Plasma Cholesterol and IL-18 Flag Persistent Fatigue in HIV Patients Whose Immunity Never Recovers

For most people living with HIV today, antiretroviral therapy (ART) does exactly what it promises: it suppresses the virus to undetectable levels and allows the immune system to rebuild itself. Yet in roughly 15 to 30 percent of treated patients, particularly those who began treatment after years of untreated infection, CD4+ T-cell counts never climb back to healthy territory. These individuals are known as immunological non-responders, or INRs, and they face a substantially higher risk of opportunistic infections, long-term complications, and death compared with patients whose immunity recovers fully. Now, a team of researchers in China has turned its attention to one of the most common and debilitating complaints in this population: fatigue.

Fatigue is far from a trivial symptom. International studies suggest that between 33 and 88 percent of people living with HIV experience it, compared with only 7 to 42 percent of the general population. It is characterized by an overwhelming sense of tiredness, low energy, and weakness that worsens with activity and undermines normal functioning. In people with HIV, fatigue can feed depression, erode medication adherence, encourage drug resistance, and degrade physical, psychological, and social well-being. Despite this burden, the biological roots of HIV-associated fatigue remain murky, and conflicting reports exist on whether fatigue tracks with CD4 counts or viral load, at least partly because fatigue is usually measured subjectively. Identifying objective biomarkers has therefore become a pressing clinical goal.

A new study published in Frontiers in Cellular and Infection Microbiology takes a step toward that goal by combining untargeted metabolomics with high-throughput cytokine profiling. The researchers, led by Baixue Li and Quansheng Feng of Chengdu University of Traditional Chinese Medicine, recruited 35 HIV immunological non-responders from disease control centers in Liangshan Prefecture, Sichuan Province, between January 2022 and December 2023. All participants had been on ART for more than three years, maintained viral suppression below 50 copies per milliliter, and had CD4+ T-cell counts that remained at or below 350 cells per microliter throughout long-term follow-up. The cohort was divided into 18 fatigued and 17 non-fatigued individuals, matched for baseline CD4 count, viral load, age, and sex, with ten HIV-negative healthy controls added for comparison. People co-infected with hepatitis B or C were excluded.

Fatigue status was assessed with a modified version of the Functional Assessment of Chronic Illness Therapy-Fatigue Scale, a 15-item questionnaire scored from 0 to 60, with a threshold of 30 points separating the fatigued from the non-fatigued group. The fatigued participants averaged a score of 37.40, while the non-fatigued averaged 11.35. Beyond the questionnaire, the team collected a rich clinical picture using the Medical Outcomes Study HIV health survey, symptom inventories, routine blood work, liver and kidney function tests, and detailed immunophenotyping. The results painted a consistent picture: fatigued INRs reported significantly poorer quality of life, clustered symptoms including loss of appetite, fever, sore throat, recurrent diarrhea, and oral ulcers, and showed notable differences in red blood cell indices and the liver enzyme gamma-glutamyl transpeptidase. Importantly, CD4 counts at the time of sampling did not differ significantly between the two groups, meaning the molecular signatures the team went on to find could not simply be explained by current immune cell numbers.

The metabolomic arm of the study relied on ultra-performance liquid chromatography coupled to high-resolution mass spectrometry. Plasma samples were processed with rigorous quality control, and more than 8,000 metabolite features showed relative standard deviations below 20 percent, indicating excellent analytical stability. Multivariate statistics, including principal component analysis and partial least squares discriminant analysis, revealed clear separation between the plasma metabolite profiles of fatigued and non-fatigued INRs. To guard against mistaking drug or dietary compounds for biology, the team explicitly excluded exogenous substances such as efavirenz metabolites, caffeine derivatives, piperine, and artemisinin from all fatigue-related interpretation. This caution proved important: efavirenz, for example, was dramatically elevated in both INR groups relative to healthy controls, a signal reflecting antiretroviral exposure rather than fatigue itself.

When the two INR groups were compared directly, five metabolites stood out. The fatigued cohort showed significant up-regulation of 4-ketopimelic acid, cholesterol, and 1-hexadecyl-sn-glycero-3-phosphocholine, alongside down-regulation of DL-arginine and D-pipecolic acid, all at p-values below 0.05. Cholesterol was among the most striking findings, with a log2 fold change of 1.84. Cis-9-palmitoleic acid, a monounsaturated fatty acid, was also elevated. Pathway enrichment analysis of these altered metabolites pointed toward efferocytosis, steroid hormone synthesis, fat digestion and absorption, cholesterol metabolism, and atherosclerosis. The researchers also flagged enrichment of the necroptosis pathway, a form of programmed inflammatory cell death, though they were careful to note that this was a pathway-level inference from enrichment analysis rather than a direct experimental measurement, and it remains a hypothesis requiring validation in the laboratory.

The cytokine analysis added an immunological dimension to the metabolic story. Using a Luminex multiplex assay, the team quantified 42 inflammatory proteins in each plasma sample. Three cytokines differed significantly between fatigued and non-fatigued INRs: interleukin-18, interferon-induced protein 10 (IP-10, also known as CXCL10), and macrophage migration inhibitory factor (MIF). MIF, a multifunctional pro-inflammatory cytokine, was elevated in the fatigued group, suggesting persistent inflammatory activity. The pattern for IL-18 and IP-10 was more subtle and arguably more interesting. Both were lower in fatigued patients than in their non-fatigued counterparts, yet both remained significantly higher than in healthy controls. The authors interpret this intermediate state as a possible adaptive down-regulation of the immune system in response to chronic stimulation, an incomplete suppression that may still disrupt energy metabolism and neurotransmitter balance while leaving patients in an immune-activated but functionally impaired condition.

Integrating the two data layers, the researchers constructed protein-metabolite interaction networks and cross-referenced KEGG pathway enrichments from the metabolome and the cytokinome. Three pathways emerged in the overlap: necroptosis, lipid metabolism and atherosclerosis, and cancer-related signaling. The atherosclerosis connection is clinically provocative, because previous work has shown that INRs carry a two- to three-fold higher risk of cardiovascular disease than patients with normal immune reconstitution, and their arterial plaques tend to be more unstable. Within the network, cytokines such as IL-18, IP-10, IL-1 alpha, and CXCL8 interacted strongly with metabolism-related proteins, and cholesterol and sphingomyelin emerged as central metabolic nodes. The authors propose a potential vicious cycle linking dysregulated lipid metabolism, immune cell death, and chronic low-grade inflammation as a mechanism sustaining fatigue, though they emphasize that this model is hypothesis-generating rather than experimentally proven.

Two candidates rose to the top as potential biomarkers: cholesterol and IL-18. However, the team is candid about the limits of this claim. IL-18 concentration showed no statistically significant association with cholesterol levels in any of the three groups, with only a non-significant positive trend among fatigued INRs, and the combined discriminatory performance of the two markers, measured by metrics such as area under the curve, sensitivity, and specificity, was never formally validated. The study itself is small, single-center, and cross-sectional, which precludes causal inference. Residual confounding by ART regimen, particularly efavirenz exposure, cannot be fully excluded with 17 to 18 participants per group, and the fatigued group had lower CD4 counts at the start of therapy, leaving pre-treatment disease severity as a competing explanation. The modified fatigue questionnaire also lacks formal psychometric validation.

Even with these caveats, the study opens a genuinely new window onto a neglected problem. It suggests that fatigue in immunological non-responders is not merely a psychological complaint but a measurable biological state, marked by lipid disturbances, red blood cell changes, and an immune system caught between overactivation and exhaustion. The convergence with findings from long COVID and chronic fatigue syndrome research, where similar cytokine patterns have been reported, hints at shared mechanisms of inflammation-driven fatigue across conditions. The authors call for large, multicenter, longitudinal studies that incorporate transcriptomics, proteomics, and gut microbiome data to test whether the cholesterol-IL-18 axis can eventually support risk stratification and personalized management. For now, the message is that the blood of a fatigued INR tells a molecular story that questionnaires alone cannot capture, and that story may one day guide treatments aimed at restoring both immune balance and energy.

Subject of Research: Plasma metabolic and inflammatory biomarkers of fatigue in HIV immunological non-responders on long-term antiretroviral therapy

Article Title: Integrated multi-omics reveals plasma biomarkers of fatigue in HIV immunological non-responders

Article References: Li, B., Chen, X., Liu, J., Wen, L., Shen, Z., Zhou, H., Wu, P., & Feng, Q. (2026). Integrated multi-omics reveals plasma biomarkers of fatigue in HIV immunological non-responders. Frontiers in Cellular and Infection Microbiology, 16, Article 1932594. https://doi.org/10.3389/fcimb.2026.1932594

Image Credits: AI Generated

DOI: 10.3389/fcimb.2026.1932594

Keywords: HIV, immunological non-responders, fatigue, metabolomics, cytokines, cholesterol, interleukin-18, MIF, IP-10, necroptosis, lipid metabolism, antiretroviral therapy