Periodontitis and cardiovascular outcome – a prospective follow-up of the PAROKRANK cohort

Background

There is an association between periodontitis (PD) and myocardial infarction (MI). Whether that is related to shared risk factors or it is causal is debated. The Periodontitis and Its Relation to Coronary Artery Disease (PAROKRANK) case-control study previously reported on an independent association between PD and a first MI.

Purpose

This follow-up of the PAROKRANK study tests the hypothesis that PD increases the risk for new cardiovascular (CV) events.

Methods

Between 2010–2014 805 patients (age <75 years; females 19%) with a first MI and 805 controls without MI, matched for gender, age (mean 62±8 years) and living area underwent a CV and dental examination including panoramic x-ray. The PD was categorised in three grades: healthy (≥80% remaining alveolar bone height), moderate (79–66%) or severe (<66%). The composite primary endpoint was the first of all-cause death, non-fatal MI or stroke, or severe heart failure until December 2018. The first of CV-death, non-fatal MI or stroke, or severe heart failure served as a secondary CV-endpoint. Data on outcomes were provided via linkage to the National Patient Registries and the Cause of Death Registry. Cumulative event rates, stratified by PD status at baseline in the combined cohort of cases and controls, were calculated using logistic regression and the Kaplan-Meier method.

Results

A total of 1587 participants with evaluated PD-status were followed for a mean of 6.2 (range 0.2–8.5) years. The total number of primary events and CV-events was 205 and 158 respectively. The number of deaths was 68. Baseline PD-status was healthy in 985 (mean age 60.4 years), moderate in 489 (mean age 65.1 years) and severe in 113 (mean age 64.3 years) participants. The figure presents the time to primary event by the three PD grades in the combined cohort (patients and controls). Time to the primary endpoint differed between the three PD grades (log-rank test 0.0148), however, significant only for patients (log-rank test patients vs. controls: 0.0382 vs. 0.608). Replacing the primary endpoint with the secondary CV-endpoint just changed the outcome slightly (log-rank 0.0976), possibly due to a low number of CV-deaths within the CV-event. Compared to participants without PD, the presence of PD at baseline was associated with the primary endpoint in the total cohort (Odds Ratio (OR): 1.49; 95% Confidence Interval (CI): 1.11–2.00) as well as the CV-endpoint (OR 1.42; 95% CI 1.02–1.98). PD was not associated with total mortality (OR 1.57; 95% CI 0.97–2.56).

Conclusion

In this up to 8-years follow up of the PAROKRANK cohort there was a graded increase in the risk for new CV-events by the presence of PD. This was in particular seen in the MI-patients. Together with the previous case-control based report from PAROKRANK, on an association between PD and a first MI, the findings during the follow-up supports the assumption that there may be a causal relationship between PD and CV-disease.

Funding Acknowledgement

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): The PAROKRANK study was supported by grants from AFA Insurance, Swedish Heart-Lung Foundation, Swedish Research Council, Swedish Society of Medicine, Stockholm County Council (ALF project and Steering committee KI/SLL for odontological research), and The Baltic Child Foundation. Figure 1