In the present study, patients diagnosed with periodontal disease were 4.4 times (95% CI, 3.74–5.17) to be positively diagnosed with COVID-19 compared to patients without periodontal disease. It is known that periodontal disease shares multiple risk factors with COVID-19; thus, univariate logistic regression was performed to adjust for potential confounders and comorbidities such as age, gender, smoking status, obesity, diabetes, and cardiovascular disease.10,14 The association remained similar and robust, P value < 0.0001, after adjustment for age (OR = 4.34; 95% CI, 3.68–5.09), gender (OR = 4.46; 95% CI, 3.79–5.23), and smoking status (OR = 4.77; 95% CI, 4.04–5.59). The association markedly decreased but remained robust, with P values < 0.0001 after adjusting for race (OR = 2.83; 95% CI, 2.40–3.32), obesity (OR = 2.53; 95% CI, 2.14–2.98), diabetes (OR = 3.32; 95% CI, 2.81–3.90), and cardiovascular disease (OR = 2.68; 95% CI, 2.27–3.14). These findings clearly illustrate that periodontal disease is significantly associated with COVID-19 infection.
Selection of covariates for statistical adjustment was important to ensure limited interference with the odds ratios from characteristics and conditions common to both COVID-19 and PD. Regarding race, studies have shown that African American individuals are at a significantly higher risk of both COVID-19 and periodontal disease.16,17 Similarly, gender has been shown to be a significant risk factor for both conditions.18,19 Males have been shown to have a significantly higher prevalence of periodontal disease and severe COVID-19 cases. Furthermore, many studies have shown that since the start of the pandemic, COVID-19 has severely affected obese individuals and individuals with diabetes.20 Lastly, cardiovascular disease follows the same trend across both conditions. The presence of these covariates in the calculations without adjustment allows introduces confounding elements into the study. Our findings support those of many prior authors showing an association between COVID-19 infection and African American race, male gender, diabetes, obesity, and cardiovascular disease. After adjustment for these proven confounders, we still see a significant association between periodontal disease and COVID-19 infection.
Studies reporting on the association between COVID-19 and periodontal disease are limited. Table 3 reports the currently published odds ratios regarding the interaction between COVID-19 and periodontal disease. All six of the studies identified report a connection between the two conditions whether through an increased risk of COVID-19 infection or increased severity of infection. A study done by Gupta et al. utilized hospital records from Chandigarh, India to report significantly increased risk of COVID-19-related outcomes among patients with PD.21 Marouf et al. reported similar results stating that periodontitis is associated with multiple COVID-19 complications.22 Anand et al. reported that severe periodontitis results in 11.75 increased risk of COVID-19 infection.23 Two studies done by Larvin et al. reported that periodontal disease increased hospital admission, mortality, and when associated with obesity resulted in a 1.33 times higher risk of infection.24,25 A study done early in the pandemic by Katz et al. reported that COVID-19 was not associated with increased COVID-19 risk (OR = 1.1; 95% CI, 0.3–4.2).26
The study done by Marouf et al. utilized a case-control design to investigate whether those individuals with severe periodontitis were more likely to develop COVID-19 related complications.22 This study utilized archived radiographs as the primary form of PD diagnosis. Research has shown that dental radiographs should be used as a guiding instrument in periodontal treatment, but their use as a periodontal diagnosis measure may be misleading.27 The study done by Larvin et al. reported insufficient evidence to link periodontal disease with an increased risk for COVID-19 infection.24 This study utilized self-reported oral health indicators as a surrogate measure for periodontal disease. Self-reported measures such as the ones used in the study by Larvin et al. are subject to bias and bring the validity of the findings into question.
This current study utilized a database of medical charts in a large medical center that allowed for the inclusion of a large sample size which is a strength of this study. To the best of our knowledge, the present study was conducted on the largest database of patient data when compared to the other published investigations on the topic. As depicted in Table 3, half of the currently published studies on the topic utilized a case-control design which inherently limits the capabilities of analyzing a large number of study subjects. The largest of the three studies was done by Larvin et al. which included a total of 1,616 COVID-19 positive patients from the UK Biobank cohort.24 Larvin et al. went on to publish another study utilizing the UK Biobank to perform a retrospective longitudinal which analyzed a sample of 14,466 COVID-19 positive patients. Our utilization of the UF Health i2b2 database allowed for queries of a total patient population of 1,314,925 individuals. The UF Health patient population contained over 38,000 COVID-19 positive patients which significantly increases the validity and power of the findings.
The cross-sectional design of this study does not allow for the assessment of causation. One drawback of this study is the inability to perform multivariate regression. The i2b2 database is a powerful tool for the construction and completion of cross-sectional studies, but the platform does not allow for a query of more than three conditions at a time thus limiting the power of statistical adjustment. Furthermore, the de-identified nature of the i2b2 database does not allow for the assessment of confounders such as socioeconomic status and educational attainment which have been shown to have an association with both COVID-19 and periodontal disease.28–30 In addition, the i2b2 platform does not capture the severity of PD but it categorized the diagnoses in a binary mode of yes/no. These limitations point to the necessity for large-scale longitudinal studies to determine the association between COVID-19 and periodontal disease.
Periodontal disease has been linked to multiple respiratory infections including bacterial pneumonia and chronic obstructive pulmonary disease (COPD) through the alteration of mucosal surfaces in the respiratory tract and the direct aspiration of oral bacterium into the lungs.14 Scannapieco states that dental plaque may also act as a reservoir for respiratory pathogens which may be part of the mechanism for this association between periodontal disease and respiratory infection. Thus, oral hygiene and periodontal disease management is vital for the improvement of the prognosis of respiratory infections (including COVID-19).
In conclusion, the present study has demonstrated that periodontal disease is strongly associated with COVID-19 infection (P < 0.0001), even after controlling for multiple confounders including age, race, gender, and other comorbidities. Future steps include the development of a study to investigate the causation of the relationship between periodontal disease and respiratory diseases. The COVID-19 pandemic will hopefully come and go, but the findings of this study emphasize the importance of a further understanding into the biological mechanisms connecting oral health to respiratory infections.
We would like to thank the UF Health Office of the Chief Data Officer and the University of Florida Integrated Data Repository (IDR) for the curation of the i2b2 data set which was used in this project.