Table 10: Risk of All-Cause Mortality in Patients with Asymptomatic Celiac Disease.

Study N Follow-up	Study Design Statistical Analysis	Patient Population Symptoms	CD diagnosis criteria Outcome and covariates ascertainment	All-Cause Mortality OR, RR, HR, SMR (95% CI) GFD effects	Participation Rate Withdrawals
Corrao et al. (20) (2001) N= 67 asymptomatic CD N= 123 asymptomatic CD (parents and siblings’ cohorts) Period: 1962-1998	• Study Design Cohort, Multicentre Mean f-up: 6.0 yrs • Recruitment CD cases: Consecutive CD diagnosed between 1962-1994 from GI clinics. Relatives’ cohorts: subjects diagnosed at gastroenterology clinics. Controls: general population mortality rate used as comparison, age, sex-adjusted • Analysis Poisson regression SMR (observed/ expected), age and sex adjusted	• Celiac disease > 18 yrs Female: 76% • Controls Not applicable	• CD diagnosis Small bowel biopsy • Mortality ascertainment Mortality and covariates ascertained through patient interviews and review of medical records or city registries. Cause defined according to ICD-9	• All-cause mortality (asymptomatics) SMR 1.2 (0.1, 7.0) Events: 1 (expected: 0.8), p .99 Relatives of CD patients Fathers SMR: 0.8 (0.3, 1.7) Events 7 (expected 8.4) Similar for mothers, brothers and sisters of CD patients • GFD effects Adherence and effects in asymptomatics not reported.	• Participation Rate CD cohort: not reported Relatives cohort: 8/873 (0.9%) • Withdrawals CD cohort: 50 (4.7%) Relatives’ cohorts: not reported
Johnston et al. (24) (1998) N= 13 asymptomatic CD (1983 cohort) Period: 1983-1995	• Study Design Cohort Mean f-up: 11.6 yrs • Recruitment Patients randomly identified from the general population for a cardiovascular study were used for this study. CD cases: positive serology Controls: expected mortality rate in the general population, adjusted for age and sex • Analysis Observed events compared to expected according to age, calendar yr. Poisson distribution	• Celiac disease Mean age: 60.1 yrs Female: 52.8% • Controls Not applicable	• CD diagnosis Serology (IgA AGA, ARA, EMA). Small bowel biopsy performed in 20/102 • Mortality ascertainment Outcome information obtained from patient interviews and mortality registry	• All-cause Mortality in subjects with at least 1 positive serologic test RR 0.92 (0.5, 1.6) Events: 13 (expected 14.1) • GFD effects Not evaluated	• 30/102 (29.4%) did not participate • 52/72 (72.2%) did not do a small bowel biopsy
Lohi et al. (21) (2009) N= 6,849 204 asymptomatic CD (IgA tTG) 74 asymptomatic CD (IgA EMA) Period: 1978-2005	• Study Design Retrospective cohort F-up: up to 28 yrs • Recruitment Patients randomly identified from the general population for a population-based heart survey were used for this study. CD cases: positive serologic test. Controls: Negative serology or positive only to 1st IgA tTG •Analysis Cox regression, RR adjusted for multiple factors	• Celiac disease Mean age ± SD: 49.2 ± 11.8 yrs (EMA) 59.1 ± 14.2 yrs (tTG) Female: 72% (EMA) 61.5% (tTG) •Controls Mean age ± SD: 51.1 ± 14.1 yrs Female: 53.7%	• CD diagnosis Positive serology in 2 tests. • Mortality ascertainment Information obtained from medical records and national mortality registry	• All-cause mortality by serologic test IgA tTG positive RR: 1.18 (0.99, 1.42) Events: 128/204 (CD) 2,941/6,783 (controls) IgA EMA positive RR: 0.91 (0.59, 1.38) Events: 23 /74 (CD) 3,46/6,913 (controls) • GFD effects Not evaluated	• 10% did not participate in survey • 87% with blood sample
Rubio-Tapia et al. (22)(2009) Period: 1948-1997 N=9,133	• Study Design Retrospective cohort F-up: 45 yrs • Recruitment Healthy young men from military base who were part of a cohort study for infectious diseases were used for this study CD cases: positive serologic test Controls: Seronegative for CD Matched for age, gender, enlistment status • Analysis Cox proportional hazards adjusted for age, gender, enlistment status	• Celiac disease Median age: 20 (14.3-46.4) yrs Female: 0 • Controls Not reported	• CD diagnosis Serology, both IgA tTG and IgA EMA positive • Mortality ascertainment Vital records database	• All-cause mortality HR: 3.9 (2.0, 7.5) Events: 9, 6 of known causes: Emphysema (1), lymphoma (1), larynx cancer (1), esophageal cancer (1), CV (1), not specified (1). • GFD effects Unknown (authors)	• Not reported
Metzger et al. (23) (2006) N=4,570 CD= 63 (1.4%) Period: 1989-1998	• Cases Retrospective cohort Mean f-up: 8.0 (0-8.9) yrs • Recruitment Patients randomly identified from the general population for population-based heart survey were used CD cases: positive serologic test for CD Controls: negative serologic test for CD • Analysis Cox proportional hazards adjusted for age	• Celiac disease Mean age: Men: 57.2 (53.1-61.4) yrs Women: 52.8 (46.5-59.0) yrs Female: 49.6% • Controls Mean age: Men: 49.7 (49.2-50.3) yrs Women: 49.2 (48.7-49.8) yrs Female: 49.9%	• CD diagnosis Serology IgA tTG • Mortality ascertainment Obtained from vital records database. Covariates obtained from medical records	• All-cause mortality HR: 2.53 (1.50, 4.25) Events: 15	• N=13 (0.2%)–moved away.

AGA antigliadin antibody; ARA anti-reticulin antibody; CD celiac disease; CI confidence interval; EMA endomysial antibody; f-up follow-up; GFD gluten-free diet; HR hazard ratio; IgA immunoglobulin A; N/A not applicable NHL non-Hodgkin’s lymphoma OR odds ratio; RR relative risk; SD standard deviation; SMR standardized mortality ratio; tTG tissue transglutaminase; yr year