Methods. The study, conducted between 1973 and 1994, evaluated 273 subjects (62.3 percent male) with a mean age of 59.6 years.

Results. Of the 273 subjects with 666 abutments, 74 lost 133 abutments. The most common cause of tooth loss was periodontal disease (29.3 percent) followed by periapical lesions (18.8 percent) and caries (16.5 percent). Through logistic regression, the authors found that subjects who lost teeth were more likely to have medical problems that could cause soft-tissue lesions of the oral mucosa, were less likely to use fluoride daily and were less likely to return for yearly recall visits. The authors found 22 vertical fractures in 17 subjects. ² analysis revealed that overdenture teeth in the maxillary arch that were opposed by natural teeth were more likely to experience vertical fractures.

Conclusions. In a study that followed up some patients for as long as 22 years, the rate of tooth loss was 20.0 percent. Many of these failures could have been prevented if patients had practiced better oral hygiene.

Clinical Implications. The findings suggest that if a dentist recommends over-denture therapy, the patient needs to be examined regularly to reduce the risk of experiencing caries and periodontal disease. Also, if the abutments are in the maxilla and are opposed by natural teeth, the dentist should consider using thimble crowns to reduce the risk of vertical fractures.

Patients with overdentures need to be examined regularly to reduce the risk of experiencing caries and periodontal disease.

In their 10-year follow-up study, Toolson and Taylor²¹ reported a similar pattern; four of 11 teeth were extracted because of periodontal disease and seven because of caries. However, Reitz and colleagues¹⁷ studied 35 patients with 95 overdenture abutments; 13 of these teeth were extracted, 12 of which because of periodontal disease.

Relatively few studies^9,11,15–21 have evaluated tooth loss in these populations. Therefore, we conducted a longitudinal, prospective cohort study of abutment loss in patients wearing overdentures.

	SUBJECTS, MATERIALS AND METHODS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RECALL APPOINTMENTS STATISTICAL ANALYSIS RESULTS DISCUSSION CONCLUSION REFERENCES

 
This prospective cohort study was composed of patients receiving overdenture therapy in the prosthodontics clinic at The University of Iowa College of Dentistry, Iowa City, from 1973 to 1994. Predoctoral dental students, postdoctoral students or faculty members provided all restorative care. We initially enrolled 395 patients in this study who were

– able to give informed consent;

– ambulatory and able to return to the clinic for recall appointments;

– physically able to clean their teeth or abutments.

Before reducing the abutment teeth, the dental students and dentists scaled and cleaned all teeth and taught patients how best to clean their remaining teeth. The clinicians found no periapical lesions on the abutment teeth and removed all caries. Eight patients had pockets deeper than 3 millimeters after débridement, and the clinicians referred them to the periodontics department for more extensive care, such as surgical recontouring, to establish optimal gingival health. This number of patients was too small to perform a separate statistical analysis.

We needed to refer these patients to the periodontal department because the majority were treated in our prosthodontic department by dental students in their junior year under the supervision of one of us (R.E.). The junior year program involves specialty rotations, so students brought their patients to the periodontal department to receive basic cleaning and scaling. Dr. Ettinger evaluated all patients after the dental students delivered the dentures to them.

The students reduced all of the abutment teeth to a level of 1.5 to 2.0 millimeters above the gingival margin. For most patients, endodontic therapy was completed in one visit, and the students restored the access opening with a restorative material after removing any caries in the tooth. None of the teeth had discernible periapical radiolucencies at baseline.

We used cast-gold copings only when it was impossible to prepare sound supragingival root surfaces on the abutment teeth as a result of caries. Gold copings were required for 22 of the abutments. Eighty-seven teeth did not require endodontic therapy (that is, they were vital over-denture abutments). We selected these teeth for simple recontouring without endodontic therapy, because the root canal space appeared radiographically to be substantially diminished to a level below the free gingival margin and without pulpal exposures. At the time of reduction, all vital teeth were responsive to electric pulp testing.

At the time of denture placement, a single examiner (R.E.) took baseline measurements and photographs of all patients. These measurements included abutment height above the gingival margin and periodontal probing depth. The examiner also recorded plaque levels, bleeding on probing and horizontal mobility. The dental students again showed patients how to brush their abutments, and dentists prescribed a high-concentration fluoride gel (5,000 parts per million) for daily home use. Students instructed patients to eat breakfast each morning, clean their dentures and abutments, place one drop of fluoride gel in the depression of the overdenture abutment in the denture, place the dentures in their mouths, and forgo eating or drinking for one-half hour.

	RECALL APPOINTMENTS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RECALL APPOINTMENTS STATISTICAL ANALYSIS RESULTS DISCUSSION CONCLUSION REFERENCES

 
We asked patients to return to the department of prosthodontics at six-month intervals. Table 2 shows the recall rate. At these recall appointments, Dr. Ettinger evaluated the overdenture abutment teeth with respect to the condition of the restorations, dental caries and periodontal problems. After this evaluation, the students performed appropriate maintenance care, such as scaling and teeth cleaning. Dr. Ettinger evaluated the abutments for changes (for example, caries) and, if necessary, replaced the restorations. The clinicians adjusted the dentures to eliminate pressure areas on the soft tissues and repaired or relined the dentures if necessary.

At recall appointments, the examiner evaluated the overdenture abutment teeth with respect to the condition of the restorations, dental caries and periodontal problems.

Six months after the patient’s last appointment, the department’s clerical staff sent out a recall card asking the patient to call the department to make an appointment. If no communication was received, a staff member sent out another card within two months. If the patient did not respond or the letter was returned "address unknown," the clerical staff tried to contact the patient by telephone.

To characterize the differences between subjects who lost teeth and those who did not, we compared a number of independent variables measured at baseline:

– age at which the patient received the overdenture;

– sex;

– mental status—alert, semialert, noncommunicative;

– level of cooperation—very cooperative, somewhat cooperative, uncooperative;

– auditory status—normal, aided, impaired;

– sight—normal, corrected, impaired;

– speech status—normal, speaks with difficulty, unintelligible;

– ability to ambulate—ambulatory, walks with aid, transportation required (for example, wheelchair);

– medical problems that potentially could cause oral soft-tissue lesions (for example, leukemia, anemia, lupus, diabetes), hard-tissue changes (for example, osteoporosis, rheumatoid arthritis) or loss of neurological control (for example, stroke, Parkinson’s disease, trigeminal neuralgia);

– number of medications used (for example, zero, one to three, four or more);

– medications that potentially could result in anticholinergic effects (for example, diuretics, tricyclics, phenothiazines), inflammatory lesions (for example, nonsteroidal anti-inflammatory drugs, gold salts, phenytoin, calcium channel blockers), increased risk of local infection (for example, corticosteroids, chemotherapy) or dyskinetic movements (for example, tricyclics, phenothiazines).

Dr. Ettinger measured the following variables at the time of tooth loss or at the last appointment before the teeth were extracted:

– smoking—never, past smoker, currently smoking;

– brushing habits—two or more times daily, once daily, occasionally;

– use of high-concentration fluoride at home—daily, every second day, occasionally/none;

– use of chlorhexidine—daily, occasionally, none.

To evaluate compliance, we divided the number of visits a subject had by the number of years he or she had been in the study, dichotomized the variable and compared subjects who had lost teeth with those who had not.

	STATISTICAL ANALYSIS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RECALL APPOINTMENTS STATISTICAL ANALYSIS RESULTS DISCUSSION CONCLUSION REFERENCES

 
We used SAS software version 8 (SAS Institute, Cary, N.C.) to analyze the data. We performed an assessment of associations for contingency tables and for bivariate analyses using the ² test, Fisher exact test for small sample sizes or Cochran-Mantel-Haenszel test for stratified analyses and analyses of ordinal outcomes. The non-parametric Wilcoxon-Mann-Whitney test, based on rank scores of the data, was used to compare the distribution of age between the groups.

We developed multivariate logistic regression models and included interaction terms. To identify factors associated with subjects who lost teeth and those who did not, we considered only those variables that were found to have a significant association in bivariate analyses as candidates for forward-model selection analyses (forward selection adds variables one at a time to the model as variables meet the specified significance level for entry to the model). All tests had a .05 level of statistical significance.

We generated a Kaplan-Meier survival curve²³ for overdenture abutments for up to 23 years, using any abutments remaining as a survival criterion.

	RESULTS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RECALL APPOINTMENTS STATISTICAL ANALYSIS RESULTS DISCUSSION CONCLUSION REFERENCES

 
A total of 395 subjects fulfilled the initial inclusion criteria; however, only 273 remained in the study after two years (104 failed to return for recall visits, and we had incomplete data for another 18 subjects). Table 3 shows the age and sex distribution of subjects at the time of overdenture placement. Subjects had a mean age of 59.6 years and an age range of 17 through 85 years. Significantly more men than women were in the study, especially in the group consisting of subjects 66 years and older (P = .02).

Table 5 shows the distribution of tooth loss according to years after overdenture placement. The majority of the teeth were lost during the first 10 years of denture wear. Very few abutments were lost after 10 years.

– medical problems that could cause soft-tissue lesions (P = .02);

– drugs that were likely to increase the risk of developing local infection (P = .04);

– current smoking (P = .06);

– brushing teeth only occasionally (P = .002);

– using fluoride at home only occasionally (P = .04);

– annual recall rate (P = .05).

A logistic regression (Table 6) using these six factors in a forward stepwise analysis found that only three remained in the model. Subjects who lost teeth were more likely to have medical problems that could cause soft-tissue lesions (odds ratio = 2.13), used fluoride at home only occasionally (odds ratio = 1.92) and were less likely to return for annual recall visits (odds ratio = 1.94).

– medical problems that potentially could cause soft-tissue lesions (P = .09);

– medical problems that potentially could cause neurological problems (P = .09);

– medical problems that potentially could cause dyskinetic movements (P = .08).

In this study, we found 22 vertical root fractures in 17 subjects; our clinical impression was that the fractures were more common in the maxillae of men, as well as when the abutments were opposed by natural teeth. An analysis of this clinical impression showed that 13 of the fractures were in the maxilla, 12 were in men and 13 were in abutments opposed by natural teeth. ² analysis found that only overdenture abutments in the maxillary arch (P = .04) and those opposed by natural teeth (P = .05) were at significant risk of developing vertical fractures.

However, because of the small numbers in one of the statistical cells, we used Fisher exact test; the results showed that being opposed by natural dentition was not statistically significant (P = .07).

The Kaplan-Meier survival curve (Figure) for the maxilla showed that there were no abutment losses after 16 years, and the survival rate at 16 years was 51.7 percent. The survival curve for the mandibular arch also showed that there were no abutment losses after 16 years, and the survival rate was 72.5 percent. Overall, there were no significant differences in survival rates between the arches.

View larger version (35K): [in this window] [in a new window]   Figure. Kaplan-Meier survival curves for overdenture abutment teeth.

	DISCUSSION

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RECALL APPOINTMENTS STATISTICAL ANALYSIS RESULTS DISCUSSION CONCLUSION REFERENCES

In teeth with periodontal deficits, bone loss could result in communication between the periodontal pocket and the apex. In a number of abutments in this study, restorations were lost and the patient failed to return for treatment for several weeks or months. Swanson and Madison²⁵ conducted in vitro studies in which they demonstrated that dye takes only three to seven days to reach the apex of endodontically treated teeth if the coronal restoration is missing. This suggests that in our clinical study, the periapical lesions associated with 25 abutment teeth may have been due to penetration of organisms from the oral cavity through the voids in the root canal sealer and gutta-percha after the loss of the coronal restorations.

We found it interesting that overdenture abutments in the maxillae of men that were opposed by natural teeth seemed to be more likely to be at risk of developing a vertical fracture. However, the data did not support our clinical impression that sex was a risk factor, but having abutments in the maxilla was supported by the data as a risk factor, especially when they were opposed by natural teeth. Therefore, we suggest that if overdenture abutments are used in the maxilla and are opposed by natural teeth (or by a removable partial denture), it is appropriate to put thimble crowns on these abutments to prevent the increased risk of vertical root fractures.

To our knowledge, the only study that has reported Kaplan-Meier survival data for overdenture abutments was conducted by Keltjens and colleagues.²² They reported an estimated survival rate of 89 percent after six years, which is similar to our six-year survival rate of 88.9 percent for the maxilla and 88.2 percent for the mandible.

Of the 133 abutment teeth lost in this study, all but 22 were lost as a result of caries, periodontal disease or both, diseases associated with plaque accumulation. It would have been interesting to determine the reasons for tooth loss before overdenture therapy in our subjects to see if there was any correlation between the causes before and after therapy. We did not collect these data initially, and many radiographs obtained before tooth extraction were no longer available when we re-examined patients’ dental records.

The bivariate analysis identified six significant variables, and a logistic regression based on these variables identified three risk factors for tooth loss. These risk factors were medical problems (such as leukemia, anemia, lupus, erythematosis and diabetes) that could cause soft-tissue lesions, use of fluoride at home only occasionally, and recall appointments less often than once per year. The first two factors influence the patient’s ability to deal with the stress of bacterial infections, which result in inflammation and may lead to the loss of teeth owing to periodontal disease. Infrequent recall appointments may mean that patients also are not compliant with regard to the preventive regimen (besides regular fluoride use, toothbrushing and removing dentures before sleep) required to preserve overdenture abutments.

The significant finding in the bivariate analysis of poor oral hygiene practices in this group may reflect the older age of these patients, as well as an associated decreased ability to adequately clean teeth owing to loss of fine-motor coordination or poor eyesight. Poor compliance with regard to the daily use of fluoride gel on the abutment teeth also may be explained by the subjects’ failure to act on the information given them by their dentist, because they did not believe it, or because of poor motivation or self-image or emotional problems such as depression²⁶ or alcohol abuse.²⁷ However, we did not measure any of these variables. Nevertheless, a tooth loss rate of 20.0 percent suggests that long-term success in the majority of patients can be achieved with overdenture therapy.

	CONCLUSION

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RECALL APPOINTMENTS STATISTICAL ANALYSIS RESULTS DISCUSSION CONCLUSION REFERENCES

 
In this study, we found that the rate of loss of overdenture abutment teeth was 20.0 percent. In retrospect, we believe that most of the failures could have been prevented. Caries and periodontal disease were the main causes of failure. For many of these patients, maintaining adequate oral hygiene was a problem. This study suggests that improved communication between the patient and the dentist with regard to the daily use of fluoride gel, as well as regular recall appointments, are critical to the success of over-denture therapy.

We found that vertical root fractures were associated with overdenture abutments in the maxilla that were opposed by natural teeth. We suggest that dentists place thimble crowns on such abutments to reduce the risk of vertical fracture. Our data suggest that to preserve the health of overdenture abutment teeth, patients need to use a high-concentration fluoride (5,000 ppm) gel daily, remove plaque effectively and return for recall appointments at least once per year.

View larger version (118K): [in this window] [in a new window]   Dr. Ettinger is a professor, Department of Prosthodontics and Dows Institute for Dental Research, University of Iowa, 418 Dental Science N., Iowa City, Iowa 52242-1010, e-mail "ronald-ettinger{at}uiowa.edu". Address reprint requests to Dr. Ettinger.

View larger version (124K): [in this window] [in a new window]   Dr. Qian is a statistician, Department of Preventive and Community Dentistry, University of Iowa, Iowa City.