Association between the triglyceride glucose index and low skeletal muscle mass: a cross-sectional study

Introduction

Sarcopenia is an age-related disorder characterised by generalised and continuous skeletal muscle mass loss (SMM-L) and function,1 which increases the risk of adverse consequences, including metabolic dysfunction and disability, together with reduced quality of life.2 Sarcopenia is extremely prevalent in older adults, with incidence rates ranging from 5%‒13% (60‒70 years) to 11%‒50% (over 80 years).3 In general, adults experience a 3%‒8% SMM-L per decade after the age of 30 years.4 This gradual SMM-L is the primary factor contributing to sarcopenia, and also represents the most important characteristic and pathophysiological change seen in sarcopenia.5 Skeletal muscles are the main tissues involved in insulin-induced glucose metabolism,6 insulin is the primary anabolic hormone and insulin resistance (IR) can reduce the synthesis of muscle proteins and mediate their degradation.7 Increased IR is, therefore, one of the main causes of SMM-L, and the early detection and intervention of IR is expected to reduce SMM-L.8

IR is characterised by decreased tissue responsiveness and sensitivity to circulating insulin,9 resulting in reduced glucose uptake by skeletal muscle and decreased glucose utilisation in the liver, and thus raising both the blood glucose and triglyceride levels.10 Quantification of IR is limited by the problems of invasiveness, complex procedures, high cost and ethical concerns. Therefore, non-insulin-based indices of IR have become useful tools for the early detection of IR,11 12 indices such as the triglyceride glucose index (TyGi), fasting blood glucose levels and triglyceride product levels are considered beneficial medical surrogate indices of IR.13 14 Numerous studies have described the value of TyGi in the diagnosis and risk prediction of various diseases related to IR.15–24 In addition, Kim et al found that high TyGi was associated with an increased risk of SMM-L,10 although the study was not sufficiently refined. For a comprehensive evaluation of the correlation between TyGi and SMM-L, the present study included a large sample size, together with subgroup analyses. It is hoped that the findings may lead to the identification of additional predictive indicators of SMM-L and provide a basis for the risk prediction and diagnosis of sarcopenia.

Methods

Study design

This is an analytical cross-sectional study. The data of the participants were extracted from the database of the Health Management Center of the West China Hospital of Sichuan University, from 1 January 2020 to 31 March 2022. The criteria for inclusion were age ≥18 years and completed a body composition test. The exclusion criteria were participants with severe cardiac, cerebrovascular, chronic liver, chronic kidney disease, rheumatic immune system disease, severe malnutrition, cancer, long-term corticosteroid use and/or disability, a use of diuretics, a weight change of >5% during the previous 3 months and if specific data were missing.

Collection and evaluation of TyG

Blood was collected from the anterior elbow veins of the participants after overnight fasting (minimum of 8 hours). Total cholesterol, high-density lipoprotein, low-density lipoprotein, uric acid and fasting blood glucose were evaluated on a self-analyzer (Cobas e801, Roche, Switzerland). Neutrophil, platelet, white blood cell and lymphocyte counts were measured on an autoanalyzer (Sysmex XN-9000, Japan). The specimen collection and laboratory tests were performed according to standard operating procedures. The TyGi was evaluated as follows:

TyG=ln(serum triglycerides (mg/dL))×(plasma glucose (mg/dL))/2.25

Collection and evaluation of SMM-L

Bioelectrical impedance analysis (InBody 570, Omron, Kyoto, Japan), a non-invasive method for determining SMM (InBody 570) was used for body composition assessment.25 Participants wore light indoor clothing and stood upright, arms extended and legs slightly spread. SMM and segmental body composition were calculated using the following formula to calculate the appendicular SMM index (ASMI):

ASMI=total limb lean mass/height2.

In accordance with the Asian Sarcopenia Working Group specifications, SMM-L is described as an ASMI of <5.7 kg/m2 (women) and 7.0 kg/m2 (men).26

Covariates collection

Two investigators independently extracted and assessed data such as sex, age, chronic diseases, smoking history, alcohol consumption, blood pressure and body mass index from the database, with the data subsequently checked by another investigator. Smoking history was defined as having smoked ≥100 cigarettes in one’s lifetime,27 and alcohol consumption was defined as having consumed alcoholic drinks ≥12 times during the previous year.28

Statistical evaluation

Data were entered into Excel (Microsoft, Redmond, Washington, USA) and analysed in SPSS V.21.0 (IBM, Armonk, New York, USA). Baseline analysis was performed after dividing the subjects into two different subgroups according to their sex, age and median TyGi. The descriptive data are presented as numbers and percentages for categorical variables and mean±SD for continuous variables. T-tests were used for ordinal or continuous variables, and the χ2 test for categorical variables to assess the differences between the groups.

Univariate analysis was used to determine SMM-L and its associated factors, and variance inflation factor (VIF) and tolerance values were calculated to diagnose multicollinearity. VIF <10 and values of tolerance >0.1 were used to denote an absence of multico-linearity among the dependent variables.

Multiple logistic regression models were then used to evaluate the unadjusted and adjusted relationships between SMM-L and TyGi. Adjusted variables included those that were found to be significant in the univariate analyses. In addition, subgroup analyses stratified by baseline sex and age (age <40, 40≤age<65 and age ≥65) were performed in the same manner as above.

Statistical tests were two-tailed, and statistically significant values were expressed as p<0.05.

Patient and public involvement

Patients and the public did not participate in the design, recruitment and implementation of research or the dissemination of research results.

Results

Study population characteristics

A total of 36 697 participants were initially recruited. After the exclusion of the 422 potential participants according to the exclusion criteria, 36 275 subjects were included in the study. Of these, 58.46% were male, and the mean age was 43.74±12.33 years (range 18–80 years). The demographic characteristics of all subjects are shown in table 1. Comparisons between subjects with low and high TyGi are summarised in table 2.

Table 1

Demographic characteristics of all subjects

Table 2

Comparisons between subjects with low and high TyG indices

Logistic regression analysis

A correlation between low SMM and TyGi (OR 1.05, 95% CI 1.01 to 1.10, p=0.02) was determined in the univariate analysis, and the correlation remained significant when adjusted for sex and age (OR 1.33, 95% CI 1.27 to 1.40, p<0.001). When other potential confounders were included in the model (OR 1.87, 95% CI 1.75 to 2.00, p<0.001), analogously, a significant correlation was observed in the multiple logistic regression model (table 3).

Table 3

Results of logistic regression analyses showing association between the TyG index and low skeletal muscle mass in overall subjects and subgroups

Subgroup analysis: sex

Table 3 shows that there was a significant correlation between low SMM and TyGi (OR 1.85, 95% CI 1.73 to 1.97, p<0.001) in women when analysed by univariate analysis. The correlation remained significant following correction for age (OR 1.85, 95% CI 1.72 to 1.99, p<0.001) as well as for adjustment for other potential confounders (OR 2.59, 95% CI 2.39 to 2.87, p<0.001). In men, as shown in table 3, the univariate analysis found a significant correlation between SMM-L and TyGi (OR 1.12, 95% CI 1.05 to 1.20, p<0.001) when analysed by univariate analysis. The correlation was significant following correction for age (OR 1.08, 95% CI 1.010 to 1.15, p=0.027) as well as for other potential confounders (OR 1.60, 95% CI 1.46 to 1.76. p<0.001).

Subgroup analysis: age

As shown in table 3, the univariate analysis indicated a correlation between lower SMM and TyGi in younger subjects (OR 0.83, 95% CI 0.77 to 0.89, p<0.001), which remained significant after adjustment for age and sex (OR 1.32, 95% CI 1.22 to 1.42, p<0.001). Following correction for other potential confounders, the association was still significant (OR 1.97, 95% CI 1.77 to 2.20, p<0.001).

Table 3 also shows that in middle-aged adults, there was a significant correlation between low SMM and TyGi (OR 1.18, 95% CI 1.11 to 1.25, p<0.001) when analysed by univariate analysis. In addition, following adjustment for age and sex, the correlation remained significant (OR 1.52, 95% CI 1.42 to 1.62, p<0.001) as well as after adjustment for other potential confounders (OR 1.95, 95% CI 1.77 to 2.14, p<0.001).

Table 3 also shows that there was a significant association between low SMM and TyGi in older adults (OR 1.32, 95% CI 1.14 to 1.54, p<0.001) when analysed by univariate analysis. The association remained significant after correction for sex and age (OR 1.31, 95% CI 1.12 to 1.53, p<0.001) as well as for other potential confounders (OR 1.73, 95% CI 1.38 to 1.16, p<0.001).

Discussion

The results showed a strong association between low SMM and high TyGi in the general population, even in younger and middle-aged adults, confirming that there was a significant independent relationship between low SMM and high TyGi in the overall study subjects, including the different subgroups based on age and sex.

SMM-L is an essential factor in the diagnosis of sarcopenia,26 and has been linked to disability29 30 and overall mortality31 in older adults, thus increasing both medical and socio-economic burdens.32 Therefore, maintaining skeletal muscle mass is critical to the prevention of sarcopenia, preserving physical function in older adults, and thus allowing them to maintain an independent lifestyle. Recently, IR has been reported to be an independent risk factor for SMM-L. In a prospective cohort study, older men with IR showed significantly reduced SMM during 4.6 years of follow-up.33 Another cohort study showed that the SMM-L rates in patients with diabetes treated with insulin sensitisers were significantly lower than those in untreated patients.33 Insulin represents a primary anabolic hormone and promotes protein synthesis by activation of the mammalian target of rapamycin complex-1. IR reduces the synthesis of muscle proteins and promotes protein degradation.7 Thus, IR represents an important pathophysiological mechanism underlying SMM-L. IR may also be reversible and thus anti-IR interventions could reduce SMM-L. Conversely, skeletal muscle is major insulin-sensitive tissue, contributing approximately 40% of body weight. SMM-L has also been reported to trigger IR.34 35

Currently, there is no specific method for the accurate determination of IR. It is commonly quantified using the hyperinsulinaemia-blood glucose clamp test, the gold standard method for measuring IR.36 However, this method is limited by its invasiveness, procedural complexity, high cost and ethical concerns about its clinical application. Thus, although it is used in academic research, it is currently not used in therapeutic approaches. Currently, the index for homeostasis model assessment-estimated IR (HOMA-IR) is widely used to assess IR and the function of β-cells but its value is limited to subjects receiving insulin therapy and subjects with non-functioning β-cells and thus for subjects not receiving insulin therapy or with non-functioning β-cells, it may be of limited value.37 As insulin promotes the utilisation of liver glucose together with peripheral glucose uptake, IR increases the levels of plasma glucose and facilitates the conversion of excess glucose to triglycerides in the liver.34 38 Therefore, TyGi represents a useful surrogate marker for IR, which has been demonstrated to be superior to HOMA-IR in assessing IR in patients with or without diabetes.13 The TyGi is based on sound published data and is a low-cost, convenient and simple surrogate index that does not require insulin quantification. In addition, TyGi can be determined in any subject, irrespective of their insulin therapy status.39 A recent study has also demonstrated that TyGi is a useful independent prognostic predictor in patients with cardiovascular disease, with and without diabetes, indicating possible clinical benefits in predicting the risk of cardiovascular disease.40 The TyGi is a measure of IR, and since IR and lower SMM have a pathophysiological relationship, our findings in this study demonstrate that higher TyG indices are correlated with lower SMM values.

The prevalence of SMM-L in this study was found to be 17.71%, with ageing, declining hormone levels, reduced nutritional intake and decreased ability to perform daily activities being important factors contributing to the reduction in skeletal muscle quality. Recommendations for the prevention of SMM-L are the use of resistance exercise, testosterone supplementation and increased protein intake.41 While it is well-known that men lose muscle mass with age, there are limited data on the association of SMM-L incidence with sex. In the present study, the prevalence of SMM-L was 25.6% in females and 12.1% in males, indicating that SMM-L was more prevalent in females. There may be two reasons for this observation. First, steroid hormones, such as testosterone, are responsible for the sex dimorphism associated with SMM, as androgens promote muscle cell anabolism in the release of local hormonal factors, including insulin-like growth factor-1, from myogenic fibroblasts.42 Second, there are gender differences in the social division of labour and eating habits. For example, men are more likely to engage in physical labour, while women are more likely to diet. Calorie control reduces protein intake, thus contributing to the increased prevalence of SMM-L in women compared with men.

Here, it was found that TyGi values were higher in men than in women. Men also showed a greater incidence of dyslipidemia, diabetes and hypertension, indicating that in western China, men have a greater likelihood of developing IR compared with women. According to Trends and Projections of Global Burden of Disease from 2000 to 2019,43 the obesity-related mortality rates are significantly higher in men than in women. This sex difference in favour of women may be associated with the protective effect of oestrogen against the development of metabolic diseases. Oestrogen has been shown to enhance peroxisome proliferator-activated receptor signalling in adipose tissue by decreasing triglycerides and maintaining insulin sensitivity in both skeletal muscle and liver.44 This illustrates the critical issue of developing sex-targeted strategies when addressing metabolic diseases.45

This study has several potential limitations. The first is its cross-sectional design, making it difficult to determine causative correlations. Second, it did not consider specific therapeutically relevant issues, such as physical activity, statin therapy, hypoglycaemic agents, high CHO diet and uncorrectable confounders. Third, the participants were enrolled in a single centre, potentially introducing a risk of bias. Therefore, a multicentre approach that includes relevant potential confounding factors is needed in the future.

This post was originally published on https://bmjopen.bmj.com