Meta-regression Analysis of Study Heterogeneity for Systemic Outcomes after Periodontal Therapy

Abstract

Introduction: The contribution of periodontal disease to adverse systemic consequences remains controversial. This analysis examined 2 well-investigated conditions independently and combined—adverse pregnancy outcomes and glycemic control for patients with diabetes—based on shared pathogenic mechanisms of periodontal infection and inflammation. It was proposed that inconsistencies in study design significantly contribute to outcome discrepancies found between periodontal intervention studies undergoing meta-analysis. Methods: Meta-analyses evaluating periodontal interventions on the rate of preterm birth and changes in glycated hemoglobin A1c in type 2 diabetes populations were conducted based on a systematic review of randomized controlled trials. Meta-regression covariates for exploring heterogeneity included sample size, level of medical management, and bias risk as moderator variables in a random-effects meta-regression. Results: Systematic review identified 17 studies of diabetes and 13 of pregnancy outcomes. Analyses of these studies identified 0.50% reduction in HbA1c and 0.78 odds ratio for preterm births. The heterogeneity associated with the models was high (I2 = 92.4 and I2 = 62.7%, respectively). The adjusted models evaluating each systemic condition separately accounted for 52.2% of the effect for diabetes and 81.4% for pregnancy outcome effects independently, and 63.5% collectively, across interventional studies. Conclusion: This systematic review with meta-regression analysis of heterogeneity demonstrates that disparate results seen in randomized controlled trials of periodontal therapy affecting systemic outcomes may be explained in large part by study design, specifically stringency in consideration of medical management and sample size. The potential for confounding factors to influence outcomes remains a concern in understanding the implications of oral health on systemic conditions. Knowledge Transfer Statement: The findings of this study demonstrate that much of the benefits seen from periodontal therapy on adverse systemic outcomes for diabetes and pregnancy are due to limitations in study design.