The risk of developing a new autoimmune inflammatory rheumatic disease (AIRD) is greater following a COVID-19 infection than after an influenza infection or in the general population, according to a study published March 5 in Annals of Internal Medicine . More severe COVID-19 infections were linked to a greater risk of incident rheumatic disease, but vaccination appeared protective against development of a new AIRD.
“Importantly, this study shows the value of vaccination to prevent severe disease and these types of sequelae,” Anne Davidson , MBBS, a professor in the Institute of Molecular Medicine at The Feinstein Institutes for Medical Research in Manhasset, New York, who was not involved in the study, said in an interview.
Dr. Anne Davidson
Previous research had already identified the likelihood of an association between SARS-CoV-2 infection and subsequent development of a new AIRD . This new study, however, includes much larger cohorts from two different countries and relies on more robust methodology than previous studies, experts said.
“Unique steps were taken by the study authors to make sure that what they were looking at in terms of signal was most likely true,” Alfred Kim , MD, PhD, assistant professor of medicine in rheumatology at Washington University in St. Louis, who was not involved in the study, said in an interview. Dr. Davidson agreed, noting that these authors “were a bit more rigorous with ascertainment of the autoimmune diagnosis, using two codes and also checking that appropriate medications were administered.”
More Robust and Rigorous Research
Past cohort studies finding an increased risk of rheumatic disease after COVID-19 “based their findings solely on comparisons between infected and uninfected groups, which could be influenced by ascertainment bias due to disparities in care, differences in health-seeking tendencies, and inherent risks among the groups,” Min Seo Kim , MD, of the Broad Institute of MIT and Harvard, Cambridge, Massachusetts, and his colleagues reported. Their study, however, required at least two claims with codes for rheumatic disease and compared patients with COVID-19 to those with flu “to adjust for the potentially heightened detection of AIRD in SARS-CoV-2–infected persons owing to their interactions with the health care system.”
Dr. Alfred Kim said the fact that they used at least two claims codes “gives a little more credence that the patients were actually experiencing some sort of autoimmune inflammatory condition as opposed to a very transient issue post COVID that just went away on its own.”
Dr. Alfred Kim
He acknowledged that the previous research was reasonably strong, “especially in light of the fact that there has been so much work done on a molecular level demonstrating that COVID-19 is associated with a substantial increase in autoantibodies in a significant proportion of patients, so this always opened up the possibility that this could associate with some sort of autoimmune disease downstream.”
While the study is well done with a large population, “it still has limitations that might overestimate the effect,” Kevin W. Byram , MD, associate professor of medicine in rheumatology and immunology at Vanderbilt University Medical Center in Nashville, Tennessee, who was not involved in the study, said in an interview. “We certainly have seen individual cases of new rheumatic disease where COVID-19 infection is likely the trigger,” but the phenomenon is not new, he added.
“Many autoimmune diseases are spurred by a loss of tolerance that might be induced by a pathogen of some sort,” Dr. Byram said. “The study is right to point out different forms of bias that might be at play. One in particular that is important to consider in a study like this is the lack of case-level adjudication regarding the diagnosis of rheumatic disease” since the study relied on available ICD-10 codes and medication prescriptions.
Dr. Kevin W. Byram
The researchers used national claims data to compare risk of incident AIRD in 10,027,506 South Korean and 12,218,680 Japanese adults, aged 20 and older, at 1 month, 6 months, and 12 months after COVID-19 infection, influenza infection, or a matched index date for uninfected control participants. Only patients with at least two claims for AIRD were considered to have a new diagnosis.
Patients who had COVID-19 between January 2020 and December 2021, confirmed by PCR or antigen testing, were matched 1:1 with patients who had test-confirmed influenza during that time and 1:4 with uninfected control participants, whose index date was set to the infection date of their matched COVID-19 patient.
The propensity score matching was based on age, sex, household income, urban versus rural residence, and various clinical characteristics and history: body mass index; blood pressure; fasting blood glucose; glomerular filtration rate; smoking status; alcohol consumption; weekly aerobic physical activity; comorbidity index; hospitalizations and outpatient visits in the previous year; past use of diabetes, hyperlipidemia, or hypertension medication; and history of cardiovascular disease, chronic kidney disease, chronic obstructive pulmonary disease, or respiratory infectious disease.
Patients with a history of AIRD or with coinfection or reinfection of COVID-19 and influenza were excluded, as were patients diagnosed with rheumatic disease within a month of COVID-19 infection.
Risk Varied With Disease Severity and Vaccination Status
Among the Korean patients, 3.9% had a COVID-19 infection and 0.98% had an influenza infection. After matching, the comparison populations included 94,504 patients with COVID-19 versus 94,504 patients with flu, and 177,083 patients with COVID-19 versus 675,750 uninfected controls.
The risk of developing an AIRD at least 1 month after infection in South Korean patients with COVID-19 was 25% higher than in uninfected control participants (adjusted hazard ratio [aHR], 1.25; 95% CI, 1.18–1.31; P < .05) and 30% higher than in influenza patients (aHR, 1.3; 95% CI, 1.02–1.59; P < .05). Specifically, risk in South Korean patients with COVID-19 was significantly increased for connective tissue disease and both treated and untreated AIRD but not for inflammatory arthritis.
Among the Japanese patients, 8.2% had COVID-19 and 0.99% had flu, resulting in matched populations of 115,003 with COVID-19 versus 110,310 with flu, and 960,849 with COVID-19 versus 1,606,873 uninfected patients. The effect size was larger in Japanese patients, with a 79% increased risk for AIRD in patients with COVID-19, compared with the general population (aHR, 1.79; 95% CI, 1.77–1.82; P < .05) and a 14% increased risk, compared with patients with influenza infection (aHR, 1.14; 95% CI, 1.10–1.17; P < .05). In Japanese patients, risk was increased across all four categories, including a doubled risk for inflammatory arthritis (aHR, 2.02; 95% CI, 1.96–2.07; P < .05), compared with the general population.
The researchers had data only from the South Korean cohort to calculate risk based on vaccination status, SARS-CoV-2 variant (wild type versus Delta), and COVID-19 severity. Researchers determined a COVID-19 infection to be moderate-to-severe based on billing codes for ICU admission or requiring oxygen therapy, extracorporeal membrane oxygenation, renal replacement, or CPR.
Infection with both the original strain and the Delta variant were linked to similar increased risks for AIRD, but moderate to severe COVID-19 infections had greater risk of subsequent AIRD (aHR, 1.42; P < .05) than mild infections (aHR, 1.22; P < .05). Vaccination was linked to a lower risk of AIRD within the COVID-19 patient population: One dose was linked to a 41% reduced risk (HR, 0.59; P < .05) and two doses were linked to a 58% reduced risk (HR, 0.42; P < .05), regardless of the vaccine type, compared with unvaccinated patients with COVID-19. The apparent protective effect of vaccination was true only for patients with mild COVID-19, not those with moderate to severe infection.
“One has to wonder whether or not these people were at much higher risk of developing autoimmune disease that just got exposed because they got COVID, so that a fraction of these would have gotten an autoimmune disease downstream,” Dr. Alfred Kim said. Regardless, one clinical implication of the findings is the reduced risk in vaccinated patients, regardless of the vaccine type, given the fact that “mRNA vaccination in particular has not been associated with any autoantibody development,” he said.
Though the correlations in the study cannot translate to causation, several mechanisms might be at play in a viral infection contributing to autoimmune risk, Dr. Davidson said. Given that viral nucleic acids also recognize self-nucleic acids, “a large load of viral nucleic acid may break tolerance,” or “viral proteins could also mimic self-proteins,” she said. “In addition, tolerance may be broken by a highly inflammatory environment associated with the release of cytokines and other inflammatory mediators.”
The association between new-onset autoimmune disease and severe COVID-19 infection suggests multiple mechanisms may be involved in excess immune stimulation, Dr. Davidson said. But she added that it’s unclear how these findings, involving the original strain and Delta variant of SARS-CoV-2, might relate to currently circulating variants.
The research was funded by the National Research Foundation of Korea, the Korea Health Industry Development Institute, and the Ministry of Food and Drug Safety of the Republic of Korea. The authors reported no relevant financial relationships with industry. Dr. Alfred Kim has sponsored research agreements with AstraZeneca, Bristol-Myers Squibb, and Novartis; receives royalties from a patent with Kypha Inc.; and has done consulting or speaking for Amgen, ANI Pharmaceuticals, Aurinia Pharmaceuticals, Exagen Diagnostics, GlaxoSmithKline, Kypha, Miltenyi Biotech, Pfizer, Rheumatology & Arthritis Learning Network, Synthekine, Techtonic Therapeutics, and UpToDate. Dr. Byram reported consulting for TenSixteen Bio. Dr. Davidson had no disclosures.