Serum branched-chain amino acids modifies the associations between air pollutants and insulin resistance

Growing evidence supports that air pollution exposure has become a risk factor of type II diabetes mellitus through the induction of insulin resistance (IR), but the presented findings did not provide a consistent relationship between air pollution exposure and IR in the temporal scale and the magnitude. Reported associated with IR and air pollution exposure, branched-chain amino acids (BCAAs) in blood might modify the association between air pollution exposure and IR. We took advantage of an existing panel study on elderly people who were healthy or with pre-diabetes. Amino acids were analyzed from the serum samples using a UPLC-QQQ-MS, and the homeostasis model assessment of insulin resistance (HOMA-IR) values were calculated to represent the levels of IR in each visit. Exposures to PM2.5, NO2, SO2, CO, O3, and black carbon (BC) were estimated using data from a monitoring station. Linear mixed-effects models were applied to estimate the associations between the air pollution and HOMA-IR, as well as the modifying effects of BCAAs. We found significantly higher concentrations of BCAAs in the pre-diabetic subjects than healthy ones. The concentrations of BCAAs were all significantly associated with HOMA-IR. For subjects with high-level BCAAs, HOMA-IR was positively associated with an IQR increase in PM2.5, NO2, BC, and CO at lag day 2 and in PM2.5, SO2, NO2, BC, and CO at lag day 7. While for subjects with low-level BCAAs, there was no significant association observed at any lag day except for CO at lag day 5. The study provided evidence that circulating BCAAs may modify the relationship between air pollution exposure and the level of insulin resistance in humans. [Display omitted] •Serum BCAAs appeared negatively associated with air pollution exposure.•Serum BCAAs modified the associations between air pollution and insulin resistance.•Serum BCAAs increased the risk of air pollution exposure on insulin resistance.