Positive association of the anti-aging protein α-Klotho with insulin resistance and its inverse L-shaped relationship with glycaemic control in the middle-aged and elderly population.

Abstract

Purpose: α-Klotho has been linked to insulin resistance (IR) in basic research. However, experimental evidence is inconsistent, and there is a lack of data from human research. This study seeks to elucidate the association of α-Klotho with IR in a nationwide, multiracial population.

Methods: A total of 5289 participants aged 40-79 years were included in the National Health and Nutrition Examination Survey (NHANES) spanning 2007-2016. Serum α-Klotho was measured using enzyme-linked immunosorbent assays (ELISA), and IR was evaluated by the homeostatic model assessment of insulin resistance (HOMA-IR). Weighted multivariate logistic and linear regression analysis, subgroup analysis stratified by demographic characteristics, medical condition or obesity status, and sensitivity analysis using propensity score matching (PSM) were performed. Restricted cubic splines (RCS) were performed to explore the nonlinear relationship.

Results: In the fully adjusted logistic regression model, a significant positive association was observed between log-transformed α-Klotho and IR (OR = 3.63, 95% CI: 1.56, 8.45), particularly in males or nonobese individuals (P_interaction < 0.05). In the linear regression model, log₁₀(α-Klotho) was associated with fasting blood glucose (FBG, β = 1.25, 95% CI: 0.74, 1.76) and glycosylated hemoglobin (HbA1c, β = 0.49, 95% CI: 0.20, 0.77). RCS revealed an inverse L-shaped dose-response relationship of α-Klotho with FBG and HbA1c (P_nonlinear <0.05). Beyond the inflection point of log₁₀(α-Klotho) at 2.79, β coefficients sharply rose for these glycaemic control indicators.

Conclusion: The study provides clinical evidence supporting a positive association between α-Klotho and IR. Moreover, the inverse L-shaped relationship suggests that α-Klotho should reach a certain level to predict glycaemic changes effectively.