The relationship of serum klotho levels and triglyceride glucose index-related indicators Authors: Yaoyao Zhou, Yaqi Wang, Fangli Li, Yiming Shi, Taotao Wu, Yingshuai Li Published: 2024 ABSTRACT AbstractBackgroundKlotho, an anti-aging protein, is linked to energy metabolism. There is limited research on the association of serum klotho and triglyceride glucose (TyG) index-related indicators. Our research aims to investigate the relationship of serum klotho with TyG-BMI (body mass index), TyG-WC (waist circumference), and TyG-WHtR (waist-to-height ratio).MethodsFrom 2007 to 2016, we examined 6,370 participants in the National Health and Nutrition Examination Survey (NHANES). The enzyme-linked immunosorbent assay (ELISA) was utilized to measure serum klotho. We calculated the TyG-BMI, TyG-WC, and TyG-WHtR based on fasting triglycerides, fasting glucose, BMI, WC, and WHtR. Multiple linear regression analysis was used to evaluate the association of serum klotho with TyG-BMI, TyG-WC, and TyG-WHtR. Additionally, generalized additive model (GAM) and smoothing curves were used to evaluate the linear and nonlinear relationships. A piecewise regression model was also utilized to test for threshold effects and determine the breakpoints. Finally, the potential independent associations of serum klotho with TyG-BMI, TyG-WC, and TyG-WHtR were further explored using subgroup analysis.ResultsWe observed a statistically significant difference in serum klotho levels across different quartiles of the population. Based on the multiple linear regression analysis, serum klotho levels were negatively associated with TyG-related indicators. There was a nonlinear relationship between the serum klotho and TyG-BMI, TyG-WC, and TyG-WHtR. The segmented regression analysis revealed that the breakpoints of TyG-BMI, TyG-WC, and TyG-WHtR were 5.42, 6.67, and 1.89, respectively. Subgroup analysis showed that TyG-related indicators interacted with gender and diabetes.ConclusionsIn this study, a negative and nonlinear relationship was identified between serum klotho and TyG-related indicators. Further research is needed to clarify the potential mechanisms that may link serum klotho to TyG-BMI, TyG-WC, and TyG-WHtR.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12944-024-02379-4.Keywords: Klotho, NHANES, Aging, Triglyceride-glucose (TyG) index, Insulin resistance (IR) Introduction Insulin resistance (IR) is a metabolic dysfunction characterized by the body’s reduced responsiveness to insulin, typically associated with metabolic disorders such as diabetes or obesity [1]. Aging is the gradual loss of function and adaptability in an organism, accompanied by decreased cellular function and organ degeneration [2]. Research indicates a close association between IR and aging. Prolonged IR increases oxidative stress and inflammatory responses, damaging cell structure and function, thus accelerating the aging process [3]. Additionally, IR affects the expression of genes related to aging, regulating cell apoptosis, proliferation and repair capacity, thereby influencing the overall pace of biological aging [4]. Conversely, aging can also lead to the onset and exacerbation of IR. As age advances, cells undergo changes that affect their sensitivity to insulin, which can disrupt insulin signaling pathways and contribute to the development of IR [5]. Furthermore, aging may induce changes in organismal energy metabolism, which can further exacerbate the degree of IR [6]. The klotho protein, encoded by the klotho gene, has been shown by recent research to play a significant role in modulating the aging process. Originally found in mice, the absence of the klotho gene in these animals can result in syndromes similar to those seen in human aging, including a shortened lifespan, infertility, skin atrophy, osteoporosis, etc [7]. In healthy individuals, serum klotho levels decrease with age, and this decline is further exacerbated in those with metabolic diseases [8], such as obesity [9], diabetes [1], hypertension [10], hyperlipidemia [11], and metabolic syndrome [12]. As research advances, klotho is gaining prominence for its crucial regulatory function in various physiological processes, including IR, glucose homeostasis, oxidative stress and inflammation, etc [13]. Based on the incorporation of triglycerides and fasting blood glucose (FBG), the triglyceride-glucose (TyG) index serves as a measure of IR and provides insight into an individual’s sensitivity to insulin [14]. Researchers have recently begun exploring the integration of the TyG index with other physiological indicators, such as WC, BMI, WHtR, or other simple, cost-effective, and non-invasive anthropometric measures. Furthermore, research has demonstrated a close correlation between metabolic diseases and IR. There is promising potential in integrating the TyG index with indicators (BMI, WC, WHtR) to enhance the predictive accuracy of IR or other metabolic disorders. For example, Yan et al. [15] proposed that TyG-BMI, TyG-WC, and TyG-WHtR could effectively screen for non-alcoholic fatty liver disease (NAFLD) and metabolic-associated fatty liver disease (MAFLD) at an early stage. Additionally, incorporating the HOMA-IR index allowed for metabolic risk assessment and disease monitoring for individuals with NAFLD. Wang et al. [16] identified that increase in TyG, TyG-BMI and METS-IR were associated with a higher risk of hyperuricemia (HU). Additionally, both TyG-BMI and METS-IR exhibited enhanced discriminatory capabilities for HU across both genders. Li et al. [17] demonstrated that TyG-BMI, TyG-WC, and TyG-WHtR were significant predictors of metabolic syndrome among middle-aged to elderly Chinese people. At present, research on the potential association between serum klotho and TyG-related indicators is limited. Therefore, this study utilized the NHANES database to explore the association between serum klotho and the TyG-related indicators. Materials and methods Data source The research utilized data from the National Health and Nutrition Examination Survey (NHANES), a cross-sectional study conducted by the National Center for Health Statistics (NCHS) and the Centers for Disease Control and Prevention (CDC) [18]. NHANES employed a complex stratified sampling methodology to ensure the representativeness of the sample. NHANES has been approved by the NCHS Ethics Review Board (ERB), with all participants furnishing written informed consent [19]. Data for this study was sourced from https://www.cdc.gov/nchs/nhanes/. Study population Our research incorporated data from five NHANES cycles (2007 to 2016), including demographics data, examination results, laboratory data, and questionnaire data. Overall, NHANES collected data from 50,588 qualified participants from 2007 to 2016. Based on the research criteria, the following participants were excluded: (1) absence of klotho data; (2) missing triglyceride and glucose data; and (3) lacking data on BMI, WC, and body height data. Ultimately, 6,370 participants were included for analysis. The screening flowchart is shown in Fig. 1. Flowchart of participant selection. NHANES: National Health and Nutrition Examination Survey; BMI: body mass index; WC: waist circumference Exposure and outcome variable In this study, the exposure variables included TyG-BMI, TyG-WC, and TyG-WHtR. The calculation procedure is outlined below: TyG = Ln [fasting triglycerides (mg/dL) × fasting glucose (mg/dL)/2] [20]; BMI = body mass (kg)/height2 (m2); WHtR = WC (cm)/height (cm); TyG-BMI = TyG × BMI [21]; TyG-WC = TyG × WC [21]; TyG-WHtR = TyG × WHtR [22]. Serum klotho were the outcome variable. Blood samples were obtained from individuals aged over 40, preserved in dry ice at -80 °C, and then sent to the University of Washington’s Northwest Lipid Metabolism and Diabetes Research Laboratory for examination. Each sample was analyzed twice, and the mean value was used as the final result. More detailed testing methods for serum klotho can be obtained on the NHANES website. Covariables Covariables included demographic information, examination results, laboratory data, and questionnaire data. Demographics information included gender, age, race, educational level, marital status, and poverty income ratio (PIR). Examination data included body mass, WC, body height, and hypertension. Laboratory data included triglycerides, glycosylated hemoglobin, FBG, and serum klotho. Questionnaire data included information on smoking (the average number of cigarettes per day over the past 30 days) and alcohol intake (the average number of drinks per day over the past 12 months). The assessment of hypertension: in a resting state, mean systolic pressure ≥ 140 mmHg or mean diastolic pressure ≥ 90 mmHg on three consecutive measurements, or the use of antihypertensive medications. Diabetes is diagnosed with glycosylated hemoglobin (HbA1c) ≥ 6.5%, FBG ≥ 7.0 mmol/L (126 mg/dL) [23], or the use of hypoglycemic medications. Hyperlipidemia is diagnosed if the total cholesterol ≥ 200 mg/dL, triglyceride ≥ 150 mg/dL, low density lipoprotein ≥ 130 mg/dL, high density lipoprotein ≤ 40 mg/dL for males and ≤ 50 mg/dL for females [24], or the use of cholesterol-lowering medications. Statistical analysis In accordance with CDC guidelines, this study used a complex sampling design to analyze the relationship between serum klotho and TyG-BMI, TyG-WC, and TyG-WHtR. Initially, the study population was divided into four groups based on quartiles (Q1-Q4) of serum klotho. Rao-Scott χ2 and non-parametric tests was employed to assess the statistical differences in the four groups for categorical variables. The continuous variables were presented as mean ± standard error. The missing values were imputed by multiple imputation (e.g., smoking, alcohol intake) (Supplement material S1)) [25]. Considering the skewed distribution, the natural logarithm transformation of serum klotho (Inklotho), TyG-BMI (InTyG-BMI), TyG-WC (InTyG-WC), and TyG-WHtR (InTyG-WHtR) were performed when assessing the relationship between serum klotho and TyG-related indicators. Therefore, three types of weighted linear regression models were used to investigate the relationship of serum klotho with TyG-BMI, TyG-WC, and TyG-WHtR. Subsequently, we generated a smoothed curve using a generalized additive model. Furthermore, a segmented linear regression model was used to assess the potential nonlinear relationship between serum klotho and TyG-related indicators, determine the threshold effects, and calculate the inflection points. Finally, we conducted subgroup analysis to explore the potential factors affecting the association between serum klotho and TyG-related indicators. The statistical analysis was conducted using R software (version 4.3.1) and SPSS 27.0. A P-value less than 0.05 was considered statistically significant. Results Participant baseline characteristics Our research included a sample of 6,370 individuals. We observed that gender, age, race, educational level, marriage status, PIR, smoking, alcohol intake, hypertension, diabetes, hyperlipidemia, BMI, WC, and height were statistically significant in serum klotho at different quartiles. With the increase in serum klotho concentration, BMI, WC, height, TyG-BMI, TyG-WC, and TyG-WHtR showed a decreasing trend (Table 1). Table 1Baseline characteristics of participants in the NHANES from 2007–2016CharacteristicKlotho (pg/ml)P-valueOverall(n = 6370)Q1(< 657.20)(n = 1593)Q2(657.20-807.40)(n = 1593)Q3(807.40-1002.90)(n = 1593)Q4(> 1002.90)(n = 1591)Gender< 0.001Male3066 (47.36%)846 (12.69%)820 (13.51%)733 (11.50%)667 (9.67%)Female3304 (52.64%)747 (12.43%)773 (12.85%)860 (13.68%)924 (13.68%)Age< 0.00140–603714 (65.67%)859 (15.62%)930 (17.31%)922 (16.19%)1003 (16.53%)> 602656 (34.33%)734 (9.50%)663 (9.04%)671 (8.99%)588 (6.81%)Race< 0.001Mexican American990 (6.46%)247 (1.55%)250 (1.61%)254 (1.70%)239 (1.60%)Non-Hispanic Black2805 (72.53%)749 (18.61%)754 (19.84%)712 (18.48%)590 (15.59%)Non-Hispanic White1198 (9.48%)293 (2.31%)241 (1.93%)259 (2.07%)405 (3.18%)Other1377 (11.54%)304 (2.66%)348 (2.98%)368 (2.93%)357 (2.97%)Educational level< 0.001Below high school1785 (16.71%)450 (3.97%)449 (4.50%)444 (4.07%)442 (4.17%)High school or above4585 (83.29%)1143 (21.15%)1144 (21.86%)1149 (21.11%)1149 (19.17%)Marriage status< 0.001Married/Living with partner4140 (71.22%)1033 (17.55%)1035 (19.01%)1048 (17.88%)1024 (16.78%)Widowed/Separated/Divorced/Never married2230 (28.78%)560 (7.58%)558 (7.35%)545 (7.30%)567 (6.56%)PIR (%)3.24 ± 0.003.24 ± 0.003.23 ± 0.003.25 ± 0.003.25 ± 0.00< 0.001Smoke (cigarettes/day)(5, 20)(5, 20)(5, 20)(5, 20)(5, 20)0.834Alcohol intake (alcoholic drinks/day)(1, 3)(2, 4.5)(2, 5)(1, 5)(1, 4)< 0.001Hypertension1416 (18.70%)366 (4.67%)356 (5.23%)324 (4.35%)370 (4.45%)< 0.001Diabetes1271 (14.99%)305 (3.47%)282 (3.71%)298 (3.69%)386 (4.13%)< 0.001Hyperlipidemia4378 (68.96%)1107 (17.84%)1110 (18.62%)1095 (17.41%)1066 (15.08%)< 0.001BMI (kg/m2)29.52 ± 0.0029.69 ± 0.0029.45 ± 0.0029.66 ± 0.0029.27 ± 0.00< 0.001WC (cm)101.82 ± 0.00102.86 ± 0.00101.93 ± 0.00102.09 ± 0.00100.27 ± 0.00< 0.001Height (cm)168.50 ± 0.00168.95 ± 0.00168.71 ± 0.00168.46 ± 0.00167.83 ± 0.00< 0.001TyG-BMI258.07 ± 0.00260.64 ± 0.01257.84 ± 0.01259.05 ± 0.01254.51 ± 0.01< 0.001TyG-WC889.47 ± 0.01902.37 ± 0.02891.64 ± 0.02891.03 ± 0.02871.43 ± 0.03< 0.001TyG-WHtR5.29 ± 0.005.35 ± 0.005.29 ± 0.005.30 ± 0.005.20 ± 0.00< 0.001PIR: poverty income ratio; BMI: body mass index; WC: waist circumference; TyG: triglyceride-glucose; WHtR = waist circumference (cm)/height (cm); TyG-BMI = TyG × BMI; Tyg-WC = TyG × WC; TyG-WHtR = TyG × WHtR Baseline characteristics of participants in the NHANES from 2007–2016 PIR: poverty income ratio; BMI: body mass index; WC: waist circumference; TyG: triglyceride-glucose; WHtR = waist circumference (cm)/height (cm); TyG-BMI = TyG × BMI; Tyg-WC = TyG × WC; TyG-WHtR = TyG × WHtR Multivariable regression analysis The association between serum klotho and TyG-related indicators was examined using a weighted multivariable regression model (Table 2). A significant negative association was found between serum klotho and TyG-BMI, TyG-WC, and TyG-WHtR. In model 1, serum lnklotho decreased by 0.063 pg/ml, 0.137 pg/ml, and 0.115 pg/ml with each one-unit increase in lnTyG-BMI, lnTyG-WC, and lnTyG-WHtR, respectively. In model 2, after adjustments for gender, age, race, educational level, marital status, PIR, hypertension, diabetes, and hyperlipidemia, serum lnklotho decreased by 0.065 pg/ml, 0.117 pg/ml, and 0.127 pg/ml in lnTyG-BMI, lnTyG-WC, and lnTyG-WHtR, respectively. In model 3, with all variables adjusted, serum lnklotho decreased by 0.064 pg/ml, 0.116 pg/ml, and 0.126 pg/ml in lnTyG-BMI, lnTyG-WC, and lnTyG-WHtR, respectively. Table 2Association between serum klotho and TyG-BMI, TyG-WC, TyG-WHtRModel 1Model 2Model 3β (95%CI) P-valueβ (95%CI) P-valueβ (95%CI) P-valueInTyG-BMI-0.063 (-0.114, -0.012)0.015-0.065 (-0.118, -0.012)0.017-0.064 (-0.118, -0.011)0.018Q1: < 5.37RefRefRefQ2 (5.37–5.52)-0.047 (-0.076, -0.017)0.002-0.035 (-0.064, -0.006)0.020-0.035 (-0.065, -0.006)0.018Q3 (5.52–5.68)-0.052 (-0.084, -0.020)0.002-0.041 (-0.074, -0.008)0.015-0.041 (-0.074, -0.008)0.015Q4 > 5.68-0.044 (-0.077, -0.010)0.012-0.044 (-0.080, -0.008)0.016-0.044 (-0.079, -0.008)0.017P for trend0.0060.0510.050InTyG-WC-0.137 (-0.207, -0.066)0.000-0.117 (-0.196, -0.038)0.004-0.116 (-0.195, -0.037)0.005Q1: < 6.65RefRefRefQ2 (6.65–6.78)-0.060 (-0.091, -0.030)< 0.001-0.047 (-0.077, -0.016)0.003-0.047 (-0.077, -0.016)0.003Q3 (6.78–6.90)-0.072 (-0.106, -0.039)< 0.001-0.054 (-0.089, -0.019)0.003-0.054 (-0.089, -0.020)0.003Q4 > 6.90-0.074 (-0.109, -0.040)< 0.001-0.062 (-0.099, -0.025)0.001-0.062 (-0.099, -0.025)0.001P for trend< 0.0010.0020.002InTyG-WHtR-0.115 (-0.184,-0.045)0.002-0.127 (-0.204, -0.049)0.002-0.126 (-0.203, − 0.049)0.002Q1: < 1.53RefRefRefQ2 (1.53–1.66)-0.049 (-0.081, -0.017)0.003-0.038 (-0.070, -0.007)0.018-0.039 (-0.070, -0.007)0.018Q3 (1.66–1.79)-0.062 (-0.094, -0.029)0.000-0.053 (-0.085, -0.022)0.001-0.054 (-0.085, -0.022)0.001Q4 > 1.79-0.056 (-0.092, -0.020)0.003-0.058 (-0.099, -0.017)0.006-0.057 (-0.098, -0.017)0.007P for trend< 0.0010.0040.004Model 1 adjusted for no covariables. Model 2 adjusted for gender, age, race, educational level, marriage status, PIR, hypertension, diabetes, and hyperlipidemia. Model 3 adjusted for gender, age, race, education level, marriage status, PIR, hypertension, diabetes, hyperlipidemia, smoking, and alcohol intake Association between serum klotho and TyG-BMI, TyG-WC, TyG-WHtR -0.063 (-0.114, -0.012) 0.015 -0.065 (-0.118, -0.012) 0.017 -0.064 (-0.118, -0.011) 0.018 -0.047 (-0.076, -0.017) 0.002 -0.035 (-0.064, -0.006) 0.020 -0.035 (-0.065, -0.006) 0.018 -0.052 (-0.084, -0.020) 0.002 -0.041 (-0.074, -0.008) 0.015 -0.041 (-0.074, -0.008) 0.015 -0.044 (-0.077, -0.010) 0.012 -0.044 (-0.080, -0.008) 0.016 -0.044 (-0.079, -0.008) 0.017 -0.137 (-0.207, -0.066) 0.000 -0.117 (-0.196, -0.038) 0.004 -0.116 (-0.195, -0.037) 0.005 -0.060 (-0.091, -0.030) < 0.001 -0.047 (-0.077, -0.016) 0.003 -0.047 (-0.077, -0.016) 0.003 -0.072 (-0.106, -0.039) < 0.001 -0.054 (-0.089, -0.019) 0.003 -0.054 (-0.089, -0.020) 0.003 -0.074 (-0.109, -0.040) < 0.001 -0.062 (-0.099, -0.025) 0.001 -0.062 (-0.099, -0.025) 0.001 -0.115 (-0.184,-0.045) 0.002 -0.127 (-0.204, -0.049) 0.002 -0.126 (-0.203, − 0.049) 0.002 -0.049 (-0.081, -0.017) 0.003 -0.038 (-0.070, -0.007) 0.018 -0.039 (-0.070, -0.007) 0.018 -0.062 (-0.094, -0.029) 0.000 -0.053 (-0.085, -0.022) 0.001 -0.054 (-0.085, -0.022) 0.001 -0.056 (-0.092, -0.020) 0.003 -0.058 (-0.099, -0.017) 0.006 -0.057 (-0.098, -0.017) 0.007 Model 1 adjusted for no covariables. Model 2 adjusted for gender, age, race, educational level, marriage status, PIR, hypertension, diabetes, and hyperlipidemia. Model 3 adjusted for gender, age, race, education level, marriage status, PIR, hypertension, diabetes, hyperlipidemia, smoking, and alcohol intake lnTyG-BMI, lnTyG-WC, and lnTyG-WHtR were categorized into four groups (Q1 to Q4) based on their quartile, with the first quartile group (Q1) serving as the reference category (Table 2). We observed a nonlinear relationship between serum klotho and InTyG-BMI, InTyG-WC, and InTyG-WHtR, characterized by an initial decrease followed by an increase. The dose-response association of serum Inklotho levels with InTyG-BMI, InTyG-WC, and InTyG-WHtR Generalized additive models are very sensitive in identifying both linear and nonlinear relationships. We generated three smooth curves to illustrate the potential relationships between serum Inklotho and InTyG-BMI, InTyG-WC, and InTyG-WHtR, based on model 3 (Fig. 2). In addition, we used a piecewise regression model to calculate the breakpoints of InTyG-BMI, InTyG-WC, and InTyG-WHtR, which were 5.42, 6.67, and 1.89, respectively (Table 3). The results showed that the piecewise regression model is more suitable for clarifying the relationship between serum Inklotho and InTyG-WC, and InTyG-WHtR. The smooth curve fitting for the associations between serum klotho and TyG-related indicators: (A) the relationship between serum Inklotho and InTyG-BMI; (B) the relationship between serum Inklotho and InTyG-WC; (C) the relationship between serum Inklotho and InTyG-WHtR Table 3Threshold effect analysis of the association between serum Inklotho and InTyG-BMI, InTyG-WC, TyG-WHtR in NHANES 2007–2016ModelsInTyG-BMIInTyG-WCInTyG-WHtRβ (95%CI) P-valueβ (95%CI) P-valueβ (95%CI) P-valueOne linear effect-0.060 (-0.097, -0.022)0.002-0.128 (-0.178, -0.078)< 0.001-0.135 (-0.185, -0.084)< 0.001Breakpoint (k)5.426.671.89≤ k-0.145 (-0.240, -0.051)0.003-0.268 (-0.393, -0.142)< 0.001-0.169 (-0.226, -0.112)< 0.001> k-0.021 (-0.075, 0.034)0.456-0.067 (-0.138, 0.004)0.0630.163 (-0.080, 0.407)0.189LRT test0.0530.0170.014Adjusting for gender, age, race, education level, marriage status, PIR, hypertension, diabetes, and hyperlipidemia. LRT test: Log likelihood ratio test Threshold effect analysis of the association between serum Inklotho and InTyG-BMI, InTyG-WC, TyG-WHtR in NHANES 2007–2016 -0.060 (-0.097, -0.022) 0.002 -0.128 (-0.178, -0.078) < 0.001 -0.135 (-0.185, -0.084) < 0.001 -0.145 (-0.240, -0.051) 0.003 -0.268 (-0.393, -0.142) < 0.001 -0.169 (-0.226, -0.112) < 0.001 -0.021 (-0.075, 0.034) 0.456 -0.067 (-0.138, 0.004) 0.063 0.163 (-0.080, 0.407) 0.189 Adjusting for gender, age, race, education level, marriage status, PIR, hypertension, diabetes, and hyperlipidemia. LRT test: Log likelihood ratio test Subgroup analysis Subgroup analyses were conducted based on gender, age, diabetes, hypertension, and hyperlipidemia (Table 4). The analysis results showed that the association between serum klotho and TyG-related indicators was influenced by gender and diabetes (P for interaction < 0.05). Table 4Subgroup analysis for the relationship of serum Inklotho with InTyG-BMI, InTyG-WC, TyG-WHtRSubgroupβ (95%CI)P-valueP for interactionInTyG-BMIGenderMale0.001 (-0.052, 0.054)0.9660.044Female-0.071 (-0.116, -0.026)0.002Age40–60-0.049 (-0.092, -0.006)0.0270.469> 60-0.023 (-0.079, 0.034)0.431DiabetesYes0.026 (-0.061, 0.112)0.5640.015No-0.089 (-0.128, -0.049)< 0.001HypertensionYes0.008 (-0.069, 0.084)0.8440.152No-0.053 (-0.091, -0.014)0.007HyperlipidemiaYes-0.026 (-0.068, 0.017)0.2350.382No-0.059 (-0.124, 0.005)0.070InTyG-WCGenderMale-0.025 (-0.091, 0.041)0.4560.003Female-0.159 (-0.220, -0.099)< 0.001Age40–60-0.122 (-0.178, -0.067)< 0.0010.753> 60-0.108 (-0.180, -0.035)0.004DiabetesYes0.000 (-0.118, 0.117)0.996< 0.001No-0.221 (-0.271, -0.170)< 0.001HypertensionYes-0.054 (-0.156, 0.048)0.2990.098No-0.146 (-0.195, -0.097)< 0.001HyperlipidemiaYes-0.113 (-0.167, -0.059)< 0.0010.383No-0.156 (-0.239, -0.074)< 0.001InTyG-WHtRGenderMale-0.021 (-0.088, 0.045)0.5320.002Female-0.166 (-0.224, -0.107)< 0.001Age40–60-0.091 (-0.147, -0.035)0.0020.618> 60-0.067 (-0.140, 0.006)0.072DiabetesYes-0.004 (-0.122, 0.114)0.9500.006No-0.177 (-0.229, -0.126)< 0.001HypertensionYes-0.025 (-0.127, 0.078)0.6360.115No-0.113 (-0.162, -0.064)< 0.001HyperlipidemiaYes-0.082 (-0.137, -0.027)0.0030.458No-0.119 (-0.202, -0.036)0.005 Subgroup analysis for the relationship of serum Inklotho with InTyG-BMI, InTyG-WC, TyG-WHtR Discussion Population aging poses a major public health challenge worldwide at present. The aging population often experiences declines in organ function and metabolic disorders. The potential link between the anti-aging protein klotho and metabolic diseases is also gaining attention [12]. Therefore, we used NHANES samples from 2007 to 2016 to explore the relationship between serum klotho and TyG-related indicators. After analyzing data from 6,370 participants, the key findings were as follows: (1) There were significant differences in levels of serum klotho according to the quartiles. (2) It showed a significant negative nonlinear relationship between the serum klotho and TyG-BMI, TyG-WC and TyG-WHtR. (3) The relationship between serum klotho and the TyG-related indicators was influenced by gender and diabetes. The negative association between klotho and TyG-related indicators may suggest a potential role for klotho in regulating energy metabolism and slowing down the aging process. The klotho family comprises three subtypes: α, β, and γ. In addition, klotho can also be categorized into membrane-bound (mKL), secretory or soluble (sKL), and intracellular forms, with different forms of klotho involved in different physiological processes [26]. Studies have shown that the serum klotho levels in patients with diabetes are significantly lower than those in non-diabetic individuals [27]. Deficiency of soluble klotho in circulating blood leads to increased IR, diabetic phenotype, and hepatic lipid accumulation in T2D. Soluble klotho can target the IGF1R/PI3K/AKT/mTORC1 signaling pathway and up-regulate peroxisome proliferator-activated receptor α (PPARα) in T2D, thereby improving insulin sensitivity, hepatic glucose production, and lipid homeostasis [28]. The kidney is the main expression site for klotho, which affects glucose metabolism by influencing glucose reabsorption and filtration within the kidney [29]. Furthermore, klotho can alleviate podocyte damage in diabetic kidney disease (DKD) by modulating the IGF-1R/RAC1/OLR1 signaling pathway, thereby improving high-glucose-induced glomerular deposition of oxidized low-density lipoprotein (ox-LDL) [30]. Therefore, klotho has potential in the prevention and treatment of diabetes and its complications. Elevated plasma lipoprotein levels and decreased lipid clearance lead to lipid accumulation in the body. Klotho can increase the body’s energy expenditure, reduce fat mass, and play a beneficial role in regulating lipid accumulation in the liver and adipose tissue. For example, a-klotho can improve obesity status in DIO mice by reducing lipid accumulation in the liver and adipose tissue [31]. Recombinant klotho can reduce lipid accumulation in foam cells and promote increased cholesterol efflux by inhibiting the Wnt/β-catenin pathway [32]. Furthermore, a-klotho promotes lipid oxidation and reduces adipogenesis by regulating gene expression in the liver and white adipose tissue (WAT) of DIO mice [26]. Soluble klotho can alleviate hepatic steatosis by inhibiting genes associated with lipid synthesis and uptake while promoting genes involved in lipid oxidation [28]. Klotho has a protective effect against high-fat diet (HFD) induced hepatic steatosis, reduces the weight of WAT and liver, and enhances the expression of thermogenic genes in brown adipose tissue (BAT) [33]. Although our study primarily investigated the association between serum klotho and TyG-related indicators, it offers novel insights into the prevention and treatment of metabolic diseases from the perspective of aging. Studies have shown that the level of s-klotho in the plasma of diabetic NOD mice is significantly lower than that in the non-diabetic mice [34]. A meta-analysis has demonstrated that individuals with chronic kidney disease (CKD) exhibiting reduced serum klotho levels are at an elevated risk of all-cause mortality [35]. Furthermore, polymorphisms G395A and C1818T in the klotho gene have been correlated with an elevated risk of developing T2D [36]. Klotho exerts a protective effect against diabetic retinopathy by attenuating the senescence of retinal macrophages, which is achieved by downregulating HECTD1 and diminishing IRS1 ubiquitination, and subsequent proteasomal degradation [37]. Numerous studies have also confirmed the potential mechanisms of α-klotho in alleviating diabetes and its complications [36], NAFLD [38], obesity [39], and other metabolic diseases. Therefore, serum klotho can not only serve as a potential biomarker for metabolic diseases but also as a therapeutic target. In the subgroup analysis, we observed the impact of gender and diabetes on the relationship between serum klotho and TyG-related indicators, particularly among female and non-diabetic patients. Our study results are consistent with previous research findings [40, 41]. Gender differences may be related to variations in hormone levels, fat distribution, and differences in insulin sensitivity. Women in menopause experience a decrease in estrogen levels, which can reduce sensitivity to insulin [42]. Our study subjects are primarily over the age of 40. Furthermore, the reduction in estrogen levels can also lead to the accumulation of fat in the visceral area, thereby increasing metabolic risk [43]. Therefore, the gender differences observed in our study may be due to the influence of hormonal levels on serum klotho expression. Moreover, relevant research has also confirmed that estrogen can upregulate the expression of serum klotho [44]. Qiu S [41] investigated the relationship between klotho and TyG index in populations with and without diabetes, and the results indicated that klotho is positively associated with the TyG index in diabetic populations compared to those without diabetes. This result is consistent with our study findings. Wang C [40] speculated that as the level of IR increases, there may be a saturation effect in the association between klotho and TyG. In summary, the biological mechanisms underlying the influence of the relationship between serum klotho and TyG-related indicators in the non-diabetic female population still require further research for elucidation. Our research has some advantages. Firstly, it represents the preliminary attempt to explore the association between serum klotho and indicators such as TyG-BMI, TyG-WC, and TyG-WHtR. This investigation may provide new insights into the prevention and treatment of both age-related conditions and metabolic disorders. Secondly, our study benefits from the weighted analysis of a population sample drawn from NHANES, enhancing the representativeness of the results. However, this study also has limitations. Firstly, this study is limited by its cross-sectional design, which hinders the ability to determine causality between klotho and TyG-BMI, TyG-WC, and TyG-WHtR. Secondly, our study included only the common complications of non-diabetic, hypertension, and hyperlipidemia, without incorporating other related chronic diseases. Future research should encompass a broader range of chronic conditions to fully understand the impact of different health status on the study outcomes. Finally, despite adjusting for potential confounding variables, there remains a possibility of unidentified factors that could impact the relationship between serum klotho and the TyG-BMI, TyG-WC, and TyG-WHtR. Conclusion Based on the analysis of nationally representative samples, our research revealed a negative nonlinear relationship between serum klotho and TyG-BMI, TyG-WC, and TyG-WHtR in U.S. adults. The findings indicate that the relationship between serum klotho and TyG-related indicators could offer a new perspective for anti-aging and prevention of metabolic diseases. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Abbreviations Triglyceride glucose index National Health and Nutrition Examination Survey Body mass index Waist circumference Waist-to-height ratio Enzyme-linked immunosorbent assay Generalized additive model Insulin resistance Fasting blood glucose Non-alcoholic fatty liver disease Metabolic-associated fatty liver disease National Center for Health Statistics Centers for Disease Control and Prevention Poverty income ratio Glycosylated hemoglobin Membrane-bound Secretory or soluble Peroxisome proliferator-activated receptor α Diabetic kidney disease Oxidized low-density lipoprotein White adipose tissue High fat diet Brown adipose tissue Chronic kidney disease Author contributions YY: Writing – original draft, Conceptualization, Visualization, Formal Analysis, Methodology. YQ: Writing – original draft, Formal Analysis, Software, Methodology. FL: Formal Analysis, Software, Methodology. YM: Writing – review & editing, Formal Analysis, Software, Methodology. TT: Writing – review & editing, Formal Analysis, Software, Methodology. YS: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization. Funding This project was supported by Special Project of State Key Laboratory of Dampness Syndrome of Traditional Chinese Medicine jointly established by the Ministry and the Province (No. SZ2022KF05), Beijing University of Traditional Chinese Medicine Campus level Project (No. 90010961020140), High level Key Discipline of National Administration of Traditional Chinese Medicine-Traditional Chinese constitutional medicine (No. zyyzdxk-2023251), The Joint Fund focuses on project U22A20365 (No.90030271420007), and Shenzhen Longgang District Science and Technology Development Special Fund (LGWJ2021-142). Data availability No datasets were generated or analysed during the current study. Declarations Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yaoyao Zhou and Yaqi Wang share first authorship. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement No datasets were generated or analysed during the current study.