Consequences of Success: Do More Teeth Translate into More Disease and Utilization?

190 Journal of Public Health Dentistry

Consequences of Success: Do More Teeth Translate into More Disease and Utilization?

Anil Joshi, BDS, MPH; Chester W. Douglass, DMD, PhD; Henry Feldman, PhD; Paul Mitchell, MS; Alan Jette, PhD

Abstract Objectives: Increased tooth retention coupled with increased numbers of older

adults means that the actual number of teeth at risk to dental disease will increase sharply. Whether this increase in the number of teeth will translate into more disease and utilization in unknown. The purpose of this study was to test this ”more teeth, therefore more dental disease” theory using cross-sectional data. Methods: In-home personal interview and oral examination data were obtained on a probability sample of elders aged 70 years and older living in the six New England states using the Medicare beneficiary list as a sampling frame. Data on dental utilization, number of teeth, dental caries, and periodontal disease were included in the current analysis. Results: Analysis of variance on subjects with 1-1 0 (Group 1). 1 1-24 (Group 2), and2542 (Group 3) teeth show that the extent of bleeding on probing, pocket depth, and loss of attachment all increase as numbers of teeth increase. Similarly, a greater number of restored coronal and root surfaces were found in Group 3 relative to the other two groups. Mean numbers ofdecayedand filled coronalsurfaces were 8.4 in Group 1,33.0 in Group 2, and 50.3 in Group 3. In contrast, unrestored coronal and root surfaces were s@niticantly higher in Group 1 (mean root DS=1.3) than Group 3 (mean roof DS=0.3). Utilkation patterns of tbose with successful aging dentitions (Group 3) show that they are visiting dentists more frequently than the compromisedgroup (Group 1). Conclusions: These cross-sectional data obtained from a probabilify sample of New England elders show that subjects who retained higher numbers of teeth have more periodontal disease and dental caries experience, and visit the dentist more frequently. [J Public Health Dent 1 W6;56(4):190-7]

Key Words: elders, caries, periodontal disease, tooth loss utilization, service demand.

The decline in edentulousness and increase in tooth retention have been documented thoroughly. A dramatic finding is that adultsover age 70 years have 17.2 teeth per person as compared to 7.2 in 1974 (1,2). Coupled with the increased numbers of adults living into their seventh, eighth, and ninth decades of life, these increases in tooth retention per person mean that the actual number of teeth at risk to dental disease in the adult population will be increased by an estimated 1 billion teeth over the next 30 years as current middle-aged adults become senior adults (3).

Previous reports (4,5) have proposed that the number of dentists per

1,000,000 teeth-at-risk is a more accu- rateindicator ofdentalcareneeds than the dentist-to-population ratio. This theory is based on the premise that since adults are retaining more of their natural dentition, the need for dental care is reflected in the number of teeth at risk to dental disease and not by a head count of persons. Published results of epidemiologic surveys show that tooth retention even in countries with developed economiescan be very different, implying that the dentist-to- population ratio would be insensitive to the actual dental care needs. For example, the International Collabora- tive Study (6) found that the 35-44- year-old population of Canterbury

(nonmetropoli tan areas), New Zea- land, had 16.1 teeth per person, while the same age group in Yamanashi (nonmetropolitan areas), Japan, had 26.5 teeth per person. The pattern of disease is also different. Canterbury had a mean D of 0.6 and F of 6.2, while the population of Yamanashi had a mean D of 3.0 and F of 6.1 (6). In these two communities the dentist-to-population ratio could be the same, but the need for dental care is expected to be different.

However, the validity of the “more teeth, more disease” theory proposed by Douglass and Furino is dependent on whether the retention of more teeth does, in fact, result in an increased need and demand for dental care. The purpose of this study was to test this “more teeth, moredisease” theory. We hypothesized that subjects with greater numbers of teeth would show higher levels of dental disease (unmet in part) and utilize more dental services than subjects with fewer teeth. This analysis does not consider pros- thodontic and aesthetic services; rather, the analysis focuses on dental caries and periodontal disease.

Methods Sample. The New England Eider

Dental Study (NEEDS) was conducted in 1988-91 in a geographicdly stratified random sample of community- dwelling elders aged 70 years and older, living within the six New Eng- land states (7). The sampling frame for this study consisted of the Medicare beneficiary list prepared by the Health Care Financing Administration (HCFA). The HCFA estimates that the beneficiary lists achieve an excellent sampling frame with 97percent cover- age of the 65 and older population. The sampling strategy used a two-stage

__ Send correspondence and reprint requests to Dr. Joshi, Department of Oral Health Policy and Epidemiology, Harvard School of Dental Medicine, 188 Longwood Avenue, Boston, MA 01115. Internet: [email protected] Dr. Douglass is professor at the Harvard School of Dental Medicine. Dr. Feldman is senior research scientist, Mr. Mitchell is senior research associate, and Dr. Jette is chief research scientist, all at the New England Research Institute, Boston, MA. Manuscript received: 1 /25/95; returned toauthors for revision: 4/6/95; accepted for publication: 10/26/95.

Vol. 56, No. 4, Summer 1996 191

stratified cluster sampling design. The first stage sampled cities and towns in New England stratified by population size into four categorie+large metropolitan, urban, suburban, and rural. Within these strata, cities and towns were sampled with replacement and with sample size proportional to the size of the 65+ populations within that city or town.

A total of 68 communities were se- lected: one large metropolitan (Bos ton), four urban, 13 suburban, and 50 rural. In the second stage of the sampling design, a 10 percent random sample of beneficiaries 70 years or older was obtained from HCFA for each of these 68 sampled towns,result- ing in a final sample of 3,668 individuals. The combined probability of selec- tion was the same for all participants. Thus, the sample is self-weighting. Each of these 3,668 individuals was screened for study eligibility (270 years, noninstitutionalized, free of specific conditions such as myocardial infarction within last six months, and currently residing in a sampled city/town). From this xreening 2,598 individuals were determined eligible for the study.

Telephone and Personal Inter- view. In Stage I of this study 2,057 (79.2% of eligible) participated in a telephone interview that provided information such as sociodemographics and oral hygiene maintenance behavior and dental utilization. Stage I1 of the study, which included an in-home dental examination and a personal interview, was completed by 1,151 participants, or 56 percent of those who completed stage I and 45 percent of those determined eligible for the study. Details on survey design and a comparison of respondents with non- respondents are available elsewhere (7-9).

Dental Examination. The oral examination included coronal caries, root caries, and periodontal health, and was conducted on all 1,151 study participants, of whom 729 had one or more natural teeth. However, the full- mouth periodontal examination was performed on only 554 of these dentate elders who were assessed neither to be at risk for subacute bacterial en- docarditis nor to have a prosthetic joint replacement. Gingival bleeding was scored after sweeping the periodontal probe from the midbuccal to mesiobuccal aspect of each tooth.

FIGURE 1 Number of School Years by Dentition Status Group

Proportion of Subjects 50%

41%

40%

30%

20%

10%

0%

32%

I 3 I %

19%

1 no-11 Years 0 12 Years W 13-1 5 Years W > = 16 Years

01 -10 1 1 -24 25-32

Number of Teeth

FIGURE 2 Income from All Sources by Dentition Status Group

80%

Proportion of Subjects

40%

20%

0%

44%

... . - . I . - . - . - . - - . I . - . - .

I - 29% 2 5%

21% _ . _

16%

34% I I< $15,000 O $15,000-$29,000 I > $30,000 m RefusedlDonY Know

01-10 11-24 25-32

Number of Teeth

Probing depth and distance from the free gingival margin to the cementoe- name1 junction (CEJ) were measured at four sites per tooth-midbuccal, mesiobuccal, distolingual, and the "deepest" site (if greater than the previous three anatomical sites), found by stepping the probe circumferentially around the tooth. These oral examina-

tions were conducted in the participants' homes by calibrated dentistsus- ing the visual-tactile method and a portable light lamp, explorer, and a mouth mirror (I). Data were recorded by trained interviewers-recorders on a laptop computer (10).

Four dentist-examiners were in- volved throughout the study. Initial


standardization and calibration sessions were held by the co-principal investigator (CWD), who was pre- viously trained in the examination methods by NIDR. After the study be- gan, recalibration sessions were held approximately every four months to maintain interexaminer consistency. Level of interexaminer reliability actu- ally achieved in the in-home examinations was then evaluated by conduct- ing a reexamination of a randomly se- lected quadrant in 5 percent of subjects.

The kappa statistics (11) for results of the coronal canes and root caries examinations calculated at the tooth surface level were 0.86 and 0.73, respectively, in the in-home reexamination. Intraclass correlation coefficients (12) for pocket depth and distance from the free gingival margin to the CEJ were 0.65 and 0.85, respectively.

Statistical Analysis. All analyses in this reportare restricted to the 718 subjects (63% of the sample) who had one or more natural teeth. Eleven subjects had root tips only and were excluded. A chi-square test was used to assess the statistical significance for categori- cal data. Analysis of variance was used to assess the statistical significance of differences among means of multiple groups.

To understand the relationship between number of teeth and the seven oral health measures (gingival bleeding, pocket depth, loss of attachment, restored coronal surfaces, restored root surfaces, decayed coronal surfaces, and decayed root surfaces), subjects were divided into three dentition statusgroups: subjects with 1-10 teeth, subjects with 11-24 teeth, and subjects with 25-32 teeth. This grouping can be thought of as consisting of subjects with compromised natural dentitions, moderately successful aging dentitions, and successful aging dentitions, respectively. This function-based grouping also provided three bal- anced groups relative to sample size with 200 subjects in the lower 30 percent of the distribution for number of remaining teeth, 364 in approximately the interquartile range, and 154 in the upper 22 percent of the distribution.

A separate analysis of variance model was created for each of the seven oral health measures. In each of these models an oral health measure, such as restored coronal surfaces, was used as a dependent variable and age,

TABLE 1 Bmshing and Flossing Frequency by Dentition Status Group

Oral Hygiene Frequency Practice per Day

Brushing' 2Twice Once <Once

Flossingt a c e <Once Never

Number of Teeth

1-10 1 1-24 25-32

92 (46%) 231 (65%) 109 (71%) 59 (30%) 99 (28%) 34 (22%) 42 (24%) 28 (7%) 10 (7%) 39 (20%) 125 (35%) 60 (39%) 26 (13%) 98 (27%) 60 (39%)

127 (67%) 135 (36%) 33 (21%)

*Chi-squareSl.95; P<.OOl. tChi-square=64.29; Pc.001.

TABLE 2 Mean Number of Sites with Bleeding on Probing, by Number of Teeth*

(Adjusted for Age and Sex)

Number Group of Teeth F Compar- (Toothgrp) Mean SE Statistic P-value son P-value

1-10 (1) 2.43 0.46 85.78 .om1 1 vs2 -001 11-24 (2) 6.42 0.35 1 v s 3 .001 25-32 (3) 9.63 0.63 2 v s 3 .001

*ANOVA model with number of sites with bleeding as dependent variable and age, sex, and tooth group as main effects and possible interaction effects.

TABLE 3 Mean Number of Sites with Pocket Depth 24 mm, by Number of Teeth'


Number Group of Teeth F Compar- (Toothgrp) Mean SE Statistic P-value son P-value

1-10 (1) 2.48 0.61 32.41 .001 1 vs2 .oo 1 11-24 (2) 6.96 0.47 1 vs3 .001 25-32 (3) 10.50 0.84 2 v s 3 .om

~ ~~ ~

*ANOVA model with number of sites with pocket depth 24 mm as dependent vanable and age, sex, and tooth p u p as main effects and possible mteraction effects.

TABLE 4 Mean Number of Sites with Loss of Attachment 24 mm, by Number of Teeth'


Number Group of Teeth F Compari- (Toothgrp) Mean SE Statistic P-value son P-value

1-10 (1) 8.04 0.91 31.09 .001 I vs 2 .001 11-24 (2) 16.79 0.70 1 vs 3 .001 25-32 (3) 16.37 1.25 2 v s 3 .77

'ANOVA model with number of sites with loss of attachment 14 mm as dependent variable and age, sex, and tooth group as main effects and possible interaction effects.

Vol. 56, No. 4, Summer 1996 193

sex, and dentition status (toothgrp) were used as the main effects. Interac- tion terms such a s age*sex, age*toothgrp, and sex*toothgrp also were included in the model. In some models main effects other than dentition status (such as sex) and interaction terms (sex‘toothgrp) were found to be significant. However, the purpose of including sex and age in each analysis was primarily to compute age- and sex-adjusted mean scores for the dependent variable for each of the three dentition status groups. Accord- ingly, the significance of age and sex as they relate to each dependent variable has been excluded from the results that are presented here and hence are not discussed.

- Results Number of Teeth and Socioeco-

nomic Status. The distribution of number of teeth was found to be bimo- dal with peaks at 6-7 and 22-25 teeth. The mean number of teeth was 16.9 (SD=8.0). As indicated earlier, subjects were divided into three dentition status groups, i.e., a compromised dentition with 1-10 teeth (mean=6.2), a moderately successful dentition with 11-24 teeth (mean=18.6), and a successful aging dentition with 25-32 teeth (mean=26.9).

Figures 1 and 2 show the relationship of dentition status with education and income, respectively. Only 10 percent of the compromised dentition group had a college degree, whereas in the successful agingdentition group 38 percent of the subjects had a college degree (Figure 1) . Similarly, only 8 percent of subjects in the compromised dentition group were in the income category of r$3o,OOO per year, whereas this number increased to 16 percent in the moderately successful dentition group, and to 34 percent in the successful aging dentition group (Figure 2). Ten to 14 percent of all subjects in these groups refused to report their income.

Table 1 shows data on oral hygiene practices of subjects in the three dentition status categories. Significant differences were found (Pe.001) among the three groups in brushing and floss ing frequencies per day. Forty-six percent of the compromised group reported brushing two or more times each day, 65 percent reported brushing in the moderately successful aging dentition group, and 71 percent in the

TABLE 5 Mean Number of Restored Coronal Surfaces, by Number of Teeth*


Number Group of Teeth F Compari- (Toothgrp) Mean SE Statistic P-value son P-va 1 ue

1-10 (1) 7.5 1.37 178.06 .0001 1 vs2 .001 11-24 (2) 31.8 1.17 1 vs3 -001 25-32 (3) 50.0 1.95 2 vs 3 -001

‘ANOVA model with number of restored coronal surfaces as dependent variable and age, sex, and tooth group as main effects and possible interaction terms.

TABLE 6 Mean Number of Decayed Coronal Surfaces, by Number of Teeth*

(Adjusted for Age and Sex) ~ ~~


1-10 (1) 0.93 0.17 5.25 .ax 1 vs 2 .20 11-24 (2) 1.22 0.14 1 v s 3 .03 25-32 (3) 0.30 0.24 2vs3 .001

*ANOVA model with number of decayed coronal surfaces as dependent variable and age, sex, and tooth group as main effects and possible interaction terms.

TABLE 7 Mean Number of Restored Root Surfaces, by Number of Teeth’



1-10 (1) 0.7 0.19 7.47 .NO6 1 vs 2 .002 11-24 (2) 1.6 0.16 1 vs 3 .014 25-32 (3) 1.5 0.27 2vs3 .68

‘ANOVA model with number of restored root surfaces as dependent variable and age, sex, and tooth group as main effects and possible interaction terms.

TABLE 8 Mean Number of Decayed Root Surfaces, y Number of Teeth+


Number Group of Teeth F Co m pari- (Toothgrp) Mean SE Statistic P-value son

1-10 (1) 1.30 0.20 4.09 .017 1 v s 2 .04 11-24 (2) 0.72 0.1 7 1 vs 3 .007 25-32 (3) 0.30 0.29 2vs3 -22

P-value

‘ANOVA model with number of decayed root surfaces as dependent variable and age, sex, and tooth group as main effects and possible interaction terms.


successful aging dentition group. Similarly, 21 percent of subjects in the successful group never flossed; however, in the compromised group, this proportion was more than three times higher at 67 percent.

Number of Teeth and Oral Health. Age- and sex-ad justed mean scores for the seven oral health measures are presented by dentition status group in Tables 2-8. Table 2 shows that the mean number of bleeding sites per person increased from 2.43 in the compromised dentition group to 9.63 in the successful aging group. All three dentition groups were statistically different from each other (P=.OOl). Simi- larly, Table 3 shows that the mean number of sites per person with pocket depths greater than or equal to 4 mm was largest for the successful aging group (10.5) and smallest for the compromised dentition group (2.5). Again, all dentition groups were significantly different from each other at Pc.001 level (Table 3). The mean loss of periodontal attachment also showed similar relationships (Table 4). The compromised group had significantly fewer sites with loss of attachment than the other groups. How- ever, the moderately successful (mean=16.79) and successful aging (mean=l6.37) groups were not different from each other. Both groups had approximately twice as many sites with loss of attachment of 4 mm or more than the compromised dentition group. Thus, the extent of all three variables related to periodontal disease increased with an increase in number of teeth.

Table 5 provides a comparison of the mean number of restored coronal surfaces in the three dentition status groups and shows that the number of restored surfaces increases dramati- cally with an increase in number of teeth. The mean number of restored coronal surfaces was 7.5 per person for the compromised dentition group, 31.8 for the moderately successful group, and 50.0 for the successful aging group. All three groups were significantly different from each other (P=.OOl). A comparison among the three groups relative to coronal decay, however, shows reverse findings- i.e., the compromised and moderately successful aging dentition groups had significantly higher mean numbers of coronal decayed surfaces at 0.93 and 1.22, respectively, than did the suc-

TABLE 9 Dental Utilization by Number of Teeth

Number of Teeth ~~ ~~~

Dental Utilization 1-10

When did you last <1 year 109 (55%)

>3 years 67 (34%) receive care?* 1-3 years 22 (11 9%)

How often have you Mean 1.11 received care in past SD 1.17 12 months?t

~~

1 1-24 25-32

279 (78%) 136 (89%) 45 (13%) 10 (6%) 31 (9%) 7 (5%) 1.64 1.97 1.12 1.01

*Chi-square=86.15; Pc.001. tOne-way ANOVA; F=27.5, P<.OO1.

TABLE 10 Purpose of Last Visit by Number of Teeth

Number of Teeth

Purpose of Last Visit 1-10

Regular maintenance 51 (26%) Conservative treatment 32 (16%)

Fillings Crown and bridge Root canal

Extractions 34 (17%) Dentures 68 (34%) Other 13 (6%)

11-24 25-32

184 (51 %) 100 (65% ) 81 (23%) 37 (24%)

40 (11%) 5 (3%) 26 (7%) 0 (0%) 27 (8%) 11 (7%)

cessful aging dentition group with 0.30 decayed surfaces (Table 6).

Table 7 shows data on restored root surfaces in all three groups. Similar to coronal surfaces, the number of restored root surfaces was higher for the dentition status groups with more teeth. Finally, a comparison of the three dentition groups for decayed root surfaces is presented in Table 8. These findings are similar to coronal decayed surfaces, with the highest number of decayed surfaces (mean=l.30) in the compromised group. The mean decreased to 0.72 for the moderately successful group and was the least for the successful aging group a t 0.30. The compromised group was significantly different than the other two groups. Although Group 3 had about one-half as many decayed surfaces as Group 2, the dif- ference was not statistically significant.

Number of Teeth and Dental Utili- zation. Tables9 and 10 present data for

dental utilization of subjects in the three dentition status groups. As shown in Table 9,55 percent of subjects in the compromised dentition group reported receiving dental care in the last 12 months, compared to 78 percent in the moderately successful group and 89 percent in the successful aging group. When the mean numbers of visits were compared in the three groups, the successful aging group was found to have visited for dental care approximately twice as many times as did the compromised group (P<.oS). When purpose of last dental visit was compared among the three groups (Table 101, it was found that 65 percent of the subjects with 25-32 teeth visited dentists for regular mainte nance and cleaning, as compared to only 26 percent in the compromised group. Also, percent of subjects seek- ing toothconservative treatment (fillings, root canal, crown and bridge) was higher in subjects with more teeth than the compromised group with

Vol. 56, No. 4, Summer 1996 195

1-10 teeth. Visits for extractions also differed across groups, with 17 percent of those in the compromised dentition group reporting an extraction during the last visit, while only 11 percent of subjects in the moderately successful group and 3 percent in the successful group reported having an extraction during their last dental visit.

Discussion The New England Elder Dental

Study is the only survey conducted with a representative sample of elders in an entire region of the United States. The consistent finding of higher numbers of sites with bleeding, pocket depth, and loss of attachment-a11 be- ing evidence of periodontal dis- easz-for groups with higher numbers of teeth suggests that more teeth are associated with more extensive and severe periodontal disease. Also, a significant increase in restored coronal and root surfaces, evidence of past dental caries, was found in groups who had retained high numbers of teeth compared to those who had not. Thus, in these data, the association between more teeth and more disease is consistent for both dental canes and periodontal disease.

Not only did subjects with higher numbers of teeth show evidence of higher levels of disease, but the utilization of dental services by adults with more teeth suggests that they are visiting dentists more often. Although visit data do not permit us to state that they are, necessarily, using more services, it is clear that the groups with higher numbers of teeth are visiting dentists more often for preventive and restorative treatments such as fillings, root canals, crowns, and bridges. Holtzman et al. (13) and Gift and New- man (14) also reported a positive reia- tionship between dentate status and dental utilization in elders. However, the relationship between tooth retention and increased utilization as shown here may not be directly related and in part may be the function of the sex, education, and income of the subjects.

The successful aging group (25-32 teeth) had the largest proportion of subjects in the highest income and education categories. This group was also the most compliant with recom- mended in-home oral hygiene maintenance (brushing and flossing frequencies). Despite these favorable socio-

economic and healthy behavior attrib- utes, it is instructive to note that 35 percent of subjects in the successful aging group reported that the purpose of their last dental visit was other than regular maintenance. Eighty-nine percent of all subjects in the successful aging group had visited a dentist in the last year. As hypothesized, this is the group that showed the highest levels of periodontal disease and the greatest number of filled surfaces among the three groups.

In contrast to these findings, the successful aging group had a lower level of coronal and root decay. This finding can be explained partly by the inade- quacies of the DFS index and the association of the D component of DFS with tooth loss (9). The relationship between number of teeth and coronal and root caries measured by DFS is distorted by the loss of a tooth because the number of decayed and filled surfaces on that tooth are lost also. In other words, if one looses teeth, the DFS index will not necessarily increase even when new carious lesions occur on the remaining teeth. Root caries has been shown to be a significant predic- tor of tooth loss (15). Moreover, personal dental attitudes are considered an important factor in the decision to extract teeth (15,161.

In a separate analysis, we found that the group with root caries lesions only (D) had fewer teeth (mean=12.81 than the group with no root caries (mean=19.1). Similar findings were noted for coronal surfaces, even though the number of subjects in the coronal "only D category was small because few people are free of coronal restorationsat that age included in this study. In essence, coronal and root caries as measured by DFS will tend to increase with an increase in the number of teeth, provided the teeth are not extracted when the disease oc- curs and thereby removing the evidence of past disease. Our finding that the successful aging group had the least amount of coronal and root decay is similar to that of Locker et al. (In, who reported that "older adults with good oral health had low rates of root decay." We suggest that, except for a minority healthy subgroup, all elders who retain teeth have a similar risk for dental caries; the successfuI aging group seeks dental services and therefore have more teeth and more resto- rations. The compromised group

either leaves the disease untreated or opts for extraction and therefore shows fewer teeth and more decayed surfaces, including a higher percent of teeth with decay (181, in comparison to subjects who have higher numbers of teeth.

This paper answers the question: do older adults with more teeth have more dental caries and periodontal disease? The answer is yes, on aver- age. However, to answer the question whether more teeth in middle age translates into more disease in older age, we need longitudinal data. Cross sectional data can provide only asso- ciations between studied variables, and there is no measure of time se- quence relative to the outcome and associated factors. In other words, the design of this study raises two important issues. First, these data cannot be used to calculate risk of disease. With- out this information we are unable to conclude whether more teeth will translate into more future disease because the amount of disease is a prod- uct of individual risk and number of teeth. If the risk of disease is low, then even higher numbers of teeth at risk may not result in an increase in the absolute burden of disease and the need for more dental services. How- ever, the risk of disease would have to be substantially lower to override a considerable increase in the number of teeth in the population. The conclusions of this study assume that both the compromised and successful aging dentition groups have the same caties or periodontal disease inci- dence, an assumption that might not be true.

Second, from these cross-sectional data one cannot infer whether the compromised dentition is due to defi- cient oral hygiene or the fact that subjects with compromised dentitions have fewer teeth and therefore brush and floss less frequently. Similarly, it cannot be inferred from cross-sectional data whether people who do not visit dentists are healthier (no need for care) or whether they are faltering on prevention and dental care regimens.

The indices for dental caries and periodontal disease used in this analysis have limitations in determining past disease and future services. For example, a small percent of dental res- torations found in this sample could be attributed to overzealous dentistry or a "treatment effect" and may not re-

1% Journal of Public Health Dentistry

flect actual past dental caries. Similar to the filled (F) component of the DFS index, periodontal disease indices such as gingival bleeding, pocket depth,and loss of attachment measure past disease, not current periodontal disease activity (19). Therefore, it is difficult to calculate which sites will deteriorate further and the time-frame for their deterioration. Given these un- certainties, a determination of how many services will be needed is difficult. In addition, for both dental caries and periodontal disease of a given se- verity, many treatment options are available; therefore, translation of disease obtained by caries and periodontal indices into service needs is always difficult. Finally, this study examined only caries and periodontal disease data and their possible implications for treatment needs. No prosthetic treatment needs or prosthesis-in- duced disease such as thoseassociated with ill-fitting dentures were examined. Obviously, a compromised dentition group is expected to have much more need for prosthetic services than those subjects with higher numbers of teeth.

In this report, all analyses have reported the absolute number of sites and surfaces affected by periodontal disease and dental caries, and not the percent of sites and surfaces with disease. This approach is appropriate because the purpose of the study was to determine if there is "evidence of higher dental disease experience in subjects who have retained higher numbers of teeth." Even if the risk of disease is less in the successful aging group, as stated before, it is the risk coefficient multiplied by the number of teeth that determines the amount of disease present.

Policv Imdications The supply of dental services is

heavily dependent upon dental service demands, which in turn are influ- enced, among other factors, by pa- tients' perceptions about oral health and by their ability to afford dental treatment (20-22). The findings of this study have direct implications for cal- culating dental service needs of current and future populations and the personnel to meet those needs. If percent of teeth with disease had been the unit of analysis rather than the total number of sites or surfaces with d i s ease, we expect that we would have

found a higher proportion of diseased teeth in the compromised dentition group. Such findings were reported by Winn et al. (18). However, we believe that the absolute amount of disease present in the population is the true indication of disease burden and thus has serious implications for resources allocation and training of personnel. The significance of this study is that for the first time data are presented to support the "more teeth, more d i s ease" theory. These findings caution us to not ignore the dental needsof the elderly population who are retaining higher numbers of teeth, experiencing more caries and periodontal disease, and requiring more dental services.

The dentist:teeth-at-risk ratio proposed by Douglass and Furino (4) might be relevant for estimating the future needs and demands for dental services. The dentist-to-population ratio does not consider the mean number of teeth in the population. Thus, this index is misleading as an index of treatment need, given the changing demographics and tooth retention patterns in the population and the findings here showing a correlation between the number of teeth at risk and levels of dental disease.

For middle-aged adults and older, the dentist-to-population ratio with the mean or median number of teeth (MT) in the subscript (e.g., 59100,000m:21) is more relevant than a simple dentist-to-population ratio. The dentist-to-population ratio is used for gross comparisons between re- gions and countries and thus fails whenever more precise measures are needed to calculate dental services or personnel. An associated number of teeth parameter adds value to the gross dentist-to-population measure and provides an at-a-glance view of the dentate status of the population, a major basis for a majority of dental services or personnel needs.

Today's young have less disease than yesterday's young, although they have retained more teeth. It is possible that tomorrow's elderly also will have less disease than today's elderly. Re- peated cross-sectional studies can provide information on these trends. Other studies of the relationship between number of teeth and the com- mon oral health parameters discussed in this paper should be conducted. Specifically,existing databases such as the NIDR 1985 Adult Survey (l), the

Iowa Dental Study (15), and the VA Dental Longitudinal Study (23) should be further analyzed to understand the consequences of retaining more teeth. Computer simulation studies involv- ing different levels of risk of disease and numbers of teeth also would be a useful and cost-effective technique to estimate future dental needs. The findings could have major implications for dental personnel training and provi- sion of dental services.

References 1. US Public Health Service, National Insti-

tute of Dental R e a r c h . Oral health of United States adults national findings. NIH pub no 87-2868. Washington, Dc: Government Printing Office, 1987.

2. US Public Health Service, National Cen- ter for Health Statistics. Edentulous persons, United States 1971. DHEW pub no (HRA) 74-1516,series10,no89. Washing- ton, DC Government Frinting Office, 1974.

3. Reinhardt JW, Douglass CW. The need for operative dentistry services: project- ing the effects of changing disease patterns. Oper Dent 1989;14114-20.

4. Douglas CW, Furino A. Balancing dental service requirements and supplies: the epidemiological and demographic evidence. J Am I k n t Assoc 1990;121:587- 92.

5. Douglas CW. Estimating periodontal treatment needs from epidemiological data. J Periodont 1989,60:417-19.

6. Arnlijot HA, Barnes DE, Cohen LK, Hunter PBV, Ship II. Oral health care systems: an international collaborative study. Geneva: Quintessence, World Health Organization, 1991.

7. Douglas CW, Jette A, Tennstedt S, et al. Oral health status of the elderly in New England. J Gerontol Med Sci 1993;48: M39-46.

8. Jette AM, Feldman HA, Tennstedt SL. Tobacco use: a modifiable risk factor for dental disease among the elderly. Am J Public Health 1983;83:1271-6.

9. Josh A, Douglas CW, Jette A, Feldman H. The distribution of mot caries in community-dwelling elders in New England.

10. Joshi A, M c h o t t S, Marcus P, Doug- lass CW, Jette A. Improving epidemiological data collection with computers. J Public Health Dent 1992;52 2328.

11. Cohen J. A coeffiaent of agreement for nominal scales. J Educ Psychol Meas 1960;20..746.

12. Shrout P, Fleiss J. lntraclass correlation: usgin assessingrata reliability. Psychol Bull 1979;86:420-8.

13. Holtzman JM, Berkey DB, Mann J. Pre- dicting utilization of dental services by the aged. J Public Health Dent 1990;50: 16471.

14. Gift HC, Newman JF. How older adults use oral health care services. J Am Dent

15. Hunt RJ, Hand JS, Kohout FJ. Five-year inadence of tooth 1- in Iowans aged 65 and older. Community Dent Oral

J Public Health Dent 1994;54:15-23.

A m 1993;12489-93.

Vol. 56, No. 4, Summer 19% 197

Epidemiol 1991;19:48-51. 76. Burt BA. The oral health of older Ameri-

cans. Am J Public Health 1985;75:11334 17. Locker D, Slade GD, Leake JL. Prevalence

of and factors associated with root decay in older adults in Canada. J Dent R e s

18. Winn DM, Marcus SE, Brown LJ. Oral health problems in remaining dentition of older persons in the United States [Ab- stract]. J Dent Res 1994;73(Spec Iss):372.

19. Socransky SS, Haffape AD, Goodson jM,

1989,@:768-72.

Lindhe J. New mncepts of destructive periodontal disease. J Clin Periodontol 1984;11:21-5.

20. Schonfeld WH, Warren-Hicks D. Esti- mating the need for dental care and dental health manpower in North Carolina. In: Bawden HW, DeFriese H, eds. Plan- ning for dental care on a statewide basis. Chapel Hill, N C Dental Foundation of North Carolina, 1981:115-65.

21. Weintraub JA, Connolly CN, Douglass

CW, Lambert C. Attitudes toward dental health and utilization of dental services by Massachusetts consumers. J Mass Dent Soc 1983322@8.

22. Beck J, Rickey DW. Iowa dental needs survey-an interview,. Iowa Dent J 1982;68:2%6.

23.Chauncy HH, Alman JE, Garcia RI, Douglass CW. The effects of age and dentition status on caries inadence in adults [Abtractl. J Dent Res 1990,69:212.

Consequences of Success: Do More Teeth Translate into More Disease and Utilization?

Documents

Transcript of Consequences of Success: Do More Teeth Translate into More Disease and Utilization?