Relationship between air pollution exposure and insulin resistance in Chinese middle-aged and older populations: evidence from Chinese cohort

Abstract

Aims: This study aimed to determine the relationships between mixed exposure to six air pollutants, namely, particulate matter with an aerodynamic diameter of 2.5 micrometers or less (PM_2.5), PM with an aerodynamic diameter of 10 micrometers or less (PM₁₀), sulfur dioxide (SO₂), nitrogen dioxide (NO₂), cobalt (CO) and ozone (O₃), and insulin resistance (IR) indices in Chinese middle-aged and older populations.

Methods: A total of 2,219 participants from the China Health and Retirement Longitudinal Study (CHARLS), who are followed from 2011 to 2015, were included. Surface air pollutant concentration data were obtained from the China High Air Pollutants (CHAP) database. Multivariable linear regression analysis was used to examine the longitudinal associations between different air pollutants and various IR indices. Additionally, Bayesian kernel machine regression (BKMR), weighted quantile sum (WQS) regression, and quantile-based g computation (Qgcomp) were further utilized to assess the mixed effects of the six air pollutants.

Results: Fully adjusted linear models revealed that increases in the levels of the six air pollutants (in μg/m³) were associated with higher triglyceride-glucose-body mass index (TyG-BMI; Beta = 0.027-0.128), triglyceride-glucose-waist circumference (TyG-WC; Beta = 0.155-0.674), and metabolic score for insulin resistance (METS-IR; Beta = 0.001-0.029) values during the four-year follow-up period. Further mixture analysis indicated that combined exposure to the six air pollutants had a significant cumulative effect on the increases in these three IR indices. Among the pollutants, NO₂ and O₃ were identified as the primary contributor to the cumulative effect. The result of mediation analysis supported the mediating role of BMI in the relationship between air pollution and IR (mediation proportion: 49.1%-93.5%). The results from both subgroup analysis and sensitivity analysis supported the detrimental effects of air pollution on IR.

Conclusion: Both individual and mixed exposures to air pollution were significantly associated with IR in Chinese middle-aged and older individuals, with our study providing new evidence.

Keywords: METS-IR; air pollution; insulin resistance; mixture effect; triglyceride glucose-related indicators.

Figure 1

The geographic distribution of the study population and 6 air pollutants, along with AQI, from 2011 to 2015.

Figure 2

Correlation heatmap of the six air pollutants: The numerical values and color intensity represent the magnitude of the correlation coefficient (r²), with red indicating positive correlation and blue indicating negative correlation.

Figure 3

(a) Relationship between overall air pollution exposure and METS-IR assessed by BKMR model. (b) When other air pollutants were fixed at specific exposure percentiles (25, 50 and 75th), the effect of a particular median air pollutant on METS-IR estimated using BKMR. (c) The weight of each of six air pollutants assessed by WQS model. (d) The dose–response curves between each air pollutant and METS-IR.

Figure 4

(a) Relationship between overall air pollution exposure and TyG-BMI assessed by BKMR model. (b) When other air pollutants were fixed at specific exposure percentiles (25, 50 and 75), the effect of a particular median air pollutant on TyG-BMI estimated using BKMR. (c). The weight of each of six air pollutants assessed by WQS model. (d) The dose–response curves between each air pollutant and TyG-BMI.

Figure 5

(a) Relationship between overall air pollution exposure and TyG-WC assessed by BKMR model. (b). When other air pollutants were fixed at specific exposure percentiles (25, 50 and 75th), the effect of a particular median air pollutant on TyG-WC estimated using BKMR. (c). The weight of each of six air pollutants assessed by WQS model. (d) The dose–response curves between each air pollutant and TyG-WC.