Inflammatory eye disease is a risk factor for future microscopic colitis: A nationwide population‐based matched case control study

Abstract

Background

Microscopic colitis (MC) is an inflammatory disorder of the colon. To date, the relationship between inflammatory eye diseases and MC is unclear.

Objective

To assess whether inflammatory eye disease (iridocyclitis and episcleritis) is a risk factor for MC.

Methods

We conducted a nationwide matched case control study in Sweden leveraging the ESPRESSO‐study (a Swedish database containing data on all biopsies from the gastrointestinal tract from 1965 to 2017). In total, we identified 14,338 patients with biopsy‐verified MC (diagnosed from 1981 to 2017). Patients with MC were matched (by age, sex, county and year of birth) with 68,753 controls from the general population and the occurrence of preceding inflammatory eye diseases (defined as diagnosis of episcleritis or iridocyclitis) in the two groups was compared. Multivariable adjusted odds ratios (aORs) were calculated using conditional logistic regression conditioned on the matching variables.

Results

A majority of patients with MC were women (71.9%) and the median age at MC diagnosis was 63.3 years (interquartile range (IQR) = 50.7–72.6). Some 225 (1.6%) MC patients had an earlier record of inflammatory eye disease compared with 614 (0.9%) in controls. These figures corresponded to an aOR of 1.77 (95% CI = 1.52–2.07) for inflammatory eye diseases in patients with MC. Compared to siblings, the aOR for previous inflammatory eye diseases in MC was 1.52 (95% CI = 1.17–1.98) and patients treated with budesonide, as a proxy for clinically significant disease, had a somewhat higher aOR for previous inflammatory eye diseases.

Conclusion

Inflammatory eye diseases are more common in patients subsequently being diagnosed with MC. Our findings highlight that these conditions may have shared causes and inflammatory pathways and are of clinical interest to gastroenterologists, ophthalmologists and general practitioners.

Key summary.

Summarize the established knowledge on this subject

• There is a well‐known association between inflammatory eye diseases (episcleritis and iridocyclitis) and subsequent inflammatory bowel disease.

• To the best of our knowledge, the association between inflammatory eye diseases and microscopic colitis (MC) has not been explored.

What are the significant and/or new findings of this study?

• Patients diagnosed with inflammatory eye diseases are at a 77% increased risk of subsequent MC compared with matched controls from the general population.

• Both episcleritis and iridocyclitis were significantly associated with future MC.

• Also, when using siblings as controls, the association remained significantly elevated, suggesting that our observed association cannot be explained by shared intrafamilial confounding.

INTRODUCTION

Microscopic colitis (MC) is an inflammatory disorder of the colon, ¹ typically presenting with watery diarrhea ² in middle‐aged women. ³ As the name of the disease implies, there are generally no macroscopic aberrations in the colonic mucosa ¹ and the inflammation is less severe ⁴ than in classical inflammatory bowel diseases (IBD), that is, Crohn's disease or ulcerative colitis. The diagnosis of MC is dependent on a biopsy from the colonic mucosa and based on the histopathological findings, MC is characterized as one of two subtypes: collagenous colitis (CC) or lymphocytic colitis (LC). ⁵ Studies from Sweden, as well as from the U.S. and Denmark have all noted increasing incidence rates of MC ³ , ⁶ , ⁷ over the past decades, with current estimates resembling, or even surpassing, those of ulcerative colitis and Crohn's disease. ⁸ Knowledge on extraintestinal manifestations in MC is limited and mainly relies on studies carried out on smaller study populations. ⁹ Moreover, the disease has rarely been considered to be associated with adverse health outcomes. Nevertheless, recent studies have found significant associations between MC and rheumatoid arthritis, ¹⁰ acute pancreatitis ¹¹ and major adverse cardiovascular events. ¹² Also, there is a well‐established link between MC and autoimmune disorders. ¹³ To the best of our knowledge, the relationship between MC and inflammatory eye diseases has not yet been explored.

Iridocyclitis and episcleritis are the most common eye manifestations in patients with IBD ¹⁴ , ¹⁵ , ¹⁶ and also occur more frequently in patients with celiac disease. ¹⁴ , ¹⁷ Previous research has found that inflammatory eye diseases often occur concomitantly with or even precede the onset of IBD. ¹⁸ MC and iridocyclitis share inflammatory characteristics in that they have both been linked to elevated levels of TH‐17 ¹⁹ , ²⁰ and more than one third of patients with episcleritis have been found to have an underlying inflammatory disease. ²¹ Thus, we hypothesized an association between inflammatory eye diseases and subsequent MC. To test this hypothesis, we leveraged a large nationwide histopathology cohort to elucidate the association between iridocyclitis, episcleritis and MC. Added knowledge in this field could shorten the diagnostic delay for patients with MC, and add to the scientific literature on extraintestinal inflammation in MC.

MATERIALS AND METHODS

Setting

Healthcare in Sweden is largely publicly funded and devised to offer all citizens equal access to healthcare. The data used in this study originate from Swedish nationwide healthcare registers as well as government‐maintained registers on the Swedish population. At birth or immigration, all Swedish citizens are provided with a unique personal identification number. ²² This number enabled us to gather relevant information on our study participants from various sources with virtually no missing data.

Data sources

Data on our outcomes were extracted from the Swedish National Patient Register (NPR). ²³ This register collects information on diagnoses and medical procedures using International Classification of Diseases (ICD) codes. The NPR reached nationwide coverage in 1987 and, since 2001, also included data from specialized outpatient care.

Details on the dispensed medications were retrieved from the Prescribed Drug Register. ²⁴ Since 1 July 2005, this register has almost complete coverage for dispensed medication throughout Sweden.

Matched controls were identified from the Total Population Register. ²⁵ This register maintains records of personal identity number, vital status, migration, and residence of all Swedish citizens dating back to 1968. Siblings of patients with MC were identified by linkage to the Multigeneration Register, ²⁵ a component of the Total Population Register. This register archives data on all Swedish citizens born from 1932 until the present day, along with their siblings and biological parents.

Definition of independent variable (uveitis or episcleritis)

Our exposure, inflammatory eye disease, was a composite of two predetermined disorders (Table S1), episcleritis and iridocyclitis (occurring before matching date). Of note, uveitis is an umbrella term encompassing anterior, intermediate, and posterior uveitis. IBD is generally associated with anterior uveitis, also known as iridocyclitis. ¹⁵ , ¹⁶ In secondary analysis, we also examined the association between the individual diseases that constituted our main exposure (Table S1).

The accuracy of our composite exposure has not been evaluated. However, the positive predictive value (PPV) of uveitis in the NPR is about 93%. ¹⁷

Definition of dependent variable (MC)

Sweden hosts 28 regional pathology registers, all of which use the Systematized Nomenclature of Medicine (SNOMED) system when storing information on the morphology of biopsies. In the current study, MC was defined as a SNOMED code for CC (M40600) or for LC (M47170). In a previous study, when reviewing randomly selected patient charts belonging to patients with a biopsy indicative of MC, we found a PPV of 95%. Moreover, 96% of patients had diarrhea prior to diagnosis, validating the precision of our definition in identifying clinically active disease. ²⁶

The ESPRESSO (Epidemiology Strengthened by histoPathology Reports in Sweden) study ²⁷ has gathered data on all gastrointestinal biopsies taken in Sweden from 1965 to 2017. Thus, this data source allowed us to identify all individuals with a colonic biopsy consistent with MC. Thereafter, up to five controls per patient with MC were randomly selected from the general population, thus creating a matching set. These were matched for sex, birth year, and county of residence all at the time of MC diagnosis. Full siblings (without MC) to patients with MC were also identified through the Total Population Register. ²⁵

Covariates

Baseline characteristics (sex, age, county of residence, date of death, emigration, and country of birth) for all individuals were gathered from the Total Population Register. ²⁵

Educational attainment was used as a surrogate for socioeconomic status. Educational levels were categorized as compulsory school (≤9 years), upper secondary school (10–12 years), or college (≥13 years). This data was gathered from the longitudinal integrated database for health insurance and labor market studies. ²⁸ In instances where data on the level of education was missing, the highest level of education of the parents was used. In some cases, information on the parents' education was also missing and these study participants were placed in a “missing” category.

Comprehensive information on the source and definition of our other covariates are included in Table S1. As there is no validated method to assess disease severity in MC, we regarded dispensed budesonide 7 days prior to or after MC diagnoses (ATC code A07EA06) as indicative of clinically significant disease. Information on dispensed budesonide was collected from the Prescribed Drug Register. ²⁴

Exclusion criteria

Figure 1 illustrates the exclusions. The exclusion criteria were applied consistently to all study participants. In summary, any individual with data irregularities (pertaining to emigration or death) or colectomy before the matching date was excluded from the study.

FIGURE 1.

Flowchart of inclusion of patients with biopsy‐confirmed microscopic colitis in the ESPRESSO histopathology cohort, siblings, and matched general population controls from the Swedish Total Population Register 1981–2017. ESPRESSO, epidemiology strengthened by histopathology reports.

Sensitivity analyses

First, to control for the impact of socioeconomic status, we added educational attainment as a covariate. Second, to control for the impact of other inflammatory events, a separate analysis was conducted where inflammatory comorbidities (Table S1) were added as covariates to our model. Finally, to assess the effect of familial confounding and co‐aggregation, full siblings, regardless of sex, were used as controls.

Patient and public involvement

No patient participated in the planning or design of this study.

Statistical analysis

We compared the occurrence of prior inflammatory eye diseases in patients with MC to that of the matched controls. We computed odds ratios (ORs) and confidence intervals (CIs) using conditional logistic regression. In the sensitivity analysis, we also adjusted for relevant comorbidities (for definitions see Table S1). In these analyses, the matching variables (birth year, sex, county of residence, and year of biopsy) were also added to the model in order to avoid them introducing confounding. ²⁹ Further, to avoid adjusting for mediators, comorbidities were only adjusted for if they occurred before a specific time point—namely the first date per matching set (i.e., each case and her or his controls) when an inflammatory eye disease was diagnosed. For example, if a control was diagnosed with episcleritis on January 1, 1998, any IBD diagnosed in the corresponding case and the other matched controls before this date was considered a confounder. As for matching sets with no record of an independent variable (inflammatory eye disease), the conditional logistic regression inherently disregards such sets; therefore, defining comorbidities in these instances becomes a moot issue.

Using the same model, we explored the association between MC and inflammatory eye diseases in subgroups defined by sex, age at the index date (<50 or ≥50 years), calendar year matching date (<1990, 1990–2000, 2001–2010, 2011–2017), country of birth (Nordic (Sweden, Norway, Denmark, Finland, and Iceland) or other), and educational attainment (≤9, 10–12, ≥13 years).

In our sibling‐controlled analysis, we conditioned on family, that is, patients with MC were exclusively compared to their own siblings.

To assess the impact of disease severity and dispensed budesonide, two cohorts ‐ defined by whether the exposed individuals had at least one recorded dispensation of budesonide—were created.

All statistical analyses used Stata/IC 17.0 for Mac (StataCorp).

Ethics

This study was approved by the Stockholm Ethics Review Board. Because the study was strictly register‐based, informed consent was not required. ³⁰

RESULTS

From 1981 to 2017, we identified 14,338 patients with incident MC. These MC cases were matched to 68,753 controls from the general population (Table 1). Moreover, 14,656 full siblings, without MC, were identified. As expected, a majority (71.9%) of patients with MC were women. The median age at MC diagnosis was 63.3 years (interquartile range (IQR) = 50.7–72.6). Some 94% of patients with MC were born in a Nordic country and 30.0% had an educational attainment corresponding to university or college (≥13 years of schooling).

TABLE 1.

Summary statistics for patients with microscopic colitis (MC), collagenous colitis (CC) and lymphocytic colitis (LC) and matched controls.

	Controls n [%]	MC n [%]	CC n [%]	LC n [%]
Total	68,753 [100.00]	14,338 [100.00]	4616 [100.00]	9722 [100.00]
Male	19,189 [27.9]	4025 [28.1]	1071 [23.2]	2954 [30.4]
Female	49,564 [72.1]	10,313 [71.9]	3545 [76.8]	6768 [69.6]
Age at matching date
Mean [SD] years	60.1 [16.8]	60.7 [16.9]	63.8 [15.0]	59.2 [17.6]
Median [IQR] years	62.7 [50.1–72.6]	63.3 [50.7–73.2]	65.5 [54.9–74.7]	62.0 [48.4–72.4]
<50 years	17,063 [24.8]	3453 [24.1]	792 [17.2]	2661 [27.4]
> = 50 years	51,690 [75.2]	10,885 [75.9]	3824 [82.8]	7061 [72.6]
Year of matching date
≤1990	20 [0.03]	4 [0.03]	0 [0]	4 [0.04]
1991–2000	4521 [6.6]	929 [6.5]	291 [6.3]	638 [6.6]
2001–2010	36,191 [52.6]	7533 [52.5]	2512 [54.4]	5021 [51.7]
2011–2017	28,021 [40.8]	5872 [41.0]	1813 [39.3]	4059 [41.2]
Eye disorders prior to matching date
Eye disorder a	614 [0.9]	225 [1.6]	73 [1.6]	152 [1.6]
Iridocyclitis	471 [0.7]	176 [1.2]	57 [1.2]	119 [1.2]
Episcleritis	165 [0.2]	56 [0.4]	18 [0.4]	38 [0.4]
Country of birth
Nordic	61,008 [88.7]	13,462 [93.9]	4423 [95.8]	9039 [93.0]
Other	7745 [11.3]	876 [6.1]	193 [4.2]	683 [7.0]
Education
Compulsory school (< = 9 years)	19,174 [27.9]	3781 [26.4]	1400 [30.3]	2381 [24.5]
Upper secondary school (10–12 years)	27,457 [39.9]	5851 [40.8]	1913 [41.4]	3938 [40.5]
College or university (> = 13 years)	19,670 [28.6]	4302 [30.0]	1191 [25.8]	3111 [32.0]
Missing	2452 [3.6]	404 [2.8]	112 [2.4]	292 [3.0]
Medication during follow‐up
Budesonide	178 [0.3]	6304 [44.0]	2399 [52.0]	3905 [40.2]

^a Defined as episcleritis or anterior uveitis (iridocyclitis) For definitions of outcomes and comorbidities, see Table S1 for relevant international classification of disease (ICD) codes.

Eye disease preceding MC

In total, we ascertained 839 events of earlier inflammatory eye diseases. Of these, 225 occurred in patients with subsequent MC (corresponding to 1.6%), while controls had suffered 614 events (corresponding to 0.9%) of inflammatory eye diseases. Notably, median age at iridocyclitis was lower in patients with MC (56.9, IQR = 43.4–68.3) compared with controls (60.6, IQR = 48.4–70.1).

Exposed patients with MC had a median time of 4.9 years (IQR = 2.0–7.7) since the first inflammatory eye disease diagnosis. Median time intervals for CC and LC were similar.

Taking the matching variables (birth year, sex, county, calendar year and educational attainment (as a proxy of socioeconomic status)) into account, yielded an adjusted odds ratio (aOR) of 1.77 (95% CI = 1.52–2.07) (Table 2) for inflammatory eye disease in patients with MC. Stratifying by sex, aORs were almost identical, with a somewhat broader CI for males, reflecting the female predominance in MC. Examining the association according to age at matching date, younger than 50 or older, we found a significantly higher aOR (P _{for heterogeneity} = 0.007) for patients diagnosed with MC at an earlier age, aOR 2.76 (95% CI = 1.94–3.92). We also stratified by calendar period at matching date. Owing to MC not being routinely diagnosed in Sweden before the mid‐1990s, the ORs for the two earlier calendar periods (≤1990 and 1991–2000) were largely uninformative, whereas the ORs for the latter calendar periods (2001–2010 and 2011–2017) were comparable (P _{for heterogeneity} = 0.20). Estimates for CC and LC were comparable (Table 2).

TABLE 2.

Odds ratios for inflammatory eye disease ^a among patients diagnosed with microscopic colitis (MC), collagenous colitis (CC), or lymphocytic colitis (LC) in Sweden compared with matched controls.

	Number of events (eye disease a preceding matching date)		MC	CC	LC
	Controls (n = 68,753)	Patients with MC (n = 14,338)	OR [95% CI]	HR [95% CI]	HR [95% CI]
Total	614	225	1.77 [1.52–2.07]	1.68 [1.29–2.21]	1.82 [1.50–2.19]
Sex
Males	184	67	1.73 [1.31–2.30]	1.57 [0.90–2.75]	1.80 [1.29–2.49]
Females	430	158	1.79 [1.49–2.15]	1.72 [1.27–2.34]	1.83 [1.45–2.30]
Age at start follow up (years)
<50	88	49	2.76 [1.94–3.92]	3.96 [2.03–7.74]	2.42 [1.59–3.67]
≥50	526	176	1.60 [1.35–1.91]	1.44 [1.06–1.94]	1.70 [1.37–2.10]
Year of MC diagnosis
≤1990	0	0	NA [NA–NA] b	NA [NA–NA] b	NA [NA–NA] b
1991–2000	4	4	4.99 [1.25–20.0]	9.74 [0.88–107.4]	3.35 [0.56–20.1]
2001–2010	240	97	1.97 [1.56–2.50)	1.71 [1.13–2.59]	2.13 [1.59–2.84]
2011–2017	370	124	1.61 [1.31–1.97]	1.59 [1.11–2.29]	1.61 [1.25–2.07]
Country of birth
Nordic	532	209	1.72 [1.46–2.02]	1.59 [1.20–2.12]	1.79 [1.47–2.18]
Other	1	7	8.22 [0.99–67.9]	3.00 [0.31–28.9]	NA [NA‐NA] b
Education (years)
Compulsory school (≤9)	49	46	2.11 [1.39–3.19]	1.41 [0.73–2.76]	2.78 [1.59–4.85]
Upper secondary school (10–12)	98	74	1.71 [1.25–2.33]	2.65 [1.58–4.45]	1.33 [0.90–1.98]
College or university (≥13)	64	56	1.63 [1.12–2.38]	1.77 [0.83–3.75]	1.59 [1.03–2.45]
Unknown	1	4	13.9 [1.47–131.7]	NA [NA–NA] b	9.07 [0.87–95.0]

Note: Adjusted for age, sex, county, calendar period and level of education.

^a Defined as anterior iridocyclitis or episcleritis, see Table S1 for relevant international classification of disease (ICD) codes.

^b No ORs could be due to insufficient number of events.

Secondary analyses

Comprehensive results of our secondary analyses are presented in Figure 2. In summary, both episcleritis and iridocyclitis were significantly associated with MC.

FIGURE 2.

Forest plot depicting odd ratios for earlier inflammatory eye disease among patients with microscopic colitis (MC). The number of individual diseases exceeded the total, since it was possible to have more than one eye diagnose.

Sensitivity analyses and exploratory analyses

Several sensitivity analyses were conducted to test the robustness of our results. Briefly, these analyses corroborated our main result. Definitions and complete results are outlined in Table 3. Of note, the aOR among MC patients with a recorded dispensation of budesonide was higher than for those without such a dispensation, aOR 2.03 (95% CI = 1.63–2.53) versus 1.60 (95% CI = 1.28–2.01). The difference between the two estimates, however, failed to reach significance (P _{for heterogeneity} = 0.14). Also, when adjusting for possibly confounding autoimmune disorders (listed in Table 3), the aOR for earlier inflammatory eye disease remained significantly elevated, aOR 1.64 (95% CI = 1.40–1.93).

TABLE 3.

Sensitivity analyses.

	Adjusted odds ratios and 95% confidence intervals
Adding adjustment for educational attainment b	1.76 [1.51–2.06]
Adding adjustment for potentially confounding comorbidities c	1.64 [1.40–1.93]
MC with dispensed budesonide	2.03 [1.63–2.53]
MC without dispensed budesonide	1.60 [1.28–2.01]
Full siblings (reference)	1.52 [1.17–1.98]

Note: Adjusted odds ratios for eye disease ^a in patients diagnosed with MC in Sweden from 1981 to 2017 compared to matched reference individuals.

^a Defined as episcleritis or iridocyclitis. For definitions (ICD‐codes) see Table S1.

^b Educational levels (used as a proxy for socioeconomic status) were categorized as compulsory school (≤9 years), upper secondary school (10–12 years), college (≥13 years), or missing.

^c Rheumatoid arthritis, Inflammatory bowel disease, Type 1 diabetes, celiac disease, autoimmune thyroid disease, psoriasis, multiple sclerosis, lymphoma, and psoriatic arthritis. For definitions, see Table S1.

DISCUSSION

In this nationwide population‐based matched case‐control study, including over 14,300 patients with MC we found that patients diagnosed with inflammatory eye disease (episcleritis and iridocyclitis) were at 77% increased odds of later MC. The association was robust across several sensitivity analyses and both episcleritis and iridocyclitis were significantly associated with MC. The increased risk was more pronounced in patients diagnosed with MC at a younger age.

A comparison with the literature

As far as we know, the association between inflammatory eye disease and MC has not yet been explored. However, there are a plethora of studies focusing on the association between IBD and episcleritis/iridocyclitis ¹⁴ , ¹⁵ , ¹⁶ and similar findings have also been reported for celiac disease. ¹⁷

The association between inflammatory eye disease and MC was assessed across various strata. Generally, these stratified estimates were comparable except for age groups (<50 vs. ≥ 50 years of age) where patients diagnosed with MC before the age of 50 years had significantly higher odds compared to older patients. This pattern was also consistent when looking at the association between CC and LC. Also, our findings were robust across several sensitivity analyses. First, to control for the impact of socioeconomic status, we adjusted for educational attainment; this addition to our model left our point estimate virtually unchanged. Second, to control for the potential confounding effect of comorbidities and the impact of an autoimmune predisposition, nine diseases, both linked to MC and inflammatory eye disease, were added as covariates. Again, the aOR did not change materially. Third, to assess the impact of disease intensity, we defined clinically active MC as having a record of at least one dispensation of budesonide. Compared to those without such a dispensation, the aOR was slightly higher. Finally, to control for intrafamilial confounding we used full siblings to our MC patients as controls yielding a somewhat attenuated aOR, a finding resembling results from previous research ³¹ showing that first degree relatives to patients with IBD are more prone to develop inflammatory eye diseases than the general population is. Nevertheless, the association between inflammatory eye disease and MC does not seem to be entirely due to shared genetics and early environmental factors.

We also observed similar estimates for CC and LC. Although previous studies have reported differing genetic associations for the two subtypes, ³² , ³³ , ³⁴ our finding of a materially unchanged aOR in the sibling‐controlled analyses as well as the comparable aORs for LC and CC indicates that shared inflammatory characteristics and an autoimmune predisposition, rather than genetic factors, are the main drivers of our observed association.

Biologic mechanisms

Our findings have biological plausibility. The inflammatory activity in both MC and iridocyclitis is characterized by an upregulation of the TH‐17(19, 20) pathway and some 30% of patients with episcleritis have been found to have a concomitant inflammatory condition. ²¹ Episcleritis and iridocyclitis occurring in patients with IBD are generally considered as extensions of immune responses in the gut, thus qualifying as a true extraintestinal manifestation. Contrary, other ocular manifestations such as dry eyes or cataracts occur as independent inflammatory events, potentially aided by the IBD or as side effects of the treatment. ³⁵ Moreover, in patients with IBD, episcleritis is regarded as an important indicator of the inflammatory activity in the gut, whereas iridocyclitis may signal the potential onset of IBD. ³⁶ In general, these views strengthen our definition of inflammatory eye disease and lend relevance to our findings. Moreover, the robustness of our finding when adding immune‐mediated diseases as covariates suggests that the association may not be entirely driven by an autoimmune predisposition.

Strengths and weaknesses

Our study has several strengths. First, due to the nationwide coverage of the ESPRESSO study, ²⁷ the impact of selection bias is minimized. To the same effect, the size of our cohort allowed precise computations across various strata. Also, our previous validation work, ²⁶ lends credibility to our definition of MC. Moreover, the personal identity number ²² made linkage to and retrieval of information from several registers possible, allowing us to build a large and comprehensive cohort of cases and controls. Finally, our ability to identify full siblings gave us a unique opportunity to address intrafamilial confounding.

There are also limitations. Our study was based on data originating from Swedish registers; consequently, we lacked information on potentially confounding life‐style factors such as BMI and smoking. After reviewing the literature, BMI seems unlikely to be of any major importance in this context, whereas our lack of smoking data should be kept in mind when interpreting our results. ³⁷ , ³⁸ Also, the NPR does not hold data on diagnoses from primary care, which likely has artificially deflated our number of eye inflammatory events. Primarily, this would apply to episcleritis, which is often handled in primary care. Likely, however, this misclassification is non‐differential and, thus, would not explain our observed result. In addition, as our data did not allow for a cohort study, we were unable to calculate absolute risks. Therefore, it should be stressed that only a minor proportion of patients with MC had an earlier record of inflammatory eye disease.

Moreover, in our previous validation work, LC had a somewhat lower PPV compared to CC, 85% and 95%, respectively. ²⁶ Hence, we cannot rule out that some patients, especially those with LC, had a false‐positive MC.

Finally, as our study relies on data from the Swedish population, the external validity might be limited. Therefore, our results may not be directly applicable to populations with a different panorama of lifestyle factors, comorbidities, and ethnicities.

To conclude, inflammatory eye disease is a risk factor for MC. The risk was comparable for both LC and CC and both episcleritis and iridocyclitis were associated with MC. Our findings are of clinical interest to ophthalmologists and general practitioners, but a scientific value also resides with the notion that these conditions may have shared causes.

AUTHOR CONTRIBUTIONS

David Bergman, Jonas F. Ludvigsson: Conceptualization. David Bergman, Jonas F. Ludvigsson: Data curation. David Bergman: Formal analysis. Jonas F. Ludvigsson: Funding acquisition. All authors: Investigation. All authors: Methodology. David Bergman, Jonas F. Ludvigsson: Project administration. Jonas F. Ludvigsson: Resources. Jonas F. Ludvigsson: Software. Jonas F. Ludvigsson: Supervision. David Bergman, Jonas F. Ludvigsson: Validation. David Bergman: Visualization. All authors: Writing − original draft. All authors: Writing − review and editing.

CONFLICT OF INTEREST STATEMENT

Dr Ludvigsson has coordinated a study for the Swedish IBD quality register (SWIBREG). That study received funding from the Janssen Corporation. Dr Ludvigsson has also received financial support from MSD to develop a paper reviewing national healthcare registers in China, and from Takeda for a study on celiac disease. Dr Ebrahimi has served as an advisory board member for Boehringer Ingelheim.

ETHICS APPROVAL

This study was approved by the Regional Ethics Committee, Stockholm, Sweden (Protocol no 2014/1287‐31/4, 2018/972‐32 and 2022‐05774‐02).

Supporting information

Table S1

Bergman D, Ebrahimi F, Sun J, Norin C, Mollazadegan K, Ludvigsson JF. Inflammatory eye disease is a risk factor for future microscopic colitis: a nationwide population‐based matched case control study. United European Gastroenterol J. 2024;12(8):1081–90. 10.1002/ueg2.12623

DATA AVAILABILITY STATEMENT

In accordance with Swedish regulations, the data from this study are not publicly available.