Relationship Between Disease Characteristics and Orofacial Manifestations in Systemic Sclerosis: Canadian Systemic Sclerosis Oral Health Study III

Abstract

Objective

Systemic sclerosis (SSc; scleroderma) is associated with decreased saliva production and interincisal distance, more missing teeth, and periodontal disease. We undertook this study to determine the clinical correlates of SSc with these oral abnormalities.

Methods

Subjects were recruited from the Canadian Scleroderma Research Group cohort. Detailed dental and clinical examinations were performed according to standardized protocols. Associations between dental abnormalities and selected clinical and serologic manifestations of SSc were examined.

Results

One hundred sixty-three SSc subjects were included: 90% women, mean ± SD age 56 ± 11 years, mean ± SD disease duration 14 ± 8 years, 72% with limited cutaneous disease, and 28% with diffuse cutaneous disease. Decreased saliva production was associated with Sjögren’s syndrome–related autoantibodies (β = −43.32; 95% confidence interval [95% CI] −80.89, −5.75), but not with disease severity (β = −2.51; 95% CI −8.75, 3.73). Decreased interincisal distance was related to disease severity (β = −1.02; 95% CI −1.63, −0.42) and the modified Rodnan skin thickness score (β = −0.38; 95% CI −0.53, −0.23). The number of missing teeth was associated with decreased saliva production (relative risk [RR] 0.97; 95% CI 0.94, 0.99), worse hand function (RR 1.52; 95% CI 1.13, 2.02), and the presence of gastroesophageal reflux disease (GERD; RR 1.68 [95% CI 1.14, 2.46]). No clinical or serologic variables were correlated with periodontal disease.

Conclusion

In SSc, diminished interincisal distance is related to overall disease severity. Decreased saliva production is related to concomitant Sjögren’s syndrome antibodies. Tooth loss is associated with poor upper extremity function, GERD, and decreased saliva. The etiology of excess periodontal disease is likely multifactorial and remains unclear.

INTRODUCTION

Oral abnormalities are common in systemic sclerosis (SSc; scleroderma) (1). Our first report from a study of 163 SSc patients and 231 controls demonstrated that SSc patients have less saliva production, smaller interincisal distances, more missing teeth, and more periodontal disease than controls (2). The aim of the present study was to determine which clinical and serologic aspects of SSc are associated with these abnormalities. We hypothesized a priori that 1) because Sjögren’s syndrome has been associated with SSc (3–9), saliva production would be associated with the presence of Sjögren’s syndrome–related antibodies; 2) because limitation of mouth opening is probably related to fibrosis of periorbital soft tissue, interincisal distance would be associated with the extent of skin involvement and global disease severity; 3) because tooth loss previously has been associated with dry mouth (10) and periodontal disease, missing teeth in SSc would be associated with decreased saliva and periodontal disease; 4) because tooth loss due to decay might be related to poor brushing resulting from decreased interincisal distance and/or hand contractures, missing teeth would be associated with interincisal distance and measures of hand function; 5) because there has been a suggestion that tooth loss may be related to gastroesophageal reflux disease (GERD), missing teeth in SSc would be associated with GERD; 6) because it is possible that tooth loss in SSc is a direct effect of the disease process on the periodontal membrane, missing teeth would be associated with increased severity of SSc; and 7) periodontal disease may be associated with decreased saliva production, tooth decay due to poor oral hygiene (resulting from decreased interincisal distance and/or hand contractures), or increased severity of SSc.

SUBJECTS AND METHODS

Study design and subjects

The study design and subjects have been previously described in detail (2). In brief, this multisite, cross-sectional study was conducted between 2008 and 2011 at a subset of the Canadian Scleroderma Research Group (CSRG) sites. Control patients consulting for mechanical joint problems were of same sex and similar age and were recruited from the same sites (2). The research ethics board of each participating center approved the study and all study subjects provided informed consent in compliance with the Helsinki Declaration.

Study measures

Information regarding sex, age, ethnicity, education, and smoking status was obtained by patient self-report. Medication use was recorded by the study physicians, and medications known to be associated with dry mouth according to the manufacturers’ product monographs were identified. SSc disease duration was measured as the time since onset of the first non–Raynaud’s phenomenon disease symptoms to study visit. Skin involvement was assessed using the modified Rodnan skin thickness score, ranging from 0–51 (11). Limited cutaneous SSc (lcSSc) was defined as skin involvement distal to the elbows and knees, with or without face involvement (12). Diffuse cutaneous SSc (dcSSc) was defined as skin involvement proximal to the elbows and knees, with or without truncal involvement (12). The presence of GERD was ascertained by patient self-report of a history of acid regurgitation (“I have food or acid-tasting liquid that comes back up into my mouth or nose”), nocturnal choking (“I wake up at night choking”), or heartburn (“I have a burning feeling rising from my stomach or lower chest up towards my neck”) on most days. SSc global disease severity was measured with the physician global assessment of disease severity using a 0–10 numerical rating scale, a partially validated measure of severity in SSc (13). In sensitivity analyses, the modified Medsger disease severity scale was applied (14–16). In this latter scale, a severity score ranging from 0 (normal) to 4 (end stage) is generated for each of 9 organ systems. There is no overall score for this scale; therefore, we used the 9 individual organ scores and the sum of the 9 organ scores in sensitivity analyses.

Hand involvement was assessed by 2 methods: the average fingertip-to-palm distance of the right and left third fingers with the hand fully flexed, with larger numbers indicating greater limitation (17), and the items from the Health Assessment Questionnaire (HAQ) to assess upper extremity function (dressing and grooming, eating, and grip; range 0–3).

Autoantibodies

Serum was collected on all patients recruited by the CSRG at their baseline registry visit and sent to a central laboratory (Mitogen Advanced Diagnostics Laboratory, University of Calgary). Aliquots of sera were stored at −80°C until needed. Anticentromere antibodies were detected by indirect immunofluorescence performed on HEp-2 substrate (HEp-2000, Immuno Concepts). Anti–Ro 52/tripartite motif–containing protein 21 (TRIM-21), SSA/Ro 60, SSB/La, and topoisomerase I autoantibodies were assayed by an addressable laser bead immunoassay using a commercially available kit (QUANTA Plex ENA 8, Inova Diagnostics) in a Luminex 100 platform according to protocols previously described (18). In addition, RNA polymerase III antibodies were detected by enzyme-linked immunosorbent assay (Inova Diagnostics) (19).

Dental and periodontal examinations

Eight dentists at 7 CSRG sites performed detailed, standardized dental examinations. Interincisal distance was assessed at the beginning of the examination to avoid bias induced by prolonged stretching of the mouth. Subjects were asked to open their mouths as wide as they could, and the interincisal distance was measured as the distance between the incisal edge of the lower central incisor tooth to the incisal edge of the upper central incisor (20). The same was done for edentulous patients, with dentures in the mouth. The number of missing teeth was recorded, excluding wisdom teeth. Periodontal disease was determined by assessing periodontal probing depth (PD) and clinical attachment level (CAL) (2). PD is the distance from the gingival margin to the base of the gingival sulcus (21). It was measured in millimeters with a Williams probe (Hu-Friedy 945188PW) at 6 sites on all teeth: mesiobuccal, midbuccal, distobuccal, mesiolingual, midlingual, and distolingual (22). All teeth were measured for PD twice. When 2 PD values differed by more than 2 mm, the measurement was performed a third time, and the 2 closest values were retained. The PD values at each site were averaged. The CAL is the distance on the buccal or labial surface from the cementoenamel junction to the base of the gingival sulcus measured using a Williams probe (21). This was measured once for each tooth. The presence of periodontal disease in a given tooth was defined as either a PD >3 mm or a CAL ≥5.5 mm (21,23). The extent of periodontal disease in the entire mouth was calculated as the number of teeth with either a PD >3 mm or a CAL ≥5.5 mm.

The Saxon test was used to determine the subject’s unstimulated saliva production per minute (24). Subjects were asked to keep a piece of preweighed gauze in their mouths for 1 minute. They were instructed to move it around but not to chew on it. The gauze was then reweighed to determine the quantity of unstimulated saliva present in the mouth in mg/minute. The scale (Mettler AJ50) was sensitive to 0.01 mg.

Statistical analysis

Descriptive statistics were used to summarize the baseline characteristics of the sample. Comparisons of orofacial abnormalities between dcSSc and lcSSc patients were performed using Student’s t-tests and Mann Whitney U tests, as appropriate.

Associations between sociodemographic and disease variables with saliva and interincisal distance were assessed using linear regressions. Associations between sociodemographics and disease variables with the number of missing teeth and the number of teeth with periodontal disease were assessed using generalized linear models with negative binomial distribution for count data. Multivariate analyses were adjusted for potential confounders identified from the 2002 US Oral Health annual report (25), and included age, sex, ethnicity (white versus other), education (greater versus less than high school), and current smoking. In addition, the number of teeth was added as an offset term variable in the models for periodontal disease (21,23,26). Edentulous patients were excluded from studies of this abnormality. We repeated the models using study site as a random effect, and no differences in parameter estimates were found.

In sensitivity analyses, we assessed the possibility that the effects of interincisal distance and saliva could be cumulative over time. Therefore, we expressed saliva and interincisal distance as the cumulative loss of either over the disease duration. To calculate this, we first estimated the amount of saliva and the interincisal distance that the SSc patients would have had at the onset of their disease. To estimate these values, we used the data from the control group based on age and sex. We obtained linear regression equations to predict the production of saliva and the interincisal distance at disease onset in subjects. Finally, we assumed a linear decline in saliva from the time of disease onset to the time of the examination for this study and calculated the cumulative loss of saliva by the following formula: (saliva production at disease onset − current saliva production) × (disease duration)/2. The cumulative loss of interincisal distance was calculated similarly.

All statistical analyses were performed using SAS, version 9.2. P values less than 0.05 were considered statistically significant.

RESULTS

Study subjects

The study included 163 SSc subjects (Table 1). They were mainly women (90%), the mean ± SD age was 56 ± 11 years, 93% were white, 9% were current smokers, and 49% had more than high school education. Diffuse disease was present in 28% of the subjects and the mean ± SD disease duration was 14 ± 8 years. The mean ± SD physician global assessment of disease severity was 2.9 ± 2.2.

Table 1.

Characteristics of the systemic sclerosis subjects (n = 163)^*

	Value
Sociodemographic variables
Female sex, no. (%)	146 (89.6)
White, no. (%)	150 (92.6)
Current smoker, no. (%)	15 (9.5)
Education greater than high school, no. (%)	80 (49.4)
Age, mean ± SD years	56.20 ± 10.56
Disease variables
Diffuse cutaneous disease, no. (%)	45 (27.6)
Disease duration, mean ± SD years	13.76 ± 8.42
mRSS (range 0–51)	8.48 ± 8.56
GERD, no. (%)	90 (57.0)
Fingertip-to-palm distance, mean ± SD cm	0.8 ± 1.44
Hand function (upper body HAQ; range 0–3), mean ± SD	0.78 ± 0.76
Disease severity score, mean ± SD
General	0.49 ± 0.7
GI tract	2.06 ± 0.72
Heart	0.48 ± 1.02
Joint/tendon	0.66 ± 1.12
Kidney	0.01 ± 0.08
Lung	1.27 ± 1.15
Muscle	0.16 ± 0.55
Peripheral vascular	1.29 ± 1.16
Skin	1.15 ± 0.69
Physician global assessment of severity (range 0–10), mean ± SD	2.91 ± 2.19
Autoantibodies, no. (%)
Centromere	56 (36.6)
Topoisomerase I	20 (13.0)
RNA polymerase III	36 (24.5)
Sjögren’s-related antibodies (Ro 52/TRIM-21, SSA/Ro 60, either/or SSB/La)	26 (16.9)
Oral abnormalities
Saliva, mean ± SD mg/minute	147.52 ± 95.07
Interincisal distance, mean ± SD mm	37.68 ± 8.36
Dental abnormalities
Edentulous, no. (%)	21 (12.9)
No. of missing teeth, mean ± SD	7.90 ± 9.44
No. of teeth with periodontal disease, mean ± SD†	5.23 ± 5.63

^*% values adjusted to reflect missing values.

mRSS = modified Rodnan skin thickness score; GERD= gastroesophageal reflux disease; HAQ = Health Assessment Questionnaire; GI = gastrointestinal; TRIM-21 = tripartite motif–containing protein 21.

^†Exclusion of completely edentulous patients.

Orofacial abnormalities

SSc subjects produced mean ± SD 147.5 ± 95.1 mg/minute of saliva, and their mean ± SD interincisal distance was 37.7 ± 8.4 mm. SSc patients had mean ± SD 7.9 ± 9.4 missing teeth and mean ± SD 5.2 ± 5.6 teeth with periodontal disease in their mouth. There were no differences in saliva flow, number of missing teeth, or number of teeth with periodontal disease in subjects with dcSSc or lcSSc. Mean ± SD interincisal distance was smaller in those with dcSSc (34.4 ± 7.9 mm versus 38.9 ± 8.3 mm; P = 0.003).

The use of medications associated with mouth dryness was present in 76.3% of patients and was not related to saliva production (β = 8.2, P = 0.629). Adjusting for medications associated with dryness also did not change the results of any of the models, and is therefore not included as a confounder in any of the multivariate analyses below.

Saliva production

Relationship between saliva production and Sjögren’s syndrome–related antibodies

The presence of at least one Sjögren’s syndrome–related antibody (either Ro 52/TRIM-21, SSA/Ro 60, or SSB/La) was associated with decreased saliva production in univariate analyses (P = 0.008) (Table 2). The relationship between Sjögren’s syndrome–related antibodies and decreased saliva production was confirmed in multivariate analysis (β = −43.32 [95% confidence interval (95% CI) −80.89, −5.75], P = 0.024) (Table 3).

Table 2.

Univariate relationship between sociodemographics, disease severity, and autoantibodies with dental and oral abnormalities^*

	Saliva, mg/minute (n = 162)†		Interincisal distance, mm (n = 161)†		No. of missing teeth (n = 163)‡		No. of teeth with periodontal disease (n = 142)§
	β	P	β	P	RR	P	RR	P
Sociodemographic variables
Female sex	−80.54	0.001¶	−3.30	0.147	0.86	0.671	1.10	0.774
White	41.93	0.138	5.30	0.035¶	1.05	0.898	1.37	0.411
Current smoker	34.75	0.148	−0.07	0.977	1.04	0.924	2.14	0.015¶
Education greater than high school	22.74	0.125	0.88	0.511	0.46	< 0.001¶	1.04	0.828
Age, years	−1.87	0.009¶	0.02	0.805	1.06	< 0.001¶	1.01	0.567
Disease variables
Diffuse cutaneous disease	0.57	0.973	−4.41	0.003¶	0.92	0.736	1.16	0.497
Disease duration, years	−0.97	0.277	0.08	0.280	1.01	0.268	1.00	0.953
mRSS (range 0–51)	0.76	0.394	−0.35	< 0.001¶	0.99	0.663	0.99	0.351
GERD	−16.40	0.250	−1.50	0.272	1.55	0.045¶	1.24	0.269
Fingertip-to-palm distance, cm	6.59	0.233	−2.30	< 0.001¶	0.95	0.492	0.94	0.322
Hand function (upper body HAQ; range 0–3)	2.32	0.803	−4.19	< 0.001¶	1.23	0.144	1.17	0.208
Disease severity score (range 0–4)
General	−8.31	0.439	−0.60	0.528	1.18	0.307	0.88	0.328
GI tract	−9.62	0.354	−1.72	0.062	1.25	0.230	1.06	0.637
Heart	4.29	0.566	−0.92	0.165	1.10	0.396	0.86	0.125
Joint/tendon	11.07	0.114	−3.14	< 0.001¶	0.97	0.717	0.88	0.147
Kidney	−16.05	0.869	−4.28	0.598	0.73	0.823	1.01	0.994
Lung	−1.37	0.833	−2.18	< 0.001¶	1.13	0.199	0.92	0.311
Muscle	1.50	0.914	−3.66	0.003¶	1.18	0.415	1.00	0.991
Peripheral vascular	2.32	0.725	−1.58	0.006¶	1.02	0.820	0.98	0.847
Skin	10.61	0.342	−3.48	< 0.001¶	0.85	0.347	0.91	0.483
Mean physician global assessment of severity (range 0–10)	−1.66	0.629	−1.09	< 0.001¶	1.07	0.147	0.98	0.578
Autoantibodies
Centromere	−18.23	0.266	3.65	0.010¶	1.32	0.207	1.57	0.029¶
Topoisomerase I	9.10	0.695	−4.36	0.031¶	0.86	0.636	0.62	0.129
RNA polymerase III	27.92	0.140	−2.80	0.097	1.24	0.371	0.96	0.883
Sjögren’s–related antibodies (Ro 52/TRIM-21, SSA/Ro 60, either/or SSB/La)	−55.82	0.008¶	−3.62	0.047¶	1.31	0.346	1.30	0.355
Oral abnormalities
Saliva, mg/minute	–	–	–	–	0.84	0.057	0.93	0.454
Interincisal distance, mm	2.63	0.002¶	–	–	0.99	0.696	1.01	0.261
Dental abnormalities
No. of teeth with periodontal disease#	–	–	–	–	0.94	0.002¶	–	–

^*RR = relative risk; mRSS = modified Rodnan skin thickness score; GERD = gastroesophageal reflux disease; HAQ = Health Assessment Questionnaire; GI = gastrointestinal; TRIM-21 = tripartite motif–containing protein 21.

^†Linear regression.

^‡Negative binomial regression.

^§Negative binomial regression. The number of teeth was added as an offset term and edentulous patients were also excluded.

^¶Significant at P < 0.05.

^#Exclusion of completely edentulous patients.

Table 3.

Multiple regression analyses to determine the association between antibodies and dental abnormalities, adjusted for potential confounders

	β	95% confidence interval	P
Dependent variable: saliva, mg/minute
Age	−1.10	−2.54, 0.34	0.132
Female sex	−60.07	−107.39, −12.75	0.013*
White	45.03	−12.18, 102.23	0.122
Education (greater than high school)	21.35	−6.95, 49.64	0.138
Current smoker	28.22	−18.31, 74.74	0.233
Disease duration	−0.23	−1.89, 1.42	0.780
Sjögren’s syndrome–related antibodies	−43.32	−80.89, −5.75	0.024*
Dependent variable: interincisal distance, mm
Age	−0.04	−0.19, 0.10	0.543
Female sex	−3.87	−8.86, 1.12	0.127
White	5.50	−0.09, 11.08	0.054
Education (greater than high school)	0.54	−2.29, 3.36	0.708
Current smoker	−0.68	−5.53, 4.17	0.782
Disease duration	0.09	−0.08, 0.25	0.302
Anticentromere antibody	4.26	1.32, 7.21	0.005*
Anti–topoisomerase I antibody	−3.99	−8.04, 0.06	0.053

^*Significant at P < 0.05.

Relationship between saliva production and disease severity

None of the disease severity scores was significantly associated with saliva in univariate or multivariate analyses (Table 2).

Interincisal distance

Relationship between interincisal distance and overall disease severity

In univariate analyses, interincisal distance was significantly inversely associated with disease severity assessed by the mean physician global assessment of severity (P < 0.001), higher skin scores (P < 0.001), and diffuse disease (P = 0.003) (Table 2). This was confirmed in multivariate analyses (Table 4).

Table 4.

Multiple regression analyses to determine the association between disease severity and extent of skin disease, and the dependent variable interincisal distance, adjusted for potential confounders^*

	β	95% CI	P
Association between disease severity and interincisal distance
Age	0.01	−0.13, 0.15	0.887
Female sex	−2.26	−6.88, 2.36	0.334
White	3.38	−1.73, 8.49	0.193
Education (greater than high school)	0.67	−2.03, 3.37	0.623
Current smoker	−0.17	−4.68, 4.35	0.942
Disease duration	0.07	−0.09, 0.23	0.393
Mean physician global assessment of severity	−1.02	−1.63, −0.42	0.001†
Association between skin score and interincisal distance
Age	−0.06	−0.19, 0.07	0.388
Female sex	−3.54	−8.00, 0.93	0.120
White	3.43	−1.44, 8.30	0.166
Education (greater than high school)	1.13	−1.47, 3.72	0.392
Current smoker	−1.05	−5.42, 3.31	0.634
Disease duration	0.07	−0.08, 0.22	0.360
mRSS	−0.38	−0.53, −0.23	< 0.001†
Association between diffuse cutaneous disease and interincisal distance
Age	−0.04	−0.17, 0.10	0.587
Female sex	−3.12	−7.78, 1.54	0.188
White	3.86	−1.20, 8.92	0.134
Education (greater than high school)	1.16	−1.53, 3.85	0.397
Current smoker	−0.16	−4.66, 4.34	0.944
Disease duration	0.07	−0.09, 0.23	0.406
Diffuse cutaneous disease	−5.23	−8.23, −2.23	0.001†

^*95% CI = 95% confidence interval; mRSS = modified Rodnan skin thickness score.

^†Significant at P < 0.05.

Relationship between interincisal distance and antibodies

Loss of interincisal distance was associated with the presence of anti–topoisomerase I antibodies (P = 0.031), whereas anticentromere antibodies were associated with a larger interincisal difference (P = 0.010) (Table 2). Antibodies to RNA polymerase III were not associated with any oral abnormalities. Multivariate analyses (Table 3) also showed that anticentromere antibodies were associated with a larger interincisal difference (β = 4.26 [95% CI 1.32, 7.21], P = 0.005) and a strong trend to a significant association between anti–topoisomerase I antibody with loss of interincisal distance (β = −3.99 [95% CI −8.04, 0.06], P = 0.053).

Although loss of interincisal distance was also associated with any of the Sjögren’s syndrome–associated antibodies in univariate analysis (P = 0.047) (Table 2), this relationship was not confirmed by multivariate analysis (β = 1.76 [95% CI −3.81, 7.32], P = 0.534).

Missing teeth

Relationship between missing teeth and periodontal disease

Unexpectedly, the number of missing teeth was significantly inversely associated with the number of remaining teeth with periodontal disease in univariate (P = 0.002) (Table 2) and multivariate analyses adjusted for demographics, disease duration, and disease severity (relative risk [RR] 0.95 [95% CI 0.92, 0.99], P = 0.009) (Table 5).

Table 5.

Multiple regression analyses to determine the association between periodontal disease and saliva production, and the dependent variable missing teeth, adjusted for potential confounders^*

	RR	95% CI	P
Association between periodontal disease and missing teeth
Age	1.05	1.03, 1.07	< 0.001†
Female sex	0.86	0.42, 1.78	0.684
White	1.10	0.49, 2.44	0.822
Education (greater than high school)	0.57	0.38, 0.85	0.005†
Current smoker	1.94	0.96, 3.91	0.066
Disease duration	1.00	0.98, 1.03	0.913
Mean physician global assessment of severity	1.05	0.96, 1.14	0.272
No. of teeth with periodontal disease	0.95	0.92, 0.99	0.009†
Association between saliva production and missing teeth
Age	1.07	1.05, 1.10	< 0.001†
Female sex	0.49	0.25, 0.96	0.038†
White	0.74	0.35, 1.59	0.445
Education (greater than high school)	0.52	0.35, 0.77	0.001†
Current smoker	2.17	1.12, 4.20	0.021†
Disease duration	1.00	0.98, 1.03	0.729
Mean physician global assessment of severity	1.00	0.92, 1.09	0.983
Saliva, per 10 mg/minute	0.97	0.94, 0.99	0.004†

^*RR = relative risk; 95% CI = 95% confidence interval.

^†Significant at P < 0.05.

Relationship between missing teeth and saliva production

The number of missing teeth was also associated with saliva production in multivariate analysis and was independent of disease severity (RR 0.97 [95% CI 0.94, 0.99], P = 0.004) (Table 5). An increase of 10 mg/minute of saliva decreased the number of missing teeth by 3%.

Relationship between missing teeth and interincisal distance, fingertip-to-palm distance, and upper extremity HAQ score

We did not find a relationship between interincisal distance and missing teeth in univariate (Table 2) or multivariate analyses adjusting for demographics, smoking, and disease duration (RR 0.99 [95% CI 0.96, 1.01], P = 0.214). Also, fingertip-to-palm distance and upper extremity HAQ score were not associated with the number of missing teeth in univariate analyses (Table 2). However, upper extremity HAQ score (RR 1.36 [95% CI 1.07, 1.73], P = 0.013), but not fingertip-to-palm distance (RR 1.06 [95% CI 0.92, 1.23], P = 0.405), was significantly associated with an increase in the number of missing teeth in multivariate analysis (Table 6).

Table 6.

Multiple regression analyses to determine the association between upper extremity HAQ or GERD and missing teeth, adjusted for potential confounders^*

	RR	95% CI	P
Association between hand function and missing teeth
Age	1.08	1.06, 1.10	< 0.001†
Female sex	0.68	0.35, 1.32	0.258
White	0.85	0.39, 1.84	0.672
Education (greater than high school)	0.48	0.33, 0.70	< 0.001†
Current smoker	1.79	0.92, 3.47	0.087
Disease duration	1.00	0.98, 1.02	0.967
Upper extremity HAQ	1.36	1.07, 1.73	0.013†
Association between GERD and missing teeth
Age	1.08	1.06, 1.10	< 0.001†
Female sex	0.51	0.26, 1.01	0.053
White	0.64	0.31, 1.33	0.231
Education (greater than high school)	0.50	0.34, 0.73	< 0.001†
Current smoker	1.79	0.93, 3.47	0.082
Disease duration	1.00	0.98, 1.02	0.952
GERD	1.68	1.14, 2.46	0.008†

^*HAQ = Health Assessment Questionnaire; GERD = gastroesophageal reflux disease; RR = relative risk; 95% CI = 95% confidence interval.

^†Significant at P < 0.05.

Relationship between missing teeth and SSc disease severity

Teeth may be lost directly from the disease process itself. We tested this hypothesis indirectly by assessing if worse global disease was related to missing teeth. Disease severity was not related to missing teeth in univariate (Table 2) or multivariate analyses (RR 1.02 [95% CI 0.93, 1.11], P = 0.747).

Relationship between missing teeth and GERD

GERD was significantly associated with an increase in the number of missing teeth in both univariate (P = 0.045) (Table 2) and multivariate analyses (RR 1.68 [95% CI 1.14, 2.46], P = 0.008) (Table 6).

Periodontal disease

Relationship between periodontal disease and decreased saliva production

Periodontal disease may be secondary to decreased saliva production; however, saliva production did not predict periodontal disease in either univariate (P = 0.454) (Table 2) or multivariate analyses (RR 0.99 [95% CI 0.96, 1.01], P = 0.198).

Relationship between periodontal disease and interincisal distance, fingertip-to-palm distance, and upper extremity HAQ score

Poor oral care due to decreased interincisal distance or hand disease could be related to periodontal disease. Interincisal distance was not associated with periodontal disease in univariate (P = 0.261) (Table 2) or multivariate analyses (RR 1.01 [95% CI 0.98, 1.03], P = 0.513). Upper extremity HAQ score and fingertip-to-palm distance were not related to periodontal disease in univariate analyses (P = 0.208 and P = 0.322, respectively) (Table 2) or in multivariate analyses (RR 1.13 [95% CI 0.87, 1.46], P = 0.366 and RR 0.95 [95% CI 0.83, 1.08], P = 0.392, respectively).

Relationship between periodontal disease and global disease severity

Disease severity was not related to periodontal disease in univariate analysis (Table 2) or in multivariate analysis (RR 0.99 [95% CI 0.91, 1.08], P = 0.803).

Sensitivity analyses

All analyses that used the physician global assessment of disease severity were repeated using either the multiple Medsger disease severity scores or the sum of the individual Medsger scores. All of the findings reported above were the same using the above measures of disease severity. Also, we repeated all analyses using our estimates of the cumulative loss of saliva or interincisal distance over the disease course for each subject. All of the findings reported concerning saliva production or interincisal distance were the same using these other two measures.

DISCUSSION

The purpose of this study was to determine the relationships between our previously reported findings of decreased saliva production, decreased interincisal distance, increased missing teeth, and more periodontal disease in SSc (2) and aspects of the disease itself. Additionally, we wished to investigate the relationships between reduced saliva and decreased interincisal distance and the findings of increased numbers of missing teeth and more periodontal disease.

With regard to our hypotheses, we predicted global disease severity would be associated with decreased interincisal distance. We indeed found a strong relationship between overall disease severity and extent of skin involvement with interincisal distance. Although antibody status is not a direct measure of disease severity, the decreased interincisal distance with the presence of anti–topoisomerase I antibodies is in keeping with a relationship to disease severity, as it is widely appreciated that these antibodies are associated with dcSSc (27). Indeed, dcSSc patients had a smaller mean interincisal distance than lcSSc subjects.

Conversely, we predicted and confirmed that decreased saliva production would primarily be associated with Sjögren’s syndrome–related antibodies and did not find an association between saliva production and overall disease severity. Using the American–European Consensus Group classification criteria (28), a recent study found Sjögren’s syndrome in 33.9% of 118 patients with SSc (29). Sjögren’s syndrome–related antibodies were associated with Sjögren’s syndrome in their subjects. We did not apply any criteria to our subjects to determine if they had Sjögren’s syndrome. While there have been at least 12 classification or diagnostic criteria published for Sjögren’s syndrome since 1965 (30), most require either a labial salivary gland biopsy and/or a measure of saliva production. We did not think it would be ethical to perform labial biopsies in all of our subjects. Also, to use criteria that included measures of salivary flow to determine if Sjögren’s syndrome was associated with decreased salivary flow would be circular and therefore not helpful. However, our finding that decreased saliva production was associated with Sjögren’s syndrome–related antibodies is suggestive that underlying Sjögren’s syndrome was the cause of the salivary flow abnormality.

We thought that the findings of more missing teeth and periodontal disease might be related to disease severity itself or to hand dysfunction. There was no relationship, however, between disease severity or hand involvement and periodontal disease, nor were missing teeth related to disease severity. However, upper extremity function was related to the number of missing teeth, suggesting that poor oral hygiene could be a factor in tooth loss in SSc. A recent study did find a relationship between hand function and gingivitis in SSc (31). That study employed a specific evaluation of manual dexterity to perform oral hygiene using the Toothbrushing Ability Test and assessed active gingivitis rather than periodontal tissue recession. The different methods of evaluating both hand function and periodontal disease may explain the different results from our study.

We did find that tooth loss was associated with GERD. GERD has been associated with dental erosion, especially in children (32–37). To our knowledge, there is little information about the relationship between GERD and tooth loss in adults, although in a small study, Yoshikawa et al found a small increase in missing teeth in GERD patients (36). Although our previous report did not find a relationship between SSc and tooth decay (2), we could only assess remaining teeth for decay, and therefore decay could still be a mediator between GERD and tooth loss in SSc.

We hypothesized that decreased interincisal distance and saliva production might in turn be possible causes of tooth and periodontal disease independent of disease severity. Indeed, decreased saliva was associated with the number of missing teeth. Interincisal distance, however, was not related to missing teeth or PD. We also unexpectedly found an inverse relationship between the extent of periodontal disease and the number of missing teeth. We do not have a good explanation for this observation, but wonder if the chances of finding periodontal disease in the remaining teeth decrease as the number of remaining teeth decreases.

There are some limitations to this study. Foremost is the fact that, as we reported in the first article from this study (2), the study sample was in many ways not as severely involved as the rest of our SSc cohort. The lack of an extensively validated measure of global disease severity is also a limitation, but we tried to account for this by confirming our findings regarding disease severity by using several different measures. The Medsger disease severity scales have never been adapted for use as one global measure of severity (6–9). We have previously used the physician global assessment of disease severity and have found that it performed well and as expected for a severity measure, thus suggesting construct validity (38–40).

Another study limitation is that we did not employ direct measures such as pH monitoring or endoscopy to diagnose GERD. We thought that this would not be ethical in our population. Many questionnaires have been used for the diagnosis of GERD and have been reviewed (41). The questions that we have used to diagnose GERD have been adapted from these tools, which have sensitivities of 62– 67%and specificities of 63–70% (42,43). The finding of significant a priori hypothesized relationships with the GERD variables that we did use supports the use of our measures.

Unfortunately, for logistic reasons we were not able to use one of the specific measures of hand function that have been used in SSc (44–46), but instead adopted the HAQ and used the upper extremity questions. This has been done previously (17,47,48). If our instrument was not sensitive enough, we may have missed a relationship between hand function and periodontal disease. However, we did not find any relationship between finger-to-palm distance, a validated measure of finger range of motion in SSc (17), and missing teeth or periodontal disease, and we would expect this measurement to be related to hand function. Also, this was a cross-sectional study in subjects with established disease, and we could not determine if the oral abnormalities such as missing teeth occurred before or after the disease onset.

Conversely, our study has considerable strengths. To our knowledge, this is the largest and perhaps only such study in which a careful standardized dental examination of multiple facets of oral health has been performed simultaneously on a well-characterized population of SSc subjects.

The picture that emerges from these data suggests the following possible sequence of events: diminished saliva production in SSc is related mainly to concomitant Sjögren’s syndrome. Decreased saliva, GERD, and diminished hand function, which may affect the ability to maintain good oral hygiene, are all related to missing teeth. Decreased interincisal distance, which is associated with worse global disease, is related neither to the loss of teeth nor to periodontal disease. The cause of the worse periodontal disease in SSc is likely multifactorial, but remains unclear. Although we cannot be certain, our data suggest that improving oral care, which may be impaired because of hand problems, with new brushing or flossing techniques may not be particularly effective in preventing periodontal disease, but may help prevent tooth loss. Conversely, there is a report that the use of an oscillating–rotating–pulsating toothbrush and a flosser that has a toothbrush-like handle may reduce the degree of inflammation in marginal gingiva (49). Whether stimulating saliva production would help prevent any of these problems remains to be tested.

Significance & Innovations.

• Oral abnormalities in systemic sclerosis (SSc) are decreased saliva, decreased interincisal distance, missing teeth, and periodontal disease.

• Diminished interincisal distance is associated with SSc disease severity, but decreased saliva production relates to Sjögren’s syndrome–related antibodies.

• Tooth loss is associated with poor upper extremity function, gastroesophageal reflux disease, and decreased saliva.