Interleukin-37 and interleukin-39 as novel immunometabolic biomarkers in metabolic syndrome: A cross-sectional study

Abstract

Metabolic syndrome (MetS) is characterized by abdominal obesity, dyslipidemia, hypertension, and insulin resistance, all driven by chronic low-grade inflammation. This study aimed to evaluate serum interleukin-37 (IL-37) and interleukin-39 (IL-39) levels as potential immunometabolic biomarkers in MetS. Eighty adults (40 MetS patients, 40 healthy controls) were enrolled based on NCEP ATP III criteria. Anthropometric, biochemical, and hormonal parameters were assessed. Serum IL-37 and IL-39 concentrations were measured using ELISA. Statistical analyses included t-tests, Pearson correlations, logistic regression, and receiver operating characteristic curve analysis. Both IL-37 and IL-39 levels were significantly elevated in MetS compared with controls (P <.001). Logistic regression revealed that each 1 pg/mL increase in IL-37 and IL-39 was associated with 1.01-fold (P = .017) and 1.06-fold (P = .001) higher odds of MetS, respectively. receiver operating characteristic analysis showed excellent discriminative ability for IL-39 (AUC = 0.96; 95% confidence intervals: 0.91–1.00) and good accuracy for IL-37 (AUC = 0.82; 95% confidence intervals: 0.73–0.91). Among classical parameters, waist circumference (AUC = 1.00), TG (AUC = 0.93), and fasting glucose (AUC = 0.88) showed the strongest diagnostic power. HDL cholesterol exhibited inverse association (AUC = 0.83), while systolic and diastolic blood pressure showed moderate accuracy (AUC = 0.85 and 0.71, respectively). IL-37 and IL-39 were elevated in individuals with MetS and showed strong discriminative performance, particularly IL-39. These cytokines may reflect underlying immunometabolic alterations and could provide complementary information alongside established metabolic parameters. However, given the cross-sectional design and modest sample size, these findings should be considered exploratory and require validation in larger, longitudinal studies to determine their clinical relevance.

1. Introduction

Metabolic syndrome (MetS) is a multifactorial clinical condition characterized by the coexistence of abdominal obesity, dyslipidemia, hypertension, and insulin resistance, which collectively increase the risk of type 2 diabetes and cardiovascular diseases.^[1,2] MetS carries significant clinical weight, as reflected in its global distribution. Data indicate that it affects between 12.5% and 31.4% of adults across different populations,^[3] with an estimated 25% based on the IDF definition.^[4] A meta-analysis published in 2018 in Turkey reported that, according to the International Diabetes Federation (IDF) criteria, the prevalence of MetS was approximately 43.3% (95% CI: 41.9–44.7), with rates of 50.4% in women and 35.4% in men.^[5]

Several diagnostic definitions have been proposed for MetS, with the most widely used being the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) and the IDF criteria. The NCEP ATP III definition requires the presence of any 3 out of 5 components: abdominal obesity, elevated triglycerides (TG), reduced high-density lipoprotein cholesterol (HDL-C), hypertension, and impaired fasting glucose (FG).^[6] The IDF criteria, on the other hand, require central obesity as an essential component, defined by waist circumference (WC) cutoffs that differ by sex and ethnicity, together with at least 2 additional metabolic abnormalities. These distinctions explain why prevalence rates of MetS differ considerably between populations.^[7]

Beyond epidemiological definitions, MetS is increasingly recognized as a state of chronic low-grade inflammation driven by adipose tissue dysfunction. Adipocyte hypertrophy and hypoxia promote macrophage infiltration and the release of pro-inflammatory mediators such as tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6).^[8,9] This process is accompanied by adipokine imbalance, particularly disruption of the leptin–adiponectin axis, which contributes to insulin resistance, metabolic dysregulation, and endothelial dysfunction.^[10,11] Together, these alterations underscore the role of chronic inflammation in metabolic dysregulation and cardiometabolic risk in MetS.^[12]

In the context of MetS, biomarkers have been extensively investigated to better understand its pathophysiology and to explore potential diagnostic or prognostic indicators. Classical inflammatory mediators such as IL-6, TNF-α, and C-reactive protein (CRP) have been widely studied, underscoring the role of inflammation in MetS. More recently, attention has shifted toward emerging cytokines beyond these traditional markers. In particular, the anti-inflammatory properties of interleukin-37 (IL-37) and the immunomodulatory features of interleukin-39 (IL-39) have prompted interest in their potential relevance to MetS.^[13,14]

Interleukin-37, a member of the interleukin-1 (IL-1) cytokine family, is recognized for its potent anti-inflammatory properties. It suppresses innate immune responses primarily by inhibiting signaling pathways such as nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) and mitogen-activated protein kinase, thereby reducing the production of pro-inflammatory cytokines.^[13] Recent evidence suggests that IL-37 may also be involved in modulating low-grade inflammation associated with obesity, highlighting its potential relevance in immunometabolic regulation.^[15]

Interleukin-39, on the other hand, is a more recently identified cytokine belonging to the interleukin-12 (IL-12) family, composed of the interleukin-23, p19 subunit (IL-23p19) and Epstein-Barr virus–induced gene 3 (Ebi3) subunits.^[16,17] IL-39 has been reported in experimental models to activate the signal transducer and activator of transcription 1/3 (STAT1/3) pathways, thereby promoting immune cell proliferation and modulating inflammatory responses.^[18] However, evidence regarding its expression and functional relevance in humans remains limited and inconclusive.^[19] A limited number of clinical studies have reported altered IL-39 levels in conditions such as autoimmune diseases, cardiovascular disorders, and type 2 diabetes. However, its biological role and relevance in humans remain incompletely understood.^[14,20] Therefore, further clinical investigations are needed to clarify whether IL-39 represents a meaningful immunometabolic signal in MetS.

In this context, our study evaluated IL-37 and IL-39 levels in patients with MetS, and it was considered that these cytokines may represent potential biomarker candidates reflecting the immunometabolic aspect of the syndrome. To the best of our knowledge, this is among the first attempts to investigate both IL-37 and IL-39 simultaneously in MetS, thereby addressing a gap in the current literature.

2. Material and methods

2.1. Study cohort and design

The study was performed between August 2025 and October 2025 in the Endocrinology Unit of the Faculty of Medicine, Malatya Turgut Özal University, Turkey. Individuals aged 18 to 65 years were screened, and forty patients fulfilling the diagnostic criteria for MetS according to the NCEP ATP III were included.^[21] Diagnosis required the presence of at least 3 of the following: central obesity (waist >102 cm in men or >88 cm in women), fasting TG ≥150 mg/dL, low HDL-C (<40 mg/dL in men or < 50 mg/dL in women), elevated blood pressure (≥130/85 mm Hg or current antihypertensive therapy), or FG ≥110 mg/dL.

Subjects were excluded if they had type 1 diabetes, malignant disease, chronic gastrointestinal problems, pregnancy or lactation, history of antibiotic therapy in the last month, recent weight loss or dietary treatment, use of lipid- or glucose-lowering medication other than metformin, or intake of dietary supplements with potential metabolic effects. A control group of forty healthy individuals of similar ethnic origin and demographic characteristics, without systemic illness and meeting the same exclusion rules, was also recruited. All volunteers gave written informed consent. Ethical approval was obtained from the Malatya Turgut Özal University Health Sciences Scientific Research Ethics Committee (Approval No: 2025/270, July 30, 2025), and the study conformed to the principles of the Declaration of Helsinki. The flowchart of participant recruitment for the study was shown in Figure 1.

Figure 1.

Flowchart of participant recruitment.

2.2. Demographic and clinical evaluations

Information on age, sex, smoking and alcohol consumption, medical history, and dietary practices was collected. Anthropometric data were obtained as follows: body weight was measured using a digital bioimpedance device (Tanita BC-420MA, Tokyo, Japan), and height was measured with a stadiometer. Body mass index (BMI) was calculated by dividing weight (kg) by height squared (m²). WC was measured midway between the lowest rib and the iliac crest using a rigid tape measure. Measurements were taken while participants wore light clothing and no shoes. Information on dietary habits was obtained through a brief self-reported medical history (anamnesis) focusing on general eating patterns, recent dietary changes, and adherence to any specific diet. This information was collected for descriptive purposes only and was not included as a covariate in statistical analyses.

Blood pressure was measured twice on the left arm with an automated oscillometric device (Omron M3, Omron Corporation, Tokyo, Japan) after the subject had been seated for at least 10 minutes. The mean of the 2 values was used as the systolic (SBP) and diastolic (DBP) blood pressure. To reduce variability, all assessments were performed by the same trained investigator.

2.3. Sample processing and biochemical analyses

Fasting venous blood samples were collected in the morning by an experienced phlebotomist. Each participant provided 2 serum tubes and 1 K₂EDTA tube. Glycated hemoglobin (HbA1c) was determined from EDTA blood with a dedicated analyzer.

Serum samples were allowed to clot for 20 to 30 minutes and centrifuged at 1800 g for 10 minutes. One aliquot was used for biochemical analyses, including FG, lipid profile, renal and liver function markers, and CRP, using an automated chemistry analyzer. Hormonal parameters (fasting insulin, TSH, T3, and T4) were measured by electrochemiluminescence immunoassay.

Insulin resistance was estimated using the Homeostasis Model Assessment (HOMA-IR) formula:

$${\displaystyle \begin{array}{l}{\displaystyle HOMA-IR=\frac{FI\left(\mu U/mL\right)\times FG\left(mg/dL\right)}{405}}\end{array}}$$

The second serum tube was divided into 1.5 mL microcentrifuge tube and stored at −80 °C until the measurement of IL-37 and IL-39.

On the day of analysis, IL-37 (ELK Biotechnology, Cat. No: ELK1343) and IL-39 (ELK Biotechnology, Cat. No: ELK9171) levels were determined using the Enzyme-Linked ImmunoSorbent Assay (ELISA) kit according to the manufacturer’s recommendations. The absorbance of the samples was determined with a microplate reader adjusted to 450 nm wavelength. The measurement ranges of IL-37 and IL-39 were both 7.82 to 500 pg/mL. The intra-assay coefficients of variation were <8%, and the inter-assay coefficients of variation were <10%, as reported by the manufacturer for both ELISA kits.

2.4. Statistical analysis

Statistical analyses were performed using the SPSS software (Statistical Package for the Social Sciences, version 25, Chicago). The normality of data distribution was assessed with the Kolmogorov–Smirnov test. Descriptive data were expressed as mean ± standard deviation for continuous variables and as numbers and percentages for categorical variables. A significance level of P <.05 was considered statistically significant. Considering the large number of statistical tests performed on biochemical and clinical parameters, the false discovery rate was controlled using the Benjamini–Hochberg procedure. This correction was applied to enhance the reliability of the findings and to reduce the risk of potential false-positive results.

Since the data followed a normal distribution (P >.05), parametric test methods were applied. Comparisons between 2 independent groups were conducted using the independent samples t-test. Relationships between continuous variables were examined using Pearson correlation coefficient.

Receiver operating characteristic (ROC) curve analysis was conducted to assess the ability of serum IL-37 and IL-39 concentrations to discriminate between individuals with MetS and healthy controls. The area under the curve (AUC) was calculated to quantify diagnostic accuracy, and optimal cutoff values for IL-37 and IL-39 were identified using the Youden index formula (J = sensitivity + specificity – 1).

A binary logistic regression model was established to evaluate the association between group status (MetS vs control) as the dependent variable and IL-37 and IL-39 levels as independent predictors. The goodness of fit of the model was tested using the Hosmer–Lemeshow statistic, and the model was found to adequately fit the data (χ² = 5.529, df = 8, P = .700). Regression coefficients (β), standard errors, Wald statistics, odds ratios [Exp(β)], and 95% confidence intervals (CI) were reported.

The required sample size was estimated with G*Power software (version 3.1). According to the power analysis, a total of 80 participants (40 per group) provided a statistical power of 0.95 at an alpha level of 0.05, assuming a large effect size (Cohen d = 0.75). This sample size ensured adequate precision to detect significant differences between groups with a 95% confidence level.^[22]

3. Results

3.1. Demographic and clinical characteristics of the study population

As shown in Table 1, there were no significant differences between the MetS and control groups regarding sex distribution, smoking habits, or alcohol consumption (P >.05). However, marked differences were observed in several metabolic parameters. Participants with MetS showed a substantially higher frequency of elevated FG, TG levels, blood pressure, WC, and BMI compared to healthy controls (P <.001). In contrast, low HDL-C levels were significantly more prevalent in the MetS group (87.5%) than in the control group (35%) (P <.001) (Table 1).

Table 1.

Comparison of demographic and clinical characteristics between MetS patients and healthy controls.

Variable	Group	MetS (n/ %)	Control (n/ %)		Total (n/ %)	χ2	P
Sex	Male	17 (42.50)	20 (50.00)		37 (46.25)	0.453	.501
	Female	23 (57.50)	20 (50.00)		43 (53.75)
Smoking status	Former	4 (10.00)	0 (0.00)		4 (5.00)	5.710	.058
	Current	8 (20.00)	14 (35.00)		22 (27.50)
	Never	28 (70.00)	26 (65.00)		54 (67.50)
Alcohol consumption	Yes	3 (7.50)	0 (0.00)		3 (3.75)	4.053	.132
	Occasional	0 (0.00)	1 (2.50)		1 (1.25)
	No	37 (92.50)	39 (97.50)		76 (95.00)
FG category	Low	9 (22.50)	39 (97.50)		48 (60.00)	46.875	<.001
	High	31 (77.50)	1 (2.50)		32 (40.00)
TG category	Low	6 (15.00)	40 (100.00)		46 (57.50)	59.130	<.001
	High	34 (85.00)	0 (0.00)		34 (42.50)
HDL-C category	Low	35 (87.50)	14 (35.00)		49 (61.25)	23.226	<.001
	High	5 (12.50)	26 (65.00)		31 (38.75)
BP category	Low	13 (32.50)	33 (82.50)		46 (57.50)	23.059	<.001
	Normal	0 (0.00)	1 (2.50)		1 (1.25)
	High	27 (67.50)	6 (15.00)		33 (41.25)
WC category	Low	0 (0.00)	39 (97.50)		39 (48.75)	76.098	<.001
	High	40 (100.00)	1 (2.50)		41 (51.25)
Variable		MetS (Mean ± SD)		Control (Mean ± SD)		t	P
BMI		33.16 ± 4.44		23.09 ± 2.21		12.826	<.001

P-value, statistical significance; P <.05, there is a statistical difference between the groups (bold values).

% = percentage, BMI = body mass index, BP = blood pressure, FG = fasting glucose, HDL-C = high-density lipoprotein cholesterol, MetS = metabolic syndrome, n = frequency, SD = standard deviation, t = 2 independent sample t-test, TG = triglyceride, WC = waist circumference.

Overall, these results indicate that individuals in the MetS group exhibited the typical metabolic disturbances that define MetS, confirming the validity of the group classification in this study.

3.2. Analysis of biochemical and hormonal markers

According to Table 2, individuals with MetS exhibited significantly higher serum HbA1c, fasting insulin, HOMA-IR albumin, ALT and CRP levels and lower T4 concentrations compared to healthy controls (P <.05). The remaining biochemical and hormonal parameters, including urea, creatinine, uric acid, AST, TSH, and T3, did not show statistically significant differences between groups (P >.05) (Table 2).

Table 2.

Comparison of biochemical and hormonal parameters between MetS patients and healthy controls.

Variable	Groups	Mean ± SD	t	P	FDR-adjusted p	Cohen d
HbA1c (%)	MetS	7.02 ± 2.3	4.670	<.001	<.001	1.04
	Control	5.31 ± 0.28
FI (μU/mL)	MetS	15.85 ± 6.34	6.405	<.001	<.001	1.43
	Control	8.62 ± 3.27
HOMA-IR	MetS	5.26 ± 3.27	6.345	<.001	<.001	1.42
	Control	1.89 ± 0.77
Urea (mg/dL)	MetS	25.71 ± 7.30	0.959	.341	.045	0.21
	Control	24.31 ± 5.65
Creatinine (mg/dL)	MetS	0.74 ± 0.17	−1.819	.073	.037	−0.39
	Control	0.80 ± 0.14
Uric acid (mg/dL)	MetS	4.94 ± 1.39	0.586	.559	.05	0.13
	Control	4.77 ± 1.15
Albumin (g/dL)	MetS	4.49 ± 0.35	1.991	.050	.032x	0.45
	Control	4.34 ± 0.31
ALT (U/L)	MetS	31.95 ± 22.62	2.924	.050	.034x	0.65
	Control	20.65 ± 9.25
AST (U/L)	MetS	26.15 ± 12.97	1.206	.231	.039	0.27
	Control	23.30 ± 7.41
CRP (mg/dL)	MetS	0.46 ± 0.28	4.134	<.001	<.001	0.95
	Control	0.18 ± 0.31
TSH (mU/L)	MetS	1.75 ± 1.05	−0.610	.544	.047	−0.14
	Control	1.92 ± 1.38
Free T3 (pg/mL)	MetS	3.16 ± 0.61	1.187	.239	.042	0.26
	Control	2.99 ± 0.69
Free T4 (ng/mL)	MetS	1.35 ± 0.27	−2.128	.037	.029x	−0.47
	Control	1.61 ± 0.74

P-value, statistical significance; P <.05, there is a statistical difference between the groups (bold values); FDR-adjusted P-value, statistical significance there is a statistical difference between the groups Benjamini–Hochberg false discovery rate (FDR); x, critical value >P-value (not statistically significant).

ALT = alanine aminotransferase, AST = aspartate aminotransferase, CRP = C-reactive protein, FI = fasting insulin, HbA1c = glycated hemoglobin, HOMA-IR = homeostatic model assessment of insulin resistance, MetS = metabolic syndrome, SD = standard deviation, T3 = triiodothyronine, T4 = thyroxine, TSH = thyroid-stimulating hormone.

3.3. Evaluation of MetS criteria

As summarized in Table 3, subjects with MetS exhibited significantly higher FG, TG levels, SBP and DBP, and WC values compared with controls (P <.001). Conversely, HDL-C levels were substantially lower in the MetS group (P <.001) (Table 3).

Table 3.

Comparison of MetS criteria between MetS patients and healthy controls.

Variable	Groups	Mean ± SD	t	P	FDR-adjusted p	Cohen d
FG (mg/dL)	MetS	130.35 ± 59.26	4.467	<.001	<.001	1.00
	Control	88.05 ± 8.64
TG (mg/dL)	MetS	209.72 ± 90.85	7.571	<.001	<.001	1.69
	Control	95.30 ± 29.73
HDL-C (mg/dL)	MetS	39.70 ± 5.69	−5.816	<.001	<.001	1.30
	Control	53.71 ± 14.13
SBP (mm Hg)	MetS	127.75 ± 11.43	6.593	<.001	<.001	1.47
	Control	113.75 ± 7.05
DBP (mm Hg)	MetS	83.75 ± 6.86	3.914	<.001	<.001	0.88
	Control	78.12 ± 5.96
WC (cm)	MetS	111.33 ± 8.28	14.132	<.001	<.001	3.16
	Control	75.05 ± 13.97

P-value, statistical significance; P <.05, there is a statistical difference between the groups (bold values); FDR-adjusted P value, statistical significance there is a statistical difference between the groups Benjamini–Hochberg false discovery rate (FDR); x, critical value >P-value (not statistically significant).

DBP = diastolic blood pressure, FG = fasting glucose, HDL-C = high-density lipoprotein cholesterol, MetS = metabolic syndrome, SBP = systolic blood pressure, SD = standard deviation, TG = triglycerides, WC = waist circumference.

These data confirm the expected metabolic and anthropometric abnormalities characteristic of MetS, including hyperglycemia, dyslipidemia, hypertension, and central obesity.

3.4. Assessment of serum IL-37 and IL-39 levels

As shown in Figure 2, both IL-37 and IL-39 levels were significantly elevated in the MetS group compared with healthy controls (P <.001). The mean serum IL-37 concentration in MetS patients was 405.38 ± 119.68 pg/mL, while that of controls was 264.58 ± 92.68 pg/mL. Similarly, IL-39 levels were almost doubled in the MetS group (195.77 ± 55.54 pg/mL) compared to controls (96.99 ± 30.95 pg/mL) (Fig. 2A and B).

Figure 2.

Comparative analysis of serum IL-37 and IL-39 concentrations among the study groups. The box plots represent mean values with standard deviations. Each dot indicates an individual data point. Serum IL-37 (A) and IL-39 (B) levels were significantly altered among groups, reflecting their potential involvement in the pathophysiology of MetS. Statistical analysis was performed using an independent samples t-test. (P <.05 was considered statistically significant).

3.5. Binary logistic regression analysis

As shown in Table 4, both IL-37 and IL-39 were significant predictors of MetS. The logistic regression model demonstrated good overall fit (Hosmer–Lemeshow χ² = 5.529, df = 8, P = .700).

Table 4.

Binary logistic regression analysis of IL-37 and IL-39 levels for predicting MetS.

Variable	β	S.E	W	df	P-value	Exp (β) (OR)	95% CI for Exp (β)
							Lower bound	Upper bound
IL-37	0.011	0.005	5.704	1	.017	1.011	1.002	1.021
IL-39	0.057	0.014	15.770	1	.001	1.058	1.029	1.089
Constant	−11.567	2.722	18.054	1	.001	0.001	–

P <.05, there is a statistical difference (bold values).

95% CI = confidence interval, df = degrees of freedom, Exp (β) = odds ratio, IL-37 = interleukin-37, IL-39 = interleukin-39, S.E. = standard error, W = Wald statistic, β = parameter estimate.

Higher serum IL-37 and IL-39 levels were associated with increased odds of having MetS. Specifically, each 1 pg/mL increase in IL-37 and IL-39 concentrations was associated with approximately 1.01-fold (P = .017) and 1.06-fold (P = .001) higher likelihood of MetS, respectively (Table 4). The Box–Tidwell procedure was applied, and it was confirmed that IL-37 and IL-39 met the linearity assumption (P >.05). The overall explanatory power of the model (Nagelkerke R² = 0.81) and its high predictive accuracy (91.3%) support the biological relevance of the reported odds ratios.

3.6. ROC curve analysis

ROC curve analysis was carried out to evaluate and compare the discriminative ability of serum IL-37 and IL-39 concentrations with conventional metabolic parameters -FG, TG, HDL-C, SBP, DBP, and WC- in differentiating individuals with MetS from healthy controls (Table 5 and Fig. 3). All AUC values were statistically significant (P <.05) unless otherwise stated.

Table 5.

ROC analysis results.

Variable	MetS–Control		FG		TG		HDL-C		BP		WC
	AUC (95% CI)	cutoff (Sens–Spec)	AUC (95% CI)	cutoff (Sens–Spec)	AUC (95% CI)	cutoff (Sens–Spec)	AUC (95% CI)	cutoff (Sens–Spec)	AUC (95% CI)	cutoff (Sens–Spec)	AUC (95% CI)	cutoff (Sens–Spec)
FG	0.88 (0.8–0.96)	88.5 (0.88–0.55)	0.99 (0.97–1)	96.5 (0.81–0.8)	0.85 (0.75–0.94)	98.5 (0.79–0.76)	0.71 (0.6–0.83)	98.5 (0.59–0.71)	0.7 (0.58–0.82)	92.5 (0.82–0.57)	0.87 (0.78–0.95)	96.5 (0.8–0.79)
TG	0.93 (0.88–0.99)	97 (0.93–0.55)	0.86 (0.78–0.95)	95.5 (0.93–0.56)	1 (1–1)	94.5 (1–0.5)	0.81 (0.71–0.91)	119.5 (0.73–0.74)	0.71 (0.59–0.83)	138 (0.73–0.66)	0.93 (0.88–0.99)	115 (0.88–0.77)
HDL-C	0.83 (0.74–0.93)	36.5 (0.83–0.9)	0.76 (0.65–0.88)	43.6 (0.76–0.77)	0.83 (0.73–0.93)	46.7 (0.74–0.78)	0.73 (0.62–0.84)	41.5 (0.71–0.68)	0.68 (0.55–0.81)	43 (0.76–0.66)	0.83 (0.73–0.92)	42.7 (0.8–0.77)
SBP	0.85 (0.77–0.93)	115 (0.85–0.53)	0.71 (0.59–0.82)	115 (0.95–0.54)	0.79 (0.69–0.88)	115 (0.94–0.46)	0.66 (0.54–0.78)	115 (0.78–0.39)	0.84 (0.75–0.93)	115 (0.94–0.45)	0.85 (0.77–0.93)	115 (0.95–0.54)
DBP	0.71 (0.6–0.82)	75 (0.71–0.2)	0.66 (0.54–0.79)	75 (0.93–0.21)	0.64 (0.52–0.77)	82.5 (0.41–0.8)	0.58 (0.45–0.71)	75 (0.88–0.16)	0.81 (0.7–0.92)	82.5 (0.67–0.98)	0.7 (0.59–0.82)	75 (0.93–0.21)
WC	1 (0.99–1)	93 (1–0.95)	0.88 (0.8–0.97)	94.5 (0.93–0.95)	0.95 (0.89–1)	97 (0.94–0.87)	0.77 (0.67–0.88)	86.5 (0.73–0.61)	0.8 (0.7–0.9)	94.5 (0.82–0.72)	0.99 (0.99–1)	90.5 (0.98–0.92)
IL-37	0.82 (0.73–0.91)	257 (0.82–0.55)	0.76 (0.65–0.86)	250.6 (0.9–0.54)	0.75 (0.64–0.86)	359.84 (0.62–0.8)	0.62 (0.49–0.74)	250.65 (0.78–0.45)	0.73 (0.62–0.84)	313.64 (0.79–0.68)	0.83 (0.74–0.92)	313.64 (0.76–0.74)
IL-39	0.96 (0.91–1)	102 (0.96–0.5)	0.88 (0.79–0.96)	106.9 (0.95–0.64)	0.89 (0.81–0.97)	129.77 (0.91–0.78)	0.72 (0.6–0.84)	106.92 (0.78–0.52)	0.77 (0.66–0.87)	126.17 (0.82–0.66)	0.94 (0.88–1)	128.18 (0.93–0.9)

AUC = area under the curve, CI = confidence interval, DBP = diastolic blood pressure, FG = fasting glucose, HDL-CBP = blood pressure, IL-37 = interleukin-37, IL-39 = interleukin-39, MetS = metabolic syndrome, ROC = receiver operating characteristic, SBP = systolic blood pressure, sens = sentivity, spec = specifity, TG = triglyceride, WC = waist circumference.

Figure 3.

ROC analyses. ROC curves illustrating the diagnostic performance of serum IL-37 and IL-39 levels, along with classical metabolic parameters, across 6 models: (A) patient-control differentiation, (B) presence or absence of abdominal obesity, (C) low-high blood pressure, (D) low-high FG, (E) low-high HDL-C, and (F) low-high TG. Abdominal obesity: WC ≥102 cm in men and ≥88 cm in women; high TG ≥150 mg/dL; high FG ≥100 mg/dL; low HDL: HDL <40 mg/dL in men and <50 mg/dL in women; high blood pressure: SBP ≥130 mm Hg or DBP ≥85 mm Hg. DBP = diastolic blood pressure, FG = fasting glucose, HDL-C = high-density lipoprotein cholesterol, IL-37 = interleukin-37, IL-39 = interleukin-39, SBP = systolic blood pressure, TG = triglyceride, WC = waist circumference.

Among the cytokines, IL-39 exhibited the highest diagnostic accuracy, with an AUC of 0.96 (95% CI: 0.91–1.00) for the MetS–control comparison, indicating excellent discriminative performance. The optimal cutoff value of 102 pg/mL provided a sensitivity of 96% and a specificity of 50%. IL-37 also demonstrated a strong diagnostic capability, with an AUC of 0.82 (95% CI: 0.73–0.91) and a cutoff value of 257 pg/mL (sensitivity 82%, specificity 55%). These findings support the role of IL-37 and IL-39 as inflammation-related biomarkers with potential utility in identifying MetS.

Among the classical metabolic parameters, FG, TG, and WC exhibited the highest discriminative power, with AUC values ranging from 0.88 to 1.00. WC demonstrated the strongest diagnostic accuracy, achieving an AUC of 1.00 (95% CI: 0.99–1.00), thereby confirming its significance as a key clinical indicator for MetS diagnosis. Similarly, TG (AUC = 0.93; 95% CI: 0.88–0.99) and FG (AUC = 0.88; 95% CI: 0.80–0.96) also showed strong diagnostic performance.

HDL cholesterol demonstrated an inverse diagnostic performance, showing a moderate discriminative ability with an AUC of 0.83 (95% CI: 0.74–0.93). The optimal cutoff value was 36.5 mg/dL, providing 83% sensitivity and 90% specificity for distinguishing MetS patients from healthy controls. These findings confirm that reduced HDL-C levels are strongly associated with the presence of MetS, reflecting its dyslipidemic component and impaired lipid metabolism.

Systolic blood pressure exhibited moderate diagnostic accuracy, with an AUC of 0.85 (95% CI: 0.77–0.93) at a cutoff of 115 mm Hg (sensitivity 85%, specificity 53%). This result indicates that elevated SBP is a frequent clinical feature of MetS and contributes significantly to its hemodynamic profile. DBP presented lower discriminative capacity compared to SBP, with an AUC of 0.71 (95% CI: 0.60–0.82) and a cutoff value of 75 mm Hg (sensitivity 71%, specificity 20%). Although DBP alone showed limited diagnostic precision, in combination with SBP it supports the overall cardiovascular assessment in patients with MetS.

4. Discussion

In this study, we have demonstrated that serum levels of IL-37 and IL-39 are significantly elevated in individuals with MetS compared to healthy controls. IL-39 showed high discriminative performance within this cohort (AUC = 0.96, 95% CI: 0.91–1.00), while IL-37 also demonstrated moderate accuracy (AUC = 0.82, 95% CI: 0.73–0.91). These findings support the possibility that IL-37 and IL-39 reflect immunometabolic alterations associated with MetS. The odds ratios per 1 pg/mL increase appear numerically small. However, given the wide concentration ranges observed for IL-37 and IL-39, cumulative changes across clinically relevant intervals may correspond to more substantial risk differences. In addition, the assumption of linearity in the log-odds model should be interpreted cautiously, and future studies may explore nonlinear or threshold-based associations.

Interleukin-37 is recognized as a potent anti-inflammatory cytokine that suppresses innate and adaptive immune responses through mechanisms including inhibition of NF-κB, mitogen-activated protein kinase, and mTOR signaling, and activation of Smad3 pathways.^[23] Its elevated levels in MetS patients might reflect a compensatory anti-inflammatory response to chronic metabolic stress and low-grade inflammation. Supporting this interpretation, a cross-sectional study conducted in Saudi adults also reported significantly higher circulating IL-37 levels in individuals with MetS. The finding that women exhibited higher IL-37 concentrations than men in the same study suggests that sex-related biological differences may modulate IL-37 responses, indicating that the behavior of this cytokine in metabolic disorders cannot be reduced to a single phenotype.^[24] Taken together, these observations strengthen the notion that the increase observed in our study may represent part of a protective mechanism aimed at limiting metabolic inflammation.

In experimental models of MetS, recombinant IL-37 administration improved insulin sensitivity and reduced adipose tissue inflammation.^[25] However, dose-dependent effects have also been reported, 1 recent study suggests that very high doses of IL-37 may paradoxically promote inflammatory activity in adipose tissue in diet-induced obesity.^[26] This underscores the importance of considering context, dose, and tissue microenvironment when interpreting IL-37 behavior in MetS. Some clinical studies have reported divergent patterns. For instance, a cross-sectional study analyzing cytokine profiles in MetS patients found lower IL-37 expression in certain contexts, possibly reflecting differences in patient populations or methodology.^[27] These discrepancies suggest that IL-37’s role may differ across metabolic phenotypes and underscores the need for larger, multi-center cohorts.

While IL-39 remains relatively understudied, emerging data support its relevance in metabolic and inflammatory disorders. In T2DM cohorts, IL-39 levels have been shown to correlate positively with BMI and to offer strong diagnostic discrimination (AUC ≈ 0.97).^[14] Our finding that IL-39 achieves AUC = 0.96 in distinguishing MetS aligns closely with that data, supporting a possible progression along the metabolic spectrum (from insulin resistance to overt syndrome). Moreover, IL-39 upregulation has been documented in inflammatory conditions such as inflammatory bowel disease and rheumatoid arthritis, reinforcing its role as an inflammation-responsive cytokine whose expression is modulated in different disease contexts.^[28,29] In periodontal disease coexisting with diabetes, IL-39 levels decreased following periodontal treatment, suggesting tissue-specific regulation.^[30] Taken together, these findings support the possibility that IL-39 responds to metabolic and inflammatory stress across organs and may represent a potential immunometabolic signal. However, its role in humans remains to be clarified in larger studies.

Our ROC analyses again confirm that classical measures remain powerful diagnostic predictors. WC, TG, and FG yielded AUCs between 0.88 and 1.00, consistent with their foundational role in MetS criteria.^[31] In fact, WC reached a perfect discrimination (AUC = 1.00, 95% CI: 0.99–1.00), reaffirming its central value as a clinical diagnostic marker. However, this perfect discrimination should be interpreted with caution, as WC is itself a defining component of MetS criteria. Therefore, its performance in this dataset may partly reflect criterion overlap rather than independent predictive capacity. Regarding IL-39, although it demonstrated excellent AUC performance, its relatively modest specificity (~50 %) may limit its utility as a standalone diagnostic marker. This finding suggests that IL-39 may be more appropriately considered within a multi-marker framework rather than as an isolated predictor. Larger validation cohorts are warranted to further clarify its clinical applicability. HDL-C displayed an inverse relationship with MetS, with AUC values ranging from 0.74 to 0.93, consistent with its decline in MetS subjects and reinforcing its protective metabolic role. SBP and DBP achieved modest discrimination (AUC ranges of ~0.77–0.93 for SBP and ~0.60–0.82 for DBP), suggesting that while hypertension is a component of MetS, its standalone discriminative power is lower than metabolic components. By juxtaposing cytokine markers and standard metabolic indices, we suggest that IL-37 and IL-39 may add orthogonal information reflecting immune activation rather than purely metabolic derangement to improve risk stratification and early detection.

Our data support the emerging concept of “immunometabolic remodeling” in which chronic metabolic stress promotes immune adaptation or maladaptation, ultimately influencing metabolic flux.^[32,33] Within this framework, elevated IL-39 levels may reflect a response to adipose tissue–related metabolic stress and inflammatory activation. However, this interpretation remains hypothesis-generating and requires further mechanistic validation. Conversely, IL-37 might represent a counter-regulatory cytokine that limits excessive inflammation and helps restore homeostasis. This reciprocal pattern aligns with current evidence on the crosstalk between innate immune receptors, metabolic sensors, and cytokine networks in obesity and insulin resistance.^[15,34] Moreover, previous findings in nonmetabolic contexts, such as acute respiratory distress syndrome, indicate that lower circulating IL-37 levels are associated with increased mortality, suggesting a potential systemic anti-inflammatory role.^[35] Although not directly related to MetS, these observations collectively suggest that IL-37 may serve as a general indicator of systemic inflammatory burden relevant to chronic metabolic and inflammatory diseases.

A key strength of the present study is the combined evaluation of emerging immunometabolic biomarkers (IL-37 and IL-39) alongside classical metabolic indices in a well-characterized cohort. Nevertheless, several limitations should be acknowledged. The cross-sectional design precludes causal inference, and validation in larger and multiethnic populations is required to improve generalizability. In addition, the absence of mechanistic and tissue-level analyses limits interpretation of the underlying biological pathways. Future studies incorporating adjustment for potential confounders such as BMI, CRP, medication use, and indices of insulin resistance will be important to determine whether these cytokines contribute independently to the MetS phenotype.

Moving forward, prospective cohorts should investigate whether elevated IL-37 or IL-39 predict incident MetS or its complications (cardiovascular, hepatic, renal). Interventional trials, possibly leveraging recombinant IL-37 or IL-39 modulation, may unravel causal pathways. Multi-omic analyses (transcriptome, epigenome) may clarify the downstream networks regulated by these cytokines. Finally, standardization of assay methods and normative reference ranges across populations is crucial for real-world clinical translation.

In summary, our findings indicate that circulating IL-37 and IL-39 levels are elevated in individuals with MetS and are associated with key metabolic disturbances. Rather than replacing established diagnostic parameters, these cytokines may reflect the inflammatory component of metabolic dysregulation and could offer complementary insight into immunometabolic interactions. Given the cross-sectional design and relatively modest sample size, the present results should be considered exploratory. Future large-scale, longitudinal, and mechanistic studies are required to determine whether IL-37 and IL-39 have independent clinical utility or predictive value in MetS.

Acknowledgments

The authors thank the patients and their families for participating in this study.

Author contributions

Conceptualization: Lezan Keskin, Tuğba Raika Kiran, Mehmet Erdem.

Data curation: Lezan Keskin, Feyza İnceoğlu.

Formal analysis: Lezan Keskin, Tuğba Raika Kiran, Mehmet Erdem.

Methodology: Lezan Keskin, Tuğba Raika Kiran.

Resources: Lezan Keskin, Tuğba Raika Kiran.

Software: Lezan Keskin, Mehmet Erdem, Feyza İnceoğlu.

Supervision: Lezan Keskin.

Writing – original draft: Lezan Keskin, Tuğba Raika Kiran, Mehmet Erdem.

Writing – review & editing: Lezan Keskin, Tuğba Raika Kiran, Mehmet Erdem, Feyza İnceoğlu.