Abstract
Purpose
This study aimed to investigate the association between serum insulin-like growth factor 1 (IGF1) levels and cardiovascular disease (CVD) risk, and to further evaluate the potential mediating role of triglyceride-glucose (TyG)-related indices.
Methods
A total of 3,613 participants from the English Longitudinal Study of Ageing (ELSA) were included. Serum IGF1 was categorized into quintiles. Cox proportional hazards models, Kaplan–Meier survival curves, and restricted cubic splines (RCS) were applied to assess the relationship between IGF1 and CVD as well as stroke risk. Subgroup and interaction analyses were performed across demographic and clinical strata. Causal mediation analysis was conducted to evaluate the mediating effects of TyG-related indices, including TyG-BMI, TyG-WC, and TyG-WHtR. Sensitivity analyses were carried out to test the robustness of findings.
Results
IGF1 demonstrated a U-shaped association with CVD risk. Participants with intermediate levels (Q2–Q3) had the lowest risk (e.g., Q2: HR = 0.73, 95% CI: 0.57–0.94; Q3: HR = 0.63, 95% CI: 0.46–0.85), while both the lowest and highest quintiles were associated with higher risk. Kaplan–Meier curves indicated significant differences across quintiles (log-rank P = 0.007). Mediation analyses revealed significant average causal mediation effects (ACME) for TyG-BMI, TyG-WC, and TyG-WHtR (all P < 0.05). Sensitivity analyses confirmed the robustness of results.
Conclusion
Both low and high serum IGF1 levels may increase the risk of cardiovascular disease (CVD) by modulating TyG-related metabolic indices. Incorporating these indices into risk assessment may improve risk stratification and provide a scientific basis for individualized prevention strategies.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12872-026-05539-9.
Keywords: IGF1, TyG, CVD, Longitudinal study, Insulin resistance
Introduction
Cardiovascular disease (CVD) remains one of the leading causes of global disease burden and mortality [1–3]. According to the Global Burden of Disease Study, approximately 4.58 million deaths occurred due to CVD in 2019, accounting for more than 40% of all deaths worldwide [4]. Identifying modifiable metabolic pathways at the population level is of paramount importance for preventing cardiovascular events and reducing mortality.
Insulin-like growth factor-1 (IGF1) is a key effector molecule in the growth hormone-IGF axis, playing a central role in cell proliferation, tissue growth, and metabolic regulation. IGF1 directly affects myocardial structure and function and indirectly influences cardiovascular outcomes by regulating vascular function, lipid metabolism, insulin sensitivity, and body fat/muscle distribution [5–9]. Recent large cohort studies and genetic causal inference research suggest that IGF1 and CVD/mortality risk may exhibit a nonlinear relationship: both lower and higher levels of IGF1 are associated with increased risk of adverse outcomes in the general population, forming a “U-shaped” curve. This may reflect that the low IGF1 “frailty/inflammation-insulin resistance” phenotype and the high IGF1 “anabolic/insulin resistance-promoting growth” phenotype are both unfavorable for cardiovascular health [10, 11]. Furthermore, Mendelian randomization studies have found that genetically higher IGF1 levels are associated with increased risk of type 2 diabetes and coronary heart disease, with some effects potentially mediated through the insulin resistance (IR) pathway, suggesting a causal chain of “IGF1 → insulin/insulin resistance → atherosclerosis“ [12]. Reviews of relevant mechanisms also highlight the interaction between the GH-IGF axis, insulin signaling, lipid metabolism, and cardiovascular aging [13].
Insulin resistance (IR) is another critical determinant of CVD development [14, 15]. However, direct measurement of IR is costly and has limited applicability [16]. In contrast, the triglyceride-glucose index (TyG) and its derived indicators (such as TyG-BMI, TyG-WC, and TyG-WHtR) are considered low-cost, operational alternatives for assessing IR [15, 17–19]and are closely associated with the onset and progression of CVD [20, 21]. IGF1 and TyG-related indices may interact at the metabolic level: TyG reflects lipid metabolism disorders and decreased insulin sensitivity, which are key targets regulated by the IGF1 axis [7]. If TyG-related indices play a mediating role in the “IGF1 → CVD” causal chain, their value in mechanistic research and clinical risk assessment will be further highlighted.
Currently, evidence on whether TyG-related indices mediate the relationship between IGF1 and CVD remains limited. Therefore, this study utilizes data from the English Longitudinal Study of Ageing (ELSA) to examine the association between serum IGF1 levels and CVD, and assess the potential mediating role of TyG-related indices in this relationship.
Methods
Data source and study population
This study used data from the English Longitudinal Study of Ageing (ELSA), a nationally representative longitudinal cohort study focused on the older adult population in the UK [22]. The study conducts computer-assisted face-to-face interviews (CAPI) and self-administered questionnaires every two years, with nurse visits every four years to collect biological measures. All study activities were approved by the National Research Ethics Service (London MREC: MREC 01/2/91), and all participants provided written informed consent.
A dynamic baseline cohort was constructed using Waves 4, 6, and 8. Specifically, the time point at which a participant first completed simultaneous measurements of serum IGF1, triglycerides (TG), and fasting plasma glucose (FPG) in any of these waves was defined as the individual baseline. Because participants entered the cohort in different survey years, this design represents a dynamic entry cohort. A total of 6,210 individuals with subsequent follow-up data were initially identified. After excluding participants with a history of CVD at baseline (n = 841), those with incomplete baseline or outcome data (n = 803), missing BMI/WC/height (n = 374), and missing covariates such as marital status, education, smoking, and drinking (n = 579), a final sample of 3,613 participants was included in the analysis (Fig. 1).
Fig. 1.
Flowchart of the study design
Outcome definitions
The primary outcome was total cardiovascular disease (CVD) and its specific subtypes, including angina, myocardial infarction (including heart attack or coronary thrombosis), congestive heart failure, arrhythmias, and stroke. All outcomes were based on diagnoses made by clinicians. In addition, angina, myocardial infarction, and stroke were classified as atherosclerosis-related phenotypes to facilitate further analysis.
Serum IGF1 levels
Participants were required to fast prior to the nurse visit. After a professional assessment by the nurse to confirm suitability for fasting blood collection, whole blood samples were drawn by trained nurses. Samples were sent to the Royal Victoria Infirmary laboratory in Newcastle [23] for serum separation and storage at -40 °C. Total IGF1, including both IGF-bound and free forms, was measured using the DPC Immulite 2000 system, with results reported in nmol/L.
Mediating variables
The mediators in this study were TyG-related indices, including TyG-BMI, TyG-WC, and TyG-WHtR. FPG and TG values were used to calculate the TyG index [24]. Fasting blood samples were collected by nursing professionals, and FPG and TG concentrations were measured by enzymatic colorimetry. The TyG index formula is: TyG = ln[TG (mg/dL) × FBG (mg/dL) / 2]. Anthropometric measures included height and weight measured by nurses during the outpatient visit, and body mass index (BMI, in kg/m²) was calculated by dividing weight by the square of height [25]. For WC measurements, the measurement was taken at the end of exhalation at the level of the navel [26]. WHtR was calculated by dividing WC by the individual’s height [27]. TyG indices were further adjusted by multiplying with BMI, WC, and WHtR to obtain TyG-BMI, TyG-WC, and TyG-WHtR, respectively.
Covariates
Sociodemographic, lifestyle, and clinical variables were considered as covariates to adjust for potential confounding factors. Sociodemographic variables included age, sex, race, and marital status (with marital status categorized as married/partnered and other statuses such as single, separated, divorced, or widowed), and education level (categorized as below or above high school). Lifestyle variables included smoking status (yes/no) and drinking frequency in the past 12 months (yes/no). Clinical variables included doctor-diagnosed hypertension, diabetes, and hypercholesterolemia.
Statistical analysis
Baseline characteristics were described by categorizing participants into quintiles of baseline IGF1 (Q1–Q5). Continuous variables were presented as mean ± standard deviation, and intergroup differences were assessed using the Student’s t-test. Categorical variables were presented as frequencies and percentages, with differences assessed using the Chi-square test. Follow-up time was calculated from each participant’s individual baseline, defined as the wave (Wave 4, 6, or 8) at which IGF1, triglycerides, and fasting glucose were first simultaneously measured. Participants were followed until the earliest occurrence of: (1) a first diagnosis of CVD, or(2) their last available survey assessment in Wave 9 (2019) if no CVD event occurred.
Cox proportional hazards models were used to assess the association between IGF1 and CVD risk. Three models were constructed sequentially: Model 1 (unadjusted); Model 2 adjusted for age, sex, marital status, and education; and Model 3 further adjusted for smoking, drinking, diabetes, hypertension, and hypercholesterolemia. The hazard ratio (HR) and its 95% confidence interval (CI) were used as the primary analysis metrics to evaluate the risk of CVD. CVD risk was assessed by comparing the different IGF1 quintiles with the first quintile, and trend tests were performed based on the median IGF1 levels of each quintile in all three models.We assessed multicollinearity using variance inflation factors (VIFs) and evaluated the proportional hazards assumption using Schoenfeld residual tests. The diagnostic results indicated no evidence of multicollinearity (Supplementary Table 1) and no violation of the proportional hazards assumption (GLOBAL = 14.87, df = 13, p = 0.315).
Based on minimizing the Akaike Information Criterion (AIC), we selected a 3-knot restricted cubic spline (RCS) approach to balance model flexibility and the risk of overfitting. The knots were placed at the 25th, 50th, and 75th percentiles, allowing the RCS to better capture potential nonlinear associations between serum IGF1 levels and CVD outcomes. Subgroup analyses were conducted based on age (< 60 or ≥ 60), sex (male or female), marital status (married/partnered vs. other marital status), education level (high school or below vs. above high school), drinking habits (yes/no), smoking status (yes/no), hypertension (yes/no), diabetes (yes/no), and hypercholesterolemia (yes/no). Interaction terms were included in the fully adjusted models to test for effect modification.
To investigate the potential mediating role of TyG-related indices in the association between IGF1 and cardiovascular disease (CVD), we employed a causal mediation analysis framework tailored for cohort data. The total effect of IGF1 on CVD was decomposed into the average causal mediation effect (ACME) and the average direct effect (ADE) under the potential outcomes framework proposed by Imai et al. [28]. This approach is predicated on the assumption of sequential ignorability, which requires that—conditional on measured covariates—there be no unmeasured confounding for either the exposure–mediator or mediator–outcome relationships. Both the mediator and outcome models were adjusted for major demographic and clinical confounders, including age, sex, marital status, educational attainment, smoking status, alcohol consumption, diabetes, hypertension, and hypercholesterolemia.Causal mediation analyses were implemented using the mediate function from the R mediation package, and uncertainty in ACME and ADE was quantified via 1,000 quasi-Bayesian Monte Carlo simulations to derive confidence intervals and statistical significance [29].This simulation-based estimation provides enhanced robustness when assessing the potential mediating role of TyG-related indices in the IGF1–CVD pathway.
To reinforce the credibility and stability of the findings, we conducted a series of sensitivity analyses. Because Wave 4 contributed the largest proportion of participants, all primary analyses were anchored in this wave. In addition, we performed an additional mediation-specific sensitivity analysis restricted to Wave 4 data to further evaluate the robustness of the estimated mediation effects. Missing baseline covariates were addressed through multiple imputation using the mice package to minimize potential selection bias. Finally, to mitigate reverse causality, individuals who developed CVD within the first two years of follow-up were excluded, and a 2-year landmark analysis was carried out.
Statistical analyses were performed using R Studio 4.3.0 and Zstats v1.0 (www.zstats.net). A two-sided p-value < 0.05 was considered statistically significant.
Results
Baseline characteristics of participants
According to the quintiles of insulin-like growth factor-1 (IGF1), among the 3613 participants, the average age was 62.64 (7.25) years, and 56.02% were female. The majority of participants were married or partnered (73.93%) and had completed at least high school education (72.57%). In addition, the majority had consumed alcohol in the past 12 months (92.17%) and had never smoked (88.38%). The sample included 1,082 participants (29.95%) with hypertension, 1,039 (28.76%) with hyperlipidemia, and 115 (3.18%) with diabetes. During a mean follow-up of 7.72 years, a total of 490 CVD events were recorded, including 87 strokes, 69 cases of angina, 79 myocardial infarctions, 20 cases of congestive heart failure, 279 arrhythmias, and 227 atherosclerotic phenotypes.Participants with higher IGF1 levels were generally younger, more often male, and had higher educational attainment. Compared with those with higher serum IGF1 levels, participants with lower IGF1 levels had a higher proportion of CVD events (Table 1).
Table 1.
Baseline characteristics of the study participants
| Variables | Total (n = 3613) | 1 (n = 832) | 2 (n = 889) | 3 (n = 561) | 4 (n = 625) | 5 (n = 706) | P |
|---|---|---|---|---|---|---|---|
| Age, Mean ± SD | 62.64 ± 7.25 | 64.04 ± 7.23 | 63.51 ± 7.15 | 62.49 ± 7.12 | 61.75 ± 7.19 | 60.80 ± 7.06 | <.001 |
| BMI, Mean ± SD | 27.79 ± 4.88 | 28.44 ± 5.48 | 27.86 ± 4.86 | 27.39 ± 4.71 | 27.62 ± 4.76 | 27.40 ± 4.30 | <.001 |
| Waist, Mean ± SD | 94.90 ± 13.28 | 95.24 ± 13.94 | 95.11 ± 13.25 | 93.73 ± 13.03 | 95.39 ± 13.29 | 94.74 ± 12.68 | 0.193 |
| TG, Mean ± SD | 1.49 ± 0.84 | 1.52 ± 0.92 | 1.50 ± 0.90 | 1.47 ± 0.77 | 1.52 ± 0.81 | 1.42 ± 0.76 | 0.149 |
| FBG, Mean ± SD | 5.03 ± 0.90 | 4.97 ± 0.96 | 5.03 ± 0.88 | 5.02 ± 0.82 | 5.06 ± 0.99 | 5.06 ± 0.81 | 0.257 |
| Height, Mean ± SD | 1.67 ± 0.09 | 1.64 ± 0.09 | 1.66 ± 0.09 | 1.67 ± 0.09 | 1.68 ± 0.09 | 1.69 ± 0.09 | <.001 |
| Follow-up, Mean ± SD | 7.72 ± 2.49 | 7.88 ± 2.40 | 7.74 ± 2.49 | 7.78 ± 2.51 | 7.64 ± 2.52 | 7.52 ± 2.52 | 0.061 |
| TYG, Mean ± SD | 8.56 ± 0.50 | 8.56 ± 0.52 | 8.56 ± 0.52 | 8.56 ± 0.49 | 8.60 ± 0.50 | 8.54 ± 0.48 | 0.341 |
| TYGBMI, Mean ± SD | 238.80 ± 48.79 | 244.57 ± 54.30 | 239.40 ± 48.89 | 235.40 ± 47.83 | 238.18 ± 47.57 | 234.49 ± 42.66 | <.001 |
| TYGWC, Mean ± SD | 815.25 ± 140.43 | 818.55 ± 148.41 | 816.98 ± 141.22 | 805.08 ± 138.63 | 822.44 ± 139.93 | 810.89 ± 131.12 | 0.210 |
| TYGWHtR, Mean ± SD | 489.08 ± 81.30 | 498.72 ± 88.63 | 491.32 ± 81.32 | 483.14 ± 81.18 | 488.69 ± 78.44 | 479.97 ± 73.25 | <.001 |
| Gender, n(%) | <.001 | ||||||
| Female | 2024 (56.02) | 565 (67.91) | 517 (58.16) | 325 (57.93) | 289 (46.24) | 328 (46.46) | |
| male | 1589 (43.98) | 267 (32.09) | 372 (41.84) | 236 (42.07) | 336 (53.76) | 378 (53.54) | |
| Marital status, n(%) | 0.010 | ||||||
| Married or with a partner | 942 (26.07) | 248 (29.81) | 231 (25.98) | 133 (23.71) | 172 (27.52) | 158 (22.38) | |
| Other | 2671 (73.93) | 584 (70.19) | 658 (74.02) | 428 (76.29) | 453 (72.48) | 548 (77.62) | |
| Education level, n(%) | <.001 | ||||||
| High school and above | 2622 (72.57) | 542 (65.14) | 629 (70.75) | 406 (72.37) | 485 (77.60) | 560 (79.32) | |
| Below high school | 991 (27.43) | 290 (34.86) | 260 (29.25) | 155 (27.63) | 140 (22.40) | 146 (20.68) | |
| Smoking status, n(%) | 0.093 | ||||||
| No | 3193 (88.38) | 722 (86.78) | 806 (90.66) | 494 (88.06) | 556 (88.96) | 615 (87.11) | |
| Yes | 420 (11.62) | 110 (13.22) | 83 (9.34) | 67 (11.94) | 69 (11.04) | 91 (12.89) | |
| Drinking status, n(%) | <.001 | ||||||
| No | 283 (7.83) | 93 (11.18) | 73 (8.21) | 38 (6.77) | 33 (5.28) | 46 (6.52) | |
| Yes | 3330 (92.17) | 739 (88.82) | 816 (91.79) | 523 (93.23) | 592 (94.72) | 660 (93.48) | |
| Diabetes, n(%) | 0.424 | ||||||
| No | 3498 (96.82) | 800 (96.15) | 869 (97.75) | 543 (96.79) | 604 (96.64) | 682 (96.60) | |
| Yes | 115 (3.18) | 32 (3.85) | 20 (2.25) | 18 (3.21) | 21 (3.36) | 24 (3.40) | |
| Hypertension, n(%) | 0.146 | ||||||
| No | 2531 (70.05) | 560 (67.31) | 630 (70.87) | 406 (72.37) | 427 (68.32) | 508 (71.95) | |
| Yes | 1082 (29.95) | 272 (32.69) | 259 (29.13) | 155 (27.63) | 198 (31.68) | 198 (28.05) | |
| Hyperlipidemia, n(%) | 0.245 | ||||||
| No | 2574 (71.24) | 570 (68.51) | 637 (71.65) | 405 (72.19) | 442 (70.72) | 520 (73.65) | |
| Yes | 1039 (28.76) | 262 (31.49) | 252 (28.35) | 156 (27.81) | 183 (29.28) | 186 (26.35) | |
| Stroke, n(%) | 0.136 | ||||||
| No | 3526 (97.59) | 802 (96.39) | 873 (98.20) | 549 (97.86) | 612 (97.92) | 690 (97.73) | |
| Yes | 87 (2.41) | 30 (3.61) | 16 (1.80) | 12 (2.14) | 13 (2.08) | 16 (2.27) | |
| Angine, n(%) | 0.129 | ||||||
| No | 3544 (98.09) | 807 (97.00) | 875 (98.43) | 552 (98.40) | 614 (98.24) | 696 (98.58) | |
| Yes | 69 (1.91) | 25 (3.00) | 14 (1.57) | 9 (1.60) | 11 (1.76) | 10 (1.42) | |
| Heart attack, n(%) | 0.268 | ||||||
| No | 3534 (97.81) | 813 (97.72) | 867 (97.53) | 555 (98.93) | 607 (97.12) | 692 (98.02) | |
| Yes | 79 (2.19) | 19 (2.28) | 22 (2.47) | 6 (1.07) | 18 (2.88) | 14 (1.98) | |
| Congestive heart failure, n(%) | 0.856 | ||||||
| No | 3593 (99.45) | 826 (99.28) | 886 (99.66) | 558 (99.47) | 621 (99.36) | 702 (99.43) | |
| Yes | 20 (0.55) | 6 (0.72) | 3 (0.34) | 3 (0.53) | 4 (0.64) | 4 (0.57) | |
| Arrhythmias, n(%) | 0.357 | ||||||
| No | 3334 (92.28) | 757 (90.99) | 825 (92.80) | 526 (93.76) | 573 (91.68) | 653 (92.49) | |
| Yes | 279 (7.72) | 75 (9.01) | 64 (7.20) | 35 (6.24) | 52 (8.32) | 53 (7.51) | |
| Atherosclerotic phenotypes, n(%) | 0.021 | ||||||
| No | 3386 (93.72) | 761 (91.47) | 839 (94.38) | 536 (95.54) | 584 (93.44) | 666 (94.33) | |
| Yes | 227 (6.28) | 71 (8.53) | 50 (5.62) | 25 (4.46) | 41 (6.56) | 40 (5.67) | |
| CVD, n(%) | 0.002 | ||||||
| No | 3123 (86.44) | 690 (82.93) | 782 (87.96) | 504 (89.84) | 534 (85.44) | 613 (86.83) | |
| Yes | 490 (13.56) | 142 (17.07) | 107 (12.04) | 57 (10.16) | 91 (14.56) | 93 (13.17) |
Association between IGF1 and CVD
Table 2 shows the associations between serum IGF1 levels and the risk of CVD and specific CVD outcomes. In the Cox proportional hazards model for serum IGF1 levels and CVD, after adjusting for all potential confounders (Model 3), participants with intermediate IGF1 levels had the lowest CVD risk (Q2 HR: 0.73, 95% CI: 0.57–0.94; Q3 HR: 0.63, 95% CI: 0.46–0.85), suggesting that moderate IGF1 levels may have a cardioprotective effect. We also found that participants in the second quintile of IGF1 showed a reduced risk of stroke (HR = 0.54, 95% CI: 0.29–0.99). However, we did not observe a significant association between serum IGF1 levels and other specific CVD outcomes. Furthermore, Kaplan–Meier survival curve analysis showed significant differences in the risk of CVD events across different serum IGF1 levels (log-rank test, P = 0.007). The overall trend was nonlinear: individuals in the lowest (Q1) and highest (Q5) quintiles had higher cumulative risks during follow-up, while those in the middle levels (Q3–Q4) exhibited relatively lower risks. These results suggest a U-shaped association between IGF1 and CVD (Fig. 2).
Table 2.
Associations between quintiles of serum IGF1 concentrations and CVD
| Group | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|
| HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | |
| CVD | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 0.72 (0.56 ~ 0.93) | 0.010 | 0.71 (0.55 ~ 0.92) | 0.008 | 0.73 (0.57 ~ 0.94) | 0.014 |
| Quartile 3 | 0.60 (0.44 ~ 0.82) | 0.001 | 0.62 (0.45 ~ 0.84) | 0.002 | 0.63 (0.46 ~ 0.85) | 0.003 |
| Quartile 4 | 0.88 (0.68 ~ 1.15) | 0.360 | 0.90 (0.68 ~ 1.17) | 0.423 | 0.90 (0.69 ~ 1.17) | 0.431 |
| Quartile 5 | 0.82 (0.63 ~ 1.07) | 0.138 | 0.90 (0.69 ~ 1.17) | 0.430 | 0.89 (0.68 ~ 1.17) | 0.414 |
| Atherosclerotic phenotypes | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 0.67 (0.47 ~ 0.97) | 0.032 | 0.65 (0.45 ~ 0.94) | 0.021 | 0.68 (0.47 ~ 0.98) | 0.039 |
| Quartile 3 | 0.53 (0.33 ~ 0.83) | 0.006 | 0.53 (0.33 ~ 0.83) | 0.006 | 0.54 (0.34 ~ 0.85) | 0.008 |
| Quartile 4 | 0.80 (0.54 ~ 1.17) | 0.250 | 0.77 (0.52 ~ 1.15) | 0.202 | 0.79 (0.53 ~ 1.17) | 0.236 |
| Quartile 5 | 0.71 (0.48 ~ 1.04) | 0.080 | 0.73 (0.49 ~ 1.09) | 0.126 | 0.73 (0.49 ~ 1.09) | 0.124 |
| Stroke | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 0.51 (0.28 ~ 0.94) | 0.030 | 0.52 (0.29 ~ 0.96) | 0.038 | 0.54 (0.29 ~ 0.99) | 0.046 |
| Quartile 3 | 0.59 (0.30 ~ 1.16) | 0.125 | 0.65 (0.33 ~ 1.27) | 0.210 | 0.67 (0.34 ~ 1.31) | 0.237 |
| Quartile 4 | 0.60 (0.31 ~ 1.15) | 0.122 | 0.67 (0.34 ~ 1.30) | 0.234 | 0.68 (0.35 ~ 1.32) | 0.254 |
| Quartile 5 | 0.67 (0.37 ~ 1.23) | 0.197 | 0.83 (0.44 ~ 1.54) | 0.552 | 0.82 (0.44 ~ 1.54) | 0.541 |
| Angina | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 0.54 (0.28 ~ 1.03) | 0.062 | 0.52 (0.27 ~ 0.99) | 0.050 | 0.54 (0.28 ~ 1.04) | 0.066 |
| Quartile 3 | 0.55 (0.25 ~ 1.17) | 0.119 | 0.54 (0.25 ~ 1.16) | 0.112 | 0.55 (0.26 ~ 1.19) | 0.131 |
| Quartile 4 | 0.61 (0.30 ~ 1.24) | 0.174 | 0.60 (0.29 ~ 1.23) | 0.160 | 0.61 (0.29 ~ 1.25) | 0.176 |
| Quartile 5 | 0.50 (0.24 ~ 1.05) | 0.067 | 0.51 (0.24 ~ 1.08) | 0.078 | 0.52 (0.25 ~ 1.10) | 0.086 |
| Heart attack | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 1.11 (0.60 ~ 2.04) | 0.749 | 1.00 (0.54 ~ 1.85) | 0.995 | 1.08 (0.58 ~ 2.01) | 0.803 |
| Quartile 3 | 0.47 (0.19 ~ 1.18) | 0.110 | 0.43 (0.17 ~ 1.07) | 0.071 | 0.44 (0.17 ~ 1.10) | 0.079 |
| Quartile 4 | 1.30 (0.68 ~ 2.48) | 0.421 | 1.08 (0.56 ~ 2.08) | 0.822 | 1.10 (0.57 ~ 2.12) | 0.774 |
| Quartile 5 | 0.92 (0.46 ~ 1.84) | 0.813 | 0.79 (0.39 ~ 1.60) | 0.512 | 0.79 (0.39 ~ 1.58) | 0.500 |
| Congestive heart failure | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 0.48 (0.12 ~ 1.90) | 0.294 | 0.46 (0.11 ~ 1.85) | 0.275 | 0.48 (0.12 ~ 1.94) | 0.302 |
| Quartile 3 | 0.73 (0.18 ~ 2.92) | 0.655 | 0.77 (0.19 ~ 3.11) | 0.714 | 0.75 (0.18 ~ 3.06) | 0.691 |
| Quartile 4 | 0.91 (0.26 ~ 3.23) | 0.885 | 0.86 (0.24 ~ 3.18) | 0.827 | 0.84 (0.22 ~ 3.11) | 0.789 |
| Quartile 5 | 0.83 (0.24 ~ 2.96) | 0.779 | 0.94 (0.26 ~ 3.44) | 0.923 | 0.92 (0.25 ~ 3.36) | 0.901 |
| Arrhythmia | ||||||
| Igf1 Group | ||||||
| Quartile 1 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Quartile 2 | 0.81 (0.58 ~ 1.14) | 0.228 | 0.82 (0.58 ~ 1.14) | 0.234 | 0.83 (0.59 ~ 1.16) | 0.264 |
| Quartile 3 | 0.70 (0.47 ~ 1.04) | 0.081 | 0.73 (0.49 ~ 1.10) | 0.134 | 0.74 (0.49 ~ 1.11) | 0.143 |
| Quartile 4 | 0.96 (0.67 ~ 1.36) | 0.805 | 1.00 (0.69 ~ 1.43) | 0.983 | 0.98 (0.68 ~ 1.41) | 0.925 |
| Quartile 5 | 0.88 (0.62 ~ 1.26) | 0.493 | 1.00 (0.70 ~ 1.43) | 0.995 | 0.99 (0.69 ~ 1.42) | 0.962 |
Model 1: Unadjusted
Model 2: Adjusted for age, sex, marital status, and education level
Model 3: Further adjusted for smoking status, drinking status, diabetes, hypertension, and hypercholesterolemia on the basis of Model 2
Fig. 2.
Kaplan–Meier curves illustrating the association between serum IGF1 concentrations and the cumulative incidence of CVD
Exposure-response relationship and subgroup analysis
Serum IGF1 levels showed a nonlinear association with CVD (Fig. 3a). Specifically, the exposure–response curve indicated that lower IGF1 concentrations were associated with increased CVD risk, whereas at higher IGF1 concentrations, the risk slightly increased. Serum IGF1 levels also exhibited a nonlinear pattern with arrhythmias and atherosclerotic phenotypes (p for nonlinear < 0.05), but the overall associations were not statistically significant (p for overall > 0.05) (Fig. 3f and g). For other outcomes, restricted cubic spline curves did not indicate a clear association between IGF1 concentrations and risk(Fig. 3b and e).
Fig. 3.
Restricted cubic spline (RCS) curves depicting the association between serum IGF1 concentrations and CVD risk: (a) IGF1 and CVD; (b) IGF1 and stroke; (c) IGF1 and angina; (d) IGF1 and heart attack; (e) IGF1 and congestive heart failure; (f) IGF1 and arrhythmia; (g) IGF1 and atherosclerotic phenotypes
Subgroup analyses showed that the association between IGF1 and CVD risk was generally consistent across most strata, with interaction tests mostly non-significant (all P for interaction > 0.05), supporting the robustness of the overall effect. Heterogeneity was observed only in the sex-stratified analysis, with a stronger association in males (HR = 1.02, 95% CI: 1.00–1.04, P < 0.001) (Fig. 4).
Fig. 4.
Subgroup analyses of the association between serum IGF1 concentrations and the risks of CVD adjusted for age, gender, educational level, marital status, smoking status, drinking status, diabetes, hypertension, and hypercholesterolemia
Mediation analysis
Causal mediation analysis revealed potential mediating effects of TyG-related indices in the association between IGF1 and CVD. The results showed that the average causal mediation effects (ACME) of TyG-BMI, TyG-WC, and TyG-WHtR were statistically significant (p < 0.05) (TyG-BMI: ACME = 0.064, 95% CI: 0.018–0.13; TyG-WC: ACME = 0.051, 95% CI༚0.011–0.11; TyG-WHtR: ACME = 0.053, 95% CI༚0.005–0.11). In contrast, the average direct effect (ADE) and total effect in these models were not statistically significant, suggesting that the effect of IGF1 on CVD may be primarily mediated through these pathways. Notably, no significant mediating effect was observed for TyG alone. These findings indicate that the potential association between IGF1 and CVD is more likely to depend on obesity-related indices associated with insulin resistance, rather than the simple TyG index.
Sensitivity analysis
The consistency of results across multiple validation approaches further supports the robustness of our primary findings. First, using Wave 4 participants as the main analytic sample, the Cox proportional hazards model yielded results consistent with the previous analyses (Supplementary Table 2). Second, in the mediation analysis of the IGF1–CVD relationship, TyG-related indices consistently demonstrated significant mediation effects: TyG-BMI (ACME = 0.067, 95% CI: 0.018–0.140), TyG-WC (ACME = 0.054, 95% CI: 0.008–0.120), and TyG-WHtR (ACME = 0.069, 95% CI: 0.017–0.140). Third, similar results were observed after multiple imputation for missing data(Supplementary Table 4). Additionally, after excluding CVD events occurring within the first two years of follow-up, the association between IGF1 and CVD risk remained robust (Supplementary Table 5). In the landmark analysis, the inverse associations persisted (Q2: HR = 0.70, 95% CI: 0.54–0.91, E-value = 2.21; Q3: HR = 0.64, 95% CI: 0.47–0.88, E-value = 2.50), indicating that only a relatively strong unmeasured confounder could fully explain the observed associations. Collectively, these findings further corroborate the robustness of the IGF1–CVD association and support the potential mediating role of TyG-related indices.
Discussion
This study, based on the English Longitudinal Study of Ageing (ELSA), found a nonlinear relationship between serum IGF1 levels and overall CVD risk: moderate IGF1 levels (approximately Q2–Q3) were associated with lower risk, while the lowest and highest quintiles were associated with higher risk. Mediation analysis suggested that TyG-BMI, TyG-WC, and TyG-WHtR indices may represent potential pathways linking IGF1 levels with CVD risk (ACME: 0.064, 0.051, and 0.053, all P < 0.05), implying that the association may partly reflect insulin resistance × obesity-related mechanisms. In contrast, the TyG index alone did not show a potential mediating role, indicating that combining glycemic–lipid imbalance with body fat distribution may better reflect the underlying mechanisms linking IGF1 and CVD risk [30–34].
Many epidemiological studies have investigated the relationship between IGF1 levels and the incidence of CVD [7, 35]. Large cohort and genetic studies also support a U-shaped association between IGF1 and CVD/metabolic risk, which may be partly related to insulin resistance [10–12].Existing evidence indicates that IGF1 exerts cardiovascular protective effects by promoting endothelial cell survival, enhancing nitric oxide (NO) production, and facilitating vasodilation [36]; in contrast, insufficient IGF1 levels are associated with endothelial dysfunction, reduced vascular protective capacity, and increased arterial stiffness, often accompanied by metabolic syndrome-related phenotypes such as dyslipidemia, abdominal obesity, and sarcopenia, thereby further elevating CVD risk [9, 10].Although chronically high IGF1 levels may confer protective effects under certain conditions, sustained activation of the PI3K/Akt and MAPK signaling pathways can lead to myocardial hypertrophy, fibrosis, metabolic dysregulation, and atherosclerosis, thereby increasing the risk of CVD [37].In comparison, moderate IGF1 levels help maintain insulin sensitivity, anti-inflammatory status, and endothelial function, promoting vascular repair and tissue remodeling [5, 7, 38].Overall, IGF1 exhibits a “moderate effect” on cardiovascular health, with both low and high levels potentially increasing the risk of cardiovascular events through distinct mechanisms.
In subgroup analyses, sex appeared to modulate the association between IGF1 and CVD risk. This difference may be biologically explained, for example, by the cardiovascular protective effects of estrogen in females, which improves endothelial function, modulates lipid metabolism, and suppresses inflammation, thereby lowering CVD risk [39]. Additionally, males and females differ in body fat distribution, insulin sensitivity, and metabolic response, with males potentially more susceptible to high IGF1 levels or related metabolic indices [40, 41], highlighting the need to consider sex in assessing IGF1–CVD associations for more accurate clinical risk stratification.
Composite indices combining TyG and body fat distribution (TyG-WC, TyG-BMI, and TyG-WHtR) better characterize the association between IGF1 levels and CVD risk compared with the TyG index alone (reflecting only fasting glucose and lipids). These composite indices may reflect potential metabolic pathways underlying the IGF1–CVD relationship. They are considered reliable surrogates for insulin resistance [42], suggesting that associations between IGF1 and cardiovascular health may partly operate through insulin sensitivity-related mechanisms [8].High TyG composite indices may also indicate more complex coronary artery disease, including multi-vessel or diffuse lesions, potentially increasing procedural complexity and the need for intensified or prolonged antithrombotic therapy [43]. This highlights the potential clinical relevance of TyG composite indices in CVD risk assessment.
Recent studies and meta-analyses have further emphasized the importance of combining TyG with body fat distribution indices for CVD risk evaluation [30, 33, 34, 44–46]. For example, Ren et al. reported that changes in TyG-WHtR were independently associated with CVD risk [33]; Rao et al. described the potential pathophysiological link between TyG-BMI and CVD and suggested its incorporation into risk prediction models to improve performance [34]; Tang et al. [45]found associations of TyG-BMI, TyG-WC, and TyG-WHtR with cardiac risk, with stronger links for TyG-BMI and TyG-WC. Overall, combining TyG with body fat distribution may more comprehensively reflect the joint impact of insulin resistance and adiposity on cardiovascular health, potentially improving CVD risk prediction accuracy.
The U-shaped risk pattern of IGF1 suggests that it may serve as a useful complement to traditional CVD risk prediction tools (e.g., Framingham and ASCVD scores), helping to identify high-risk subgroups at extreme IGF1 levels. Easily obtainable composite indices, including TyG-WC, TyG-BMI, and TyG-WHtR, may serve as practical markers for monitoring insulin resistance and metabolic vulnerability, thereby enhancing clinical risk stratification. Individuals with markedly low or high IGF1 levels accompanied by elevated TyG composite indices may benefit from lifestyle interventions, weight management, dietary adjustment, regular exercise, and metabolic therapies (e.g., GLP-1RA, SGLT2i). Future studies could explore combined interventions targeting IGF1 regulation and insulin sensitivity improvement to mitigate cardiometabolic risk. Taken together, IGF1 and TyG composite indices not only facilitate risk stratification but may also help identify individuals requiring closer monitoring or early intervention, providing a basis for personalized preventive strategies.
Strengths and limitations
This study innovatively explored the potential role of TyG-related composite indices in the association between IGF1 and CVD, and the use of a large longitudinal cohort with a prospective design enhances the reliability of the findings. However, several limitations should be acknowledged. First, as an observational study, residual confounding—such as diet, physical activity, and medication use—may still influence the results despite adjustment for multiple covariates. Second, IGF1 and TyG-related indices were measured only at baseline, which limits the ability to satisfy strict temporal assumptions required for causal mediation analysis. In addition, CVD outcomes were primarily based on self-reported physician diagnoses, which may miss asymptomatic events and introduce non-differential misclassification, potentially biasing the total effect toward the null. The subgroup findings should be validated in larger samples or other populations, and since the cohort consists mainly of middle-aged and older adults, generalizability may be limited. Future research should replicate these associations across different age groups and high-risk populations to further assess the robustness and broader applicability of the results.
Conclusions
This large-scale longitudinal cohort study demonstrates that both low and high serum IGF1 levels are associated with an increased risk of cardiovascular disease (CVD). We further observed that TyG-related indices, including TyG-WC, TyG-BMI, and TyG-WHtR, may mediate the association between IGF1 and CVD risk. These findings underscore the potential value of incorporating IGF1 and TyG-related indices into risk assessment frameworks, which could facilitate more precise identification of high-risk populations and provide a scientific basis for implementing individualized prevention strategies, thereby mitigating the CVD burden associated with abnormal IGF1 levels.
Supplementary Information
Acknowledgements
All authors gratefully acknowledge the original data collectors, depositors, copyright holders, and funders of the English Longitudinal Study of Ageing (ELSA).
Questionnaire and data source
The questionnaires used in this study were obtained from the English Longitudinal Study of Ageing (ELSA), a well-established and nationally representative longitudinal cohort. The ELSA questionnaires were developed by the ELSA research team and have been widely used in previous studies. Detailed information on the study design, data collection procedures, and questionnaires is publicly available on the official ELSA website (https://www.elsa-project.ac.uk) and through the UK Data Service (https://ukdataservice.ac.uk).
Clinical trial number
Not applicable.
Abbreviations
- IGF1
Insulin-like growth factor-1
- CVD
Cardiovascular diseases
- ELSA
English Longitudinal Study of Ageing
- IR
Insulin resistance
- TyG
Triglyceride and glucose index
- BMI
Body mass index
- WC
Waist circumstance
- WHtR
Waist to height ratio
- TyG-BMI
TyG index in relation to BMI
- TyG-WC
TyG index in relation to WC
- TyG-WHtR
TyG index in relation to WHtR
- ACME
Average causally mediated effect
- ADE
Average direct effect
- TE
Total effects
Authors’ contributions
Haipeng Chen drafted the manuscript and organized the preparation of figures and tables. All authors contributed to patient data collection, participated in data interpretation, and provided critical revisions. All authors read and approved the final version of the manuscript.
Funding
This work was supported by the Natural Science Foundation Project of Chongqing, Chongqing Science and Technology Commission (Award number(s): CSTB2023NSCQ-BHX0068.)
Data availability
The data that support the findings of this study are available from the English Longitudinal Study of Ageing (ELSA) through the UK Data Service. Access can be obtained via the official website: https://ukdataservice.ac.uk.
Declarations
Ethics approval and consent to participate
The English Longitudinal Study of Ageing (ELSA) was approved by the London Multicentre Research Ethics Committee. All participants provided written informed consent prior to participation. All procedures involving human participants were conducted in accordance with the ethical standards of the institutional and/or national research committee and with the Declaration of Helsinki and its later amendments.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Jin Chen, Email: chenjin@hospital.cqmu.edu.cn.
Xu Luo, Email: luoxu2009@hospital.cqmu.edu.cn.
References
- 1.Sacco RL, Roth GA, Reddy KS, Arnett DK, Bonita R, Gaziano TA, Heidenreich PA, Huffman MD, Mayosi BM, Mendis S, et al. The heart of 25 by 25: achieving the goal of reducing global and regional premature deaths from cardiovascular diseases and stroke: A modeling study from the American heart association and world heart federation. Circulation. 2016;133(23):e674–690. [DOI] [PubMed] [Google Scholar]
- 2.Dagenais GR, Leong DP, Rangarajan S, Lanas F, Lopez-Jaramillo P, Gupta R, Diaz R, Avezum A, Oliveira GBF, Wielgosz A, et al. Variations in common diseases, hospital admissions, and deaths in middle-aged adults in 21 countries from five continents (PURE): a prospective cohort study. Lancet. 2020;395(10226):785–94. [DOI] [PubMed] [Google Scholar]
- 3.Global regional. National disability-adjusted life-years (DALYs) for 359 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet. 2018;392(10159):1859–922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Roth GA, Mensah GA, Johnson CO, Addolorato G, Ammirati E, Baddour LM, Barengo NC, Beaton AZ, Benjamin EJ, Benziger CP, et al. Global burden of cardiovascular diseases and risk Factors, 1990–2019: update from the GBD 2019 study. J Am Coll Cardiol. 2020;76(25):2982–3021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lee WS, Abel ED, Kim J. New insights into IGF-1 signaling in the heart. Physiol (Bethesda). 2024;39(5):0. [DOI] [PubMed] [Google Scholar]
- 6.Troncoso R, Ibarra C, Vicencio JM, Jaimovich E, Lavandero S. New insights into IGF-1 signaling in the heart. Trends Endocrinol Metab. 2014;25(3):128–37. [DOI] [PubMed] [Google Scholar]
- 7.Macvanin M, Gluvic Z, Radovanovic J, Essack M, Gao X, Isenovic ER. New insights on the cardiovascular effects of IGF-1. Front Endocrinol (Lausanne). 2023;14:1142644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Fluca AL, Pani B, Janjusevic M, Zwas DR, Abraham Y, Calligaris M, Beltrami AP, Campos Corgosinho F, Marketou M, D’Errico S, et al. Unraveling the relationship among insulin resistance, IGF-1, and amyloid-beta 1–40: is the definition of type 3 diabetes applicable in the cardiovascular field? Life Sci. 2024;352:122911. [DOI] [PubMed] [Google Scholar]
- 9.Schiaffino S, Dyar KA, Ciciliot S, Blaauw B, Sandri M. Mechanisms regulating skeletal muscle growth and atrophy. Febs J. 2013;280(17):4294–314. [DOI] [PubMed] [Google Scholar]
- 10.Lin J, Yang L, Huang J, Liu Y, Lei X, Chen R, Xu B, Huang C, Dou W, Wei X, et al. Insulin-Like growth factor 1 and risk of cardiovascular disease: results from the UK biobank cohort study. J Clin Endocrinol Metab. 2023;108(9):e850–60. [DOI] [PubMed] [Google Scholar]
- 11.Mukama T, Srour B, Johnson T, Katzke V, Kaaks R. IGF-1 and risk of morbidity and mortality from Cancer, cardiovascular Diseases, and all causes in EPIC-Heidelberg. J Clin Endocrinol Metab. 2023;108(10):e1092–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Larsson SC, Michaëlsson K, Burgess S. IGF-1 and cardiometabolic diseases: a Mendelian randomisation study. Diabetologia. 2020;63(9):1775–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Nijenhuis-Noort EC, Berk KA, Neggers S, Lely AJV. The fascinating interplay between growth Hormone, insulin-Like growth Factor-1, and insulin. Endocrinol Metab (Seoul). 2024;39(1):83–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ormazabal V, Nair S, Elfeky O, Aguayo C, Salomon C, Zuñiga FA. Association between insulin resistance and the development of cardiovascular disease. Cardiovasc Diabetol. 2018;17(1):122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Tao LC, Xu JN, Wang TT, Hua F, Li JJ. Triglyceride-glucose index as a marker in cardiovascular diseases: landscape and limitations. Cardiovasc Diabetol. 2022;21(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Guerrero-Romero F, Simental-Mendía LE, González-Ortiz M, Martínez-Abundis E, Ramos-Zavala MG, Hernández-González SO, Jacques-Camarena O, Rodríguez-Morán M. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab. 2010;95(7):3347–51. [DOI] [PubMed] [Google Scholar]
- 17.Yang Y, Huang X, Wang Y, Leng L, Xu J, Feng L, Jiang S, Wang J, Yang Y, Pan G, et al. The impact of triglyceride-glucose index on ischemic stroke: a systematic review and meta-analysis. Cardiovasc Diabetol. 2023;22(1):2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Yu Y, Gu M, Huang H, Cheng S, Deng Y, Cai C, Chen X, Niu H, Ning X, Hua W. Combined association of triglyceride-glucose index and systolic blood pressure with all-cause and cardiovascular mortality among the general population. J Transl Med. 2022;20(1):478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Dang K, Wang X, Hu J, Zhang Y, Cheng L, Qi X, Liu L, Ming Z, Tao X, Li Y. The association between triglyceride-glucose index and its combination with obesity indicators and cardiovascular disease: NHANES 2003–2018. Cardiovasc Diabetol. 2024;23(1):8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ding X, Wang X, Wu J, Zhang M, Cui M. Triglyceride-glucose index and the incidence of atherosclerotic cardiovascular diseases: a meta-analysis of cohort studies. Cardiovasc Diabetol. 2021;20(1):76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sánchez-Íñigo L, Navarro-González D, Fernández-Montero A, Pastrana-Delgado J, Martínez JA. The TyG index May predict the development of cardiovascular events. Eur J Clin Invest. 2016;46(2):189–97. [DOI] [PubMed] [Google Scholar]
- 22.Steptoe A, Breeze E, Banks J, Nazroo J. Cohort profile: the english longitudinal study of ageing. Int J Epidemiol. 2013;42(6):1640–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Freeman Laboratories Newcastle. DPC Immulite 2000 IGF-1. In.; 2012.
- 24.Kang B, Yang Y, Lee EY, Yang HK, Kim HS, Lim SY, Lee JH, Lee SS, Suh BK, Yoon KH. Triglycerides/glucose index is a useful surrogate marker of insulin resistance among adolescents. Int J Obes (Lond). 2017;41(5):789–92. [DOI] [PubMed] [Google Scholar]
- 25.Li Y, Gui J, Liu H, Guo LL, Li J, Lei Y, Li X, Sun L, Yang L, Yuan T, et al. Predicting metabolic syndrome by obesity- and lipid-related indices in mid-aged and elderly chinese: a population-based cross-sectional study. Front Endocrinol (Lausanne). 2023;14:1201132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wang Y, Tan H, Zheng H, Ma Z, Zhan Y, Hu K, Yang Z, Yao Y, Zhang Y. Exposure to air pollution and gains in body weight and waist circumference among middle-aged and older adults. Sci Total Environ. 2023;869:161895. [DOI] [PubMed] [Google Scholar]
- 27.Zhang FL, Ren JX, Zhang P, Jin H, Qu Y, Yu Y, Guo ZN, Yang Y. Strong association of waist circumference (WC), body mass index (BMI), waist-to-Height ratio (WHtR), and waist-to-Hip ratio (WHR) with diabetes: A Population-Based Cross-Sectional study in Jilin Province, China. J Diabetes Res. 2021;2021:8812431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Imai K, Keele L, Tingley D. A general approach to causal mediation analysis. Psychol Methods. 2010;15(4):309–34. [DOI] [PubMed] [Google Scholar]
- 29.Zhang Z, Zheng C, Kim C, Van Poucke S, Lin S, Lan P. Causal mediation analysis in the context of clinical research. Ann Transl Med. 2016;4(21):425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Liu X, Tan Z, Huang Y, Zhao H, Liu M, Yu P, Ma J, Zhao Y, Zhu W, Wang J. Relationship between the triglyceride-glucose index and risk of cardiovascular diseases and mortality in the general population: a systematic review and meta-analysis. Cardiovasc Diabetol. 2022;21(1):124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Jayedi A, Soltani S, Zargar MS, Khan TA, Shab-Bidar S. Central fatness and risk of all cause mortality: systematic review and dose-response meta-analysis of 72 prospective cohort studies. BMJ. 2020;370:m3324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Li X, Tang J, Lin S, Liu X, Li Y. Mendelian randomization analysis demonstrates the causal effects of IGF family members in diabetes. Front Med (Lausanne). 2024;11:1332162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ren Q, Huang Y, Liu Q, Chu T, Li G, Wu Z. Association between triglyceride glucose-waist height ratio index and cardiovascular disease in middle-aged and older Chinese individuals: a nationwide cohort study. Cardiovasc Diabetol. 2024;23(1):247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Rao X, Xin Z, Yu Q, Feng L, Shi Y, Tang T, Tong X, Hu S, You Y, Zhang S, et al. Triglyceride-glucose-body mass index and the incidence of cardiovascular diseases: a meta-analysis of cohort studies. Cardiovasc Diabetol. 2025;24(1):34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.AsghariHanjani N, Vafa M. The role of IGF-1 in obesity, cardiovascular disease, and cancer. Med J Islam Repub Iran. 2019;33:56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wen HJ, Liu GF, Xiao LZ, Wu YG. Involvement of endothelial nitric oxide synthase pathway in IGF–1 protects endothelial progenitor cells against injury from oxidized LDLs. Mol Med Rep. 2019;19(1):660–6. [DOI] [PubMed] [Google Scholar]
- 37.Pires KM, Buffolo M, Schaaf C, David Symons J, Cox J, Abel ED, Selzman CH, Boudina S. Activation of IGF-1 receptors and Akt signaling by systemic hyperinsulinemia contributes to cardiac hypertrophy but does not regulate cardiac autophagy in obese diabetic mice. J Mol Cell Cardiol. 2017;113:39–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Khan MZ, Zugaza JL, Torres Aleman I. The signaling landscape of insulin-like growth factor 1. J Biol Chem. 2025;301(1):108047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Iorga A, Cunningham CM, Moazeni S, Ruffenach G, Umar S, Eghbali M. The protective role of Estrogen and Estrogen receptors in cardiovascular disease and the controversial use of Estrogen therapy. Biol Sex Differ. 2017;8(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Singh P, Covassin N, Marlatt K, Gadde KM, Heymsfield SB. Obesity, body Composition, and sex hormones: implications for cardiovascular risk. Compr Physiol. 2021;12(1):2949–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Kim SH, Reaven G. Sex differences in insulin resistance and cardiovascular disease risk. J Clin Endocrinol Metab. 2013;98(11):E1716–1721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Lim J, Kim J, Koo SH, Kwon GC. Comparison of triglyceride glucose index, and related parameters to predict insulin resistance in Korean adults: an analysis of the 2007–2010 Korean National health and nutrition examination survey. PLoS ONE. 2019;14(3):e0212963. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Castiello DS, Oliva A, Andò G, Niccoli G, Pelliccia F, Moscarella E, Montone RA, Gragnano F, Porto I, Calabrò P, et al. Antithrombotic therapy in complex percutaneous coronary intervention. EuroIntervention. 2025;21(18):e1051–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Miao Y, Wang Y, Wan Q. Association between TyG index with obesity indicators and coronary heart disease: a cohort study. Sci Rep. 2025;15(1):8920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Tang X, Zhang K, He R. The association of triglyceride-glucose and triglyceride-glucose related indices with the risk of heart disease in a National. Cardiovasc Diabetol. 2025;24(1):54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Zhu X, Xu W, Song T, Wang X, Wang Q, Li J, Liu X, Hao B, Chen T, Guo J. Changes in the combination of the triglyceride-glucose index and obesity indicators estimate the risk of cardiovascular disease. Cardiovasc Diabetol. 2024;23(1):192. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the English Longitudinal Study of Ageing (ELSA) through the UK Data Service. Access can be obtained via the official website: https://ukdataservice.ac.uk.