Risk factors associated with tongue lesions: a propensity score-matched case-control study

Background: to analyse the potential risk factors of tongue lesions, comparing the results with a control group. Material and Methods: An observational, case-control study was designed. The study included a case group comprising 336 patients with tongue lesions and 336 sex and age-matched controls. We recorded tobacco/alcohol habits, presence of dentures, allergies, medical conditions, and medications. Statistical analysis was performed via logistic regression models to estimate the odds ratio (OR) adjusted for gender, age, tobacco, and alcohol using propensity score-matching analysis (PSM). Results: According to the final PSM model, patients with tongue lesions were more likely to suffer from allergies (OR 2.13; 1.27-3.66) or medical conditions (OR 2.14; 1.19-3.85), and more likely to take medication (OR 1.99; 1.113.57). Elderly individuals were more prone to hairy tongue (OR 3.82; 1.53-10.47). Smoking was associated with coated tongue (OR 2.05; 1.12-3.63), hairy tongue (OR 3.77; 1.52-9.22) and median rhomboid glossitis (OR 40.49; 5.84-860.43). Allergic individuals were more likely to exhibit sublingual varices (OR 1.73; 1.02-2.88). Medical conditions increased the chances of having coated tongue (OR 2.44; 1.36-4.64) or crenated tongue (OR 2.70; 1.425.30). Arterial hypertension was associated with median rhomboid glossitis (OR 5.85; 1.08-34.18). Individuals on medication showed a higher risk of fissured tongue (OR 1.87; 1.20-2.94) and varices (OR 2.42; 1.58-3.80). Agents acting on the alimentary tract and metabolism increased the probability of fissured tongue (OR 2.31; 1.42-3.79). Conclusions: As far as we are aware, this is the first study on lingual pathology to include a PSM analysis. The results suggest that a history of allergies, the presence of medical conditions, and the use of medication are associated with increased probability of tongue lesions. The analysis of diseases and medications by subgroups requires studies matched by habits with larger sample sizes, in order to corroborate our observations.