Abstract
Objective: Personality characteristics mean that people may interpret similar symptoms differently, complicating the measurement of self-reported oral health, and so we tested the hypothesis that controlling for aspects of personality makes a difference to the association between xerostomia and oral-health-related quality of life.
Material and methods: A cross-sectional study was conducted of a representative adult population sample in Dunedin (New Zealand). Data were collected on xerostomia, OHRQoL and personality characteristics, using (respectively) the 5-item Shortened Xerostomia Inventory (SXI), the OHIP-14 and the Positive and Negative Affect Schedule (PANAS). Negative binomial regression was used to model the association between the SXI and the OHIP-14 scores, and models with and without the PANAS score were compared.
Results: The participation rate was 51.3%, with complete OHIP-14 data available for 250 individuals (56.5% female). The SXI score (mean 6.9, sd 1.8) was strongly and positively associated with the OHIP-14 score (in both models), as was the PANAS negative affect score in the second model, which also explained slightly more of the observed variance than the first model. However, the difference in model deviance fell short of the amount required to reject the hypothesis that adding the PANAS variables to the model made a significant difference.
Conclusions: Considering aspects of personality in investigating OHRQoL remains a theoretically important undertaking, but adjusting for it in analyses of associations between xerostomia and OHRQoL is unlikely to be necessary.