Introduction

Obesity, often termed the epidemic of the 21 st century, is a growing public health issue with rising prevalence and projections1,2,3. A key complication is insulin resistance (IR), a central factor in obesity-related cardiometabolic diseases4,5,6. The Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) index, commonly used to assess IR, is linked to adipose tissue dysfunction, with abdominal fat contributing to elevated HOMA-IR values7,8,9. Visceral adiposity promotes IR via secretion of pro-inflammatory cytokines and adipokines such as asprosin (ASP), disrupting insulin signaling10. Several IR markers based on biochemical and anthropometric data have been proposed11. Among these markers, the triglyceride-glucose index (TyG), calculated using fasting glucose and triglyceride levels, has emerged as a reliable, cost-effective biomarker for detecting IR12. It demonstrates superior sensitivity and specificity compared to conventional indices such as fasting insulin and HOMA-IR (7,13). Its strong correlation with the hyperinsulinemic-euglycemic clamp, the gold standard for insulin sensitivity, underscores its clinical utility14,15,16. Other indices include QUICKI and the Matsuda Index, based on fasting or OGTT data17,18. However, no consensus exists on the most accurate index across populations. Therefore, composite indices integrating lipid–glucose interactions with anthropometric parameters have been proposed to improve detection of hepatic IR and related metabolic risk. Among these, the triglyceride–glucose waist-to-hip ratio (TyG-WHR) index, which combines the TyG index with central adiposity, has shown superior performance in identifying liver-related metabolic dysfunction and early steatosis compared with TyG alone19. Incorporating TyG-WHR into ASP research may therefore provide novel insight into whether this adipokine primarily reflects hepatic rather than peripheral IR.

ASP, a glucogenic adipokine secreted by white adipose tissue during fasting, regulates hepatic glucose release and may serve as a biomarker for IR and body composition10,20,21,22. Elevated ASP correlates with IR and is influenced by fat and lean mass, both of which modulate metabolic responses and may confound IR indices23,24,25. Integrating ASP with body composition measures may improve IR assessment. Existing literature remains inconclusive, necessitating further research on the interplay between ASP, IR, and body composition. Additionally, oxidative stress (OS) is increasingly linked to obesity, IR, and metabolic dysfunction26. Excess reactive oxygen species (ROS) promotes inflammation, insulin signaling impairment, and may enhance ASP expression, worsening metabolic outcomes27,28. Assessing OS markers alongside ASP, IR indices, and body composition could yield a more comprehensive view of metabolic health in obesity.

This study aims to evaluate ASP and OS parameters in overweight and obese individuals, focusing on muscle mass, fat mass, and resting metabolic rate (RMR). It also explores the association between ASP, OS, and IR, incorporating body composition as a confounding factor. Analyses were stratified by sex to capture potential differences. By simultaneously examining ASP levels, OS markers, and detailed body composition metrics, this research seeks to advance the understanding of obesity-related metabolic disturbances and support the development of more targeted and individualized therapeutic approaches.

Results

Biochemical profiling and measurement analysis

Glucose and insulin levels, both fasting and post-load, were elevated in the O1 group compared to controls, with further increases in the O2 group. C-peptide and HbA1c concentrations rose progressively with body mass (Table 1). All groups showed significant differences in anthropometric and body composition parameters, including BMI, fat and muscle mass, total body water (TBW), waist-to-hip ratio (WHR), VAT, adipose tissue %, RMR, and android/gynoid fat (p < 0.0001) (Table 3). Significant variations in IR indices (HOMA-IR, TyG, QUICKI-I, Matsuda-I) were also noted between groups (Table 1).

Table 1 Characterization of biochemical parameters, body composition, and IR indices across study groups.
Full size table
Table 3 The evaluation of body composition and ASP concentration parameters in the study groups stratified by sex.
Full size table

ASP levels were significantly elevated in O1 and O2 versus controls and differed between O1 and O2 (p < 0.0001). Oxidative status did not differ significantly between O1 and O2 (Table 2).

Table 2 Concentrations of ASP and OS parameters in the CG, O1 and O2 groups.
Full size table

Although BMI did not differ between sexes, significant differences were observed in body composition (Table 4). Men had greater muscle mass, TBW, and RMR, whereas women exhibited higher fat mass, adipose tissue %, and gynoid fat (Table 3). In the O2 group, android fat differed significantly by sex (p < 0.0001). ASP levels were higher in women, reaching statistical significance in the O1 group only (p < 0.01) (Table 3).

Table 4 The spearman correlation coefficients between ASP, insulin and glucose concentrations, HOMA-IR, Ty-G, QUICKI and Matsuda index and body composition parameters in the CG, O1 and O2 groups.
Full size table

Correlations

In the CG group, ASP positively correlated with fasting insulin, insulin at 60 and 120 min (OGTT), and TG levels (Table 4). In O1, ASP showed negative correlations with fasting glucose (R = − 0.23, p = 0.02), glucose at 60 min (R = − 0.23, p = 0.01), and 240 min (R = − 0.28, p = 0.0008). Positive correlations were observed with insulin at 120 min (R = 0.39, p < 0.001) and 180 min (R = 0.29, p = 0.008). In O2, ASP correlated positively with fasting insulin and insulin at 60 min post-glucose. In O1, ASP also showed a negative correlation with LDL and a positive one with HDL (Table 4). In CG, ASP correlated positively with gynoid fat and negatively with TBW and skeletal mass. In both O1 and O2, ASP correlated positively with body weight, adipose tissue %, total fat, and gynoid fat, and negatively with skeletal muscle mass, TBW, and RMR. Across all groups, ASP showed positive correlations with HOMA-IR and TyG, and negative correlations with QUICKI and Matsuda (Table 4).

In O1 and O2 groups, TOC showed moderate positive correlations with adipose tissue %, fat mass, gynoid fat, and ASP levels, and negative correlations with skeletal muscle mass, TBW, and RMR (Table 5). TAC was negatively correlated with adipose tissue % in both groups, and with fat and gynoid fat mass in O1. Positive correlations between TAC and skeletal muscle mass, TBW, RMR, and glucose at 60 min were observed in both groups. In O1, TAC also correlated positively with glucose at 240 min (R = 0.36, p < 0.05) (Table 5).

Table 5 The spearman correlation coefficients between OS parameters, body composition, resting metabolic rate, glucose, and ASP in the CG, O1 and O2 groups.
Full size table

Discussion

Given the heterogeneity of previous findings, our study contributes to elucidating the potential role of ASP in metabolic dysfunction29. ASP, considered alongside measures of IR and body composition, shows potential as an early marker of metabolic dysregulation. In normal-weight subjects, it was positively linked to fasting and OGTT-induced insulin responses. Our results corroborate previous studies demonstrating elevated ASP levels in overweight and obese individuals (p < 0.0017, p < 0.0001; respectively)29,30,31. As fat tissue increases, significant relationships between ASP levels and body composition parameters emerge. Consistent with Suder A. et al. research, in our study ASP exhibited a positive correlation with increased body weight (p < 0.001), as well as adipose tissue percentage, total fat mass, and gynoid fat mass in both overweight and obese individuals (p < 0.001)32. These results support Cantay et al., who identified white adipose tissue adipocytes as the main source of ASP33. In normal-weight individuals, ASP correlated positively with gynoid fat mass and inversely with total body water and muscle mass, suggesting regional fat distribution influences ASP secretion independently of obesity. Reduced lean mass may indicate early metabolic impairment. Despite similar BMI, men and women showed distinct, sex-specific body composition patternsConsistent with literature data in our groups muscle mass was higher in men (CG: p < 0.003; O1: p < 0.0001; O2: p < 0.0001), while women had higher fat mass, % adipose tissue, and gynoid fat mass (CG: p < 0.05; O1: p < 0.0001; O2: p < 0.0001)34,35,36. Additionally, in the O2 group, statistically significant differences in android fat were noted between men and women (p < 0.0001), consistent with the research by Camilleri G. et al.37. In contrast to Mirr M. et al., although BMI was similar between men and women, we observed significantly higher ASP levels only in the overweight male group (p < 0.01)38. As noted by Li et al., ASP regulation may be modulated by sex hormones, potentially contributing to differences between men and women39. Although the study controlled for menstrual cycle phase and oral contraceptive use, the predominance of postmenopausal women suggests minimal hormonal influence. Despite similar BMI, variations in fat distribution and sex-specific features of white adipose tissue may underlie observed differences in ASP levels.Furthermore, we observed statistically significant differences in RMR between the O1 and O2 groups compared to healthy controls, as well as between the O1 and O2 groups themselves (p < 0.0001 for all comparisons). This is consistent with previous studies indicating that men typically have a higher RMR due to their greater muscle mass, which requires more energy to maintain40,41.

Beyond its association with traditional IR indices, asprosin also demonstrates significant relationships with markers of inflammation and liver function. These associations appear to vary across different metabolic states, suggesting that asprosin may reflect not only IR but also systemic inflammation and hepatic dysfunction. This aligns with previous findings indicating that ASP promotes hepatic glucose production and is linked to proinflammatory cytokine activation20,21. Taken together, these observations support the potential role of ASP as an integrated biomarker of obesity-related metabolic risk, encompassing endocrine, inflammatory, and hepatic components.

Contrary to earlier reports suggesting reduced ASP levels in advanced obesity due to inflammatory and hormonal dysregulation, our findings show consistent positive correlations of ASP with hepatic enzymes and inverse associations with CRP in overweight and obese, which reflects observations from recent studies linking ASP with liver injury, low-grade inflammation, and metabolic risk25,42,43, our findings demonstrate a positive association between ASP concentrations and increasing body mass and adiposity. Chen et al. suggested that mitochondrial dysfunction, common in obesity, may impair the physiological response to ASP and limit its role in energy metabolism44. In both O1 and O2 groups, ASP correlated negatively with skeletal muscle mass, TBW, and RMR, linking elevated ASP to greater adiposity and reduced lean mass. This supports evidence that excess adiposity promotes muscle loss and dehydration through inflammatory and hormonal pathways45. We also observed correlations between TOC, TAC, and metabolic parameters - ASP was positively associated with TOC, which in turn correlated with fat mass and gynoid fat mass, indicating links between ASP, OS, adiposity, and metabolic dysfunction28,46,47.Previous studies reported negative correlations between TAC and adiposity markers, suggesting reduced antioxidant defenses with greater fat mass due to chronic inflammation48. In contrast, our results showed positive associations of TAC with skeletal muscle mass, TBW, and RMR, supporting a protective role in maintaining lean mass and metabolic function. Although TAC did not correlate directly with ASP, it was positively linked to glucose levels in later OGTT stages, possibly reflecting enhanced glucose clearance during early metabolic disturbance. In obesity, persistent inflammation and OS may impair these defenses, driving further metabolic decline49Multiple studies have shown significant associations between ASP and IR markers in both diabetic and non-diabetic populations50,51,52. Our results reveal a heterogeneous relationship between ASP and IR indices, particularly the TyG index, across BMI groups, consistent with evidence highlighting TyG’s superiority over HOMA-IR in assessing IR7,13,53,54. In normal-BMI individuals, ASP correlated positively with HOMA-IR and negatively with QUICKI and Matsuda, supporting its link to IR, as also proposed by Rohoma et al.55. The positive association with TyG underscores ASP’s involvement in hepatic IR56,57,58. Similarly, Zhong et al. observed elevated ASP and correlations with TyG, though only in type 2 diabetes patients59.Our results extend this association to non-diabetic subjects. Nevertheless, in this study, ASP showed a negative correlation with fasting glucose as well as at 60 and 240 min during the OGTT in the O1 group. It also exhibited a positive correlation with fasting insulin and at 120 and 180 min in the O1 group, indicating its involvement in hyperinsulinemia associated with IR29,60,61. In overweight individuals, ASP was positively associated with hyperinsulinemia, supporting its role as both a marker and mediator of IR, with potential for antibody-based therapies62. In the O1 group, correlations between ASP and HOMA-IR, QUICKI, and Matsuda indices were attenuated, suggesting early alterations in metabolic regulation. By contrast, the stronger association with the TyG index indicated a persistent link to lipid metabolism and hepatic function, consistent with reports that advancing obesity shifts dysfunction from insulin sensitivity toward inflammation and lipid dysregulation63. In obese individuals, ASP correlated with fasting and 60-minute insulin, supporting a role in insulin secretion through increased glucose demand and feedback regulation64,65. Similar observations were made by Camilleri et al., who demonstrated in animal models that ASP initially enhanced insulin action and reduced resistance, but later acted as a pro-inflammatory mediator in adipose tissue, thereby exacerbating IR37,66. In the O2 group, the correlation between ASP and HOMA-IR weakens significantly, suggesting that the body may be developing resistance not only to insulin but also to the effects of ASP. This is consistent with the findings of Al-Sulaiti et al., who suggested that higher levels of ASP may be less effective in promoting insulin sensitivity in obesity due to metabolic adaptations and the presence of chronic low-grade inflammation67. However, the most notable finding in this group is the strongest correlation between ASP and TyG. Notably, in our cohort the correlation of ASP with the composite TyG-WHR index was stronger than with TyG alone, underscoring that ASP may preferentially mirror hepatic IR driven by central adiposity. In addition, we observed significant associations between ASP and liver transaminases as well as CRP. These findings suggest that circulating ASP not only reflects impaired lipid–glucose handling but also integrates signals of hepatic injury and systemic inflammation. Clinically, this combined profile is highly relevant, since patients with obesity are at increased risk of non-alcoholic fatty liver disease and cardiometabolic complications, where hepatic IR, low-grade inflammation, and elevated liver enzymes often coexist. Thus, ASP’s parallel correlations with TyG-WHR, CRP, and transaminases strengthen its potential utility as a biomarker for liver-related metabolic dysfunction and may support its role in early risk stratification and therapeutic targeting. This suggests that, despite a weaker overall relationship with IR markers, ASP remains significantly involved in liver IR and dyslipidemia in individuals with obesity. This observation highlights the crucial role of ASP in metabolic dysfunction, particularly in obesity-related complications such as hypertriglyceridemia and fatty liver disease [68]. Our results suggest that TyG-WHR may serve as a more reliable indicator of liver-related IR in obese individuals, particularly when assessing ASP’s contribution to metabolic dysregulation. The stronger correlation between ASP and TyG and TyG-WHR in the O2 group suggests that, while ASP levels may be less effective at regulating insulin sensitivity, they still have a significant relationship with lipid metabolism and hepatic-related IR. This aligns with studies by Son et al., which found that TyG may better reflect IR in individuals with obesity than traditional markers like HOMA-IR13. Similarly, Kim et al. support the idea that the TyG index is particularly useful for assessing ASP’s impact on liver metabolism in obese populations, as it provides a comprehensive view of both glucose and lipid metabolism53. Clinically, the decreasing correlation between ASP and IR markers with increasing BMI suggests a diminished role of ASP in glucose and lipid regulation in advanced obesity. This may reflect metabolic adaptation and chronic low-grade inflammation. As such, ASP alone may have limited utility as a marker of IR in severe obesity. However, its consistent association with TyG highlights its potential relevance in targeting hepatic IR and dyslipidemia. To our knowledge, this is the first study to demonstrate that asprosin is preferentially linked with hepatic IR markers, such as TyG-WHR and liver transaminases, rather than solely with peripheral insulin sensitivity, highlighting a novel mechanistic aspect of its role in obesity.

Conclusions

ASP is associated with key features of metabolic dysfunction, including increased adiposity, reduced muscle mass, and impaired insulin sensitivity. While correlated with traditional IR indices (HOMA-IR, QUICKI, Matsuda) and with the proposed composite marker TyG-WHR, these associations weakened in obesity, suggesting possible ASP resistance. In contrast, its strong correlation with the TyG and TyG-WHR index across BMI groups points to a role in hepatic IR. Associations with OS markers further support a link between metabolic inflammation and endocrine function. ASP may thus serve as a relevant marker of liver-specific metabolic impairment in obesity, warranting further investigation.

Materials and methods

Study design

This study recruited a total of 150 participants, divided into three groups: 50 individuals with obesity (BMI > 30 kg/m²) (O2), 50 overweight individuals (BMI > 25 kg/m²) (O1), and 50 healthy volunteers with a BMI below 25 kg/m² (CG). Anthropometric measurements, including height and weight, were carefully taken using standardized instruments to ensure precision and accuracy. Body mass index (BMI) was calculated as the ratio of body weight (in kilograms) to the square of height (in meters squared). To further assess metabolic function, all participants underwent an OGTT, with glucose and insulin levels measured at specified intervals: 0, 60, 120, 180, and 240 min. The samples and controls were analyzed using the blind analysis method in a single run to eliminate bias. Glycated hemoglobin (HbA1c) was measured using the Bio-Rad D10 dual HbA2/F/A1c platform, employing the CE-HPLC method for high accuracy.

Body composition was assessed using dual-energy X-ray absorptiometry (DXA) (Lunar iDXA (GE Healthcare, Chicago, USA). Participants were non-smokers, abstinent from alcohol abuse, and free from medications (e.g., hypoglycemics, immunosuppressants) or conditions affecting oxidative stress. Eligibility was confirmed through medical history, physical examination, and documentation review. Venous blood (5.5 mL) was collected, centrifuged, and serum was aliquoted and stored at − 80 °C. The procedures were approved by the Local Ethics Committee of the Medical University of Bialystok, Poland, and written informed consent was obtained from each participant (APK.002.364.2021). All methods were carried out in accordance with relevant guidelines and regulations.

Chemical identity

Oxidative status was evaluated by measuring total oxidative capacity (TOC) and total antioxidant capacity (TAC) using a photometric assay (PerOx TOS/TAC kit; KC5100, Germany). ASP levels were determined by ELISA (Cloud-Clone Corp., SEA332Hu). Glucose, total cholesterol (CHOL), high density cholesterol (HDL), low density cholesterol (LDL), and triglycerides (TG) were measured via enzymatic colorimetric methods (Roche C111, Switzerland). Insulin and C-peptide were assessed using electrochemiluminescence (ECLIA; Roche E411, UK), while glycated hemoglobin (HbA1c) was analyzed via CE-HPLC (Bio-Rad D10 platform). All samples and controls were processed blindly in a single run to reduce bias.

IR indices were calculated as follows:

HOMA-IR = fasting glucose (mmol/l)×fasting insulin (µU/mL)​/22.5.

TyG index = ln (fasting triglycerides [mg/dl]×fasting glucose [mg/dl])/2.

QUICKI Index = 1/(log(fasting insulin (µU/mL)) + log(fasting glucose (mg/dL)).

Matsuda Index = 10,000/(fasting glucose(mg/dL)×fasting insulin(µU/mL))×(glucose (mg/dL) at 120 min×insulin (µU/mL) at 120 min)​

TyG-WHR index = ln (fasting triglycerides [mg/dl]×fasting glucose [mg/dl])/2 x WHR (cm)

Bioelectrical impedance analysis

The Bioelectrical Impedance Analysis (BIA) method was employed to assess body composition using the medical body analyzer INBODY 220 (Biospace, Korea). This device enables the measurement of body mass, total body water (TBW), fat mass, skeletal muscle mass, BMI and RMR.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9.0. As data were not normally distributed (Shapiro–Wilk test), nonparametric tests were used. The Mann–Whitney (**) and Kruskal–Wallis (*) tests assessed inter-group differences (p < 0.05). Spearman correlation was applied to evaluate relationships between variables.