(005) Cohort studies on 71 different adverse health outcomes among people with atopic eczema in UK primary care data

Background: Atopic eczema may be related to multiple subsequent adverse health outcomes, however, high quality evidence from studies with designs and data sources suitable for hypothesis-testing is limited. The existing evidence base, encompassing hundreds of studies employing heterogenous approaches to design, analysis and data management, makes results difficult to compare, and has been slow and expensive to build.

Objectives: To generate evidence on multiple adverse health outcomes associated with eczema using an efficient, comprehensive, and systematic approach.

Methods: We conducted 71 cohort studies using primary care electronic health records data from Clinical Practice Research Datalink Aurum (1997 - 2023), with cohort sizes of up to 3.6 million with eczema matched (on age, sex and general practice) to 16.8 million without eczema. Applying an outcome-wide study design and confounding-adjustment strategy, we fitted Cox models, estimating hazard ratios (HRs) for each outcome comparing people with eczema to people without. We also assessed the effect of eczema severity, estimated absolute effects, and conducted a range of sensitivity analyses.

Results: Eczema was associated with outcomes with adjusted HRs (99% confidence intervals) of up to 4.02 (3.95-4.10) for food allergy, 2.02 (1.99-2.05) for allergic conjunctivitis, and 1.93 (1.91-1.94) for allergic rhinitis. Outcomes not included in current guidelines, where results showed strong associations, dose-response relationships, and large rate differences were: irritable bowel syndrome (1.31 [1.29-1.33]), oesophageal ulcer and gastritis (1.26 [1.24-1.27]), gastro oesophageal reflux disease (1.25 [1.24-1.26]), thromboembolic disease, obesity (1.22 [1.21-1.23]), gastritis and duodenitis and peripheral neuropathies (1.21 [1.20-1.22]). Associations with large hazard ratios but lower rate differences included: Hodgkin’s lymphoma (1.83 [1.64-20.4]), Crohn’s disease (1.62 [1.54-1.69]), coeliac disease (1.43 [1.38-1.48]), autoimmune liver disease (1.35 [1.54-1.69]), ulcerative colitis (1.41 [1.35-1.47]), liver fibrosis, sclerosis and cirrhosis (1.27 [1.22-1.33]), non-Hodgkin’s lymphoma (1.26 [1.21-1.32]), fatty liver (1.25 [1.22-1.27]), and oesophageal varices (1.21 [1.11-1.32]).

Conclusions: We identified several novel strong associations between eczema and adverse health outcomes, including some where individuals with eczema, clinicians, and guideline authors may benefit from increased awareness of these risks. Results closely matched those from previous studies specifically designed to investigate cancer, fracture, cardiovascular and mental illness outcomes, suggesting our approach to studying multiple outcomes produces valid estimates.