Comorbidity Before and After a Diagnosis of Inflammatory Bowel Disease

Charles N. Bernstein; Zoann Nugent; Seth Shaffer; Harminder Singh; Ruth Ann Marrie

Disclosures

Aliment Pharmacol Ther. 2021;54(5):637-651. 

In This Article

Discussion

As compared to age, sex and geographically matched controls without IBD, we found an increased prevalence of almost all diseases that make up the CCI in persons with IBD before diagnosis, and an increased incidence of these diseases post-diagnosis. The increase in most of the comorbidities prior to IBD diagnosis would not be secondary to a case ascertainment bias. The comorbidities diagnosed prior to IBD diagnosis as opposed to post-IBD diagnosis might be even stronger evidence that IBD is a multisystem disease, and extends to a much greater extent than what has been seen with other more classical immune-mediated extra-intestinal diseases. While some diseases were more common in persons with CD, only chronic pulmonary disease post-diagnosis was more common in UC suggesting that CD is more of a multisystem disease. For most comorbidities, there was no predilection by sex except for an increase in peripheral vascular disease in females and an increase in connective tissue disease/rheumatic disease for males, but several of the comorbidities were more commonly diagnosed in persons under age 25 than at other age groups. Having found these comorbidities diagnosed at a younger median age in persons with IBD than in controls suggests that IBD patients are harbouring risk factors for other systemic diseases and health care providers must be vigilant in preventative and diagnostic measures.

Vascular diseases were increased both pre-IBD diagnosis and post-IBD diagnosis. Elsewhere, vascular diseases have also been shown to be increased in persons with IBD. Using a large commercial database (Explorys, IBM Watson) that aggregates electronic medical records from 26 nationwide health care systems, adults with IBD (n = 290 430) between 2013 and 2018 were examined for risk of myocardial infarction.[16] The prevalence of Cardiac was higher in IBD vs persons without IBD; UC, OR = 2.09 [95% CI 2.04, 2.13], and CD, OR = 2.79 [95% CI 2.74, 2.85]. The highest OR for Cardiac was in persons aged 30–34 years (12.05, 95% CI 11.16, 13.01). Common cardiovascular risk factors were more prevalent in individuals with IBD compared with those without IBD: such as dyslipidaemia, hypertension, diabetes, obesity and smoking. In the adjusted analysis of all patients, being male conferred a higher risk of Cardiac (OR 1.78 [95% CI, 1.77, 1.79]). In a population-based Korean study, Cardiac risk was higher in CD patients than in controls (HR 1.80, 95% CI, 1.47–2.21), with the strongest risk in patients aged less than 40 years and among females.[17]

Using the UMIBDED, our research group previously reported an increased risk of ischemic heart disease in IBD including both CD and UC and males and females (incidence rate ratio, IRR = 1.26, 95% CI 1.11, 1.44), whereas cerebrovascular disease was increased in only CD (IRR 1.32, 95% CI 1.05, 1.66).[18] A meta-analysis through 2015 reported a modest increase in the risk of cerebrovascular disease incidence (HR = 1.29; 95% CI, 1.16–1.43) with an increase in both CD and UC.[19] In addition, this risk was increased in both males and females. Two other meta-analyses concluded that cardiovascular/cerebrovascular disease risk was increased in IBD, particularly among women, in accordance with our present results.[20,21] In a population-based study from Taiwan there was an increased rate of peripheral vascular disease in IBD,[22] while in a smaller population-based study from Catalonia Spain there was not an increased risk of peripheral vascular disease in IBD.[23]

Clinical disease activity in IBD is a known risk factor for vascular events.[24] The increase in vascular disease in IBD and other chronic immune-mediated diseases is not fully explained by the traditional vascular disease risk factors. The increase in vascular diseases pre-IBD diagnosis may be mediated through systemic inflammation-inducing arterial vascular wall rigidity.[25] Immunomodulatory drugs such as antibodies to tumour necrosis factor used in treating IBD could reduce this arterial vascular wall rigidity.[25]

We found an increase in chronic pulmonary disease both pre and post-IBD diagnosis. Previously, we reported an increased risk of asthma and bronchitis in IBD.[26] Possible etiologic reasons for their co-occurrence include shared embryology, shared genetics, shared dysregulated protease activity and shared smoking histories at least in COPD and IBD.[27–30] Bronchiolitis and bronchiectasis have been reported as extra-intestinal diseases in IBD, but some drugs used to treat IBD may also cause pulmonary disorders.[31–35] Subclinical radiographic and pulmonary function test abnormalities have also been reported in IBD.[36] An increased risk of IBD in persons with chronic obstructive pulmonary disease (COPD) has been reported[37,38] and the co-occurrence of IBD and COPD increased mortality compared to persons with COPD without IBD.[39,40]

We found that connective tissue disease/rheumatic disease had among the strongest increased risks both pre and post-IBD diagnosis of the comorbidities we assessed. There is a considerable literature regarding arthritis being comorbid in IBD.[41] There is an increase in peripheral arthritis and spondyloarthropathy.[26,42] There may not be an increased risk of rheumatoid arthritis or osteoarthritis in IBD; however, it is difficult to identify specific arthropathy diagnoses using administrative data.

Peptic ulcer disease was increased in our study pre and post-IBD diagnosis. Increased peptic ulcer disease was also evident in a Swiss study.[2] Peptic ulcer disease perhaps more than any of the other comorbidities might represent a diagnostic or surveillance bias. Some peptic ulcer diagnoses may be misdiagnoses in persons with IBD. Considering that paediatric CD is more likely to be diagnosed with upper GI disease[43] this may be initially diagnosed as peptic ulcer disease. Alternatively, it may be increasingly diagnosed in IBD since persons with IBD are more likely to undergo endoscopy and cross-sectional imaging.

We found an increased risk of liver disease both prior to and after diagnoses of IBD, especially in persons younger than 25 years. Primary sclerosing cholangitis (PSC) is the best known of the associated liver diseases with IBD. Most patients with PSC have IBD and up to 3% of persons with IBD may have PSC.[25] However, chronic liver disease that is not PSC is common in IBD.[44] The CCI distinguishes mild from moderate to severe chronic liver disease however, this can be a difficult distinction using administrative data, so we assessed all liver disease comorbidity collectively. As many as one in five persons with IBD may have serum liver test abnormalities.[45] Often this is proven to be secondary to non-alcoholic fatty liver disease (NAFLD). It has been reported that IBD patients develop NAFLD with fewer metabolic risk factors than non-IBD patients with NAFLD.[46]

We found an increased risk of renal disease pre and post-IBD with a greater increase in CD than UC. We also found that post-IBD the increased HR diminished with advancing age. This may reflect the relative paucity of renal disease in the younger general population. However, persons with IBD diagnosed with renal disease were diagnosed at a significantly younger age than controls diagnosed with renal disease. Elsewhere, increased renal disease in IBD has been reported mostly as case reports or case series.[47] Despite multiple reports of 5-ASA-induced nephrotoxicity, two recent reports on the topic did not suggest that 5-ASA was a risk factor for nephrotoxicity.[48,49] IgA nephropathy is thought to be the most common renal disease, followed by tubulointerstitial nephritis.[50] In a population-based study from Korea the incidence of end-stage renal disease was significantly higher in an IBD cohort compared to controls, after multivariable adjustments for age, sex, place of residence, income, comorbidities and medication use (adjusted HR = 3.03; 95% CI: 1.77–5.20).[51] Interestingly, those with CD who were treated with 5-ASA had a significantly lower incidence of end-stage renal disease than those who were treated without 5-ASA (0.40 vs 1.68 per 1000 person-years; adjusted HR = 0.43; 95% CI: 0.19–0.99).

We found an increased risk of cancer pre and post-IBD with a diminishing HR post-diagnosis with advancing age. As with renal disease, this may reflect the relative reduction in cancer in younger adults in the general population. Cancer is increased in IBD but the cancers that can be expected at an increased rate are mostly limited to colorectal cancer,[51] with increases in rarer cancers such as small bowel adenocarcinoma, cholangiocarcinoma, and anal cancers.[52] Lymphomas, skin cancers and cervical dysplasias can be increased in persons with IBD using immunomodulating therapies.[53–55]

In our study, diabetes was increased in CD and only among persons older than 50 years. Perhaps this may reflect corticosteroid use in CD, which, while used across the age spectrum in IBD may have a diabetogenic impact when increasing age reduces insulin sensitivity. While type I diabetes may develop in persons with IBD as a co-immune disease, diabetes in one study was not increased in IBD but was increased in another study.[2,16] There has been some concern that the use of some diabetes medications, dipeptidyl peptidase 4 inhibitor (DPP-4-I), may lead to an increased risk of IBD although data have been conflicting.[56–58] In a recent meta-analysis, using a random-effects analysis, DPP-4-I do not appear to increase the risk of developing IBD, while there was an increased risk using a fixed-effects model.[59]

We found that females with CD had an increased risk of dementia post-IBD diagnosis. We have recently reported that persons with IBD may have cognitive impairment compared to expected outcomes in the general population.[60] A large Swiss study reported no increase dementia in persons with IBD.[2] In a population-based study from Taiwan the overall incidence of dementia among persons with IBD was significantly elevated compared to controls (5.5% vs 1.4%, HR 2.54, 95% CI 1.91, 3.37).[61] Patients with IBD were diagnosed with dementia at 76.2 years vs 83.5 years among controls. Alzheimer's was the most common type of dementia. There was no difference in dementia rate between UC and CD or between males and females. There is evidence that chronic inflammation is a driver of cognitive decline and other evidence that alterations in the gut microbiome may enhance the development of amyloid plaques in the brain.[62–65]

Prior to and post-IBD diagnosis we found an increased RR and HR, respectively, for paraplegia/hemiplegia. This was unexpected but the excess risk in IBD persisted even when persons with cerebrovascular disease were excluded from the analysis. There is little published literature on hemiplegia or paraplegia in IBD. There is no evidence HIV is increased in persons with IBD.[2,66]

There are several limitations to our study. One main limitation is the assignation of comorbid diagnoses based on single health system contacts for the various diseases. This enhances sensitivity at the expense of specificity. This is typically how CCI is applied to all large administrative databases so CCI provides an estimate of potential comorbidity as opposed to rigorously proven comorbidity. In a previous study assessing for other chronic diseases in IBD using the University of Manitoba IBD Epidemiology Database, undertaking sensitivity analyses for one vs five health system contacts did not appreciably change the direction of magnitude of calculated risks in IBD compared to controls.[42] Further, several of our findings have been replicated elsewhere in the literature. Finally, mislabelling of diagnoses in our administrative data should be similarly evident in our cases and controls and hence should not impact on RR or HR.

We are reporting that persons with IBD have higher rates of cardiac, cerebrovascular disease, and peripheral vascular disease. There is biological plausibility to these associations. Whereas in the middle of the 20th century IBD patients often presented with malnutrition, current day IBD is more likely to be associated with features of metabolic syndrome and even obesity. This may also translate to greater prevalence of chronic liver disease other than PSC perhaps related to increased rates of NAFLD. An increase in pulmonary and especially rheumatological diseases in IBD has been widely reported. The increased findings of increased chronic kidney disease and dementia post-diagnosis are perhaps the most novel of our findings. Our results serve as reminders for clinicians to be aware of potential for various comorbidities when undertaking screening blood work, chest radiography and discussing the risks of smoking with IBD patients.

In conclusion, persons with IBD have a higher comorbidity burden than persons without IBD. This is identified through the CCI and through assessment of the individual diseases making up the CCI. The burden of disease for persons with IBD is compounded by classic immuno-inflammatory extra-intestinal diseases, by an increase in psychiatric diseases, venous thromboembolic diseases and these other major categories of disease. All of this may contribute to the increased disability in IBD, mortality in CD and overall reduction in QOL.[67,68] Hence, optimal care plans for persons with IBD should include attention to general health in its entirety and an assessment for other comorbidities that include just about every other organ system.

processing....