Abstract
Objective
To determine the prevalence of comorbidities in RA, discover which comorbidities might predispose to developing RA, and identify which comorbidities are more likely to develop after RA.
Patients and Methods
We performed a case-control study using a single-center biobank, identifying 821 cases of RA (143 incident RA) between January 1 2009 and February 28 2018, defined as two diagnosis codes plus a disease-modifying antirheumatic drug. We matched each case to three controls based on age and sex. Participants self-reported presence and onset of 74 comorbidities. Logistic regression models adjusted for race, body mass index (BMI), education, smoking, and Charlson comorbidity index.
Results
After adjustment for confounders and multiple comparisons, eleven comorbidities were associated with RA, including epilepsy (OR 2.13, p=.009), obstructive sleep apnea (OSA) (OR 1.49, p=.001), and pulmonary fibrosis (OR 4.63, p<.001), but cancer was not. Inflammatory bowel disease (IBD) (OR 3.82, p<.001), type 1 diabetes (OR 3.07, p=.01), and venous thromboembolism (VTE) (OR 1.80, p<.001), occurred more often before RA diagnosis compared to controls. In contrast, myocardial infarction (OR 3.09, p<.001) and VTE (OR 1.84, p<.001) occurred more often after RA diagnosis compared to controls. Analyses restricted to incident RA cases and their matched controls mirrored these results.
Conclusion
IBD, type 1 diabetes, and VTE might predispose to RA development, while cardiovascular disease, VTE, and OSA may result from RA. These findings have important implications for RA pathogenesis, early detection, and recommended screening.
Introduction
Excess comorbidity is associated with poorer outcomes in rheumatoid arthritis (RA) patients, including worse physical disability,1 functional decline,2 lower remission rates,3 poorer quality of life,4 and increased mortality2,5. Understanding the relationship between RA and comorbidities might help direct interventions to improve these outcomes. Furthermore, better understanding individual comorbidities and when they develop relative RA might provide important clues to disease pathogenesis.6
While several large-scale studies of comorbidities in RA patients have been performed,7–9 several gaps in knowledge remain. First, most existing studies are cross-sectional studies with no information on the timing of comorbidity development relative to RA. Second, a few studies have examined whether RA predisposes to individual comorbidities, but were unable to account for confounding factors such as smoking. Further, results from these studies sometimes conflict. For example, some studies show an increase in obstructive sleep apnea (OSA) among RA patients,10,11 while others do not.12 Finally, only a handful of studies have investigated whether certain comorbidities might predispose to RA despite existing evidence that depression,13 OSA,14 inflammatory bowel disease (IBD),15 and atopic dermatitis16 may increase risk of developing RA.
To address these gaps, we leveraged the Mayo Clinic Biobank, which contains data on seventy-four comorbidities and their age of onset.17 Our aims were threefold. First, we aimed to assess the prevalence of a wide array of comorbidities in patients with RA. Second, we aimed to determine which comorbidities might predispose to developing RA. Finally, we aimed to identify which comorbidities may be more likely to develop after RA. We hypothesized that pulmonary diseases and other autoimmune diseases would increase the risk of developing RA, while RA would predispose to cardiovascular disease and psychiatric diseases.
Patients and Methods
Study Design
This case-control study used questionnaire data from the Mayo Clinic Biobank.17 It followed a pre-determined protocol, received approval from the Mayo Clinic and Olmsted County Institutional Review Boards (17–010806; 060-OMC-17) which waived the need for informed consent, and complies with the Declaration of Helsinki. This manuscript also follows the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) reporting guidelines for observational studies.18 To improve quality of questionnaire data, all questionnaires were visually examined for errors and omissions. Those with more than ten errors were returned to participants for correction. A computer program also evaluated for logical errors and incorrect skip patterns, flagging records for manual verification.
Participants
The biobank’s 55,898 participants were recruited mainly from primary care locations at the Mayo Clinic in Minnesota and Florida starting in 2009. Eligibility criteria included age 18 or older, ability to communicate in English, capacity to consent, and residence in the United States. Approximately 29% of those invited chose to participate and completed a baseline questionnaire. Of those, 77% completed the follow-up questionnaire sent approximately four years later.
RA Cases
We identified RA cases using a rules-based algorithm combining two occurrences of RA diagnosis codes (714.0 or 714.9) at least 30 days apart plus use of a disease-modifying antirheumatic drug (DMARD) occurring at any point in the electronic health record (EHR) since it became available in 1999.19 This definition had a positive predictive value (PPV) of 95% for RA by American College of Rheumatology 2010 criteria.20 Of the identified cases of RA, 143 had no RA self-report on the baseline or follow-up questionnaire but met the criteria for RA based on two diagnosis codes and a DMARD at least 90 days after the most recent questionnaire (presumed to be incident RA). Seropositive RA was defined as positivity for either rheumatoid factor (RF) or cyclic citrullinated peptide (CCP) antibodies, collected from the EHR.
We defined the index date as the date of RA diagnosis. Age at RA diagnosis came mainly from the self-reported onset of RA (≤ 19, 20–49, 50–64, 65–79, or 80+ years), using the midpoint of the selected age category (33%), the midpoint between the start of the age category and age at the baseline questionnaire (48%), or age at the second RA diagnosis code for incident RA cases or those cases where the first two diagnosis codes for RA occurred in an earlier age category than self-report (19%).
Controls
Controls included participants without self-reported RA and no diagnosis codes for RA. We matched each RA case to three controls based on age at the time of baseline questionnaire (within 5 years), sex, recruitment year (within five years), recruitment location (Minnesota or Florida), and distance from recruitment location (within 500 miles). The index date for controls was assigned as the same date from the matched case.
Comorbidities
This study’s 74 comorbidities were self-reported on the baseline questionnaire, along with age of diagnosis. Any comorbidities that developed in an age window before the index date of RA diagnosis were included in the “before RA” analyses. Similarly, those that developed in an age window after the index date of RA diagnosis were included in the “after RA” analyses. Of note, the “rheumatologic autoimmune disorder” comorbidity came from the “rheumatologic” section of the questionnaire and prompted participants with the examples of lupus and scleroderma.
Covariates
Confounder variables in this study included age (continuous) at baseline questionnaire, sex, body mass index (BMI) (continuous), race (non-Hispanic white versus all others), education (bachelor’s degree or higher versus all others), smoking status (ever versus never), and Charlson comorbidity index (CCI).21 They were collected primarily from the baseline questionnaire. EHR data was used to fill in missing BMI values or smoking status within one year of the baseline questionnaire, as well as CCI comorbidities five years before to seven days after the baseline questionnaire.
Statistical Analysis
Chi square tests compared proportions between groups, while Wilcoxon rank-sum tests compared continuous variables. Unconditional logistic regression models calculated odds ratios (ORs) and 95% confidence intervals (CI) for the outcome variable adjusting for age, sex, BMI, race, education, smoking, and CCI. To account for the multiple comparisons in this study, we pre-selected a final significance threshold of 0.01.
In the first set of models to determine prevalence of comorbidities, the exposure variables of interest were individual comorbidities accrued by the time of the baseline questionnaire, and the outcome variable was RA case status. In the second set of models to determine which comorbidities might predispose to RA, the exposure variables of interest were individual comorbidities that occurred BEFORE the RA diagnosis age window, and the outcome variable was RA case status. Finally, in the third set of models to investigate which comorbidities develop more often after RA, the exposure variable was RA vs. control status (reference), and the outcome variables were individual comorbidities that developed after the RA diagnosis age window.
Any possibly significant associations (unadjusted p<.05) were investigated in two sensitivity analyses. The first calculated the main analyses among the 143 RA cases that developed after the baseline questionnaire (incident RA), along with their corresponding controls. The second calculated the main analyses among the 250 RA cases with known serological status, comparing those who were seropositive to those who were seronegative to determine whether comorbidity development differs by serological status. Participants with missing data for a particular comorbidity were excluded from that model. The percent missing for each individual model ranged from 0.9% to 4.2%. We performed analyses using SAS version 9.4 (SAS Institute Inc., Cary, NC).
Results
This study included 821 RA cases and 2,455 controls, for a total sample of 3,276 participants (Table 1). Mean age at the index date was 50 years (standard deviation [SD] 16 years). While RA cases did have a higher raw number of comorbidities than controls at the time of the baseline questionnaire (median 5 [IQR 3,8] vs median 4 [IQR 2,6], p<.001), they did not have more comorbidities at the index date of RA diagnosis (median 1 [IQR 0,3] vs median 1 [IQR 0,2], p = .49).
Table 1.
Demographic characteristics of RA cases and controls at baseline questionnaire
| Number (%) |
||||
|---|---|---|---|---|
| Characteristic | RA cases (N = 821) | Incident RA (N = 143) | Controls (N = 2455) | p-valuea |
| Age, years (mean ±SD) | 62 (±14) | 61 (±15) | 62 (±14) | .97 |
| Female sex | 600 (73) | 95 (66) | 1792 (73) | .96 |
| Body mass index, kg/m2 (mean ±SD) | 30 (±7) | 31 (±6) | 29 (±7) | .03 |
| Race white, non-Hispanic | 804 (98) | 142 (99) | 2393 (98) | .46 |
| Education bachelor’s degree or higher | 324 (40) | 62 (43) | 1182 (48) | .002 |
| Ever smoker | 402 (49) | 69 (48) | 993 (40) | < .001 |
| Charlson comorbidity index (median, IQR) | 1 (0,3) | 1 (0,3) | 1 (0,2) | < .001 |
IQR = interquartile range, kg = kilograms, m = meters, RA = rheumatoid arthritis, SD = standard deviation
comparing all RA cases to controls
Prevalence and Timing of Comorbidities
To address our first aim, we calculated the prevalence of each of the 74 comorbidities in RA cases and controls at the time of the baseline questionnaire (Table 2). Data for the rare diseases (prevalence < 1% in controls) are included in the supplement (Supplemental Table 1). After adjusting for confounders, 11 comorbidities were associated with RA (p<.01), including cataracts, epilepsy, fibromyalgia, GERD, IBD, MI, OSA, OA, pulmonary fibrosis, other rheumatic autoimmune disorder, and VTE (Table 2). The timing of comorbidity development relative to RA diagnosis varied significantly (Figure 1). None of the individual 21 cancers were more common in RA cases compared to controls, even when combining all cancers together (RA cases 31%, controls 32%, p=.80).
Table 2.
Prevalence of comorbidities at the time of baseline questionnaire, and odds ratios for RA
| N (%) |
|||||
|---|---|---|---|---|---|
| Comorbidity | RA cases (N = 821) | Controls (N = 2455) | p-value | Unadjusted OR for RA (95% CI) | Adjusted OR for RA (95% CI) a |
| Abnormal distance vision | 150 (19) | 524 (22) | .06 | ||
| ADHD | 21 (3) | 61 (3) | .89 | ||
| Alcoholism | 24 (3) | 82 (3) | .57 | ||
| Anxiety | 181 (23) | 486 (20) | .14 | ||
| Asthma | 136 (17) | 325 (13) | .02 | 1.31 (1.05,1.63) | 1.22 (0.98,1.53) |
| Atrial fibrillation | 89 (11) | 232 (10) | .23 | ||
| Barrett’s esophagus | 18 (2) | 38 (2) | .21 | ||
| Bleeding disorder | 22 (3) | 51 (2) | .25 | ||
| Breast cancer | 40 (5) | 163 (7) | .08 | ||
| Cataracts | 278 (35) | 699 (29) | .003 | 1.30 (1.09,1.53) | 1.38 (1.12,1.69) |
| Celiac disease | 15 (2) | 21 (1) | .02 | 2.16 (1.09,4.19) | 2.19 (1.09,4.28) b |
| Congestive heart failure | 30 (4) | 57 (2) | .03 | 1.62 (1.02,2.52) | 1.59 (1.00,2.49) |
| COPD | 47 (6) | 88 (4) | < .001 | 1.64 (1.13,2.34) | 1.48 (1.01,2.13) |
| Depression | 231 (29) | 595 (25) | .02 | 1.24 (1.03,1.48) | 1.15 (0.95,1.39) |
| Endometriosis | 94 (11) | 242 (10) | .18 | ||
| Epilepsy | 22 (3) | 29 (1) | .003 | 2.32 (1.31,4.04) | 2.13 (1.19,3.74) b |
| Fibromyalgia | 114 (15) | 144 (6) | < .001 | 2.70 (2.07,3.50) | 2.67 (2.03,3.49) |
| GERD | 304 (38) | 675 (28) | < .001 | 1.55 (1.31,1.84) | 1.47 (1.23,1.74) |
| Glaucoma | 40 (5) | 113 (5) | .72 | ||
| Hepatitis | 22 (3) | 59 (4) | .65 | ||
| Hyper/hypothyroidism | 150 (19) | 371 (16) | .03 | 1.26 (1.02,1.56) | 1.28 (1.03,1.58) |
| Hyperlipidemia | 321 (40) | 1035 (43) | .12 | ||
| Hypertension | 361 (45) | 971 (40) | .01 | 1.23 (1.05,1.45) | 1.19 (1.00,1.42) |
| Inflammatory bowel disease | 37 (5) | 40 (2) | < .001 | 2.85 (1.81,4.50) | 2.63 (1.66,4.18) b |
| Irritable bowel syndrome | 114 (14) | 291 (12) | .12 | ||
| Lazy eye | 41 (5) | 112 (5) | .57 | ||
| Lymphoma | 15 (2) | 30 (1) | .19 | ||
| Macular degeneration | 28 (4) | 78 (3) | .73 | ||
| Melanoma | 28 (4) | 76 (3) | .67 | ||
| Migraine headaches | 191 (24) | 511 (21) | .11 | ||
| Myocardial infarction | 63 (8) | 88 (4) | < .001 | 2.26 (1.61,3.15) | 2.15 (1.52,3.04) |
| Nonmelanoma skin cancer | 121 (15) | 358 (15) | .88 | ||
| Obesity (BMI ≥ 30) | 336 (41) | 909 (37) | .05 | 1.18 (1.00,1.38) | 1.10 (0.94,1.30) |
| Obstructive sleep apnea | 164 (20) | 352 (15) | < .001 | 1.49 (1.21,1.83) | 1.44 (1.15,1.80) |
| Osteoarthritis | 370 (50) | 915 (38) | < .001 | 1.59 (1.35,1.88) | 1.70 (1.42,2.04) |
| Other cancer | 16 (2) | 42 (2) | .62 | ||
| Other liver disease | 31 (4) | 59 (3) | .04 | 1.60 (1.02,2.47) | 1.42 (0.89,2.23) |
| Other mental illness | 14 (2) | 47 (2) | .73 | ||
| Prostate cancer | 22 (3) | 83 (4) | .34 | ||
| Pulmonary fibrosis | 16 (2) | 10 (0.4) | < .001 | 4.90 (2.24,11.2) | 4.63 (2.11,10.7) b |
| Rheum. autoimmune dzc | 123 (16) | 95 (4) | < .001 | 4.43 (3.35,5.88) | 4.44 (3.33,5.94) |
| Stroke | 18 (2) | 27 (1) | .02 | 2.03 (1.10,3.68) | 1.97 (1.06, 3.58) |
| TIA | 37 (5) | 79 (3) | .07 | ||
| Type 1 diabetes | 18 (2) | 35 (2) | .13 | ||
| Type 2 diabetes | 77 (10) | 241 (10) | .72 | ||
| Venous thromboembolism | 82 (10) | 132 (6) | < .001 | 1.97 (1.47,2.62) | 1.85 (1.37,2.49) |
ADHD = attention-deficit/hyperactivity disorder, BMI = body mass index, CI = confidence interval, COPD = chronic obstructive pulmonary disease, dz = disease, GERD = Gastroesophageal reflux disease, OR = odds ratio, RA = rheumatoid arthritis, rheum. = rheumatologic, TIA = transient ischemic attack
Adjusting for age, sex, race, BMI (except in obesity model), education, smoking, Charlson comorbidity index (excluded for disease contained within it). Bold values are statistically significant to p < .01.
Excluding race due to sample size
i.e. systemic lupus erythematosus or scleroderma
Figure 1.
Timing of comorbidities in RA cases (N=821), relative to RA diagnosis, according to age windows on baseline questionnaire
Comorbidities Before RA
Four comorbidities did occur more commonly before RA compared to controls, suggesting a predisposition to RA development. These included IBD, OA, type 1 diabetes, and VTE (Table 3). Hyper/hypothyroidism, other rheumatic autoimmune disorder, and stroke also tended to occur more commonly before RA diagnosis in RA cases relative to controls (p<.05 but >.01) (Table 3). All other comorbidities had no propensity to occur more commonly before RA diagnosis relative to controls according to the age windows on the questionnaire (Supplemental Table 2).
Table 3.
Comorbidities that occurred more often before RA diagnosis relative to controls
| N (%) |
||||
|---|---|---|---|---|
| Comorbidity | RA cases (N = 821) | Controls (N = 2455) | Unadjusted OR for RA (95% CI) | Adjusted OR for RA (95% CI) a |
| GERD | 86 (13) | 219 (10) | 1.31 (1.00,1.71) | 1.23 (0.93,1.60) |
| Hyper/hypothyroidism | 66 (9) | 150 (7) | 1.37 (1.01,1.85) | 1.39 (1.02,1.89) |
| Inflammatory bowel disease | 15 (2) | 12 (0.5) | 3.84 (1.79,8.41) | 3.82 (1.77,8.41) b |
| Osteoarthritis | 84 (16) | 241 (12) | 1.43 (1.09,1.87) | 1.54 (1.15,2.05) |
| Rheum. autoimmune dzc | 17 (3) | 29 (1) | 2.00 (1.07,3.62) | 2.02 (1.07,3.69) |
| Stroke | 8 (1.0) | 9 (0.4) | 2.70 (1.01,7.07) | 2.75 (1.03,7.26) b |
| Type 1 diabetes | 10 (1) | 10 (0.4) | 3.03 (1.24,7.40) | 3.07 (1.24,7.60) b |
| Venous thromboembolism | 27 (4) | 48 (2) | 1.91 (1.41,2.57) | 1.80 (1.31,2.45) |
BMI = body mass index, CI = confidence interval, dz = disease, GERD = Gastroesophageal reflux disease, OR = odds ratio, RA = rheumatoid arthritis, rheum. = rheumatologic
Adjusting for age, sex, race, BMI, education, smoking, Charlson comorbidity index (excluded for disease contained within it). Bold values are statistically significant to p<.01.
Excluding race due to sample size
i.e. systemic lupus erythematosus or scleroderma
Comorbidities After RA
MI and VTE tended to occur with increased frequency after RA diagnosis relative to controls (Table 4). CHF and OSA also trended towards significance with increased occurrence in RA cases relative to controls after index date (p<.05 but >.01) (Table 4). Hyperlipidemia was less common in RA cases compared to controls (OR 0.63, 95% CI 0.46 to 0.84, p=.003). All other comorbidities had no propensity to occur after RA diagnosis relative to controls (Supplemental Table 3).
Table 4.
Comorbidities that occurred more often after RA diagnosis relative to controls
| N (%) |
||||
|---|---|---|---|---|
| Comorbidity | RA cases (N = 821) | Controls (N = 2455) | Unadjusted OR for comorbidity (95% CI) | Adjusted OR for comorbidity (95% CI) a |
| Congestive heart failure | 15 (2) | 21 (1) | 2.20 (1.11,4.26) | 2.26 (1.13,4.42) b |
| COPD | 20 (3) | 34 (1) | 1.80 (1.02,3.12) | 1.66 (0.93,2.90) b |
| GERD | 64 (11) | 161 (8) | 1.37 (1.00,1.85) | 1.32 (0.96,1.79) |
| Lymphoma | 7 (0.9) | 8 (0.3) | 2.66 (0.93,7.43) | 2.37 (0.82,6.69) b |
| Myocardial infarction | 29 (4) | 28 (1) | 3.27 (1.93,5.55) | 3.09 (1.78,5.36) |
| Obstructive sleep apnea | 51 (7) | 113 (5) | 1.44 (1.02,2.02) | 1.42 (1.00,2.03) |
| Venous thromboembolism | 14 (2) | 22 (1) | 1.95 (1.41,2.66) | 1.84 (1.33,2.54) |
BMI = body mass index, CI = confidence interval, COPD = chronic obstructive pulmonary disease, GERD = Gastroesophageal reflux disease, OR = odds ratio, RA = rheumatoid arthritis
Adjusting for age, sex, race, BMI, education, smoking, Charlson comorbidity index (excluded for disease contained within it). Bold values are statistically significant to p<.01.
Excluding race due to sample size
Sensitivity Analyses
In the subset of RA cases with incident RA, several comorbidities were again significant including fibromyalgia, IBD, OA, pulmonary fibrosis, rheumatologic autoimmune disorder, and VTE at any time, as well as IBD and OA before RA. However, several were not significant including cataracts, GERD, MI, and OSA at any time, and VTE before RA development (Table 5). Despite its smaller sample size, the incident cohort also had some statistically significant associations which were not present in the full cohort. These included hyper/hypothyroidism both at any time and before RA diagnosis, and other rheumatologic autoimmune disorder before RA diagnosis (Table 5). In general, the association between RA and comorbidities did not differ by serological status (Supplemental Table 4). Characteristics of participants missing two or more of the 74 comorbidities are shown in Supplemental Table 5.
Table 5.
Analyses among the subset of cases who developed RA after the questionnaire (incident RA cohort) and their matched controls
| Any Time |
Before RA diagnosis |
|||||
|---|---|---|---|---|---|---|
| Comorbidity | RA cases (N = 143) | Controls (N = 426) | p-value | RA cases (N = 143) | Controls (N = 426) | p-value |
| Asthma | 21 (15) | 50 (12) | .28 | |||
| Cataracts | 42 (30) | 119 (28) | .66 | |||
| Celiac disease | 3 (2) | 3 (1) | .15 | |||
| Congestive heart failure | 3 (2) | 8 (2) | .84 | |||
| COPD | 5 (4) | 12 (3) | .65 | |||
| Depression | 33 (24) | 75 (18) | .14 | |||
| Epilepsy | 3 (2) | 2 (1) | .07 | |||
| Fibromyalgia | 13 (9) | 15 (4) | .008 | |||
| GERD | 43 (31) | 107 (25) | .22 | 27 (22) | 72 (19) | .45 |
| Hyper/hypothyroidism | 29 (21) | 42 (10) | .001 | 22 (17) | 32 (8) | .004 |
| Hypertension | 58 (41) | 169 (40) | .08 | |||
| Inflammatory bowel disease | 8 (6) | 3 (1) | <.001 | 6 (4) | 2 (1) | .001 |
| Myocardial infarction | 3 (2) | 14 (3) | .49 | |||
| Obesity (BMI ≥ 30) | 65 (46) | 147 (35) | .02 | |||
| Obstructive sleep apnea | 23 (16) | 62 (15) | .62 | |||
| Osteoarthritis | 69 (49) | 125 (30) | < .001 | 42 (37) | 68 (19) | < .001 |
| Other liver disease | 7 (2) | 5 (4) | .18 | |||
| Pulmonary fibrosis | 2 (2) | 0 (0) | .01 | |||
| Rheum. autoimmune disorder | 13 (9) | 9 (2) | < .001 | 9 (7) | 4 (1) | < .001 |
| Stroke | 1 (1) | 4 (1) | .79 | 1 (0.7) | 3 (0.7) | .99 |
| Type 1 diabetes | 3 (2) | 2 (0.5) | .07 | 3 (2) | 2 (0.5) | .07 |
| Venous thromboembolism | 13 (10) | 20 (5) | .05 | 8 (6) | 15 (4) | .23 |
BMI = body mass index, COPD = chronic obstructive pulmonary disease, GERD = gastroesophageal reflux disease, RA = rheumatoid arthritis, rheum. = rheumatologic
Discussion
In this study of 74 comorbidities and the timing of their development, we discovered that comorbidities accumulate in an accelerated fashion after RA diagnosis. In addition, autoimmune diseases and epilepsy may predispose to RA development, while heart disease, VTE, and OSA might develop as a result of RA. These findings have important implications for understanding RA pathogenesis, promoting earlier detection of RA, and screening for comorbidities among RA patients.
This study first confirmed that current RA patients have more comorbidities than their matched counterparts without RA. This higher comorbidity burden is well-described and leads to worse outcomes.1–5 An interesting and surprising finding, however, was that RA cases did not have more comorbidities than controls before RA occurred. RA diseases processes, treatments, or consequences may therefore play important roles in the development of certain comorbidities. Investigating the relationship between RA and comorbidities thus becomes even more important, both to understand and to prevent comorbidity development.
Another novel finding from this study was that VTE was associated in the time period before RA development. Previous studies have shown that RA is associated with increased VTE risk at any time, with a similar magnitude to the hazard ratio of 1.9 found in this study.22,23 The mechanism for this association is thought to be chronic inflammation and/or DMARD use after RA occurs.24 A new finding, however, was that VTE occurred more commonly in RA cases before RA onset, even in the incident RA cohort. These results suggest systemic inflammation may start before RA becomes clinically apparent. Alternatively, this could be explained by increased healthcare utilization occurring after VTE or RA diagnosis, or there may be another underlying mechanism shared by both diseases.
Autoimmune conditions did predispose to RA as hypothesized, and the strongest of these associations was with IBD. RA was associated with IBD at any time but even more so in the time period before RA diagnosis, with odds nearly four times higher in RA cases compared to controls. While this association might have resulted from small sample size or contamination with spondyloarthropathy, the significant findings were present even in the smaller incident RA cohort. Further, recent population-based studies in Denmark and Korea also found an increase in RA among patients with IBD.15,25 Monitoring patients with IBD for RA may be warranted.
The association between type 1 diabetes and RA was also strong, especially in the time period before RA diagnosis. While this finding was not significant in the incident cohort, likely due to limited sample size, a trend was still present. The mechanism for this association may include shared genetic risk.26 Interestingly, parents with RA are also more likely to have offspring with type 1 diabetes.27 Existing literature on this topic is limited, with one prior study showing increased type 1 diabetes in seropositive RA patients.28 Our observed association regarding type 1 diabetes before RA is novel and suggests a predisposition to RA among individuals who have type 1 diabetes. All of the other autoimmune conditions in this study were also associated with RA at some time point as well, perhaps due to shared genetic and epigenetic influences leading to loss of tolerance and clinical overlap syndromes.29 Overall, the strong associations between RA and autoimmune diseases highlight the importance of increased suspicion of RA in patients with autoimmune diseases, and vice versa.
This study also found that epilepsy may predispose to RA. Epilepsy was associated with RA in the full cohort but not the incident cohort. Existing literature is mixed, with many studies showing an association,30,31 but others showing none.32 Nevertheless, there is a plausible mechanism for association between the two diseases. Impaired regulation of inflammation is known to be critical in the development of epilepsy, and a recent study also suggested that autoimmune conditions like RA may predispose to epilepsy.33 In fact, some argue that epilepsy itself may be autoimmune,31 and as discussed above, people with autoimmune diseases may be predisposed to RA. Epilepsy did occur before RA in nearly two thirds of cases in this study, suggesting it might be a useful marker to indicate increased predisposition to RA.
Cardiovascular disease, including MI and CHF, was more common after developing RA. This was hypothesized and previously described.34,35 Lower risk of developing hyperlipidemia was also observed as previously described.36 Cataracts were also associated with RA in the full cohort but not the incident cohort, likely since patients with RA have often been treated with glucocorticoids (and those with incident RA had not). Overall, these findings serve as a positive control and provide reassurance about the accuracy of the study findings.
OSA was also associated with RA at any time and trended toward occurring more often after RA diagnosis in cases than controls. These findings were true even after adjusting for key confounders including BMI. However, the results were not significant in the incident cohort. Previous studies have also found an increased risk of OSA in RA patients,10,11 though others have found non-statistically significant associations with OSA.12 Providers should have a high index of suspicion for OSA in RA patients due to its large health impacts and consider routinely screening for it in their RA patients.
Strengths of this study include its lengthy list of comorbidities, delineation of time association between comorbidities and RA, and thorough adjustment for confounders. The risk of RA-specific recall bias in this study is also low since participants were asked about all comorbidities and exposures, both related and unrelated to RA. Additionally, the incident RA subgroup provides even further support for study findings.
There are also several important limitations. First, using a convenience sample for selection creates potential for selection bias and limits generalizability, which was somewhat mitigated by recruitment in primary care divisions and at multiple geographic locations. The degree and direction of any remaining bias is unknown, as volunteers might have been healthier or sicker than the general population of RA patients. Second, the timing of diagnosis for RA and each comorbidity relied on patient report using wide age ranges and thus it was not possible to accurately order comorbidities that develop shortly before or after RA diagnosis. The wide age windows also precluded our ability to determine the exact timing of comorbidity onset relative to RA. Third, reliance on self-report for presence or absence of comorbidities leads to misclassification bias, the degree of which is unknown for each of the comorbidities and likely varies for each. Future studies verifying our results using an alternative method such as diagnosis codes are warranted. Fourth, any observed increased comorbidity risk may have resulted from diagnostic access bias from possible increased healthcare utilization among participants with RA. Fifth, as in any observational study, there may be residual confounding that was unmeasured. While we adjusted for smoking status, this was only available at the time of questionnaire by self-report, and thus may not accurately represent smoking exposure at the time of comorbidity development. Finally, data on serological status was limited due to lack of availability.
Conclusion
In conclusion, we found an association between RA and pre-existing comorbidities such as IBD, type 1 diabetes, epilepsy, and VTE, which provides opportunities for earlier detection of RA. In addition, the associations with cardiovascular disease, VTE, and OSA after RA diagnosis emphasize the importance of screening for these comorbidities among RA patients. Future studies should investigate why the observed overlap between RA and these comorbidities occurs and examine whether heightened screening for RA and its comorbidities leads to improved outcomes.
Supplementary Material
Acknowledgments
Financial Support: Rheumatology Research Foundation Resident Research Preceptorship and K Supplement Award. National Institute for Arthritis and Musculoskeletal Skin Diseases (K23 AR069688, L30 AR066953, R03 AR075886, P30 AR070253, and P30 AR072577). The funders had no role in study design, analysis, manuscript writing, or decision to submit.
Abbreviations
- BMI
body mass index
- CCI
Charlson comorbidity index
- CI
confidence interval
- GERD
gastroesophageal reflux disease
- IBD
inflammatory bowel disease
- MI
myocardial infarction
- OR
odds ratio
- OSA
obstructive sleep apnea
- RA
rheumatoid arthritis
- VTE
venous thromboembolism
Footnotes
Disclosures: The authors have no conflicts of interest related to this manuscript.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Radner H, Smolen JS, Aletaha D. Impact of comorbidity on physical function in patients with rheumatoid arthritis. Ann Rheum Dis. 2010;69(3):536–541. [DOI] [PubMed] [Google Scholar]
- 2.Norton S, Koduri G, Nikiphorou E, Dixey J, Williams P, Young A. A study of baseline prevalence and cumulative incidence of comorbidity and extra-articular manifestations in RA and their impact on outcome. Rheumatology (Oxford). 2013;52(1):99–110. [DOI] [PubMed] [Google Scholar]
- 3.Ellerby N, Mattey DL, Packham J, Dawes P, Hider SL. Obesity and comorbidity are independently associated with a failure to achieve remission in patients with established rheumatoid arthritis. Ann Rheum Dis. 2014;73(11):e74. [DOI] [PubMed] [Google Scholar]
- 4.Geryk LL, Carpenter DM, Blalock SJ, DeVellis RF, Jordan JM. The impact of co-morbidity on health-related quality of life in rheumatoid arthritis and osteoarthritis patients. Clin Exp Rheumatol. 2015;33(3):366–374. [PMC free article] [PubMed] [Google Scholar]
- 5.Pedersen JK, Holst R, Primdahl J, Svendsen AJ, Horslev-Petersen K. Mortality and its predictors in patients with rheumatoid arthritis: a Danish population-based inception cohort study. Scand J Rheumatol. 2018;47(5):371–377. [DOI] [PubMed] [Google Scholar]
- 6.Gabriel SE, Michaud K. Epidemiological studies in incidence, prevalence, mortality, and comorbidity of the rheumatic diseases. Arthritis Res Ther. 2009;11(3):229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jin S, Li M, Fang Y, et al. Chinese Registry of rheumatoid arthritis (CREDIT): II. prevalence and risk factors of major comorbidities in Chinese patients with rheumatoid arthritis. Arthritis Res Ther. 2017;19(1):251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Jeong H, Baek SY, Kim SW, et al. Comorbidities of rheumatoid arthritis: Results from the Korean National Health and Nutrition Examination Survey. PLoS One. 2017;12(4):e0176260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.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(1):e000267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Shen TC, Hang LW, Liang SJ, et al. Risk of obstructive sleep apnoea in patients with rheumatoid arthritis: a nationwide population-based retrospective cohort study. BMJ Open. 2016;6(11):e013151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Reading SR, Crowson CS, Rodeheffer RJ, Fitz-Gibbon PD, Maradit-Kremers H, Gabriel SE. Do rheumatoid arthritis patients have a higher risk for sleep apnea? J Rheumatol. 2009;36(9):1869– 1872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wilton KM, Matteson EL, Crowson CS. Risk of Obstructive Sleep Apnea and Its Association with Cardiovascular and Noncardiac Vascular Risk in Patients with Rheumatoid Arthritis: A Population-based Study. J Rheumatol. 2018;45(1):45–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Vallerand IA, Lewinson RT, Frolkis AD, et al. Depression as a risk factor for the development of rheumatoid arthritis: a population-based cohort study. RMD Open. 2018;4(2):e000670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chung WS, Lin CL. Sleep disorders associated with risk of rheumatoid arthritis. Sleep Breath. 2018. [DOI] [PubMed] [Google Scholar]
- 15.Halling ML, Kjeldsen J, Knudsen T, Nielsen J, Hansen LK. Patients with inflammatory bowel disease have increased risk of autoimmune and inflammatory diseases. World J Gastroenterol. 2017;23(33):6137–6146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schmitt J, Schwarz K, Baurecht H, et al. Atopic dermatitis is associated with an increased risk for rheumatoid arthritis and inflammatory bowel disease, and a decreased risk for type 1 diabetes. J Allergy Clin Immunol. 2016;137(1):130–136. [DOI] [PubMed] [Google Scholar]
- 17.Olson JE, Ryu E, Johnson KJ, et al. The Mayo Clinic Biobank: a building block for individualized medicine. Mayo Clin Proc. 2013;88(9):952–962. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014;12(12):1495–1499. [DOI] [PubMed] [Google Scholar]
- 19.Chung CP, Rohan P, Krishnaswami S, McPheeters ML. A systematic review of validated methods for identifying patients with rheumatoid arthritis using administrative or claims data. Vaccine. 2013;31 Suppl 10:K41–61. [DOI] [PubMed] [Google Scholar]
- 20.Kronzer VL, Crowson CS, Sparks JA, Vassallo R, Davis JM, III. Investigating asthma, allergic disease, passive smoke exposure, and risk of rheumatoid arthritis. Arthritis Rheumatol. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373– 383. [DOI] [PubMed] [Google Scholar]
- 22.Holmqvist ME, Neovius M, Eriksson J, et al. Risk of venous thromboembolism in patients with rheumatoid arthritis and association with disease duration and hospitalization. JAMA. 2012;308(13):1350–1356. [DOI] [PubMed] [Google Scholar]
- 23.Kim SC, Schneeweiss S, Liu J, Solomon DH. Risk of venous thromboembolism in patients with rheumatoid arthritis. Arthritis Care Res (Hoboken). 2013;65(10):1600–1607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Davies R, Galloway JB, Watson KD, Lunt M, Symmons DP, Hyrich KL. Venous thrombotic events are not increased in patients with rheumatoid arthritis treated with anti-TNF therapy: results from the British Society for Rheumatology Biologics Register. Ann Rheum Dis. 2011;70(10):1831–1834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bae JM, Choo JY, Kim KJ, Park KS. Association of inflammatory bowel disease with ankylosing spondylitis and rheumatoid arthritis: A nationwide population-based study. Mod Rheumatol. 2017;27(3):435–440. [DOI] [PubMed] [Google Scholar]
- 26.Dieude P, Teixeira VH, Pierlot C, Cornelis F, Petit-Teixeira E. Testing for linkage and association with rheumatoid arthritis a ptpn22 promoter polymorphism reported to be associated and linked with type 1 diabetes in the Caucasian population. Ann Rheum Dis. 2008;67(6):900–901. [DOI] [PubMed] [Google Scholar]
- 27.Hemminki K, Li X, Sundquist J, Sundquist K. Familial association between type 1 diabetes and other autoimmune and related diseases. Diabetologia. 2009;52(9):1820–1828. [DOI] [PubMed] [Google Scholar]
- 28.Liao KP, Gunnarsson M, Kallberg H, et al. Specific association of type 1 diabetes mellitus with anti-cyclic citrullinated peptide-positive rheumatoid arthritis. Arthritis Rheum. 2009;60(3):653– 660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhang P, Lu Q. Genetic and epigenetic influences on the loss of tolerance in autoimmunity. Cell Mol Immunol. 2018;15(6):575–585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Chang KH, Hsu YC, Chang MY, et al. A Large-Scale Study Indicates Increase in the Risk of Epilepsy in Patients With Different Risk Factors, Including Rheumatoid Arthritis. Medicine (Baltimore). 2015;94(36):e1485. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ong MS, Kohane IS, Cai T, Gorman MP, Mandl KD. Population-level evidence for an autoimmune etiology of epilepsy. JAMA Neurol. 2014;71(5):569–574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Gaitatzis A, Carroll K, Majeed A, J WS. The epidemiology of the comorbidity of epilepsy in the general population. Epilepsia. 2004;45(12):1613–1622. [DOI] [PubMed] [Google Scholar]
- 33.Rana A, Musto AE. The role of inflammation in the development of epilepsy. J Neuroinflammation. 2018;15(1): 144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Gabriel SE, Crowson CS, O’Fallon WM. Comorbidity in arthritis. J Rheumatol. 1999;26(11):2475–2479. [PubMed] [Google Scholar]
- 35.Wright K, Crowson CS, Gabriel SE. Cardiovascular comorbidity in rheumatic diseases: a focus on heart failure. Heart Fail Clin. 2014;10(2):339–352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Myasoedova E, Crowson CS, Kremers HM, Fitz-Gibbon PD, Therneau TM, Gabriel SE. Total cholesterol and LDL levels decrease before rheumatoid arthritis. Ann Rheum Dis. 2010;69(7):1310–1314. [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.