Results
We included 9247 IBD cases (4253 CD, 4994 UC) and 85 691 controls in the analyses. The median (interquartile range [IQR]) age at diagnosis of IBD was 36 (25, 52) and was significantly younger in persons with CD (33 years [23,49]) than in UC (39 years [27,54] P < 0.0001). Forty-seven per cent of all IBD cases were male. IBD cases and controls had similar Socioeconomic Factor Index scores (Table 1).
Overall the RRs and HRs were highly correlated (r = +0.69, P = 0.013). Similar patterns were evident in CD (r = +0.69, P = 0.013) and UC (+0.71, P = 0.0094). When data were compared between disease types RR (P = 0.012) and HR (P = 0.0015) are higher in CD than UC. RR and HR for males and females did not differ significantly overall.
Pre-IBD Diagnosis Comorbidities
All comorbidities were increased pre diagnosis except for diabetes and dementia (Table 2). Peptic ulcer disease had the highest RR pre-diagnosis of all the comorbidities [RR = 2.01, CI 1.82, 2.23]. For all of Cardiac, connective tissue disease/rheumatic disease, peptic ulcer disease, and renal disease the RR was greater in CD than in UC. For peripheral vascular disease, chronic pulmonary disease, liver disease, and cancer the increase was similar in CD and UC. In the case of cerebrovascular disease, paraplegia/hemiplegia, diabetes and dementia there was no increase in either CD or UC. The increased RR was greater in females compared to males for cerebrovascular disease, while for all other comorbidities there was no difference in RR by sex.
Post-IBD Diagnosis Comorbidities
All comorbidities were increased post-IBD diagnosis (Table 3, Figure 1, Figures S1-S12). There was a significantly increased HR in CD for cerebrovascular disease [1.49 (1.26–1.75)], paraplegia/hemiplegia [1.66 CI (1.19–2.30)] diabetes [1.14 (CI 1.03–1.27)], and dementia [1.44 (CI 1.16–1.78)], but no increase was evident for any of these conditions in UC. For chronic pulmonary disease, the increase in UC was significant [1.12 (1.003–1.26)] but the increase for CD was not significant. For all other comorbidities the increased HR was evident in both CD and UC, however, was increased to a greater extent in CD for peptic ulcer disease and renal disease. The greatest increased HR post-diagnosis was seen in connective tissue disease/rheumatic disease [HR = 2.49 (CI 2.21–2.79)] and especially for connective tissue disease/rheumatic disease in CD [HR = 2.88 (CI 2.43–3.39)]. There was a greater increase in HR for females in peripheral vascular disease [1.83 (1.57, 2.13)] vs males [1.25 (CI 1.11, 1.42)] and for connective tissue disease/rheumatic disease the HR was increased in males [2.93 (CI 2.41–3.57)] compared to females [2.28 (CI 1.97–2.64)]. For all other comorbidities with an increased HR post-IBD diagnosis, the HR was similar by sex. Compared to other age groups persons under age 25 had the greatest increased HR for peripheral vascular disease, connective tissue disease/rheumatic disease, peptic ulcer disease, and liver disease, and did not have an increased HR for chronic pulmonary disease, or paraplegia/hemiplegia. The largest HR for a comorbidity post-IBD diagnosis was for connective tissue disease/rheumatic disease in persons under age 25 [5.70 (CI 3.43–9.49)]. There was an increase in all comorbidities in adults over age 25 however for diabetes an increased HR was evident only for those over age 50 and for dementia there was an increase in HR for persons 25–50 years for all IBD and 25–65 years for CD.
Figure 1.
Hazard ratios of comorbid diseases in IBD and Crohn's disease and ulcerative colitis compared to matched controls
The number of persons with IBD needed to present with a specific comorbidity was lowest for peptic ulcer disease by 10 years at 21 and was lowest for connective tissue disease/rheumatic disease by 20 years at 11 (Table 4). For all comorbidities except diabetes, the median age at diagnosis was lower in IBD compared to the age of first diagnosis in controls (Table 5).
As an association with paraplegia/hemiplegia was unexpected, we explored whether persons with these health system contacts had prior contacts for cerebrovascular disease, muscular dystrophy, cerebral palsy and motor vehicle accident. For each of the latter three diagnoses the incidence was less than six and hence, cannot be reported. However, of 184 persons with IBD without cerebrovascular disease or paraplegia/hemiplegia prior to diagnosis of IBD, 56% were subsequently diagnosed with paraplegia/hemiplegia alone, 15% with paraplegia/hemiplegia followed by cerebrovascular disease and 29% with cerebrovascular disease followed by paraplegia/hemiplegia, compared to 1155 matched controls with 43%, 25% and 32% in those categories (P = 0.0018). There was no difference between IBD and controls for age at diagnosis of paraplegia/hemiplegia. If persons with prior cerebrovascular disease are excluded from an analysis for paraplegia/hemiplegia the HR for IBD was 1.48 (CI 1.22, 1.79), for CD was 1.58 (CI 1.19, 2.10) and for UC was 1.41 (CI 1.09, 1.82) (Figure 1).
Aliment Pharmacol Ther. 2021;54(5):637-651. © 2021 Blackwell Publishing