Insulin resistance as potential mediator linking ApoB/ApoA1 to MAFLD,but not inflammation

Abstract

Aims:

Metabolic-associated fatty liver disease (MAFLD) is closely associated with insulin resistance (IR) and systemic inflammation. Apolipoprotein A1 (ApoA1) and Apolipoprotein B (ApoB), as notable non-traditional lipid markers, have demonstrated distinct advantages in identifying risks related to metabolic syndrome and coronary atherosclerosis, yet its association with MAFLD and the mediating roles of IR/inflammation remain unclear.

Methods:

This retrospective investigation involved 1061 participants, categorized into a non-MAFLD group (n = 529) and an MAFLD group (n = 532). Univariate and multivariate logistic regression models, Spearman’s correlation, and mediation analysis were utilized to explore the intricate associations between ApoB/ApoA1, MAFLD, inflammation, and IR.

Results:

The MAFLD group exhibited markedly elevated levels of neutrophils/lymphocytes, neutrophils/platelets, systemic immune inflammation index, systemic inflammation response index, pan-immune-inflammation value and triglyceride-glucose index (TyG), TyG body mass index (TyGBMI), and metabolic score for insulin resistance (METS-IR) compared to the non-MAFLD group. Logistic regression analysis revealed that ApoB/ApoA1, TyG, TyGBMI, and METS-IR were markedly linked to MAFLD risk. Spearman’s correlation analysis identified substantial positive links between ApoB/ApoA1 and TyG (r = 0.45), METS-IR (r = 0.47), and TyGBMI (r = 0.42). Mediation analysis, adjusted for confounding variables, revealed that TyG, TyGBMI, and METS-IR mediated 70.4%, 100%, and 100% of the association between ApoB/ApoA1 and MAFLD, respectively.

Conclusion:

Our findings clarify the complex interrelationships between ApoB/ApoA1, MAFLD risk, inflammation, and IR, and for the first time, demonstrate that IR may act as a key potential mediator in the link between ApoB/ApoA1 and MAFLD, rather than systemic inflammation. This suggests that IR may serve a more prominent role than chronic systemic inflammation in the association between lipid metabolism and MAFLD risk, and intervening in IR may be more effective than anti-inflammatory therapy in blocking the progression from lipid metabolism disorders to MAFLD.

Keywords

apolipoprotein inflammation insulin resistance metabolic-associated fatty liver disease

Introduction

Apolipoprotein B (ApoB) and Apolipoprotein A1 (ApoA1) are key lipid metabolism markers, with the ApoB/ApoA1 ratio reflecting the balance between atherogenic and anti-atherogenic lipoproteins. This ratio remains unaffected by statin therapy, providing a more accurate assessment of lipid abnormalities’ impact on cardiovascular health compared to conventional lipid markers. Several studies^1–4 have highlighted the distinct advantages of ApoB/ApoA1 in identifying risks associated with coronary atherosclerotic disease, myocardial infarction, and ischemic stroke while also serving as an indicator for cardiovascular and overall mortality. In addition, a prominent study has demonstrated that ApoB levels may function as an indicator of metabolic syndrome (MetS) risk among Chinese adults.⁵ Nevertheless, investigations examining the connection between ApoA1, ApoB, and metabolic-associated fatty liver disease (MAFLD) risk remain scarce.

MAFLD has become an increasingly prevalent and significant cause of chronic liver disease, characterized by a broad range of risk factors and complex pathophysiological mechanisms. It is well established that lipid metabolism disorders, insulin resistance (IR),⁶ and chronic inflammation⁷ serve essential functions in the pathogenesis of MAFLD. In 2020, an international panel renamed non-alcoholic fatty liver disease (NAFLD)/steatohepatitis (NASH) as MAFLD, revised the diagnostic criteria, and further emphasized its relationship with overweight, IR, type 2 diabetes mellitus (T2DM), lipid dysregulation, and metabolic inflammation.^8–10

IR, as the cornerstone of MAFLD pathogenesis,¹¹ is characterized by reduced insulin-mediated glucose uptake and utilization.¹² Peripheral IR impairs the antilipolytic effect of insulin in adipose tissue, leading to excessive hepatic fatty acid accumulation, hepatocellular dysfunction, inflammation, and oxidative stress, thus promoting MAFLD progression.¹³ Beyond IR, systemic inflammation exacerbates steatosis by inducing oxidative stress, promoting hepatocyte apoptosis, tissue damage, mitochondrial dysfunction, and free fatty acid (FFA) accumulation.^14,15 Specifically, novel inflammatory markers such as neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and systemic immune inflammatory index (SII) can quantify the inflammatory immune balance of the body,¹⁶ while elevated levels of cytokines such as interleukin-8 (IL-8) and tumor necrosis factor alpha (TNF-α) are significantly correlated with the severity of steatohepatitis.¹⁷ IR surrogate indices, including triglyceride-glucose index (TyG), TyG body mass index (TyGBMI),¹⁸ and metabolic score for insulin resistance (METS-IR)¹⁹ quantify IR through glycometabolic parameters, indirectly reflecting MAFLD progression.

MAFLD is a complex disorder involving lipid dysregulation, obesity, IR, chronic inflammation, and genetic factors. Despite numerous studies and proposed mechanisms, how these factors interact in MAFLD pathogenesis remains unclear. Notably, whether IR and inflammation mediate the link between apolipoproteins and MAFLD risk is unproven. Thus, this study aimed to further investigate the intricate associations among ApoB/ApoA1, MAFLD risk, inflammation, and IR, and to examine the contributions of IR and inflammation to the link between ApoB/ApoA1 and MAFLD risk, to provide new therapeutic insights into the interplay of lipid metabolism, IR, inflammation, and MAFLD.

Methods

Subject inclusion

This retrospective analysis encompassed individuals who underwent medical evaluations at the Health Examination Center of Fuyang People’s Hospital between January 2023 and July 2024. A total of 1061 participants were included and categorized into two groups: a non-MAFLD group (n = 529) and an MAFLD group (n = 532), based on the diagnostic criteria for MAFLD.

Inclusion criteria

Participants were eligible for inclusion if they: (1) Underwent a comprehensive health examination at the Health Examination Center of Fuyang People’s Hospital between January 2023 and July 2024. (2) Had complete medical records, including demographic data (age, gender, height, weight), medical history (T2DM, hypertension), laboratory test results (blood routine, liver and kidney function, glucose and lipid metabolism indicators, etc.), and liver-biliary-pancreatic-splenic ultrasound findings. (3) Met the diagnostic criteria for MAFLD.

Exclusion criteria

Participants were excluded if they: (1) Had incomplete medical records (e.g., missing key laboratory parameters, ultrasound data, or demographic information). (2) Had a history of other chronic liver diseases, including viral hepatitis (hepatitis B or C), autoimmune liver disease, cirrhosis, or liver cancer. (3) Had excessive alcohol consumption or a history of drug-induced fatty liver (e.g., long-term use of glucocorticoids, methotrexate). (4) Had severe systemic diseases, such as end-stage renal disease, malignant tumors, or acute infections, which could affect lipid metabolism, glucose homeostasis, or inflammatory markers. (5) Were pregnant or lactating women.

The diagnostic criteria for MAFLD in the Chinese population are based on the English version of the Chinese MAFLD Prevention and Treatment Guidelines.²⁰ MAFLD is characterized by hepatic steatosis accompanied by metabolic dysfunction. The diagnosis of MAFLD requires the exclusion of excessive alcohol intake (defined as weekly ethanol consumption ⩾210 g for males and ⩾140 g for females) and other conditions that may lead to fatty liver, along with the presence of at least one component of MetS. The components of MetS are as follows: (1) Overweight/obesity: body mass index (BMI) ⩾24.0 kg/m², or waist circumference ⩾90 cm (for males) and ⩾85 cm (for females), or increased body fat percentage. (2) Hypertension/elevated blood pressure: Blood pressure ⩾130/85 mmHg or the use of antihypertensive medication. (3) Prediabetes or T2DM: Fasting blood glucose (FBG) ⩾6.1 mmol/L, or 2-h postprandial blood glucose ⩾7.8 mmol/L, or glycosylated hemoglobin, Type A1C ⩾5.7% (⩾39 mmol/mol), or a history of T2DM, or homeostasis model assessment of IR index (HOMA-IR) ⩾2.5. (4) Elevated triglycerides (TG): Fasting serum TG ⩾1.70 mmol/L or use of lipid-lowering medication. (5) Decreased high-density lipoprotein cholesterol (HDL-C): Serum HDL-C ⩽1.0 mmol/L (for males) and ⩽1.3 mmol/L (for females) or use of lipid-lowering medication.

Data collection

Two physicians extracted data from the hospital’s electronic database, including: demographics and medical history: age, gender, height, weight, T2DM, hypertension. Laboratory parameters: white blood cells, neutrophils, lymphocytes, monocytes, platelets, homocysteine, ApoA1, ApoB, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma-glutamyl transferase (γ-GT), FBG, uric acid (UA), total cholesterol (TC), TG, HDL-C, low-density lipoprotein cholesterol (LDL-C), and liver-biliary-pancreatic-splenic ultrasound results. The inflammatory immune biomarkers included the neutrophils/lymphocytes (NLR), platelets/lymphocytes (PLR), neutrophils/platelets (NPR), systemic immune inflammation index (SII), systemic inflammation response index (SIRI), and pan-immune-inflammation value (PIV).^16,21,22 Surrogate indices for IR included the TyG, TyGBMI, and METS-IR.^23,24 The formulas were as follows: SII = Platelets * Neutrophils/Lymphocytes, SIRI = Neutrophils * Monocytes/Lymphocytes, PIV = Neutrophils * Monocytes * Platelets/Lymphocytes, TyG = ln [TG * (FBG/2)], TyGBMI = TyG * BMI, and METS-IR = ln (2 * FBG + TG) * BMI/ln (HDL-C), BMI = Weight (kg)/Height (m)². For unit conversion, FBG, TG, and HDL-C were measured in mg/dl.

Statistical analysis

Statistical analysis was executed utilizing SPSS 26.0 software (IBM Corporation, IBM SPSS Statistics 26.0). The independent samples t test or the Mann–Whitney U test was utilized to compare groups. Continuous variables were tested for normality using a combination of visual inspection (histograms, Q–Q plots) and formal statistical tests (Kolmogorov-Smirnov test for n ⩾ 50). Non-parametric tests (Mann–Whitney U test for two-group comparisons) were used when variables violated the normality assumption. Categorical variables were compared using the Chi-square test or Fisher’s exact test, as appropriate. Univariate and multivariate logistic regression analysis were executed to ascertain independent risk factors linked to the development of MAFLD. Spearman’s correlation analysis was employed to investigate the relationships between ApoB/ApoA1 ratios and inflammatory markers, as well as IR surrogate indicators. Finally, mediation models were constructed to assess whether inflammation and IR parameters mediated the relationship between ApoB/ApoA1 and MAFLD risk, with the mediation effects quantified by calculating the mediation percentage, defined as the ratio of the indirect effect to the total effect. Mediation analysis was performed using the “bruceR” package in R software, with the Bootstrap method applied. (The development and maintenance of the ‘bruceR’ package mainly rely on academic research teams and are open source community projects, without any commercial companies as’ manufacturers’). The confidence interval (CI) of indirect effects was estimated via 5000 repeated samplings. Statistical significance was considered at p < 0.05.

Results

Characteristics of participants

This retrospective analysis encompassed 1061 subjects, comprising 529 cases in the non-MAFLD group and 532 in the MAFLD group. As depicted in Table 1, the MAFLD group exhibited markedly higher proportions of males, T2DM, and hypertension, as well as increased values for age, BMI, systolic and diastolic blood pressure (SBP and DBP), when compared to the non-MAFLD group. Furthermore, the MAFLD group demonstrated markedly elevated levels of white blood cells, neutrophils, lymphocytes, monocytes, platelets, homocysteine, ApoB, ApoB/ApoA1, ALT, AST, ALP, γ-GT, FBG, UA, TC, TG, LDL-C, NLR, NPR, SII, SIRI, PIV, TyG, TyGBMI, and METS-IR, with all differences reaching statistical significance. In contrast, levels of ApoA1, HDL-C, and PLR were markedly lower in the MAFLD group (p < 0.001, p < 0.001, p < 0.01, respectively).

Table 1.

Baseline characteristics and hematological parameters of participants.

Characteristics	Total (n = 1061)	Non-MAFLD (n = 529)	MAFLD (n = 532)	p
Gender, n(%)				<0.001
Male	596(56.2)	239(45.2)	357(67.1)
Female	465(43.8)	290(54.8)	175(32.9)
T2DM, n(%)	290(27.3)	78(14.7)	212(39.8)	<0.001
Hypertension, n(%)	312(29.4)	96(18.1)	216(40.6)	<0.001
Age (years)	52(41, 59)	48(38, 58)	54(46, 60)	<0.001
BMI (kg/m²)	25.0(22.6, 27.5)	22.8(21.2, 24.5)	27.2(25.3, 29.3)	<0.001
SBP (mmHg)	131(119, 145)	122(111, 132)	141(130, 153.3)	<0.001
DBP (mmHg)	81(75, 89)	77(71, 83)	87(80, 92)	<0.001
Leukocyte (×10⁹/L)	6.0(5.0, 7.0)	5.6(4.7, 6.5)	6.4(5.4, 7.3)	<0.001
Neutrophil (×10⁹/L)	3.4(2.7, 4.1)	3.1(2.5, 3.8)	3.6(2.9, 4.4)	<0.001
Lymphocyte (×10⁹/L)	2.0(1.6, 2.3)	1.9(1.6, 2.3)	2.1(1.7, 2.5)	<0.001
Monocyte (×10⁹/L)	0.4(0.3, 0.4)	0.3(0.3, 0.4)	0.4(0.3, 0.5)	<0.001
Platelet (×10⁹/L)	232(196, 271)	229(193, 268)	235(199, 276)	0.071
Homocysteine (μmol/L)	11.2(9.3, 13.3)	9.3(8.0, 10.9)	13.0(11.5, 16.6)	<0.001
ApoA1 (g/L)	1.5(1.3, 1.6)	1.5(1.4, 1.6)	1.4(1.3, 1.6)	<0.001
ApoB (g/L)	1.0(0.8, 1.1)	0.9(0.8, 1.0)	1.0(0.9, 1.2)	<0.001
ApoB/ApoA1	0.7(0.6, 0.8)	0.6(0.5, 0.7)	0.7(0.6, 0.8)	<0.001
ALT (U/L)	19.2(14.3, 28.8)	16.6(12.3, 22.3)	22.8(16.6, 34.6)	<0.001
AST (U/L)	20.6(17.3, 25)	19.4(16.8, 22.9)	21.8(18, 27)	<0.001
ALP (U/L)	70.8(58.8, 84.4)	65.3(54.8, 81.5)	75.9(64.0, 88.9)	<0.001
γ-GT (U/L)	21.8(15, 36)	16.6(12.3, 25)	28.6(20.3, 49.2)	<0.001
FBG (mmol/L)	5.4(5.0, 6.0)	5.2(4.9, 5.5)	5.6(5.2, 6.4)	<0.001
UA (μmol/L)	315.9(255.9, 385.4)	285.8(235.6, 343.5)	346.7(286.9, 416.7)	<0.001
TC (mmol/L)	4.9(4.3, 5.6)	4.8(4.2, 5.3)	5.0(4.4, 5.7)	<0.001
TG (mmol/L)	1.3(0.9, 2.0)	1.0(0.8, 1.4)	1.7(1.2, 2.5)	<0.001
HDL-C (mmol/L)	1.3(1.1, 1.5)	1.4(1.2, 1.6)	1.1(1.0, 1.3)	<0.001
LDL-C (mmol/L)	2.7(2.2, 3.2)	2.6(2.2, 3.1)	2.8(2.2, 3.3)	0.002
NLR	1.7(1.3, 2.2)	1.6(1.3, 2.1)	1.8(1.4, 2.2)	0.001
PLR	116.7(94.8, 144.9)	121.1(98.0, 148.3)	113.9(92.5, 140.4)	0.009
NPR	0.01(0.01, 0.02)	0.01(0.01, 0.02)	0.02(0.01, 0.02)	<0.001
SII	387.7(288.7, 535.7)	370.2(280, 492.4)	402.3(305.6, 557.3)	<0.001
SIRI	0.6(0.4, 0.8)	0.5(0.4, 0.7)	0.7(0.5, 0.9)	<0.001
PIV	134.6(91.3, 203.8)	121.9(83.0, 175.9)	154.5(102.3, 237.5)	<0.001
TyG	8.0(7.5, 8.4)	7.6(7.3, 8.0)	8.3(7.9, 8.7)	<0.001
TyGBMI	201.5(174.8, 227.8)	174.9(159.5, 191.2)	224.8(208.7, 246.1)	<0.001
METS-IR	29.3(25.0, 33.4)	25.1(22.9, 27.9)	33.1(30.3, 36.4)	<0.001

Continuous variables are presented as median (IQR); categorical variables are presented as n (%). Reference interval: Leukocyte: 3.5–9.5 × 10⁹/L; Neutrophil: 1.8–6.3 × 10⁹/L; Lymphocyte: 1.1–3.2 × 10⁹/L; Monocyte: 0.1–0.6 × 10⁹/L; Platelet: 125–350 × 10⁹/L; Homocysteine: 5–15 μmol/L; ApoA1: 1.0–1.6 g/L; ApoB: 0.8–1.1 g/L; ALT: 9–50 U/L; AST: 15–40 U/L; ALP: 45–125 U/L; γ-GT: 10–60 U/L; FBG: 3.9–6.1 mmol/L; UA: 208–428 μmol/L; TC: <5.18 mmol/L; TG: <1.7 mmol/L; HDL-C: 1.16–1.42 mmol/L; LDL-C: <3.37 mmol/L. (The reference intervals are based on the clinical laboratory standards of Fuyang People’s Hospital, which conform to the general reference ranges widely used in clinical practice across China.)

ALP, alkaline phosphatase; ALT, alanine aminotransferase; ApoA1, apolipoprotein A1; ApoB, apolipoprotein B; AST, aspartate aminotransferase; BMI, body mass index; DBP, diastolic blood pressure; FBG, fasting blood glucose; HDL-C, high-density lipoprotein cholesterol; IQR, interquartile range; LDL-C, low-density lipoprotein cholesterol; MAFLD, metabolic-associated fatty liver disease; METS-IR, metabolic score for insulin resistance; NLR, neutrophils/lymphocytes; NPR, neutrophils/platelets; PIV, pan-immune-inflammation value; PLR, platelets/lymphocytes; SBP, systolic blood pressure; SII, systemic immune inflammation index; SIRI, systemic inflammation response index; T2DM, type 2 diabetes mellitus; TC, total cholesterol; TG, triglycerides; TyG, triglyceride-glucose index; TyGBMI, TyG body mass index; UA, uric acid; γ-GT, gamma-glutamyl transferase.

Independent risk factors linked to MAFLD occurrence

Through univariate (Supplemental Table 1) and multivariate (Table 2) logistic regression analysis, after adjusting for potential confounding factors, as shown in Table 2, ApoB/ApoA1, TyG, TyGBMI, and METS-IR were markedly associated with MAFLD risk, all functioning as independent risk factors. Specifically, A unit increase in ApoB/ApoA1 was linked to a 9.73-fold heightened probability of MAFLD (odds ratio (OR) 10.73, 95% CI 3.5, 33.69). An elevate in TyG by 1 unit was connected with a 4.48-fold rise in the risk of MAFLD (OR 5.48, 95% CI 3.65, 8.44). For each unit elevated in TyGBMI, the risk of MAFLD increased by 0.07-fold (OR 1.07, 95% CI 1.06, 1.09). A unit elevate in METS-IR was connected with a 0.54-fold rise in the risk of MAFLD (OR 1.54, 95% CI 1.43, 1.66). Separate models for each indicator were constructed, controlling for gender, age, homocysteine levels, SBP and DBP, ALT, AST, alkaline phosphatase, γ-GT, and UA levels.

Table 2.

Multivariate logistic regression analysis of MAFLD risk factors.

Variable	b	SE	OR (95% CI)	p	Wald
ApoB/ApoA1	2.373	0.577	10.73(3.5, 33.69)	<0.001	16.932
NLR	0.08	0.116	1.08(0.86, 1.36)	0.492	0.471
PLR	0	0.002	1(1.0, 1.01)	0.946	0.005
NPR	20.119	16.223	5.46 × 1010(0.5.98 × 1022)	0.215	1.538
SII	0.001	0	1.001(1, 1.002)	0.032	4.607
SIRI	0.287	0.233	1.33(0.85, 2.12)	0.218	1.514
PIV	0.002	0.001	1.002(1, 1.004)	0.020	5.406
TyG	1.702	0.213	5.48(3.65, 8.44)	<0.001	63.552
TyGBMI	0.072	0.006	1.07(1.06, 1.09)	<0.001	126.344
METS-IR	0.429	0.038	1.54(1.43, 1.66)	<0.001	125.077

b, regression coefficient, representing the change in the log-odds of MAFLD occurrence per unit increase in the independent variable; SE, indicating the variability of the regression coefficient estimate; a smaller value reflects a more precise estimate; OR, quantifying the strength of association between the independent variable and MAFLD risk. An OR >1 suggests the variable is a risk factor, while OR <1 indicates a protective factor; 95% CI, representing the range within which the true OR is likely to fall. If the interval does not include 1, the association is statistically significant; Wald, Chi square value, used to test the statistical significance of the regression coefficient; a larger value indicates stronger evidence for a significant association between the independent variable and MAFLD.

95% CI, 95% confidence interval; ApoB/ApoA1, apolipoprotein B/apolipoprotein A1; MAFLD, metabolic-associated fatty liver disease; METS-IR, metabolic score for insulin resistance; NLR, neutrophils/lymphocytes; NPR, neutrophils/platelets; OR, odds ratio; PIV, pan-immune-inflammation value; PLR, platelets/lymphocytes; SE, standard error; SII, systemic immune inflammation index; SIRI, systemic inflammation response index; TyG, triglyceride-glucose index; TyGBMI, TyG body mass index.

ApoB/ApoA1 ratio is markedly linked to MAFLD risk

The previous logistic regression analysis has shown a significant association between the ApoB/ApoA1 ratio and MAFLD risk. To further clarify the dose-effect relationship of this association, we divided the ApoB/ApoA1 ratio into three groups by tertiles: low tertile (T1, 0.211–0.586), middle tertile (T2, 0.586–0.744), and high tertile (T3, 0.744–1.484). As illustrated in Table 3, the percentages of MAFLD patients in the T1, T2, and T3 groups were 21.6%, 31%, and 47.4%, respectively. In comparison to the T1 group, the ORs for the T2 and T3 groups were 1.72 (95% CI: 1.27, 2.33) and 5.26 (95% CI: 3.80, 7.26), respectively. These findings indicated that the ApoB/ApoA1 ratio is significantly positively correlated with MAFLD risk; that is, the risk of MAFLD increases markedly with the elevation of the ApoB/ApoA1 ratio.

Table 3.

ApoB/ApoA1 ratio is markedly linked to MAFLD risk.

ApoB/ApoA1 tertiles	Total, n(%)	Non-MAFLD, n(%)	MAFLD, n(%)	Statistic	p
ApoB/ApoA1 tertiles	Total, n(%)	Non-MAFLD, n(%)	MAFLD, n(%)	111.557	<0.001
T1 (0.211–0.586)	350(33)	235(44.4)	115(21.6)
T2 (0.586–0.744)	361(34)	196(37.1)	165(31)
T3 (0.744–1.484)	350(33)	98(18.5)	252(47.4)
ApoB/ApoA1 tertiles	b	SE	OR (95% CI)		p
	Reference
T1 (0.211–0.586)
T2 (0.586–0.744)	0.542	0.155	1.72(1.27, 2.33)		<0.001
T3 (0.744–1.484)	1.659	0.165	5.26(3.80, 7.26)		<0.001

95% CI, 95% confidence interval; ApoB/ApoA1, apolipoprotein B/apolipoprotein A1; b, regression coefficient; MAFLD, metabolic-associated fatty liver disease; OR, odds ratio; SE, standard error; T1, low tertile; T2, middle tertile; T3, high tertile.

Correlations between ApoB/ApoA1, systemic inflammatory immune indicators, and IR surrogate markers

Spearman correlation analysis was executed to investigate the links among the ApoB/ApoA1 ratio, systemic inflammation-immune markers, and IR surrogate indicators, aiming to further clarify their intricate interactions. As depicted in Figure 1, Spearman correlation analysis revealed strong positive correlations between ApoB/ApoA1 and IR indices: TyG (r = 0.45), METS-IR (r = 0.47), and TyGBMI (r = 0.42; all p < 0.001). Weak positive correlations were observed with SIRI (r = 0.10, p < 0.05), PIV (r = 0.17, p < 0.001), and SII (r = 0.13, p < 0.01). No significant correlations were found with NPR, NLR, or PLR (all p > 0.05). IR markers showed weak positive correlations with NPR, SIRI, PIV, and SII (r = 0.10–0.23, all p < 0.05), and weak negative correlations with PLR (r = −0.13 to −0.15, p < 0.001).

Figure 1.

Spearman related heatmap, blue represents positive correlation between indicators, and red represents negative correlation between indicators.

Mediation analysis of IR and inflammation in the ApoB/ApoA1-MAFLD risk association

Mediation models were used to evaluate the mediating roles of inflammatory markers (NLR, PLR, NPR, SII, SIRI, PIV) and IR indices (TyG, TyGBMI, METS-IR) in the ApoB/ApoA1-MAFLD risk association. Before adjusting for confounding factors (as presented in Supplemental Table 2), PIV mediated 4.5% of the association between ApoB/ApoA1 and MAFLD, while the IR indices TyG, TyGBMI, and METS-IR mediated 73.7%, 100%, and 100% of the association, respectively. After adjusting for confounding factors (as shown in Table 4), none of the inflammatory markers showed mediating effects. However, the three IR indices remained significant mediators: TyG mediated 70.4%, and TyGBMI as well as METS-IR exhibited complete mediation (100% and 100%, respectively; as illustrated in Figure 2). The mediation analysis results suggest IR serves a pivotal mediating function in the ApoB/ApoA1-MAFLD risk association, with TyGBMI and METS-IR exhibiting complete mediation effects (100%). In contrast, inflammation appears to lack a mediating role in this relationship.

Table 4.

Systemic inflammatory immune markers and IR alternative markers mediation models of the relationship between ApoB/ApoA1 and MAFLD (after controlling for confounding factors).

Mediators	Type	Effect	95% CI	SE	Z	p	Proportion of mediation
NLR	Indirect(ab)	−0.002	(−0.012, 0.003)	0.004	−0.581	0.561	—
	Direct(c’)	0.510	(0.372, 0.637)	0.070	7.320	0.000
	Total(c)	0.508	(0.367, 0.636)	0.070	7.299	0.000
PLR	Indirect(ab)	−0.003	(−0.015, 0.002)	0.004	−0.769	0.442	—
	Direct(c’)	0.513	(0.368, 0.630)	0.066	7.759	0.000
	Total(c)	0.510	(0.363, 0.627)	0.067	7.646	0.000
NPR	Indirect(ab)	−0.011	(−0.028, 0.000)	0.007	−1.476	0.140	—
	Direct(c’)	0.518	(0.378, 0.640)	0.067	7.698	0.000
	Total(c)	0.507	(0.367, 0.633)	0.068	7.405	0.000
SII	Indirect(ab)	0.011	(−0.000, 0.028)	0.007	1.500	0.134	—
	Direct(c’)	0.497	(0.356, 0.626)	0.068	7.272	0.000
	Total(c)	0.508	(0.366, 0.633)	0.068	7.454	0.000
SIRI	Indirect(ab)	0.002	(−0.010, 0.017)	0.007	0.299	0.765	—
	Direct(c’)	0.505	(0.363, 0.625)	0.069	7.284	0.000
	Total(c)	0.507	(0.365, 0.629)	0.070	7.283	0.000
PIV	Indirect(ab)	0.021	(0.004, 0.046)	0.011	1.908	0.056	—
	Direct(c’)	0.488	(0.347, 0.617)	0.069	7.027	0.000
	Total(c)	0.508	(0.372, 0.636)	0.068	7.482	0.000
TyG	Indirect(ab)	0.350	(0.280, 0.418)	0.035	9.956	0.000	70.4%
	Direct(c’)	0.147	(0.002, 0.278)	0.071	2.080	0.038
	Total(c)	0.496	(0.361, 0.607)	0.064	7.732	0.000
TyGBMI	Indirect(ab)	0.494	(0.401, 0.568)	0.043	11.555	0.000	100.0%
	Direct(c’)	0.022	(−0.082, 0.130)	0.055	0.410	0.682
	Total(c)	0.517	(0.371, 0.634)	0.066	7.796	0.000
METS-IR	Indirect(ab)	0.537	(0.441, 0.604)	0.043	12.578	0.000	100.0%
	Direct(c’)	−0.021	(−0.128, 0.088)	0.055	−0.382	0.703
	Total(c)	0.516	(0.359, 0.627)	0.068	7.644	0.000

Indirect(ab): mediation effect; Direct(c’): direct effect; Total(c): total effect; Effect: The estimated value of the effect variable, indicating the strength of the influence of the independent variable on the dependent or mediating variable; Z: test statistic; Proportion of mediation = indirect(ab)/total(c) * 100%. Confounding variables included age, gender, type 2 diabetes, systolic, and diastolic blood pressure.

95% CI: 95% confidence interval; METS-IR, metabolic score for insulin resistance; NLR, neutrophils/lymphocytes; NPR, neutrophils/platelets; PIV, pan-immune-inflammation value; PLR, platelets/lymphocytes; SE, standard error; SII, systemic immune inflammation index; SIRI, systemic inflammation response index; TyG, triglyceride-glucose index; TyGBMI, TyG body mass index.

Figure 2.

IR surrogate indicators (TyG, TyGBMI, and METS-IR) mediation models of the relationship between ApoB/ApoA1 and MAFLD.

Discussion

This represents the first clinical investigation to comprehensively describe the intricate relationships between apolipoproteins, six systemic inflammatory-immune markers, three IR surrogate indices, and the prevalence of MAFLD. The study revealed that inflammatory-immune indicators (NLR, NPR, SII, SIRI, PIV) and IR surrogate indices (TyG, TyGBMI, METS-IR) were markedly elevated in individuals with MAFLD. Further mediation analysis revealed that IR indices markedly mediated the relationship between ApoB/ApoA1 and MAFLD prevalence, with TyGBMI and METS-IR showing complete mediation effects, whereas none of the inflammatory-immune markers exhibited mediating effects. It provides the first evidence suggesting that IR may serve as a pivotal link in the association between ApoB/ApoA1 and MAFLD risk, while systemic inflammatory responses appear to play no mediating role. These findings offer valuable insights for clinicians to enhance their comprehension of IR and inflammatory pathways in MAFLD, and they suggest that IR seems to play a more important role than systemic inflammation in the association between lipid metabolism and MAFLD risk, revealing the unique pathological pathway through which lipid metabolism disorders drive MAFLD via IR.

In recent years, non-insulin-based methods for assessing IR, encompassing the TyG index, TyG BMI, and METS-IR, have garnered increased attention due to their advantages, including lower cost, ease of calculation, and independence from insulin measurement.^23–25 The core finding of this study is that three IR surrogate indicators, TyG, TyGBMI, and METS-IR, all act as key mediators in the ApoB/ApoA1-MAFLD association. From a metabolic mechanism perspective, an elevated ApoB/ApoA1 ratio reflects an imbalance between atherogenic lipoproteins (such as LDL-C) and anti-atherogenic lipoproteins (HDL-C). ApoB, as a core marker of total atherogenic lipid particles, encompasses triglyceride-rich lipoproteins (TRLs), including chylomicron remnants and very low-density lipoprotein (VLDL), which play a crucial role in metabolic abnormalities. TRLs are all ApoB-containing lipoproteins, and their elevated concentrations are closely associated with genetic defects. For instance, mutations in genes such as lipoprotein lipase (LPL), Apolipoprotein C3 (APOC3), or Angiopoietin-like 3 (ANGPTL3) can lead to impaired clearance of TRLs,²⁶ which is consistent with the mechanism underlying the elevated ApoB/ApoA1 ratio observed in this study. Meanwhile, increased TRL concentrations are often accompanied by decreases in HDL-C and ApoA1.²⁷ Severe hypertriglyceridemia is closely associated with MAFLD, and the accumulation of TRLs is a key driver of hepatic fat deposition. Furthermore, the impact of IR on TRL metabolism exacerbates this process. Specifically, IR can reduce TRL lipolysis by inhibiting LPL activity,²⁶ and impaired TRL clearance directly increases hepatic fat fraction.²⁷ This forms a vicious cycle of “IR-TRL accumulation-increased hepatic fat.” In addition, the imbalance between ApoB and ApoA1 may promote IR through several pathways: on one hand, excessive ApoB-containing lipoproteins (e.g., VLDL) transport FFAs to the liver, induce ectopic fat accumulation, and impair insulin signaling pathways via PKC-ε and JNK pathways.^6,13 On the other hand, the accumulation of intrahepatic TG (positively correlated with ApoB/ApoA1) can impair mitochondrial function through “lipotoxicity” and reduce insulin-mediated glucose uptake.¹¹ Secondly, a decrease in ApoA1, the main apolipoprotein of HDL, can weaken the insulin-sensitizing effect by reducing adiponectin secretion.^5,28 This imbalance may be related to the disruption of HDL-mediated lipid exchange due to impaired TRL lipolysis, and the ApoB/ApoA1 ratio directly reflects such metabolic imbalance. In this study, indicators such as TyG and TyGBMI showed a high proportion of mediating effects on the association between ApoB/ApoA1 and MAFLD, further demonstrating the core link of IR in the pathological chain of “lipid metabolism disorder-MAFLD.” It is worth noting that METS-IR and TyGBMI exhibit a complete mediating effect (100%), which may be attributed to their integration of multi-dimensional parameters such as BMI and glycolipid metabolism. Previous studies have shown that METS-IR outperforms traditional HOMA-IR in predicting the risk of MAFLD,²⁹ while TyGBMI, by combining BMI with the TyG index, more accurately reflects the degree of IR in obesity.³⁰ The results of our study further support the key value of these novel IR indicators in the pathological mechanism of MAFLD. These mechanisms collectively support the conclusion of this study that the ApoB/ApoA1 ratio contributes to the development of MAFLD by influencing TRL metabolism and exacerbating IR.

NLR, PLR, NPR, SII, SIRI, and PIV represent novel systemic inflammatory immune markers procured from peripheral blood cell counts, offering insights into the overall inflammatory-immune status of the body. These markers provide a comprehensive representation of the equilibrium between host immunity and inflammation. MetS, including MAFLD, is often considered a chronic inflammatory condition. Previous studies have established that systemic immune-inflammatory biomarkers (SII, NLR, PLR, and (lymphocytes/monocytes) LMR) are strongly linked to the risk of developing NAFLD.^16,31 Our study found that NLR, NPR, SII, SIRI, and PIV in patients with MAFLD were significantly increased, and logistic regression showed that SII and PIV were associated with the risk of MAFLD, which supported that MAFLD had the characteristics of a chronic inflammatory state. However, mediation analysis showed that all inflammatory markers did not mediate the ApoB/ApoA1-MAFLD association; we speculate that the underlying mechanisms may include the following. First, inflammation may be a secondary pathological response driven by IR. According to the “second hit” theory of MAFLD, IR induced by lipotoxicity is the initiating factor (the first hit), while inflammation is the subsequent damage result (the second hit).^7,15 ApoB/ApoA1 elevation promotes hepatic lipid deposition and induces IR. Under IR conditions, excessive lipolysis in adipose tissue releases FFAs, which activate liver macrophages (Kupffer cells) and leading to the secretion of pro-inflammatory cytokines such as TNF-α and IL-8.¹⁷ Although inflammatory response is involved in the progression of MAFLD, this study suggests that it is not a direct mediator of ApoB/ApoA1’s influence on MAFLD, but rather a downstream event triggered by IR. Secondly, the “non-specificity” of peripheral blood inflammatory biomarkers may limit the detection of their mediating effects. Hepatic inflammation in MAFLD is characterized by localized macrophage activation and pro-inflammatory cytokine secretion, whereas the peripheral blood markers assessed in this study (such as NLR, PLR) primarily reflect the systemic immune-inflammatory state and may not accurately represent the intrahepatic local inflammatory environment.^7,17,28 Further research may be needed in the future to investigate the association between the local immune-inflammatory status within the liver and the development of MAFLD.

The findings of this study are complementary to previous reports. On one hand, multiple studies have confirmed that IR is a central driving factor for MAFLD, while our study is the first to quantify the mediation contribution of IR in the “dyslipidemia-MAFLD” pathway through mediation analysis. On the other hand, although MAFLD has been regarded as a systemic chronic inflammatory state, our study is the first to clarify that it has no mediating contribution in the association of ApoB/ApoA1-MAFLD. IR is the main mediator of the effect of ApoB/ApoA1 on MAFLD, which may be due to its role in initiating hepatic lipid accumulation. Inflammation functions as a secondary amplifier that exacerbates liver damage, yet our findings reveal that it does not serve as a link between lipid metabolism disorders and the pathogenesis of MAFLD. The traditional view holds that IR and inflammation synergistically drive MAFLD. However, the complete mediating effect of IR observed in this study suggests that in the ApoB/ApoA1-MAFLD pathway, inflammation may be a downstream event of IR rather than a parallel mechanism. Thus, intervening in IR may be more effective than anti-inflammatory therapy in blocking the progression from lipid metabolism disorders to MAFLD. In summary, this study has established a hierarchical model for MAFLD risk. Lipid metabolism imbalance (ApoB/ApoA1) is the basis, IR is the core mediator, and inflammation may act as a secondary amplifier. These insights establish ApoB/ApoA1 as a potential screening tool and IR as a key therapeutic target, providing new evidence for understanding the pathological mechanisms and clinical practice of MAFLD. In clinical practice, interventions targeting IR, such as improving insulin sensitivity, may prove far more effective in halting the progression of MAFLD among individuals with a high ApoB/ApoA1 ratio.

Several limitations exist within this investigation. As a single-center retrospective investigation with a relatively small sample size, its findings may be prone to unmeasured biases and confounding variables, hindering the establishment of definitive causal relationships. In addition, this study did not conduct pre-sample size calculation and efficacy analysis. As a retrospective single-center study, the sample size was naturally determined by 1061 participants who met the inclusion criteria at the physical examination center of Fuyang People’s Hospital from January 2023 to July 2024, rather than calculated based on statistical efficacy formulas. This design may lead to insufficient inspection efficiency. Consequently, larger-scale multicenter prospective cohort studies, in conjunction with further basic research—including animal models and in vitro cellular experiments—are essential to unravel the underlying molecular signaling mechanisms of the disease.

Supplemental Material

sj-docx-1-tae-10.1177_20420188251378318 – Supplemental material for Insulin resistance as potential mediator linking ApoB/ApoA1 to MAFLD, but not inflammation

Supplemental material, sj-docx-1-tae-10.1177_20420188251378318 for Insulin resistance as potential mediator linking ApoB/ApoA1 to MAFLD, but not inflammation by Mengying Yang, Xiaoman Liu, Qianqian Li, Jun Liu and Baogui Wang in Therapeutic Advances in Endocrinology and Metabolism

Footnotes

Appendix

Acknowledgements

None.

Declarations

ORCID iD

Baogui Wang

Supplemental material

Supplemental material for this article is available online.

References

Faaborg-Andersen

Liu

Subramaniyam

, et al. U-shaped relationship between apolipoprotein A1 levels and mortality risk in men and women. Eur J Prev Cardiol 2023; 30(4): 293–304.

Nishiyama

Funamizu

Iwata

, et al. Low apolipoprotein A1 was associated with increased risk of cancer mortality in patients following percutaneous coronary intervention: a 10-year follow-up study. Int J Cancer 2022; 151(9): 1482–1490.

Johannesen

CDL

Mortensen

Langsted

, et al. ApoB and non-HDL cholesterol versus LDL cholesterol for ischemic stroke risk. Ann Neurol 2022; 92(3): 379–389.

Marston

Giugliano

Melloni

GEM

, et al. Association of apolipoprotein B-containing lipoproteins and risk of myocardial infarction in individuals with and without atherosclerosis: distinguishing between particle concentration, type, and content. JAMA Cardiol 2022; 7(3): 250–256.

Peng

, et al. Serum apolipoprotein B is associated with increased risk of metabolic syndrome among middle-aged and elderly Chinese: a cross-sectional and prospective cohort study. J Diabetes 2019; 11(9): 752–760.

Sakurai

Kubota

Yamauchi

, et al. Role of insulin resistance in MAFLD. Int J Mol Sci 2021; 22(8): 4156.

Luo

Chen

Fang

, et al. Study on inflammation and fibrogenesis in MAFLD from 2000 to 2022: a bibliometric analysis. Front Endocrinol (Lausanne) 2023; 14: 1231520.

Eslam

Sarin

Wong

, et al. The Asian Pacific Association for the study of the liver clinical practice guidelines for the diagnosis and management of metabolic associated fatty liver disease. Hepatol Int 2020; 14(6): 889–919.

Rinella

Lazarus

Ratziu

, et al. A multisociety Delphi consensus statement on new fatty liver disease nomenclature. Hepatology 2023; 78(6): 1966–1986.

10.

Loomba

Friedman

Shulman

GI.

Mechanisms and disease consequences of nonalcoholic fatty liver disease. Cell 2021; 184(10): 2537–2564.

11.

Nogueira

Cusi

Role of insulin resistance in the development of nonalcoholic fatty liver disease in people with type 2 diabetes: from bench to patient care. Diabetes Spectr 2024; 37(1): 20–28.

12.

Lee

S-H

Park

S-Y

Choi

CS.

Insulin resistance: from mechanisms to therapeutic strategies. Diabetes Metab J 2022; 46(1): 15–37.

13.

Ahmed

Mohamed

Farrag

MA.

Luteolin loaded on zinc oxide nanoparticles ameliorates non-alcoholic fatty liver disease associated with insulin resistance in diabetic rats via regulation of PI3K/AKT/FoxO1 pathway. Int J Immunopathol Pharmacol 2022; 36: 3946320221137435.

14.

Song

Zhang

Wang

, et al. A novel immune-related genes signature after bariatric surgery is histologically associated with non-alcoholic fatty liver disease. Adipocyte 2021; 10(1): 424–434.

15.

Farzanegi

Dana

Ebrahimpoor

, et al. Mechanisms of beneficial effects of exercise training on non-alcoholic fatty liver disease (NAFLD): roles of oxidative stress and inflammation. Eur J Sport Sci 2019; 19(7): 994–1003.

16.

Liu

Tang

Liu

, et al. Systemic immune-inflammatory biomarkers (SII, NLR, PLR and LMR) linked to non-alcoholic fatty liver disease risk. Front Immunol 2024; 15: 1337241.

17.

du Plessis

van Pelt

Korf

, et al. Association of adipose tissue inflammation with histologic severity of nonalcoholic fatty liver disease. Gastroenterology 2015; 149(3): 635–648.e14.

18.

Gao

Chen

Zhou

, et al. The association between different insulin resistance surrogates and all-cause mortality and cardiovascular mortality in patients with metabolic dysfunction-associated steatotic liver disease. Cardiovasc Diabetol 2025; 24(1): 200.

19.

Song

Lee

, et al. Identification of pediatric MASLD using insulin resistance indices. JHEP Rep 2025; 7(7): 101419.

20.

Fan

J-G

X-Y

Yang

R-X

, et al. Guideline for the prevention and treatment of metabolic dysfunction-associated fatty liver disease (Version 2024). J Clin Transl Hepatol 2024; 12(11): 955–974.

21.

Tang

Zhai

Song

, et al. Association between complete blood count-derived inflammatory markers and the risk of frailty and mortality in middle-aged and older adults. Front Public Health 2024; 12: 1427546.

22.

Dong

Jiang

, et al. Interpretable machine learning analysis of immunoinflammatory biomarkers for predicting CHD among NAFLD patients. Cardiovasc Diabetol 2025; 24(1): 263.

23.

Chen

Hou

, et al. Association between triglyceride-glucose related indices and mortality among individuals with non-alcoholic fatty liver disease or metabolic dysfunction-associated steatotic liver disease. Cardiovasc Diabetol 2024; 23(1): 232.

24.

Peng

Xiang

Pan

, et al. METS-IR/HOMA-IR and MAFLD in U.S. adults: dose-response correlation and the effect mediated by physical activity. BMC Endocr Disord 2024; 24(1): 132.

25.

Yan

Wang

, et al. Association between triglyceride glucose index and risk of cerebrovascular disease: systematic review and meta-analysis. Cardiovasc Diabetol 2022; 21(1): 226.

26.

Malick

Rosenson

RS.

Severe hypertriglyceridemia: existing and emerging therapies. Pharmacol Ther 2023; 251: 108544.

27.

Malick

Waksman

, et al. Clinical trial design for triglyceride-rich lipoprotein-lowering therapies: JACC Focus Seminar 3/3. J Am Coll Cardiol 2023; 81(16): 1646–1658.

28.

Ballantyne

CM.

Metabolic inflammation and insulin resistance in obesity. Circ Res 2020; 126(11): 1549–1564.

29.

Lee

Park

Lee

, et al. The usefulness of metabolic score for insulin resistance for the prediction of incident non-alcoholic fatty liver disease in Korean adults. Clin Mol Hepatol 2022; 28(4): 814–826.

30.

Xie

Peng

, et al. Diagnostic value of triglyceride-glucose index and related parameters in metabolism-associated fatty liver disease in a Chinese population: a cross-sectional study. BMJ Open 2023; 13(9): e075413.

31.

Tang

Chen

Zhou

, et al. NLR contributed to the diagnosis and detection of nonalcoholic fatty liver disease: a meta-analysis. Clin Res Hepatol Gastroenterol 2022; 46(3): 101847.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB