Skip to main content
Download PDF
  • Research article
  • Open access
  • Published:

The plasma metabolome of juvenile idiopathic arthritis varies according to subtype and underlying inflammatory status

Pediatric Rheumatology volume 22, Article number: 113 (2024) Cite this article

Abstract

Background

Juvenile idiopathic arthritis (JIA) is challenging to classify and effectively monitor due to the lack of disease- and subtype-specific biomarkers. A robust molecular signature that tracks with specific JIA features over time is urgently required, and targeted plasma metabolomics may reveal such a signature. The primary aim of this study was to characterise the differences in the plasma metabolome between JIA patients and non-JIA controls and identify specific markers of JIA subtype. We also assessed the extent to which these signatures are due to underlying inflammation as assessed by glycoprotein acetyls (GlycA) and high-sensitivity C-Reactive Protein (hsCRP) levels.

Methods

Targeted nuclear magnetic resonance (NMR) metabolomic profiles of plasma of 72 children with JIA and 18 controls were assessed cross-sectionally. Associations between 71 metabolomic biomarkers and JIA, JIA subtype, disease activity status, and inflammation markers (GlycA and hsCRP) were assessed using multivariable linear regression models.

Results

JIA was associated with higher GlycA (mean difference = 0.93 standard deviations, 95% confidence interval = [0.370, 1.494], Padj = 0.039) and docosahexaenoic acid (1.06, [0.51, 1.60], Padj = 0.021), and lower acetate (-0.92, [-1.43, -0.41], Padj = 0.024) relative to controls. This variation was largely driven by systemic JIA (sJIA), with 24 of 71 total biomarkers significantly different (Padj <0.05) relative to controls. There were no specific differences identified in oligoarticular (oJIA) or polyarticular (rheumatoid factor positive or negative) JIA relative to controls. Despite being generally highly correlated with hsCRP (r > 0.70), GlycA, but not hsCRP, was positively associated with active disease in sJIA (0.22, [-0.40, -0.04], Padj = 0.018), and 6 of 24 sJIA-associated markers were associated with GlycA levels. Only 1 sJIA-associated biomarker, histidine, was associated with hsCRP levels.

Conclusion

Differences in the plasma NMR metabolomic profiles are apparent in children with sJIA, but not other JIA subtypes, relative to non-JIA controls. These findings suggest a potential utility for classifying and monitoring JIA through metabolomic profiling, with chronic inflammation, measured by GlycA, potentially playing a role in at least some of these metabolomic differences.

Key messages

• Children with JIA showed limited metabolomic differences in plasma compared to non-JIA controls.

• Subtype-associated analysis identified exclusive metabolomic variation in sJIA.

• Some metabolomic differences in sJIA may be driven by chronic inflammation.

• GlycA may better reflect disease activity and metabolomic differences in sJIA than other inflammatory markers such as hsCRP.

Introduction

Juvenile idiopathic arthritis (JIA) is a heterogenous group of chronic rheumatic diseases in children, categorised into subtypes such as oligoarticular (oJIA, 27–60%), polyarticular (pJIA, 2–30%, Rheumatoid factor-negative more common than Rheumatoid factor-positive) and systemic JIA (sJIA, 10–20%), based on numbers of affected joints and presence of Rheumatoid factor (RF) and systemic symptoms [1,2,3,4]. The response rates of drug treatment between JIA subtypes vary, and there is a lack of specific biomarkers to distinguish JIA subtypes and to monitor disease activity. A key feature of JIA is inflammation, with certain subtypes distinguished by whether the inflammation is systemic affecting multiple organs or localized to the joints and surrounding tissues [1, 2, 5,6,7]. The absence of distinctive JIA-associated molecules hinders effective disease classification and monitoring [4, 6]. As such, approaches for profiling the molecular signatures of JIA and its subtypes are becoming acknowledged as a potential way to achieve more personalised treatment, and therefore improve clinical outcomes [4,5,6].

The plasma metabolome is a functional readout of metabolic activities in the body. Alterations in the plasma metabolome have previously been associated with chronic inflammatory conditions in childhood such as Inflammatory Bowel Disease (IBD) [8] and juvenile Systemic Lupus Erythematosus (jSLE) [9]. The metabolomic differences in adults with RA and children with jSLE relative to controls have also been characterised, some of which may promote inflammation [9, 10]. While several studies have begun to characterise the metabolome in JIA [11,12,13], these have yet to identify subtype-differentiating signatures or markers predictive of disease course and long-term outcome.

Untargeted metabolomics measures a wide spectrum of metabolites in a sample without prior knowledge of their identity [14]. However, due to a wide variety of metabolites, including uncharacterised molecules, it can require extensive validation to confirm their identities [14, 15]. Also, it is a relative quantification of metabolites, which could be less suitable for precise measurements for JIA subtype categorisation and monitoring [14]. In contrast, targeted metabolomics focuses on quantifying specific, predefined metabolites of interest within a sample, providing absolute concentrations [14]. Nuclear magnetic resonance (NMR)-based targeted metabolomics is increasingly recognised as having utility in identifying disease biomarkers across a range of conditions including adult RA [10, 16,17,18,19], and providing insight into disease processes [16]. Furthermore, several NMR-based metabolomic markers are now approved for clinical use in a range of settings [20].

Among the biomarkers identified through NMR-based approaches, glycoprotein acetyls (GlycA) is a promising marker of cumulative inflammation, based on glycosylation of several acute-phase proteins across multiple inflammatory pathways [21,22,23]. It is more stable over time and thought to better reflect chronic inflammation than conventional inflammatory markers such as high sensitivity C-reactive protein (hsCRP) or erythrocyte sedimentation rate (ESR) [21, 22, 24]. Studies in adult RA have shown that GlycA is closely associated with disease activity as well as hsCRP and ESR, which are widely used markers for monitoring disease activity in RA [25, 26].

Here, we characterised NMR metabolomic biomarkers in plasma of children with four JIA subtypes, oJIA, pJIA(RF-), pJIA(RF+), and sJIA, with the aim of characterising molecular signatures of JIA generally, and those associated with specific JIA subtypes. We also investigated the relationship between disease activity status, inflammation measured by GlycA and hsCRP, and these metabolomic differences.

Methods

Participants

A total of 72 children diagnosed with JIA and 18 hospital controls were recruited between 2013 and 2017 as part of the CLARITY (ChiLdhood Arthritis Risk factor Identification sTudY) cohort and biobank (Table 1) [27]. CLARITY is a cross-sectional biobank of JIA and healthy control children, which aims to investigate both genetic and environmental risk factors that impact on JIA risk. All CLARITY protocols were approved by the Human Research Ethics Committee of the Royal Children’s Hospital, Melbourne, Australia (HREC no. 27127Q) [27]. All participants provided informed consent [27]. All JIA cases (≤ 18 years) were recruited from the Royal Children’s Hospital (RCH) and diagnosed with JIA by a paediatric rheumatologist by ILAR classification [3, 27]. Active disease in cases was defined as the presence of at least one active joint or systemic features and inactive disease was assessed using the Wallace criteria as no active disease for at least 6 months on treatment or 12 months off treatment [28]. Controls (≤ 9 years) were recruited from the Day Surgery Unit and consisted of children undergoing minor surgery unrelated to immune-mediated inflammatory conditions.

Table 1 Baseline demographics and clinical characteristics of patients (n = 90)a
Full size table

NMR targeted metabolomics

We used a high-throughput proton NMR metabolomics platform (Nightingale Health, Helsinki, Finland) to quantify metabolites within plasma (https://research.nightingalehealth.com/). The platform generates 249 total measures including 39 clinically validated blood biomarkers for lipid and glucose metabolism, ketone bodies, amino acids and a marker of chronic inflammation (GlycA) [20, 21]. Plasma samples (EDTA tube collection) from the 90 participants were shipped on dry ice to Nightingale Health (Helsinki, Finland) for metabolomic quantification.

Preparation of data

Due to the high degree of correlation between markers (natural unit Spearman’s rank correlation) (Supplementary Figure S1), 249 initial measures (Supplementary Data D1) were filtered down to 71 primary measures prior to statistical analysis. Consistent with previous studies with this metabolomics platform [29,30,31], lipoprotein subclasses, relative lipoprotein lipid concentrations groups, and ratio measures expressed as a percentage were removed due to their high correlation with other measures. Data for 3-hydroxybutyrate were also removed due to the high percentage (55%) of missing data. Since the fasting status of participants was unknown, fasting-sensitive measures including glucose and lactate were also removed prior to statistical analysis.

Statistical analysis

All analyses were conducted using R (version 3.6.3). Metabolomic measures were natural log-transformed and scaled to a standard distribution, with effect sizes reported in standard deviation units. Linear regression models adjusted for age and sex were used to identify (i) JIA-associated biomarkers independent of subtypes, (ii) subtype-associated biomarkers, and (iii) inflammation (hsCRP and GlycA) associated biomarkers. Discovery regression models from the ggforestplot package were used for analysis. For JIA-associated biomarkers, disease status (control group (0) and JIA group (1) was used as a predictor. Estimates were reported as an adjusted mean difference (AMD) and the associated 95% confidence interval (CI). To control the false discovery rate (FDR), we applied the Benjamini-Hochberg (BH) procedure [32] with padj values below 0.05 considered statistically significant. This corresponds to a stringency level of 5% for the false discover rate (FDR). Volcano and forest plots were generated using ggplot2. Biomarkers of interest were visualised using box-and-whisker plots. In secondary analyses, we considered models stratifying participants by disease activity status (active disease only and inactive disease only models). For subtype-associated biomarkers, each subtype [sJIA, oJIA, pJIA(RF+) and pJIA(RF-)] was independently compared to the control group using the same approach as above. For inflammation analyses, we first investigated whether there were differences in GlycA and hsCRP levels between disease status (inactive, active and unknown) in JIA cases and within each subtype using pairwise Student’s t-tests. We also calculated the Pearson’s correlation between GlycA and hsCRP for each subtype. We investigated inflammation-associated metabolomic differences using linear regression modelling with GlycA or hsCRP as the exposure and each of the 70 non-GlycA metabolomic measures as the outcome, adjusted for age and sex. Biomarkers of interest were visualised with scatter plots with linear fit lines.

Data availability

Mean metabolomic biomarker levels per group for 71 primary measures are provided in Supplementary data D2. Individual participant data from CLARITY cannot be made freely available online per the approved ethics for this study. Interested parties can access the data used in this study upon reasonable request, with approval by the CLARITY data custodians.

Results

Children with JIA show limited variation in the plasma metabolome relative to non-JIA controls

Linear regression was performed to identify differences in metabolomic profiles between the JIA and control groups, independently of subtype. Compared to controls, JIA status was associated with higher GlycA (mean difference = 0.93 standard deviations, 95% confidence interval = [0.370, 1.49], Padj =0.039) and DHA (1.06, [0.51, 1.6], Padj =0.021), and lower acetate (0.92, [-1.43, -0.41], Padj =0.024) (Fig. 1A & Supplementary Data D3). Acetate showed the clearest separation between the JIA and control groups with regards to interquartile range (Fig. 1B). Weak associations were found for several other biomarkers, particularly amino acids and high-density lipoprotein (HDL), however, these differences were not statistically significant following correction for multiple testing (Padj >0.05; Fig. 1C, Supplementary Data D3). In stratified models limited to active JIA patients only, the difference in GlycA was more marked (1.26, [0.69, 1.82], Padj = 0.004), while acetate levels remained significantly lower (-0.92, [-1.40, -0.43], Padj = 0.023), and DHA, while still elevated (0.87, [0.25, 1.48]), was no longer significant (Padj = 0.141 (Fig. 1D & Supplementary Data D4). In JIA cases with inactive disease (n = 13), only DHA was still higher relative to controls amongst JIA-associated markers (1.35, [0.52, 2.17], Padj = 0.041) (Fig. 1E & Supplementary Data D5). In addition, alanine, HDL particle size (HDL-size), phenylalanine, and HDL free cholesterol (HDL-FC) levels were also higher in inactive JIA, while acetoacetate and acetone were lower (Fig. 1E).

Fig. 1
figure 1

JIA-associated metabolomic biomarkers (A) Volcano plot for the estimated metabolomic difference between JIA group (n = 72) and controls (n = 18) from adjusted linear regressions model. A dotted line indicates padj value cut-off of 0.05. Biomarkers with padj<0.05 are labelled above the dotted line. Biomarkers with raw p < 0.05 are labelled below the dotted line. (B) Dot plots of JIA-associated markers between JIA and non-JIA controls. The statistical comparison was performed with the Student’s T-test. padj values are marked above dendrogram. All models were adjusted for age and sex. The distribution difference between JIA group and controls for all primary metabolomic measures are shown in Supplementary Figure S2. (C) Forest plot for the estimated metabolomic difference between JIA group and control (circle points) from adjusted linear regression models. Error bars are 95% confidence intervals. Closed points represent padj values < 0.05 (BH). (D) Volcano plot for the estimated metabolomic difference between active JIA group (n = 43) and controls (n = 18) from adjusted linear regressions model. A dotted line indicates padj value cut-off of 0.05. Biomarkers with padj<0.05 are labelled above the dotted line. Biomarkers with raw p < 0.05 are labelled below the dotted line. (E) Volcano plot for the estimated metabolomic difference between inactive JIA group (n = 13) and controls (n = 18) from adjusted linear regressions model. A dotted line indicates padj value cut-off of 0.05. Biomarkers with padj<0.05 are labelled above the dotted line. Biomarkers with raw p < 0.05 are labelled below the dotted line

Full size image

Subtype-specific analysis identifies widespread metabolomic variation in systemic JIA

Linear regression was also used to identify metabolome differences associated with each of four JIA subtypes [sJIA, oJIA, pJIA(RF+) and pJIA(RF-)] relative to controls. No significant markers were identified in oJIA, pJIA(RF+) or pJIA(RF-) following adjustment for multiple testing (Fig. 2A & Supplementary Data D6). In contrast, 24 of 71 total measures were different in the sJIA subtype compared to controls (Padj <0.05; Fig. 2A, Supplementary Figure S3). This included amino acids (lower histidine, higher tyrosine), lower acetate, several fatty acid measures including higher omega-3, omega-6, linoleic acid (LA) and DHA, higher inflammatory GlycA and several lipid and cholesterol measures. These were higher levels of apolipoprotein A1 (ApoA1), sphingomyelins and various high-density lipoprotein (HDL) measures, including HDL-FC, HDL-size, HDL-associated lipids (HDL-L), HDL cholesterol (HDL-C), HDL cholesterol esters (HDL-CE), HDL particles (HDL-P) and HDL phospholipids (HDL-PL).

Fig. 2
figure 2

Subtype-associated analysis between four JIA subtypes (colored by subtype) A. Volcano plot for the estimated metabolomic difference between each subtype and control from adjusted linear regressions model (sJIA: n = 22, oJIA: n = 18, pJIA(RF-): n = 18 and pJIA(RF+): n = 14) and non-JIA controls (n = 18). (A) dotted line indicates padj value cut-off of 0.05. Biomarkers with padj<0.05 in these subtype analyses are labelled. (B) Volcano plot for the estimated metabolomic difference between each subtype and control from adjusted linear regressions model in active subgroup (sJIA: n = 11, oJIA: n = 10, pJIA(RF-): n = 13 and pJIA(RF+): n = 9) and non-JIA controls (n = 18). A dotted line indicates padj value cut-off of 0.05. sJIA-associated biomarkers (n = 24) are labelled. (C) Forest plot for the estimated metabolomic difference between each subtype and control in 24 sJIA-associated biomarkers (circle points) from adjusted linear regression models (n = 90). Error bars are 95% confidence intervals. Closed points represent padj<0.05 (BH). All models were adjusted for participants age and sex. Metabolomic difference between each subtype and control for all primary metabolomic measures are shown in Supplementary Figure S3

Full size image

In models including only participants with active disease, GlycA and DHA remained significantly associated with active sJIA, while evidence for differences in several lipid markers, including HDL-P, Phosphatidylcholine (Phosphatidylc), omega-3 FA, and acetate was reduced (Fig. 2B & Supplementary Data D7). No evidence of metabolome changes in inactive sJIA patients was found (Supplementary Data D8). Whilst evidence for metabolomic differences was much weaker in non-sJIA subtypes, the estimated differences for a small number of measures showed a similar direction and/or magnitude of effect across all four subtypes, particularly acetate. (Fig. 2C).

GlycA is more strongly associated with sJIA disease activity than hsCRP

Across all subtypes, hsCRP levels tended to be higher in the active group compared to the inactive group (Fig. 3A-D). However, these differences were not statistically significant in sJIA (Fig. 3A). Additionally, only one patient for each of pJIA RF- and RF + had inactive status, preventing statistical comparison (Fig. 3B and C). Similarly, the trends of GlycA levels were higher in active patients than inactive (Fig. 3E and H). Notably, GlycA levels were significantly higher in the active sJIA disease group compared to inactive group (0.25 mmol/L higher, [-0.40, -0.04], Padj =0.018) (Fig. 3E), and were higher in active oJIA compared to inactive (0.50 mmol/L higher, [0.03, 0.36], Padj=0.027) (Fig. 3H). There was a strong positive correlation between hsCRP and GlycA across all JIA subtypes (sJIA: Pearson’s correlation (R) = 0.79 [Fig. 3I], pJIA(RF-): R = 0.79 [Fig. 3J], pJIA(RF+): R = 0.89 [Fig. 3K], and oJIA: R = 0.84 [Fig. 3L]).

Fig. 3
figure 3

Boxplots of hsCRP (A-D) and GlycA (E-H) by disease activity groups (inactive, active, no record) across different JIA subtypes. Padj value is corrected by BH and the significance between groups are marked by p-values. The scatter plots (I-L) show Pearson’s correlation between hsCRP and GlycA levels with a linear fitted line, demonstrating a strong positive correlation across all subtypes. R-value and Padj value are marked. (A, E, I: sJIA, B,F, J: pJIA(RF+), C, G,K: pJIA(RF-), D, H,L: oJIA.)

Full size image

Marker of chronic (GlycA), rather than acute (hsCRP), inflammation captures more variation in the metabolome of sJIA

Next, we wanted to investigate the extent to which differences observed in the metabolome between sJIA and control groups are associated with underlying inflammation. We assessed the relationship between GlycA with each of the other 70 metabolomic measures in all participants (n = 90). GlycA was associated with 13 biomarkers (Padj<0.05) (Fig. 4A, Supplementary Data D9). Of these, 6 biomarkers (polyunsaturated fatty acids (PUFA), sphingomyelins, low density lipoprotein triglycerides (LDL-TG), acetate, histidine, and omega-6 fatty acids) also overlapped with the 24 sJIA-associated biomarkers identified in subtype analysis (Fig. 2A). Correlation plots of GlycA and each of the 13 GlycA-associated biomarkers are shown in Fig. 4B. We also assessed the relationship between hsCRP and the same 70 biomarkers (Fig. 4C & Supplementary Data D10). Correlation plots of hsCRP and each of the 13 GlycA-associated biomarkers are shown in Fig. 4D. hsCRP was associated with 4 of the same biomarkers as GlycA, generally with a similar strength of correlation: albumin (GlycA R=-0.60; hsCRP R=-0.56), glutamine (GlycA R=-0.44; hsCRP R=-0.42), histidine (GlycA R=-0.58; hsCRP R=-0.50), and citrate (GlycA R=-0.53; hsCRP R=-0.42 (Fig. 4B and D).

Fig. 4
figure 4

Biomarkers that are associated with chronic inflammation (GlycA) (A) Volcano plot for the estimated association between level of 70 primary metabolomic measures and GlycA (n = 90) from adjusted linear regressions model. A dotted line indicates padj value cut-off of 0.05. Biomarkers with padj<0.05 are labelled. (B) A scatter plot for 13 GlycA-associated biomarkers (n = 90). The distribution in the whole cohort were described by scatter plot (black dots) and the associations between GlycA and potential biomarkers were illustrated with a linear fitted line (blue). (C) Volcano plot for the estimated association between level of 70 primary metabolomic measures and hsCRP (n = 90) from adjusted linear regressions model. A dotted line indicates padj value cut-off of 0.05. Biomarkers with padj<0.05 are labelled. (D) A scatter plot for 4 hsCRP-associated biomarkers (n = 90). The distribution in the whole cohort were described by scatter plot (black dots) and the associations between hsCRP and potential biomarkers were illustrated with a linear fitted line (blue)

Full size image

Discussion

In this study, we used targeted metabolomics to identify metabolic differences between JIA patients and non-JIA controls in plasma. The main benefits of the NMR-based targeted platform used here include clinically validated biomarkers, absolute quantification of metabolites, and high reproducibility owing to no batch effects [33]. Another strength of our approach was the analysis of four JIA subtypes. A previous untargeted liquid chromatography/mass spectrometry metabolomics study reported some differences in metabolites between JIA (n = 20) and controls (n = 20) but did not investigate JIA subtype-specific differences [12].

Of the 71 metabolic biomarkers investigated, three showed clear differences between children with JIA relative to controls, some of which differed by disease status at the time of blood collection. GlycA—a marker of cumulative inflammation—showed the largest difference in the active JIA group compared to controls. Previous studies have similarly reported that GlycA levels are elevated in individuals with SLE and RA and are associated with the severity of these diseases [34, 35]. We found elevated DHA levels in plasma in sJIA, in contrast to a previous JIA study that reported lower levels of polyunsaturated fatty acids in relation to active disease [36]. However, the difference in DHA between JIA cases and controls appeared specific to JIA cases with inactive disease, potentially reflecting medication use [37, 38]. Participants in our study received medications such as corticosteroids, NSAIDs and disease modifying anti rheumatoid drugs (DMARDs). In this cross-sectional study, the impact of medications on biomarkers is difficult to determine due to the non-random assignment of medication use, affected by factors including disease status, timing of diagnosis, and medication intolerance. Our findings suggest that active and inactive disease status in JIA is associated with distinct metabolic profile, with a chronic inflammatory marker, GlycA, prominent in active disease and lipid/amino acid metabolism differences becoming more prominent in remission.

In subtype-specific analysis, metabolic variation was most evident for sJIA, which has the most severe inflammation and is classified as an autoinflammatory disease (driven by innate immunity), as opposed to other subtypes that are primarily autoimmune (driven by adaptive immunity) in nature [1, 4, 39, 40]. ​While HDL has anti-inflammatory properties [41], our results showed that HDL-associated markers and ApoA1 were significantly increased in total sJIA and remained significant in active sJIA group. This contrasts with a meta-analysis that found no significant differences in HDL levels between JIA patients and healthy controls, though the included studies had a high degree of heterogeneity (I2 = 98.0%), indicating substantial variability between studies [42]. Furthermore, increased HDL levels in sJIA patients identified in our study might reflect specific subgroups not captured in the meta-analysis, such as patients with certain disease characteristics or those undergoing particular treatments. Unlike other childhood inflammatory conditions such as jSLE [43], we did not see strong evidence for LDL measures except for LDL-TG differing between sJIA and controls.

The absence of clear metabolic differences in plasma of children with oJIA and pJIA subtypes may reflect that sJIA is characterised by systemic inflammation affecting the circulation, in contrast to oJIA and pJIA, which show localized inflammation at the joints [44, 45]. It is possible that the metabolome of non-systemic JIA might show greater variation compared to controls within the affected joints than in plasma, however this requires investigation.

We found that hsCRP and GlycA were highly correlated across all JIA subtypes. However, only GlycA was significantly elevated in the active disease group compared to the inactive group in sJIA and oJIA subtypes. Further, we found that the level of chronic inflammation (as measured by GlycA) may explain some of the metabolic differences observed in sJIA relative to other JIA subtypes. GlycA was positively associated with 4 sJIA-associated biomarkers (LDL-TG, PUFA, omega-6 and sphingomyelins) and negatively associated with 2 (acetate and histidine). This negative association is consistent with previous evidence that acetate and histidine have anti-inflammatory properties and regulate the production of pro-inflammatory cytokines and the activation of inflammatory cells [46]. While histidine-rich glycoprotein (HRG) has been identified as a biomarker in adult RA [47], acetate has not previously been linked to arthritis. The weaker association of hsCRP and sJIA- associated metabolomic changes in JIA suggests the chronic aspects of inflammation may be more relevant to metabolic alterations in sJIA than acute inflammation.

Overall, our findings indicate that only some of the JIA-associated metabolic biomarkers were also inflammation-associated, suggesting that there are metabolic signatures in sJIA that are independent of inflammation. Further, as GlycA is a composite marker of several inflammatory processes in the body, such as elevated inflammatory cytokines and neutrophils, it is difficult to determine which of these cellular processes are contributing most to metabolomic differences [24, 48]. Future research is needed to characterize other soluble inflammatory mediators such as cytokines and inflammatory cells in JIA, and to explore opportunities to use metabolic and inflammatory biomarkers to better distinguish JIA subtypes, particularly sJIA, from other inflammatory conditions such as jSLE [49].

Strengths and limitations

This study investigated the association between the metabolome and JIA, as well as chronic inflammation, using a targeted NMR metabolomic approach. Our pediatric JIA cohort is large compared to previous metabolomic studies in JIA and included four distinct subtypes. Certain limitations merit consideration. NMR-based metabolomics is generally less sensitive and less comprehensive than mass-spectrometry based platforms, raising the possibility that untargeted metabolomics profiling may identify circulating biomarkers not measured here [50]. Our control cohort was younger on average than our JIA cohort, which may influence the results. However, age was included as a covariate in all statistical models. The limited number of rarer subtypes, such as only 14 cases of pJIA (RF+), affected the statistical power for relevant subtype analyses. Of note, the pJIA (RF+) group in this cohort were all female and older on average than the other subtypes, in line with the known higher prevalence of girls (8–9 females to 1 male) and age range (generally 10 to 13 years of age) [51]. Our results showed no significant changes in the plasma metabolome of pJIA(RF+) patients compared to controls. Hormonal and metabolic changes typically associated with puberty in this age range [52] may be involved, however this was not formally tested.

The lack of data for the medication history, including the duration of medication and symptom improvement after the treatment, means we were unable to account for these potential confounding factors. Additionally, the broad “active disease” category used in this study included patients with varying levels of disease severity. This wide range of clinical features, particularly for non-sJIA patients, may have diminished our ability to identify effects of disease activity on metabolomic associations. Future studies should investigate more detailed categories of disease activity to better understand the relationship between these biomarkers and disease severity.

Conclusion

The plasma NMR metabolome of systemic JIA is different to non-JIA controls, where as other JIA subtypes show limited evidence of metabolomic disruption compared to controls. Some of the sJIA-associated metabolomic variation tracks with underlying inflammatory status. These findings suggest a potential clinical application for classifying and monitoring JIA with targeted metabolomics with inflammation. Furthermore, these data support GlycA as a novel marker for monitoring disease activity in sJIA.

Data availability

The datasets analysed during the current study are available from the corresponding author on reasonable request, following approval from CLARITY data custodians.

Abbreviations

JIA:

Juvenile idiopathic arthritis (JIA)

ILAR:

International League of Associations for Rheumatology

RF:

Rheumatoid factor

oJIA:

Oligoarticular JIA

pJIA:

Polyarticular JIA

pJIA(RF-):

RF negative pJIA

pJIA(RF+):

RF positive pJIA

RA:

Rheumatoid arthritis

sJIA:

Systemic JIA

DMARDs:

Disease modifying anti-rheumatic drugs

IBD:

Inflammatory Bowel Disease

SLE:

Systemic Lupus Erythematosus

NMR:

Nuclear magnetic resonance

GlycA:

Glycoprotein acetyls

CRP:

C-reactive protein

CLARITY:

ChiLdhood Arthritis Risk factor Identification sTudY

HREC:

Human Research Ethics Committee

RCH:

Royal Children’s Hospital

AMD:

Adjusted mean difference

CI:

Confidence interval

FDR:

False discovery rate

BH:

Benjamini-Hochberg

p adj :

Adjusted P value

HDL:

High-density lipoprotein

Phe:

Phenylalanine

PUFA:

Polyunsaturated fatty acids

LDL-TG:

Low density lipoprotein triglycerides

NSAIDs:

Non-steroid anti-inflammatory drugs

DHA:

Docosahexaenoic acid

References

  1. Ravelli A, Martini A. Juvenile idiopathic arthritis. Lancet. 2007;369(9563):767–78.

    Article  CAS  PubMed  Google Scholar 

  2. Martini A, Lovell DJ, Albani S, Brunner HI, Hyrich KL, Thompson SD, et al. Juvenile idiopathic arthritis. Nat Rev Dis Primers. 2022;8(1):5.

    Article  PubMed  Google Scholar 

  3. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001. J Rheumatol. 2004;31(2):390–2.

    PubMed  Google Scholar 

  4. Prakken B, Albani S, Martini A. Juvenile idiopathic arthritis. The Lancet. 2011;377(9783):2138–49. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/S0140-6736(11)60244-4

  5. Martini A. Are the number of joints involved or the presence of psoriasis still useful tools to identify homogeneous disease entities in juvenile idiopathic arthritis? J Rheumatol. 2003;30(9):1900–3.

    PubMed  Google Scholar 

  6. Martini A. New takes on categorization and treatment. Nat Rev Rheumatol. 2012;8(2):67–8.

    Article  PubMed  Google Scholar 

  7. Giancane G, Muratore V, Marzetti V, Quilis N, Benavente BS, Bagnasco F, et al. Disease activity and damage in juvenile idiopathic arthritis: methotrexate era versus biologic era. Arthritis Res Ther. 2019;21(1):168.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Daniluk U, Daniluk J, Kucharski R, Kowalczyk T, Pietrowska K, Samczuk P, et al. Untargeted metabolomics and inflammatory markers profiling in Children with Crohn’s Disease and Ulcerative Colitis—A preliminary study. Inflamm Bowel Dis. 2019;25(7):1120–8.

    Article  PubMed  Google Scholar 

  9. Robinson GA, Peng J, Pineda-Torra I, Ciurtin C, Jury EC. Metabolomics Defines Complex Patterns of Dyslipidaemia in Juvenile-SLE patients Associated with inflammation and potential Cardiovascular Disease Risk. Metabolites. 2021;12(1):3.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Kapoor SR, McGrath CM, Fitzpatrick MA, Young SP. Metabolomics in rheumatology. Rheumatology. 2015;kev334.

  11. Giacalone VD, Cammarata-Mouchtouris A, Moncada-Giraldo D, Shenoy SPV, Ponder LA, Gergely TR, et al. Immunometabolic analysis of Synovial Fluid from Juvenile Idiopathic Arthritis patients. Immunohorizons. 2022;6(11):768–78.

    Article  CAS  PubMed  Google Scholar 

  12. Lewis KA, Osier N, Carrasco R, Chiou J, Carter P, Garcia A, et al. Serine, N-acetylaspartate differentiate adolescents with juvenile idiopathic arthritis compared with healthy controls: a metabolomics cross-sectional study. Pediatr Rheumatol. 2022;20(1):12.

    Article  Google Scholar 

  13. Funk RS, Becker ML. Metabolomic profiling identifies exogenous and microbiota-derived metabolites as markers of Methotrexate Efficacy in Juvenile Idiopathic Arthritis. Front Pharmacol. 2021;12.

  14. Johnson CH, Ivanisevic J, Siuzdak G. Metabolomics: beyond biomarkers and towards mechanisms. Nat Rev Mol Cell Biol. 2016;17(7):451–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Ivanisevic J, Elias D, Deguchi H, Averell PM, Kurczy M, Johnson CH, et al. Arteriovenous blood metabolomics: a readout of Intra-tissue Metabostasis. Sci Rep. 2015;5(1):12757.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Zhang A, Sun H, Qiu S, Wang X. NMR-based metabolomics coupled with pattern recognition methods in biomarker discovery and disease diagnosis. Magn Reson Chem. 2013;51(9):549–56.

    Article  CAS  PubMed  Google Scholar 

  17. Cedeno M, Murillo-Saich J, Coras R, Cedola F, Brandy A, Prior A, et al. Serum metabolomic profiling identifies potential biomarkers in arthritis in older adults: an exploratory study. Metabolomics. 2023;19(4):37.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Xu L, Chang C, Jiang P, Wei K, Zhang R, Jin Y et al. Metabolomics in rheumatoid arthritis: advances and review. Front Immunol. 2022;13.

  19. Gupta L, Ahmed S, Jain A, Misra R. Emerging role of metabolomics in rheumatology. Int J Rheum Dis. 2018;21(8):1468–77.

    Article  PubMed  Google Scholar 

  20. Julkunen H, Cichońska A, Tiainen M, Koskela H, Nybo K, Mäkelä V, et al. Atlas of plasma NMR biomarkers for health and disease in 118,461 individuals from the UK Biobank. Nat Commun. 2023;14(1):604.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Otvos JD, Shalaurova I, Wolak-Dinsmore J, Connelly MA, Mackey RH, Stein JH, et al. GlycA: a Composite Nuclear magnetic resonance biomarker of systemic inflammation. Clin Chem. 2015;61(5):714–23.

    Article  CAS  PubMed  Google Scholar 

  22. Lorenzo C, Festa A, Hanley AJ, Rewers MJ, Escalante A, Haffner SM. Novel protein glycan–derived markers of systemic inflammation and C-Reactive protein in relation to Glycemia, insulin resistance, and insulin secretion. Diabetes Care. 2017;40(3):375–82.

    Article  CAS  PubMed  Google Scholar 

  23. Lawler PR, Akinkuolie AO, Chandler PD, Moorthy MV, Vandenburgh MJ, Schaumberg DA, et al. Circulating N-Linked glycoprotein acetyls and longitudinal mortality risk. Circ Res. 2016;118(7):1106–15.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Ritchie SC, Würtz P, Nath AP, Abraham G, Havulinna AS, Fearnley LG, et al. The Biomarker GlycA is Associated with chronic inflammation and predicts long-term risk of severe infection. Cell Syst. 2015;1(4):293–301.

    Article  CAS  PubMed  Google Scholar 

  25. Bartlett DB, Connelly MA, AbouAssi H, Bateman LA, Tune KN, Huebner JL, et al. A novel inflammatory biomarker, GlycA, associates with disease activity in rheumatoid arthritis and cardio-metabolic risk in BMI-matched controls. Arthritis Res Ther. 2016;18(1):86.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Ormseth MJ, Chung CP, Oeser AM, Connelly MA, Sokka T, Raggi P, et al. Utility of a novel inflammatory marker, GlycA, for assessment of rheumatoid arthritis disease activity and coronary atherosclerosis. Arthritis Res Ther. 2015;17(1):117.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Ellis JA, Ponsonby AL, Pezic A, Chavez RA, Allen RC, Akikusa JD, et al. CLARITY – ChiLdhood Arthritis Risk factor identification sTudY. Pediatr Rheumatol. 2012;10(1):37.

    Article  Google Scholar 

  28. LURATI A, SALMASO A, GERLONI V. Accuracy of Wallace Criteria for Clinical Remission in Juvenile Idiopathic Arthritis: a Cohort Study of 761 consecutive cases. J Rheumatol. 2009;36(7):1532–5.

    Article  PubMed  Google Scholar 

  29. Mansell T, Saffery R, Burugupalli S, Ponsonby AL, Tang ML, O’Hely M et al. Early life infection and proinflammatory, atherogenic metabolomic and lipidomic profiles in infancy: a population-based cohort study. Elife. 2022;11.

  30. Ellul S, Wake M, Clifford SA, Lange K, Würtz P, Juonala M, et al. Metabolomics: population epidemiology and concordance in Australian children aged 11–12 years and their parents. BMJ Open. 2019;9(Suppl 3):106–17.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Juonala M, Ellul S, Lawlor DA, Santos Ferreira DL, Carlin JB, Cheung M, et al. A Cross-cohort Study examining the associations of Metabolomic Profile and subclinical atherosclerosis in children and their parents: the Child Health CheckPoint Study and Avon Longitudinal Study of Parents and children. J Am Heart Assoc. 2019;8(14):e011852.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Benjamini Y, Hochberg Y. Controlling the false Discovery rate: a practical and powerful Approach to multiple testing. J Roy Stat Soc: Ser B (Methodol). 1995;57(1):289–300.

    Article  Google Scholar 

  33. Würtz P, Kangas AJ, Soininen P, Lawlor DA, Davey Smith G, Ala-Korpela M. Quantitative serum nuclear magnetic resonance metabolomics in large-scale epidemiology: a primer on -Omic technologies. Am J Epidemiol. 2017;186(9):1084–96.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Jury EC, Peng J, Van Vijfeijken A, Martin Gutierrez L, Woodridge L, Wincup C et al. Systemic lupus erythematosus patients have unique changes in serum metabolic profiles across age associated with cardiometabolic risk. Rheumatology. 2023.

  35. Fang XY, Zhang J, Qian TT, Gao P, Wu Q, Fang Q, et al. Metabolomic profiles, polygenic risk scores and risk of rheumatoid arthritis: a population-based cohort study in the UK Biobank. RMD Open. 2023;9(4):e003560.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Gorczyca D, Postępski J, Czajkowska A, Paściak M, Prescha A, Olesińska E, et al. The profile of polyunsaturated fatty acids in juvenile idiopathic arthritis and association with disease activity. Clin Rheumatol. 2017;36(6):1269–79.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Shen CC, Yao TC, Yeh KW, Huang JL. Association of disease activity and anti-rheumatic treatment in juvenile idiopathic arthritis with serum lipid profiles: a prospective study. Semin Arthritis Rheum. 2013;42(6):590–6.

    Article  CAS  PubMed  Google Scholar 

  38. Priori R, Scrivo R, Brandt J, Valerio M, Casadei L, Valesini G, et al. Metabolomics in rheumatic diseases: the potential of an emerging methodology for improved patient diagnosis, prognosis, and treatment efficacy. Autoimmun Rev. 2013;12(10):1022–30.

    Article  CAS  PubMed  Google Scholar 

  39. Cimaz R. Systemic-onset juvenile idiopathic arthritis. Autoimmun Rev. 2016;15(9):931–4.

    Article  CAS  PubMed  Google Scholar 

  40. Nigrovic PA. Autoinflammation and autoimmunity in systemic juvenile idiopathic arthritis. Proc Natl Acad Sci. 2015;112(52):15785–6.

  41. Barter PJ, Nicholls S, Rye KA, Anantharamaiah GM, Navab M, Fogelman AM. Antiinflammatory properties of HDL. Circ Res. 2004;95(8):764–72.

    Article  CAS  PubMed  Google Scholar 

  42. Zhao WJ, Deng JH, Li CF. Lipid profiles in patients with juvenile idiopathic arthritis: a systematic literature review and meta-analysis. Lipids Health Dis. 2023;22(1):136.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Robinson GA, Waddington KE, Coelewij L, Peng J, Naja M, Wincup C, et al. Increased apolipoprotein-B:A1 ratio predicts cardiometabolic risk in patients with juvenile onset SLE. EBioMedicine. 2021;65:103243.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Vastert SJ, Kuis W, Grom AA. Systemic JIA: new developments in the understanding of the pathophysiology and therapy. Best Pract Res Clin Rheumatol. 2009;23(5):655–64.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Lin YT, Wang CT, Gershwin ME, Chiang BL. The pathogenesis of oligoarticular/polyarticular vs systemic juvenile idiopathic arthritis. Autoimmun Rev. 2011;10(8):482–9.

    Article  CAS  PubMed  Google Scholar 

  46. Thalacker-Mercer AE, Gheller ME. Benefits and adverse effects of histidine supplementation. J Nutr. 2020;150:S2588–92.

    Article  Google Scholar 

  47. Kim D, Mun S, Lee J, Park A, Seok A, Chun YT, et al. Proteomics analysis reveals differential pattern of widespread protein expression and novel role of histidine-rich glycoprotein and lipopolysaccharide-binding protein in rheumatoid arthritis. Int J Biol Macromol. 2018;109:704–10.

    Article  CAS  PubMed  Google Scholar 

  48. Gruppen EG, Connelly MA, Dullaart RPF. Higher circulating GlycA, a pro-inflammatory glycoprotein biomarker, relates to lipoprotein-associated phospholipase A2 mass in nondiabetic subjects but not in diabetic or metabolic syndrome subjects. J Clin Lipidol. 2016;10(3):512–8.

    Article  PubMed  Google Scholar 

  49. Shimizu M, Takei S, Mori M, Yachie A. Pathogenic roles and diagnostic utility of interleukin-18 in autoinflammatory diseases. Front Immunol. 2022;13.

  50. Emwas AHM. In. The strengths and weaknesses of NMR spectroscopy and Mass Spectrometry with Particular Focus on Metabolomics Research. 2015. pp. 161–93.

  51. Oberle EJ, Harris JG, Verbsky JW. Polyarticular juvenile idiopathic arthritis - epidemiology and management approaches. Clin Epidemiol. 2014;6:379–93.

    PubMed  PubMed Central  Google Scholar 

  52. Zhou F, Mao J, Jin Z, Zhu L, Li X. Multi-omic analysis of precocious puberty girls: pathway changes and metabolite validation. Front Endocrinol (Lausanne). 2024;15.

Download references

Acknowledgements

We thank the participants and their families in the CLARITY study for the generous contribution they have made to this project.

Funding

The establishment work and infrastructure for CLARITY was provided by the Murdoch Children’s Research Institute (MCRI). BN is supported by an NHMRC (Australia) Investigator Grant (APP1173314). Murdoch Children’s Research Institute is supported by the Victorian Government’s Operational Infrastructure Program.

Author information

Author notes

  1. Boris Novakovic and Toby Mansell are co-senior authors.

Authors and Affiliations

  1. Infection, Immunity and Global Health Theme, Murdoch Children’s Research Institute, Parkville, VIC, 3052, Australia

    Jooa Kwon, Melanie R. Neeland, Jane Munro, Richard Saffery, Boris Novakovic & Toby Mansell

  2. Department of Paediatrics, University of Melbourne, Parkville, VIC, 3052, Australia

    Jooa Kwon, Melanie R. Neeland, Jane Munro, Richard Saffery, Boris Novakovic & Toby Mansell

  3. Northern Health Research Development and Governance Unit, Epping, VIC, 3076, Australia

    Justine A. Ellis

  4. School of Health & Biomedical Sciences, RMIT University, Melbourne, VIC, 3000, Australia

    Justine A. Ellis

  5. Royal Children’s Hospital Melbourne, Parkville, VIC, 3052, Australia

    Jane Munro

Authors
  1. Jooa Kwon
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Melanie R. Neeland
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Justine A. Ellis
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Jane Munro
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Richard Saffery
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Boris Novakovic
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Toby Mansell
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

JK, MN, RS, BN, TM conceptualised and designed the study. JK led the analysis. JK, RS, TM were major contributors to interpretation of data and drafting of the manuscript. All authors provided expert advice and review. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Toby Mansell.

Ethics declarations

Ethics approval and consent to participate

All CLARITY protocols were approved by the Human Research Ethics Committee of the Royal Children’s Hospital, Melbourne Australia (HREC no. 27127Q). All participants provided informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

: Supplementary Figure S1. Heatmap for the correlation between 249 NMR biomarkers. The heatmap represents a correlation matrix computed using Spearman’s rank correlation coefficient, denoted as r, ranging from − 1 (perfect negative correlation) to 1 (perfect positive correlation), with 0 indicating no correlation. Each cell in the heatmap corresponds to the correlation between two biomarkers. The colour scale from blue to red represents the strength and direction of the correlation (Dark blue: -1 < r<-0.8, light blue: -0.79 < r<-0.5, white: -0.09 < r < 0, yellow: 0.11 < r < 0.49, red: 0.5 < r < 1).

Supplementary Material 2

: Supplementary Figure S2. The distribution difference between JIA group and controls for all primary metabolomic measures (n = 90). Student’s T test for JIA-associated markers between JIA group and controls. padj values were marked above dendrogram. All models were adjusted for participants age and sex.

Supplementary Material 3

: Supplementary Figure S3. Forest plot for the estimated metabolomic difference between each subtype and control for biomarkers not associated with sJIA (46 biomarkers) from adjusted linear regression models (n = 90). Error bars are 95% confidence intervals. Closed points represent padj values < 0.05 (BH). All models were adjusted for participants age and sex.

Supplementary Material 4: Supplementary Data D1-D10.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kwon, J., Neeland, M.R., Ellis, J.A. et al. The plasma metabolome of juvenile idiopathic arthritis varies according to subtype and underlying inflammatory status. Pediatr Rheumatol 22, 113 (2024). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12969-024-01041-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12969-024-01041-8

Keywords

Pediatric Rheumatology

ISSN: 1546-0096