Increased NAFLD risk in newly diagnosed patients with RA during the first 4 years of follow-up: a nationwide, population-based cohort study

Abstract

Background

Although the non-alcoholic fatty liver disease (NAFLD) is prevalent in the general population, NAFLD risk in newly diagnosed rheumatoid arthritis (RA) has rarely been explored. In this population-based cohort, we examined NAFLD risk in patients with RA and identified the potential risk factors.

Design

Retrospective study.

Setting

Taiwan.

Participants

2281 newly diagnosed patients with RA and selected 91 240 individuals without RA to match with patients with RA (1:40) by age, gender, income status and urbanisation level of the residence.

Outcomes

In this retrospective study using the 2000–2018 claim data from two-million representative Taiwanese population, we identified and compared the incidence rates (IRs) of NAFLD and alcoholic fatty liver disease (AFLD) between RA and non-RA groups. Using multivariable regression analyses, we estimated adjusted HR (aHR) of NAFLD development in patients with RA compared with individuals without RA, with 95% CIs.

Results

The incidences of NALFD and AFLD were not significantly different between individuals with RA and without RA during the 17-year follow-up period. However, patients with RA had significantly increased NAFLD risk during the first 4 years after RA diagnosis, with IR ratio of 1.66 fold (95% CI 1.18 to 2.33, p<0.005), but the risk was reduced after the first 4 years. Multivariable regression analyses revealed that aHR was 2.77-fold greater in patients not receiving disease-modifying anti-rheumatic drugs therapy than in non-RA subjects (p<0.05). Old age, women, low-income status and obesity could significantly predict NAFLD development.

Conclusions

We demonstrated elevated risk of NAFLD in patients with RA during the first 4 years after RA diagnosis, and old age, women, low-income status and obesity were significant predictors of NAFLD.

Correspondence to Dr Der-Yuan Chen; [email protected]

STRENGTHS AND LIMITATIONS OF THIS STUDY

• Patients with rheumatoid arthritis (RA) had a significantly higher incidence rate of non-alcoholic fatty liver disease (NAFLD) than the non-RA subjects during the first 4 years after RA diagnosis but not after 4 years.

• The adjus HR was 2.77-fold higher in patients with RA without disease-modifying anti-rheumatic drugs therapy than in the non-RA subjects.

• The old age, women, low-income status and obesity were significant predictors of NAFLD development in RA.

• The accumulative incidence of alcoholic fatty liver disease was not significantly different between the RA and non-RA groups during the 17-year follow-up period.

• The absence of data regarding individual RA disease activity, which was related to the risk of NAFLD, is another important limitation.

Introduction

Non-alcoholic fatty liver disease (NAFLD), the most common cause of chronic non-viral liver disease in the Western countries,¹ is characterised by the presence of steatosis in more than 5% of hepatocytes in individuals with minimal or no alcohol consumption.¹ NAFLD represents histological change ranging from indolent and simple steatosis to non-alcoholic steatohepatitis (NASH),¹ which may progress to cirrhosis, liver failure and hepatocellular carcinoma.^{2 3} According to the different ages, gender and ethnicity, 17%–46% of adults could be affected by NAFLD.^{4 5} The data from Korea,⁶ China⁷ and Taiwan⁸ indicated a similarly high prevalence of NAFLD (11%–45%) among northern Asian populations. NAFLD is often associated with obesity, metabolic syndrome (MetS), insulin resistance, type II diabetes mellitus (DM) and genetic susceptibility.^9–12 However, the aetiopathogenesis of NAFLD is complex and not fully elucidated.

Rheumatoid arthritis (RA) is an inflammatory autoimmune disease characterised by chronic synovitis and bone destruction and is often associated with poor life quality.¹³ Several epidemiological studies found a high prevalence of NAFLD-related risk factors (eg, MetS, dyslipidaemia and DM) in patients with RA.^{10 14–16} Besides, the therapy with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) may be another risk factor for NAFLD in RA. Methotrexate (MTX), the most commonly used csDMARDs in RA, has been reported to be associated with increased NAFLD risk.^17–19 Accordingly, a high prevalence of ultrasound-proven NAFLD was observed in 20.3%–24.2% of patient with RA.^{20 21} However, Wagan et al demonstrated no significant association between treatment with MTX or other csDMARDs and the occurrence of NAFLD in the Pakistani cohort of RA.²⁰ To the best of our knowledge, very few studies have investigated the incidence of NAFLD in a large cohort of patient with RA.^20–22

The Taiwanese National Health Insurance Research Database (NHIRD) has recently become a great tool for population-based, longitudinal epidemiologic studies. In the present study, we aimed to use the NHIRD to examine the NAFLD incidence rates (IRs) in newly diagnosed patients with RA during a 17-year follow-up period. We also identified the potential risk factors predictive of the occurrence of NAFLD in patients with RA.

Methods

Patient and public involvement

None.

Study design

This investigation was a nationwide, population-based, age-matched and gender-matched cohort study, as shown in figure 1.

Figure 1. Flow chart of case selection in this study. The newly diagnosed patients with rheumatoid arthritis (RA) with new-onset NAFLD, and age-matched and sex-matched non-RA control subjects were selected from the Taiwan National Health Insurance research database. NAFLD, non-alcoholic liver disease.

Data source

The study analysed the claim data from 2000 to 2018 extracted from the Taiwanese NHIRD. Initiated in Taiwan in 1995, the National Health Insurance (NHI) is a compulsory insurance programme currently covering 99% of the Taiwanese population. The Bureau of NHI routinely audits and monitors claims data to identify and remove fraudulent data. The National Health Research Institute (NHRI) manages the NHIRD and releases claims data for research purposes. The NHIRD contains registration files and comprehensive claims data for all ambulatory and inpatient services, including information regarding demographics, residence, prescription medications, diagnoses, medical expenditures, surgeries, procedures and examinations. The diagnoses in the NHIRD were made according to the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), and ICD-10-CM. All patients with major illnesses, including malignancies and RA, are registered in the Registry of Catastrophic Illness Database (RCIPD) if the categorisation is agreed on, regardless of disease severity, after a thorough review of original charts by two independent specialists. A catastrophic illness certificate is then issued to these patients, who are then exempt from the expenses of medical services.

The NHRI randomly selected two million representative population who were NHI beneficiaries in 2000 to construct a representative longitudinal health insurance database (LHID2000). We selected patients with RA and individuals without RA from the LHID2000 and extracted their claims data between 2000 and 2018.

Identification of patients with RA from the entire population of Taiwan

From 2000 to 2018 LHID2000, patients were identified to have RA if they were registered in RCIPD of the NHIRD with a diagnosis of RA (ICD-9-CM codes 714.0, 714.30–714.33; ICD-10-CM codes M05.7- M05.9, M06.0, M06.2, M06.3, M06.8, M06.9, M08), which two board-certified rheumatologists had verified. We also ensured that the diagnosis of RA was made according to the 1987 American College of Rheumatology criteria²³ or the 2010 American College of Rheumatology/European League Against Rheumatism classification criteria for RA.²⁴ We enrolled all the patients who were newly diagnosed with RA from 2002 to 2018. Exclusion criteria included any diagnosis of acute/subacute liver necrosis, liver abscess/sequelae of chronic liver disease (ICD-9-CM codes 570–573; ICD-10-CM codes K70-K77), hepatitis B (ICD-9-CM codes 070.2, 070.3, 070.42, 070.52, V02.61; ICD-10-CM codes B16, B18.0, B18.1, B19.1, Z22.51), hepatitis C (ICD-9-CM codes 070.41, 070.44, 070.51, 070.54, 070.7, v02.62; ICD-10-CM codes B18.2, B18.2, Z22.52), or alcohol related disorders (ICD-9-CM codes 305.0, 303, 291; ICD-10-CM codes F10) before the index date. The index date for the patients with RA was defined as the date of acquisition of the catastrophic illness certificate.

Identification of individuals without RA from the two million representative population of Taiwan

Individuals without outpatient or inpatient diagnosis of RA from 2000 to 2018 were considered as individuals without RA. From the LHID2000, we randomly selected and matched individuals without RA (in a ratio of 1:40) to patients with RA by age, gender, year of index date, level of insured amount (≤ 21 900 New Taiwan Dollars) and urbanisation level of the residence. The level of urbanisation for an individual’s residence is categorised according to the density of the population (people/km²), the proportion of agricultural workers, the proportion of people with an educational level above college, the proportion of elderly individuals aged >65 years, and the number of physicians per 100 000 subjects.²⁵ The index date for individuals without RA was defined as either the first visit or first ambulatory care for any reason. The exclusion criteria for the control group were diagnosis of acute/subacute liver necrosis, liver abscess/sequelae of chronic liver disease, other liver disorders, hepatitis B, hepatitis C, alcohol-related disorders and alcoholic liver disease before the index date.

Definition of NAFLD and AFLD

Participants who had at least three ambulatory visits or one inpatient visit with a diagnosis of NAFLD (ICD-9-CM codes 571.8, 571.9; ICD-10-CM code, K75.81) or AFLD (ICD-9-CM code 571.0; ICD-10-CM K70.0) were identified as patients with NAFLD or AFLD, respectively. The diagnosis of NAFLD was based on ultrasound evidence of fatty liver and the exclusion of both secondary causes (eg, viral hepatitis) and excessive alcohol consumption (≥ 30 g/d for male and 20 g/d for female).^{26 27} The diagnosis of AFLD was made if a patient with alcohol-use disorder presenting with hepatic steatosis on ultrasound and/or elevation in liver enzymes (aspartate aminotransferase level higher than alanine aminotransferase level) and the absence of secondary liver diseases.²⁸

Outcome

The primary outcome was the time from newly diagnosed RA to the diagnosis of NAFLD. Data were censored on 31 December 2018 (the last date of data collection) or the date of withdrawal from the NHI for any reason, whichever came first.

Potential confounders

The potential confounders for this study included age, gender, level of insured amount (as a proxy of economic status), level of urbanisation, healthcare utilisation (ie, the history of hospitalisation and number of ambulatory visits due to any reason within 1 year before the index date) and comorbidities. We considered comorbidities correlated with both RA and NAFLD as covariates, including DM, hypertension, hyperlipidaemia, obesity, cardiovascular disease, cerebral vascular disease, gallbladder disease, osteoarthritis, malignancies, depression and gout. The inclusion criteria for covariates in the individual were at least three ambulatory visits or ≥1 inpatient visit with the corresponding diagnoses based on ICD-9-CM and ICD-10-CM within 1 year before the index date.

Negative control outcome

We use negative control outcome (or falsification endpoint) to detect unmeasured and residual confounding.²⁹ After searching for a supplementary outcome variate that is not likely to be affected by the exposure in the original study, we initiated a negative control outcome analysis of AFLD, which is unlikely associated with RA according to current knowledge.

Statistical analysis

Categorical variables are presented as percentages of subjects, whereas continuous data are presented as mean±SD. We used Pearson’s χ² test to evaluate differences between categorical variables and the Student’s t-test for continuous variables. The IRs of RA-NAFLD were taken as the number of subjects newly diagnosed with RA-NAFLD per 100 000 person years (total person years (py)). The IRs, incidence rate ratios (IRRs) and their corresponding 95% CIs were calculated. To determine the association between NAFLD and RA or RA subgroups, we conducted a Kaplan-Meier (KM) estimate and log-rank test. HR with 95% CIs were then calculated from a Cox proportional hazard model after adjusting for potential confounders, such as age, gender, economic status, level of urbanisation, healthcare utilisation and selected comorbidities (diabetes, obesity and obesity-related comorbidities). Covariates with a p value<0.05 in the univariable models were investigated in the multivariable models. A two-sided p value of <0.05 was considered statistically significant.

Results

Baseline characteristics

We identified 2281 patients with RA and 91 240 matched individuals without RA for our analysis. Table 1 shows the demographic status and the comorbidities of the patients with RA and non-RA control subjects. The distributions of mean age, gender, economic status and urbanisation level were not significantly different between RA and non-RA groups. Compared with individuals without RA, patients with RA had a significantly higher frequency of hospitalisation or outpatient visits and a higher prevalence or number of comorbidities, including hypertension, obesity, osteoarthritis, depression and gout.

Table 1

Demographic status and comorbidities of patients with rheumatoid arthritis (RA) and non-RA control subjects

Variables	RA	Non-RA
	(n=2281)	(n=91 240)	P value
Age, years	52.0±15.8	52.0±15.8	1.000
Age group			1.000
Age<30	201 (8.8)	8040 (8.8)
30≤Age < 45	471 (20.6)	18 840 (20.6)
45≤Age < 65	1127 (49.4)	45 080 (49.4)
65≤Age	482 (21.1)	19 280 (21.1)
Gender			1.000
Female	1771 (77.6)	70 840 (77.6)
Male	510 (22.4)	20 400 (22.4)
Low Income (≤US$ 730 /month)	1240 (54.4)	49 600 (54.4)	1.000
Urbanisation			1.000
Urban	667 (29.2)	26 680 (29.2)
Suburban	1095 (48.0)	43 800 (48.0)
Rural	519 (22.8)	20 760 (22.8)
Medical utilisation
Number of hospitalisations	468 (20.5)	7004 (7.7)	<0.001
Number of outpatient visits, mean±SD	31.6±20.7	15.5±15.9	<0.001
Comorbidities
Number of comorbidities	0.9±1.1	0.5±0.9	<0.001
Diabetes mellitus	180 (7.9)	6934 (7.6)	0.604
Obesity	1076 (47.2)	23 557 (25.8)	<0.001
Hypertension	439 (19.2)	15 827 (17.3)	0.018
Hyperlipidaemia	171 (7.5)	6350 (7.0)	0.320
Cardiovascular disease	120 (5.3)	4339 (4.8)	0.263
Cerebral vascular accident	52 (2.3)	2558 (2.8)	0.133
Gallbladder disease	9 (0.4)	280 (0.3)	0.456
Osteoarthritis	664 (29.1)	3948 (4.3)	<0.001
Cancers	54 (2.4)	1930 (2.1)	0.409
Depression	58 (2.5)	1321 (1.4)	<0.001
Gout	191 (8.4)	1230 (1.3)	<0.001

Data are shown as mean±SD, number (%).

Comparison of incidences of NAFLD between patients with RA and individuals without RA

In this population-based cohort study, we conducted KM analysis to compare the cumulative incidence of NALFD between patients with RA and individuals without RA during the whole follow-up period (ie, 0–17 years). It showed that the cumulative incidence of NAFLD in patients with RA was not significantly different from individuals without RA (online supplemental figure S1). Before conducting Cox regression analysis, we tested the assumption of the proportional hazard model and found a significant violation of the proportional hazard assumption during the whole study period (p=0.02). Using KM estimate, the univariable and multivariable Cox regression models, we could separate the survival analyses based on a cut-off at 4 years (ie, 0–4 years and after 4 years). During the first 4 years, a total of 35 newly diagnosed patients with RA developed NAFLD during 7970 py of follow-up (IR 439 per 100 000 py). In comparison, 862 individuals without RA contracted NAFLD during 325 511 py (IR 265 per 100 000 py). The IRR was 1.66-fold greater in the RA group than in the non-RA group (95% CI 1.18 to 2.33, p<0.005, table 2). The KM analysis showed that the incidence of NAFLD was significantly higher in patients with RA than in individuals without RA during the first 4 years of follow-up (p=0.003, figure 2A) but not during the period after 4 years (figure 2B).

Figure 2. Cumulative NAFLD incidence in patients with RA during the first 4 years and the 4-17 year follow-up. RA, rheumatoid arthritis; NAFLD, non-alcoholic liver disease.

Table 2

Incidence rates of NAFLD in patients with RA and individuals without RA during the whole follow-up period

Variables	Total	Event (%)	Total PY	Incidence rate (/105 years)	IRR (95% CI)	P value
Non-alcoholic fatty liver disease
Whole follow-up period (1–17 years)
Non-RA	91 240	1751 (1.92)	786 794	224	Ref.
RA	2281	50 (2.19)	18 558	269	1.21 (0.91 to 1.60)	0.183
During the first 4 years
None RA	91 240	862 (0.94)	325 511	265	Ref.
RA	2281	35 (1.53)	7970	439	1.66 (1.18 to 2.33)	0.003
After 4 years, up to 17 years
None RA	69 650	889 (1.28)	461 282	193	Ref.
RA	1712	15 (0.88)	10 588	142	0.74 (0.44 to 1.22)	0.237
Alcoholic fatty liver disease
Whole follow-up period (1–17 years)
Non-RA	91 240	107 (0.12)	799 737	13	Ref.
RA	2281	3 (0.13)	18 921	16	1.19 (0.38 to 3.73)	0.772

IRR, incidence rate ratio; NAFLD, non-alcoholic fatty liver disease; PY, person years; RA, rheumatoid arthritis.

Comparison of AFLD incidence between patients with RA and individuals without RA

As shown in online supplemental figure S2, the incidence of AFLD was not significantly different between the RA and non-RA groups during the 17-year follow-up period. The separate KM figures for the cumulative incidences of AFLD in both groups were shown in online supplemental figure S3A (for the first 4 years) and (online supplemental figure S3B) (for the follow-up period after 4 years), both with a p value>0.05 in the log-rank test.

Risk of NAFLD in patients with RA compared with the individuals without RA

The associations between the covariates and NAFLD development were analysed using univariable and multivariable Cox proportional regression. The results are summarised in table 3. After adjusting for potential confounders, the risk of NAFLD was significantly increased in patients with RA. Variables with significant associations (p<0.05) in the univariable analysis result were used to build the multivariable regression model. Multivariable Cox regression analyses revealed that the adjusted HR (aHR) was 2.77-fold higher in early patients with RA without DMARDs therapy than in the non-RA subjects (p<0.05). Besides, other significant risk factors of NAFLD for patients with RA included old age, female gender, low economic status and obesity.

Table 3

Cox-regression analysis of associations between variables and NAFLD risk during the first 4-year follow-up period

	Univariable		Multivariable
Variables	HR (95% CI)	P value	HR (95% CI)	P value
Non-RA	Ref.		Ref.
RA	1.66 (1.18 to 2.32)	0.003
RA subgroups
RA-No DMARDs use	3.54 (1.32 to 9.44)	0.012	2.77 (1.04 to 7.42)	0.042
RA-csDMARDs alone	1.60 (1.11 to 2.32)	0.013	1.40 (0.96 to 2.03)	0.079
RA-b/tsDMARDs	1.07 (0.27 to 4.30)	0.920	0.93 (0.23 to 3.73)	0.919
Corticosteroids use#
0 mg/day	Ref.		Ref.
0<Steroid ≤ 5 mg/day	0.74 (0.38 to 1.46)	0.386
More than 5 mg/day	1.76 (0.86 to 3.62)	0.123	2.11 (1.001 to 4.45)	0.049
Age	1.02 (1.02 to 1.03)	<0.001	1.02 (1.01 to 1.02)	<0.001
Male	0.82 (0.69 to 0.97)	0.018	0.79 (0.67 to 0.94)	0.007
Low income (≦US$ 730 /month)	1.33 (1.16 to 1.53)	<0.001	1.24 (1.07 to 1.42)	0.003
Urbanisation
Urban	Ref.
Suburban	1.17 (1.001 to 1.37)	0.049
Rural	1.19 (0.99 to 1.43)	0.063
Medical utilisation
Number of hospitalisations	1.79 (1.47 to 2.18)	<0.001	1.33 (1.08 to 1.63)	0.006
Number of outpatient visits	1.80 (1.33 to 2.43)	<0.001	1.38 (1.01 to 1.87)	0.044
Comorbidity
Diabetes mellitus	1.90 (1.57 to 2.31)	<0.001
Obesity	2.09 (1.83 to 2.39)	<0.001	1.47 (1.25 to 1.73)	<0.001
Hypertension	1.82 (1.57 to 2.11)	<0.001
Hyperlipidemia	1.52 (1.22 to 1.89)	<0.001
Cardiovascular disease	1.80 (1.41 to 2.29)	<0.001
Cerebral vascular accident	1.22 (0.83 to 1.77)	0.310
Gallbladder disease	2.30 (1.03 to 5.12)	0.042
Osteoarthritis	2.14 (1.72 to 2.66)	<0.001
Cancers	2.00 (1.41 to 2.85)	<0.001
Depression	1.23 (0.75 to 2.02)	0.409
Gout	1.49 (0.96 to 2.32)	0.078

b/tsDMARDs, biologic/targeted synthetic DMARDs; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; DMARDs, disease-modifying antirheumatic drug; NAFLD, non-alcoholic fatty liver disease; RA, rheumatoid arthritis.

Discussion

A total of 2281 newly diagnosed patients with RA were identified, and other 91 240 non-RA subjects matched for age and gender in the ratio of 1:40 (figure 1) were selected from the two million Taiwanese population. In this large cohort for investigating NAFLD risk in RA, we demonstrated that patients with RA had a significantly higher IR of NAFLD (1.66-fold greater) than the non-RA subjects during the first 4 years after RA diagnosis but not after 4 years. The multivariable Cox regression analysis revealed that the aHR was 2.77-fold greater in early patients with RA not receiving DMARDs than in the non-RA subjects. Older age, female gender, low economic status and obesity were significant risk factors for developing NAFLD in newly diagnosed patients with RA. These findings suggest that newly diagnosed patients with RA, particularly those who were in the early RA stage, female, or obese, have a greater risk of developing NAFLD compared with the non-RA subjects.

Previous studies have shown that patients with RA had an elevated prevalence of MetS,^{14 30} which was also common in NAFLD.¹⁰ Besides, RA and NAFLD shared similar pathogenic cytokines, such as tumour necrosis factor (TNF)-α.^{13 31–33} These findings suggest an increased risk of NAFLD in patients with RA. Like the previous reports,^20–22 we revealed significantly higher IR (1.66, 95% CI 1.18 to 2.33) of NAFLD in newly diagnosed patients with RA than in the non-RA group in a nationwide cohort study. The KM analysis also revealed a significantly higher cumulative incidence of NAFLD in newly diagnosed RA than in the non-RA group during the first 4 years of follow-up. These observations indicate an increased risk of NAFLD in newly diagnosed patients with RA.

To avoid the influence of csDMARDs or biologic/targeted synthetic DMARDs (b/tsDMARDs) on the occurrence of NAFLD, we extracted a subgroup of patients with RA before treatment with DMARDs, and observed a further higher aHR, up to 2.77-fold, of NAFLD in this group. Although the use of non-steroidal anti-inflammatory drugs (NSAIDs), such as celecoxib and valdecoxib, could attenuate NAFLD through autophagy-mediated mechanisms,^{34 35} we revealed a further increase of NAFLD risk in patients treated with NSAIDs with/without corticosteroids. Our results also support the findings that NSAIDs may accelerate NAFLD through its enteropathy effects.³⁶ Besides, corticosteroids can potentially modify the pathological processes in all stages of NAFLD, thus increasing hepatic steatosis.³⁷ Accordingly, this study’s time-dependent Cox regression analysis revealed an elevated risk of NAFLD in patients with RA receiving glucocorticoids at a dose of >5 mg daily. However, Erre et al revealed that the use of corticosteroids was not significantly associated with moderate-to-severe hepatic steatosis in patients with RA.³⁸ Despite the conflicting results regarding the impact of corticosteroid use on NAFLD risk, we observed an elevated risk of NAFLD in early and newly diagnosed patients with RA before using the csDMARDs.

Several studies have revealed that csDMARDs use is associated with NAFLD development in patients with RA.^17–19 Sakthiswary et al revealed a cumulative dose of MTX as an independent risk factor for NAFLD with liver dysfunction.¹⁷ Mori et al also identified 51 patients, 6.0% of 846 MTX-treated patients with RA, who developed transaminitis; 42 of those had ultrasound-proven NAFLD, and 32 had histological abnormality.¹⁸ The 2021 ACR guideline for the treatment of RA recommended that MTX use should be restricted to patients with NAFLD with normal liver enzymes and liver function tests.³⁹ However, the results of previous studies showed no significant impact of MTX use on the occurrence of NAFLD²⁰ or the the prevalence of hepatic steatosis in a real-world nested case-control study.²² The time-dependent Cox regression analysis revealed non-significant lower HR of NAFLD development in our MTX-treated patients than those without MTX treatment (p=0.079). These discrepancies in the prevalence of NAFLD among MTX-treated patients are probably related to the differences in screening methods for NAFLD or characteristics of the enrolled participants with RA.

Given that both RA and NAFLD may share common pathogenic mechanisms and proinflammatory cytokines,^{13 31–33} cytokine-targeting biologics therapy may impact the prevalence of NAFLD. Wandrer et al revealed that TNF receptor-1 signalling was involved in the development of liver fibrosis in NAFLD, and the blockade of TNFR1 could reduce liver steatosis.⁴⁰ Verhoeven et al also demonstrated the treatment with TNF-α inhibitors would be safe on the risk of NAFLD in immune-mediated inflammatory disease.⁴¹ On the contrary, two previous studies did not observe a beneficial effect of TNF-α inhibitors on preventing NAFLD development.^{42 43} In the present study, we revealed no significant impact of bDMARDs, mainly TNF-α inhibitors, on NAFLD risk in patients with RA. Since the available data are scarce regarding the effects of non-TNF-α inhibitors or Janus kinase inhibitors on the risk of NAFLD, there is a need for future large cohort studies that recruit more patients with RA treated with b/tsDMARDs.

Many studies have addressed the potential factors contributing to NAFLD development.^{6–8 44 45} Herein, the multivariable regression analysis revealed the significant risk factors for NAFLD in newly diagnosed patients with RA: old age, female gender, low economic status and obesity comorbidities. The prevalence of NAFLD or NASH was higher in older patients, who had significantly more risk factors such as hypertension, obesity, diabetes and hyperlipidaemia, than in younger groups.^{6 8 45 46} Female gender was a significant predictor of NAFLD in our study. Although the data of available studies are inconclusive,^{45 47} NAFLD and NASH were reported to be more common in females.^{48 49} The prevalence of NAFLD worldwide is positively linked to economic status.⁴⁷ Our results showed that low economic status was a significant predictor of NAFLD, resonating with a study from Sri Lanka also revealed a high prevalence of NAFLD (18%) in low-income population.⁵⁰ Besides, our results were consistent with the findings of recent studies showing obesity as the significant risk factors for NAFLD in patients with RA,^{18 21} and increased NAFLD prevalence in line with obesity.^{6 7 45}

Some limitations should be addressed in the present study. The retrospective nature of our study did not allow for obtaining all the needed information from the enrolled patients. Although Taiwan NHIRD does not contain numerical data regarding individual health status (eg, body mass index (BMI) and waist circumference), we reveal obesity, based on the measured BMI, as a significant risk factors of NAFLD for patients with RA. The results were consistent with previous findings that obesity was closely associated with NAFLD.^{51 52} Our NHIRD neither contains detailed status of smoking that is significantly related to NAFLD risk.⁵³ The 1:40 matching of the groups simply based on two factors (age and gender) is also a limitation. In addition, the absence of data regarding individual RA disease activity, which was related to the risk of NAFLD, is another important limitation. Future large, hospital-based studies are required to validate our findings.

The major strength of this study was the utilisation of the nationwide database, which provides detailed medical care records and is widely deemed a reliable instrument for epidemiological research.⁴⁹ The NHIRD contains detailed pharmacy claims, registration files and original claims records of reimbursement for each study subject, with the diagnoses based on ICD-9-CM and ICD-10-CM codes. The large sample size of the NHIRD (23 million enrollees) and a long-term (17 years) design enhance this study’s statistical power and accuracy. We also evaluated the effect of commonly used medications on the risk of NAFLD in newly diagnosed patients with RA (online supplemental table S1).

Conclusion

An elevated risk of NAFLD was observed in patients with RA during the first 4 years after RA diagnosis, particularly in those before treatment with DMARDs. The old age, women, low-income status, and obesity were significant predictors of NAFLD development.

The study is based on data from the NHIRD, which is provided by the National Health Insurance Administration and the Ministry of Health and Welfare and managed by the National Health Research Institutes. The interpretation and conclusions do not represent those of the National Health Insurance Administration, the Ministry of Health and Welfare, or the National Health Research Institutes. The authors would like to thank the Biostatistics Task Force of Taichung Veterans General Hospital, Taichung, Taiwan, for statistical support.

Data availability statement

Data may be obtained from a third party and are not publicly available. All data relevant to the study are included in the article or uploaded as supplementary materials.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

The present study has been approved by the Institutional Review Board (IRB) of Taichung Veterans General Hospital (IRB number: CE20295A). The requirement for informed consent was waived because personal details were completely anonymised before data analysis.

Footnote

W-LH and H-HC contributed equally.

Contributors All authors made substantive intellectual contributions to the present study and approved the final manuscript. W-LH and H-HC conceived of the study, designed the study, supported the NHIRD application for the data extraction, and analysed data. P-KC and T-LL conceived of the study, analysed data with statistical analysis, and editing the manuscript. S-HC and Y-MC contributed to cleaning and analysing the extracted data. C-HL and K-TT analysed data with statistical analysis. D-YC generated the original hypothesis, conceived of the study, analysed data, drafted and revised the manuscript.

Funding This study received support from the grants of TCVGH-VHCY118604, TCVGH-VHCY1088605, TCVGH-VHCY118604, TCVGH-VHCY1128603.

Competing interests None declared.

Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ Chalasani N, Younossi Z, Lavine JE, et al. The diagnosis and management of non-alcoholic fatty liver disease: practice guideline by the American Association for the study of liver diseases,American college of Gastroenterology, and the American Gastroenterological Association. Hepatology 2012; 55: 2005–23. doi:10.1002/hep.25762

² Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural history of Non‐Alcoholic fatty liver disease and Non‐Alcoholic Steatohepatitis in adults. Aliment Pharmacol Ther 2011; 34: 274–85. doi:10.1111/j.1365-2036.2011.04724.x

³ Arab JP, Arrese M, Trauner M. Recent insights into the pathogenesis of Nonalcoholic fatty liver disease. Annu Rev Pathol 2018; 13: 321–50. doi:10.1146/annurev-pathol-020117-043617

⁴ Lonardo A, Nascimbeni F, Targher G, et al. AISF position paper on Nonalcoholic fatty liver disease (NAFLD): updates and future directions. Digestive and Liver Disease 2017; 49: 471–83. doi:10.1016/j.dld.2017.01.147

⁵ Farrell GC, Wong VW-S, Chitturi S. NAFLD in Asia--as common and important as in the West. Nat Rev Gastroenterol Hepatol 2013; 10: 307–18. doi:10.1038/nrgastro.2013.34

⁶ Park SH, Jeon WK, Kim SH, et al. Prevalence and risk factors of non-alcoholic fatty liver disease among Korean adults. J of Gastro and Hepatol 2006; 21: 138–43. doi:10.1111/j.1440-1746.2005.04086.x

⁷ Fan JG, Farrell GC. Epidemiology of non-alcoholic fatty liver disease in China. Journal of Hepatology 2009; 50: 204–10. doi:10.1016/j.jhep.2008.10.010

⁸ Chen C, Huang M, Yang J, et al. Prevalence and etiology of elevated serum alanine aminotransferase level in an adult population in Taiwan. J of Gastro and Hepatol 2007; 22: 1482–9. doi:10.1111/j.1440-1746.2006.04615.x

⁹ Cusi K. Role of obesity and Lipotoxicity in the development of Nonalcoholic Steatohepatitis: pathophysiology and clinical implications. Gastroenterology 2012; 142: 711–25. doi:10.1053/j.gastro.2012.02.003

¹⁰ Hamaguchi M, Kojima T, Takeda N, et al. The metabolic syndrome as a Predictor of Nonalcoholic fatty liver disease. Ann Intern Med 2005; 143: 722. doi:10.7326/0003-4819-143-10-200511150-00009

¹¹ Lomonaco R, Ortiz-Lopez C, Orsak B, et al. Effect of Adipose tissue insulin resistance on metabolic parameters and liver histology in obese patients with Nonalcoholic fatty liver disease. Hepatology 2012; 55: 1389–97. doi:10.1002/hep.25539

¹² Hui Y, Yu-Yuan L, Yu-Qiang N, et al. Effect of peroxisome Proliferator-activated receptors-Γ and Co-Activator-1Α genetic Polymorphisms on plasma adiponectin levels and susceptibility of non-alcoholic fatty liver disease in Chinese people. Liver Int 2008; 28: 385–92. doi:10.1111/j.1478-3231.2007.01623.x

¹³ Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis. Lancet 2016; 388: 2023–38. doi:10.1016/S0140-6736(16)30173-8

¹⁴ Cai W, Tang X, Pang M. n.d. Prevalence of metabolic syndrome in patients with rheumatoid arthritis: an updated systemic review and meta-analysis. Front Med; 9. doi:10.3389/fmed.2022.855141

¹⁵ Siebert S, Lyall DM, Mackay DF, et al. Characteristics of rheumatoid arthritis and its association with major comorbid conditions: cross-sectional study of 502 649 UK Biobank participants. RMD Open 2016; 2: e000267. doi:10.1136/rmdopen-2016-000267

¹⁶ Nicolau J, Lequerré T, Bacquet H, et al. Rheumatoid arthritis, insulin resistance, and diabetes. Joint Bone Spine 2017; 84: 411–6. doi:10.1016/j.jbspin.2016.09.001

¹⁷ Sakthiswary R, Chan GYL, Koh ET, et al. Methotrexate-associated Nonalcoholic fatty liver disease with Transaminitis in rheumatoid arthritis. ScientificWorldJournal 2014; 2014: 823763. doi:10.1155/2014/823763

¹⁸ Mori S, Arima N, Ito M, et al. Non-alcoholic Steatohepatitis-like pattern in liver biopsy of rheumatoid arthritis patients with persistent Transaminitis during low-dose methotrexate treatment. PLoS ONE 2018; 13: e0203084. doi:10.1371/journal.pone.0203084

¹⁹ Ezhilarasan D. Hepatotoxic potentials of methotrexate: understanding the possible Toxicological molecular mechanisms. Toxicology 2021; 458. doi:10.1016/j.tox.2021.152840

²⁰ Wagan AA, Bhutoo AQ, Khan D, et al. Fatty liver in Pakistani cohort with rheumatoid arthritis. Pak J Med Sci 2020; 36: 723–8. doi:10.12669/pjms.36.4.1984

²¹ Wu T, Zou YW, Ma JD, et al. The characteristics of non-alcoholic fatty liver disease and its associated factors in patients with rheumatoid arthritis. Zhonghua Yu Fang Yi Xue Za Zhi 2022; 56: 574–82. doi:10.3760/cma.j.cn112150-20210706-00647

²² Choi Y, Lee CH, Kim IH, et al. Methotrexate use does not increase the prevalence of hepatic steatosis: a real-world retrospective nested case-control study. Clin Rheumatol 2021; 40: 2037–45. doi:10.1007/s10067-020-05456-y

²³ Arnett FC, Edworthy SM, Bloch DA, et al. The American rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 315–24. doi:10.1002/art.1780310302

²⁴ Aletaha D, Neogi T, Silman AJ, et al. Rheumatoid arthritis classification criteria: an American college of rheumatology/European League against rheumatism collaborative initiative. Annals of the Rheumatic Diseases 2010; 69: 1580–8. doi:10.1136/ard.2010.138461

²⁵ Lin YJ, Tian WH, Chen CC. Urbanization and the utilization of outpatient services under national health insurance in Taiwan. Health Policy 2011; 103: 236–43. doi:10.1016/j.healthpol.2011.08.007

²⁶ Ratziu V, Bellentani S, Cortez-Pinto H, et al. Marchesini G: A position statement on NAFLD/NASH based on the EASL 2009 special conference. J Hepatol 2010; 53: 372–84. doi:10.1016/j.jhep.2010.04.008

²⁷ Lin S, Huang J, Wang M, et al. Comparison of MAFLD and NAFLD diagnostic criteria in real world. Liver Int 2020; 40: 2082–9. doi:10.1111/liv.14548

²⁸ Singal AK, Bataller R, Ahn J, et al. ACG clinical guideline: alcoholic liver disease. Am J Gastroenterol 2018; 113: 175–94. doi:10.1038/ajg.2017.469

²⁹ Nørgaard M, Ehrenstein V, Vandenbroucke JP. Confounding in observational studies based on large health care databases: problems and potential solutions-a primer for the clinician. Clin Epidemiol 2017; 9: 185–93. doi:10.2147/CLEP.S129879

³⁰ Hallajzadeh J, Safiri S, Mansournia MA, et al. Metabolic syndrome and its components among rheumatoid arthritis patients: A comprehensive updated systematic review and meta-analysis. PLoS One 2017; 12: e0170361. doi:10.1371/journal.pone.0170361

³¹ Furst DE, Emery P. Rheumatoid arthritis pathophysiology: update on emerging cytokine and cytokine-associated cell targets. Rheumatology (Oxford) 2014; 53: 1560–9. doi:10.1093/rheumatology/ket414

³² Crespo J, Cayón A, Fernández-Gil P, et al. Gene expression of tumor necrosis factor alpha and TNF-receptors, P55 and P75, in Nonalcoholic Steatohepatitis patients. Hepatology 2001; 34: 1158–63. doi:10.1053/jhep.2001.29628

³³ Kugelmas M, Hill DB, Vivian B, et al. Cytokines and NASH: a pilot study of the effects of lifestyle modification and vitamin E. Hepatology 2003; 38: 413–9. doi:10.1053/jhep.2003.50316

³⁴ Liu C, Liu L, Zhu H-D, et al. Celecoxib Alleviates Nonalcoholic fatty liver disease by restoring Autophagic Fux. Sci Rep 2018; 8: 4108. doi:10.1038/s41598-018-22339-0

³⁵ Park SY, Cho W, Abd El-Aty AM, et al. Valdecoxib attenuates lipid-induced hepatic steatosis through Autophagy-mediated suppression of Endoplasmic Reticulum stress. Biochem Pharmacol 2022; 199. doi:10.1016/j.bcp.2022.115022

³⁶ Utzeri E, Usai P. Role of non-Steroidal anti-inflammatory drugs on intestinal permeability and Nonalcoholic fatty liver disease. World J Gastroenterol 2017; 23: 3954–63. doi:10.3748/wjg.v23.i22.3954

³⁷ Woods CP, Hazlehurst JM, Tomlinson JW. Glucocorticoids and non-alcoholic fatty liver disease. The Journal of Steroid Biochemistry and Molecular Biology 2015; 154: 94–103. doi:10.1016/j.jsbmb.2015.07.020

³⁸ Erre GL, Castagna F, Sauchella A, et al. Prevalence and risk factors of moderate to severe hepatic steatosis in patients with rheumatoid arthritis: an Ultrasonography cross-sectional case–control study. Ther Adv Musculoskelet Dis 2021; 13. doi:10.1177/1759720X211042739

³⁹ Fraenkel L, Bathon JM, England BR, et al. The 2021 American college of rheumatology guideline for the treatment of rheumatoid arthritis. Arthritis Rheumatol 2021; 73: 1108–23. doi:10.1002/art.41752

⁴⁰ Wandrer F, Liebig S, Marhenke S, et al. TNF-receptor-one inhibition reduces liver steatosis, hepatocellular injury and Fbrosis in NAFLD mice. Cell Death Dis 2020; 11: 212. doi:10.1038/s41419-020-2411-6

⁴¹ Verhoeven F, Weil-Verhoeven D, Prati C, et al. Safety of TNF inhibitors in rheumatic disease in case of NAFLD and cirrhosis. Semin Arthritis Rheum 2020; 50: 544–8. doi:10.1016/j.semarthrit.2020.03.013

⁴² Feagins LA, Flores A, Arriens C, et al. Nonalcoholic fatty liver disease: a potential consequence of tumor necrosis factor-inhibitor therapy. Eur J Gastroenterol Hepatol 2015; 27: 1154–60. doi:10.1097/MEG.0000000000000421

⁴³ Zekić T, Benić MS, Radić M. Treatment of rheumatoid arthritis with conventional, targeted, and biological disease-modifying Antirheumatic drugs in the setting of liver injury and non-alcoholic fatty liver disease. Rheumatol Int 2022; 42: 1665–79. doi:10.1007/s00296-022-05143-y

⁴⁴ Fu C-C, Chen M-C, Li Y-M, et al. The risk factors for ultrasound-diagnosed non-alcoholic fatty liver disease among adolescents. Ann Acad Med Singap 2009; 38: 15–7.

⁴⁵ Younossi ZM. Non-alcoholic fatty liver disease: a global public health perspective. J Hepatol 2019; 70: 531–44. doi:10.1016/j.jhep.2018.10.033

⁴⁶ Frith J, Day CP, Henderson E, et al. Non-alcoholic fatty liver disease in older people. Gerontology 2009; 55: 607–13. doi:10.1159/000235677

⁴⁷ Zhu J-Z, Dai Y-N, Wang Y-M, et al. Prevalence of Nonalcoholic fatty liver disease and economy. Dig Dis Sci 2015; 60: 3194–202. doi:10.1007/s10620-015-3728-3

⁴⁸ Summart U, Thinkhamrop B, Chamadol N, et al. Gender differences in the prevalence of Nonalcoholic fatty liver disease in the northeast of Thailand: a population-based cross-sectional study. F1000Res 2017; 6: 1630. doi:10.12688/f1000research.12417.2

⁴⁹ Younossi Z, Tacke F, Arrese M, et al. Global perspectives on non-alcoholic fatty liver disease and non-alcoholic Steatohepatitis. Hepatology 2019; 69: 2672–82. doi:10.1002/hep.30251

⁵⁰ Pinidiyapathirage MJ, Dassanayake AS, Rajindrajith S, et al. Non-alcoholic fatty liver disease in a rural, physically active, low-income population in Sri Lanka. BMC Res Notes 2011; 4: 513. doi:10.1186/1756-0500-4-513

⁵¹ Negi CK, Babica P, Bajard L, et al. Insights into the molecular targets and emerging Pharmacotherapeutic interventions for Nonalcoholic fatty liver disease. Metabolism 2022; 126: 154925. doi:10.1016/j.metabol.2021.154925

⁵² Tang X, Shi Y, Du J, et al. Clinical outcome of non-alcoholic fatty liver disease: an 11-year follow up study. BMJ Open 2022; 12: e054891. doi:10.1136/bmjopen-2021-054891

⁵³ Rezayat AA, Moghadam MD, Nour MG, et al. Association between smoking and non-alcoholic-fatty liver disease: a systemic review and meta-analysis. SAGE Open Med 2018; 6.