Periodontitis and incident diabetes

Background + Aims

• Diabetes is a growing global health issue, with an estimated 415 million cases in 2015 and projected to rise to 642 million by 2040.
• Type 2 diabetes mellitus (T2DM) accounts for 90% of cases, posing a significant economic burden, particularly in the UK.
• The bi-directional link between diabetes and periodontitis is well-established, with individuals with diabetes having a 3-fold higher risk of developing periodontitis.
• Additionally, periodontitis can impair glycaemic control, with studies showing that successful periodontal treatment can reduce HbA1c levels by around 0.5%.
• While periodontitis is hypothesised as a potential risk factor for developing T2DM due to its inflammatory impact, evidence from population studies remains limited.
• Systematic reviews have identified a small number of studies showing an association between severe periodontitis and incident T2DM, but findings remain inconclusive.
• This study aimed to investigate whether baseline periodontitis increases the risk of developing T2DM in a cohort of older, diabetes-free men in Northern Ireland.

Materials + Methods

• This study used data from the PRIME study, a longitudinal cohort of 2748 men in Northern Ireland aged 50-60, recruited between 1991 and 1994.
• Of the 2010 men who attended a follow-up screening between 2001-2003, 1400 dentate men underwent periodontal examinations.
• Participants with pre-existing or undiagnosed diabetes at baseline were excluded.
• Periodontal assessments measured probing pocket depths (PPD) and clinical attachment levels (CAL) using CDC/AAP classifications to define periodontitis severity.
  • Periodontitis was categorised into no/mild or moderate/severe for analysis.
  • Severe periodontitis was classified as the presence of at ≥2 interproximal sites with CAL ≥6mm on different teeth, along with at ≥1 interproximal site with PPD ≥5mm.
  • Moderate periodontitis was defined as ≥2 interproximal sites with CAL ≥4mm on different teeth, or at ≥2 interproximal sites with PPD ≥ 5mm on different teeth.
• Data collected included demographics, medical history, lifestyle factors, and fasting blood samples for cholesterol and CRP.
• Participants were followed until April 2010, with diabetes diagnoses validated through general practitioners based on WHO guidelines.
• Baseline comparisons used t-tests, chi-square tests, and Mann-Whitney U-tests.
• Kaplan-Meier plots and log-rank tests assessed cumulative diabetes incidence by periodontal status.
• Cox proportional hazards models estimated the risk of incident diabetes, adjusting for confounders in four models: demographics, health measures, socio-economic factors, and CRP levels.
• A dose-response analysis tested the trend in diabetes risk across periodontitis severity levels.
• Statistical significance was set at p<0.05.

Results

• A total of 1331 men were included in the study, with a median follow-up of 7.8 years.
• During this period, 80 men (6.0%) were diagnosed with T2DM and 113 men died.
• Moderate or severe periodontitis was present in 41% of the cohort, with a greater prevalence among those who developed T2DM (53%) compared to those who did not (41%), a significant difference (p = 0.04).
• Men with moderate/severe periodontitis at baseline had fewer teeth, higher CRP levels, greater smoking exposure, more hypertension, lower socio-economic status, and less frequent dental visits compared to those with no/mild periodontitis.
• Kaplan-Meier analysis indicated a higher cumulative incidence of T2DM in men with baseline moderate/severe periodontitis (log rank p = 0.026).
• The unadjusted hazard ratio (HR) for developing T2DM in men with moderate/severe periodontitis was 1.69 (95% CI: 1.07–2.67, p = 0.02), which remained significant after adjusting for age, BMI, smoking, socio-economic factors, and CRP (HR 1.69, 95% CI: 1.06–2.69, p = 0.03).
• A dose-response relationship showed that men with severe periodontitis had a higher risk of T2DM compared to those with no/mild periodontitis (HR 1.85, p = 0.023).
• Age and BMI were also significant predictors of T2DM, with BMI showing the strongest association (HR 1.21 per kg/m², p < 0.001).

Limitations