Abstract
Background
Although approximately one-third of patients with rheumatoid arthritis (RA) have comorbid nonalcoholic fatty liver disease (NAFLD), the clinical characteristics of these patients remain unclear. The aim of this study was to clarify the clinical characteristics of RA patients classified as having NAFLD based on imaging, with a focus on composite disease activity assessment.
Methods
This retrospective cohort study analysed data from 349 RA patients enrolled in the Nagahama Riumachi Cohort between March 2017 and August 2023. Patients with liver diseases other than NAFLD were excluded. NAFLD was classified based on findings from abdominal computed tomography and/or ultrasonography performed during the cohort period. Disease activity was assessed using the Clinical Disease Activity Index (CDAI), Simplified Disease Activity Index (SDAI), Disease Activity Score in 28 joints using C-reactive protein (DAS28-CRP), and Disease Activity Score in 28 joints using erythrocyte sedimentation rate (DAS28-ESR). Functional status was evaluated using the Health Assessment Questionnaire Disability Index (HAQ-DI). Factors associated with NAFLD were examined by logistic regression, and repeated measurements during follow-up were analysed using linear mixed-effects models.
Results
Seventy-three patients (20.9%) were classified as having NAFLD and 276 (79.1%) as not having NAFLD. Patients in the NAFLD group were younger, had a higher proportion of men, and had a higher prevalence of body mass index (BMI) ≥ 25 kg/m². There were no significant differences between the groups in disease activity assessed by CDAI, SDAI, or DAS28-CRP. However, the NAFLD group had significantly lower DAS28-ESR and erythrocyte sedimentation rate (ESR) values despite similar C-reactive protein (CRP) levels. HAQ-DI was lower in the NAFLD group, especially in the lower-limb domain. In multivariable analysis, younger age and higher (BMI) were independently associated with NAFLD, but ESR and HAQ-DI were not significantly associated. In linear mixed-effects models, NAFLD classification was not significantly associated with ESR, although ESR tended to be lower in the NAFLD group.
Conclusions
Among patients with RA classified as having NAFLD, DAS28-ESR may underestimate disease activity compared with CRP-based or clinical indices. In these patients, DAS28-CRP, CDAI, and SDAI may provide complementary information.
Keywords: Rheumatoid arthritis, Nonalcoholic fatty liver disease, DAS28-ESR, DAS28-CRP, Erythrocyte sedimentation rate, Disease activity
Background
Nonalcoholic fatty liver disease (NAFLD) is a condition characterized by excessive fat accumulation in the liver in people who consume little or no alcohol and have no secondary causes of hepatic steatosis [1, 2]. It is recognized as a significant public health problem worldwide, including in Japan [3]. Nonalcoholic steatohepatitis (NASH), a more severe form within the NAFLD spectrum, often progresses to cirrhosis and is associated with an increased risk of hepatocellular carcinoma [3]. The presence of advanced fibrosis in patients with NASH markedly increases the risk of liver-related mortality compared with patients with non-NASH NAFLD [4]. Since NAFLD has a significant impact on patient prognosis through liver dysfunction, its accurate diagnosis is important. In 2020, acknowledging the heterogeneity of metabolic fatty liver diseases, an expert panel proposed a new nomenclature and diagnostic criteria, changing NAFLD to metabolic dysfunction–associated fatty liver disease (MAFLD) to achieve a more precise and inclusive diagnosis [5]. MAFLD is diagnosed when there is histological, imaging, or biomarker evidence of hepatic steatosis, together with at least one of the following: overweight/obesity, type 2 diabetes mellitus, or metabolic dysregulation [5–7]. Thus, metabolic disorders such as obesity and diabetes are well-recognized risk factors for NAFLD.
Rheumatoid arthritis (RA) is a systemic autoimmune disorder characterized by chronic inflammation that primarily targets the synovial lining of joints, resulting in persistent, symmetric synovitis and ultimately leading to progressive disability and increased mortality [8]. Patients with RA often exhibit metabolic disturbances such as dyslipidaemia, elevated body mass index (BMI), and type 2 diabetes, which are closely associated with the development of NAFLD [9]. It was reported that one-third of RA patients had NAFLD, which was comparable to its overall prevalence in the general population [10]. Methotrexate is a standard treatment for patients with RA and may cause liver damage. Therefore, the coexistence of NAFLD in patients with RA may be related to the progression of pre-existing fatty liver to fibrotic NASH [11].
However, despite the importance of liver damage in patients with RA, the clinical characteristics of patients with RA and NAFLD remain unclear. We therefore investigated the clinical features of these patients using data from the Nagahama Riumachi Cohort (“Riumachi” means rheumatism in Japanese), with a particular focus on composite disease activity measures.
Methods
Study population
Patients who fulfilled the American College of Rheumatology 1987 or 2010 classification criteria for RA [12, 13] were registered in a prospective study named the Nagahama City Hospital Riumachi (NCH-R) Cohort [14]. Among 562 patients with RA registered in the NCH-R Cohort from March 2017 to August 2023, 365 underwent computed tomography (CT) or abdominal ultrasonography during the study period. Sixteen patients were excluded because of other causes of hepatitis, such as hepatitis C virus, hepatitis B virus, and primary biliary cholangitis, and 349 patients were included in this study (Fig. 1). We retrospectively analysed the clinical data recorded at study entry and at each visit in the database.
Fig. 1.
Flow diagram of patient selection for the Nagahama Riumachi cohort study. Among 562 patients with rheumatoid arthritis (RA), those with other causes of liver disease, such as hepatitis C virus (HCV) infection,hepatitis B virus (HBV) infection, or primary biliary cholangitis (PBC), were excluded. Finally, 349 patients (276 without non-alcoholicfatty liver disease [NAFLD] and 73 with NAFLD) were included in the present analysis
This retrospective cohort was designed to reflect routine clinical practice at a rheumatology center; therefore, no upper age limit was applied. Baseline characteristics and other cross-sectional analyses were based on data obtained at cohort entry.
Patients were classified according to whether NAFLD was identified by abdominal imaging at any time during the cohort period and the timing of imaging differed among patients. Repeated-measures analyses were performed separately using longitudinal follow-up data collected during the observation period. Because abdominal imaging was not routinely performed at cohort entry and was not repeated in a uniform manner during follow-up, these analyses should be interpreted as comparisons based on NAFLD classification in this cohort rather than as true longitudinal analyses of persistent NAFLD confirmed from baseline.
Among 562 patients with rheumatoid arthritis (RA), those with other causes of liver disease, such as hepatitis C virus (HCV) infection, hepatitis B virus (HBV) infection, or primary biliary cholangitis (PBC), were excluded. Finally, 349 patients (276 without non-alcoholic fatty liver disease [NAFLD] and 73 with NAFLD) were included in the present analysis.
Acquisition of clinical and laboratory data
Baseline demographic, clinical and laboratory data, including age, sex, disease duration, medication, erythrocyte sedimentation rate (ESR), serum C-reactive protein (CRP) level, anti-cyclic citrullinated peptide antibodies, swollen joint count, tender joint count, physician’s global assessment of RA activity, and patient’s global assessment of RA activity, were retrospectively abstracted from the electronic medical records.
Radiographic stage and functional class were assessed according to the Steinbrocker classification. Stage III/IV and class III/IV were defined as advanced structural damage and impaired functional status, respectively [15]. Use of oral prednisolone (PSL), methotrexate (MTX), other conventional synthetic DMARDs (csDMARDs), and biologic DMARDs (bDMARDs) was confirmed from the medical records. Disease activity was assessed using the Disease Activity Score 28 based on C-reactive protein (DAS28-CRP) and erythrocyte sedimentation rate (DAS28-ESR), the Clinical Disease Activity Index (CDAI), the Simplified Disease Activity Index (SDAI), the Health Assessment Questionnaire Disability Index (HAQ-DI), and the modified HAQ (mHAQ). In this study, clinical remission was defined using standard cut-offs for each composite index: DAS28-CRP < 2.3 [16], DAS28-ESR < 2.6 [17], CDAI ≤ 2.8 [18], and SDAI ≤ 3.3 [18, 19]. For functional assessment, HAQ-DI ≤ 0.5 was defined as functional remission based on previous reports [20]. In this study, BMI ≥ 25 kg/m² was used as the cutoff according to the criteria of the Japan Society for the Study of Obesity (JASSO) [21].
Longitudinal follow-up and missing data
For longitudinal comparisons, the NAFLD-positive and NAFLD-negative groups were compared at each yearly time point using the available data at that time point. Patients were followed for up to 6 years after cohort entry. Outcome measures were summarized at yearly time points. Follow-up retention was 74.2% in the NAFLD group and 73.2% in the non-NAFLD group at 3 years, and 39.3% and 27.5%, respectively, at 6 years. Missing data were not imputed. Linear mixed-effects models were additionally used for ESR, DAS28-ESR, and DAS28-CRP to account for within-patient correlation over time. NAFLD classification in this cohort, time, and their interaction were included as fixed effects. Age, sex, BMI, baseline MTX use, and baseline PSL use were included as covariates, and patient ID as a random intercept. All available repeated measurements were used without imputation. Baseline covariates were defined using the month 0 data.
Statistical analysis
Statistical analyses were performed with EZR version 1.68 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R version 4.5.1 (The R Foundation for Statistical Computing, Vienna, Austria) [22]. Baseline data were summarized using descriptive statistics. Medians and interquartile ranges (IQR) were calculated for continuous variables, and proportions were calculated for categorical variables. Associations between NAFLD classification and clinical variables were examined using univariable and multivariable logistic regression analyses. Age, BMI, HAQ-DI, and ESR were treated as continuous variables in the regression analyses to avoid arbitrary dichotomization. For follow-up analyses, outcomes at each yearly time point were first summarized descriptively using available-case data. In addition, linear mixed-effects models were fitted for ESR, DAS28-ESR, and DAS28-CRP to account for within-patient correlation across repeated measurements over time. Classification as having NAFLD, time, and their interaction were included as fixed effects, with age, sex, BMI, baseline MTX use, and baseline PSL use as covariates, and patient ID as a random intercept.
Results
Baseline characteristics according to NAFLD classification within the cohort
At the time of entry into this cohort, the participants’ median age was 66 [IQR; 58–75] years, median disease duration was 5 [0–13] years. The median CDAI score was 9.7 [3.7–19.8]. RA patients with comorbid NAFLD (NAFLD group) numbered 73 and those without (non-NAFLD group) numbered 276. The comorbidity rate of NAFLD among RA patients was 20.9%. (Table 1).
Table 1.
Comparison of the clinical characteristics of RA patients with and without NAFLD
| Total(n = 349) | NAFLD(-) (n = 276) |
NAFLD(+) (n = 73) |
p-value | |
|---|---|---|---|---|
| Age, years, median (IQR) | 66 (58–75) | 68 (59–76) | 62 (55–67) | < 0.001 |
| > 75 years, number (%) | 91 (26.1) | 83 (30.1) | 8 (11.0) | < 0.001 |
| Male, number (%) | 112 (32.0) | 81 (29.3) | 31 (42.5) | 0.036 |
| RA duration, years, median (IQR) | 5 (0–13) | 5 (0–13) | 6 (1–11) | 0.631 |
| Anti-CCP(+), number (%) | 265 (75.9) | 213 (80.1) | 52 (74.3) | 0.32 |
| Stage Ⅲ/Ⅳ, number (%) | 118 (33.8) | 99 (35.9) | 19 (26.0) | 0.16 |
| Class Ⅲ/Ⅳ, number (%) | 27 (7.7) | 24 (8.7) | 3 (4.1) | 0.32 |
| CDAI, median (IQR) | 9.7 (3.7–19.8) | 9.5 (3.9–19.9) | 10.4 (2.9–19.0) | 0.729 |
| SDAI, median (IQR) | 10.3 (4.1–20.9) | 10.3 (4.3–21.0) | 10.4 (3.2–20.3) | 0.637 |
| DAS28-CRP, median (IQR) | 2.93 (1.88–4.20) | 2.94 (1.89–4.18) | 2.87 (1.81–4.24) | 0.72 |
| DAS28-ESR, median (IQR) | 3.74 (2.58–5.16) | 3.85 (2.70–5.15) | 3.10 (2.02–5.20) | 0.066 |
| HAQ-DI, median (IQR)) | 0.38 (0.13–1.25) | 0.50 (0.13–1.25) | 0.25 (0-0.88) | 0.03 |
| mHAQ, median (IQR) | 0.13 (0-0.63) | 0.25 (0-0.63) | 0.13 (0-0.50) | 0.027 |
| CRP, mg/dl, median (IQR) | 0.29 (0.07–1.27) | 0.30 (0.07–1.33) | 0.27 (0.10–0.95) | 1 |
| ESR (1 h), mm, median (IQR) | 21.0 (9.0–53.0) | 25.5 (10.0-59.3) | 12.0 (6.0–37.0) | 0.002 |
| BMI, kg/m2, median (IQR) | 22.4 (20.1–25.6) | 21.9 (19.8–24.4) | 25.3 (22.1–28.0) | < 0.001 |
| BMI ≥ 25 kg/m², number (%) | 93 (26.6) | 56 (20.3) | 37 (50.7) | < 0.001 |
| Hypertension, number (%) | 123 (35.2) | 93 (33.7) | 30 (41.1) | 0.271 |
| Type 2 diabetes, number (%) | 37 (10.6) | 25 (9.1) | 12 (16.4) | 0.086 |
| dyslipidaemia, number (%) | 68 (19.5) | 49 (17.8) | 19 (26.0) | 0.134 |
| Smoking, number (%) | 76 (21.8) | 58 (21.0) | 18 (24.7) | 0.525 |
| PSL use, number (%) | 97 (27.8) | 76 (27.5) | 21 (28.8) | 0.883 |
| MTX use, number (%) | 162 (46.4) | 123 (44.6) | 39 (53.4) | 0.189 |
| Other csDMARDs use, number (%) | 141(40.4) | 113(40.9) | 28(38.3) | 0.689 |
| bDMARDs use, number (%) | 141(40.4) | 107(38.9) | 34(45.9) | 0.288 |
Values are expressed as median (interquartile range) or number (%). P values indicate comparisons between patients with and without NAFLD. Steinbrocker stage: radiographic stage I-IV; stage III/IV indicates advanced structural joint damage. Steinbrocker class: functional class I-IV; class III/IV indicates impaired functional status [15].
RA rheumatoid arthritis, BMI body mass index, CRP C-reactive protein, ESR erythrocyte sedimentation rate, CDAI Clinical Disease Activity Index, SDAI Simplified Disease Activity Index, DAS28 28-joint Disease Activity Score, HAQ Health Assessment Questionnaire, mHAQ modified Health Assessment Questionnaire, OR odds ratio, CI confidence interval, PSL prednisolone, MTX methotrexate
The NAFLD group was significantly younger than the non-NAFLD group (62 years vs. 68 years, p < 0.001) and had a relatively higher proportion of men (42.5% vs. 29.3%, p = 0.036), although women remained the majority in both groups. BMI ≥ 25 kg/m² was also more common in the NAFLD group (50.7% vs. 20.3%, p < 0.001). In addition, type 2 diabetes tended to be more common in the NAFLD group (16.4% vs. 9.1%, p = 0.086). Baseline characteristics are shown in Table 1. Logistic regression analyses of factors associated with NAFLD classification are presented in Table 2.
Table 2.
Comparison of remission between RA patients with and without NAFLD
| NAFLD(-) | NAFLD(+) | OR (95%CI) | p-value | |
|---|---|---|---|---|
| DAS28-CRP (Remission) | 115 (41.7) | 31 (42.5) | 1.03 (0.59–1.80) | 0.895 |
| DAS28-ESR (Remission) | 41 (16.4) | 19 (28.4) | 2.01 (1.01–3.92) | 0.035 |
| CDAI (Remission) | 57 (20.7) | 18 (24.7) | 1.26 (0.64–2.38) | 0.522 |
| SDAI (Remission) | 56 (20.3) | 19 (26.0) | 1.38 (0.71–2.60) | 0.336 |
| HAQ (Remission) | 145 (52.5) | 48 (65.8) | 1.73 (0.98–3.11) | 0.048 |
The table shows the number (%) of patients achieving remission according to each index (DAS28-CRP, DAS28-ESR, CDAI, SDAI, and HAQ-DI), together with odds ratios (ORs) and 95% confidence intervals (CIs) for NAFLD (+) versus NAFLD (−). Remission was defined as described in the Methods section
Comparison of disease activity and laboratory markers between patients with and without NAFLD
Remission rates for each indicator between groups showed no significant statistical differences in disease activity measured by CDAI, SDAI, and DAS28-CRP, whereas the remission rate based on DAS28-ESR was higher in the NAFLD group (28.4% vs. 16.4%; OR, 2.01 [1.01–3.92]; p = 0.035) (Table 3). These findings suggest that the discrepancy between ESR-based and CRP-based or clinical assessments was greater in patients classified as having NAFLD in this cohort. In addition, HAQ-DI and mHAQ were significantly lower in the NAFLD group (Table 1). Because HAQ-DI was lower in the NAFLD group, we further examined individual HAQ-DI items and found that scores for lower-limb activities were lower in the NAFLD group (Table 4).
Table 3.
Univariate and multivariate logistic regression analyses of factors associated with NAFLD in patients with RA
| Univariate | Multivariate | |||
|---|---|---|---|---|
| OR (95%CI) | p-value | OR (95%CI) | p-value | |
| Age | 0.97(0.94–0.99) | 0.015 | 0.96 (0.94–0.99) | 0.008 |
| Male sex | 1.66 (0.94–2.92) | 0.08 | 1.55 (0.85–2.83) | 0.155 |
| Anti-CCP(+) | 1.37 (0.79–2.39) | 0.26 | ||
| Stage Ⅲ/Ⅳ | 0.64 (0.33–1.17) | 0.16 | ||
| Class Ⅲ/Ⅳ | 0.46 (0.09–1.57) | 0.32 | ||
| DAS-CRP | 3.06(1.49–6.29) | 0.002 | ||
| DAS-ESR | 0.33 (0.16–0.69) | 0.003 | ||
| HAQ-DI | 0.58 (0.37–0.89) | 0.013 | 0.75 (0.47–1.20) | 0.236 |
| mHAQ | 0.49(0.27–0.88) | < 0.001 | ||
| CRP | 0.95(0.87–1.03) | 0.219 | ||
| ESR(1 h) | 0.98 (0.97–0.99) | 0.004 | 0.99 (0.98–1.01) | 0.41 |
| BMI | 1.20 (1.10–1.30) | < 0.001 | 1.18 (1.10–1.27) | < 0.001 |
| Hypertension | 1.37 (0.78–2.40) | 0.271 | 1.59 (0.85–2.96) | 0.146 |
| Type 2 diabetes | 1.97 (0.85–4.35) | 0.086 | 1.99 (0.84–4.70) | 0.115 |
| dyslipidaemia | 1.63 (0.83–3.09) | 0.134 | ||
| Smoking | 1.23 (0.63–2.32) | 0.525 | ||
| Prednisolone use | 1.06 (0.57–1.94) | 0.883 | ||
| Methotrexate use | 1.43 (0.83–2.48) | 0.189 |
Multivariable logistic regression model adjusted for age, sex, body mass index (BMI), hypertension, type 2 diabetes mellitus, HAQ-DI, and ESR, selected based on clinical relevance
Table 4.
Item-level comparison of Health Assessment Questionnaire Disability Index (HAQ-DI) scores between RA patients with and without NAFLD.
| Are you able to: | NAFLD(-) | NAFLD(+) | p-value | ||
|---|---|---|---|---|---|
| HAQ1 | Dressing and Grooming | Dress yourself, including tying shoelaces and doing buttons | 185/49/34/8 | 54/16/3/0 | 0.078 |
| HAQ2 | Shampoo your hair | 207/35/19/13 | 58/13/2/0 | 0.088 | |
| HAQ3 | Arising | Stand up from an armless chair | 177/56/33/10 | 48/24/1/0 | < 0.01 |
| HAQ4 | Get in and out of bed | 139/86/43/7 | 45/21/7/0 | 0.23 | |
| HAQ5 | Eating | Cut your meat | 193/46/25/11 | 56/12/4/1 | 0.59 |
| HAQ6 |
Lift a full cup or glass to your mouth |
228/30/12/5 | 66/6/1/0 | 0.47 | |
| HAQ7 | Open a new carton of milk | 125/72/34/44 | 44/20/4/5 | 0.034 | |
| HAQ8 | Walking | Walk outdoors on flat ground | 193/52/24/7 | 61/11/1/0 | 0.04 |
| HAQ9 | Climb up five stairs | 160/61/28/26 | 52/18/3/0 | < 0.01 | |
| HAQ10 | Hygiene | Wash and dry your entire body | 196/38/28/13 | 54/15/3/1 | 0.14 |
| HAQ11 | Take a bath | 215/30/13/17 | 63/6/4/0 | 0.1 | |
| HAQ12 | Get on and off the toilet | 230/29/11/4 | 63/8/2/0 | 0.93 | |
| HAQ13 | Reach | Reach and get down a 5 lb object | 146/50/28/51 | 47/14/6/6 | 0.14 |
| HAQ14 |
Bend down to pick up clothing from the floor |
199/36/21/18 | 60/8/4/1 | 0.26 | |
| HAQ15 | Grip | Open push-button car doors | 215/32/18/10 | 67/4/1/1 | 0.074 |
| HAQ16 | Open jars which have been previously | 157/55/32/30 | 46/16/7/4 | 0.52 | |
| HAQ17 | Turn taps on and of | 163/55/26/27 | 49/13/8/3 | 0.4 | |
| HAQ18 | Activity | Run errands and shop | 204/28/19/23 | 67/4/2/0 | < 0.01 |
| HAQ19 | Get in and out of a car | 210/38/21/3 | 69/3/1/0 | < 0.01 | |
| HAQ20 | Do chores such as vacuuming, housework or light gardening | 184/42/20/28 | 58/13/2/0 | < 0.01 | |
For each HAQ item (HAQ1–HAQ20), the distribution of responses (0/1/2/3) in the NAFLD (−) and NAFLD (+) groups and the corresponding p value for between-group comparison are presented. Items related to lower limb function and daily activities are included
Factors associated with NAFLD
In univariable logistic regression analysis, younger age, male sex, lower ESR, lower DAS28-ESR, lower HAQ-DI, and higher BMI were significantly associated with classification as having NAFLD. In multivariable logistic regression analysis, age, sex, BMI, hypertension, diabetes mellitus, HAQ-DI, and ESR were included in the model. Younger age (OR 0.96, 95% CI 0.94–0.99, p = 0.008) and higher BMI (OR 1.18, 95% CI 1.10–1.27, p < 0.001) were independently associated with NAFLD, whereas sex, hypertension, diabetes mellitus, HAQ-DI, and ESR were not (Table 2).
Longitudinal changes in RA disease activity during follow-up according to NAFLD classification in this cohort
Next, we examined disease activity measures, including CDAI, SDAI, DAS28-CRP, DAS28-ESR, and HAQ-DI, from cohort entry up to 6 years of follow-up. Figure 2 shows yearly trends in disease activity and functional measures, whereas Table 3 shows remission status at the comparison point.
Fig. 2.
Yearly trends in disease activity and functional measures during follow-up according to NAFLD classification in this cohort. CDAI, SDAI, DAS28-CRP, DAS28-ESR, HAQ-DI, and mHAQ are shown at yearly time points in patients with and without NAFLD asidentified by abdominal imaging during the cohort period. Values at each time point are based on available-case data only
In descriptive comparisons, ESR, DAS28-ESR, and HAQ-DI tended to be lower in patients classified as having NAFLD, whereas DAS28-CRP did not differ significantly between the groups (Fig. 2). We also performed linear mixed-effects model analyses for ESR, DAS28-ESR, and DAS28-CRP to account for repeated measurements over time. In models adjusted for age, sex, BMI, baseline MTX use, and baseline PSL use, month was significantly associated with decreases in ESR, DAS28-ESR, and DAS28-CRP during follow-up. Classification as having NAFLD was not significantly associated with ESR, although ESR tended to be lower in this group (coefficient, -5.11; p = 0.095). No significant associations were found for DAS28-ESR (coefficient, -0.205; p = 0.182) or DAS28-CRP (coefficient, 0.038; p = 0.762). The interaction between NAFLD classification and time was not significant in any model, suggesting that the between-group differences did not show a clear change over time (Table 5).
Table 5.
Longitudinal associations of NAFLD status with ESR, DAS28-ESR, and DAS28-CRP in linear mixed-effects models
| Outcome | Term | β coefficient | 95% CI | p value |
|---|---|---|---|---|
| ESR | NAFLD | -5.11 | -11.10 to 0.87 | 0.095 |
| Month | -0.068 | -0.118 to -0.018 | 0.008 | |
| Age | 0.621 | 0.443 to 0.799 | < 0.001 | |
| Male sex | -1.48 | -6.06 to 3.10 | 0.527 | |
| BMI | -0.627 | -1.193 to -0.062 | 0.031 | |
| MTX use | -3.53 | -7.98 to 0.92 | 0.121 | |
| PSL use | -1.6 | -6.27 to 3.08 | 0.504 | |
| NAFLD × Month | -0.057 | -0.160 to 0.045 | 0.273 | |
| DAS28-ESR | NAFLD | -0.205 | -0.506 to 0.095 | 0.182 |
| Month | -0.011 | -0.0138 to -0.0082 | < 0.001 | |
| Age | 0.0185 | 0.0099 to 0.0271 | < 0.001 | |
| Male sex | -0.194 | -0.416 to 0.027 | 0.086 | |
| BMI | -0.0029 | -0.0303 to 0.0245 | 0.837 | |
| MTX use | -0.121 | -0.336 to 0.093 | 0.269 | |
| PSL use | -0.15 | -0.375 to 0.075 | 0.192 | |
| NAFLD × Month | -0.0044 | -0.0102 to 0.0013 | 0.130 | |
| DAS28-CRP | NAFLD | 0.038 | -0.206 to 0.282 | 0.762 |
| Month | -0.0112 | -0.0137 to -0.0086 | < 0.001 | |
| Age | 0.0104 | 0.0036 to 0.0172 | 0.003 | |
| Male sex | -0.036 | -0.211 to 0.139 | 0.687 | |
| BMI | 0.0116 | -0.0100 to 0.0333 | 0.293 | |
| MTX use | -0.113 | -0.283 to 0.058 | 0.196 | |
| PSL use | -0.0468 | -0.226 to 0.132 | 0.609 | |
| NAFLD × Month | -0.0021 | -0.0072 to 0.0030 | 0.425 |
Linear mixed-effects models included NAFLD classification in this cohort, month, and their interaction as fixed effects, adjusted for age, sex, BMI, baseline MTX use, and baseline PSL use, with patient ID as a random intercept. β coefficients, 95% CI, and p values are shown
CDAI, SDAI, DAS28-CRP, DAS28-ESR, HAQ-DI, and mHAQ are shown at yearly time points in patients with and without NAFLD as identified by abdominal imaging during the cohort period. Values at each time point are based on available-case data only.
Discussion
In this retrospective cohort, patients classified as having NAFLD based on abdominal imaging during the cohort period had lower ESR values and lower DAS28-ESR scores than those without NAFLD, whereas DAS28-CRP did not differ between the groups. In linear mixed-effects models, classification as having NAFLD was not significantly associated with ESR, although ESR tended to be lower in this group. No significant association was found for DAS28-ESR or DAS28-CRP, and no significant interaction with time was observed.
These findings do not suggest that DAS28-ESR is problematic only in patients with NAFLD. In this cohort, the discrepancy between ESR-based and CRP-based or clinical composite measures appeared greater in the NAFLD-classified group. In such cases, DAS28-CRP, CDAI, and SDAI may be useful for additional assessment. Remission rates differed across DAS28-CRP, DAS28-ESR, CDAI, SDAI, and HAQ-DI. This is not unexpected, because these indices are based on different components and different remission thresholds. In our cohort, DAS28-CRP showed higher remission rates than the other disease activity indices in both groups. In addition, HAQ-DI reflects functional status rather than inflammatory disease activity alone and may be influenced by age, frailty, structural damage, and comorbidities. Therefore, HAQ-DI remission should not be interpreted as equivalent to remission defined by composite disease activity measures. In this study, patients with NAFLD had lower HAQ-DI scores, particularly in items related to activity and lower limb function (Table 4), despite similar disease activity. Although this may appear counterintuitive given the higher BMI in the NAFLD group, one possible explanation is that these patients were younger and included fewer individuals aged > 75 years.
NAFLD was present in about one-fifth of our cohort, and patients with NAFLD were more likely to have BMI ≥ 25 kg/m² (Table 1), which is consistent with previous reports [4, 5]. Because no upper age limit was set, this cohort included older adults seen in routine clinical practice. However, fatty liver in older adults may be influenced by age-related and metabolic factors, and the findings related to age should be interpreted with caution. In multivariable logistic regression analyses, younger age and higher BMI were independently associated with NAFLD, whereas HAQ-DI and ESR were not. These findings suggest that the associations of lower HAQ-DI and lower ESR with NAFLD in univariable analyses may have been influenced, at least in part, by differences in demographic and metabolic characteristics.
ESR is influenced not only by inflammation but also by non-inflammatory factors such as age, sex, anemia, obesity, and metabolic syndrome [23]. NAFLD is characterized by chronic low-grade systemic inflammation and metabolic abnormalities, including obesity and metabolic syndrome, which may affect fibrinogen levels and red blood cell properties [24, 25]. A similar dissociation between ESR and other markers has been described in conditions such as macrophage activation syndrome, in which hypofibrinogenemia leads to an inappropriately low ESR despite marked inflammation [26]. These findings suggest that ESR-based indices such as DAS28-ESR may be less reliable in some clinical settings. However, because fibrinogen and detailed coagulation/hemostatic parameters were not measured in the present study, any mechanistic explanation for the lower ESR values remains speculative and should be interpreted cautiously.
This study has several limitations. First, the retrospective single-center design limits causal inference and may introduce selection and information bias. Second, NAFLD was defined by imaging findings without histological confirmation, and disease severity such as steatohepatitis or fibrosis could not be assessed. Therefore, the impact of disease severity on ESR remains unclear. Third, abdominal imaging was not systematically performed at study entry, and the timing of imaging varied across participants. In addition, repeat abdominal imaging was not uniformly available during follow-up. Therefore, these analyses should be interpreted as comparisons based on NAFLD classification in this cohort, rather than as longitudinal effects of baseline-confirmed persistent NAFLD. Fourth, detailed metabolic and mechanistic data, including lipid profiles, glucose-related or insulin resistance-related measures, longitudinal changes in body weight, fibrinogen, and coagulation parameters, were not uniformly available. Therefore, the mechanism underlying the lower ESR values remains unclear. Finally, because this study was conducted in a single Japanese hospital-based cohort, the findings may not be fully generalizable to other populations or healthcare settings.
Conclusions
In this retrospective cohort, patients with RA classified as having NAFLD based on abdominal imaging during the cohort period were younger, had a higher BMI, and included a higher proportion of men, although women remained the majority in both groups. These patients had lower ESR and DAS28-ESR values, whereas DAS28-CRP, CDAI, and SDAI were similar between the groups. In linear mixed-effects models, ESR also tended to be lower in patients classified as having NAFLD. These findings suggest that DAS28-ESR may underestimate disease activity in some patients with RA classified as having NAFLD, and that other disease activity measures such as DAS28-CRP, CDAI, and SDAI may also be informative in these patients.
Acknowledgements
We thank the members of the medical laboratory in our hospital for their helpful input in this study, and David Price of English Services for Scientists based in Hiroshima for proofreading.
Abbreviations
- CDAI
Clinical Disease Activity Index
- CRP
C-reactive protein
- DAS28-CRP
Disease Activity Score 28-C-reactive protein
- DAS28-ESR
Disease Activity Score 28-erythrocyte sedimentation rate
- ESR
Erythrocyte sedimentation rate
- HAQ-DI
Health Assessment Questionnaire Disability Index
- IQR
Interquartile range
- MAFLD
Metabolic dysfunction–associated fatty liver disease
- NAFLD
Nonalcoholic fatty liver disease
- NASH
Nonalcoholic steatohepatitis
- NCH-R
Nagahama City Hospital Riumachi
- RA
Rheumatoid arthritis
- SDAI
Simplified Disease Activity Index
Authors' contributions
HU conceived the study design; YS analysed the data; YS and HU wrote the manuscript. All authors read and approved the final manuscript.
Funding
This research received no external funding.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study was approved by the Ethics Committee of Nagahama City Hospital (approval number: H28-38). Patients gave verbal and written informed consent.
Consent for publication
Not applicable.
Competing interests
M.T. belongs to the department financially supported by two local governments in Japan (Nagahama City, Shiga and Toyooka City, Hyogo) and Asahi Kasei Pharma Corp. M.T. has received research grants and/or speaker fees from AbbVie GK, Asahi Kasei Pharma Corp., Astellas Pharma Inc., Ayumi Pharmaceutical Corp., Chugai Pharmaceutical Co., Ltd., Daiichi Sankyo Co., Ltd., Mitsubishi Tanabe Pharma Corp., Taisho Pharmaceutical Co., Ltd., Teijin Pharma, Ltd. A.M. has received research grants from Asahi Kasei Pharma Corp., Chugai Pharmaceutical Co., Ltd., Taisho Pharmaceutical Co., Ltd., Eisai Co., Ltd., and Mitsubishi Tanabe Pharma Corp, and has received speaker fees from Astellas Pharma Inc., Bristol Myers Squibb Company., Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Eli Lilly Japan K.K.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella M, et al. The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases. Hepatology. 2018;67(1):328–57. 10.1002/hep.29367. [DOI] [PubMed] [Google Scholar]
- 2.Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M. Global epidemiology of nonalcoholic fatty liver disease: Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology. 2016;64(1):73–84. 10.1002/hep.28431. [DOI] [PubMed] [Google Scholar]
- 3.Tokushige K, Ikejima K, Ono M, Eguchi Y, Kamada Y, Itoh Y, et al. Evidence-based clinical practice guidelines for nonalcoholic fatty liver disease/nonalcoholic steatohepatitis 2020. J Gastroenterol. 2021;56(11):951–63. 10.1007/s00535-021-01796-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sanyal AJ, Van Natta ML, Clark J, Neuschwander-Tetri BA, Diehl AM, Dasarathy S, et al. Prospective study of outcomes in adults with nonalcoholic fatty liver disease. N Engl J Med. 2021;385(17):1559–69. 10.1056/NEJMoa2029349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Eslam M, Newsome PN, Sarin SK, Anstee QM, Targher G, Romero-Gomez M, et al. A new definition for metabolic dysfunction-associated fatty liver disease: An international expert consensus statement. J Hepatol. 2020;73(1):202–9. 10.1016/j.jhep.2020.03.039. [DOI] [PubMed] [Google Scholar]
- 6.Bedogni G, Miglioli L, Masutti F, Tiribelli C, Marchesini G, Bellentani S. Prevalence of and risk factors for nonalcoholic fatty liver disease: The Dionysos nutrition and liver study. Hepatology. 2005;42(1):44–52. 10.1002/hep.20734. [DOI] [PubMed] [Google Scholar]
- 7.Hamaguchi M, Kojima T, Takeda N, Nakagawa T, Taniguchi H, Fujii K, et al. The metabolic syndrome as a predictor of nonalcoholic fatty liver disease. Ann Intern Med. 2005;143(10):722–8. 10.7326/0003-4819-143-10-200511150-00009. [DOI] [PubMed] [Google Scholar]
- 8.Scott DL, Wolfe F, Huizinga TWJ. Rheumatoid arthritis. Lancet. 2010;376(9746):1094–108. 10.1016/S0140-6736(10)60826-4. [DOI] [PubMed] [Google Scholar]
- 9.Barbarroja N, Ruiz-Ponce M, Cuesta-López L, Pérez-Sánchez C, López-Pedrera C, Arias-de la Rosa I, et al. Nonalcoholic fatty liver disease in inflammatory arthritis: Relationship with cardiovascular risk. Front Immunol. 2022;13:997270. 10.3389/fimmu.2022.997270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zamani M, Alizadeh-Tabari S, Chitkara P, Singh S, Loomba R. Prevalence of nonalcoholic fatty liver disease in patients with rheumatoid arthritis: A systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2023;21(11):2880–e28921. 10.1016/j.cgh.2023.02.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mori S, Arima N, Ito M, Fujiyama S, Kamo Y, Ueki Y, et al. Incidence, predictive factors and severity of methotrexate-related liver injury in rheumatoid arthritis: A longitudinal cohort study. Rheumatol Adv Pract. 2020;4(1):rkaa020. 10.1093/rap/rkaa020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum. 1988;31(3):315–24. 10.1002/art.1780310302. [DOI] [PubMed] [Google Scholar]
- 13.Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham CO 3rd, et al. 2010 rheumatoid arthritis classification criteria: An American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 2010;62(9):2569–81. 10.1002/art.27584. [DOI] [PubMed] [Google Scholar]
- 14.Naito R, Taniguchi M, Onizawa H, Nakajima T, McCracken K, Mori M, et al. Upper extremity joint tenderness as a practical indicator for assessing presenteeism in rheumatoid arthritis patients: A cross-sectional observational study. PLoS ONE. 2025;20(1):e0318047. 10.1371/journal.pone.0318047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Steinbrocker O, Traeger CH, Batterman RC. Therapeutic criteria in rheumatoid arthritis. J Am Med Assoc. 1949;140(8):659–62. [DOI] [PubMed] [Google Scholar]
- 16.Inoue E, Yamanaka H, Hara M, Tomatsu T, Kamatani N. Comparison of Disease Activity Score (DAS)28-erythrocyte sedimentation rate and DAS28-C-reactive protein threshold values. Ann Rheum Dis. 2007;66(3):407–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Landewé R, van der Heijde D, van der Linden S, Boers M. Twenty-eight-joint counts invalidate the DAS28 remission definition but not the DAS28 low disease activity state: A comparison between the DAS28 and the original DAS remission definition. Arthritis Rheum. 2006;54(12):3719–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhang B, Combe B, Rincheval N, Felson DT. Validation of ACR/EULAR definition of remission in rheumatoid arthritis from RA practice: The ESPOIR cohort. Arthritis Res Ther. 2012;14(3):R156. 10.1186/ar3896. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Aletaha D, Ward MM, Machold KP, Nell VP, Stamm T, Smolen JS. Remission and active disease in rheumatoid arthritis: Defining criteria for disease activity states. Arthritis Rheum. 2005;52(9):2625–36. [DOI] [PubMed] [Google Scholar]
- 20.Sugihara T, Ishizaki T, Hosoya T, Iga S, Yokoyama W, Hirano F, Miyasaka N, Harigai M. Structural and functional outcomes of a therapeutic strategy targeting low disease activity in patients with elderly-onset rheumatoid arthritis: A prospective cohort study (CRANE). Rheumatology (Oxford). 2015;54(5):798–807. 10.1093/rheumatology/keu395. [DOI] [PubMed] [Google Scholar]
- 21.Ogawa W, Ito C, Kitahara A, et al. Definition, criteria, and core concepts of guidelines for the management of obesity disease in Japan. Endocr J. 2024;71(3):223–31. 10.1507/endocrj.EJ23-0593. [DOI] [PubMed] [Google Scholar]
- 22.Kanda Y. Investigation of the freely available easy-to-use software EZR for medical statistics. Bone Marrow Transpl. 2013;48(3):452–8. 10.1038/bmt.2012.244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Alende-Castro V, Alonso-Sampedro M, Vazquez-Temprano N, Tuñez C, Rey D, García-Iglesias C, et al. Factors influencing erythrocyte sedimentation rate in adults: New evidence for an old test. Med (Baltim). 2019;98(34):e16816. 10.1097/MD.0000000000016816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Haukeland JW, Damas JK, Konopski Z, Loberg EM, Haaland T, Goverud I, et al. Systemic inflammation in nonalcoholic fatty liver disease is characterized by elevated levels of CCL2. J Hepatol. 2006;44(6):1167–74. 10.1016/j.jhep.2006.02.011. [DOI] [PubMed] [Google Scholar]
- 25.George MD, Giles JT, Katz PP, England BR, Mikuls TR, Michaud K, et al. Impact of obesity and adiposity on inflammatory markers in patients with rheumatoid arthritis. Arthritis Care Res (Hoboken). 2017;69(12):1789–98. 10.1002/acr.23229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Eloseily EMA, Minoia F, Crayne CB, Beukelman T, Ravelli A, Cron RQ. Ferritin to erythrocyte sedimentation rate ratio: Simple measure to identify macrophage activation syndrome in systemic juvenile idiopathic arthritis. ACR Open Rheumatol. 2019;1(6):345–9. 10.1002/acr2.11048. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.