Familial clustering of clinically measured oral health in co-resident three-generation families: a family-linkage study using KNHANES data

Abstract Background Oral diseases cluster in families and follow a strong social gradient, yet how far this familial patterning extends across generations, and whether it reflects shared socioeconomic position, is poorly understood. Most evidence is limited to parents and children and to self-reported oral health. Using clinically measured periodontitis and caries experience, we examined whether clinically measured oral health is concordant between successive generations within co-resident three-generation families, and whether any such association is attenuated after adjustment for the available socioeconomic indicators. Methods Using the Korea National Health and Nutrition Examination Survey (2007–2018), 966 families were linked across three generations, 960 of them among the 2867 households carrying a three-generation composition code; the analysis is therefore not population-representative, and linked and unlinked households differed modestly in income, size and residence. Periodontitis (Community Periodontal Index ≥ 3; grandparent–parent dyad only) and caries experience (decayed, missing and filled teeth, DMFT) were analysed for successive generational dyads by logistic and negative binomial regression with family-clustered standard errors, before and after multivariable adjustment including household income and education; educational mobility was examined as an effect modifier. Results Concordance was strongest in the two adult generations: for total DMFT the parent–offspring association was weaker and compatible with the null, as was a direct grandparent–offspring analysis. In the grandparent–parent dyads, periodontitis in the two generations was associated (adjusted odds ratio 2.57, 95% CI 1.85 to 3.57; adjusted prevalence ratio 1.62 from a log-binomial model; adjusted predicted prevalence 22.9% vs. 38.7%), as was caries experience (count ratio 1.18, 95% CI 1.09 to 1.27 per 10 additional teeth). Neither association was materially attenuated after adjustment for current household income and educational attainment. Because upward mobility was nearly universal (about 79%), the data were largely uninformative about effect modification. Conclusions Clinically measured oral health was associated between successive adult generations of co-resident families, and this clustering was not materially attenuated by current household income or educational attainment. Because the sample comprised co-resident families that differed from those that could not be linked, the findings are not population-representative and are hypothesis-generating; they support further evaluation of family-centred prevention.