View larger version (111K): [in this window] [in a new window]   Figure 1. Dental erosion and dental caries. Top. Extensive loss of enamel with exposed underlying dentin (yellow) characteristic of dental erosion (arrows). Bottom. Dental caries. Soft, discolored, irregular, carious lesions on interproximal surfaces of maxillary teeth (arrow). Note thick, edematous, reddened gingival margins secondary to plaque accumulation.

View larger version (100K): [in this window] [in a new window]   Figure 2. Dental erosion. Mandibular facial surfaces showing smooth, glistening, yellow, "dished-out" erosion (arrows).

Both extrinsic and intrinsic factors contribute to dental erosion. Extrinsic factors include acidic foodstuffs, beverages, and snacks and exposure to acidic contaminants in the workplace [3-7]. The pattern of damage depends on the exposure. For example, frequent ingestion of citrus fruits such as lemons has been associated with erosion of the facial surface of the anterior teeth. Exposure to aerosolized industrial acids in the workplace (as in lead battery factories) may result in the erosion of surfaces exposed to the atmosphere [8, 9]. There have also been anecdotal reports of swimmers developing dental erosion from improperly maintained water in gas-chlorinated swimming pools [10].

Intrinsic factors that have been recognized to contribute to dental erosion include impaired salivary production [11, 12], altered buffering capacity [13], and frequent vomiting of gastric contents [14-19]. Bulimia nervosa is a common problem that affects between 4% and 10% of adolescent and college-aged women [20, 21]. Jones and Cleaton-Jones [22] found dental erosion in 69% of bulimic patients and postulated that it was caused by the exposure of enamel and dentin to acidic gastric contents. Erosions were seen on the lingual surfaces of maxillary teeth in 50% of their patients and on the facial surfaces of the central incisors and canines in 70% of their patients. A similar pattern of tooth loss has been reported in healthy patients with radiographic evidence of hiatal hernia [23-25]. This has been anecdotally attributed to gastroesophageal reflux disease; however, physiologic studies evaluating the relation between gastroesophageal reflux and dental erosion are lacking. Because of the frequency with which gastroesophageal reflux occurs in the general population (7% have it daily and 36% have it at least once monthly [26]) and the potentially detrimental effects of acid reflux into the oral cavity, we evaluated the relation between gastroesophageal reflux disease and idiopathic dental erosion using state-of-the-art ambulatory esophageal pH testing.

Methods

Top Methods Results Discussion Author & Article Info References

Study Samples

Dental Group

Patients with idiopathic dental erosion were identified by two faculty members at the University of Alabama School of Dentistry. Patients presenting with a history suggestive of bulimia were excluded from the study. Seventeen patients with dental erosion were identified over an 18-month period. Five of these patients elected not to participate in the study when nasal intubation with an esophageal pH electrode was discussed.

Gastroenterology Group

Patients with suspected reflux disease who had recently had ambulatory 24-hour esophageal pH monitoring at the University of Alabama at Birmingham were offered a free, complete dental examination. This group of patients was divided into those without reflux (those with normal results on 24-hour esophageal pH testing), those with abnormal amounts of distal gastroesophageal reflux, and those with abnormal amounts of proximal gastroesophageal reflux. The first 10 patients in each of these subgroups who agreed to the dental evaluation were included in our analysis. Dental examinations were done as described below; both dentists were blinded to the results of the pH testing.

Questionnaire

Patients in both study groups completed a questionnaire designed to identify possible causes of erosion. They were asked about occupational exposures, frequency of dental examinations, bruxism, brushing and flossing habits, use of fluoridated water, dental fluoride treatments, and the frequency and quantity of consumption of sweets, citrus fruits, and soft drinks. Excessive intake was arbitrarily defined as more than five soft drinks per day or more than five servings of fruit per day. Patients in the dental group answered a second questionnaire about the presence and frequency of typical (heartburn, regurgitation, chest pain, and dysphagia) and atypical gastroesophageal reflux symptoms (coughing, wheezing, hoarseness, and asthma) and medications used to control these symptoms.

Dental Examination

All patients had a thorough oral examination done by two dental school faculty members. The first five patients were evaluated by both dentists to ensure similar scoring. Each of three surfaces (facial, occlusal, and lingual) of every tooth, excluding the wisdom teeth, was given an individual score on the erosion scale described by Eccles and Jenkins [27] (Table 1). A cumulative score per patient and individual average scores for the lingual, occlusal, and facial surfaces were determined. Two standard scoring systems that reflected tooth loss and tooth surface restoration, some of which may be secondary to acid reflux, were determined: the number of decayed, missing, or filled teeth and the number of decayed, missing, or filled tooth surfaces. Additionally, a plaque index [28] and a gingival index [20] Table 1 were calculated on the basis of each of the three surfaces of six teeth. The plaque index is an indicator of oral hygiene; the gingival index is an indicator of periodontal or soft tissue disease. Both indexes tend to be abnormal in patients with dental caries.

Saliva Analysis

As part of the dental examination, saliva from the patients in the dental group was collected by expectoration over a 10-minute period in the basal state and after stimulation by chewing on paraffin [30]. Samples were immediately analyzed for buffering capacity using the Dentobuff Strip (Orion Diagnostica, Finland) and for pH using the Corning pH meter (Corning, New York). They were then frozen at –20°C until analyzed for calcium and phosphate levels. Total calcium levels were determined using spectrophotometry after chromophore formation using o-cresolpthalein complexone [31]. The assay of inorganic phosphate levels relied on the formation of a complex of phosphate ion using a molybdate compound [32]. Saliva samples from 10 age- and sex-matched persons without dental erosion served as controls.

Ambulatory Esophageal pH Study

Patients discontinued therapy with histamine-2 antagonists for 3 days and proton pump inhibitors for 7 days before esophageal studies were done. Esophageal manometry was done using nasal intubation with a round, polyvinyl catheter (diameter, 4.5 mm [Arndorfer Specialties, Inc., Greendale, Wisconsin]) continuously perfused with distilled water by a low-compliance, pneumohydraulic capillary-infusion system (Arndorfer Specialties, Inc.). The location and length of the lower and upper esophageal sphincters were determined by the station pull-through technique [33]. Immediately after esophageal manometry, 24-hour esophageal pH monitoring was done using a monocrystalline catheter (diameter, 2.5 mm) with two antimony pH electrodes (Synectics, Irving, Texas) calibrated at pH 7.0 and pH 1.0 before onset and at the completion of each study. The distal pH electrode was placed 5 cm above the lower esophageal sphincter; the proximal electrode was placed just distal to the upper esophageal sphincter. A reference electrode was placed on the anterior chest wall. Both the distal and proximal electrodes were connected to a digital data logger (Synectics) that stores pH data every 4 seconds for 24 hours. Patients were sent home with instructions to record meal times, time of assuming the supine posture for sleep, and time of arising in the morning. They were encouraged to do normal daily activities without restrictions on diet or smoking. Patients returned the following morning for removal of the pH catheter. Abnormal distal esophageal 24-hour acid exposure times were defined as those exceeding the 95th percentile values for either percentage of time that upright pH is less than 4.0 (8.3%) or the percentage of time that supine pH is less than 4.0 (3.0%), which were obtained from previous studies in 110 healthy volunteers [34]. Similarly, proximal esophageal pH data were considered abnormal if they exceeded the 95th percentile values for either percentage of time that upright pH is less than 4.0 (1.7%) or the percentage of time that supine pH is less than 4.0 (0.6%); these percentile values were obtained from a study of 20 healthy volunteers [35].

Statistical Analysis

Dental examination scores (erosion score; decayed, missing, or filled teeth index; decayed, missing, or filled surfaces index; and plaque and gingival indices) are presented as means with 95% CIs. Acid reflux variables are presented as means and ranges of values. Saliva data are presented as means and either ranges or 95% CIs. The Fisher exact test was used to compare the prevalence of dental erosion among the gastroenterology subgroups.

After the dental and esophageal pH data were accumulated, all patients from both groups who were found to have dental erosion were combined so that the correlation between dental lesions and acid reflux variables could be assessed. Data from the proximal and distal esophageal electrodes were calculated separately. Two erosion ratios were calculated by dividing the cumulative erosion score by 1) the number of existing tooth surfaces and 2) the number of individual tooth surfaces affected by some degree of erosion. Both distal and proximal acid exposure times in the upright and supine positions were correlated with the cumulative erosion score; the two erosion ratios; the decayed, missing, or filled teeth index; the decayed, missing, or filled surfaces index; the plaque index; and the gingival index. A P value of less than 0.05 was considered significant. Our a priori hypothesis was that proximal acid exposure times, as measures of microaspiration, would correlate with erosion scores.

Results

Top Methods Results Discussion Author & Article Info References

Dental Patients

Twelve dental patients (6 men and 6 women; mean age, 54 years) agreed to participate in the study. By definition, all had dental erosion involving, on average, 82% of their teeth and 66% of their tooth surfaces, as detailed in Table 2. Furthermore, nearly one half of the affected tooth surfaces had erosions involving the dentin (Eccles and Jenkins grades 2 or 3) (Figure 3). The mean erosion scores for the tooth surfaces were 37.5 for occlusal surfaces, 31.7 for lingual surfaces, and 29.0 for facial surfaces. The mean decayed, missing, or filled teeth and decayed, missing, or filled surfaces index scores were 14.7 (95% CI, 2.1 to 27.3) and 57.3 (CI, 0.0 to 128.4), respectively. The mean plaque index score was 0.7 (CI, 0.0 to 1.9), and the mean gingival index score was 1.1 (CI, 0.1 to 2.1). None of the patients had cavities detected by observation and transillumination.

View larger version (22K): [in this window] [in a new window]   Figure 3. Presence and severity of erosion. Percentage of tooth surfaces affected by erosion and severity according to the Eccles and Jenkins scale in the dental group and in the gastroenterology subgroups. In patients with distal and proximal reflux on 24-hour esophageal pH monitoring, 21% and 29% of tooth surfaces, respectively, were affected by erosion. In the single patient without reflux who had erosion, only 5% of tooth surfaces were affected.

Ten of the 12 patients (83%) had abnormal amounts of gastroesophageal reflux on 24-hour esophageal pH testing. Nine patients (75%) had abnormal distal reflux values. Of these 9 patients, 8 had increased upright reflux values ranging from 8.3% to 32.0% (mean, 14.0% [normal, <8.3%]) and 3 had increased supine reflux values ranging from 8.7% to 13.1% (normal, <3.0%). Seven patients (58%) had abnormal proximal reflux values; of these, 6 had increased upright reflux values that ranged from 2.2% to 13.6% (mean, 3.6% [normal, <1.7%]), and 3 had increased supine reflux values that ranged from 0.7% to 7.2% (normal, <0.6%).

Dental histories did not indicate any unusual or excessive dietary habits or occupational exposures associated with erosion formation. Ten of the 12 patients (83%) had typical gastroesophageal reflux symptoms. Six patients had heartburn several times a day, 1 had it several times a week, and one had it at least once a week. Regurgitation was reported by 9 patients; the frequency ranged from once daily to several times a week. Two patients did not have heartburn or regurgitation. Seven patients (58%) reported atypical gastroesophageal reflux symptoms. Three had chronic cough, 5 reported chronic hoarseness, and 3 had a history of asthma or bronchitis. Only 1 patient was free of both typical and atypical gastroesophageal reflux symptoms. Eight patients (67%) were taking antacids, histamine-2 antagonists, or omeprazole.

One patient had symptomatic xerostomia and failed to produce any saliva during the test period; another did not complete the saliva collection after erosions were detected. Saliva studies were done in the other 10 patients. Unstimulated salivary flow rates were normal in all patients; the mean flow rate was 0.46 mL/min (range, 0.3 mL/min to 1.1 mL/min [normal, > 0.25 mL/min]) [36]. Stimulated flow rates were normal in all patients but 1; the mean flow rate was 1.35 mL/min (range, 0.3 mL/min to 2.0 mL/min [normal, > 0.7 mL/min]) [30, 36]. Unstimulated salivary pH was normal in all patients (mean, 6.97 [range, 6.40 to 7.48]; normal pH, > 6.0). The salivary pH appropriately increased with stimulation in all patients but 1. There were no abnormalities of salivary buffering capacity. The mean salivary calcium ion concentration was actually higher in the age- and sex-matched controls (2.02 mg/dL [CI, 0.63 to 3.41 mg/dL]) than in the dental erosion group (0.97 mg/dL [CI, 0.0 to 2.00 mg/dL]; P < 0.001). The mean salivary inorganic phosphate concentration in patients with dental erosion (14.58 mg/dL [CI, 4.51 to 24.65 mg/dL]) did not differ from that in controls (13.96 mg/dL [CI, 9.69 to 18.23 mg/dL]; P = 0.59).

Frequency of Dental Erosion in the Gastroenterology Group

Of the 10 patients without reflux (9 women and 1 man; mean age, 44 years) who had normal proximal and distal acid exposure times, only 1 (10%) had dental erosion. Of 10 patients (6 women and 4 men; mean age, 49 years) with abnormal distal acid exposure times, 4 (40%) had erosion. Of 10 patients (7 women and 3 men; mean age, 54 years) with both proximal and distal gastroesophageal reflux, 7 (70%) had dental erosion. Thus, the patients with reflux (55% [11 of 20 patients]) had a higher prevalence of dental erosion than the patients without reflux (10% [1 of 10 patients]; P = 0.02, Fisher exact test). In addition, all 10 patients with proximal reflux had upright reflux. Five of the 10 patients with distal reflux had upright reflux.

The frequency of severe erosion (Eccles and Jenkins grade 2 and 3) was compared among the three gastroenterology subgroups. The patients with proximal reflux had 7% of their teeth affected and the patients with distal reflux had 3% of their teeth affected. The one participant without reflux who had dental erosion had only grade 1 lesions (Figure 3). Erosion scores in the gastroenterology group for the lingual (mean, 17.0 [CI, 0.2 to 33.8]) and occlusal surfaces (mean, 17.7 [CI, 0.0 to 39.3]) were higher than those for the facial surfaces (mean, 8.3 [CI, 0.2 to 30.3]; P < 0.05). Decayed, missing, and filled surfaces and decayed, missing, and filled teeth index scores and the plaque and gingival index scores did not differ statistically between the gastroenterology and dental groups (Table 2). However, 5 of 30 patients (16.7%) in the gastroenterology group had surfaces with caries, but only 3 of these patients had associated dental erosion. In addition, the questionnaires answered by the dental group did not show any unusual occupational exposures or excessive dietary habits associated with dental erosion formation.

Correlation Analysis

Patients with dental erosion in the dental group (n = 12) and the gastroenterology group (n = 12) were combined for a correlation analysis of pH values compared with erosion scores; erosion ratios; decayed, missing, or filled teeth indexes; decayed, missing, or filled surfaces indexes; and plaque and gingival indexes. The ratios of erosion score to existing surfaces and erosion score to affected surfaces correlated strongly with the cumulative erosion score (r = 0.99 [P < 0.001] and r = 0.75 [P < 0.001], respectively). Thus, the erosion score was the only variable used for reporting the results of subsequent analysis.

Proximal upright reflux correlated with the cumulative erosion score (r = 0.55 [P < 0.01]) Figure 4, left). A subgroup of patients (n = 12) with abnormal amounts of proximal upright reflux (their pH was less than 4.0 more than 1.7% of the time) showed an even stronger correlation with the erosion score (r = 0.84 [P < 0.001]) Figure 4, right). No other reflux variables correlated with the erosion score.

View larger version (14K): [in this window] [in a new window]   Figure 4. Correlation of erosion scores and proximal upright reflux. Relation between cumulative erosion scores and percentage of time that proximal upright pH was less than 4.0 in all patients with dental erosion (left) and in the subgroup (n = 12) with abnormal amounts of proximal upright reflux (right). Amounts of proximal upright reflux were defined as abnormal if the percentage of time that pH was less than 4.0 was more than 1.7%. Linear regression analysis found a relation (P < 0.01 and P < 0.001, respectively) between erosion and proximal reflux in both groups.

No correlation was found between the various pH values and the indices of decayed, missing, and filled teeth or decayed, missing, and filled surfaces. Similarly, no correlation was found between the measures of dental hygiene (plaque index and gingival index) and gastroesophageal reflux. There were no correlations between erosion scores and salivary flow rates, salivary pH, or buffering capacity.

Discussion

Top Methods Results Discussion Author & Article Info References

Ours is the first study to establish a relation between dental erosion and gastroesophageal reflux disease using state-of-the-art ambulatory pH monitoring. Previous studies evaluating this relation have been limited to case reports or case studies in which gastroesophageal reflux was inferred as a cause on the basis of the presence of hiatal hernia or esophagitis [14, 15, 23-25]. We found abnormal amounts of gastroesophageal reflux in 83% of patients with dental erosion who were referred by dentists for pH testing. The relation between reflux and erosion was confirmed by studying a general sample of persons with reflux symptoms in which 40% (4 of 10) of patients with distal reflux and 70% (7 of 10) of patients with proximal reflux, compared with only 10% (1 of 10) of patients without reflux (P = 0.02), were found to have dental erosion.

According to their histories, no dental patient in our study had a vomiting disorder. The mean age (54 years) and the sex distribution of the group would be unusual for a bulimic population. Dietary histories did not indicate that excessive consumption of acidic foods or beverages were possible extrinsic causes of dental erosion. Salivary flow rates, pH, buffering capacity, and calcium and phosphate levels were normal in nearly all patients, confirming the results of previous studies [30, 36, 37]. Therefore, although dental erosion is probably a multifactorial process, acid reflux appears to be the primary factor in the development of dental erosion in these patients.

The reported prevalence of dental erosion in the general population varies considerably. In a survey of 1345 male students in the United States, Pindborg [1] found a 2.1% prevalence of erosion. In contrast, Sognnaes and colleagues [38] observed that 18% of a random sample of 10 000 extracted teeth had some form of erosion. More recently, Lussi and colleagues [39] found that as many as 42.6% of Swiss adults 46 to 50 years of age had at least one severe erosion on an occlusal surface and had 3.9 teeth affected by erosion. Jarvinen and colleagues [18] studied the prevalence of erosion in 109 patients with gastrointestinal symptoms and found that 7 of 35 patients (20%) with reflux esophagitis or duodenal ulcer had dental erosion. We found the prevalence to be even higher in a representative sample of patients with reflux defined by pH testing: Eleven of 20 patients (55%) had dental erosion.

Overall, erosion scores were higher in the dental group than in the gastroenterology group. Similarly, as indicated by the Eccles and Jenkins scale, erosion was generally more severe in the dental group (Figure 3). These differences are probably due to selection biases because the dentists initially saw in their clinics the more severe and obvious cases of dental erosion. Both the dental and the gastroenterology groups had a pattern of higher erosion scores on the lingual and occlusal surfaces of the teeth, confirming the results of previous studies [39, 40]. This pattern reached statistical significance in the gastroenterology group (P < 0.05). Overall, oral hygiene was good and there was little periodontal disease in the two groups. Furthermore, there were surprisingly few dental caries lesions: None were found in the dental group, and only five were found in the gastroenterology group. This raises the intriguing possibility that acid reflux may play a role in preventing the formation of dental caries; it probably does so by inhibiting bacterial growth in the mouth.

The probable mechanism of damage to the dentition is gastroesophageal reflux high into the proximal esophagus spilling across the upper esophageal sphincter and into the oral cavity; this mechanism is similar to the process suspected to operate in ear, nose, and throat and asthma presentations of gastroesophageal reflux disease [41-47]. That this mechanism is operative is supported by the absence of other intrinsic or extrinsic factors and by our correlation studies, which show a strong relation between the severity of dental erosion and the degree of proximal upright reflux. Proximal supine reflux, however, did not correlate with erosion scores, even though upper esophageal sphincter pressures were generally lowest during periods of sleep [48]. This may represent a type II error due to the small number (n = 7) of patients with proximal supine reflux studied, because a trend (r = 0.31 [P = 0.15]) was evident even with these numbers.

Several possible limitations of our study should be noted. Not all patients in the dental group had 24-hour esophageal pH testing. A selection bias toward patients with reflux may have occurred. However, this is unlikely because 1) patients refused the test uniformly because they did not like the idea of nasal intubation; 2) reflux symptoms alone were unlikely to determine whether patients were tested because two patients were asymptomatic; and 3) even if all study dropouts had normal amounts of gastroesophageal reflux, the prevalence rate for acid reflux disease would still be unusually high at 59% (10 of 17 patients). Larger numbers of patients and groups of patients with more severe gastroesophageal reflux disease (for example, patients with esophagitis who have antireflux surgery) need to be studied to better define the prevalence of this process. Despite the small size of our sample, a type I error for the overall relation we found between dental erosion and proximal upright reflux is unlikely. A study with a larger sample will probably also find a relation between dental erosion and proximal supine reflux. Unfortunately, testing the hypothesis that excessive amounts of proximal esophageal acid reflux can cause dental erosion will be difficult. Aggressive acid suppression can resolve both classic reflux symptoms and atypical presentations such as chest pain, asthma, hoarseness, and coughing. However, acid suppression should not cause regression of dental erosion lesions, although further damage and tooth loss can, we hope, be prevented with early detection.

Our results indicate that dental erosion is a common physical finding in patients with gastroesophageal reflux disease and that it should now be considered an atypical manifestation of this disease. The high prevalence of reflux disease [26] in the general population suggests that this could be a frequently overlooked problem. Dental erosion can be diagnosed not only by dentists but by primary care physicians and gastroenterologists who are familiar with its physical characteristics. Early diagnosis and suppression of refluxed acid through lifestyle changes and medications could potentially prevent further damage and tooth loss. For patients with signs of erosion, a dental referral is appropriate for evaluation and restoration of lost tooth structure and institution of preventive dental measures. Thus, not only is it important for the dentist to be familiar with and inquire about typical and atypical reflux symptoms, but the primary care physician and the gastroenterologist need to pay more attention to the often neglected oral examination.