Immunometabolic drivers of metabolic dysfunction-associated steatotic liver disease in non-obese people living with human immunodeficiency virus

Abstract

BACKGROUND

Metabolic dysfunction-associated steatotic liver disease (MASLD) is common in people living with human immunodeficiency virus (PLWH), but its drivers in non-obese individuals are unclear. A low cluster of differentiation (CD) 4/CD8 ratio marks immune senescence, yet the role of a normalized ratio in MASLD risk in non-obese PLWH remains unexplored.

AIM

To investigate MASLD prevalence, risk factors, and short-term integrase strand transfer inhibitors (INSTIs) impact on steatosis in non-obese PLWH.

METHODS

This prospective study enrolled virologically suppressed PLWH who had received antiretroviral therapy for ≥ 6 months. All participants underwent clinical assessments, biochemical tests, and vibration-controlled transient elastography. Multivariate logistic regression identified independent risk factors for MASLD. Changes in parameters across the INSTIs-based, non-INSTIs-based, and switching to the INSTIs-based regime groups in non-obese PLWH were compared using mixed-effects models from baseline to month 12.

RESULTS

Among 181 enrolled participants [median age, 40 years (interquartile range: 35-48); 5.0% female], the overall prevalence of MASLD was 46.4%. Notably, among all patients with MASLD (n = 84), 28.6% (24/84) were non-obese. In the non-obese subgroup, higher body mass index [odds ratio (OR) = 2.14, 95% confidence interval (CI): 1.32-3.47], alanine aminotransferase (ALT) (OR = 1.04, 95%CI: 1.003-1.069), triglyceride-glucose index (OR = 3.29, 95%CI: 1.04-10.40), CD4/CD8 ratio (OR = 8.55, 95%CI: 1.57-46.66) and lower high-density lipoprotein level (OR = 0.029, 95%CI: 0.001-0.688) were independently associated with increased MASLD risk. 12-month longitudinal analysis revealed no significant group-by-time interaction effects for ultrasound attenuation parameter, liver stiffness measurement, and ALT.

CONCLUSION

MASLD is prevalent in non-obese PLWH. Higher CD4/CD8 ratio associates with MASLD, indicating that immune reconstitution does not ensure metabolic health. INSTIs were not associated with short-term hepatic steatosis progression.

Key Words: Human immunodeficiency virus; Metabolic dysfunction-associated steatotic liver disease; Integrase strand transfer inhibitors; Antiretroviral therapy; Transient elastography

Core Tip: In virologically suppressed people living with human immunodeficiency virus (PLWH), metabolic dysfunction-associated steatotic liver disease (MASLD) is highly prevalent even in non-obese individuals. We report a paradoxical finding: A higher cluster of differentiation (CD) 4/CD8 ratio, a marker of favorable immune recovery, was independently associated with increased MASLD risk in non-obese PLWH (odds ratio = 8.55). This challenges the view that immune reconstitution is the sole therapeutic endpoint. Instead, metabolic risk stratification including routine steatosis assessment is warranted even in immunologically favorable, non-obese individuals. Integrase strand transfer inhibitors-based antiretroviral therapy demonstrated short-term hepatic safety.

INTRODUCTION

Human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) remains a critical global public health challenge, with approximately 39 million people living with HIV (PLWH) worldwide[1]. While antiretroviral therapy (ART) has transformed HIV into a manageable chronic condition, PLWH now face a rising burden of non-AIDS-defining comorbidities[2,3], among which chronic liver disease is a leading cause of morbidity and mortality[4].

Metabolic dysfunction-associated steatotic liver disease (MASLD) is more prevalent in PLWH than in the general population[5,6], indicating HIV-specific pathophysiological mechanisms beyond traditional risk factors[7]. Notably, a substantial proportion of MASLD cases in PLWH occur in non-obese individuals[8,9]. In addition to direct viral and ART effects, persistent immune dysregulation and chronic low-grade inflammation despite virologic suppression are considered key drivers[7,10,11]. In parallel, elevated systemic inflammation markers, such as neutrophil-to-lymphocyte ratio, have been shown to predict adverse outcomes in MASLD[12], highlighting the clinical relevance of immune activation in this disease.

The cluster of differentiation (CD) 4/CD8 ratio, a simple clinical marker, reflects immune reconstitution status. While its normalization typically indicates immunological success, it has also been paradoxically associated with chronic low-grade inflammation (inflammaging) and an elevated risk of non-AIDS events, including cardiometabolic diseases[13-15]. This suggests that a “successful” numerical immune recovery may mask persistent functional immune disturbances, potentially involving dysregulated cytokine profiles, altered T-cell subsets (e.g., senescent or exhausted T cells), and systemic metabolic inflammation[16,17]. However, the specific relationship between the CD4/CD8 ratio and risk of MASLD, particularly in the non-obese PLWH, remains unclear and mechanistically unexplored.

Concurrently, the metabolic impact of modern ART, particularly integrase strand transfer inhibitors (INSTIs), warrants investigation. Although INSTIs are associated with weight gain[18-20], their direct effect on hepatic steatosis is inconsistent[21-24]. Therefore, it is crucial to distinguish the contribution of the underlying HIV-related immunometabolic disturbances from the potential effects of specific ART drugs.

Therefore, this study aimed to: (1) Investigate the prevalence and independent risk factors of MASLD in a cohort of non-obese Chinese HIV patients through cross-sectional analysis, with a particular focus on the role of the CD4/CD8 ratio; and (2) Longitudinally assess the short-term impact of different ART regimens, particularly INSTIs-based therapy, on hepatic steatosis and related metabolic and immune parameters in this population.

MATERIALS AND METHODS

Study design and participants

This was a prospective, single-center, observational study with a stratified design, including a cross-sectional and longitudinal cohort. The cross-sectional cohort consecutively enrolled PLWH from the specialist clinic of the Peking Union Medical College Hospital between June 2021 and June 2023 (ClinicalTrials.gov ID: NCT05330923). All the participants underwent vibration-controlled transient elastography (VCTE) for hepatic steatosis. The inclusion criteria were as follows: (1) Diagnosis of HIV infection according to the Chinese Guidelines for the Diagnosis and Treatment of HIV/AIDS (2021 Edition)[25]; (2) Receipt of ART for at least six months with sustained virological suppression (HIV-1 RNA < 20 copies/mL); and (3) Completion of VCTE examination at our hospital at the time of study enrollment. The exclusion criteria were: (1) Age < 18 years; (2) Significant alcohol intake[26]; (3) Active co-infection with hepatitis B or C; (4) Pregnancy; (5) Poor compliance and inability to follow up regularly or take medicine on time; (6) Severe hepatic/renal dysfunction (Child-Pugh class B/C, estimated glomerular filtration rate < 60 mL/minute/1.73 m²); and (7) Active malignant tumor or other life-threatening diseases.

From the above cross-sectional cohorts, all non-obese participants [defining body mass index (BMI) < 25 kg/m² as non-obese][27-29] were selected and included in a 12-month longitudinal follow-up study to form a longitudinal sub-cohort. All enrolled non-obese participants were categorized into three groups based on their baseline ART regimen: The INSTIs group (maintaining an INSTIs-based regimen), the non-INSTIs group (those remaining on nucleoside reverse transcriptase inhibitors or protease inhibitors-containing regimens), and the switch group (transitioning from non-INSTIs to an INSTIs-based regimen at baseline). The study complied with the Declaration of Helsinki and the 2018 Declaration of Istanbul, and was approved by the Ethics Committee of Peking Union Medical College Hospital, Chinese Academy of Medical Sciences (approval No. HS-3401D). Written informed consent was obtained from all the participants.

Clinical data collection

Data were collected cross-sectionally at baseline and longitudinally during follow-up. Baseline data (cross-sectional cohort): For all enrolled patients, we collected demographic characteristics (sex, age, weight, height, current medical history, past medical history, comorbidities, medication history, and calculated BMI), HIV-related details (ART regimens, HIV duration, HIV viral load), metabolic indicators [fasting blood glucose, total cholesterol (TC), triglycerides (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL), TC to HDL ratio (TC/HDL), aspartate aminotransferase, alanine aminotransferase (ALT), and uric acid (UA)], and immunological indicators (CD4 cell count and CD4/CD8 ratio). Follow-up data (longitudinal cohort): For non-obese HIV patients, data on demographic characteristics, HIV-related details, laboratory parameters, and repeated VCTE examinations were collected at 6- and 12-month visits (with a permissible window of ± 3 months).

Assessment of hepatic steatosis

Hepatic steatosis and fibrosis were assessed using VCTE (iLivTouch, Wuxi Haishi Medical Technology Co., Ltd., China) to measure the ultrasound attenuation parameter (UAP) and liver stiffness measurement (LSM). Five trained operators performed all examinations. They received standardized equipment operation and interpretation training before the start of the study. The participants were blinded to their clinical data. Following the standard protocol, at least ten valid measurements were obtained from the right liver lobe via the intercostal space. The median value was used for analysis, provided that the examination met the quality criteria [interquartile range (IQR)/median < 30% and success rate ≥ 60%]. The final analysis used the median of all valid measurements. Representative VCTE image was shown in Supplementary Figure 1.

Definition

UAP cutoffs determined the prevalence of hepatic steatosis: Mild for 269 dB/m > UAP ≥ 244 dB/m, moderate for 296 dB/m > UAP ≥ 269 dB/m, and severe for UAP ≥ 296 dB/m. This study is based on the 2023 international expert consensus and uses the latest term and diagnostic criteria of “MASLD”[30]. MASLD was defined as hepatic steatosis with at least one cardiometabolic risk factor[30]. Cardiometabolic risk factors include[28-30]: (1) BMI ≥ 25 kg/m²; (2) Fasting serum glucose ≥ 5.6 mmol/L or a diagnosis of type 2 diabetes or treatment for type 2 diabetes; (3) Blood pressure ≥ 130/85 mmHg or specific antihypertensive treatment; (4) Plasma TG ≥ 1.7 mmol/L or treatment for dyslipidemia; and (5) Plasma HDL ≤ 1.0 mmol/L for males and ≤ 1.3 mmol/L for females or treatment with lipid-lowering agents. Participants with HIV-1 RNA levels of < 20 copies/mL were defined as having viral suppression. The CD4/CD8 ratio is the ratio of the absolute count of CD4 cells to the absolute count of CD8 cells in peripheral blood. Triglyceride-glucose (TyG) = ln [TG (mg/dL) × fasting blood glucose (mg/dL)/2].

Statistical analysis

Data were analyzed using IBM SPSS software (v.26.0, Chicago, IL, United States) and GraphPad Prism software (v.8.3.0). Normality was assessed using the Kolmogorov-Smirnov test. Continuous variables are presented as mean ± SD or median (IQR) and compared using Student’s t-test or the Mann-Whitney U test. Categorical data were reported as absolute n (%) and compared using the χ² or Fisher’s exact test. To identify independent risk factors for MASLD, variables with P < 0.1 in univariable analysis or of clinical relevance were included in a multivariable logistic regression model, as specified in the footnotes of each table. Stratified analysis by CD4/CD8 ratio (cutoff = 1) was performed to examine whether the association between risk factors and MASLD differed by immune status[15]. In the longitudinal analysis of the non-obese sub-cohort, linear mixed-effects models were employed to evaluate changes in outcomes over time (baseline and 6 and 12 months) across the three ART groups, efficiently handling incomplete follow-up data. Statistical significance was set at P < 0.05. Sample size was based on consecutive enrollment.

Sensitivity analyses for missing data

Sensitivity analyses were conducted to assess the robustness of the longitudinal findings. First, to evaluate the impact of missing data, we performed multiple imputation by chained equations, generating five imputed datasets. The imputation model included all longitudinal outcome variables (UAP, LSM, BMI, weight, ALT, TC, TG, HDL, LDL, glucose, UA, TyG, TC/HDL, CD4 cell count, CD4/CD8 ratio) used in the primary analysis. Linear mixed-effects models were refitted to each imputed dataset, and the range of P values for the three fixed effects (group, time, group × time interaction) is reported in Supplementary Table 1. Consistency between the original and imputed analyses was further confirmed by comparing the estimated marginal means and their 95% confidence interval (CI) (Supplementary Table 2). Second, to assess potential selection bias due to attrition, we compared baseline characteristics between: (1) Participants who completed the 12-month follow-up (having both baseline and 12-month UAP/LSM data, n = 64) and those who did not (n = 51); and (2) Participants with complete UAP/LSM data at all three time points (n = 45) and those with any missing data (n = 70). Results are presented in Supplementary Tables 3 and 4, respectively.

RESULTS

Baseline characteristics and prevalence of MASLD

A total of 181 PLWH were enrolled in this study (Figure 1). The median age was 40 (35-48) years, and 9 (5.0%) were female (Table 1). Compared with the non-MASLD group, patients in the MASLD were older and had a higher body weight, BMI, UAP, LSM, and levels of ALT, TG, glucose, UA, TyG index, and TC/HDL ratio, but lower HDL levels (all P < 0.05; Table 1). Hypertension and antihypertensive drug use were more prevalent in the MASLD group. Immunologically, a higher proportion of patients with MASLD had a current CD4 cell count of ≥ 350 cells/μL.

Figure 1 Study flow chart. HIV: Human immunodeficiency virus; VCTE: Vibration-controlled transient elastography; INSTI: Integrase strand transfer inhibitors.

Table 1 Demographic and clinical characteristics of the total population, mean ± SD/n (%)/median (interquartile range).

Variables	Total (n = 181)	Non-MASLD (n = 97)	MASLD (n = 84)	P value
Age, years	40 (35-48)	39 (34-46)	41 (36-52)	< 0.05
Gender				0.57
Male	172 (95.0)	93 (95.9)	79 (94.1)
Female	9 (5.0)	4 (4.1)	5 (6.0)
Weight, kg	72.9 ± 12.6	65.5 ± 8.9	81.5 ± 10.6	< 0.001
BMI, kg/m2	24.0 ± 3.6	21.7 ± 2.1	26.7 ± 3.1	< 0.001
ALT, U/L	25.0 (17.8-34.0)	21.5 (16.0-30.0)	28.0 (21.0-41.8)	< 0.001
Glucose, mmol/L	5.3 (5.0-5.7)	5.2 (5.0-5.5)	5.5 (5.1-5.9)	< 0.01
UA, μmol/L	366.5 ± 90.9	350.8 ± 83.6	384.2 ± 95.9	< 0.05
LDL, mmol/L	2.9 ± 0.7	2.8 ± 0.7	2.9 ± 0.7	0.57
HDL, mmol/L	1.2 (1.0-1.4)	1.3 (1.2-1.5)	1.1 (0.9-1.2)	< 0.001
TG, mmol/L	1.4 (1.0-2.2)	1.2 (0.9-1.9)	1.5 (1.1-2.8)	< 0.01
TC, mmol/L	4.7 (4.0-5.4)	4.6 (4.0-5.3)	4.9 (4.0-5.7)	0.16
TyG index	7.1 (6.8-7.7)	7.0 (6.6-7.5)	7.3 (6.9-7.8)	< 0.01
TC/HDL	4.0 (3.4-4.8)	3.6 (2.9-4.2)	4.5 (3.6-5.2)	< 0.001
UAP, dB/m	253.6 (226.4-272.1)	227.1 (216.1-240.2)	276.1 (262.5-291.1)	< 0.001
LSM, kPa	5.4 (4.61-6.6)	5.1 (4.4-5.9)	5.9 (5.0-7.1)	< 0.001
Duration, months	63.2 (48.7-93.2)	69.3 (50.4-93.5)	60.8 (48.6-95.1)	0.72
Initial HVL, copies/mL	50149.0 (22723.0-185078.5)	50149.0 (23000.0-175952.0)	49541.5 (22418.3-222100.3)	0.91
Initial CD4 cell count, cells/μL	276.1 ± 177.7	262.5 ± 174.5	292.2 ± 181.2	0.28
CD4 cell count, cells/μL	562.0 (399.5-677.0)	562.0 (378.0-694.0)	568.0 (439.3-665.0)	0.43
CD4 cell count ≥ 300 cells/μL				< 0.05
No	53 (29.3)	35 (36.1)	18 (21.4)
Yes	128 (70.7)	62 (63.9)	66 (78.7)
CD4/CD8 ratio	0.8 (0.6-1.1)	0.7 (0.5-1.1)	0.8 (0.6-1.1)	0.30
Group				0.26
INSTIs	76 (42.0)	42 (43.3)	34 (40.5)
Switch	32 (17.8)	13 (13.4)	19 (22.6)
Non-INSTIs	73 (40.3)	42 (43.3)	31 (36.9)
Drinking				0.86
No	175 (96.7)	94 (96.9)	81 (96.4)
Yes	6 (3.3)	3 (3.1)	3 (3.6)
Smoking				0.48
No	178 (98.3)	96 (99.0)	82 (97.6)
Yes	3 (1.7)	1 (1.0)	2 (2.4)
Hypertension				0.82
No	154 (85.1)	89 (91.8)	65 (77.4)
Yes	27 (14.9)	8 (8.3)	19 (22.6)
Diabetes				< 0.01
No	173 (95.6)	95 (97.9)	78 (92.9)
Yes	8 (4.4)	2 (2.1)	6 (7.1)
Lipid lowering drug				0.05
No	171 (94.5)	92 (94.9)	79 (94.1)
Yes	10 (5.5)	5 (5.2)	5 (6.0)
Antihypertensive drug				0.86
No	158 (87.3)	92 (94.9)	66 (78.6)
Yes	23 (12.7)	5 (5.2)	18 (21.4)
Hypoglycemic drug				0.48
No	172 (95.0)	95 (97.9)	77 (91.7)
Yes	9 (5.0)	2 (2.1)	7 (8.3)

Due to rounding, percentages may not add up to 100%. MASLD: Metabolic dysfunction-associated steatotic liver disease; BMI: Body mass index; ALT: Alanine aminotransferase; UA: Uric acid; LDL: Low-density lipoprotein; HDL: High-density lipoprotein; TG: Triglycerides; TC: Total cholesterol; TyG: Triglyceride-glucose; CD: Cluster of differentiation; UAP: Ultrasound attenuation parameter; LSM: Liver stiffness measurement; HVL: Human immunodeficiency virus viral load; INSTIs: Integrase strand transfer inhibitors-based regime group; Switch: Switching-to-integrase strand transfer inhibitors-based regime groups; Non-INSTIs: Non-integrase strand transfer inhibitors-based regime group.

Characteristics of patients with MASLD among non-obese PLWH

Given the notable proportion of MASLD among non-obese individuals, we performed subgroup analysis. Among non-obese patients, those with MASLD presented with higher body weight, BMI, ALT, TG, glucose, TyG index, TC/HDL ratio, and UAP but lower HDL levels (all P < 0.05; Table 2). They also exhibited a more favorable immunological profile, including a higher CD4/CD8 ratio and a greater proportion with a CD4 cell count of ≥ 350 cells/μL.

Table 2 Demographic and clinical characteristics of non-obese participants with metabolic dysfunction-associated steatotic liver disease and non-metabolic dysfunction-associated steatotic liver disease, mean ± SD/n (%)/median (interquartile range).

Variables	Non-MASLD (n = 91)	MASLD (n = 24)	P value
Age, years	39 (33-46)	39 (36-49)	0.35
Gender			0.30
Male	87 (95.6)	24 (100.0)
Female	4 (4.4)	0 (0.0)
Weight, kg	64.9 ± 8.7	70.9 ± 5.9	< 0.001
BMI, kg/m2	21.5 (20.1-23.4)	23.2 (22.8-24.4)	< 0.001
ALT, U/L	21.0 (16.0-29.0)	31.0 (21.3-54.3)	< 0.01
Glucose, mmol/L	5.2 (5.0-5.5)	5.5 (5.1-5.9)	< 0.05
UA, μmol/L	349.5 ± 85.1	377.2 ± 71.5	0.15
LDL, mmol/L	2.8 ± 0.7	2.8 ± 0.7	0.82
HDL, mmol/L	1.3 (1.2-1.5)	1.0 (0.9-1.2)	< 0.001
TG, mmol/L	1.2 (0.9-1.9)	1.9 (1.1-3.2)	< 0.05
TC, mmol/L	4.6 (4.1-5.3)	4.9 (4.1-5.8)	0.25
UAP, dB/m	228.5 ± 20.9	264.2 ± 11.9	< 0.001
LSM, kPa	5.1 (4.4-5.9)	5.1 (4.4-6.7)	0.29
Duration, months	67.5 (50.2-93.1)	75.6 (58.4-107.9)	0.17
Group			0.58
INSTIs	40 (44.0)	8 (33.3)
Switch	13 (14.3)	5 (20.8)
Non-INSTIs	38 (41.8)	11 (45.8)
TyG index	7.1 ± 0.7	7.5 ± 0.7	< 0.01
TC/HDL	3.6 (2.9-4.2)	4.6 (3.6-5.5)	< 0.001
Initial HVL, copies/mL	46899.5 (21446.5-141112.5)	30761.5 (14159.0-177755.8)	0.48
Initial CD4 cell count, cells/μL	269.2 ± 176.7	279.1 ± 125.1	0.80
CD4 cell count, cells/μL	559.0 (376.0-682.0)	519.0 (428.8-672.8)	0.51
CD4 cell count ≥ 350 cells/μL			< 0.01
No	33 (36.3)	1 (4.2)
Yes	58 (63.7)	23 (95.8)
CD4/CD8 ratio	0.7 (0.5-1.1)	0.9 (0.6-1.4)	< 0.05
Drinking			0.84
No	88 (96.7)	23 (95.8)
Yes	3 (3.3)	1 (4.2)
Smoking			0.61
No	90 (98.9)	24 (100.0)
Yes	1 (1.1)	0 (0.0)
Hypertension			< 0.05
No	85 (93.4)	19 (79.2)
Yes	6 (6.6)	5 (20.8)
Diabetes			0.46
No	89 (97.8)	24 (100.0)
Yes	2 (2.2)	0 (0.0)
Lipid lowering drug			0.14
No	87 (95.6)	21 (87.5)
Yes	4 (4.4)	3 (12.5)
Antihypertensive drug			< 0.05
No	88 (96.7)	20 (83.3)
Yes	3 (3.3)	4 (16.7)
Hypoglycemic drug			0.46
No	89 (97.8)	24 (100.0)
Yes	2 (2.2)	0 (0.0)

Due to rounding, percentages may not add up to 100%. MASLD: Metabolic dysfunction-associated steatotic liver disease; BMI: Body mass index; ALT: Alanine aminotransferase; UA: Uric acid; LDL: Low-density lipoprotein; HDL: High-density lipoprotein; TG: Triglycerides; TC: Total cholesterol; TyG: Triglyceride-glucose; CD: Cluster of differentiation; UAP: Ultrasound attenuation parameter; LSM: Liver stiffness measurement; HVL: Human immunodeficiency virus viral load; INSTIs: Integrase strand transfer inhibitors-based regime group; Switch: Switching-to-integrase strand transfer inhibitors-based regime groups; Non-INSTIs: Non-integrase strand transfer inhibitors-based regime group.

Risk factors of MASLD among total population

In the regression analysis, we found that high BMI [odds ratio (OR) = 2.58; 95%CI: 1.85-3.60], high TyG (OR = 3.27; 95%CI: 1.26-8.48), ALT level (OR = 1.03; 95%CI: 1.002-1.058), and low HDL (OR = 0.17; 95%CI: 0.049-0.59) were independently associated with an increased risk of MASLD (Figure 2).

Figure 2 Adjusted multivariate logistic regression analysis of metabolic dysfunction-associated steatotic liver disease risk factors in overall human immunodeficiency virus patients. Variables included in the model are gender, age, group, body mass index, alanine aminotransferase, high-density lipoprotein, triglyceride-glucose, glucose, cluster of differentiation (CD) 4 cell count ≥ 350 cells/μL, and CD4/CD8 ratio. OR: Odds ratio; CI: Confidence interval; BMI: Body mass index; HDL: High-density lipoprotein; ALT: Alanine aminotransferase; TyG: Triglyceride-glucose.

Risk factors of MASLD among non-obese PLWH

In regression analysis, we found that high BMI (OR = 2.14, 95%CI: 1.32-3.47), ALT level (OR = 1.04, 95%CI: 1.003-1.069), TyG level (OR = 3.29, 95%CI: 1.04-10.40), CD4/CD8 ratio (OR = 8.55, 95%CI: 1.57-46.66), and low HDL level (OR = 0.03, 95%CI: 0.001-0.688) were independently associated with an increased risk of MASLD among non-obese PLWH (Table 3).

Table 3 Multivariable analysis of factors associated with metabolic dysfunction-associated steatotic liver disease among non-obese people living with human immunodeficiency virus.

Variables	Univariate analysis		Multivariable analysis
	P value	Crude OR (95%CI)	P value	Adjusted OR (95%CI)
Age	0.34	1.05 (0.96-1.14)
Gender	1.00	137533244.9 (0-.)1
Group
INSTIs	0.75
Switch	0.46	1.85 (0.36-9.48)
Non-INSTIs	0.67	1.56 (0.20-11.92)
BMI	0.004	2.03 (1.26-3.26)	0.002	2.14 (1.32-3.47)
ALT	0.027	1.04 (1.00-1.07)	0.032	1.04 (1.003-1.069)
HDL	0.015	0.01 (0-0.37)	0.028	0.03 (0.001-0.688)
Glucose	0.20	0.68 (0.38-1.22)
TyG index	0.06	3.31 (0.93-11.75)	0.043	3.29 (1.04-10.40)
CD4 cell count ≥ 350 cells/μL	0.19	5.88 (0.42-100.00)
CD4/CD8 ratio	0.19	4.21 (0.49-35.84)	0.013	8.55 (1.57-46.66)

¹The extreme odds ratio for gender results from complete separation caused by the limited number of female participants (n = 4, all in the non-metabolic dysfunction-associated steatotic liver disease group). The upper bound of the 95% confidence interval is not estimable.

Variables included gender, age, group, body mass index, alanine aminotransferase, high-density lipoprotein, triglyceride-glucose, glucose, cluster of differentiation (CD) 4 cell count ≥ 350 cells/μL, and CD4/CD8 ratio. OR: Odds ratio; CI: Confidence interval; BMI: Body mass index; ALT: Alanine aminotransferase; HDL: High-density lipoprotein; TyG: Triglyceride-glucose; CD: Cluster of differentiation; INSTIs: Integrase strand transfer inhibitors-based regime group; Switch: Switching-to-integrase strand transfer inhibitors-based regime groups; Non-INSTIs: Non-integrase strand transfer inhibitors-based regime group.

In the non-obese HIV subgroup, further stratification according to immune status revealed distinct risk profiles (Table 4). In patients with a CD4/CD8 ratio ≥ 1, only a higher BMI (P = 0.008, OR = 3.01, 95%CI: 1.34-6.76) was independently associated with MASLD. In contrast, among those with a CD4/CD8 ratio < 1, MASLD risk was associated with higher BMI (P = 0.026, OR = 1.73, 95%CI: 1.07-2.80), lower HDL (P = 0.016, OR = 0.01, 95%CI: 0.00-0.40), and CD4 cell count ≥ 350 cells/μL (P = 0.031, OR = 13.00, 95%CI: 1.27-125.0).

Table 4 Multivariable analysis of factors associated with liver steatosis among non-obese people living with human immunodeficiency virus with cluster of differentiation 4/cluster of differentiation 8 ratio ≥ 1 and < 1.

Variables	CD4/CD8 ratio ≥ 1		CD4/CD8 ratio < 1
	P value	Adjusted OR (95%CI)	P value	Adjusted OR (95%CI)
BMI	0.008	3.01 (1.34-6.76)	0.026	1.73 (1.07-2.80)
HDL			0.016	0.01 (0.00-0.40)
CD4 cell count ≥ 350 cells/μL			0.031	13.00 (1.27-125.00)

Variables included gender, age, group, body mass index, alanine aminotransferase, high-density lipoprotein, triglyceride-glucose, glucose, and cluster of differentiation 4 cell count ≥ 350 cells/μL in models 1 and 2. OR: Odds ratio; CI: Confidence interval; CD: Cluster of differentiation; BMI: Body mass index; HDL: High-density lipoprotein.

Changes of hepatic fat, metabolic, and immunological parameters in different ART regimens

Longitudinal analysis over 12 months revealed that most parameters (UAP, LSM, ALT, CD4/CD8 ratio, CD4 cell count) showed no significant changes over time or between groups (all P > 0.05; Table 5 and Figure 3). However, significant time effects were observed for BMI (P < 0.01) and TyG (P < 0.05), and significant group effects were noted for HDL (P < 0.01) and glucose (P < 0.05). Importantly, no significant group-by-time interaction effects were found for any parameter (all P > 0.05; Table 5), indicating that the trajectories of change over the 12-month period did not differ significantly among the INSTIs, switch, and non-INSTIs groups.

Figure 3 Longitudinal changes in hepatic, metabolic, and immunological parameters among non-obese people with human immunodeficiency virus (12-month follow-up). A: Ultrasound attenuation parameter; B: Liver stiffness measurement; C: Alanine aminotransferase; D: Body mass index; E: Glucose; F: High-density lipoprotein; G: Triglyceride-glucose; H: Cluster of differentiation (CD) 4 cell count; I: CD4/CD8 ratio. Data are presented as mean ± SD. The mixed-effects model analysis showed that none of the unmarked indicators were statistically significant. INSTIs: Integrase strand transfer inhibitors-based regime group; Switch: Switching-to-integrase strand transfer inhibitors-based regime groups; Non-INSTIs: Non-integrase strand transfer inhibitors-based regime group; UAP: Ultrasound attenuation parameter; LSM: Liver stiffness measurement; ALT: Alanine aminotransferase; BMI: Body mass index; HDL: High-density lipoprotein; TyG: Triglyceride-glucose; CD: Cluster of differentiation.

Table 5 Longitudinal changes in key metabolic, immunological, and hepatic parameters across three antiretroviral therapy regimens (mixed-effects model results), mean ± SD.

Characteristics	Group	Baseline	Month 6	Month 12	P value for time effect	P value for group effect	P value for time × group effect
UAP, dB/m	INSTIs	237.6 ± 24.5 (n = 47)	238.4 ± 24.3 (n = 30)	236.1 ± 23.5 (n = 21)	0.8	0.087	0.583
	Switch	239.1 ± 24.4 (n = 19)	243.3 ± 19.0 (n = 15)	237.1 ± 20.6 (n = 15)
	Non-INSTIs	233.2 ± 24.0 (n = 49)	224.7 ± 27.1 (n = 24)	228.6 ± 28.8 (n = 28)
LSM, kPa	INSTIs	5.4 ± 1.4 (n = 47)	6.1 ± 3.2 (n = 30)	5.8 ± 0.9 (n = 21)	0.911	0.872	0.552
	Switch	6.0 ± 2.9 (n = 19)	6.4 ± 2.6 (n = 15)	6.2 ± 3.7 (n = 15)
	Non-INSTIs	5.3 ± 1.3 (n = 49)	5.4 ± 1.2 (n = 24)	5.5 ± 1.1 (n = 28)
ALT, U/L	INSTIs	24.9 ± 16.0 (n = 43)	26.1 ± 16.2 (n = 28)	31.4 ± 16.7 (n = 25)	0.587	0.19	0.999
	Switch	28.0 ± 15.7 (n = 17)	26.4 ± 12.8 (n = 16)	29.6 ± 24.5 (n = 12)
	Non-INSTIs	33.3 ± 26.5 (n = 43)	33.2 ± 28.7 (n = 30)	34.6 ± 26.9 (n = 22)
BMI, kg/m2	INSTIs	22.0 ± 2.3 (n = 47)	22.7 ± 2.2 (n = 30)	22.5 ± 2.3 (n = 21)	< 0.01	0.258	0.057
	Switch	22.0 ± 1.8 (n = 19)	22.7 ± 1.6 (n = 15)	23.1 ± 1.6 (n = 15)
	Non-INSTIs	21.6 ± 1.9 (n = 49)	22.0 ± 1.6 (n = 23)	22.2 ± 1.8 (n = 28)
HDL, mmol/L	INSTIs	1.5 ± 0.8 (n = 43)	1.6 ± 0.9 (n = 29)	1.5 ± 0.7 (n = 24)	0.323	< 0.01	0.979
	Switch	1.4 ± 0.6 (n = 17)	1.4 ± 0.6 (n = 17)	1.2 ± 0.2 (n = 10)
	Non-INSTIs	1.2 ± 0.3 (n = 42)	1.3 ± 0.6 (n = 31)	1.2 ± 0.3 (n = 24)
Glucose, mmol/L	INSTIs	5.2 ± 0.5 (n = 43)	5.3 ± 0.4 (n = 29)	5.0 ± 0.9 (n = 22)	0.056	< 0.05	0.357
	Switch	5.9 ± 2.0 (n = 17)	5.5 ± 0.5 (n = 17)	5.3 ± 0.3 (n = 9)
	Non-INSTIs	5.6 ± 1.3 (n = 43)	5.5 ± 0.4 (n = 31)	5.5 ± 0.3 (n = 19)
TyG index	INSTIs	7.1 ± 0.7 (n = 40)	7.4 ± 0.9 (n = 28)	6.9 ± 0.5 (n = 17)	< 0.05	0.572	0.057
	Switch	7.2 ± 0.9 (n = 15)	7.0 ± 0.4 (n = 16)	6.7 ± 0.3 (n = 8)
	Non-INSTIs	7.2 ± 0.6 (n = 42)	7.1 ± 0.4 (n = 25)	7.1 ± 0.6 (n = 14)
CD4/CD8 ratio	INSTIs	0.8 ± 0.4 (n = 43)	0.8 ± 0.4 (n = 29)	0.8 ± 0.4 (n = 25)	0.324	0.452	0.322
	Switch	1.0 ± 0.5 (n = 17)	1.0 ± 0.5 (n = 16)	0.9 ± 0.4 (n = 12)
	Non-INSTIs	0.9 ± 0.4 (n = 44)	0.8 ± 0.3 (n = 29)	0.9 ± 0.4 (n = 22)
CD4 cell count, cells/μL	INSTIs	522.7 ± 204.6 (n = 42)	570.9 ± 188.2 (n = 29)	524.8 ± 180.2 (n = 27)	0.381	0.101	0.81
	Switch	647.9 ± 241.0 (n = 17)	656.7 ± 262.7 (n = 16)	613.5 ± 224.2 (n = 13)
	Non-INSTIs	533.9 ± 224.4 (n = 44)	534.7 ± 226.9 (n = 29)	487.5 ± 229.5 (n = 24)

Number of patients with non-missing data at each time point. UAP: Ultrasound attenuation parameter; LSM: Liver stiffness measurement; ALT: Alanine aminotransferase; BMI: Body mass index; HDL: High-density lipoprotein; TyG: Triglyceride-glucose; CD: Cluster of differentiation; INSTIs: Integrase strand transfer inhibitors-based regime group; Switch: Switching-to-integrase strand transfer inhibitors-based regime groups; Non-INSTIs: Non-integrase strand transfer inhibitors-based regime group.

Sensitivity analysis

The results of the sensitivity analyses supported the robustness of the primary findings. First, the P values from the original mixed-model analysis fell within the range of those obtained from the five imputed datasets for all fixed effects (Supplementary Table 1). Furthermore, the pooled estimated marginal means and their 95%CIs were nearly identical to those from the original analysis (Supplementary Table 2). Second, baseline comparisons between participants with complete vs incomplete UAP/LSM data under two definitions (Supplementary Tables 3 and 4) showed no significant differences in the primary exposure (CD4/CD8 ratio) or the primary outcomes (UAP and LSM), despite minor differences in some secondary variables (initial viral load, TC/HDL, ART regimen). Taken together, these analyses suggest that the observed associations particularly between CD4/CD8 ratio and MASLD are unlikely to have been substantially biased by the pattern of missing data in this cohort.

DISCUSSION

This study assessed the prevalence and risk factors for MASLD among PLWH in China, particularly in non-obese individuals, and preliminarily explored the short-term effects of INSTIs. Our main findings were as follows: (1) The overall prevalence of MASLD was 46.4%, with a substantial proportion (28.6%) occurring in non-obese individuals; (2) In non-obese PLWH, a higher CD4/CD8 ratio a marker of favorable immune reconstitution was paradoxically associated with increased MASLD risk (OR = 8.55); and (3) Over 12 months, INSTIs-based regimens did not lead to significant differential changes in hepatic steatosis or key metabolic parameters across ART regimens (all group-by-time interaction P > 0.05). Importantly, clinical markers of “success” such as viral suppression or CD4/CD8 normalization do not always align with biological health.

The non-obese MASLD phenotype in PLWH

Although obesity is the main driver of MASLD in the general population, up to 28.6% of the MASLD cases in this study occurred in non-obese PLWH. This proportion is similar to that reported in Western PLWH using comparable controlled attenuation parameter thresholds (≥ 248 dB/m) and BMI cutoffs (< 25 kg/m²) (24.2%)[31], but slightly higher than that observed in liver biopsy studies of the general Asian population (21.6%)[32]. This confirms a distinct non-obese MASLD subgroup, where persistent immune-metabolic dysregulation may elevate risk even without caloric excess. Consistent with this, the large 2000 HIV cohort study from the Netherlands recently reported that among virologically suppressed PLWH (defined as BMI < 23 kg/m² for Asian descent and < 25 kg/m² for other descent), liver steatosis affected approximately onefifth of lean individuals, and importantly, steatosis in lean PLWH was associated with both CD4 and CD8 counts a pattern not observed in overweight/obese participants[9]. This further suggests that in non-obese PLWH, immune reconstitution status may be particularly relevant to liver steatosis.

The CD4/CD8 ratio paradox: A dual-pathway hypothesis

The central finding driving this hypothesis is that in non-obese PLWH, a higher CD4/CD8 ratio was independently associated with increased MASLD risk (OR = 8.55). Moreover, subgroup analysis revealed distinct risk profiles: In patients with CD4/CD8 ratio ≥ 1, only higher BMI was an independent risk factor; In those with ratio < 1, risk was associated with higher BMI, lower HDL, and higher CD4 cell count (≥ 350 cells/μL). To explain these contrasting patterns, we propose a “dual-pathway” hypothesis based on immune-metabolic imbalance.

Interestingly, our finding appears to contrast with a recent Thai cohort study of older PLWH (median age 54 years), which reported that a higher CD4/CD8 ratio was associated with a lower risk of MAFLD (OR = 0.32)[33]. This discrepancy likely reflects differences in study populations: Their cohort was older, had higher BMI, and included 37% female participants, whereas our cohort consisted exclusively of non-obese, predominantly male, younger PLWH. Beyond demographic differences, older age and higher BMI in the Thai cohort may be associated with more advanced immunosenescence and distinct adipose tissue distribution, potentially altering the relationship between immune status and metabolic risk. However, the relationship between immune recovery and metabolic complications is not unidirectional. A pilot study from India similarly observed that a higher median CD4 cell count was associated with MASLD compared to those without fatty changes on ultrasound[34]. Thus, the paradoxical association between CD4/CD8 normalization and MASLD risk may be particularly evident in non-obese individuals, highlighting the importance of population-specific immunometabolic phenotyping.

The following sections elaborate on each pathway

Pathway 1 (CD4/CD8 ratio ≥ 1): In this pathway corresponding to our subgroup with CD4/CD8 ratio ≥ 1, where only higher BMI was an independent risk factor we propose a “pro-inflammatory immune reconstitution-driven” mechanism. Individuals with quantitative immune reconstitution may still harbor persistent chronic low-grade inflammation[10,11,13]. This chronic inflammation promotes insulin resistance and hepatic fat deposition[10,11,13], making the liver abnormally sensitive to metabolic loads. Thus, even a modest increase in BMI can trigger steatosis in an already primed inflammatory milieu, explaining why BMI remains the dominant risk factor in this subgroup despite the underlying immune activation. Recent advances in immunometabolism have revealed a bidirectional crosstalk between innate and adaptive immunity in MASLD: While innate immune cells drive steatosis, the transition to steatohepatitis requires the involvement of T and B cells, which become metabolically reprogrammed by the lipotoxic environment[35]. In particular, a population of auto-aggressive CXCR6⁺ CD8⁺ T cells has been identified in non-alcoholic steatohepatitis that directly kills hepatocytes in a metabolic-sensitive manner[36]. Consistent with this, a recent single-cell atlas of human livers with metabolic dysfunction-associated steatohepatitis (MASH) demonstrated enrichment of exhausted CD8⁺ T cells and immunosuppressive S100A9⁺ macrophages, reinforcing the concept that chronic inflammation reshapes the hepatic immune landscape toward a pro-fibrotic and dysfunctional state[37]. This observation supports our hypothesis that a normalized CD4/CD8 ratio might be accompanied by dysfunctional, pro-inflammatory T-cell subsets that promote hepatic steatosis. Therefore, MASLD risk in this subgroup may partly stem from a dysregulated inflammatory microenvironment accompanying quantitative immune recovery a pathway we term “pro-inflammatory immune reconstitution-driven”. Indeed, elevated levels of the inflammasome-derived cytokine interleukin-18 (IL-18) have been shown to correlate positively with liver enzymes and hepatic steatosis in PLWH[38]. We acknowledge that our study did not directly measure inflammatory markers [e.g., high-sensitivity C-reactive protein (hs-CRP), IL-6] or perform immune cell phenotyping; this is a key gap in validating this pathway.

Pathway 2 (CD4/CD8 ratio < 1): In this pathway corresponding to our subgroup with CD4/CD8 ratio < 1, where risk was associated with higher BMI, lower HDL, and higher CD4 cell count (≥ 350 cells/μL) we propose an “immunodeficiency with metabolic dysregulation” mechanism. Patients with a persistently low CD4/CD8 ratio are often in a state of profound immunodeficiency[39]. In this setting, the immune system may exhibit a “low-reactivity” state that temporarily dampens immune-mediated hepatic inflammation. However, the absence of robust systemic inflammation does not imply metabolic health. Emerging evidence indicates that immunodeficiency itself can serve as an early risk factor for insulin resistance, independent of inflammatory markers[40], and that even in the context of low CD4/CD8 ratios, chronic CD8⁺ T cell activation persists and is independently associated with visceral fat accumulation and lipodystrophy in virologically suppressed PLWH[41]. Furthermore, adipose tissue in HIV infection is enriched for activated and late-differentiated CD8⁺ T cells, which can directly impair adipocyte function and promote local inflammation[42,43].

Crucially, in this subgroup, a higher CD4 cell count (≥ 350 cells/μL) was associated with increased MASLD risk a finding that aligns with our main observation that better immune recovery, even if partial, is linked to higher MASLD risk. This suggests that as immune function begins to recover, an underlying metabolic vulnerability becomes unmasked. Supporting this interpretation, the TyG index a validated surrogate for insulin resistance that we identified as an independent risk factor for MASLD in our cohort (OR = 3.29) has recently been shown to correlate positively with CD8⁺ T-cell counts and negatively with the CD4/CD8 ratio in PLWH; moreover, the TyG index was closely associated with hepatic steatosis (area under the curve = 0.743 at week 52)[44]. Taken together, these observations indicate that in patients with a low CD4/CD8 ratio, MASLD risk is driven primarily by lipid metabolism disorders and adipose tissue dysfunction, with partial immune recovery acting as a permissive factor that exposes this metabolic vulnerability. Subgroup analysis supported this distinction: In patients with CD4/CD8 ratio ≥ 1, only higher BMI was independently associated with MASLD; In those with ratio < 1, risk was associated with higher BMI, lower HDL, and CD4 cell count ≥ 350 cells/μL though wide confidence intervals due to limited sample size warrant cautious interpretation.

Future directions to verify the hypothesis

To empirically test these pathways, future studies should prioritize measurements that directly address each pathway. For pathway 1 (pro-inflammatory immune reconstitution), priority should be given to circulating inflammatory cytokines (IL-6, IL-1β, tumor necrosis factor-α, IL-18), monocyte activation markers (sCD14, CD163), and T-cell exhaustion/senescence markers (programmed death 1⁺, T cell immunoglobulin and mucin domain-containing protein 3⁺, CD57⁺, CD28^-) in CD8⁺ T-cell subsets[17,36,45,46]. For pathway 2 (immunodeficiency with metabolic dysregulation), priority should be given to metabolomic profiling (lipid subspecies, ceramides), adipokine measurements (leptin, adiponectin), and assessment of adipose tissue function and insulin resistance. Additionally, exploring the role of fatty acid transporters such as CD36 in mediating lipid-induced inflammation may uncover novel therapeutic targets[47]. Beyond classical immunology, other molecular mechanisms (e.g., long noncoding RNAs[48], antimicrobial peptides[49]) and advanced methodologies (e.g., in-cell nuclear magnetic resonance spectroscopy[50], unmodified RNA analysis[51]) may offer further insights into the immune-metabolic crosstalk observed in this study.

Longitudinal findings and clinical implications

Although previous studies have shown that ART drugs are associated with chronic inflammation[52], our 12-month follow-up revealed no significant group-by-time interaction effects for hepatic steatosis or metabolic parameters across ART regimens (Table 5). This negative result is noteworthy given prior reports linking INSTIs to weight gain and metabolic disturbances. Several explanations may account for this discrepancy. First, the 12-month follow-up may be insufficient to capture the long-term metabolic effects of ART, particularly the lagged effects of weight change on hepatic steatosis. Second, our cohort consisted of non-obese, virally suppressed PLWH with relatively stable metabolic profiles, whereas weight gain and metabolic changes associated with INSTIs may be more pronounced in overweight/obese individuals or those with baseline metabolic syndrome. Third, the absence of differential trajectories across ART regimens suggests that the CD4/CD8 ratio-MASLD association reflects long-term immunometabolic dysregulation inherent to the host or HIV infection itself, rather than a short-term effect attributable to contemporary ART. The observed significant time effects (e.g., on BMI and TyG) likely reflect cohort-wide trends or subtle lifestyle changes, rather than a drug-specific impact. Similarly, the group effects (e.g., on HDL and glucose) were largely attributable to baseline differences that remained stable, as the mixed-effects model showed no significant group-by-time interactions for these parameters (Table 5).

Our findings have direct clinical implications. First, proactive metabolic monitoring including assessment of hepatic steatosis (e.g., by VCTE or ultrasound), glucose-lipid profiles, and body composition should be integrated into routine HIV care, analogous to the management of other modifiable risk factors in chronic liver disease[53]. Second, based on our findings, we propose that non-obese PLWH with a normalized CD4/CD8 ratio (≥ 1) may warrant more aggressive screening for hepatic steatosis, as they exhibit a paradoxical increase in MASLD risk despite favorable immune recovery. Third, for non-obese PLWH with persistently low CD4/CD8 ratio (< 1) and dyslipidemia (low HDL, high TG), management should focus on metabolic risk factor modification (weight control, lipid management) in addition to immune recovery. Successful immune reconstitution does not guarantee metabolic health, and non-obese PLWH warrant particular attention.

Limitations

This study had several limitations. Firstly, the main analysis had a cross-sectional design, which cannot fully rule out residual confounding from unmeasured factors, nor can it establish a causal relationship between the CD4/CD8 ratio and MASLD. To address this, we are conducting a multicenter study aimed at establishing a long-term cohort to validate the predictive value of the CD4/CD8 ratio for MASLD incidence and progression and to clarify the temporal relationship between its changes and dynamic changes in liver fat content. Secondly, this study had a relatively small sample size from a single center, and the longitudinal follow-up period was only 12 months. This duration may be too short to observe significant changes in hepatic steatosis, especially considering the potential lag effects of metabolic alterations and weight change on liver fat accumulation. Some missing data occurred due to the corona virus disease 2019 pandemic. Although we used mixed-effects models and sensitivity analyses to reduce this impact, the results should still be interpreted with caution. Additionally, the multiple imputation models did not adjust for all variables that differed between completers and non-completers, which is a minor limitation. Thirdly, the applicability of the MASLD definition in our non-obese cohort should be considered. While we adopted the latest 2023 criteria and excluded significant alcohol consumption and viral hepatitis, it remains possible that hepatic steatosis in some individuals could be attributed to other etiologies not fully captured, such as drug-induced liver injury or rapid weight changes. Although our clinical records did not indicate such histories among participants, this potential confounding should be acknowledged. Fourthly, the cohort in this study was predominantly male (95%), which reflects the demographic reality of the HIV outpatient population at our center, where men who have sex with men constitute the majority. This severely limits the generalizability of our findings to female PLWH. We explicitly state that the conclusions of this study are primarily applicable to male PLWH, and their applicability to females requires further validation in studies with balanced gender representation. Fifthly, regarding the mechanistic explanation for the association between a higher CD4/CD8 ratio and increased MASLD risk, our discussion proposes hypotheses involving inflammatory dysregulation vs immune deficiency with metabolic disturbances. However, we acknowledge that this remains speculative due to the lack of direct measurements of systemic inflammatory markers (e.g., hs-CRP, IL-6) or detailed immune cell phenotyping (e.g., exhausted T cells, monocyte activation markers) in this study. This constitutes a key gap, and future studies should incorporate these biomarkers to verify the proposed inflammatory pathways. Finally, the liver fat content was assessed using VCTE rather than the gold standard for liver biopsy. While VCTE is a validated non-invasive tool, it does not assess MASH. However, all VCTE examinations were performed by experienced operators under standardized conditions, and the steatosis diagnostic threshold set by the iLivTouch device (UAP ≥ 244 dB/m) has been validated in Chinese population studies (area under the curve = 0.88)[54], which enhances the reliability of the assessment.

CONCLUSION

In this single-center study of the Chinese HIV population, MASLD presented a heavy burden, and there existed a high-risk subgroup of non-obese MASLD. The exploratory association between a higher CD4/CD8 ratio and increased MASLD risk in non-obese individuals highlights a potential dissociation between numerical immune recovery and metabolic health, warranting further investigation into the underlying immunometabolic dysregulation. Over a 12-month period, INSTIs-based regimens were not associated with hepatic steatosis progression. Our findings emphasize the importance of screening for MASLD in all PLWH, regardless of BMI, and call for integrated care models that address both immunological and metabolic health.