Evaluating the ecological hypothesis: early life salivary microbiome assembly predicts dental caries in a longitudinal case-control study

Table 3 Odds ratios (ORs) and 95% confidence intervals (CIs) for community state types (CSTs) and future early childhood caries case status from logistic regressions stratified by visit (12- or 24-month visit) among 189 children in a nested case-control study selected from the Center for Oral Health Research in Appalachia 2 cohort study

Characteristic	Unadjusted					Adjusted¹					Sensitivity²
Characteristic	N³	Case N	OR	95% CI	p value	N³	Case N	OR	95% CI	p value	N⁴	Case N	OR	95% CI	p value
12-month visit CST					< 0.001					< 0.001					< 0.001
Gemella ASV2-H. parainfluenzae-Neisseria ASV9	81	28	—	—		81	28	—	—		80	27	—	—
H. parainfluenzae-Neisseria ASV9	9	2	0.54	0.08, 2.42		9	2	0.39	0.05, 2.03		8	2	0.28	0.03, 1.82
Streptoccous ASV1 dominated w/ G. elegans	18	12	3.79	1.32, 11.9		18	12	4.32	1.34, 15.4		17	11	4.93	1.40, 19.1
Streptococcus ASV8-Neisseria ASV12	42	33	6.94	3.02, 17.3		42	33	7.67	2.97, 21.8		40	31	8.18	3.00, 24.7
Neisseria ASV12-Veillonella ASV5	2	2	NR⁵	NR⁵		2	2	--	--		2	2	NR⁵	NR⁵
Streptococcus ASV1 dominated w/ Gemella ASV2	4	3	NR⁵	NR⁵		4	3	--	--		4	3	NR⁵	NR⁵
24-month visit CST					< 0.001					< 0.001					< 0.001
H. parainfluenzae-Neisseria ASV9	67	24	Ref	—		67	24	—	—		66	23	—	—
Streptococcus ASV8-Neisseria ASV12	20	14	4.18	1.48, 13.1		20	14	3.34	1.09, 11.2		20	14	4.93	1.44, 18.8
Neisseria ASV12-Veillonella ASV5	40	30	5.38	2.31, 13.4		40	30	4.71	1.91, 12.4		37	28	7.12	2.57, 22.0
Gemella ASV2-H. parainfluenzae-Neisseria ASV9	6	0	NR⁵	NR⁵		6	0	0.00			6	0	NR⁵	NR⁵

¹ Model adjusted for maternal education, count of emerged primary teeth, breastfeeding, antibiotic exposure within 3 months and visit of case diagnosis
²Model adjusted for all variables listed above + weekly frequency child consumes fruit juice (beverage composed of 100% fruit juice) and brushing/wiping of child’s teeth
³To ensure the salivary bacteriome is prospectively predicting future early childhood caries diagnosis, cases which were diagnosed at 12 months and corresponding controls (N = 6) were excluded from 12-month models. Similarly, cases and diagnosed at 24 months and corresponding controls (N = 37 at 24 months) were excluded from 24-month models. Includes duplicate records for 1 child selected as a control at 36 months and a case at 60 months, and 1 child selected as a control for both 36- and 60-month risk sets. 1 child with missing birth delivery mode excluded. N = 3 children with unassigned CST at 12 months
⁴Sample size as described in footnote 3, but additionally children with missing dietary or oral hygiene data (n = 5 at 12 months, n = 4 at 24 months) were excluded
⁵Because community state type assignment correlated with sample age, some community state types had very small cell counts at the 12- and 24-month visits. If a cell count for a community state type was < 10, we do not report the ORs or 95% CI (not reported, NR) since these estimates are likely unstable. We use the CST with the largest cell count as the reference category in each visit strata