Periodontal status is associated with coronary artery disease defined by myocardial perfusion scintigraphy: a cross-sectional study

Periodontal disease has been increasingly linked to systemic inflammation and cardiovascular disorders. Chronic periodontal inflammation has been proposed to be linked to the development and progression of coronary artery disease, yet clinical evidence linking periodontal parameters with objective myocardial ischemia remains limited. This study aimed to determine whether periodontal status is associated with the presence of coronary artery disease, using myocardial perfusion scintigraphy (MPS) as an objective assessment tool. This cross-sectional study included 272 consecutive adult patients referred for evaluation of suspected myocardial ischemia. Exclusion criteria included prior diagnosis of diabetes mellitus, active or former smoking, obesity (BMI ≥ 30 kg/m²), dyslipidemia, hypertension under pharmacological treatment, a history of coronary artery bypass grafting (CABG), malignancy, chronic kidney disease and liver failure to minimize confounding. MPS using Tc-99 m-sestamibi with SPECT imaging was used to assess coronary artery disease status and left ventricular ejection fraction. Periodontal assessment included plaque index (PI), gingival index (GI), probing depth (PD), clinical attachment loss (CAL), and bleeding on probing. To address near-complete separation between groups on PI and GI, Firth penalized logistic regression was used for CAD analyses, and robust regression was used for ejection fraction analyses, conducted in both the total cohort and separately within each CAD subgroup. Of 272 participants, 150 had coronary artery disease (CAD) and 122 did not. CAD patients were significantly older (median 61.0 [IQR: 54.3–68.0] vs. 58.5 [IQR: 47.0–63.8] years, p < 0.001) and had markedly lower post-stress ejection fraction (median: 51.0% [IQR: 48.0–55.0] vs. 67.0% [IQR: 66.0–69.8], p < 0.0001). All periodontal indices were significantly higher in the CAD group, including PD (5.20 [IQR: 4.30–6.00] vs. 2.10 [IQR: 1.52–2.60] mm) and CAL (3.55 [IQR: 2.56–4.29] vs. 1.04 [IQR: 0.74–1.30] mm; both p < 0.0001). CAL showed the strongest inverse correlation with EF in the total cohort ( r = − 0.609, p < 0.0001), followed by PI ( r = − 0.590), PD ( r = − 0.587), GI ( r = − 0.576), and BOP ( r = − 0.348); however, this association was attributable to between-group compositional differences, within the CAD-negative subgroup, PD showed a modest positive correlation with EF ( r = 0.235, p = 0.009), opposite to the total-cohort direction, and no inverse periodontal–EF association was observed in either subgroup. In Firth penalized logistic regression employed to address near-complete separation between groups on PI and GI, age ( p < 0.001), PI ( p = 0.002 in Model 1; p = 0.008 in Model 2), PD ( p < 0.001 in Model 1; p = 0.037 in Model 3), and CAL ( p < 0.001 in Model 2; p = 0.037 in Model 3) were independently associated with CAD. GI was independently associated in Model 2 only ( p = 0.037); PI and GI were not significant in Model 3. Total-cohort robust regression models showed significant inverse associations between periodontal parameters and EF (R² = 0.43–0.44); however, within-group analyses demonstrated that no consistent inverse periodontal–EF association was present in either subgroup; within the CAD-positive subgroup all periodontal parameters were non-significant (all p > 0.05), and within the CAD-negative subgroup PD showed a modest positive association with EF in Models 1 and 3 (β ≈ 1.61, p ≈ 0.022), opposite to the total-cohort direction, confirming the compositional nature of the pooled association. Compromised periodontal health is significantly associated with CAD prevalence; all periodontal indices were markedly elevated in CAD-positive patients and periodontal parameters were independently associated with CAD status in Firth penalized logistic regression. The total-cohort inverse association between periodontal parameters and post-stress ejection fraction was driven by between-group compositional differences and did not reflect an independent within-person periodontal effect on myocardial function. These results are best interpreted as hypothesis-generating evidence of a periodontal–cardiovascular association, pending confirmation in prospective studies conducted in representative populations.