An International Reporting Registry of Patients with Celiac Disease and COVID-19: Initial Results from SECURE-CELIAC.

Surveillance Epidemiology Under Research Exclusion for Celiac Disease (SECURE-CELIAC) is an international, de-identified adult and pediatric database created to monitor and report on the severity of coronavirus disease 2019 (COVID-19) outcomes in patients with celiac disease (CD).

Methods

The SECURE-CELIAC registry (https://covidceliac.org/) was modeled on (and with support from) a comparable registry for patients with inflammatory bowel disease (https://covidibd.org/). It was established on March 31, 2020, and promoted via physician email lists, national societies, and social media. Clinicians worldwide were asked to report all confirmed cases of COVID-19 (through viral polymerase chain reaction swab or serology) in patients with CD using a secure online data entry platform. Clinicians were counseled to report cases only after a minimum of 7 days and resolution of illness or death. A choropleth map to illustrate geographic differences in reported cases of COVID-19 in those with CD was created using QGIS 3.44. An interactive online website was created using ArcGIS Online and ArcGIS Pro 2.4.1 to visualize current data by time, country, age, sex, hospitalizations, and deaths (https://arcg.is/1PyiXD). This study was approved by the institutional review board at Columbia University Medical Center. Statistical analyses were conducted using SAS v9.4 (SAS Software, Cary, NC). We aimed to assess for CD-specific factors associated with severe COVID-19 outcomes, defined as hospitalization or death. Continuous and categorical variables were analyzed using the Mann-Whitney U test and the Fisher exact test, respectively. Multivariable analysis was performed using logistic regression.

Results

Between March 31, 2020, and February 23, 2021, 123 confirmed cases of COVID-19 were reported by 60 physicians in the registry, including patients from 21 countries across 5 continents (Supplementary Figure 1). Demographic and clinical characteristics are detailed in Supplementary Table 1. Median age was 38 years. The majority of patients were female (76.4%), White (90.2%), diagnosed via duodenal biopsy (86.2%), and either on a strictly gluten-free diet or with rare inadvertent gluten exposure (82.1%). A total of 39.0% had at least 1 comorbidity, while 8.9% were on at least 1 immunosuppressive medication at time of COVID-19 diagnosis. A total of 36.6% of reported COVID-19 cases experienced new gastrointestinal symptoms such as diarrhea, nausea, vomiting, and abdominal pain. A total of 13.9% of cases (n = 17) met the primary outcome of hospitalization or death. A total of 11.4% (n = 14) were hospitalized, 0.8% (n = 1) required intensive care, and 2.4% (n = 3) died from complications related to COVID-19. Patients who met the primary outcome tended to be older (46 years vs 38 years; P = .030), be Hispanic/Latino (35.3% vs 21.9%; P = .010), and have a history of cardiovascular disease or stroke (23.5% vs 1.0%; P = .001). There was a nonsignificant relationship between the primary outcome and presence of ≥1 comorbidity (58.8% vs 36.5%; P = .109), immunosuppressive treatments (17.7% vs 7.7%; P = .185), strictly gluten-free diet (52.9% vs 66.7%; P = .286), elevated tissue transglutaminase levels prior to COVID-19 diagnosis (11.8% vs 10.5%; P = 1.000), Marsh 3 villous atrophy on follow-up biopsy prior to COVID-19 diagnosis (17.7% vs 8.6%; P = .531), and refractory CD (17.7% vs 5.7%; P = .163). On multivariable analysis (Table 1 ), age (adjusted odds ratio, 1.04; 95% confidence interval [CI], 1.002–1.09; P = .039) and presence of new gastrointestinal symptoms on diagnosis of COVID-19 (adjusted odds ratio, 3.78; 95% confidence interval, 1.04–13.76; P = .044) were associated with increased odds of hospitalization or death.

Table 1

Logistic Regression for COVID-19 Outcomes in the SECURE-CELIAC Cohort

Characteristics	Unadjusted OR (95% CI; P Value)	Adjusted OR (95% CI; P Value)a
Age	1.04 (1.01–1.07; P = .023)	1.04 (1.002–1.09; P = 0.039)b
Male	2.03 (0.68–6.11; P = .206)	3.66 (0.88–15.30; P = .075)
Non-Whitec	2.57 (0.61–10.85; P = .199)	1.71 (0.30–9.93; P = .549)
Hispanic/Latino	3.55 (0.99–12.73; P = .052)	4.39 (0.94–20.57, p =0.060)
Strictly gluten-free diet	0.56 (0.20–1.58; P = .276)	0.34 (0.09–1.26; P = .107)
Number of comorbidities	1.27 (0.83–1.94; P = .263)	0.81 (0.42–1.57; P = .527)
Immunosuppressive treatment	2.57 (0.61–10.86; P = .199)	2.29 (0.34–15.28; P = .393)
GI symptoms	2.74 (0.90–8.34; P = .075)	3.78 (1.04–13.76; P = .044)b

CI, confidence interval; COVID-19, coronavirus disease 2019; GI, gastrointestinal; OR, odds ratio; SECURE-CELIAC, Surveillance Epidemiology Under Research Exclusion for Celiac Disease.

ORs were adjusted for other variables in the table.

Statistically significant.

The non-white patient group included patients who are Asian, Black, biracial, and other. Please see Supplementary Table 1 for more information.

Supplementary Table 1

Outcomes by Demographic and Clinical Characteristics of the SECURE-CELIAC Cohort

	Total (N = 123)	Hospitalized or Death (n = 17)
Age, y	38 (9–78)	46 (24–75)
<18 y	4 (3.3)	0 (0)
18–39 y	61 (49.6)	5 (29.4)
40–64 y	45 (36.6)	9 (53.0)
≥65 y	13 (10.6)	3 (17.7)
Sex
Female	94 (76.2)	11 (64.7)
Male	28 (22.8)	6 (35.3)
Other	1 (0.8)	0 (0)
Race
White	111 (90.2)	14 (82.4)
Asian	3 (2.4)	0 (0)
Black	1 (0.8)	1 (5.9)
Biracial	1 (0.8)	0 (0)
Other	6 (4.9)	2 (11.8)
Unknown	1 (0.8)	0 (0)
Ethnicity
Hispanic/Latino	29 (23.6)	6 (35.3)
Not Hispanic/Latino	74 (60.2)	5 (29.4)
Unknown	20 (16.3)	6 (35.3)
Years since diagnosis
<1 y	14 (11.4)	0 (0)
1–5 y	33 (26.8)	2 (11.8)
6–10 y	29 (23.6)	4 (23.5)
11–20 y	32 (26.0)	6 (35.3)
21–30 y	4 (3.3)	0 (0)
>30 y	9 (7.3)	3 (17.7)
Unknown	2 (1.6)	2 (11.8)
Adherence to gluten-free diet
Strictly gluten-free diet	79 (64.2)	9 (52.9)
Usually gluten-free with rare unintentional gluten consumption	22 (17.9)	2 (11.8)
Usually gluten-free with rare intentional gluten consumption	3 (2.4)	0 (0)
Gluten-free most of the time	1 (0.8)	1 (5.9)
Gluten-free sometimes	2 (1.6)	1 (5.9)
Trying to follow a gluten-free diet but not always sure	6 (4.9)	1 (5.9)
Unrestricted gluten but other foods restricted	2 (1.6)	1 (5.9)
Unrestricted diet	7 (5.7)	2 (11.8)
TTG level in preceding year
Normal	38 (30.9)	5 (29.4)
Elevated	13 (10.6)	2 (11.8)
Unknown	72 (58.5)	10 (58.8)
Follow-up histology
Normal villi (Marsh 0, 1, or 2)	39 (31.7)	5 (29.4)
Villous atrophy (Marsh 3)	12 (9.8)	3 (17.7)
Unknown	72 (58.5)	9 (52.9)
Refractory Celiac Disease
Type 1	4 (3.3)	1 (5.9)
Type 2	5 (4.1)	2 (11.8)
No	82 (66.7)	9 (52.9)
Unknown	32 (26.0)	5 (29.4)
Immunosuppressive medications
Steroids	9 (7.3)	3 (17.7)
Immunomodulators (e.g. azathioprine)	1 (0.8)	0 (0)
Biologics (eg, TNF inhibitor)	2 (1.6)	0 (0)
None	110 (90.2)	14 (82.4)
Comorbidities	48 (39.0)	10 (58.8)
Chronic lung disease (asthma, COPD, etc.)	18 (14.6)	1 (5.9)
Hypertension	11 (8.9)	3 (17.7)
Diabetes	4 (3.3)	1 (5.9)
Cardiovascular disease/stroke	5 (4.1)	4 (23.5)
Chronic liver disease	2 (1.6)	0 (0)
Chronic renal disease	1 (0.8)	0 (0)
Cancer	1 (0.8)	0 (0)
Autoimmune condition (psoriasis, Hashimoto thyroiditis, etc.)	8 (6.5)	3 (17.7)
Gastrointestinal symptoms	45 (36.6)	9 (52.9)
Abdominal pain	24 (19.5)	3 (17.7)
Diarrhea	36 (29.3)	5 (29.4)
Nausea/vomiting	19 (15.5)	2 (11.8)
Other (eg, bloating, anorexia)	8 (6.5)	2 (11.8)
Hospitalized	14 (11.4)	14 (82.4)
LOS, d	7 (3–100)	7 (3–100)
ICU and/or ventilator	1 (0.8)	1 (5.9)
Death from COVID-19	3 (2.4)	3 (17.7)

Values are median (range) or n (%).

COPD, chronic obstructive pulmonary disease; COVID-19, coronavirus disease 2019; ICU, intensive care unit; LOS, length of stay; SECURE-CELIAC, Surveillance Epidemiology Under Research Exclusion for Celiac Disease; TNF, tumor necrosis factor; TTG, tissue transglutaminase.

Discussion

In this international health professional–reported registry of patients with CD who contracted COVID-19, we observed a hospitalization rate of 12% and a case fatality rate of 2.5%. COVID-19 hospitalization and case fatality rates in the general population are estimated at 20% , and 1%–11%,4, 5, 6, 7 respectively, suggesting that patients with CD are not at increased risk of hospitalization or death. Observed and expected deaths by age are reported in Supplementary Table 2. We found that although CD serologies and histology were not predictive of outcomes, CD patients with increased age and new gastrointestinal symptoms had higher odds of adverse outcomes from COVID-19; similar findings have been reported in individuals without CD.7, 8, 9 A recent population-based study of CD patients in Sweden found no increased risk of COVID-19–related outcomes among people with CD. To our knowledge, this is the first study evaluating predictors of severe of COVID-19 outcomes among patients with CD. Strengths include its international nature and reliance on clinician reporting, likely enhancing the veracity of both CD and COVID-19 diagnoses. Limitations include selection bias either secondary to access to COVID-19 testing or cases that escape provider notification. However, as these cases tend to be milder, they are unlikely to affect our study findings. Language barriers may have limited participation from non-English speaking countries.

In conclusion, patients with CD are not at increased risk of hospitalization or death from COVID-19. Increased age and new gastrointestinal symptoms are associated with adverse COVID-19 outcomes in patients with CD.

Supplementary Table 2

Observed and Expected Deaths by Age for the SECURE-CELIAC Cohort vs General Population

Age Group	SECURE-CELIAC Cases	SECURE-CELIAC Observed Deaths	SECURE-CELIAC Case Fatality Ratio	Case Fatality Population in General Populationa	SECURE-CELIAC Expected Deaths	Observed-to-Expected Deaths
<10 y	1	0	0.00%	0.05%	0.0005	0
10–19 y	10	0	0.00%	0.03%	0.003	0
20–29 y	23	0	0.00%	0.04%	0.0092	0
30–39 y	31	1	3.23%	0.12%	0.0372	26.88
40–49 y	27	2	7.41%	0.23%	0.0621	32.21
50–59 y	14	0	0.00%	0.67%	0.0938	0
60–69 y	9	0	0.00%	2.42%	0.2178	0
70–79 y	8	0	0.00%	8.38%	0.6704	0
80+ y	0	0	0.00%	15.10%	0	0

SECURE-CELIAC, Surveillance Epidemiology Under Research Exclusion for Celiac Disease.

Estimates are based on population-level data from 9 countries: China, France, Germany, Italy, the Netherlands, South Korea, Spain, Switzerland, and the United States.