A U-shaped association between Lipid Accumulation Product and Periodontitis among U.S Adults: A cross-sectional study from NHANES 2009-2014

Aim: This study explores the characterization of a nonlinear dose-response relationship linking lipid accumulation product (LAP), a metabolic dysregulation biomarker, to periodontitis. Methods: This investigation leveraged the NHANES 2009–2014 epidemiological dataset, employing a multi-cycle cross-sectional design. We comprised 4,853 demographically stratified participants with an average age of 51.63 years (SD = 14.35). Periodontitis was defined according to clinical periodontal examinations, and LAP was derived not only from anthropometric indices but also integrated biochemical determinants, specifically waist circumference and fasting serum triglyceride concentrations. This study deployed a generalized additive model to elucidate the nonlinear dose-response interplay between LAP and periodontitis, not only adjusting for sociodemographic confounders (age, gender, race/ethnicity, educational attainment, marital status) but also incorporating socioeconomic determinants (poverty-to-income ratio thresholds) and behavioral-metabolic covariates (BMI, smoking, alcohol use, hypertension, diabetes). A two-piecewise linear regression model was further implemented, detecting the threshold impact of LAP on periodontitis. Results: The study found a nonlinear "U-shaped" relationship between LAP and periodontitis. When LAP was less than 123.49, the odds ratio (OR) for periodontitis was 0.59 (95% CI: 0.35–0.99, p <0.05). However, when LAP was greater than 123.49, the OR for periodontitis significantly increased to 7.69 (95% CI: 1.13–52.34, p <0.05). The threshold effect of LAP on periodontitis was identified at 123.49. Conclusion: This research reveals a nonlinear "U-shaped" association between LAP and periodontitis. Specifically, the risk of periodontitis rises notably when LAP surpasses 123.49. However, the wide CIs (particularly for high LAP) and the cross-sectional design employed in this research restrict our capacity to draw inferences about causality. Also, the possibility of unmeasured confounding factors ought to be taken into account. Long-term longitudinal studies in the future are required to validate these discoveries and delve into the fundamental mechanisms.