Association Between Triglyceride-Glucose Index and Development of Asthma in US Adolescents: A Cross-Sectional Study Authors: Chuhan Cheng, Liyan Zhang Published: 2024 ABSTRACT AbstractPurposeMetabolic abnormalities, such as insulin resistance (IR) and dyslipidemia, have been linked to an increased risk of asthma. The triglyceride-glucose index (TyG), a metric indicating metabolic dysfunction, exhibits correlations with metabolic syndrome and IR. However, little research has been conducted on the relationship between TyG and asthma in the pediatric population. Therefore, we aimed to investigate the relationship between TyG and asthma among adolescents.MethodsData from the National Health and Nutrition Examination Survey between 2007 and 2012 was analyzed in this cross-sectional study. The association between TyG and asthma was evaluated using various statistical methods, including multivariate logistic regression analysis, restricted cubic spline (RCS) analysis, threshold effects analysis, and subgroup analysis.ResultsA total of 1,629 adolescent participants were enrolled in the study, consisting of 878 (53.9%) males and 751 females (46.1%), with a mean age of 15.5 years. After adjusting for all covariates in the multivariate logistic regression, the adjusted odds ratio (OR) for TyG and asthma in the highest quintile (Q5, > 8.65) was 4.26 (95% confidence interval [CI], 1.54, 11.81; P = 0.005) compared to the TyG in the second quintile (Q2, 7.68–7.96). Additionally, the multivariate RCS analysis revealed a non-linear relationship between TyG and asthma (P = 0.003). In the threshold analysis, the adjusted OR of asthma was 0.001 (95% CI, 0, 0.145; P = 0.007) in participants with a TyG < 7.78, and the adjusted OR of asthma was 3.685 (95% CI, 1.499, 9.058; P = 0.004) in participants with a TyG ≥ 7.78. Subgroup analysis did not show any interactive role for TyG and asthma.ConclusionsIn US adolescents, a U-shaped association was observed between asthma and the TyG, with a critical turning point identified at around 7.78.Keywords: Dyslipidemia, insulin resistance, metabolic disease, asthma, adolescent INTRODUCTION Asthma, characterized by airway inflammation and hyperresponsiveness, is the most prevalent chronic respiratory condition during childhood. Over recent decades, the prevalence of asthma has experienced a steady rise, placing considerable strain on healthcare systems and adversely impacting the well-being of affected individuals. Among children, the overall prevalence of asthma was reported to be at 9.1%, while among adolescents, it was slightly elevated at 11.0%.1 While numerous risk factors for asthma have been identified, emerging evidence suggests a potential association between metabolic disturbances and the development of asthma. Assessing insulin resistance (IR) and metabolic syndrome in the general population presents challenges due to the invasiveness of the gold standard diagnostic method, the high insulin-normal glucose clamp.2 As a result, alternative laboratory measures have garnered interest, with the triglyceride-glucose index (TyG), which has emerged as a promising biomarker for assessing metabolic syndrome, offering valuable insights into IR, metabolic dysregulation, and systemic inflammation.3,4,5 Derived from fasting glucose and triglyceride levels,3 a higher level of TyG is associated with cardiovascular outcomes and incidence of diabetes.6,7 Notably, TyG has emerged as a superior biomarker compared to the homeostatic model assessment of insulin resistance (HOMA-IR), owing to its simplicity and strong correlation with metabolic abnormalities.8 Although the association between TyG and metabolic disorders is well-established, its potential relationship with asthma remains relatively unexplored, particularly in the adolescent population. This cross-sectional study aims to investigate the association between TyG and asthma among US adolescents, utilizing data from the National Health and Nutrition Examination Survey (NHANES). The hypothesis posits a dose-response connection between asthma and TyG, suggesting that both low and high levels of TyG are linked to an elevated risk of asthma. MATERIALS AND METHODS Data source and study population This cross-sectional study used data from 2007–2012 NHANES, a national survey conducted by the National Center for Health Statistics (NCHS) aimed at evaluating the health and nutritional status of the noninstitutionalized population in the US.9 NHANES gathered demographic data and a diverse range of health-related information through household visits, screening assessments, and laboratory analyses conducted at a mobile examination center. Approval for the NHANES study protocol was granted by the Research Ethics Review Board at the NCHS, and all participants provided written informed consent upon enrollment.10 No further Institutional Review Board approval was necessary for secondary analysis. The present study included individuals aged between 12 and 20 years who completed the survey. Participants who lacked information about asthma, fasting triglyceride, and fasting glucose were eliminated. Asthma assessment To identify participants with asthma, we evaluated their responses to the questions “Has a doctor or other health professional ever told you have asthma?” and “Do you still have asthma?” in the Medical Condition Questionnaire. Only those who answered ‘Yes’ to both questions were classified as having asthma, while those who answered ‘No’ to either question were classified as non-asthmatic. TyG measurement Enzymatic techniques were employed with an autoanalyzer to assess plasma fasting glucose and serum fasting triglyceride levels. TyG was computed using the formula3: Ln [Fasting Glucose (mg/dL) × Fasting Triglycerides (mg/dL)/2]. Participants were then stratified into quintiles (Q1–Q5) according to their TyG levels for analysis. Covariates definitions A variety of covariates were obtained according to prior literature requirements,11 including sex, age, race/ethnicity, body mass index (BMI) category, family income, smoke exposure, family history of asthma, waist circumference, systolic blood pressure (SBP), diastolic blood pressure (DBP), white blood cells (WBCs), eosinophils percent (EOPC), hemoglobin (HGB), high-density lipoprotein cholesterol (HDL-C), glycohemoglobin (HbA1c), diabetes, HOMA-IR, onset age of asthma, and asthma treatment. The categorization of race/ethnicity consisted of 5 groups: Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black, and other races. BMI categories were determined using age- and sex-specific growth charts as advised by the Centers for Disease Control and Prevention.12 These categories were defined as follows: underweight (less than 5th percentile), normal weight (5th to 84.9th percentile), overweight (85th to 94.9th percentile), and obese (more than 95th percentile). The poverty income ratio (PIR) divided family income into 3 categories: low (PIR ≤ 1.3), medium (1.3–3.5), and high (PIR > 3.5).13 Smoke exposure was identified by the level of serum cotinine into 3 groups: non-exposure (< 1 ng/mL), heavy exposure (1–10 ng/mL), and active smokers (> 10 ng/mL).14 HOMA-IR was computed using the formula15: [Fasting Insulin (mU/L) × Fasting Glucose (mmol/L)]/22.5. Subsequently, in subgroup analysis, HOMA-IR was categorized into 2 groups: < 3 denoted normal, while ≥ 3 indicated IR.16 The receipt of asthma treatment was ascertained based on the responses to the question: “During the past 3 months, have you taken medication prescribed by a doctor or other healthcare professionals for asthma?” Those who answered affirmatively were classified as having received asthma treatment, whereas those who responded in the negative were classified as having not received treatment for asthma.17 Multiple single imputation was utilized for 392 participants (2.4%) with missing data pertaining to the mentioned covariates. During each step of the round-robin imputation process, a Bayesian Ridge model was employed as the estimator.18 Statistical analysis The present study is a secondary analysis of a publicly accessible database. Continuous variables are expressed as mean (standard deviation) or median (interquartile range), while categorical variables are presented as percentages. To compare the baseline characteristics among different TyG groups, statistical differences were assessed using t-tests or one-way analysis of variance for continuous variables, and χ2 or Fisher’s exact test for categorical variables. We conducted multivariate logistic regression to assess the odds ratio (OR) and 95% confidence interval (CI) for the association between TyG and asthma. Model 1 was adjusted for sociodemographic factors, comprising sex, age, race/ethnicity, BMI category, and PIR. In Model 2, additional adjustments were made for family history of asthma, smoke exposure, SBP, DBP, WBC, EOPC, and HGB. Lastly, Model 3 included full adjustments with additional variables, such as HDL-C, HbA1c, diabetes status, asthma treatment, and HOMA-IR. We employed restricted cubic splines (RCSs) with 4 knots to explore linearity and evaluate the dose-response relationship between TyG and asthma, while adjusting for covariates in Model 3. Additionally, we conducted a piecewise logistic regression model to assess the threshold association between TyG and asthma, with the aim of identifying specific cut-off points where the risk of asthma undergoes a significant change as a function of varying TyG levels and to ascertain the TyG level that is associated with an elevated risk of asthma, thereby providing guidance for clinical assessments and interventions. Furthermore, to assess the consistency of the association between TyG and asthma across various subgroups, interaction and subgroup analyses were conducted based on sex, BMI category, family asthma history, and HOMA-IR. Finally, several sensitivity analyses were conducted. First, participants with missing covariates were excluded to ensure consistency in the patterns observed in the single imputation analysis, multivariable logistic regression and RCS were utilized to further investigate these trends. Second, in order to evaluate the correlation between TyG and current asthma status, we conducted an analysis to determine the relationship between TyG and forced expiratory volume in the first second (FEV1). Subsequently, the participants were screened for asthma to further investigate the relationship between TyG and the duration of asthma, as well as the receipt of asthma treatment. All statistical analyses were conducted using R Statistical Software (Version 4.2.2; The R Foundation, Vienna, Austria; http://www.R-project.org) and the Free Statistics analysis software (Version 1.9; Beijing Free Clinical Medical Technology Co., Ltd., Beijing, China). Statistical significance was defined as a 2-sided P value of < 0.05. RESULTS Study population Data from 3 NHANES release cycles (2007–2008, 2009–2010, 2011–2012) were extracted for analysis. Out of the 30,442 participants who completed the survey, 17,713 were aged 20 years or older. Among the remaining 12,729 respondents, those with incomplete answers for asthma (1,320), and those with missing data for fasting triglyceride and fasting glucose (n = 9,780) were excluded. In all, a total of 1,629 participants met the inclusion criteria of the present study, of which 189 reported having asthma (Fig. 1). Flowchart of this study. Baseline characteristics We enrolled a total of 1,629 patients with a mean age of 15.5 (± 2.3 years). Among these participants, 29.6% were identified as non-Hispanic white ethnicity and 53.9% were male. The overall prevalence of asthma in the study population was 11.6%, with an onset age of asthma of 5 years. Table 1 presents the demographic characteristics of the participants stratified by quintiles of TyG. The groups categorized based on TyG levels exhibited significant differences in terms of sex, race/ethnicity, BMI category, PIR, smoke exposure, waist circumference, SBP, WBC, HGB, HDL-C, HbA1c, diabetes, and HOMA-IR (all P < 0.05). Conversely, there were no significant variations in the distribution of patient characteristics including age, DBP, and EOPC across the different TyG groups (all P > 0.05). Table 1Baseline characteristics categorized by TyGVariablesTyGP valueTotal (n = 1,629)Q1 (< 7.68) (n = 324)Q2 (7.68–7.96) (n = 317)Q3 (7.97–8.22) (n = 333)Q4 (8.23–8.56) (n = 329)Q5 (> 8.65) (n = 326)Sex0.045Male878 (53.9)157 (49.5)170 (51.1)177 (54.6)176 (53.5)198 (60.7)Female751 (46.1)160 (50.5)163 (48.9)147 (45.4)153 (46.5)128 (39.3)Age (yr)15.5 ± 2.315.5 ± 2.215.5 ± 2.315.3 ± 2.215.6 ± 2.315.7 ± 2.40.151Race/Ethnicity< 0.001Mexican American376 (23.1)62 (19.6)77 (23.1)57 (17.6)87 (26.4)93 (28.5)Other Hispanic192 (11.8)24 (7.6)38 (11.4)31 (9.6)44 (13.4)55 (16.9)Non-Hispanic White482 (29.6)94 (29.7)96 (28.8)62 (19.1)119 (36.2)111 (34)Non-Hispanic Black420 (25.8)106 (33.4)77 (23.1)146 (45.1)48 (14.6)43 (13.2)Other/Multi-racial159 (9.8)31 (9.8)45 (13.5)28 (8.6)31 (9.4)24 (7.4)BMI category< 0.001Underweight45 (2.8)13 (4.1)7 (2.1)10 (3.1)10 (3)5 (1.5)Normal weight959 (58.9)200 (63.1)203 (61)232 (71.6)187 (56.8)137 (42)Overweight289 (17.7)64 (20.2)53 (15.9)50 (15.4)64 (19.5)58 (17.8)Obese336 (20.6)40 (12.6)70 (21)32 (9.9)68 (20.7)126 (38.7)PIR0.030Low (< 1.3)713 (43.8)129 (40.7)125 (37.5)146 (45.1)145 (44.1)168 (51.5)Median (1.3–3.5)589 (36.2)123 (38.8)136 (40.8)114 (35.2)110 (33.4)106 (32.5)High (> 3.5)327 (20.1)65 (20.5)72 (21.6)64 (19.8)74 (22.5)52 (16)Smoke exposure0.041Non-exposure (< 1 ng/mL)1,276 (78.3)261 (82.3)264 (79.3)263 (81.2)255 (77.5)233 (71.5)Heavy exposure (1–10 ng/mL)155 (9.5)25 (7.9)30 (9)31 (9.6)28 (8.5)41 (12.6)Active smokers (> 10 ng/mL)198 (12.2)31 (9.8)39 (11.7)30 (9.3)46 (14)52 (16)Family asthma0.255Yes511 (31.4)104 (32.8)119 (35.7)98 (30.2)98 (29.8)92 (28.2)No1,118 (68.6)213 (67.2)214 (64.3)226 (69.8)231 (70.2)234 (71.8)Waist circumference (cm)81.8 ± 14.778.5 ± 12.381.3 ± 14.276.3 ± 10.282.8 ± 14.489.8 ± 17.4< 0.001SBP (mmHg)109.2 ± 10.5108.4 ± 9.8108.9 ± 10.1108.4 ± 10.3109.3 ± 10.7111.1 ± 11.40.006DBP (mmHg)58.6 ± 12.558.2 ± 12.358.8 ± 12.358.7 ± 11.158.4 ± 12.759.1 ± 13.90.893WBC (1,000 cells/uL)6.4 ± 1.96.0 ± 1.76.3 ± 1.85.8 ± 1.76.7 ± 1.87.3 ± 2.3< 0.001EOPC (%)2.6 (1.7, 4.3)2.6 (1.6, 3.9)2.6 (1.7, 4.3)2.8 (1.7, 4.6)2.4 (1.6, 4.0)2.7 (1.7, 4.2)0.663HGB (g/dL)14.1 ± 1.413.9 ± 1.314.0 ± 1.313.7 ± 1.314.2 ± 1.414.4 ± 1.4< 0.001HDL-C (mg/dL)52.7 ± 12.156.1 ± 11.853.7 ± 11.758.3 ± 11.350.6 ± 10.545.1 ± 10.5< 0.001HbA1c (%)5.3 ± 0.55.2 ± 0.35.3 ± 0.35.2 ± 0.35.3 ± 0.65.3 ± 0.80.007Diabetes< 0.001Yes11 (0.7)0 (0)2 (0.6)0 (0)1 (0.3)8 (2.5)No1,608 (98.7)316 (99.7)329 (98.8)323 (99.7)328 (99.7)312 (95.7)Borderline10 (0.6)1 (0.3)2 (0.6)1 (0.3)0 (0)6 (1.8)HOMA-IR2.7 (1.9, 4.2)2.3 (1.7, 3.4)2.7 (1.9, 4.1)2.2 (1.5, 3.1)2.9 (2.0, 4.3)4.1 (2.7, 6.9)< 0.001Onset age of asthma (yr)5.0 (2.0, 9.0)6.0 (2.0, 10.0)6.0 (2.0, 9.0)6.0 (3.0, 10.0)4.0 (1.0, 8.0)3.0 (1.0, 8.0)< 0.001Asthma treatment0.661Yes128 (7.9)28 (8.8)23 (6.9)21 (6.5)30 (9.1)26 (8)No1,501 (92.1)289 (91.2)310 (93.1)303 (93.5)299 (90.9)300 (92)Asthma0.347Yes189 (11.6)32 (10.1)33 (9.9)36 (11.1)41 (12.5)47 (14.4)No1,440 (88.4)285 (89.9)300 (90.1)288 (88.9)288 (87.5)279 (85.6)Values are presented as number (%), mean ± standard deviation or mean (range).TyG, triglyceride-glucose index; BMI, body mass index; PIR, ratio of income to poverty; SBP, systolic blood pressure; DBP, diastolic blood pressure; WBC, white blood cell; EOPC, eosinophils percent; HGB, hemoglobin; HDL-C, high-density lipoprotein cholesterol; HbA1c, glycohemoglobin; HOMA-IR, homeostatic model assessment of insulin resistance. Baseline characteristics categorized by TyG Values are presented as number (%), mean ± standard deviation or mean (range). TyG, triglyceride-glucose index; BMI, body mass index; PIR, ratio of income to poverty; SBP, systolic blood pressure; DBP, diastolic blood pressure; WBC, white blood cell; EOPC, eosinophils percent; HGB, hemoglobin; HDL-C, high-density lipoprotein cholesterol; HbA1c, glycohemoglobin; HOMA-IR, homeostatic model assessment of insulin resistance. Association between TyG and asthma The univariate logistic regression analysis revealed that race/ethnicity, BMI category, smoke exposure, family history of asthma, WBC, EOPC, diabetes, and HOMA-IR were significantly associated with asthma (Supplementary Table S1). In the multivariate logistic regression analysis, after adjusting for all potential covariates, individuals with the highest TyG values (> 8.65) exhibited a 4.26-fold increased risk of asthma compared to those with lower TyG values in Q2 (7.68–7.96) (OR, 4.26; 95% CI, 1.54, 11.81; P = 0.005) (Table 2). In order to investigate the presence of a dose-response relationship between TyG and the risk of asthma, an RCS analysis was conducted. After controlling for potential confounding factors, a nonlinear association between TyG and asthma was identified (P for nonlinearity = 0.003) (Fig. 2). Notably, in the threshold analysis, the OR of developing asthma was 0.001 (95% CI, 0, 0.145; P = 0.007) among the participants with a TyG < 7.78. When TyG exceeded 7.78, the OR for the risk of developing asthma shifted to 3.685 (95% CI, 1.499, 9.058; P = 0.004). This finding suggests a U-shaped relationship between TyG levels and the incidence of asthma; both low and high levels of TyG were associated with an elevated risk of asthma (Table 3). Table 2Association between triglyceride-glucose index and asthmaQuintilesTotalNo. (%)CrudeModel 1Model 2Model 3OR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P valueQ2 (7.68–7.96)31732 (10.1)1.00 (Reference)1.00 (Reference)1.00 (Reference)1.00 (Reference)Q1 (< 7.68)32436 (11.1)1.11 (0.67, 1.84)0.6761.08 (0.65, 1.81)0.7641.22 (0.7, 2.13)0.4792.61 (0.99, 6.86)0.051Q3 (7.97–8.22)33333 (9.9)0.98 (0.59, 1.64)0.9371.08 (0.64, 1.82)0.7801.05 (0.6, 1.84)0.8531.87 (0.68, 5.1)0.225Q4 (8.23–8.56)32941 (12.5)1.27 (0.78, 2.07)0.3431.45 (0.87, 2.4)0.1531.53 (0.88, 2.64)0.1292.54 (0.94, 6.88)0.066Q5 (> 8.65)32647 (14.4)1.5 (0.93, 2.42)0.0971.63 (0.98, 2.71)0.0611.85 (1.05, 3.25)0.0334.26 (1.54, 11.81)0.005P for trend1,6290.1110.0460.0590.112Model 1 was adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio; Model 2 was adjusted for Model 1 + family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin; Model 3 was adjusted for Model 2 + high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, homeostatic model assessment of insulin resistance.OR, odds ratio; CI, confidence interval. Association between triglyceride-glucose index and asthma Model 1 was adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio; Model 2 was adjusted for Model 1 + family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin; Model 3 was adjusted for Model 2 + high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, homeostatic model assessment of insulin resistance. OR, odds ratio; CI, confidence interval. Association between TyG and asthma odds ratio. Solid and dashed lines represent the predicted value and 95% confidence intervals. They were adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and homeostatic model assessment of insulin resistance. Only 95% of the data is shown. TyG, triglyceride-glucose index. Table 3Threshold effect analysis of the relationship between TyG and asthmaTyGAdjusted modelOR (95% CI)P value< 7.780.001 (0, 0.145)0.007≥ 7.783.685 (1.499, 9.058)0.004Likelihood ratio test< 0.001Adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and homeostatic model assessment of insulin resistance. Only 95% of the data is displayed.TyG, triglyceride-glucose index; OR, odds ratio; CI, confidence interval. Threshold effect analysis of the relationship between TyG and asthma Adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and homeostatic model assessment of insulin resistance. Only 95% of the data is displayed. TyG, triglyceride-glucose index; OR, odds ratio; CI, confidence interval. Stratified analyses based on additional variables and sensitivity analyses Stratified analyses revealed that the no significant interactions across all subgroups when stratified by sex, BMI category, family history of asthma, and HOMA-IR (Supplementary Fig. S1). In addition, sensitivity analyses were performed by excluding participants with missing variables. Results remained constant after adjusting the model for multivariate logistic regression and RCS analysis (Supplementary Table S2, Supplementary Fig. S2). To further assess the robustness of the main findings, we calculated E values to evaluate the potential impact of unmeasured confounding variables. The results indicated that the association between TyG and asthma was robust unless the OR for the risk of asthma due to an unmeasured confounder reached 7.987. Subsequently, we explored the correlation between FEV1 and TyG, thereby evaluating the association between the current severity of asthma and TyG. After adjusting for all covariates in the multiple linear regression analysis, a negative correlation was identified between FEV1 and TyG, with a reduction of 91.62 mL for each unit increase in TyG (β = −91.62; 95% CI, −155.2, −28.04; P = 0.005) (Supplementary Table S3). This finding suggests that elevated TyG levels may be associated with diminished lung function. Moreover, asthmatics were selected for further investigation of the relationship between the TyG and the duration of asthma. After adjusting for all covariates, a positive correlation was observed between the TyG and the duration of asthma (β = 0.39; 95% CI, −1.22, 2; P = 0.634). However, the association was not statistically significant (Supplementary Table S4). We ultimately examined the correlation between asthma treatment status and the TyG within this population. In the multivariate logistic regression analysis, asthmatics with the highest TyG quintile (Q5, > 8.65) exhibited an OR value of 4.09 (95% CI, 1.11, 15.1; P = 0.034), compared to those with lower TyG quintile (Q2, 7.70–7.96) (Supplementary Table S5). These findings suggest that individuals with higher TyG values are less likely to be receiving asthma treatment, indicating a significant relationship between elevated TyG levels and untreated asthma. DISCUSSION This cross-sectional study of American adolescents demonstrated a U-shaped relationship between TyG and asthma in US adolescents, with a turning point of 7.78. These findings suggest that both elevated and diminished TyG levels are linked to an elevated risk of asthma. A grasp of this threshold is of paramount importance for clinicians, as it enables the identification of patients at elevated risk of asthma based on their TyG levels. This, in turn, facilitates the implementation of more targeted preventive measures and treatments. Furthermore, subgroup analysis within the study population did not reveal any significant interactions between different subgroups. Sensitivity analysis revealed that this association remained robust. Additionally, we observed a seemingly contradictory result in the additional analysis, where higher TyG levels were associated with both an increased risk of asthma and a higher likelihood of untreated asthma. This paradox can be explained by considering socioeconomic and healthcare access disparities. Individuals with higher TyG may have poorer overall metabolic health, often linked to lower socioeconomic status and limited healthcare access.19,20 Consequently, this population may have more severe asthma that remains untreated due to barriers in accessing healthcare services.21,22 Over the past few years, there has been a growing emphasis on exploring the correlation between metabolic markers and respiratory conditions, notably asthma. A study conducted by Michael in adults found that IR was common among asthmatics and was associated with a poor response to bronchodilators and corticosteroids.16 Forno et al.23 reported that IR and metabolic syndrome had a significant relationship with worsened lung function in US adolescents, and this effect was even worse among those who had asthma. A Danish cohort study demonstrated IR as an important predictor for the onset of asthma-like symptoms.24 In addition, several previous studies have indicated a correlation between serum lipids and asthma.25,26,27 These studies have suggested a strong association between asthma and metabolic abnormalities. TyG, encompassing both IR and dyslipidemia components, offers a more comprehensive assessment of metabolic dysregulation. Consequently, it holds promise as a valuable marker to explore the association between metabolic dysfunction and asthma. However, the relationship between TyG and asthma has only been explored in a limited number of studies, and the conclusions drawn from these studies have been inconsistent. Staggers et al.28 reported that TyG was associated with the risk of asthma exacerbation utilizing data from the Veterans Health Administration. Wu et al.11 discovered TyG as a potential risk factor for compromised respiratory function, displaying a robust correlation with restrictive spirometry pattern and various respiratory symptoms. Nevertheless, their findings did not reveal a noteworthy association between the index and asthma. This inconsistency with our present study might stem from variations in study populations. While Wu et al.11 concentrated on adults aged ≥ 40 years, our investigation centered on adolescents. In contrast to the prior research, our study unveiled a non-linear correlation between TyG and asthma. We found that both low and high TyG levels were linked to heightened asthma risk. Further multicenter randomized controlled trials are imperative to elucidate the precise impact of TyG on asthma among adolescents in the future. While the precise mechanism underlying the U-shaped relationship between TyG and asthma remains incompletely understood, our findings align with biological plausibility supported by existing evidence. First, elevated TyG levels reflect IR, a condition characterized by impaired insulin action and compensatory hyperinsulinemia. IR may affect airway smooth muscle, leading to increased contractility, proliferation, and remodeling.29 This can result in airway hyperresponsiveness and bronchoconstriction, key features of asthma pathophysiology. In addition, hypoglycemia can induce increased inflammatory cytokines and leukocytosis,30,31 suggesting a link between hypoglycemia and inflammation. This may contribute to airway hyperresponsiveness and worsen asthma control. Furthermore, dyslipidemia, indicated by the triglyceride levels in TyG, may also contribute to asthma risk. Triglycerides may modulate immune cell function and polarization, leading to dysregulated immune responses in asthma.32,33 In conclusion, TyG, serves as a comprehensive biomarker of metabolic dysregulation, has the potential to link metabolic syndrome and asthma. The strengths of our study lie in the utilization of a relatively large population-based cohort, which provided ample statistical power to investigate the relationship between TyG and asthma in adolescents. Additionally, through rigorous statistical analyses, including multivariate logistic regression, RCS analysis, and stratified analysis, the present study offers compelling evidence regarding the correlation between TyG levels and the risk of asthma. However, there are certain limitations to our study. Firstly, despite extensive efforts to adjust for numerous important confounders, the possibility of unmeasured confounders linked to asthma risk, such as allergen, and rhinitis, which were not included in the 2007–2012 NHANES database, cannot be entirely ruled out. Nevertheless, the sensitivity analysis utilizing the E-value methodology yielded E-values higher than the reported strength of associations between allergens and asthma,34,35,36 indicating that allergens are less likely to diminish the association in our study. We acknowledge the potential influence of residual measured or unmeasured confounding on our findings. Furthermore, the cross-sectional design of the study hinders our ability to establish a causal link between TyG and asthma. Lastly, the study was limited to participants of American population, thus the generalizability of our findings to non-American populations may be limited. Consequently, well-designed multicenter randomized controlled trials are imperative to validate our results. In summary, The present study revealed a U-shaped relationship between the TyG and asthma in American adolescents, with a cut-off point of approximately 7.78. The presence of a single cut-off value in a U-shaped curve underscores the importance of considering both ends of the spectrum when assessing asthma risk. This finding highlights the index as a potential epidemiologic biomarker to link metabolic irregularity and asthma within adolescents, as it allows for more precise identification of individuals at higher risk and enables more targeted preventive and therapeutic strategies. Further exploration is warranted to elucidate the intricate relationships between TyG levels and asthma. ACKNOWLEDGMENTS This research was sponsored by Key Clinical Specialty Discipline Construction Program of Fuzhou, Fujian, P.R.C. Footnotes Disclosure: There are no financial or other issues that might lead to conflict of interest. SUPPLEMENTARY MATERIALS Supplementary Table S1 Association of covariates and asthma risk Supplementary Table S2 Association between triglyceride-glucose index and asthma after deletion of missing covariates Supplementary Table S3 Association between forced expiratory volume in the first second and TyG Supplementary Table S4 Association between duration of asthma and TyG Supplementary Table S5 Association between asthma treatment status and triglyceride-glucose index within asthmatics Supplementary Fig. S1 Forest plot of multivariable logistics analysis between triglyceride-glucose index and asthma. Except for the stratification component itself, each stratification factor was adjusted for age, sex, race/ethnicity, BMI category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and HOMA-IR. Supplementary Fig. S2 Association between TyG and asthma odds ratio after deletion of missing covariates. Solid and dashed lines represent the predicted value and 95% confidence intervals. They were adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and homeostatic model assessment of insulin resistance. Only 95% of the data is shown. References Associated Data Supplementary Materials Supplementary Table S1 Association of covariates and asthma risk Supplementary Table S2 Association between triglyceride-glucose index and asthma after deletion of missing covariates Supplementary Table S3 Association between forced expiratory volume in the first second and TyG Supplementary Table S4 Association between duration of asthma and TyG Supplementary Table S5 Association between asthma treatment status and triglyceride-glucose index within asthmatics Supplementary Fig. S1 Forest plot of multivariable logistics analysis between triglyceride-glucose index and asthma. Except for the stratification component itself, each stratification factor was adjusted for age, sex, race/ethnicity, BMI category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and HOMA-IR. Supplementary Fig. S2 Association between TyG and asthma odds ratio after deletion of missing covariates. Solid and dashed lines represent the predicted value and 95% confidence intervals. They were adjusted for age, sex, race/ethnicity, body mass index category, poverty income ratio, family asthma, smoke exposure, systolic blood pressure, diastolic blood pressure, white blood cell, eosinophils percent, hemoglobin, high-density lipoprotein cholesterol, glycohemoglobin, diabetes, asthma treatment, and homeostatic model assessment of insulin resistance. Only 95% of the data is shown.