DOI: https://doi.org/10.21203/rs.3.rs-2525210/v1
The association between triglyceride glucose (TYG) index and depression is unclear. We conducted this analysis to explore whether higher TYG index is associated with a higher odd of depression.
The objective was to investigate the relationship between TYG and abdominal aortic calcification (AAC) in people over 60 years old.
The National Health and Nutrition Examination Survey data were analyzed using logistic regression models to examine the independent association between TYG index and the Kauppila AAC-24 score.
A total of 1,408 people took part in our study. Participants with higher TYG quartiles had higher AAC scores. SAAC was defined as a Kauppila score > 6, and the prevalence of SAAC was 17.0%. After adjusting for relevant covariates, the multivariable-adjusted odds ratios (ORs) and 95% confidence intervals (CI) for participants in the third and fourth quartiles were 1.970 (1.232–3.150, P = 0.005) and 2.261 (1.404–3.644, P < 0.001). Subgroup analyses indicated that the positive association between TYG and SAAC persisted across population subgroups.
Triglyceride glucose index was negatively correlated with SAAC in the elderly.
The pathogenesis of vascular calcification has been a research hotspot for centuries. In the past decade, the research on vascular calcification has become more intensive as vascular calcification markers predict poor outcomes for many patients, especially those with coronary vessel disease (CVD) or chronic kidney disease (CKD) [1]. Most of the initial focus of these studies was on coronary artery calcification, but recently, abdominal aortic calcification has received increasing attention as a possible risk factor for cardiovascular disease [2]. At present, abdominal aortic calcification (AAC) is mainly related to age, gender, and traditional cardiovascular risk factors [3]. However, risk factors for abdominal aortic calcification in the elderly are rarely reported.
Insulin resistance (IR) is a metabolic disorder caused by impaired tissue responsiveness to insulin stimulation, which is mainly manifested as dysfunction of glucose and lipid metabolism [4]. Recently, the triglyceride and glucose (TYG) index has been proposed as a simple surrogate for insulin resistance (IR) [5, 6]. The TYG index has been reported to be associated with stroke, carotid atherosclerosis, microvascular and macrovascular injury, and coronary artery disease [7–9]. Studies have been conducted to link abdominal aortic calcification scores with TgY in adults [10, 11]. However, no study has investigated the association between the TYG index and the risk of abdominal aortic calcification in the elderly. Therefore, we conducted a cross-sectional study to investigate the association between the TYG index and a severe AAC (SAAC) score in an elderly population in the United States, thereby providing a clinical reference.
Data were obtained from the 2013–2014 NHANES cycle. Details regarding the study design and protocol of the NHANES have been reported previously [12]. Briefly, NHANES is a representative US non-institutional civilian resident survey with a stratified, multistage, and probability sampling design. In the NHANES 2013–2014 cohort, a total of 10175 participants completed the study. From these data, patients younger than 60 years of age were first excluded. We subsequently excluded participants with missing data on triglycerides or fasting glucose (N = 147) or an AAC score (N = 67). Finally, 1408 participants were included in this cross-sectional study. A detailed algorithm describing participant selection is shown in Fig. 1. The ethics review board at the National Center for Health Statistics reviewed and approved the NHANES protocol, and all participants provided written informed consent before data collection.
Detailed information on the AAC measurement can be found at https://wwwn.cdc.gov/Nchs/Nhanes/2013-2014/DXXAAC_H.htm. In brief, AAC was measured by dual-energy X-ray absorptiometry (DXA; Hologic Discovery Model A densitometers, Hologic, Inc., Marlborough, Ma, USA) from lumbar lateral (vertebral BMD at L1-L4) scanning, and the Kauppila scoring system was used for quantification [13]. AAC greater than 0 was diagnosed as AAC, and AAC greater than 6 was diagnosed as SAAC [14]. All methods were performed in accordance with the Declaration of Helsinki.
LN[triglycerides (mg/dl)* fasting blood glucose (mg/dl)/2] was used to calculate the TYG index. After overnight fasting, blood was drawn in the morning to measure triglyceride and glucose levels. The concentrations of triglycerides and fasting glucose were measured using an enzymatic assay in an automatic biochemistry analyzer. Serum triglyceride levels were measured using the Roche Modular P and Roche Cobas 6000 chemistry analyzers. Fasting plasma glucose was measured by the hexokinase-mediated reaction on Roche/Hitachi Cobas C-501 chemistry analyzers.
Potential covariates were selected based on clinical relevance and statistical significance, including demographics, comorbidities, lifestyle variables, body mass index (BMI), serum bone mineral metabolism markers, and others. Demographic, comorbidity, and lifestyle information was obtained using questionnaires from interviews completed by trained interviewers in the participants' homes. Medical examination and laboratory evaluation of MEC yielded BMI and other biochemical parameters.Races are classified as non-Hispanic whites, non-Hispanic blacks, Mexican Americans, other Hispanics, or others. The education level is divided into below high school, high school, equivalent education, and college. Hypertension is defined as an ASBP or ADBP of 140/90 mmHg, as currently taking blood pressure medication, or as previously diagnosed by a doctor or other health professional. Diabetes was defined as fasting blood sugar > 7 mmol/L, random blood sugar ≥ 11.1 mmol/L, or A1c ≥ 6.5%, or use of hypoglycemic drugs, or a history of diabetes. The history of CVD was determined from self-reported congestive heart failure, coronary heart disease, angina, heart attacks, and stroke. INDFMPIR is an index for the ratio of family income to poverty. Smoking is defined as smoking at least 100 cigarettes in your life. Serum bone mineral metabolism indicators included total 25-hydroxyvitamin D, serum calcium, and serum phosphorus.
Results were described as weighted mean standard deviations (SE) for continuous variables and frequencies (weighted percentages) for categorical variables according to the NHANES analysis guidelines. The TYG index was analyzed as a continuous variable and a quartile variable. Based on the nature of the data, we performed chi-square, ANOVA, or Kruskal-Wallis H tests to determine differences between participants in different TYG quartiles. Three logistic regression models were developed to assess the association between TYG and SAAC. Model 1 is an unadjusted model. Age, sex, and race were adjusted in Model 2. Model 3 was adjusted for age, sex, race, BMI, education, PIR, hypertension, diabetes, CVD, smoking, 25-hydroxyvitamin D, serum calcium, and serum phosphorus. The potential nonlinear relationship between TYG and SAAC was analyzed by restricted cubic spline (RCS). We also calculated Spearman correlation coefficients to assess the association of TYG with some cardiovascular risk factors. Subgroup analyses were performed on the basis of Model 3 to assess whether gender, hypertension, diabetes, CVD, BMI, and smoking affected the relationship between TYG and SAAC. P value < 0.05 was statistically significant. We performed all statistical analyses using R software (Version 4.1, Vienna, Austria) and IBM SPSS Statistics Version 23.0( Chicago, IL, USA).
The baseline characteristics of the included participants are shown in Table 1. Except for age, gender, CVD, smoking, serum calcium and phosphorus, and 25 (OH) D, there were significant differences in TYG among the four quartiles. Participants with higher TYG scores (Q 3 and Q 4) were more likely to be hypertensive, diabetic, poor, and obese compared to those with lower TYG scores (quartiles 1 and 2). In addition, poor education was more common among participants with higher TYG scores. More importantly, we observed that participants with higher TYG quartiles tended to have higher SAAC rates (interquartile range: 1:13.6%, interquartile range: 2:15.6%, interquartile range: 3:19.3%, interquartile range: 4:19.3%, P = 0.014).
| Variable | Total(n = 1408) | Q1(n = 352) | Q2(n = 352) | Q3(n = 352) | Q4(n = 352) | p-Value |
|---|---|---|---|---|---|---|
| Age(years) | 69.61 ± 6.72 | 69.93 ± 6.81 | 69.91 ± 6.75 | 69.48 ± 6.76 | 69.11 ± 6.54 | 0.310 |
| Male ,n (%) | 676(48.0%) | 161(45.7%) | 172 (48.9%) | 163(46.3%) | 180(51.1%) | 0.454 |
| Hypertension, n (%) | 883(62.7%) | 198(56.3%) | 209(59.4%) | 236(67.0%) | 240(68.2%) | 0.001 |
| Diabetes, n (%) | 317(22.5%) | 34(9.7%) | 51(14.5%) | 85(24.1%) | 147(41.8%) | < 0.001 |
| CVD, n (%) | 302(21.4%) | 74(21.0) | 69(19.6) | 68(19.3) | 91(25.9) | 0.124 |
| Education level n (%) | 0.015 | |||||
| Less than high school | 163(11.6) | 25(7.1) | 42(11.9) | 41(11.6) | 55(15.6) | |
| High school diploma or GED | 191(22.7) | 49(13.9) | 40(11.4) | 48(13.6) | 54(15.3) | |
| More than high school | 1054(74.9) | 278(79.0) | 270(76.7) | 263(74.7) | 243(69.0) | |
| Race, n (%) | < 0.001 | |||||
| Mexican American | 163(11.6) | 25(7.1) | 39(11.1) | 43(12.2) | 56(15.9) | |
| Other Hispanic | 128(9.1) | 24(6.8) | 31(8.8) | 37(10.5) | 36(10.2) | |
| Non-Hispanic white | 687(48.8) | 163(46.3) | 170(48.3) | 180(51.1) | 174(49.4) | |
| Non-Hispanic black | 275 (19.5) | 115(32.7) | 70(19.9) | 56(15.9) | 34(9.7) | |
| Other races | 155(11.0) | 25(7.1) | 42(11.9) | 36(10.2) | 52(14.8) | |
| Smoking, n (%) | 688(48.9%) | 159(45.2%) | 172(48.9%) | 182(51.7%) | 175(49.7%) | 0.164 |
| PIR | 2.61 ± 1.59 | 2.78 ± 1.61 | 2.69 ± 1.59 | 2.57 ± 1.57 | 2.40 ± 1.56 | 0.015 |
| BMI kg/m2 | 28.14 ± 5.24 | 26.30 ± 5.11 | 27.79 ± 5.01 | 29.00 ± 5.51 | 29.48 ± 4.72 | < 0.001 |
| Ca, mg/dL | 9.49 ± 0.37 | 9.47 ± 0.40 | 9.45 ± 0.33 | 9.50 ± 0.36 | 9.52 ± 0.39 | 0.093 |
| Phosphorus ,mg/dl | 3.79 ± 0.56 | 3.82 ± 0.54 | 3.79 ± 0.51 | 3.79 ± 0.57 | 3.77 ± 0.60 | 0.741 |
| 25(OH)D, nmol/l | 77.17 ± 31.63 | 79.54 ± 33.19 | 77.99 ± 32.31 | 76.03 ± 32.05 | 75.12 ± 28.74 | 0.246 |
| SAAC | 239(17.0) | 48(13.6) | 55(15.6) | 68(19.3) | 68(19.3) | 0.014 |
Values are given as mean ± standard deviation or numbers and percentages. Q1: TYG < 8.36; Q2: 8.36–8.79; Q3:8.79–9.29; Q4: TYG > 9.29. BMI,body mass index; CVD, cardiovascular ; GED, general educational development; PIR, poverty income ratio; AAC, abdominal aortic calcification.
The relationship between the SAAC and the TYG index is shown in Table 2 as continuous and categorical variables. When TYG index was analyzed as a continuous variable, in the univariate logistic regression model, each unit increase in TYG index was associated with a higher probability of SAAC (OR = 1.297, 95% CI: 1.070–1.572). This association remained statistically significant in models 2 and 3. When TYG was used as a quartile based categorical variable and the first quartile was used as a reference, participants in the third and fourth quartiles had a higher risk of SAAC in all three models. Univariate analysis showed that the SAAC with 95% CI in the quartile of TYG increase was 1.516 (1.103–2.270) and 1.516 (1.103–2.270), respectively. After adjusting for race, age, and sex, the OR (95% CI) for SAAC in the Q3 and Q4 quartiles were 1.687 (1.103–2.579) and 1.798 (1.175–2.751), respectively. After all variables were adjusted, the results showed that the OR (95% CI) of SAAC in the Q3 and Q4 was 1.970 (1.232–3.150) and 2.261 (1.404–3.644), respectively.
| TgY index | Model1 | Modle2 | Modle3 | |||
|---|---|---|---|---|---|---|
| OR(95%CI) | p | OR(95%CI) | p | OR(95%CI) | p | |
| Continuous | 1.297(1.070–1.572) | 0.008 | 1.445(1.176–1.776) | < 0.001 | 1.620(1.280–2.050) | < 0.001 |
| Categorical | ||||||
| Q1 | Reference | Reference | Reference | |||
| Q2 | 0.947(0.699–1.284) | 0.453 | 1.205(0.788–1.866) | 0.403 | 1.404(0.876–2.250) | 0.159 |
| Q3 | 1.516(1.103–2.270) | 0.043 | 1.687(1.103–2.579) | 0.016 | 1.970(1.232–3.150) | 0.005 |
| Q4 | 1.516(1.103–2.270) | 0.043 | 1.798(1.175–2.751) | 0.007 | 2.261(1.404–3.644) | 0.001 |
Data are presented as odds ratios, 95% CIs (confidence intervals), and p-value. Model 1 adjusted for none. Model
2 adjusted for age, sex, and race. Model 3 adjusted for all covariates. TYG triglyceride glucose index
As shown in Fig. 2, we also used a Restricted Cubic Spline to visualize the association between TYG and SAAC. In the curve, we found that when TYGf was in the Q1-Q2 interval (TYG < 8.79), OR less than 1, Thereafter, the OR increased sharply until TYG exceeded 8.8.
Subgroup analyses were conducted by gender, hypertension, diabetes, CVD, BMI, and smoking. TYG was further treated as a continuous variable. As shown in the forest plot (Fig. 3), there was a significant positive correlation between TYG and SAAC. Women (OR = 1.406, 95% CI: 1.176–1.680), those with BMI 28 (OR = 1.784, 95% CI: 1.297–2.454), those without CVD (OR = 1.412, 95% CI: 1.105–1.802), and nonsmokers (OR = 1.120, 95% CI: 1.006–1.246) had these positive associations. In addition, we found that the association between TYG and SAAC was more pronounced in people without CVD (P for interaction = 0.005).
Using a representative national sample of older adults in the United States, we found that a higher TYG index was independently associated with increased odds of SAAC, showing a nearly linear dose-dependent relationship, after adjusting for factors including demographics, cardiovascular risk factors, and multiple other potential covariates. Subgroup analysis showed that the direction of the relationship between the TYG index and the SAAC in different subgroups was consistent with the trend of the study population.
In recent years, the TYG index has been suggested as a surrogate marker of insulin resistance (IR) [15, 16]. IR, a state of reduced sensitivity and responsiveness to insulin action, has been identified as a hallmark of T2DM [17]. There is increasing evidence that IR and its related diseases are associated with the development of CVD in both diabetic and non-diabetic patients [18]. IR patients are known to be prone to multiple metabolic disorders, such as hyperglycemia, dyslipidemia, and hypertension, all of which are strongly associated with adverse CVD outcomes [19]. Therefore, IR is not only considered a pathogenic cause of cardiovascular disease but also a predictor of cardiovascular disease in the general population and in patients with diabetes. Therefore, it is particularly important to develop convenient and reliable screening tools to detect IR and predict cardiovascular risk. Previously, the homeostasis model assessment-estimated insulin resistance index (HOMA-IR) was a widely used measure of -cell function and IR, but it had limited utility in subjects on insulin therapy or without -cell function [20]. To address this limitation, the TYG index was developed, and the TYG index was used to evaluate insulin resistance. And it has been proven to be superior to HOMA-IR in evaluating IR in diabetic and non-diabetic patients. [21] Many diseases, including hypertension, myocardial infarction, peripheral artery disease, and COVID-19, have been shown to be excellent predictors of TYG [22–25].
Vascular calcification (VC) is defined as the deposition of minerals in the form of calcium phosphate complexes in the vasculature. Although VC is thought to be a normal part of the aging process, certain pathological processes, such as diabetes, hypertension, chronic kidney disease (CKD), and/or rare genetic diseases, may also play a role [26].VC and atherosclerotic vascular disease have an inseparable relationship [27]. In addition, arterial stiffness, which represents VC dysfunction, is known to be an independent predictor of cardiovascular mortality [28]. Elastin loss is accompanied by medial calcification, and elastin degradation is thought to further promote the osteogenic process in aortic tissue [29]. Over the years, studies have revealed various mechanisms of vascular calcification, such as induction of bone formation, apoptosis, altered Ca/P balance, and loss of inhibition. [30] From the above, we can infer that the underlying mechanism of the association between the TYG index and AAC may be related to IR, which involves functional and structural damage of the arterial wall, including impaired vasodilation caused by chemical mediators, reduced arterial wall distensibility (arterial stiffness), vascular calcification, and increased arterial wall thickness [31, 32]. Studies have shown that vascular disease associated with insulin resistance begins early in life. Children and adolescents with insulin resistance exhibit impairment of the arterial system compared to adolescents without insulin resistance, suggesting that insulin resistance plays a crucial role in the development of initial vascular damage [33, 34]. In adults with T2D pre-clinical onset, asymptomatic subjects are chronically characterized by insulin resistance. Latent vascular dysfunction begins to develop at this stage, so that patients with T2D are at increased cardiovascular risk before the disease is diagnosed [35].
Stratified analyses showed associations that were generally consistent with the main findings. However, we found that some subgroups, such as men, smokers, and patients with low BMI and CVD, had lower ORS than the corresponding subgroups but did not reach statistical significance. Male, smoking, hypertension, and diabetes are typical risk factors for CVD, and the presence of these factors may change and weaken the effect of IR on AAC [36]. In addition, only 70 SAAC participants were included in the subgroup of diabetic patients after stratification. As the reduction in sample size leads to potential bias, the results need to be validated in a larger population in the future. To our knowledge, this is the first study to examine the association between the TYG index and the SAAC in a large and representative national sample of older adults in the United States. Our study has the strengths of a rigorous protocol and quality control, a large representative sample, standardized measures of vascular calcification, and the integration of data on many important covariates from the NHANES studies. However, this study has several limitations. Firstly, as a cross-sectional observational study, a causal relationship between the TYG index and the SAAC cannot be determined. Secondly, the adjustment is partial, and residual confounders should always be ruled out. Finally, because all participants were US residents, generalization of the results to other populations with different demographics may be limited.
In our study, we demonstrated that the TYG index is independently associated with the risk of SAAC in older adults. Our findings may provide support for further large-scale prospective studies to clarify the precise causality of this relationship.
Acknowledgements We thank the public availability of NHANES data and thank all NHANES participants and staff for their valuable efforts.
Authors’ contributions DKP, JLG and YQG contributed to the study design.JYW, WZM and ZXS preformed the data analysis. DKP wrote the manuscript. YQG and JMG critically revised and edited the manuscript for important intellectual content. All authors reviewed and approved the final manuscript.
Funding: This research was funded by the National Key R&D Program of China (2021YFC2500500)
Availability of data and materials The data that support the findings of this study are available from the corresponding author upon reasonable request.
Code availability Not applicable.
Declarations Conflicts of interest We declare that we have no competing interests.
Ethical approval The study was approved (or granted exemption) by the appropriate institutional and/or national research ethics committee.
Consent to participate Not applicable.
Consent for publication Not applicable.