Diagnostic modulation of subgingival proteomic biomarkers by age and smoking habits in periodontitis.

backgroundAlthough age and smoking influence the periodontal proteome, their impact on subgingival biomarkers for diagnosing periodontitis remains unclear. This multicenter study assessed their influence on subgingival proteins for disease detection.

methodsGingival crevicular fluid (GCF) samples were collected from 44 healthy subjects and 40 with periodontitis stages III-IV. Samples were analyzed using sequential window acquisition of all theoretical mass spectra (SWATH-MS). Proteins were identified using UniProt, and diagnostic accuracy was evaluated using generalized additive models (GAM). Models included unadjusted proteins, proteins adjusted for age and smoking (years smoking and pack-year index), besides clinical variables alone. Three-fold cross-validation controlled overfitting.

resultsAge and smoking followed the expected epidemiological distribution, with most controls <45 years (88.6%) and periodontitis patients ≥45 years (80.0%), including higher proportions of smokers (34.1% vs 42.5%) and ex-smokers (4.5% vs 25.0%). Eight proteins were evaluated (GAPDH, keratin, type II cytoskeletal 6A, ZG16B, plasma protease inhibitor C1, carbonic anhydrase 1, and hemoglobin subunits -Hb- alpha, beta, and delta). Unadjusted models showed accuracies (ACC) of 64.3-91.7% (sensitivity/specificity: 42.5-95.0%/68.2-88.6%). Age adjustment consistently improved the diagnostic performance across all proteins (ACC: 88.1-96.4%; sensitivity/specificity: 87.5-97.5%/81.8-97.7%). According to smoking, performance improved for six proteins (81.0-92.9%; 67.5-95.0%/72.7-93.2%) but decreased for GAPDH considering years smoking and pack-year index (71.4-85.7%; 72.5-87.5%/70.5-84.1%) and for Hb beta considering years smoking (72.6%; 57.5%/86.4%).

conclusionAge and smoking affect the diagnostic accuracy of subgingival protein biomarkers for periodontitis. The influence of age is consistent, yielding exceptional predictive values, while the impact of smoking is more variable and biomarker-dependent. However, these results require external validation. PLAIN LANGUAGE SUMMARY: Periodontitis is a chronic inflammatory disease characterized by the progressive destruction of the supporting structures of the teeth. Previous investigations have demonstrated that certain gingival crevicular fluid (GCF) proteins exhibit high diagnostic capability for detecting periodontitis. However, individual risk factors, such as age and smoking habits, may affect the reliability of these protein-based biomarkers. In this multicenter study, we analyzed GCF samples from periodontally healthy and periodontitis subjects using the proteomic technique sequential window acquisition of all theoretical mass spectra (SWATH-MS). We evaluated the diagnostic accuracy of eight proteins, both unadjusted and adjusted for age and smoking habits (years of smoking and cigarette consumption), as well as the clinical variables alone. The results showed that age consistently improved the accuracy of all proteins in distinguishing between periodontal health and periodontitis, achieving outstanding diagnostic performance. In contrast, the impact of smoking status was lower and more protein-specific. Some biomarkers showed increased results in relation to tobacco habits, while others were less accurate when considering this risk factor. These findings highlight the influence of the patient's age and smoking status on the diagnostic accuracy of GCF proteins for detecting periodontitis.