The purpose of the present investigation was to relate clinical characteristics at a site to the frequency of detection, absolute counts and proportions of 14 subgingival species. Subgingival plaque samples were removed by curette from the mesial surface of 2299 teeth in 3 healthy and 87 subjects with periodontal attachment loss. Samples were dispersed,diluted and plated on Trypticase soy agar supplemented with 5% sheep blood. After 7 days of anaerobic incubation, colonies were lifted onto nylon filters, lysed and the DNA fixed to the filters. Digoxygenin-labeled DNA probes were used to identify colonies of each test species. Measurements of pocket depth, attachment level, recession, redness, bleeding on probing and suppuration were made at each sampled site. Total viable counts at sites ranged from 10(3) to > 10(8) and were strongly related to pocket depth. Mean total counts at sites < 3 mm averaged 4.6 x 10(6), while mean counts at sites > 7 mm averaged 2.0 x 10(7). Species enumerated and % of sites colonized were as follows; V. parvula 44; S. sanguis II 36; B. intermedius I 33; C. ochracea 31; B. intermedius II 30; S. sanguis I 29; B. gingivalis 27; S. intermedius 25; P. micros 24; W. recta 23; F. nucleatum ss vincentii 18; B. forsythus 15; A. actinomycetemcomitans serotype a 10; A. actinomycetemcomitans serotype b 8. Counts of B. intermedius II were higher at sites which exhibited gingival redness while B. intermedius I was higher at sites which bled on probing. A. actinomycetemcomitans serotype b was more frequent and at higher mean % at sites without recession. The opposite was true for S. sanguis II. B. gingivalis was somewhat more prevalent and at higher levels at suppurating sites. B. gingivalis, B. intermedius I and II and B. forsythus were found more frequently and at higher levels at sites with deeper pockets, while, V. parvula was less prevalent at sites with pocket depths < 4 mm. B. gingivalis, B. intermedius I and A. actinomycetemcomitans serotype b increased with increasing pocket depth in both localized and widespread disease subjects, but mean counts were higher in the localized disease subjects at any pocket depth. Only W. recta was found at higher levels at deep sites in widespread disease subjects when compared with similar sites in localized disease subjects. No suspected pathogens were detected in 38% of shallow sites, 31% of intermediate sites and 22% of deep sites. 2/3 of deep pockets, but less than 1/2 of shallow pockets harbored at least 2 of the suspected pathogens.
