Effect of Passive Smoking on Salivary Biomarkers Related to Periodontitis

Objectives: Several reports have documented the harmful effects of passive smoking on periodontal health. However, the association between passive smoking and biomarkers related to development of periodontitis has not been investigated. The present study examined the effect of passive smoking on salivary markers related to periodontitis. Methods: Lifestyle including smoking status was evaluated via a self-administered questionnaire in 273 Japanese workers. Moreover, percentage of pocket depth ≥3.5 mm (%PPD) served as a periodontal parameter. Salivary marker levels were determined by enzyme assay or ELISA. Proportions of six periodontal pathogens in saliva were assessed employing a real-time PCR method. Non-, passive and active smokers were defined as subjects exhibiting salivary cotinine levels of 0, 1-7 and ≥8 ng/ml, respectively. Results: Compared with non-smokers, multiple logistic regression analysis revealed significantly higher odds ratios of periodontitis risk in passive and active smokers after adjustment for other lifestyle factors, and their odds ratios were 3.3 (95%CI; 1.0-10.5) and 5.0 (95%CI; 1.6-15.9), respectively. Levels of salivary markers including IL-1β, albumin, AST and lactoferrin were elevated significantly in passive smokers in comparison to non-smokers (P<0.03); in contrast, no meaningful differences in the levels of markers such as PGE₂, MMP-9 and S-IgA, and proportions of periodontal pathogens were observed. For non- and passive smokers, multiple linear regression analyses were performed for each marker utilizing age, gender, cotinine level and %PPD as independent variables. Salivary cotinine level was independently associated with IL-1β, albumin and AST (P<0.05). Conclusion: We conclude that passive smoke exposure leads to elevation of IL-1β, albumin and AST levels in saliva. This conclusion suggests the involvement of a host-defense mechanism in development of periodontitis as a result of passive smoke exposure. This work was supported by Health and Labour Sciences Research Grants, Japan.