLUSION 55 56 57 This study reinforces DE as strong comorbidity of chronic and a severe form of GORD. Patients with DE’s had severe 58 59 oral symptoms and relatively more compromised oral health than did those without DE. We found a positive correlation 60 between severe form of GORDFor peer and review oral only manifestations, - http://bmjopen.bmj.com/site/about/guidelines.xhtml such as xerostomia, aphthoid mucosal ulceration, gingivitis, BMJ Open Page 16 of 22 and angular cheilitis. Further, unhealthy dietary patterns and frequent nausea/vomiting increases the likelihood of BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 developing chronic GORD & DE, hence leading towards development of oral soft-tissue lesions, with increasing 2 severity, and compromised oral health. We recommend timely dental visits for evaluation of oral health, to assess the 3 4 degree of oral manifestations, and prevent its progression. We also urge gastroenterologists to perform routine oral 5 6 examinations and make referrals to a dentist to avoid worsening of the oral conditions caused by GORD. 7 8 Acknowledgments 9 10 11 The authors acknowledge the participating hospitals and respective gastroenterologists for supervising this research 12 13 and providing ethical and institutional/departmental permission to conduct this study, namely: Drs. Tayyab Saeed 14 Akhtar (Holy Family Hospital), Farzana Memon (CHK), Saeed Qureshi (CHK), Mashood (PIMS), Waseem Said 15 16 (PIMS), Waqar (JPMC), MasroorFor (JPMC). peer The study review would not have been only possible without the efforts of the following 17 18 clinicians who were trained and calibrated as the data collection team at four study centers: Drs. Emmad Naeem Khan, 19 Sumbul Zakir, Iqra Raja, Rimsha Hashmi, Javeria Ahmed, Qurratul-Ain, Ovais Khan Muhammad, Hafsa Mahida, 20 21 Yumna Shaheen Ali, Sajida Jhamro. We would also like to acknowledge the efforts of our study volunteers: Barha 22 23 Kabir, Tooba Mohtashim, Maryam Nazeer, Saad Aziz, Prem Shankar, Sham Kumar, Sanjay Kumar, and Orangzeb 24 25 Khatri. We would also like to thank Editage for English language editing. Also, thanks for the kind timely support from 26 http://www.phsrg.com for the upgrading of this manuscript. This research would not have been possible without the 27 28 valuable support and guidance of Professor Dr. Ambrina Qureshi (Chairperson community dentistry department, 29 30 DUHS). 31 32 Declaration of interests

33 http://bmjopen.bmj.com/ 34 35 All authors have completed the ICMJE uniform disclosure form at "http://www.icmje.org/coi_disclosure.pdf" and 36 37 declare no support from any organization for the submitted work; no financial relationships with any organizations that 38 might have an interest in the submitted work in the previous three years; no other relationships or activities that could 39 40 appear to have influenced the submitted work.

41 on October 1, 2021 by guest. Protected copyright. 42 43 Funding 44 45 None 46 47 48 Data sharing statement 49 50 All data relevant to this study are included in the article. No additional data available. 51 52 53 Author contributions 54 55 IW, study conception, data collection/analysis/interpretation, manuscript writing and editing; JA, data collection and 56 57 data entry/analysis; AY, study design/analysis, manuscript editing; AR, statistical analysis and proofreading; TA, 58 59 supervision of gastroenterology research (clinical supervisor) and proofreading; QA, data analysis, re-editing & peer 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml review (internal medicine expert); ZK, statistical interpretation and peer-review (dentistry expert). Page 17 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 License to Publisher 2 The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, a 3 4 worldwide license to the Publishers and its licensees in perpetuity in all forms. 5 6 7 8 REFERENCES 9 10 1 Patel A, Amaechi BT, Brady C. Prevention and Control of Dental Erosion: Gastroesophageal Reflux Disease 11 Management. In: Dental Erosion and Its Clinical Management. Cham: : Springer International Publishing 12 2015. 203–24. doi:10.1007/978-3-319-13993-7_12 13 14 2 Vakil N, van Zanten S V., Kahrilas P, et al. The Montreal Definition and Classification of Gastroesophageal 15 Reflux Disease: A Global Evidence-Based Consensus. The American Journal of Gastroenterology 16 For peer review only 17 2006;101:1900–20. doi:10.1111/j.1572-0241.2006.00630.x 18 19 3 Pace F, Pallotta S, Tonini M, et al. Systematic review: gastro-oesophageal reflux disease and dental lesions. 20 Alimentary Pharmacology & Therapeutics 2008;27:1179–86. doi:10.1111/j.1365-2036.2008.03694.x 21 22 4 Lazarchik D, Filler SJ. Effects of gastroesophageal reflux on the oral cavity. The American Journal of 23 medicine 1997;103:5:107-13 24 25 5 Meurman JH, Toskala J, Nuutinen P, et al. Oral and dental manifestations in gastroesophageal reflux disease. 26 Oral Surgery, Oral Medicine, Oral Pathology 1994;78:583–9. doi:10.1016/0030-4220(94)90168-6 27 28 6 Habesoglu TE, Habesoglu M, Sürmeli M, et al. Histological changes of rat soft palate with exposure to 29 experimental laryngopharyngeal reflux. Auris Nasus Larynx 2010;37:730–6. doi:10.1016/J.ANL.2010.03.009 30 31 7 Watanabe M, Nakatani E, Yoshikawa H, et al. Oral soft tissue disorders are associated with gastroesophageal 32 reflux disease: retrospective study. BMC Gastroenterology 2017;17:92. doi:10.1186/s12876-017-0650-5

33 http://bmjopen.bmj.com/ 34 8 Deppe H, Mücke T, Wagenpfeil S, et al. Erosive esophageal reflux vs. non erosive esophageal reflux: oral 35 findings in 71 patients. BMC Oral Health 2015;15:84. doi:10.1186/s12903-015-0069-8 36 37 9 Järvinen V, Meurman JH, Hyvärinen H, et al. Dental erosion and upper gastrointestinal disorders. Oral 38 39 Surgery, Oral Medicine, Oral Pathology 1988;65:298–303. doi:10.1016/0030-4220(88)90113-2 40 10 Friesen LR, Bohaty B, Onikul R, et al. Is histologic esophagitis associated with dental erosion: a cross-

41 on October 1, 2021 by guest. Protected copyright. 42 sectional observational study? BMC Oral Health 2017;17:116. doi:10.1186/s12903-017-0408-z 43 44 11 Yoshikawa H, Furuta K, Ueno M, et al. Oral symptoms including dental erosion in gastroesophageal reflux 45 disease are associated with decreased salivary flow volume and swallowing function. Journal of 46 Gastroenterology 2012;47:412–20. doi:10.1007/s00535-011-0515-6 47 48 12 Ranjitkar S, Smales RJ, Kaidonis JA. Oral manifestations of gastroesophageal reflux disease. Journal of 49 Gastroenterology and Hepatology 2012;27:21–7. doi:10.1111/j.1440-1746.2011.06945.x 50 51 13 Korsten MA, Rosman AS, Fishbein S, et al. Chronic xerostomia increases esophageal acid exposure and is 52 associated with esophageal injury. The American Journal of Medicine 1991;90:701–6. doi:10.1016/S0002- 53 9343(05)80058-0 54 55 14 Daley TD, Armstrong JE. Oral Manifestations of Gastrointestinal Diseases. Canadian Journal of 56 Gastroenterology 2007;21:241–4. doi:10.1155/2007/952673 57 58 15 Picos AM, Poenar S, Opris A, et al. Prevalence of dental erosions in GERD: a pilot study. Clujul medical 59 60 (1957) 2013;86:344–6.For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 22 16 Warsi I, Younus A, Rasheed A, et al. Oral health-related quality of life in patients with upper gastrointestinal

and hepatic disorders in Pakistan: validation of the Oral Health Impact Profile-14 in the Urdu language. BDJ BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 open 2018;4:17036. doi:10.1038/s41405-018-0002-8 2 3 17 Barron RP, Carmichael RP, Marcon MA, et al. Dental erosion in gastroesophageal reflux disease. Journal 4 Canadian Dental Association 2003;69:84–9. 5 6 18 Tripathi M, Streutker CJ, Marginean EC. Relevance of histology in the diagnosis of reflux esophagitis. Annals 7 of the New York Academy of Sciences 2018;1434:94–101. doi:10.1111/nyas.13742 8 9 19 Petersen P, Baez R. Oral Health Surveys Basic Methods 5th Edition. 5th ed. World Health Organization 2013. 10 http://www.icd.org/content/publications/WHO-Oral-Health-Surveys-Basic-Methods-5th-Edition-2013.pdf 11 12 (accessed 2 Jan 2019). 13 14 20 Silva MAGS, Damante JH, Stipp ACM, et al. Gastroesophageal reflux disease: New oral findings. Oral 15 Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology 2001;91:301–10. 16 doi:10.1067/moe.2001.111139For peer review only 17 18 21 Ranjitkar S, Kaidonis JA, Smales RJ. Gastroesophageal reflux disease and tooth erosion. International journal 19 of dentistry 2012;2012:479850. doi:10.1155/2012/479850 20 21 22 Milani DC, Venturini APC, Callegari-Jacques SM, et al. Gastro-oesophageal reflux disease and dental erosions 22 in adults. European Journal of Gastroenterology & Hepatology 2016;28:797–801. 23 doi:10.1097/MEG.0000000000000622 24 25 23 Roelen CA, Bültmann U, van Rhenen W, et al. External validation of two prediction models identifying 26 employees at risk of high sickness absence: cohort study with 1-year follow-up. BMC Public Health 27 28 2013;13:105. doi:10.1186/1471-2458-13-105 29 30 24 Vinesh E, Masthan K, Kumar MS, et al. A Clinicopathologic Study of Oral Changes in Gastroesophageal 31 Reflux Disease, Gastritis, and Ulcerative Colitis. The journal of contemporary dental practice 2016;17:943–7. 32 25 Petruzzi M, Lucchese A, Campus G, et al. Oral stigmatic lesions of gastroesophageal reflux disease (GERD). 33 http://bmjopen.bmj.com/ 34 Rev Med Chil 2012;140(7):915-18. 35 36 26 Roesch-Ramos L, Roesch-Dietlen F, María Remes-Troche J, et al. Dental erosion, an extraesophageal 37 manifestation of gastroesophageal reflux disease. The experience of a center for digestive physiology in 38 Southeastern Mexico. Rev Esp Enferm Dig 2014;106(2):92-7 39 40 27 Wang H-Y, Leena KB, Plymoth A, et al. Prevalence of gastro-esophageal reflux disease and its risk factors in

41 a community-based population in southern India. BMC Gastroenterology 2016;16:36. doi:10.1186/s12876- on October 1, 2021 by guest. Protected copyright. 42 016-0452-1 43 44 28 Javadzadeh F, Rafeey M. 715 Dental Erosion and Gastroesophageal Reflux Disease (GERD) in Children. 45 46 Archives of Disease in Childhood 2012;97:A206–A206. doi:10.1136/archdischild-2012-302724.0715 47 48 29 Ness-Jensen E, Lagergren J. Tobacco smoking, alcohol consumption and gastro-oesophageal reflux disease. 49 Best Practice & Research Clinical Gastroenterology 2017;31:501–8. doi:10.1016/J.BPG.2017.09.004 50 51 30 Mantonanaki M, Koletsi-Kounari H, Mamai-Homata E, et al. Dental erosion prevalence and associated risk 52 indicators among preschool children in Athens, Greece. Clinical Oral Investigations 2013;17:585–93. 53 doi:10.1007/s00784-012-0730-4 54 55 31 Meurman JH, Rantonen P. Salivary flow rate, buffering capacity, and yeast counts in 187 consecutive adult 56 patients from Kuopio, Finland. European Journal of Oral Sciences 1994;102:229–34. doi:10.1111/j.1600- 57 0722.1994.tb01185.x 58 59 32 Kongara K, VarilekFor G, peer Soffer review EE. only Salivary - http://bmjopen.bmj.com/site/about/guidelines.xhtml Growth Factors and Cytokines Are Not Deficient in Patients with 60 Gastroesophageal Reflux Disease or Barrett’s Esophagus. Digestive Diseases and Sciences 2001;46:606–9. Page 19 of 22 BMJ Open doi:10.1023/A:1005615703009 BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 33 Rourk RM, Namiot Z, Edmunds MC, et al. Diminished luminal release of esophageal epidermal growth factor 2 in patients with reflux esophagitis. The American journal of gastroenterology 1994;89:1177–84. 3 4 34 Kao CH, Ho YJ, ChangLai SP, et al. Evidence for decreased salivary function in patients with reflux 5 esophagitis. Digestion 1999;60:191–5. doi:10.1159/000007658 6 7 35 Ahmad M, Bhayat A, Zafar M, et al. The Impact of Hyposalivation on Quality of Life (QoL) and Oral Health 8 in the Aging Population of Al Madinah Al Munawarrah. International Journal of Environmental Research and 9 Public Health 2017;14:445. doi:10.3390/ijerph14040445 10 11 36 Jarvinen VK, Rytomaa II, Heinonen OP. Risk Factors in Dental Erosion. Journal of Dental Research 12 1991;70:942–7. doi:10.1177/00220345910700060601 13 14 37 Di Fede O, Di Liberto C, Occhipinti G, et al. Oral manifestations in patients with gastro-oesophageal reflux 15 16 disease: a single-centerFor case-control peer study. Journal review of Oral Pathology only & Medicine 2008;37:336–40. 17 doi:10.1111/j.1600-0714.2008.00646.x 18 19 38 Rajalalitha P, Vali S. Molecular pathogenesis of oral submucous fibrosis - a collagen metabolic disorder. 20 Journal of Oral Pathology and Medicine 2005;34:321–8. doi:10.1111/j.1600-0714.2005.00325.x 21 22 39 Pindborg JJ, Sirsat SM. Oral submucous fibrosis. Oral Surgery, Oral Medicine, Oral Pathology 1966;22:764– 23 79. doi:10.1016/0030-4220(66)90367-7 24 25 40 Hogaboam CM, Steinhauser ML, Chensue SW, et al. Novel roles for chemokines and fibroblasts in interstitial 26 fibrosis. Kidney International 1998;54:2152–9. doi:10.1046/J.1523-1755.1998.00176.X 27 28 41 Scott DR, Simon RA. Supraesophageal reflux disease: A review of the literature. Allergy and Asthma 29 Proceedings 2014;35:104–10. doi:10.2500/aap.2014.35.3719 30 31 42 Khan S, Prashanth S, Rao P, et al. Pathogenesis of oral submucous fibrosis. Journal of Cancer Research and 32 Therapeutics 2012;8:199. doi:10.4103/0973-1482.98970

33 http://bmjopen.bmj.com/ 34 43 Ismail-Beigi F, Horton PF, Pope CE. Histological Consequences of Gastroesophageal Reflux in Man. 35 36 Gastroenterology 1970;58:163–74. doi:10.1016/S0016-5085(70)80004-X 37 38 44 Pauwels A. Dental erosions and other extra-oesophageal symptoms of gastro-oesophageal reflux disease: 39 Evidence, treatment response and areas of uncertainty. United European Gastroenterology Journal 40 2015;3:166–70. doi:10.1177/2050640615575972

41 on October 1, 2021 by guest. Protected copyright. 42 45 Sreebny LM, Valdini A. Xerostomia. Archives of Internal Medicine 1987;147:1333. 43 doi:10.1001/archinte.1987.00370070145022 44 45 46 Corrêa MCCSF, Lerco MM, Cunha M de LR de S da, et al. Salivary parameters and teeth erosions in patients 46 with gastroesophageal reflux disease. Arquivos de Gastroenterologia 2012;49:214–8. doi:10.1590/S0004- 47 28032012000300009 48 49 47 Niaz M, Naseem M, Siddiqui S, et al. An outline of the oral health challenges in “Pakistani” population and a 50 discussion of approaches to these challenges. J Pak Dent Assoc 2013;22. 51 52 48 Stephan RM. Intra-Oral Hydrogen-Ion Concentrations Associated With Dental Caries Activity. Journal of 53 54 Dental Research 1944;23:257–66. doi:10.1177/00220345440230040401 55 56 49 Lussi A, Jaeggi T, Zero D. The role of diet in the aetiology of dental erosion. Caries research 2004;38 Suppl 57 1:34–44. doi:10.1159/000074360 58 59 50 Chakraborti S, Pruthi S, Nirupama M. Evaluation of gastric biopsies in chronic gastritis: Grading of 60 inflammation by VisualFor peer Analogue review only Scale. - http://bmjopen.bmj.com/site/about/guidelines.xhtml Medical Journal of Dr DY Patil University 2014;7:463. doi:10.4103/0975-2870.135268 BMJ Open Page 20 of 22 51 Al-Saadi AM, Al-Khayat JQ, Muhammad IM, et al. The role of Helicobacter pylori in esophagitis and peptic

ulcer disease in Iraq. Saudi Med J 2004;25:1216–22.www.smj.org.sa (accessed 2 Jan 2019). BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 1 2 52 Schechter RB, Lemme EMO, Coelho HSM. Gastroesophageal reflux in cirrhotic patients with esophageal 3 varices without endoscopic treatment. Arquivos de Gastroenterologia 2007;44:145–50. doi:10.1590/S0004- 4 28032007000200012 5 6 53 Harchandani N. Oral health challenges in Pakistan and approaches to these problems. Pakistan Oral & Dental 7 Journal 2012;32:497–501. 8 9 54 AMA. AMA Position Statement Social Determinants of Health and the Prevention of Health Inequities 2007. 10 2007. 11 12 https://ama.com.au/system/tdf/documents/AMA_Position_Statement_on_the_Social_Determinants_of_Health 13 _and_the_Prevention_of_Health_Inequities_2007_0.pdf?file=1&type=node&id=40625 (accessed 2 Jan 2019). 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

1 2 3 Risk factors, associated oral manifestations, and oral health impact of

4 gastro-oesophageal reflux disease: A multicentre, cross-sectional study BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 5 6 in Pakistan 7 8 GORD-study STROBE checklist of items, that should be included in reports of cross-sectional studies. 9 10 Items ticked in ‘blue’ () are included in manuscript that has been submitted, while items not ticked are not 11 applicable for our study. 12 Item 13 No Recommendation 14 15 Title and abstract  1  (a) Indicate the study’s design with a commonly used term in the title or the 16 abstract 17  (b) Provide in the abstract an informative and balanced summary of what was 18 Fordone peer and what was review found only 19 20 Introduction 21 Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 22 Objectives 3 State specific objectives, including any prespecified hypotheses 23 24 Methods 25 Study design 4 Present key elements of study design early in the paper 26 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, 27 28 exposure, follow-up, and data collection 29 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of 30 participants 31 32 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect 33 modifiers. Give diagnostic criteria, if applicable 34 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 35 measurement assessment (measurement). Describe comparability of assessment methods if there 36 is more than one group

37 http://bmjopen.bmj.com/ 38 Bias 9 Describe any efforts to address potential sources of bias 39 Study size 10 Explain how the study size was arrived at 40 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, 41 42 describe which groupings were chosen and why 43 Statistical methods 12  (a) Describe all statistical methods, including those used to control for 44 confounding 45 on October 1, 2021 by guest. Protected copyright. 46  (b) Describe any methods used to examine subgroups and interactions 47 (c) Explain how missing data were addressed 48 (d) If applicable, describe analytical methods taking account of sampling strategy 49 (e) Describe any sensitivity analyses 50 51 Results 52 Participants 13*  (a) Report numbers of individuals at each stage of study—eg numbers potentially 53 eligible, examined for eligibility, confirmed eligible, included in the study, 54 55 completing follow-up, and analysed 56  (b) Give reasons for non-participation at each stage 57 (c) Consider use of a flow diagram 58 59 Descriptive data 14*  (a) Give characteristics of study participants (eg demographic, clinical, social) 60 and information on exposures and potential confounders (b) Indicate number of participants with missing data for each variable of interest

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 22

1 2 Outcome data 15* Report numbers of outcome events or summary measures 3 Main results 16  (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates

4 BMJ Open: first published as 10.1136/bmjopen-2017-021458 on 30 March 2019. Downloaded from 5 and their precision (eg, 95% confidence interval). Make clear which confounders 6 were adjusted for and why they were included 7  (b) Report category boundaries when continuous variables were categorized 8 (c) If relevant, consider translating estimates of relative risk into absolute risk for a 9 10 meaningful time period 11 Other analyses  17  Report other analyses done—eg analyses of subgroups and interactions, and 12 sensitivity analyses 13 14  Discussion 15 Key results  18  Summarise key results with reference to study objectives 16 Limitations  19  Discuss limitations of the study, taking into account sources of potential bias or 17 18 Forimprecision. peer Discuss review both direction and only magnitude of any potential bias 19 Interpretation  20  Give a cautious overall interpretation of results considering objectives, 20 limitations, multiplicity of analyses, results from similar studies, and other relevant 21 evidence 22 23 Generalisability  21  Discuss the generalisability (external validity) of the study results 24  Other information 25 Funding 22 Give the source of funding and the role of the funders for the present study and, if 26 27 applicable, for the original study on which the present article is based 28 29 30 I declare that the checklist maintained in reference to the submitted original research article, is correct to the best of my 31 32 knowledge. For any concerns, you can direct them to me directly. 33 34 Thank you for your consideration. I look forward to hearing from you. 35 36 Sincerely,

37 http://bmjopen.bmj.com/ 38 39 Dr. Ahmed Ibrahim Warsi 40 BDS, Implantologist, MMSCI (HMS) 41 42 Research Fellow | Forsyth Institute 43 Harvard Medical School 44 Email: [email protected] 45 Contact: (+1)832-458-5039 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml