To the Editor: Asthma affects over 300 million people globally and imposes substantial burdens. 1 It has a notable hereditary component, with heritability ranging from 70% to 90%. With the emergence of summary statistics from large-scale genome-wide association studies (GWAS), polygenic risk scores (PRSs) are increasingly used to assess the genetic risk of asthma. Obesity and insulin resistance (IR) are prevalent in patients with asthma and are associated with airway hyperreactivity, poor lung function, more asthma attacks, and a weak response to treatment. 2 The triglyceride–glucose (TyG) index, which measures IR using a combination of fasting blood glucose and triglyceride levels, has been associated with severe exacerbation of asthma. However, the relationship between the TyG index and the risk of onset of asthma is unclear. The TyG-body mass index (BMI) index is a new metabolic indicator that considers the relationship between IR and obesity in cardiopulmonary diseases and might be a better predictor. This study aimed to determine if the TyG-BMI value is associated with the risk of asthma onset in a large prospective cohort and to explore its interaction with genetic factors. We analyzed data from the UK Biobank, a cohort of 40–69-year-old UK residents enrolled between 2006 and 2010. After excluding individuals with baseline asthma (n = 56, 515), missing covariates (n = 21, 841), incomplete TyG-BMI data (n = 62, 778), and incomplete genetic information (n = 4710), the final study cohort included 356, 329 individuals. Ethical approval was obtained from the North West Multicenter Research Ethics Committee (No. 21/NW/0157). All participants had provided informed consent. The TyG-BMI index was calculated as TyG × BMI, where TyG = ln (triglycerides mg/dL × fasting glucose mg/dL/2). 3 Participants were stratified into quartiles based on their TyG-BMI value. The genetic risk of asthma was assessed using a validated polygenic risk score (PRSAsthma) derived from UK Biobank GWAS data, with participants categorized into low, intermediate, and high genetic risk tertiles. Incident adult-onset asthma (AOA) was defined via hospital records, self-reports, and death registries (Data-Field 42014). Covariates included age, sex, education, employment, ethnicity, smoking, alcohol consumption, Townsend deprivation index (TDI), eosinophil count, albumin, hemoglobin, hypertension, and history of myocardial infarction. Statistical analyses included Kaplan–Meier survival curves and Cox proportional hazards models to evaluate associations between TyG-BMI, PRSAsthma, and risk of AOA. Two adjusted models were used: model 1 (age, sex) and model 2 (model 1 + socioeconomic and clinical covariates). Nonlinear relationships were examined by restricted cubic spline (RCS) analysis. Stratified analyses assessed effect modification by genetic risk. Additive interactions were evaluated using the “epiR, ” package in R (www. r-project. org). Selection bias in the TyG-BMI groups was minimized by 1: 1 propensity score matching using the “MatchIt” package in R. Sensitivity analyses were adjusted for particulate matter with an aerodynamic diameter <2. 5 µm (PM2. 5), forced expiratory volume in 1 second (FEV1), and vitamin D, and tested correlations between TyG-BMI and inflammatory markers (C-reactive protein, white blood cell count, neutrophil count, lymphocyte count, platelet count). All analyses were conducted in R version 4. 4. 1 (https: //www. r-project. org/), with P <0. 05 considered statistically significant. Over a mean follow-up of 15. 05 years, 10, 123 participants experienced incident AOA. Supplementary Table 1, https: //links. lww. com/CM9/C786 shows the baseline characteristics of the participants according to the TyG-BMI quartile. Participants with high TyG-BMI values were predominantly older, male, white, and had lower educational levels, higher TDI scores, and lower employment rates. They also had higher smoking and alcohol consumption rates and increased peripheral blood eosinophil counts. Furthermore, they had higher BMI, fasting blood glucose and triglyceride levels and were more prone to metabolic comorbidities (P <0. 001). First, we examined whether PRSAsthma was associated with the incidence of AOA. As expected, the distribution of PRSAsthma was shifted to the right in participants with asthma Supplementary Figure 1A, https: //links. lww. com/CM9/C786. The incidence of AOA was higher in participants with the highest PRSAsthma than in those with an intermediate genetic risk, and those with the lowest PRSAsthma had a lower incidence of AOA Supplementary Figure 1B, https: //links. lww. com/CM9/C786; P <0. 0001, log-rank test. After adjustment for covariates, there was a significantly increased risk of AOA in the higher PRSAsthma tertile Supplementary Figure 1C, https: //links. lww. com/CM9/C786. These results justified the inclusion of PRSAsthma as a measure of genetic risk for AOA in subsequent analyses. The risk of incident AOA based on the TyG-BMI quartile was determined by Kaplan–Meier survival analysis Figure 1A. After full adjustment, we found an association between the baseline TyG-BMI value and the risk of incident AOA Figure 1B. The risk of incident AOA increased with increasing TyG-BMI quartile (P for trend <0. 001), indicating a graded association between the TyG-BMI and risk of incident AOA. Furthermore, when TyG-BMI was treated as a continuous variable, a 1-standard deviation increase in TyG-BMI was associated with a 15% greater risk of incident AOA (hazard ratio 1. 15; 95% confidence interval 1. 13–1. 18, P <0. 001). Supplementary Table 2, https: //links. lww. com/CM9/C786 shows the results of the stratified analyses based on genetic risk, which indicate that the relationship between TyG-BMI and the risk of incident AOA is consistent regardless of genetic risk. Furthermore, restricted cubic spline analysis revealed a positive but nonlinear growth relationship between TyG-BMI and the risk of AOA (nonlinear, P <0. 001) Supplementary Figure 2, https: //links. lww. com/CM9/C786. Figure 1: Risk of incident AOA according to baseline TyG-BMI levels and PRSAsthma. (A) Kaplan-Meier analysis of asthma-free probability according to tertiles of TyG-BMI levels. (B) Effect of TyG-BMI on the hazard ratio for AOA after adjustment for age, sex, race, education, employment, smoking status, alcohol status, TDI, eosinophil, hemoglobin, albumin, hypertension, and heart attack. (C) Joint associations of TyG-BMI and PRSAsthma with the risk of incident AOA. Age, sex, race, education, employment, smoking status, alcohol status, TDI, eosinophil, hemoglobin, albumin, hypertension, and heart attack were adjusted. AOA: Adult-onset asthma; CI: Confidence interval; HbA1c: Hemoglobin A1C; HR: Hazard ratio; IR: Insulin resistance; PRS: Polygenic risk score; SD: Standard deviation; TDI: Townsend deprivation index; TyG-BMI: Triglyceride-glucose-body mass index. We then examined the interaction and joint effects of TyG-BMI and PRSAsthma. The joint associations of TyG-BMI and PRSAsthma with the risk of incident AOA are shown in Figure 1C. The risk of incident AOA was significantly higher in participants with a high PRSAsthma and in those with the highest TyG-BMI quartile than in those with a low PRSAsthma and those with the lowest TyG-BMI quartile (hazard ratio 2. 56; 95% confidence interval 2. 30–2. 85, P <0. 001). No significant additive interaction was found between TyG-BMI and genetic risk Supplementary Table 3, https: //links. lww. com/CM9/C786. The association between TyG-BMI and the risk of incident AOA was consistent with the main analysis in the following sensitivity analyses: exclusion of participants who developed incident AOA within the first 3 or 5 years of follow-up Supplementary Table 4, https: //links. lww. com/CM9/C786; additional adjustment for PM2. 5, FEV1, and serum vitamin D level based on the adjustments made in model 2 Supplementary Table 5, https: //links. lww. com/CM9/C786; stratification by age, sex, education, employment, race, alcohol consumption, hypertension, hemoglobin level, and albumin level to assess potential effect modification Supplementary Figure 3, https: //links. lww. com/CM9/C786; correlation between TyG-BMI and risk of AOA with additional adjustment for inflammatory markers based on model 2 Supplementary Table 6, Supplementary Table 7, https: //links. lww. com/CM9/C786; and establishment of a well-matched cohort at baseline by 1: 1 propensity score matching (Q1–Q2 vs. Q3–Q4) with nearest-neighbor matching Supplementary Table 8, https: //links. lww. com/CM9/C786. We investigated the effects of TyG-BMI, the genetic risk of asthma, and their interaction effect on the risk of AOA in the UK Biobank cohort. Most of the previous studies have focused on the individual effects of obesity and IR without considering the interaction effect between genetic risk, obesity, blood lipids and IR on the risk of asthma. Moreover, although previous studies have explored the association between IR indicators (e. g. , TyG, homeostatic model assessment for insulin resistance HOMA-IR, and HbA1c) and symptoms or exacerbations of asthma, they were mainly cross-sectional or focused on patients with asthma, thereby failing to address the need for early prediction of asthma onset. 4 This study suggests that TyG-BMI is associated with the risk of incident AOA regardless of the genetic predisposition to asthma and highlights the potential of TyG-BMI to be an effective predictor of asthma onset, facilitating timely intervention for individuals with elevated levels. Mechanically, the TyG-BMI is closely related to three key characteristics of asthma, namely, persistent inflammation, airway remodeling, and airway hyperreactivity. Notably, TyG-BMI incorporates triglyceride levels, a factor that can regulate immune cell function and polarization, ultimately resulting in perturbed immune responses in asthma. 5 In our study, TyG-BMI was positively correlated with inflammatory markers. It is well known that obesity and hyperinsulinemia contribute to increased airway resistance and remodeling by deposition of adipose tissue and contraction of airway smooth muscle. This study has several limitations. First, it was observational, so the association between TyG-BMI and asthma risk cannot be interpreted as causal. Second, the UK Biobank data are primarily for middle-aged or older European adults, which may limit the generalizability of our results to other populations. Third, we cannot exclude the possibility of unmeasured or residual confounding factors, especially fractional exhaled nitric oxide and serum total and allergen-specific IgE. Fourth, TyG-BMI was only assessed at baseline, and changes over time were not considered. Nevertheless, this study provides compelling evidence that a high TyG-BMI is a reliable early indicator of the risk of AOA. IR, which is associated with poorer asthma outcomes, is a modifiable feature of asthma. 2 Patients with elevated TyG-BMI may reduce their risk of AOA by weight loss, lifestyle improvements, lipid-lowering interventions, and the use of antihyperglycemic agents. Further research is needed to clarify the causal relationship between TyG-BMI and AOA and to define target populations for clinical trials of metabolic intervention to reduce the incidence of asthma. In conclusion, our study suggests that the TyG-BMI has the potential to serve as a risk stratification tool for predicting incident AOA regardless of the genetic risk of asthma. As an easily calculated index, the TyG-BMI offers clinicians a valuable tool for the prediction of asthma risk. Improving IR and reducing blood lipid levels could be a promising strategy for the reduction of the risk of asthma. Acknowledgment We would like to thank the UK Biobank participants and team. This research was conducted under application number 105139. We also thank Liwen Bianji (Edanz) for editing the English text of this manuscript. Funding This study was supported by the National Natural Science Foundation of China (Nos. 82070070, 82270079 and 82300096). Conflicts of interest None.
Yu et al. (Wed,) studied this question.