Association between periodontitis and diabetes

Background + Aims

• Periodontitis and type 2 diabetes share inflammatory pathways, but whether periodontal disease contributes to rising blood glucose before diabetes develops remains uncertain.
• Following adults over time can help establish whether periodontal disease is associated with early deterioration in glucose regulation.
• This study aimed to:
  • Investigate whether periodontitis was associated with different patterns of HbA1c between ages 26 and 38, after accounting for other factors influencing metabolic health.

Materials + Methods

• This longitudinal study used the Dunedin birth cohort in New Zealand, with assessments at ages 26, 32 and 38. Periodontal data at all 3 ages were available for 799 participants, with 784 included in fully adjusted analyses.
• Researchers measured HbA1c (an indicator of average blood glucose over approximately 2-3 months).
• Periodontitis was defined as at ≥1 site with clinical attachment loss ≥4 mm, indicating loss of the supporting attachment around a tooth. A stricter definition requiring ≥2 affected sites was also examined.
• Group-based trajectory modelling grouped participants with similar HbA1c patterns over time into low, medium and high trajectories. These were statistical groupings, rather than diagnostic categories for diabetes.
• Analyses accounted for sex, socioeconomic status, smoking and waist-to-height ratio, a measure of abdominal obesity.

Results

• HbA1c increased across all 3 trajectory groups as participants aged.
• Periodontitis prevalence increased from 17.6% at age 26 to 33.9% at age 38, using the one-site definition.
• Periodontitis at age 26 did not significantly predict membership of the highest HbA1c trajectory after adjustment: OR 1.77; 95% CI 0.69–4.52.
• Smoking and abdominal obesity at age 26 were associated with higher odds of belonging to the highest trajectory.
• Periodontitis at ages 32 and 38 was associated with small upward shifts in estimated HbA1c. In the highest trajectory, the adjusted increase was 0.56 mmol/mol (95% CI 0.15–0.97) – approximately 0.05% This was statistically significant, although small in magnitude.
• A separate analysis examining average attachment loss found no association with HbA1c. Using the stricter periodontitis definition or replacing waist-to-height ratio with BMI produced similar overall findings.

Limitations