This population-based, cross-sectional study extracted data from the US Nationwide Inpatient Sample (NIS) database, which is the largest all-payer, inpatient care database in the US, including about 8 million hospital stays each year (19). The database is administered by the Healthcare Cost and Utilization Project (HCUP) of the US National Institutes of Health (NIH) and developed through a Federal-State-Industry partnership and sponsored by the Agency for Healthcare Research and Quality (AHRQ). All admitted patients are initially considered for inclusion. The continuous, annually updated NIS database derives patient data of about 1,050 hospitals from 44 States in the US, representing a 20% stratified sample of the US community hospitals as defined by the American Hospital Association.

All data were obtained through request to the Online HCUP Central Distributor (available at: https://www.distributor.hcup-us.ahrq.gov/), which administers the database (certificate # HCUP-833FWV78H). This study conforms to the NIS data-use agreement with HCUP. As this study analyzed secondary data of the NIS database, patients' data were not used directly used. This study was approved by the Institutional Review Board of the Chang Gung Medical Hospital (202100234B0). As we conducted this retrospective analysis of the NHIRD and all the individual information were de-identified, we could not obtain informed consent from the recruited patients.

Adults ≥20 years old admitted to US hospitals between 2005 and 2014 with a primary or secondary diagnosis of SLE or RA were identified in the NIS database through the International Classification of Diseases, Ninth Revision (ICD-9) diagnostic codes (code 710.0 or 714.0). Patients with congenital heart disease, asthma, HIV, heart surgery, and cancer were excluded. Patients with incomplete data for outcomes and main variables of interest were also excluded from the study cohort.

HF was identified by ICD-9 code 428.x, which covered any type of HF regardless of ejection or reduced ejection fractions. MI was identified through code 410.x, which covered all types of MI. Since no unique code for “sarcopenia” is included in the ICD-9 coding system, sarcopenia was defined using the code 728.2x (muscular wasting and disuse atrophy) in accordance with a previous study (20).

Patients' baseline characteristics included age, gender, race (ethnicity), household income level, insurance status (primary payer), admission type, smoking, alcohol consumption, diabetes, hypertension, and hyperlipidemia. In addition, hospital characteristics, including hospital bed size, location, hospital region, and length of stay, were also extracted from the database as part of the comprehensive data available for all patients in the NIS database. Length of stay is calculated by subtracting the admission date from the discharge date. Details on hospital characteristics are documented on the NIS webpage (https://www.hcup-us.ahrq.gov/db/vars/hosp_locteach/nisnote.jsp).

Descriptive statistics of the patients are presented as unweighted counts (n) and weighted percentage (%) or mean ± standard error (SE). Differences between the groups were evaluated using PROC SURVEYFREQ and SURVEYREG for analysis of categorical and continuous data. Logistic regression analysis was performed using PROC SURVEYLOGISTIC to determine factors associated with the presence of HF and MI. Variables that were significant in univariate regression analysis were adjusted in multivariable analysis. Since the NIS database covers 20% of samples of the USA annual inpatient admissions, weighted samples (DISCWT), stratum (NIS_STRATUM), cluster (HOSPID) were used to generate national estimates for all analyses. All p-values were two-sided and p < 0.05 was considered statistically significant. All statistical analyses were carried out with SAS 9.4. A two-tailed P-value < 0.05 was considered statistically significant.

The detailed patient selection process is shown in Figure 1. Data of 1,225,241 patients aged 20 years or older who were diagnosed with SLE, RA or both were extracted. After excluding patients with congenital heart diseases (n = 3,906), asthma (n = 127,207), HIV (n = 1,334), previous cardiac procedure (n = 57,216), cancer history (n = 100,882), and incomplete data (n = 153,497), finally 781,199 patients (representing 3,865,575 US inpatients) were included as the study cohort (Figure 1).

A total of 0.1% patients had sarcopenia, 25.6% had SLE alone, 71.2% had RA alone, and 3.2% had overlapping diagnoses between SLE and RA. Mean age was 63.2, and the majority of patients were female (79.1%), White (71.0%), non-smokers (99.9%) and did not consume alcohol (99.6%). Most patients were admitted emergently (77.8%), with income levels at the lowest quartile (29.6%), and insurance covered by Medicare/Medicaid (72.6%) (Table 1).

Characteristics of patients with or without HF are summarized in Table 1. Among all patients, 127,812 (16.4%) had HF and 653,387 (83.6%) did not. Patients with HF had higher proportions of sarcopenia (0.13% vs. 0.10%, p-value = 0.003) and RA (75.9% vs. 70.3%, p < 0.001) than those without HF. Also, HF patients were older (74.1 vs. 61.7 years, p < 0.001), stayed longer in the hospital (6.2 vs. 4.9 days, p-value < 0.001), included more males (24.7% vs. 20.2%), were White (71.8% vs. 70.8%) or Black race (18.2% vs. 15.4%), had lower household income (Q1:31.9% vs. 29.1%; Q2: 26.0% vs. 25.7%), insurance covered by Medicare/Medicaid (87.0% vs. 69.7%), emergent admission (89.0% vs. 75.6%), were non-drinkers (99.8% vs. 99.5%) and had more diabetes, hypertension and hyperlipidemia (all p-value < 0.001) (Table 1).

Characteristics of patients with or without MI are summarized in Table 2. Among all patients, 12,781 (1.6%) had MI, and 768,418 (98.4%) did not. RA occurred more frequently in MI patients (79.6% vs. 71.1%, p-value < 0.001) than in those without MI. The incidence of sarcopenia was not significantly different between patients with or without MI (Table 2).

As shown in Table 3, prevalence of sarcopenia among HF patients was 0.1, 0.1, and 0.2% among patients diagnosed with SLE alone, RA alone, and both SLE and RA, respectively. In addition, prevalence of sarcopenia among MI patients was 0.1, 0.1 and 0% among patients diagnosed with SLE alone, RA alone and both, respectively.

The results of univariate and multivariate regression analysis on the associations between sarcopenia, type of autoimmune disease and HF are summarized in Table 4. After adjusting for relevant confounders, no significant associations were found between sarcopenia and odds for HF (aOR: 1.12, 95% CI: 0.94–1.33). Patients with RA had significantly lower odds of HF occurrence (aOR = 0.77, 95% CI: 0.76–0.79) than SLE patients.

The results of univariate and multivariate regression analysis on associations between sarcopenia, type of autoimmune disease and MI are summarized in Table 5. Univariate analysis revealed no associations between sarcopenia and odds for MI. After adjusting for relevant confounders, RA patients had significantly lower odds for MI (aOR = 0.86, 95% CI: 0.82–0.91) than SLE patients.

The major strengths of this study were the use of the large national representative database of the US, which provided good statistical power. In addition, traditional risk factors for MI or HF were considered and carefully adjusted. However, the study still has several limitations. First, the study uses secondary cross-sectional data and retrospective analysis, both of which may limit generalization to other populations or locations outside the US and prevent long-term follow-up of patients. Also, the lack of information regarding disease duration or activity for SLE and RA populations hindered further adjustment in the analyses. Subgroup analyses based on different severity levels, ejection fraction type and stage of HF, or types of MI (i.e., STEMI, NSTEMI) were also not performed. Lastly, although important, medications used and laboratory parameter values were not provided for patients in the NIS database and could not be included in data analysis, which may have shed light on individual differences between SLE/RA patients. Prospective studies are needed to help elucidate such differences.