Subgingival temperature Relation to microbial counts
A. D. Haffajee, S. S. Socransky, C. Smith, S. Dibart and J. M. Goodson Forsyth Dental Center, Boston, MA, USA
Haffajee AD, Socransky SS, Smith C, Dibart S and Goodson JM: Subgingival temperature (III). Relation to microbial counts, J Clin Periodontol 1992; 19: 417-422. Abstract. The present investigation examined the relationship of selected bacterial species and subgingival temperature. 35 subjects were measured at 6 sites per tooth for clinical parameters and subgingival temperature. Measurements were repeated for 21 subjects at 2 month intervals providing a total of 66 subjects visits. At each visit, subgingival plaque samples were taken from the mesial aspect of each tooth and anaerobically dispersed, diluted and plated on non-selective media. After anaerobic incubation, colonies were lifted to nylon filters and specific species detected using digoxigenin-labeled whole chromosomal DNA probes. Species enumerated were; A, actinomycetemcomitans serotypes a and b, B, forsythus, B, gingivalis, B. intermedius I and II, C. ochracea, E nucleatum ss. vincentii, P. micros, S. intermedius, S, sanguis I and II, V. parvula and W, recta. Total viable counts and counts of Capnocytophaga sp. were determined directly from the primary isolation plates. A total of 1581 samples were evaluated. Subject visits with higher mean subgingival temperatures had significantly higher mean %s of B, intermedius I and P, micros, and lower mean %s of Capnocytophaga sp. Sites with higher subgingival temperatures had elevated proportions of B, intermedius I and II, A. actinomycetemcomitans serotype a and B, gingivalis more frequently than sites with lower temperatures, while Capnocytophaga sp. were elevated more often at cooler sites. 43 of the subject visits had follow up attachment level measurements at 2 months. The 1026 microbial samples and the subgingival temperature measurements from these visits were related to longitudinal attachment change. Discriminant analysis indicated that combinations of temperature and microbial parameters were more effective in predicting subjects at risk and sites at risk for new attachment loss than either group of variables alone.
Key words: subgingival temperature; periodontitis; bacteria. Accepted for publication 22 Aprii 1991
Previous studies (Rung et al. 1990, Haf- of temperatures observed stimulated marized previously (Haffajee et al. fajee et al. 1992a) demonstrated that speculation that subgingival species 1992a). Data were available from 35 subgingival temperature could be con- may play some role in "causing" the subjects who had at least 1 set of cliniveniently measured using a commer- elevated or depressed temperatures or cal, temperature and microbiological cially available temperature analyzer that different subgingival temperatures measurements. 14 subjects had 1 set, 16 (Periotemp™, ABIODENT^^, Dan- might affect the microbial composition had 2, 2 had 3, 1 had 4 and 2 had 5 vers, MA). The data indicated that there of subgingival plaque and the biological sets of measurements. Thus, data from a was an anterior-posterior temperature activities of the colonizing organisms. total of 66 subject visits were available gradient in subgingival sites and that Temperature changes are more likely for analysis. 43 of the 66 visits had an temperature differed at healthy and dis- mediated by alterations in the local in- additional pair of attachment level eased sites. Further, sites and subjects fiammatory state; however, other expla- measurements taken 2 months later prowith higher temperatures appeared to nations are possible. Thus, the purpose viding longitudinal data for certain be at greater risk for new attachment of the present investigation was to de- analyses. Clinical and temperature loss than sites or subjects with lower termine the relationship of site and sub- measurements were performed as desubgingival temperatures (Haffajee et ject subgingival temperatures to the scribed previously (Haffajee et al. : . ^ , al. 1992b). One surprising finding was levels of selected subgingival bacterial 1992a), species. the wide range of temperatures observed; both in terms of mean "whole Microbiological monitoring mouth" subgingival temperature and Subgingival plaque samples were taken temperature at individual sites. The iVIateriai and Methods Subject population and clinical from the mesial surface of all teeth exmean subgingival temperatures of subcluding third molars (maximum 28 jects differed by as much as 3°C while monitoring site temperatures on occasion could dif- The clinical characteristics of the sub- samples per subject) at each of the 66 fer by as much as 6°C. The wide range jects examined in this study were sum- subject visits. Prior to sample taking.
418
Haffajee et al.
supragingival plaque Was removed by colony morphology) were performed using sterile curettes and discarded. In- on the same plate. The availability of dividual Gracey curettes were used to total viable counts permitted the counts remove subgingival plaque samples of each species to be expressed as % of from the base of the pocket (or sulcus) the total cultivable microbiota. A total to the gingival margin from the mesial of 1581 plaque samples were analyzed aspect of each tooth. The samples were for their microbial content. placed into separate tubes containing 10 ml pre-reduced anaerobically sterilized Data analysis Ringer's solution and immediately were dispersed by sonic oscillation, serially The Kruskal-Wallis test was employed diluted and plated on Trypticase soy to seek differences in the levels of blood agar plates prior to anaerobic in- subgingival species at sites or in subjects cubation (Dzink et al. 1988). Sample grouped on the basis of subgingival taking, dispersion and plating were temperature. Regression analysis was completed within 1 hour to optimize employed to explore the relationships sample recovery. After 7 days, colony between the mean percentages of milifts were prepared and the content of crobial species at the sampled sites in 10 suspected pathogens {Actinobacillus each subject and the outcome variables; actinomycetemcomitans serotypes a & b, sublingual temperature, mean subgingiBacteroides forsythus, Bacteroides gin- val temperature and mean difference begivalis, Bacteroides intermedius I and II, tween subgingival and sublingual temFusobacterium nucieatum ss vincentii, peratures. Spearman rank correlation Peptostreptococcus micros. Streptococ- coefficients were computed for each cus intermedius and Wolinella recta) and temperature-microbial species pair. 4 suspected beneficial species {Capnocy- Log-linear analysis was employed to tophaga ochracea. Streptococcus sanguis examine relationships among 3 variI and II and Veillonella parvula) were ables. Discriminant analysis was emdetermined using digoxigenin-labeled ployed when multiple variables were DNA probes (Gunaratnam et al. 1992). used to "predict" a dichotomous outTotal viable counts and counts of total come variable; i.e. disease activity in a Capnocytophaga species (as determined subject or at a site. Capnocytophaga
36.0
S. intermedius
MEAN SUBGINGIVAL
Relations of microbial counts and mean subject subgingival temperature
%s of Capnocytophaga sp. averaged across all sampled sites in a subject were negatively related to the mean subgingival temperatures of the subject, while mean percentages of B. intermedius I, E nucieatum ss vincentii and P. micros were positively related to this parameter (Fig. 1). The relationships were statistically significant, although considerable variation can be observed in the scatter plots. When subject visits were divided into 3 approximately equal groups according to mean subgingival temperature, the suspected periodontal pathogens B. intermedius I and P. micros were found at significantly higher levels in subjects with higher mean subgingival temperatures, while the suspected beneficial group of Capnocytophaga species were found at higher levels in subjects with lower mean subgingival temperatures (Table 1). The differences in levels of F. nucieatum ss vincentii approached but did not achieve significance. There were no statistically significant correlations between percentages of any of the microbial groups and either sublingual temperature or the difference between the sublingual and the mean of the subgingival temperatures. Relation of microbial counts and subgingival temperature at sites
TEMPERATURE
°C
Results
34.7 -j
33.4
F. nucieatum ss vincentii
36.0 -r
P. micros
0
1
0
s>°
34.7 i
o o
t
^
^
0 0
y = 34.8 + O.lOx rs= 0.31
o
33.4 -
0
3
6
9
12
% OF TOTAL VIABLE COUNT
Eig. 1. Scatter plots of the mean % of the total viable count of 4 species (x-axis) and the mean subgjngival temperature of the subject (y-axis). Each circle represents the mean % of the total cultivable count averaged across all sampled sites (maximum 28 sites) and the mean subgingival temperature averaged across all sites (maximum 168 sites) in a subject.
Microbiological and temperature data were available from 1581 sites taken at 66 subjects visits. Of the 14 species evaluated by DNA probes, 4 showed significant associations with temperature at the site (Table 2). B. intermedius I and II, A. actinomycetemcomitans serotype a and B. gingivalis were present at levels > 1% more often in hotter sites than in cooler sites. Capnocytophaga sp. at levels > 5% were present more frequently in cooler sites than hotter sites. Similarly, significant differences were also observed when the mean %s of the total cultivable microbiota for the same species were compared using the same temperature thresholds. Fig. 2 is an example of the differences observed for B. intermedius. The levels of 5. intermedius I are positively associated with pocket depth (Socransky et al. 1991). Further, pocket depth and temperature have been shown to be positively associated (Kung et al. 1990, Haffajee et al. 1992a). There-
Bacteria and subgingival temperature Table 1. Mean %s of subgingival species at subject visits with mean subgingival temperatures of 35.2°C Mean subgingival temperature °C ;: < 34.6 >=35.2
A. D. Haffajee, S. S. Socransky, C. Smith, S. Dibart and J. M. Goodson Forsyth Dental Center, Boston, MA, USA
Haffajee AD, Socransky SS, Smith C, Dibart S and Goodson JM: Subgingival temperature (III). Relation to microbial counts, J Clin Periodontol 1992; 19: 417-422. Abstract. The present investigation examined the relationship of selected bacterial species and subgingival temperature. 35 subjects were measured at 6 sites per tooth for clinical parameters and subgingival temperature. Measurements were repeated for 21 subjects at 2 month intervals providing a total of 66 subjects visits. At each visit, subgingival plaque samples were taken from the mesial aspect of each tooth and anaerobically dispersed, diluted and plated on non-selective media. After anaerobic incubation, colonies were lifted to nylon filters and specific species detected using digoxigenin-labeled whole chromosomal DNA probes. Species enumerated were; A, actinomycetemcomitans serotypes a and b, B, forsythus, B, gingivalis, B. intermedius I and II, C. ochracea, E nucleatum ss. vincentii, P. micros, S. intermedius, S, sanguis I and II, V. parvula and W, recta. Total viable counts and counts of Capnocytophaga sp. were determined directly from the primary isolation plates. A total of 1581 samples were evaluated. Subject visits with higher mean subgingival temperatures had significantly higher mean %s of B, intermedius I and P, micros, and lower mean %s of Capnocytophaga sp. Sites with higher subgingival temperatures had elevated proportions of B, intermedius I and II, A. actinomycetemcomitans serotype a and B, gingivalis more frequently than sites with lower temperatures, while Capnocytophaga sp. were elevated more often at cooler sites. 43 of the subject visits had follow up attachment level measurements at 2 months. The 1026 microbial samples and the subgingival temperature measurements from these visits were related to longitudinal attachment change. Discriminant analysis indicated that combinations of temperature and microbial parameters were more effective in predicting subjects at risk and sites at risk for new attachment loss than either group of variables alone.
Key words: subgingival temperature; periodontitis; bacteria. Accepted for publication 22 Aprii 1991
Previous studies (Rung et al. 1990, Haf- of temperatures observed stimulated marized previously (Haffajee et al. fajee et al. 1992a) demonstrated that speculation that subgingival species 1992a). Data were available from 35 subgingival temperature could be con- may play some role in "causing" the subjects who had at least 1 set of cliniveniently measured using a commer- elevated or depressed temperatures or cal, temperature and microbiological cially available temperature analyzer that different subgingival temperatures measurements. 14 subjects had 1 set, 16 (Periotemp™, ABIODENT^^, Dan- might affect the microbial composition had 2, 2 had 3, 1 had 4 and 2 had 5 vers, MA). The data indicated that there of subgingival plaque and the biological sets of measurements. Thus, data from a was an anterior-posterior temperature activities of the colonizing organisms. total of 66 subject visits were available gradient in subgingival sites and that Temperature changes are more likely for analysis. 43 of the 66 visits had an temperature differed at healthy and dis- mediated by alterations in the local in- additional pair of attachment level eased sites. Further, sites and subjects fiammatory state; however, other expla- measurements taken 2 months later prowith higher temperatures appeared to nations are possible. Thus, the purpose viding longitudinal data for certain be at greater risk for new attachment of the present investigation was to de- analyses. Clinical and temperature loss than sites or subjects with lower termine the relationship of site and sub- measurements were performed as desubgingival temperatures (Haffajee et ject subgingival temperatures to the scribed previously (Haffajee et al. : . ^ , al. 1992b). One surprising finding was levels of selected subgingival bacterial 1992a), species. the wide range of temperatures observed; both in terms of mean "whole Microbiological monitoring mouth" subgingival temperature and Subgingival plaque samples were taken temperature at individual sites. The iVIateriai and Methods Subject population and clinical from the mesial surface of all teeth exmean subgingival temperatures of subcluding third molars (maximum 28 jects differed by as much as 3°C while monitoring site temperatures on occasion could dif- The clinical characteristics of the sub- samples per subject) at each of the 66 fer by as much as 6°C. The wide range jects examined in this study were sum- subject visits. Prior to sample taking.
418
Haffajee et al.
supragingival plaque Was removed by colony morphology) were performed using sterile curettes and discarded. In- on the same plate. The availability of dividual Gracey curettes were used to total viable counts permitted the counts remove subgingival plaque samples of each species to be expressed as % of from the base of the pocket (or sulcus) the total cultivable microbiota. A total to the gingival margin from the mesial of 1581 plaque samples were analyzed aspect of each tooth. The samples were for their microbial content. placed into separate tubes containing 10 ml pre-reduced anaerobically sterilized Data analysis Ringer's solution and immediately were dispersed by sonic oscillation, serially The Kruskal-Wallis test was employed diluted and plated on Trypticase soy to seek differences in the levels of blood agar plates prior to anaerobic in- subgingival species at sites or in subjects cubation (Dzink et al. 1988). Sample grouped on the basis of subgingival taking, dispersion and plating were temperature. Regression analysis was completed within 1 hour to optimize employed to explore the relationships sample recovery. After 7 days, colony between the mean percentages of milifts were prepared and the content of crobial species at the sampled sites in 10 suspected pathogens {Actinobacillus each subject and the outcome variables; actinomycetemcomitans serotypes a & b, sublingual temperature, mean subgingiBacteroides forsythus, Bacteroides gin- val temperature and mean difference begivalis, Bacteroides intermedius I and II, tween subgingival and sublingual temFusobacterium nucieatum ss vincentii, peratures. Spearman rank correlation Peptostreptococcus micros. Streptococ- coefficients were computed for each cus intermedius and Wolinella recta) and temperature-microbial species pair. 4 suspected beneficial species {Capnocy- Log-linear analysis was employed to tophaga ochracea. Streptococcus sanguis examine relationships among 3 variI and II and Veillonella parvula) were ables. Discriminant analysis was emdetermined using digoxigenin-labeled ployed when multiple variables were DNA probes (Gunaratnam et al. 1992). used to "predict" a dichotomous outTotal viable counts and counts of total come variable; i.e. disease activity in a Capnocytophaga species (as determined subject or at a site. Capnocytophaga
36.0
S. intermedius
MEAN SUBGINGIVAL
Relations of microbial counts and mean subject subgingival temperature
%s of Capnocytophaga sp. averaged across all sampled sites in a subject were negatively related to the mean subgingival temperatures of the subject, while mean percentages of B. intermedius I, E nucieatum ss vincentii and P. micros were positively related to this parameter (Fig. 1). The relationships were statistically significant, although considerable variation can be observed in the scatter plots. When subject visits were divided into 3 approximately equal groups according to mean subgingival temperature, the suspected periodontal pathogens B. intermedius I and P. micros were found at significantly higher levels in subjects with higher mean subgingival temperatures, while the suspected beneficial group of Capnocytophaga species were found at higher levels in subjects with lower mean subgingival temperatures (Table 1). The differences in levels of F. nucieatum ss vincentii approached but did not achieve significance. There were no statistically significant correlations between percentages of any of the microbial groups and either sublingual temperature or the difference between the sublingual and the mean of the subgingival temperatures. Relation of microbial counts and subgingival temperature at sites
TEMPERATURE
°C
Results
34.7 -j
33.4
F. nucieatum ss vincentii
36.0 -r
P. micros
0
1
0
s>°
34.7 i
o o
t
^
^
0 0
y = 34.8 + O.lOx rs= 0.31
o
33.4 -
0
3
6
9
12
% OF TOTAL VIABLE COUNT
Eig. 1. Scatter plots of the mean % of the total viable count of 4 species (x-axis) and the mean subgjngival temperature of the subject (y-axis). Each circle represents the mean % of the total cultivable count averaged across all sampled sites (maximum 28 sites) and the mean subgingival temperature averaged across all sites (maximum 168 sites) in a subject.
Microbiological and temperature data were available from 1581 sites taken at 66 subjects visits. Of the 14 species evaluated by DNA probes, 4 showed significant associations with temperature at the site (Table 2). B. intermedius I and II, A. actinomycetemcomitans serotype a and B. gingivalis were present at levels > 1% more often in hotter sites than in cooler sites. Capnocytophaga sp. at levels > 5% were present more frequently in cooler sites than hotter sites. Similarly, significant differences were also observed when the mean %s of the total cultivable microbiota for the same species were compared using the same temperature thresholds. Fig. 2 is an example of the differences observed for B. intermedius. The levels of 5. intermedius I are positively associated with pocket depth (Socransky et al. 1991). Further, pocket depth and temperature have been shown to be positively associated (Kung et al. 1990, Haffajee et al. 1992a). There-
Bacteria and subgingival temperature Table 1. Mean %s of subgingival species at subject visits with mean subgingival temperatures of 35.2°C Mean subgingival temperature °C ;: < 34.6 >=35.2
