Abstract
Background/Aim: Prurigo nodularis (PN) is a chronic inflammatory skin disorder characterized by intense pruritus and nodular lesions. Emerging evidence suggests PN may be associated with systemic conditions, including liver diseases. This study aimed to investigate the relationship between PN and non-alcoholic fatty liver disease (NAFLD) and liver fibrosis/cirrhosis.
Patients and Methods: This study was conducted using the TriNetX Research Network. Adults diagnosed with PN between 2005 and 2018 were compared with a propensity score-matched control group without PN. Patients with prior liver disease or neoplasms were excluded. The outcomes of interest were incident NAFLD, liver fibrosis, and cirrhosis, assessed using ICD-10-CM codes. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated and sensitivity and stratification analyses were conducted to evaluate the robustness of the findings.
Results: Among 28,390 PN patients and matched controls, PN was associated with an elevated risk of NAFLD (HR=1.27, 95% CI=1.17-1.38) and liver fibrosis/cirrhosis (HR=2.01, 95% CI=1.65-2.45) over a 15-year follow-up. Stratified analyses revealed higher risks in males and younger patients (18–64 years). Sensitivity analyses confirmed consistent findings across various definitions, follow-up durations, and active comparators.
Conclusion: PN is associated with an increased risk of NAFLD, liver fibrosis, and cirrhosis. These findings highlight the need for monitoring and proactive management of liver health in PN patients. Further research is warranted to elucidate the mechanisms underlying this association and explore potential therapeutic strategies.
- Prurigo nodularis
- non-alcoholic fatty liver disease
- liver cirrhosis
- cohort
- epidemiology
- electronic medical records
Introduction
Prurigo nodularis (PN) is a persistent inflammatory skin condition marked by the presence of numerous itchy nodules, primarily located on the symmetrical extensor areas of the limbs and the trunk. The prevalence of PN was estimated at 72 per 100,000 individuals of adults aged 18 to 64 years in the United States. The mean and median ages of PN patients were 50.9 and 54 years, with a slightly higher tendency in females and darker skin color (1, 2). Intense pruritus is one of the typical features in PN, with other potential sensations including burning or stinging pain. The considerable physical and psychological burden can adversely affect patients’ quality of life (3).
Although the exact mechanisms of PN remain unclear, various potential etiologic factors have been identified, including dermatological, metabolic, infectious, psychiatric, neurologic, and medical causes (3). Patients with PN exhibited heightened expression of calcitonin-gene-related peptide (CGRP) and an increased density of nerve fibers containing substance P (SP). Besides, infiltration of eosinophils, T lymphocytes, and mast cells with proinflammatory cytokines played a crucial role in pruritus and erythema of PN lesions (4).
Non-alcoholic fatty liver disease (NAFLD) has recently been considered a multisystem disease for increasing evidence of the extra-hepatic organs effect (5). Incidence rates of NAFLD are estimated to be 20/10,000 person-years, and its prevalence is approximately 30~40% in males and 15~20% in females (6, 7). Several chronic comorbidities are associated with NAFLD, involving type 2 diabetes mellitus, cardiovascular disease, chronic kidney disease, sleep apnea, malignancy, osteoporosis, endocrinopathies, and psoriasis (8, 9).
The development of NAFLD involves several mechanisms. NAFLD pathogenesis begins with liver lipid accumulation, which can advance to liver fibrosis, cirrhosis, and hepatocellular carcinoma. Poor nutrition and genetic predisposition increase hepatic lipid storage, while inflammation of adipose tissue and visceral fat promotes insulin resistance and stellate cell activation, leading to extracellular matrix buildup. Additionally, dietary imbalances may cause dysbiosis, disrupting short-chain fatty acid pathways and fostering a pro-inflammatory liver state (10). The significant increase in cytokines, aggregation of lipid products, and collagen deposits of stellate cells result in liver damage, which progresses to NAFLD and other chronic liver diseases.
Studies exploring comorbidities of PN are still limited. A multi-institution study in Korea indicated comorbidities such as chronic kidney disease, dyslipidemia, type 2 diabetes mellitus, arterial hypertension, autoimmune thyroiditis, non-Hodgkin’s lymphoma, and atopic dermatitis (11). Another study suggested PN is highly related to liver dysfunction, providing genetic evidence for a potential skin-liver axis. However, the study is limited by its smaller subjects, single recruitment source, and cross-sectional study design (12). Consequently, our study aims to investigate the mechanisms of PN, NAFLD, liver fibrosis and cirrhosis, thereby enhancing the understanding of the PN-chronic liver diseases connection and providing valuable evidence for clinical diagnosis and treatment.
Patients and Methods
This population-based retrospective cohort study utilized data from the TriNetX Research Network, a global repository of electronic medical records integrating clinical data from over 220 healthcare organizations (HCOs) across 17 countries, which has been widely used in real-world evidence studies to evaluate healthcare outcomes (13-16). Specifically, we accessed a subset of the Global Collaboration Network, which included data from 131 HCOs and over 150 million patients. This dataset encompasses comprehensive clinical information, including diagnoses, procedures, and pharmacological interventions. Disease diagnoses were identified using ICD-10-CM codes, pharmacological treatments were classified using Anatomical Therapeutic Chemistry (ATC) codes, and procedural interventions were defined via Current Procedural Terminology (CPT) codes. Table I provides a detailed list of the codes used in the study for disease identification and outcomes.
Utilized proxy codes.
Participants were enrolled during the study period from January 1, 2005, to December 31, 2018. The PN cohort included individuals with a confirmed diagnosis of PN prior to the index date, while the control cohort comprised individuals without a prior PN diagnosis who had undergone routine medical examinations. To mitigate potential bias, participants under 18 years of age or those with a documented history of neoplasms or liver disease (including NAFLD, liver cirrhosis or fibrosis, autoimmune hepatitis, hepatitis B, or hepatitis C) before the index date were excluded from the study. This approach minimized confounding factors associated with underlying liver pathology or frailty.
The primary exposure was the presence of PN, as determined by ICD-10-CM codes, while the control group included patients with no prior diagnosis of PN. The primary outcome was defined as the onset of NAFLD and secondary outcome was defined as incident liver cirrhosis or fibrosis, with events occurring within the first three months post-index date excluded to reduce reverse causation. The follow-up period extended up to 15 years from the index date, with patients observed for liver disease outcomes during this timeframe.
Propensity score matching (PSM) was utilized to minimize potential confounding by ensuring that key covariates were balanced between the PN and control groups in a 1:1 ratio. Covariates considered in the matching process included demographic characteristics (age, sex, race), socioeconomic status, lifestyle factors, comorbidities, medical utilization status, body mass index (BMI), and laboratory data. The standardized mean difference (SMD) was used to evaluate balance between the matched groups, with SMD values exceeding 0.1 considered indicative of significant imbalance. Following PSM, a total of 28,390 PN patients and 28,390 matched controls were included in the analysis (Figure 1).
Patient selection process. In the current study, chronic liver diseases include non-alcoholic fatty liver diseases, liver cirrhosis and fibrosis, alcohol related disorders, autoimmune hepatitis and viral hepatitis.
All statistical analyses were conducted using the proprietary analytical platform of the TriNetX Research Network’s. Hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) for the development of liver disease were evaluated via the Cox proportional hazards model. The models were adjusted for potential confounders identified through PSM. Subgroup analyses were also performed to evaluate stratified risks based on age and sex. Several sensitivity analyses were conducted to further validate the findings of the study. These included varying the washout period, exploring different claims-based algorithms, adjusting follow-up duration, and testing alternative comparators. Detailed description of sensitivity analyses was reported in Table II. These analyses were performed to account for potential biases and to ensure the robustness of the results across different methodological approaches. This study was approved by the Institutional Review Board of Chung Shan Medical University Hospital (CS1-25002).
Description of all applied sensitivity analysis models.
Results
Several confounders involving age, sex, race, lifestyle, comorbidities, medical utilization status, and laboratory data were significantly different between the PN cohorts and controls. Before matching, PN patients were prone to have comorbidities such as hypertension, hyperlipidemia, diabetes mellitus, and chronic kidney disease. Furthermore, higher alanine aminotransferase (ALT), total bilirubin, C-reactive protein (CRP), and BMI were observed in PN groups than in control groups. After matching, the difference between the two groups was presented as insignificant. PN patients had a mean age of 47.0 and were mostly female (61.4%) and white (51.0%) (Table III).
Baseline characteristics of study subjects (before and after propensity score matching).
With a follow-up of 15 years since the index date and the wash-out period of 3 months, NAFLD risk was significantly higher in PN patients than in controls (HR=1.27, 95% CI=1.17-1.38). Similarly, PN patients were more likely to have liver cirrhosis and fibrosis (HR=2.01, 95% CI=1.65-2.45). In sensitivity analyses, we applied various definitions of PN, follow-up time, wash-out period, matching covariates, and active comparators to validate the findings. Across these analyses, a consistent trend was observed (Figure 2, Figure 3). Notably, compared to patients with psoriasis and alopecia areata (AA), patients with PN were associated with an elevated risk of NAFLD, with HRs of 1.11 (95% CI=1.03-1.21) and 1.28 (95% CI=1.15-1.42), respectively (Figure 2).
Forest plot of NAFLD risk in prurigo nodularis patients. The crude model represents the analysis without propensity score matching. In Matching model 1, covariates include age at index, sex, race, socioeconomic status and substance abuse status. In Matching model 2, covariates include age at index, sex, race, comorbidity status (including diabetes mellitus, hypertension, hyperlipidemia, chronic kidney disease). In sensitivity analyses, Algorithm 1 included patients with more than 2 PN diagnoses in the PN group; Algorithm 2 included in the PN group patients diagnosed with PN with more than 2 visit records and at least one impatient visit; Algorithm 3 included in the PN group patients diagnosed with PN with more than 2 visit records and a record of systemic corticosteroid, pimecrolimus or tacrolimus. Detailed information of sensitivity analysis was presented in Table II. PN, Prurigo nodularis; NAFLD, non-alcoholic fatty liver diseases; PSO, psoriasis; AA, alopecia areata; CI, confidence interval.
Forest plot of liver cirrhosis and fibrosis risk in prurigo nodularis patients. The crude model represents the analysis without propensity score matching. In Matching model 1, covariates include age at index, sex, race, socioeconomic status and substance abuse status. In Matching model 2, covariates include age at index, sex, race, comorbidity status (including diabetes mellitus, hypertension, hyperlipidemia, chronic kidney disease). In sensitivity analyses, Algorithm 1 included patients with more than 2 PN diagnoses in the PN group; Algorithm 2 included in the PN group patients diagnosed with PN with more than 2 visit records and at least one impatient visit; Algorithm 3 included in the PN group patients diagnosed with PN with more than 2 visit records and a record of systemic corticosteroid, pimecrolimus or tacrolimus. Detailed information of sensitivity analysis was presented in Table II. PN, Prurigo nodularis; PSO, psoriasis; AA, alopecia areata; CI, confidence interval. Crude model: analysis did not undergo propensity score matching.
Stratified analyses by age and sex demonstrated that among PN patients, males had a 36% higher risk of NAFLD compared to male controls, while females had a 16% higher risk of NAFLD compared to female controls. When stratified by age, both young and elderly PN patients had higher risk of NAFLD (41% and 36%, respectively) compared to PN-free controls in the respective age subgroup. By race, White (HR=1.32, 95% CI=1.18-1.48), Black (HR=1.41, 95% CI=1.12-1.77), and Asian (HR=1.66, 95% CI=1.10-2.49) patients with PN had elevated risk of NAFLD while comparing with PN-free controls in the respective racial subgroup. Similar results were observed for liver cirrhosis and fibrosis. The risk of liver cirrhosis and fibrosis was higher in both males (HR=2.69, 95% CI=1.95-3.70) and females (HR=1.64, 95% CI=1.26-2.13) with PN while comparing with PN-free controls. Likewise, both young (18-64 years) and elderly (≥65 years) PN patients had higher risk of liver cirrhosis and fibrosis than controls, which was 141% and 99%, respectively. Significantly elevated risk of liver cirrhosis in PN patients was also observed in White (HR=1.49, 95% CI=1.14-1.94) and Black (HR=2.90, 95% CI=1.79-4.71) subgroups, while comparing with PN-free controls in the same racial groups. After excluding patients with atopic dermatitis, the association between PN and NAFLD remained (Table IV).
Stratification analysis of age and sex in the risk of non-alcoholic fatty liver disease (NAFLD) and liver cirrhosis and fibrosis in prurigo nodularis (PN) patients and control individuals.
Discussion
In this large, population-based cohort study, we demonstrated a higher risk of NAFLD, liver cirrhosis, and fibrosis in PN patients compared to PN-free individuals (control group). The accuracy and robustness of these findings were also confirmed by sensitivity and subgroup analyses. Although the exact pathogenesis of PN is not yet fully understood, there is evidence indicating that PN is related to neuroimmune dysregulation. PN patients exhibit dermal neuronal hyperplasia, while PN lesions are characterized by loss of epidermal pan-neuronal marker protein gene product (PGP) 9.5 and nerve growth factor (NGF) (17). This deficiency leads to an abnormal accumulation of PGP 9.5 immuno-reactive cells and eosinophils in the dermis, triggering inflammatory reactions with nerve fibers. Increased SP nerve fibers in this region amplify the local immune response contributing to allergic reaction, with eosinophils further exacerbating this process. SP, CGRP, IL-4, and vasoactive intestinal polypeptide (VIP) released by eosinophils are responsible for PN lesions (17, 18). CGRP triggers the Th2-related pathway, which increases IL-31 expression by Th2 cells and causes local pruritus (19). In addition, activation of Th17- and Th22- related pathways contributes to keratinocyte hyperplasia, impaired epidermal differentiation, keratinocyte proliferation, and enhanced inflammatory responses, with IL-17 and IL-22 playing key roles in these processes in PN (20).
A cross-sectional study has provided genetic evidence for the PN-liver axis (12). Via GWAS, the study revealed seven genes linked to SNPs overlapping between PN and liver disease. Among those genes, androgen receptor (AR) gene, is known for activating the TGF-β1 pathway in the liver of hepatocellular carcinoma progression (12, 21). TGF-β1 is responsible for fibroblast proliferation and activation of hepatic stellate cells, which leads to the NAFLD progression with fibrosis (22). Others like ZEB2, EDIL3, and MACROD2 genes are involved in the epithelial-to-mesenchymal transition (EMT) in liver fibrosis and malignancy, and their dysregulations are also observed in local skin lesions (12). These findings suggest a genetic relationship between PN and chronic liver diseases. However, though TriNetX database was additionally utilized for validation in this study, the lack of sensitivity and stratification analysis massively limited the findings.
The involvement of Th17 and Th22 pathways is also evident in NAFLD. Th17 strengthens the adipose tissue inflammation (23). Th17-mediated adipose tissue inflammation contributes to NAFLD progression among hepatocytes, hepatic stellate cells (HSC), and Kupffer cells (24, 25). A recent study suggested a cluster of inflammatory hepatic CXCR3+ Th17 cells (ihTh17) that exacerbate NAFLD. These cells accumulate in the steatotic liver microenvironment, increasing chromatin accessibility, glycolytic output, and IL-17A production (26, 27). Conversely, the role of IL-22 is still being debated (28, 29). IL-22 mitigates palmitate-induced lipotoxicity in vitro through phosphoinositide 3-kinase (PI3K)/Akt-mediated suppression of JNK activity. However, in the presence of IL-17, this effect is not observed (23, 30).
Similar skin diseases, such as psoriasis and AA may help to elucidate the mechanisms linking PN and NAFLD. A bidirectional association between psoriasis and NAFLD has been mentioned in several studies (31). Elevated TNF-α in psoriasis patients stimulates keratinocyte proliferation, angiogenesis, and lipid accumulation in the liver. IL-1 and IL-6 promote keratinocyte proliferation, insulin resistance, and liver inflammation through mitogen-activated protein (MAP) and E-26 transformation specific-related gene (ERG) pathways (32). Some studies revealed that IL-22 mediates IL-23-induced dermal inflammation, leading to NAFLD progression by the Th17 pathway (23-25, 33). On the other hand, a recent study by Tehranchinia Z et al. found a higher frequency of fatty liver in AA patients, although the difference did not reach statistical significance (34). Decreased PON1 levels in AA patients may decline the potential prevention effect of lipid peroxidation, thereby promoting the lipid storage of the liver (35, 36). These findings imply that overlapping immune-inflammatory and metabolic pathways, involving cytokine dysregulation and impaired antioxidant defense, may account for the increased occurrence of chronic liver diseases such as NAFLD in individuals with psoriasis and AA.
In PN patients, both IL-17 and IL-22 increase in the skin lesions area, resulting in keratinocyte proliferation and exacerbation of inflammatory responses (20). In addition, IL-22 activation through the Th17 pathway may further exacerbate inflammation, resembling the pathogenic mechanisms observed in the psoriasis–NAFLD association, although the precise role of IL-22 remains under debate (12, 23, 27). Furthermore, the study by Marani et al. provides additional evidence supporting a PN-liver axis, showing that genetic factors in PN patients may influence the TGF-β1 signaling pathway and epithelial-to-mesenchymal transition, thereby contributing to chronic liver disease (12). Together, these findings suggest that PN may share common immunological and genetic mechanisms with other inflammatory skin disorders that predispose patients to chronic liver diseases.
Some data in our study warrants further consideration. In our study, we observed a numerically higher risk of liver cirrhosis and fibrosis than general NAFLD among PN patients. This could be attributed to the chronic inflammatory microenvironment of lipid accumulation in PN patients. Considering the fibroblast proliferation and activation of hepatic stellate cells, the exacerbating effect of this mechanism may be more significant in advanced liver diseases than in early NAFLD, resulting in a stronger association between PN and liver fibrosis and cirrhosis than in general NAFLD (12, 22). Another noteworthy point is the greater risk of liver cirrhosis and fibrosis in male PN patients than in females. Owing to the higher prevalence of NAFLD in males than females, we suggested this observation may aligned with the PN patients, which is demonstrated in our study. More studies are needed to confirm our supposition.
Study limitations. First, despite our matching efforts, the study samples were relatively small and predominantly composed of White and Black individuals, which may affect the generalizability of our findings. Second, due to the observational design of the study, causal relationships between PN, NAFLD, liver cirrhosis, and fibrosis cannot be established. Third, the severity of PN, NAFLD, and advanced liver diseases could not be assessed due to limited available data. Therefore, we could not evaluate the risks associated with different PN stages. However, we conducted sensitivity analyses using different definitions of PN, and the results were consistent supporting the robustness of our findings.
In conclusion, we report a long-term, real-world association between PN and liver diseases, providing new evidence of this association to the growing body of skin-liver axis studies. These real-world findings may increase the awareness of liver comorbidities in PN patients and highlight novel insights for future research on skin-liver interactions.
Footnotes
Authors’ Contributions
All Authors participated in manuscript preparation and approved of the submitted version. Study conception and design: Chang HC, Lin CY, Lu HY, Chiu TM, Lo SW, Wu CL, Liao WC, Fang YJ, Gau SY. Data acquisition: Chang HC, Gau SY. Data analysis and demonstration: Gau SY, Fang YJ, Lu HY and Chang HC. Original draft preparation: Chang HC, Lin CY, Lu HY, Chiu TM, Lo SW, Wu CL, Liao WC, Fang YJ, Gau SY.
Data Availability
Data in this study were retrieved from TriNetX Research Network. All data available in the database were administrated by the TriNetX platform. Detailed information can be retrieved at the official website of the research network (https://trinetx.com).
Conflicts of Interest
The Authors have no conflicts of interest to declare.
Funding
This study was partially funded by Chung Shan Medical University Hospital (CSH-2025-C-007).
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model (ChatGPT-4o, OpenAI) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.
- Received September 19, 2025.
- Revision received November 1, 2025.
- Accepted November 11, 2025.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
Jump to section
Related Articles
Cited By...
- No citing articles found.