Variance components of gingival thickness

Authors
Müller HP, Könönen E
Year
2005
Journal
Journal of Periodontal Research, 40, 239-244
DOI
10.1111/j.1600-0765.2005.00798.x
Abstract

OBJECTIVES: Distinct periodontal phenotypes have been identified by cluster analysis, which is an explorative method with very low external validity. The aim of the present study was to investigate variance components of facial gingival thickness in young adults with mild gingivitis.

MATERIAL AND METHODS: Thirty-three non-smoking females, 18-23 years of age, with mild or moderate plaque-induced gingivitis participated. Gingival thickness was measured at every tooth present by use of ultrasound technology to the next 0.1 mm with a lowest measurement of 0.5 mm. Periodontal probing depth and clinical attachment level were measured with a pressure-controlled probe. Gingival bleeding index was assessed after probing on a 0-2 scale, where 1 was slight, and 2 was profuse bleeding on probing. The Silness-Loe plaque index was recorded. Multilevel variance components and random intercept models were built.

RESULTS: A 2-level (subject, tooth) variance component model of gingival thickness without any explanatory variable revealed an intercept (mean) of 0.93 +/- 0.02 mm. Subject variation of gingival thickness amounted to 4.2% of the total variance. Addition of tooth- and subject-related covariates to the model revealed, after adjusting for tooth type, an association with periodontal probing depth (estimated coefficient 0.067 +/- 0.025), and considerable association with average bleeding index (-0.395 +/- 0.149) and plaque index (0.125 +/- 0.048). Variation at the tooth level was drastically reduced; subject variation amounted to 5.2%.

CONCLUSION: Gingival thickness is mainly associated with tooth-related variables. Bleeding tendency is higher if gingiva is thin. Subject variability related to periodontal phenotype may add to the total variance, however, to a very low extent.

Number of levels
2
Model data structure
Response types
Multivariate response model?
No
Longitudinal data?
No
Substantive discipline
Paper submitted by
Hans-Peter Müller, Institute of Clinical Dentistry, Tromsø UNiversity, hans-peter.muller@uit.no
Edit this page