Introduction

The global issue of population aging is intensifying, with projections indicating that 1.4 billion people will reach the age of 60 or more by 20301. This demographic shift poses a significant burden on healthcare systems worldwide. Aging is a complex process characterized by a gradual decline in the functional capacities of various organs and systems2. However, aging does not progress uniformly over time, as individuals age at different rates and exhibit varying susceptibilities to chronic diseases and mortality3. Accelerated aging, in particular, is causally linked to many diseases and adverse health outcomes4. Therefore, gaining insight into the biological aging process is of critical importance.

Insulin resistance (IR) is a metabolic dysfunction associated with various chronic illnesses5,6,7,8. However, the euglycemic-hyperinsulinemic clamp, considered the gold standard for identifying IR, is not widely used in clinical practice due to its inherent limitations9. Consequently, various alternative methods have been developed10. Among these, TyG is a reasonably accurate alternative marker11. The significant correlation between TyG and its combined obesity-related parameters and various age-related diseases has been revealed12,13,14,15. Vitamin E, a colorless lipid-soluble antioxidant, serves as an essential micronutrient with pleiotropic effects in human physiology16. Substantial evidence has established associations between vitamin E intake and multiple health outcomes17,18,19.

To date, no studies have investigated the relationship between TyG and its obesity-related parameters (including TyG-BRI) with accelerated phenotypic aging, nor the potential mediating role of vitamin E intake. The present study aims to: (1) examine the relationship between these indicators and biological age acceleration in middle-aged and elderly populations; (2) evaluate the mediating contribution of vitamin E intake; and (3) assess their value in the early diagnosis of accelerated aging.

Methods

Participants

This study utilized data collected from the NHANES between 2005 and 2010, comprising a total of 31,034 participants. Our research applied specific exclusion criteria, which included individuals aged under 45 or over 79 years (N = 22543), as well as those with missing data on phenotypic age (PhenoAge) (N = 4689), TyG and its obesity-related indicators (N = 160), vitamin E intake (N = 81) and covariates (N = 447). Ultimately, a total of 3114 subjects were included in the analysis (Fig. 1).

Fig. 1
figure 1

The research flow chart of study participants. NHANES: National Health and Nutrition Examination Survey; PhenoAge: phenotypic age; TyG: triglyceride glucose index.

Full size image

Definitions of TyG, its related indicators, and PhenoAge

TyG is calculated using fasting triglycerides (TG) and fasting glucose data obtained from the Laboratory Data section. The formula for the TyG is as follows: TyG = Ln [fasting TG (mg/dL) × fasting glucose (mg/dL)/2].

Data on height, weight, and waist circumference (WC) for all participants were obtained from the Body Measures section. WHtR = WC (m)/height (m); BMI = weight (kg)/height2 (m2); BRI = 364.2 − 365.5 × √(1 − [WC (m)/2π]2/[0.5 × height (m)]2); TyG-WHtR = TyG × WHtR; TyG-BMI = TyG × BMI; TyG-BRI = TyG × BRI.

PhenoAge is emerging indicator currently utilized to represent biological age. The required nine clinical biochemical indicators were obtained from the Laboratory Data section and the calculation formulas are as follows20:

$$\:\text{P}\text{h}\text{e}\text{n}\text{o}\text{A}\text{g}\text{e}\:=\:141.50\:+\:\frac{\text{l}\text{n}[-0.00553\:\times\:\:\text{l}\text{n}\:(1\:-\:\text{x}\text{b}\left)\right]}{0.09165}$$

where

xb = − 19.907 − 0.0336 × albumin + 0.0095 × creatinine + 0.1953 × glucose + 0.0954 × ln (C-reactive protein [CRP]) − 0.0120 × lymphocyte percentage + 0.0268 × mean cell volume + 0.3306 × red cell distribution width + 0.00188 × alkaline phosphatase + 0.0554 × white blood cell count + 0.0804 × chronological age.

PhenoAgeAccel is defined as the deviation of expected PhenoAge based on chronological age. Individuals were classified into two distinct states based on whether PhenoAgeAccel exceeded zero: accelerated aging and delayed aging21.

Vitamin E intake

Information on vitamin E intake was obtained through 24-hour dietary recall interviews. The total vitamin E intake for each participant was precisely determined by summing two specific components: “Vitamin E as alpha-tocopherol” and “Added alpha-tocopherol (Vitamin E)”.

Covariates

Based on previous studies, several demographic variables, lifestyle behaviors, and disease statuses were included as covariates. Age, in addition to being treated as a continuous variable, was categorized into two groups: middle-aged (45–59 years), and elderly (60–79 years). Race was classified into four categories: Mexican, White, Black, and Other. Educational level, marital status, and the poverty income ratio (PIR) were also taken into consideration. Alcohol drinking was defined as having consumed at least 12 alcoholic drinks in the past year. Smoking history was categorized as either smoker and non-smoker. Physical inactivity was defined as engaging in less than 10 min per week of moderate–vigorous recreational activity. Diagnoses of diabetes, hypertension, cardiovascular disease (CVD) and cancer were assessed using self-reported information. Hypolipidemic agent usage was determined based on the reported utilization of antihyperlipidemic agents in the Prescription Medications dataset. We also considered laboratory indicators such as blood urea nitrogen (BUN) and CRP.

Statistical analysis

Statistical analyses were performed using EmpowerStats (version 4.2) and Stata (version 16.0), employing the appropriate NHANES sample weights. Continuous variables were expressed as mean ± standard deviation (SD), while categorical variables were presented as numbers (weighted percentages). Baseline characteristics of continuous variables were analyzed using either the t-test or the Mann-Whitney U test, whereas categorical variables were analyzed using the Chi-squared test or Fisher’s exact test. Multivariate linear regression was employed to investigate the relationships between TyG and its related metrics with PhenoAgeAccel. Multivariable logistic regression was conducted to evaluate the associations between these indices and accelerated aging, estimating ORs and 95% CIs. Model 1 was adjusted for sex, age, race, education level, marital status, PIR, smoking history, alcohol drinking, and recreational activity. Model 2 included further adjustments for BUN, CRP, hypolipidemic agent usage, history of diabetes, hypertension, CVD and cancer. Subgroup analyses were conducted to further explore potential modifying factors. Subsequently, formal mediation analyses were performed to quantify the mediating contribution of vitamin E intake. ROC curves were employed to evaluate the diagnostic value, calculating the AUC to quantify the predictive ability of TyG and its related indicators in detecting accelerated aging. A two-tailed P-value of < 0.05 was considered statistically significant.

Results

Participants’ characteristics

A total of 3114 subjects were included in the final analysis, among whom 27.71% (863/3114) exhibited accelerated aging. The baseline characteristics of the participants are presented in Table 1. Compared to individuals with delayed aging, those with accelerated aging were generally older (59.34 ± 9.87 years), predominantly male (57.98%), had lower educational attainment (54.68%), were more likely to be inactive (61.17%), and had a higher likelihood of smoking (60.94%), while their alcohol consumption was lower (68.45%). Additionally, a range of health conditions was more prevalent in this group, including hypertension (59.21%), diabetes (32.03%), CVD (19.79%), cancer (16.92%) and hypolipidemic agent use (35.44%).

Table 1 The clinical characteristics of the study population.
Full size table

Associations between TyG and its related metrics with accelerated phenotypic aging

Table 2 presented that in the partially adjusted model, TyG, TyG-WHtR, TyG-BMI, and TyG-BRI, were positively correlated with PhenoAgeAccel, and all indices were also positively associated with the risk of accelerated aging. These associations remained statistically significant in Model 2. The corresponding β/OR values for the indicators as continuous variables were as follows: TyG, β = 2.67, 95% CI: 2.27, 3.07, OR = 2.00, 95% CI: 1.63, 2.46; TyG-WHtR, β = 1.87, 95% CI: 1.61, 2.13, OR = 1.85, 95% CI: 1.59, 2.15; TyG-BMI/100, β = 2.59, 95% CI: 2.20, 2.99, OR = 2.46, 95% CI: 1.95, 3.11; TyG-BRI/10, β = 0.80, 95% CI: 0.67, 0.93, OR = 1.30, 95% CI: 1.21, 1.40. Analysis also revealed a positive gradient in both PhenoAgeAccel and accelerated aging risk across ascending tertiles of these parameters, though not all positive associations with accelerated aging reached statistical significance. Specifically, when comparing the highest tertiles (T3) to the lowest tertiles (T1) in Model 2, the following results were observed: TyG, β = 2.59, 95% CI: 1.99, 3.19; OR = 2.52, 95% CI: 1.81, 3.51; TyG-WHtR, β = 3.55, 95% CI: 2.92, 4.18; OR = 3.22, 95% CI: 2.28, 4.54; TyG-BMI, β = 3.01, 95% CI: 2.41, 3.62; OR = 2.64, 95% CI: 1.90, 3.67; TyG-BRI, β = 3.08, 95% CI: 2.45, 3.70; OR = 3.01, 95% CI: 2.13, 4.25.

Table 2 Associations between TyG, TyG-WHtR, TyG-BMI, and TyG-BRI with accelerated phenotypic aging.
Full size table

Subgroup analyses

Subgroup analyses further revealed the potential effects of TyG and its related indices on PhenoAgeAccel (Fig. 2). The robustness of the observed associations was confirmed across subgroups.

Fig. 2
figure 2

Forest plot of the subgroup analyses on the correlation between TyG and its related parameters with PhenoAgeAccel in the fully adjusted model.

Full size image

Mediation analysis

Mediation analysis results, as depicted in Fig. 3, revealed that vitamin E intake exhibited significant mediating effects on the relationships between TyG and its related indices with PhenoAgeAccel. The mediation proportions were 1.73%, 2.06%, 1.60%, and 2.19%, respectively.

Fig. 3
figure 3

Mediation effects of vitamin E intake on the associations between TyG (A), TyG-WHtR (B), TyG-BMI (C), and TyG-BRI (D) and PhenoAgeAccel in the fully adjusted model. *P < 0.05, **P < 0.001.

Full size image

Predictive value in accelerated aging

The ROC curves in Fig. 4 illustrated that TyG-WHtR demonstrates superior diagnostic accuracy in identifying accelerated aging, achieving an AUC of 0.694 (95% CI: 0.673, 0.715). This performance outperformed TyG (AUC = 0.650, 95% CI: 0.628, 0.671), TyG-BMI (AUC = 0.678, 95% CI: 0.656, 0.700), and TyG-BRI (AUC = 0.681, 95% CI: 0.660, 0.703).

Fig. 4
figure 4

Receiver operating characteristic curves of TyG, TyG-WHtR, TyG-BMI, and TyG-BRI in all participants.

Full size image

Discussion

The findings of this study, based on a representative sample of U.S. middle-aged and elderly population, revealed that approximately 27.71% of participants exhibited accelerated aging, which was often associated with higher TyG obesity-related parameters levels. Notably, vitamin E intake demonstrated a significant mediating effect on the observed associations. Furthermore, TyG-WHtR exhibited the best performance in identifying accelerated aging, surpassing the diagnostic effectiveness of TyG, TyG-BMI, and TyG-BRI. However, the overall predictive performance of these indices remains moderate.

IR refers to a decreased sensitivity and responsiveness of the body to insulin22 and it is believed to play a pivotal role in phenotypic aging and age-related events23. And TyG is a superior tool for assessing IR compared to other indices. Its association with various age-related diseases has been extensively evaluated in previous studies, including adverse cardiovascular events12,24 respiratory diseases and impaired lung function25 and peripheral artery disease26. Undoubtedly, TyG is also associated with aging and has recently been identified as a potential predictor of accelerated aging27. In fact, researches have attempted to establish the association between TyG and aging-related biomarkers. Among these, Klotho—a metabolism-related anti-aging protein that plays a critical role in aging—has garnered particular attention. Reports focusing on this research subject are ongoing, and the conclusions are consistent28,29,30. Furthermore, traditional aging-related biomarkers, such as telomere length, have also been shown to be associated with TyG31. Obesity is another significant factor contributing to accelerated aging. Nevalainen et al. highlighted the harmful effects of obesity in driving accelerated epigenetic aging32. More recently, a report employing Mendelian randomization analysis identified a correlation between obesity, BMI, and telomere length shortening33.

Combining TyG with obesity-related metrics as a predictive approach for diseases appears to be a growing research focus. This strategy has been shown to improve diagnostic accuracy for chronic conditions34,35,36. Zhou and colleagues explored the use of TyG adiposity-related indicators in association to Klotho levels37which provided initial inspiration for our study. In addition to earlier-developed biological age prediction markers3PhenoAgeAccel has emerged as a novel parameter reflecting accelerated biological aging.

To clarify the relationship between TyG and its obesity-related metrics and accelerated phenotypic aging, we integrated these indicators into both linear and logistic regression models for a comprehensive evaluation and obtained all positive correlation results. In response, we have attempted to propose some hypotheses and explanations based on existing investigations and findings. Firstly, reduced insulin sensitivity leads to elevated blood glucose levels, while impaired hepatic glucose metabolism increases the likelihood of conversion into triglycerides. And TyG provides an objective and reliable reflection of these metabolic states. IR-induced glucose metabolism imbalance triggers inflammation38 which is a key mechanism of aging mobilized by age-related core disorders39,40. Oxidative stress-mediated DNA and mitochondrial damage accelerates cellular aging41. Metabolic abnormalities caused by IR are also major drivers of cardiovascular aging6. The pathophysiological outcomes of sustained IR inevitably increase the burden on the heart and kidneys42. Aging, in turn, may reduce overall metabolic activity, further impairing insulin signaling pathways43. Additionally, IR can affect the expression of aging-related genes, accelerating the progression of biological aging in individuals37. Secondly, obesity is another important factor in age-related decline in physiological functions. Obesity induces oxidative stress and inflammation, which accelerate systemic aging processes44. Moreover, both obesity itself and obesity-related conditions have been closely linked to accelerated aging, as well as aging-related illnesses45. Aging and obesity may also share overlapping cellular pathways and molecular mechanisms46. Thirdly, IR, obesity, and aging are interrelated and mutually influential. Obesity often leads to inflammation in adipose tissue, promoting the development of lipid-induced IR and diabetes47,48. A hallmark of aging in adipose tissue is the progression of inflammation, and as the organism ages, pathological changes in adipose tissue can cause secondary damage to vital organs, such as the liver, further exacerbating IR47. Existing studies have suggested that IR may partially mediate the obesity-accelerated aging association. For instance, in Lundgren’s study, the correlation between IR-related markers and accelerated aging diminished after controlling for BMI as a variable48. Other research has proposed that obesity, particularly abdominal obesity, partially explains the effect of IR on cellular aging and plays a key role in various cascading reactions49. A report by Ehrhardt et al. emphasized that both obesity and aging serve as central factors in the development of IR50. Similar perspective has been presented in another review report, which emphasizes that both aging and obesity exert substantial effects on metabolism through various mechanisms51. However, our current study does not further investigate the causal relationships among these factors.

Vitamin E, a health-beneficial antioxidant, exists in eight homologous forms categorized into tocopherols and tocotrienols, with α-tocopherol being the predominant form in the human body52,53. Vitamin E is absorbed through the intestinal lumen, incorporated into chylomicrons, secreted into circulation, and transported via lipoproteins54. Functioning as an antioxidant, vitamin E mitigates oxidative stress, thereby preventing lipid peroxidation products from impairing the integrity of β-cells, providing protective effects on pancreatic cells55. Moreover, vitamin E has been shown to activate calcium-dependent endopeptidases, ultimately promoting extracellular insulin secretion56. Accumulating evidence indicates that vitamin E effectively reduces glycated hemoglobin, IR, and its alternative markers57,58. Foods rich in vitamin E can reduce the risk of developing diabetes, and vitamin E supplementation can also lower fasting blood glucose levels in diabetic patients58,59. Recently, Xu et al. found a significant negative correlation between vitamin E levels and TyG60. On the other hand, a survey indicated an inverse relationship between vitamin E intake and accelerated aging61. As early as 1991, Cutler et al. demonstrated that lifespan in mammalian species is associated with higher levels of alpha-tocopherol62. Subsequent studies further reported elevated levels of vitamin E in healthy centenarians63. The significance of vitamin E in slowing down aging and preventing degenerative diseases is widely recognized64. Aging in humans is a complex phenomenon, and nutritional deficiencies are considered a critical factor contributing to the development of many age-related diseases. In contrast to the accelerated aging effects induced by high-fat diets, vitamin E can mitigate the progression of aging through its antioxidant properties, by maintaining cellular homeostasis and integrity65,66. Additionally, vitamin E can enhance immune responses, although the underlying mechanisms are evidently complex67. Even before the findings by Ma et al., researchers had demonstrated that vitamin E can slow down cellular aging through its relationship with telomere length68. Our study successfully established a bridge between vitamin E intake and the relationship between TyG and PhenoAgeAccel.

In summary, we propose an innovative approach to quantify accelerated phenotypic aging in middle-aged and elderly individuals using TyG and its obesity-related indices, which is feasible for implementation in routine health examinations. Additionally, we identified the mediating contribution of vitamin E intake in these associations. These findings provide valuable insights for developing strategies to mitigate the human aging process. However, several limitations should be acknowledged. As mentioned earlier, establishing a causal relationship was not achieved in our study. While we employed certain statistical methods to control for confounding factors, it does not mean we achieved complete control. Furthermore, the mediating effects observed were relatively weak, and the involvement of other micronutrients with stronger effects cannot be ruled out. Additionally, the assessment of vitamin E intake in this study is based solely on a 24-hour dietary recall, which may not adequately represent the individual’s serum vitamin E levels. Lastly, as our study used data from NHANES, the generalizability of our findings to developing countries remains uncertain, and caution is warranted in this regard.

Conclusions

Among individuals aged 45 and older, those with higher TyG and its obesity-related metrics tend to exhibit accelerated phenotypic aging. Notably, vitamin E intake mediates the relationships between these parameters and PhenoAgeAccel.