Effects of individual characteristics on healthy oral mucosa autofluorescence spectra.

Autofluorescence spectroscopy is a tool for detecting tissue alterations in vivo. In a previous study, we found spectral differences between clinically normal mucosa of different patient groups. These are possibly caused by associated patient characteristics. In the present study, we explore the influences of volunteer characteristics on healthy oral mucosa autofluorescence. Autofluorescence spectra were recorded in 96 volunteers with no clinically observable oral lesions. We applied principal components analysis to extract the relevant information. We used multivariate linear regression techniques to estimate the effect of volunteer characteristics on principal component scores. Statistically significant differences were found for all factors but age. Skin color strongly affected autofluorescence intensity. Gender differences were found in blood absorption. Alcohol consumption was associated with porphyrin-like peaks. However, all differences but those associated with skin color were of the same order of magnitude as standard deviations within categories. The effects of volunteer characteristics on autofluorescence spectra of the oral mucosa are measurable. Only the effects of skin color were large. Therefore, in lesion classification, skin color should be taken into account.