Introduction

            In 1971, BAER described the characteristics of juvenile periodontitis as being rapid progression of the disease, distinctive radiographic pattern of bone loss around molars and incisors, age of onset during early puberty, a 3:1 female/male ratio, familial background, and that the primary dentition was not involved (1). During the years, concepts such as involvement of primary dentition (2,3) and female/male ratio (4,5) have been argued, and new data concerning racial predilection (4-6) and strong immune response against specific bacteria (7) has been added.   

            The disease has been divided in two entities, localized juvenile periodontitis (LJP) (6) and generalized juvenile periodontitis (GJP) (8-10). The major difference between the two forms is the more localized pattern of bone loss in LJP, in contrast to the more generalized pattern of the latter. Disagreement still remains as to whether the two forms of disease are different manifestations of the same disease or two distinct entities.   

            The prevalence of LJP has been reported to be 0.1% to 1% in industrialized countries and somewhat higher in third-world countries (11,12). Also, reports from populations coming from different secluded areas have indicated prevalence higher than 3% (13,14). In particular, studies performed by GJERMO et al. (11) suggested a very high prevalence of juvenile periodontitis in the area of Belo Horizonte, Brazil. Among 304 teenagers, 8 were suggested to have the disease, i.e. a prevalence of 2.63%.   

            Localized juvenile periodontitis has frequently been associated with Actinobacillus actinomycetemcomitans (A.a.), which is a gram-negative, capnophilic, cocobacillary rod (15,16). The prevalence of A.a. in LJP ranges from 89% to 100% (15-17) but still a few studies report no association between A.a. and LJP (18,19). A.a. infection is also known to induce a strong immune response, and LJP patients often show elevated serum antibody titers to the bacterium (7,20). It is not established whether there is a causal relationship between LJP and A.a. or not, but some reports indicate that the cure of the disease is associated with the elimination of the bacterium, and that the recurrence of the disease is associated with the recurrence of detectable levels of A.a. (21,22). In addition, there are reports suggesting a familial aggregation of the disease (16,23).   

            The aim of this study was to assess the prevalence of LJP in four locations in Brazil, including the area previously studied by GJERMO et al. (11). Furthermore we wanted to study the possible association between LJP and presence of A.a. and the distribution of this bacterium among the patients’ family members.    

Materials and methods

The study population

            A total of 7,843 teenagers from different suburban areas in Brazil were screened. The areas were situated within three Brazilian cities: Rio de Janeiro, Belo Horizonte and Votorantim. The two first cities have a population of more than 10 and 5 million, respectively, and Votorantim has more than 500,000 inhabitants. The examinees were 12 - 19 yr old and came from 16 public schools and 2 charity institutions (Table 1). In general, the teenagers lived in areas close to the schools or institutions, and the socio-economic status of the sample population was low, with an average monthly family income of < 300  USD. A possible exception was the school in Votorantim with a slightly better socio-economic situation among its students. There were no marked racial differences between them, and the only recognized difference was the geographical localization.   

Identification of cases   

            Initially, the 7,843 subjects were screened for pockets >5mm adjacent to first molars, resulting in a selection of 119 subjects for further examination. The sensitivity of the screening procedure tested in 83 subjects, not included in the study, was 91% on the subject level. Those with pockets >5mm in any first molar were included in the study as LJP-patients if they, in addition, matched with the following established criteria:    

a) presence of at least one radiographically diagnosed infrabony periodontal lesion involving molars and/or incisors. The lesions had to be confirmed by means of bilateral bitewing radiographs for the posterior area, and periapical radiographs for the front teeth.   

b) clinical loss of attachment > 2mm in at least one site.    

c) no recognized systemic disease.   

d) willingness to sign an informed consent.         

One adult member of each family signed the consent forms.   

Clinical examination   

            The identified patients (25) were submitted to full mouth periapical radiographs and clinical recording of periodontal conditions at 4 sites per tooth: Plaque Index (P.I.)  (24), Gingival Index (G.I.) (25), Probing Periodontal Attachment Level (P.A.L.) (26) and Probing Pocket Depth (P.P.D.)  (26).  

Microbiology   

            All LJP patients, their available family members, and a group of 46 subjects, 12-40 yr, living in the same area as one of the subpopulations from Belo Horizonte and not related to any of the LJP patients, were sampled for microbiological detection of A.a. In order to participate in the microbiological sampling, the subjects should not have had any antibiotic treatment in the last 6 months. One proband had received antibiotic treatment 8 months before the sampling and was thus not excluded from the sampling. All examinations and sampling were performed by one investigator.   

            Microbiologic samples were obtained by inserting sterile paper points to the bottom of all sites that showed radiographic bone loss. After 10 s the points were transferred to tubes containing 10 ml of VMG III solution (27) and processed separately for each site. In the family members, sterile paper points were inserted in all mesial sites of all present teeth and additional distal sites with a probing pocket depth > 5mm for sampling.  After 10 s the points were transferred to the transport medium.  Upper and lower samples were pooled separately. Samples from edentulous or very young family members were taken from the inside aspects of the cheeks and from the tongue with sterile gauze swabs and pooled in separate tubes. All samples were then transported to the laboratory and processed for culturing the same day.    

            The tubes were vortexed at maximal setting for 30 s, and aliquots of 200 ml were seeded on parallel TSBV plates (28). The plates were incubated for 48-72 h in CO₂ gas-pouches (BBL-CO₂ GasPak; Becton Dickinson Microbiology Systems, Cockeysville, USA) at  35-37°C.   

            After 48-72 h, suspected A.a. isolates were harvested upon a preliminary assessment based on colony morphology, Gram stain and catalase test (28). The RNA -23r Gap test (29) was later used to confirm the identification.   

            Patients and family members with clinical and radiographic signs of LJP who were culture negative for A.a. were sampled three times, with intervals of one week between the sampling, to confirm the negative findings.   

Data Analysis

            Prevalence and 99% confidence intervals were calculated (30).   

Results

            The prevalence of LJP in the total screened population sample was 0.3% (Table 2). A separate analysis of the subpopulations showed that the school in Votorantim (referred to as Votorantim) and one charity institution in Belo Horizonte (called B.H.*) yielded higher prevalence values than did the other sampling sites (Table 1). All the other subpopulations of Belo Horizonte yielded prevalence < 0.5%. and were grouped under the name of B.H. The subpopulation from the school in the city of Rio de Janeiro was called Rio (Table 2).   

            A.a. was found in 20 of the 25 patients with LJP, in 35.5% of the parents, and in 43.9% of the siblings. Only in 1 of the 22 families involved, A.a. could not be detected in any family member. The distribution of the LJP patients according to sex, age and presence or absence of detectable A.a. is shown in Table 4, and shows no large variations between the groups. Among non-LJP individuals coming from the same area as the B.H.* subpopulation approximately one third was infected with A.a. (Table 3).   

            In order to study the association between A.a. infection and periodontal disease, the family members of the LJP patients were distributed according to a dichotomization of periodontal status ( > 2 pockets >5mm or < 2 pockets >5mm) and presence or absence A.a. According to such distribution, the presence of A.a. in the family members without LJP diagnosis did not seem to be associated with their periodontal disease status (Table 4).   

            In three families, more than one sibling showed radiographic and clinical signs of LJP, although not always being culture positive for A.a.   

            In one family, a brother (12 yr) and a sister (18 yr) had both the disease and A.a., and in 3 out of 4 families where the probands showed no detectable A.a., at least one member harbored the bacterium.   

Discussion   

            The population examined was a non-random sample from different areas of Brazil and their representativity may be argued. In particular, that would be the case for the sample from Rio. However, the data from that area was used only to calculate the total sample prevalence. No separate conclusions on the prevalence of LJP in Rio were drawn.   

            The sensitivity of the screening procedure for selecting subjects with one or more pockets >5mm adjacent to a first molar (91%) suggests that some cases might have been overlooked and that the real prevalence may be slightly higher than reported.   

            There are differences in reported LJP prevalence, varying with sample size and origin of the population. Although there is a relative consensus and similarity between studies regarding clinical and radiographic characteristics of LJP, the same does not apply to the age limit of the patients. Some studies only include patients up to 19 yr old while others extend this limit to 30 yr of age. Moreover, the studies where the prevalence of the disease has been reported to be higher than 1% have been based on small samples (< 500). This suggests that at least part of the differences in reported LJP prevalence may be due to the different criteria related to age of the populations, and uncertainties related to assessments of prevalence in small populations, in particular when the actual prevalence is low. However, the two subpopulations with high prevalence - Votorantim and B.H.* - were for these kind of studies relatively large, and would support the notion that variations in the prevalence of LJP may exist in different population groups within a limited geographical area as for instance in Belo Horizonte. The subjects from B.H.* all lived in an area secluded from all other participants in Belo Horizonte as far as water supply, housing conditions, infrastructure and demographic profiles are concerned. Similar high prevalence of LJP was reported by GJERMO et al. (11) from an area in Belo Horizonte, which was in part covered by the present study. The former study was performed more then 10 yr ago and under different circumstances, which makes direct comparison of results difficult.   

            In spite of the reported strong association between LJP and A.a. (31), 5 out of 25 LJP patients in the present study were culture negative for the bacterium, even after three separate samplings. Several other studies also report LJP patients who were culture negative for A.a. (31,34,35). This means that either the bacterium is absent or under detectable levels. Some authors have suggested that this can be due to a “burn-out” in the A.a. population when the patients become older, or to a possible bacteriophage infection which may have killed the bacteria (36,37). During the culturing procedures of our wild type strains, we often experienced that we were not able to perform more than two transfers before we lost the bacterium, suggesting bacteriolysis due to phage infection (37). Another explanation for being culture negative and having the disease could be that the bacterium is not implicated in the disease etiology, or that other ‘specific’ organisms may be responsible for the disease picture.   

            Several studies show that A.a. is most frequently found in subjects with age around puberty, but it has also been reported that children and adults can harbor the bacterium (38,39). This is frequently reported in families where one of the members suffers from adult periodontitis (34), or in secluded areas where the prevalence of early-onset periodontitis is high (13,40). In our study, no large variations regarding sex and age were noticed among LJP patients and siblings (Table 5).   

            Although several studies have reported that LJP is strongly associated with A.a., and also that the bacterium is not a part of the common periodontal flora of healthy patients (31,32), there are a few studies which have reported that many people can harbor the bacterium and not present any clinical signs of periodontal disease (33,34). In our study, 52% of A.a. infected siblings, with age between 3 and 30 yr, showed no or minor clinical signs of periodontal disease (<2 pockets >5mm). In addition, the high prevalence of A.a. (35%) in non-LJP patients and its similarity to the numbers shown within the LJP family members (40%) suggests that the differences in the previously reported prevalence variations might be due to differences in the sensitivity of the microbiological techniques applied. Another speculation would be that the prevalence of the bacterium is somewhat higher in some areas than in others. If the latter speculation is true, and assuming that the presence of A.a. increases the risk of developing periodontal disease, populations with a higher prevalence of the bacterium would experience an increased risk of developing periodontal disease.   

            Studies by PREUS et al. (34) and PETIT et al. (41) have suggested a familial aggregation of periodontal disease and a transmission pattern of the bacterium within family members. This could be an explanation for the cases of families having more than one member affected with the disease. The high prevalence of A.a. among families and non-LJP individuals in our study, and the fact that they were not related to each other but came from the same area, may suggest an alternative unknown transmission pattern. Further studies on this subject are needed.      

Acknowledgements: This work was supported by CNPq grant 200136-93/3. We gratefully acknowledge Splice do Brasil, Estacao Saude and Colgate-Palmolive do Brasil for extraordinary support. We also would like to thank Dr. Milton Uzeda and Dr. Marcia Meyer for their valuable help.

Table 1 - The distribution of the sample population according to schools, charity institutions, and number of LJP cases.   

BH - Belo Horizonte   

Vot - Votorantim   

RJ - Rio de Janeiro   

* - Charity institution 

Table 2 - The prevalence of LJP (%) in the studied Brazilian populations.   

** - Confidence Interval at 99% calculated by multiplying the standard error of the frequencies by 2.576.   

Table 3 -The prevalence (%) of subjects with A.a. among the probands, their families and selected non-LJP individuals.   

* The non-LJP subjects consist of a group of individuals, 12-40 yr, coming from the same area as the BH* population.    

Table 4 - The presence of A.a. in patients’ family members according to periodontal disease status.   

Table 5 - Distribution of the LJP patients and siblings according to sex, presence of detectable A.a. and age.