To investigate the association between incident dementia and previous number of teeth measured over a long interval.DESIGN:
Retrospective analysis of a 37-year cohort study.SETTING:
Prospective Population Study of Women in Gothenburg.PARTICIPANTS:
Women with (n = 158) and without (n = 539) dementia in 2000 to 2005.MEASUREMENTS:
Tooth counts in 1968–69, 1980–81, and 1992–93. Covariates included age, education, stroke, myocardial infarction, diabetes mellitus, smoking status, blood pressure, body mass index, and cholesterol level.RESULTS:
After adjustment for age, odds ratios (ORs) for dementia in 2000–05, comparing first with fourth tooth count quartiles, were 1.81 (95% confidence interval (CI) = 1.03–3.19) for tooth counts measured in 1968, 2.25 (95% CI = 1.18–4.32) for those in 1980, and 1.99 (0.92–4.30) for those in 1992. After further adjustment for education, ORs were 1.40 (95% CI = 1.03–3.19) for 1968, 1.96 (95% CI = 0.98–3.95) for 1980, and 1.59 (95% CI = 0.71–3.53) for 1992, and after additional adjustment for vascular risk factors, ORs were 1.38 (95% CI = 0.74–2.58) for 1968, 2.09 (95% CI = 1.01–4.32) for 1980, and 1.61 (95% CI = 0.70–3.68) for 1992.CONCLUSION:
In most of the analyses, lower tooth count was not associated with dementia, although a significant association was found for one of the three examinations. Further research may benefit from more-direct measures of dental and periodontal disease.