Evidencetabellen en risk of bias tabellen

Scoglio, 2003

Type of study^[1]:

cohort

Setting and country:

Celiac clinic, Italy

Funding and conflicts of interest:

Not descripted

Inclusion criteria:

biopsy for positive serological tests for CD (AEA or

guinea pig-TGA or both) and/or a suspect CD, based on GI

symptoms.

Exclusion criteria:

n.a.

N studies =50 adults

Prevalence: -

Mean age ± SD: 30.4, range 18-69 years

Sex: 37/50 females (74%)

Other important characteristics:

-

Describe index test:

- transglutaminase antibodies (TGA)

- antiemdomysial antibodies (AEA)

Cut-off point(s):

Data not provided.

Comparator test^[2]:

n.a.

Cut-off point(s):

 n.a.

Describe reference test^[3]:

biopsy

Cut-off point(s):

Biopsy criteria for diagnosis

Time between the index test en reference test:

Yes, blood sample, later biopsy and analysis of blood.

For how many participants were no complete outcome data available?

N =0 (0%)

Reasons for incomplete outcome data described?

N.a.

Outcome measures and effect size (include 95%CI and p-value if available)⁴:

TGA:

Sensi; 100.0 (100.0–100.0)

Spec; 81.3 (62.1–100.4)

Ppv; 91.9 (83.1–100.7)

Npv; 100.0 (100.0–100.0)

Accuracy; 94.0 (87.4–100.6)

AEA:

Sensi; 97.1 (91.4–102.7)

Spec; 87.5 (71.3–103.7)

Ppv; 94.3 (86.6–102.0)

Npv; 93.3 (80.7–106.0)

Accuracy; 94.0 (87.4–100.6)

Only 50 adults

* Commercial kits are used.

Abrams, 2006

Type of study:

cohort

Setting and country:

Celiac disease center, USA

Funding and conflicts of interest:

-

Inclusion criteria:

IgA anti-tTG (human) antibody

determination performed before upper endoscopy for duodenal

biopsies.

Exclusion criteria:<16 years, selective IgA deficiency, on gluten-free diet, were taking immunosuppressants

at the time of initial evaluation, or had initial serologic testing performed at more than one laboratory

N= 122

Prevalence: 102/122 (84%)

Mean age ± SD:

44.5 (15.4)

Sex:  69/122 (57%) F

Other important characteristics:

-

Describe index test:

anti-tissue transglutaminase (tTG)

antibody test.

Cut-off point(s):

Data not provided.

Comparator test:

N.a.

Cut-off point(s):

 N.a.

Describe reference test^[4]:

biopsy

Cut-off point(s):

Biopsy criteria for diagnosis

Time between the index test en reference test:

Antibody test before biopsy

For how many participants were no complete outcome data available?

N =0 (0%)

Reasons for incomplete outcome data described?

N.a.

Outcome measures and effect size (include 95%CI and p-value if available):

TGA:

Sensi; 70.6 (60.8 – 79.2)

Spec; 65.0 (408 – 84.6)

Ppv; 91.1 (84.8 – 95.0)

Npv; 30.2 (21.8 – 40.2)

Accuracy; 69.7 (60.7 – 77.7)

High prevalence

tTG IgA not specific for celiac disease.

* Commercial kits are used.

Schyum, 2013

Type of study:

SR of 10 studies;

A: Veermersch, 2010 ; case-control

B: Sugai, 2006; case-control

C: Sugai, 2010; cohort

D: Niveloni, 2007; cohort

E: Volta, 2011; case-control

F: Kaukinen, 2007; case-control

G: Hopper, 2008; case-control

H: Dahle,2010; cohort

I: Tonutti,2009; case-control

J: Villalta, 2010; case-control

Setting and country:

See article for information per study

Funding and conflicts of interest:

None.

Inclusion criteria:

Cohort study or case-control, published between 2006-2013, serologic tests,

Exclusion criteria:

Studies including children, unclear study design, no biopsy.

N studies = 10

Prevalence: 39.1% to 44.9%, in the 3 cohort studies

Mean age ± SD:

-

Sex: % M / % F -

Other important characteristics: -

1. Describe index test:

endomysial antibodies (EMA) IgA

2. Describe index test:

EMA IgG

3. Describe index test:

Tissue transglutaminase antibodies (tTG) IgA

4. Describe index test:

tTG IgG

5. Describe index test:

deamidated gliadin peptide (DGP) IgA.

6. Describe index test:

deamidated gliadin peptide (DGP) IgG.

Cut-off point(s):

The 3 cohort studies used 20U/ml as cut-off, however, Dahle used the cut-off of 5 U/ml for the tTG IgA

assay.

No information was provided about the case-control studies.

Describe reference test^[5]:

biopsy

Cut-off point(s):

Biopsy criteria for diagnosis

Time between the index test en reference test:

Unclear

For how many participants were no complete outcome data available?

N 0 (0%)

Reasons for incomplete outcome data described?

N.a.

Outcome measures and effect size (include 95%CI and p-value if available):

* median (range)

EMA IgA (E – H):

Sensi 0.835 (0.61–0.937)

Spec 1 (0.98–1)

Ppv 0.82 (0.644–1)

Npv 0.969 (0.944–0.994)

EMA IgG: (J)

Sensi 0.758

Spec -

Ppv -

Npv -

tTG IgA (A – H):

Sensi0.93 (0.76–0.968)

Spec 0.952 (0.909–0.98)

Ppv 0.937 (0.286–0.969)

Npv 0.968 (0.963–0.996)

tTG IgG (A, J):

Sensi 0.628 (0.414–0.842)

Spec 0.988

Ppv -

Npv -

DGP IgA (A-F, I):

Sensi 0.91 (0.69–0.984)

Spec 0.969 (0.903–0.98)

Ppv 0.945 (0.892–0.989)

Npv 0.972 (0.852–0.989)

DGP IgG (A- G, I-J):

Sensi 0.882 (0.754–0.967)

Spec 0.996 (0.985–1)

Ppv 1 (0.982–1)

Npv 0.959 (0.869–0.97)

* Different kits are used; in the individual studies it is described that standard cut-offs according to the manufacturer were used.

Shomaf, 2017

Type of study:

cohort

Setting and country:

Hospital, Jordan

Funding and conflicts of interest:

This study was supported by a grant from the Deanship of Scientific Research/The University of Jordan. None conflict of interest.

Inclusion criteria:

Biopsies were performed to exclude coeliac disease and had serological test.

Exclusion criteria: -

N= 133

Prevalence: 97/133 (73%)

Mean age ± SD:

32.3 (14.7)

Sex:  97/133 (73%) F

Other important characteristics:

-

Describe index test:

anti-tissue transglutaminase (tTG)

antibody test.

Cut-off point(s):

20 IU/mL

Comparator test:

N.a.

Cut-off point(s):

 N.a.

Describe reference test^[6]:

biopsy

Cut-off point(s):

Biopsy criteria for diagnosis

Time between the index test en reference test:

Antibody test before biopsy

For how many participants were no complete outcome data available?

N =0 (0%)

Reasons for incomplete outcome data described?

N.a.

Outcome measures and effect size (include 95%CI and p-value if available):

TGA:

Sensi; 85.6 (77.0 – 91.9)

Spec; 66.7 (49.0 – 81.4)

Ppv; 87.4 (81.2 – 91.7)

Npv; 63.2 (50.0 – 74.5)

Accuracy; 80.5 (72.7 –86.8)

The differences in sensitivity and specificity can be related to the levels of the antibodies titre and the cut-off points of the different assays used.

* commercial kit was used.

Scoglio, 2003

Was a consecutive or random sample of patients enrolled?

Yes

Was a case-control design avoided?

Yes

Did the study avoid inappropriate exclusions?

Yes

Were the index test results interpreted without knowledge of the results of the reference standard?

No

If a threshold was used, was it pre-specified?

Commercial kit was used, but cut-off unknown.

Is the reference standard likely to correctly classify the target condition?

Yes

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear, first blood sample, later biopsy and analysis of blood

Was there an appropriate interval between index test(s) and reference standard?

Yes

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

Yes

Are there concerns that the included patients do not match the review question?

No

Are there concerns that the index test, its conduct, or interpretation differ from the review question?

Maybe, cut-off not known.

Are there concerns that the target condition as defined by the reference standard does not match the review question?

No

CONCLUSION:

Could the selection of patients have introduced bias?

No

RISK: Low

CONCLUSION:

Could the conduct or interpretation of the index test have introduced bias?

Yes, cut-off unknown

RISK: HIGH

CONCLUSION:

Could the reference standard, its conduct, or its interpretation have introduced bias?

Yes, it is not known if the reference standard results were interpretated without knowledge of the index test

RISK: HIGH

CONCLUSION

Could the patient flow have introduced bias?

No

RISK: LOW

Abrams, 2006

Was a consecutive or random sample of patients enrolled?

Yes

Was a case-control design avoided?

Yes

Did the study avoid inappropriate exclusions?

Yes

Were the index test results interpreted without knowledge of the results of the reference standard?

No

If a threshold was used, was it pre-specified?

Commercial kit was used, but cut-off unknown.

Is the reference standard likely to correctly classify the target condition?

Yes

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear, first blood sample, later biopsy and analysis of blood

Was there an appropriate interval between index test(s) and reference standard?

Yes

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

Yes

Are there concerns that the included patients do not match the review question?

No

Are there concerns that the index test, its conduct, or interpretation differ from the review question?

Maybe, cut-off not known.

Are there concerns that the target condition as defined by the reference standard does not match the review question?

No

CONCLUSION:

Could the selection of patients have introduced bias?

No

RISK: Low

CONCLUSION:

Could the conduct or interpretation of the index test have introduced bias?

Yes, cut-off unknown

RISK: HIGH

CONCLUSION:

Could the reference standard, its conduct, or its interpretation have introduced bias?

Yes, it is not known if the reference standard results were interpretated without knowledge of the index test

RISK: HIGH

CONCLUSION

Could the patient flow have introduced bias?

No

RISK: LOW

Schyum, 2013

Was a consecutive or random sample of patients enrolled?

Yes

Was a case-control design avoided?

No

Did the study avoid inappropriate exclusions?

Yes

Were the index test results interpreted without knowledge of the results of the reference standard?

Yes

If a threshold was used, was it pre-specified?

Only cut-off values of the cohort studies were provided, not for the case-control studies.

Is the reference standard likely to correctly classify the target condition?

Yes

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear

Was there an appropriate interval between index test(s) and reference standard?

Yes

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

No, subgroups per test.

Are there concerns that the included patients do not match the review question?

No

Are there concerns that the index test, its conduct, or interpretation differ from the review question?

Maybe, cut-off not known.

Are there concerns that the target condition as defined by the reference standard does not match the review question?

No

CONCLUSION:

Could the selection of patients have introduced bias?

Yes, some of the included studies had a case-control design.

RISK: LOW

CONCLUSION:

Could the conduct or interpretation of the index test have introduced bias?

Yes, only cut-offs for the cohort studies were provided.

RISK: HIGH

CONCLUSION:

Could the reference standard, its conduct, or its interpretation have introduced bias?

Probably, it is unknown whether outcomes of the index test were known.

RISK: UNCLEAR

CONCLUSION

Could the patient flow have introduced bias?

No

RISK: LOW

Shomaf, 2017

Was a consecutive or random sample of patients enrolled?

Yes

Was a case-control design avoided?

Yes

Did the study avoid inappropriate exclusions?

Yes

Were the index test results interpreted without knowledge of the results of the reference standard?

Yes

If a threshold was used, was it pre-specified?

yes

Is the reference standard likely to correctly classify the target condition?

Yes

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear

Was there an appropriate interval between index test(s) and reference standard?

unclear

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

No, subgroups per test.

Are there concerns that the included patients do not match the review question?

No

Are there concerns that the index test, its conduct, or interpretation differ from the review question?

No

Are there concerns that the target condition as defined by the reference standard does not match the review question?

No

CONCLUSION:

Could the selection of patients have introduced bias?

No

RISK: LOW

CONCLUSION:

Could the conduct or interpretation of the index test have introduced bias?

No

RISK: Low

CONCLUSION:

Could the reference standard, its conduct, or its interpretation have introduced bias?

Probably, it is unknown whether outcomes of the index test were known.

RISK: UNCLEAR

CONCLUSION

Could the patient flow have introduced bias?

No

RISK: LOW

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Lähdeaho, 2011

Type of study:

Prospective cohort study

Setting and country:

Hospital, Finland

Funding and conflicts of interest:

Non-commercial funding

No conflict of interest

Inclusion criteria:

Biopsy-proven celiac

disease in patients aged 18-75 years adhering to a strict gluten-free diet for at least two years and being in clinical remission as judged by an interview, clinical examination and on-site rapid celiac autoantibody testing

Exclusion criteria:

Patients with a history of regular dietary transgressions, and those suffering from a condition carrying a risk at endoscopy, or using immunosuppressive medication.

N total at baseline:

Total: 25

Important prognostic factors²:

Female; n(%)

21 (81%)

Age; median (range), years 49 (21-68)

Reason for suspicion of celiac disease at diagnosis, n(%)

Abdominal symptoms

23 (88%)

Anemia or malabsorption 16 (62%)

Extraintestinal manifestations

9 (35%)

Screening in risk groups of celiac disease

1 (4%)

Duration of gluten free diet; median (range), years 11 (2-34)

Describe intervention:

Gluten challenge with a moderate (3-5 g) of gluten daily.

Describe  control:

Gluten challenge with a  low (1-3 g) amount of gluten daily

Length of follow-up:

12 weeks

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size

Baseline vs. Gluten challenge

IgA anti-tTG , MD (95%CI)

99.00 (24.07 to 173.93)

Gluten challenge vs. after gluten challenge

IgA anti-tTG , MD (95%CI)

34.20 (8.30 to 60.10)

Histology

Low dose

Vh:Cd ratio between baseline and after gluten challenge, MD (95% CI)

-1.09 (-1.64 to -0.54

Moderate dose

Vh:Cd ratio between baseline and after gluten challenge, MD (95% CI)

-1.30 ( -1.93 to -0.67

Symptoms, n(%)

15/21 (71%) experienced mild symptoms after gluten challenge

CSI score, MD (95%CI)

-0.04 ( -5.76 to 5.68)

GSRS score, MD (95%CI)

10.45 (95%CI 0.47 to 20.2

The authors conclude that Low amounts of gluten can cause significant mucosal deterioration in the majority of the

patients.

Leffler, 2013

Type of study:

Non-randomized trial

Setting and country:

Hospital, USA

Funding and conflicts of interest:

Funded through the National

Institute of Health (non-commercial)

No conflicts of interest

Inclusion criteria:

Subjects with biopsy proven coeliac disease in remission were enrolled. Remission was defined as: (1) Immunoglobulin (IgA)-

tTG levels within normal limits (<10 ELISA units),

(2) on a strictly gluten-free

diet for at least 12 months with a Celiac Dietary Adherence Score of <18 and (3) absence of both symptoms typical of

coeliac disease and symptoms which prompted initial diagnosis

in the participant.

Exclusion criteria:

History of food

intolerances other than to gluten and lactose, any food allergy, a history of severe acute symptomatic reactions to sporadic

gluten ingestion, a history of chronic active gastrointestinal

disease other than coeliac disease, or exposure to corticosteroids

or other immunosuppressive agents within the prior 3 months

N total at baseline:

Total: 20

High gluten: 10

Low gluten: 10

Important prognostic factors²:

Female; n(%)

19 (95.0%)

Age, years; mean(SD)

43.3 (13.7)

Age of coeliac diagnosis, years; mean (SD)

37.5 (14.4)

Age of coeliac diagnosis, years; mean (SD)

37.5 (14.4)

Months on GFD; Mean (SD)

65.0 (38.9)

Symptoms at diagnosis; n(%)

Gastrointestinal

13 (65.0%)

Extra-intestinal

13 (65.0%)

Anaemia

9 (45.0%)

Asymptomatic

4 (20.0%)

Describe intervention:

14-day gluten challenge of 7.5 g

Describe  control:

14-day gluten challenge of 3 g

Length of follow-up:

28 days

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size

Baseline vs. Gluten challenge

IgA anti-tTG , MD (95%CI)

2.86 (-4.4 to 10.1)

Gluten challenge vs. after gluten challenge

IgA anti-tTG , MD (95%CI)

29.76 (-0.9 to 60.4)

Histology

Vh:Cd ratio between baseline and after gluten challenge, MD (95% CI)

-1.09 (-1.64 to -0.54

The authors conclude that 14day gluten challenge at >3 g of gluten/day induces

histological and serological changes in the majority of

adults with coeliac disease. These data permit accurate design of clinical trials and indicate that many individuals

will meet coeliac diagnostic criteria after a 2-week gluten challenge

Leonard (2021)

Type of study:

Randomized double-blind controlled trial

Setting and country:

Hospital, USA

Funding and conflicts of interest:

Funded through Oxford PharmaGenesis, Oxford, UK and  by Takeda Pharmaceutical Company

Authors are linked to Takeda Pharmaceuticals, Glutenostics LLC, Anokion and HealthMode Inc, Biomedal S.L.,

Cour Pharmaceuticals, and Glutenostics LLC, Alba Therapeutics, Mead Johnson Nutrition, NextCure and

Viome SAB, Cour

Pharmaceuticals, ImmunogenX, Innovate, invicro, Merck Research Laboratories

Inclusion criteria:

HLA-DQ2.5 and/or HLA-DQ8–positive adults with biopsyproven

CeD in clinical and histologic remission on a GFD for 12 months were enrolled. Patients selected for inclusion

had records of diagnostic pathology reviewed at screening, and were required to have no ongoing signs or symptoms

that, in the investigator’s opinion, were due to CeD.

Exclusion criteria:

History of food intolerances/

allergy other than to gluten and lactose; severe

acute reactions to sporadic gluten ingestion; chronic active gastrointestinal disease other than CeD; or exposure to corticosteroids or other immunosuppressive agents within the prior 3 months

N total at baseline:

Total: 14

High gluten: 7

Low gluten: 7

Important prognostic factors²:

Female, n (%)

11 (78.6)

Age, y, mean (SD)

43.7 (18.8)

Years since diagnosis, mean (SD)

5.2 (4.7)

d2.1 (1.6–2.6)

IEL, median (IQR)

26.1 (20.6–29.8)

CDSD, overall mean (SD) 1.7 (2.1)

Describe intervention:

14-day gluten challenge of 10 g

Describe  control:

14-day gluten challenge of 3 g

Length of follow-up:

15 days

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size

Baseline vs. 15 days of Gluten challenge

IgA anti-tTG , MD (95%CI)

0.90 (-2.61 to 4.41)

Gluten challenge vs. after 42 days of gluten challenge

IgA anti-tTG , MD (95%CI)

7.70 (4.76 to 10.64)

Histology

Low dose

Vh:Cd ratio between baseline and after gluten challenge, MD (95% CI)

-0.10 ( -0.61 to 0.41).

Moderate dose

Vh:Cd ratio between baseline and after gluten challenge, MD (95% CI)

-1.90 (-1.21 to -2.59).

The authors conclude that selected CD biomarkers are sensitive and

responsive to gluten exposure, providing the potential for less-invasive, lower-dose, and shorter-duration gluten ingestion.

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Leonard, 2021

Patients were block-randomized by site in a 1:1 ratio

concurrently to either 3 g gluten/d or 10 g gluten/d for 14 days, using unique randomization sequence numbers and the

relevant gluten dose.

Likely

Care providers, the investigator, and outcomes assessor were all blinded after assignment to interventions.

Unlikely

Gluten doses were supplied in premeasured packets blinded to dose patients. 

Unlikely

Care providers, the investigator, and outcomes assessor were all blinded after assignment to interventions.

Unlikely

Care providers, the investigator, and outcomes assessor were all blinded after assignment to interventions.

Unlikely

The trial was registered at ClinicalTrials.gov under identifier NCT03409796

Unlikely

Minimal loss of participants after randomization in both study arms.

Unlikely

No mention of ITT-analysis but minimal loss of participants after randomisation.

Study reference

(first author, year of publication)

Bias due to a non-representative or ill-defined sample of patients?¹

(unlikely/likely/unclear)

Bias due to  insufficiently long, or   incomplete follow-up, or differences in follow-up between treatment groups?²

(unlikely/likely/unclear)

Bias due to ill-defined or inadequately measured outcome ?³

(unlikely/likely/unclear)

Bias due to inadequate adjustment for all  important prognostic factors?⁴

(unlikely/likely/unclear)

Lähdeaho, 2011

unlikely

unclear

Unlikely

not applicable

Leffler, 2013

unlikely

unclear

Unlikely

not applicable

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Szakacs, 2017

SR and meta-analysis of  observational studies

Literature search between until 30 December 2016

Study design:

Included 61 observational studies

Setting and Country:

See Table 1 (Szakacs, 2017)

Source of funding and conflicts of interest:

See table S5 (Szakacs 2017) for the Quality assessment of each study included

Inclusion criteria SR:

experimental (randomized or non-randomized, controlled or uncontrolled

clinical trials) and observational studies (cohort, cross-sectional and case-control studies)

carried out either in a prospective or a retrospective manner without respect to the

primary objectives of the studies. The PICO items of children-to-adults comparison were as follows: (P) celiac patients previously

subjected to GFD with control biopsy, (I-C) adults and children, and (O) mucosal recovery

ratios.

Exclusion criteria SR:

etters, comments, conference abstracts, editorials,

and reviews.

Important patient characteristics at baseline:

See table S3 for individual study characterisitcs included in the analyses (Szakacs, 2017)

Describe intervention

Adherence to a gluten free diet during follow-up to assess mucosal recovery

Describe  control

Diagnoses of CD by biopsy at baseline before starting on a gluten-free diet

End-point of follow-up

Individual studies varied from 11 to 118.4 months of follow-up. See table S3 (Szakacs 2017)

For how many participants were no complete outcome data available?

See table S5 (Szakacs 2017) for the Quality assessment of each study included

Outcome

Complete mucosal recovery ratio of all studies (95%CI):

0.36 (0.28 to 0.44)

Complete mucosal recovery ratio in adults (95%CI):

0.65 (0.44 to 0.85)

Complete mucosal recovery ratio in children (95%CI):

0.24 (0.15 to 0.33)

Complete mucosal recovery ratio after 12 months follow-up (95%CI):

0.38 (0.28 to 0.48)

Facultative:

The authors conclude that Several celiac patients fail to achieve complete mucosal recovery even if a strict dietary regimen is followed.

Despite the prescribed GFD, complete mucosal recovery ratios were low (36%), even in patients being on a strict diet (47%).

Nikniaz, 2019

SR and meta-analysis of  observational studies

Literature search between 1 January 1995 and 15 January 2017

A: Vuoristo, 1993

B: Smecoule, 1997

C: Ciacci, 2002

D: Murray, 2004

E: Capristo, 2005

F: Lanzini, 2006

G: Dickey, 2006

H: Brar, 2006

I: Capristo, 2009

J: Lewis, 2009

K: Passananti, 2011

L: Zanini, 2013

M:  Casella, 2012

N: Kabbani, 2012

O: De March, 2013

P: Pekki, 2014

Q: Riezzo, 2014

R: Tortora, 2015

S: Newnham, 2015

T: Barone, 2015

U: Talaie, 2015

Study design:

Cohort (A-U)

Setting and Country:

A: Hospital, Finland

B: Hospital, Argentina

C: Hospital, USA

D: Hospital, USA

E: Hospital, Italy

F: Hospital, Italy

G: Hospital, UK

H: Hospital, USA

I: Hospital, Italy

J: Hospital, UK

K: Hospital, Italy

L: Hospital, Italy

M: Hospital, Italy

N: Hospital, USA

O: Hospital, USA

P: Hospital, Finland

Q: Hospital, Italy

R: Hospital, Italy

S: Hospital, Australia

T: Hospital, Italy

U: Hospital, Iran

Source of funding and conflicts of interest:

Not reported outcomes of the Newcastle-Ottawa assessment are also not reported

Inclusion criteria SR:

Controlled studies with case-control or cohort study designs assessing

Body mass index, total cholesterol, LDL-C, HDL-C, and triglyceride in adults with CD

Exclusion criteria SR:

conference proceedings or

book chapters and unpublished work.

Important patient characteristics at baseline:

n; duration follow-up

A: 6

B: 25

C:  390

D:  21

E:  18

F: 44

G:  371

H: 132

I: 26

J: 100

K: 47

L: 715

M: 987

N: 679

O: 20

P: 26

Q: 20

R: 98

S: 98

T: 3

U: 5

Describe intervention

Follow-up of the same diagnosed CD after afherence to a gluten-free diet during follow-up.

Describe  control

Newly diagnosed CD in patients starting gluten free diet

End-point of follow-up

A: 8 months

B: 37 months

C:  2 years

D: 6 months

E: 1 year

F: 1 year

G: 2 years

H: 20,5 months

I: 1 year

J: 1 year

K: 5 years

L:  1-5 years

M: 30 months

N: 39.5 months

O: 6-8 months

P:  1 year

Q: 1 year

R: 1 year

S: 1-5 year

T: 24.3

U: 6 months

For how many participants were no complete outcome data available?

(intervention/control)

A-U: not reported

Outcome

BMI; mean difference (95%CI)

0.265 kg/m² (0.177 to 0.352)

TC; mean difference (95%CI)

0.172 mmol/L (-0.068 to 0.412)

HDL-C; mean difference (95%CI)

0.367 mmol/L (0.181 to 0.554)

LDL-C; mean difference (95%CI)

0.271 mmol/L (-0.540 to 0.107)

Triglyceride; mean difference (95%CI)

0.367 mmol/L (0.181 to 0.554)

Facultative:

The authors conclude that a gluten-free diet has a positive effect on BMI and HDL-C.

Note:

Outcomes of the Newcastle-Ottawa assessment are not reported

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Mora, 2001

Type of study:

Longitudinal cohort study

Setting and country:

Hospital, Italy

Funding and conflicts of interest:

Not reported

Inclusion criteria:

White children and adolescents who were newly diagnosed

as having celiac disease were eligible for the study. The diagnosis was made by following the recommendations of the European

Society of Paediatric Gastroenterology and Nutrition

Exclusion criteria:

Subjects who had been immobilized or hospitalized in the preceding 6mo, were taking medications known to affect bone metabolism, had any chronic illness other than type 1 diabetes, or had one or

more nontraumatic fractures were excluded

N total at baseline:

Total: 19

Important prognostic factors²:

Age, years (SD):

14.2 (2.6)

Weight, kg (SD):

42.7 (10.7)

Height, cm (SD):

151.5 (12.0)

Describe intervention:

Gluten-free diet

Describe  control:

Gluten-containing diet

Length of follow-up:

Short-term follow-up: 1.1 year on a gluten free diet.

Long-term follow-up: 4.3 years on a gluten free diet.

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size

1.1 year follow-up:

Spinal BMD; mean difference (95%CI)

0.06 g/cm2 (-0.08 to 0.03)

Total body BMD; mean difference (95%CI)

0.04 g/cm2 (-0.05 to -0.02)

4.3 year follow-up:

Spinal BMD; mean difference (95%CI)

0.19 g/cm2 (-0.22 to -0.16)

Total body BMD; mean difference (95%CI)

-0.13 g/cm2 (-0.15 to -0.12)

The authors conclude that bone mineral density improves significantly with a gluten-free diet

Kalayci, 2001

Type of study:

Longitudinal cohort study

Setting and country:

Hospital, Turkey

Funding and conflicts of interest:

Not reported

Inclusion criteria:

Children were eligible for the study if they were diagnosed for CD by intestinal biopsy. The diagnosis was made by following the recommendations of the European

Society of Paediatric Gastroenterology and Nutrition.

Exclusion criteria:

Children with other diseases known to affect BMD were excluded

N total at baseline:

Total: 32

The patients with CD were evaluated under 2 subgroups, 16 patients with recent diagnosis (group 1a) and 16 patients who follow their diet strictly (group 1b).

Important prognostic factors²:

Group  1a

BMD z-score (SD)

-1.55 + 1.36

Group  1B

BMD z-score (SD)

-0.67 + 0.85

Describe intervention:

Patients already on a strict gluten-free diet

Describe  control:

Patients with recent diagnosis starting a gluten-free diet

Length of follow-up:

1 year.

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size

Group 1a

BMD z-score ; mean difference (95%CI)

-0.88 (-1.67 to -0.09)

Group 1b

BMD z-score ; mean difference (95%CI)

-0.30 (-0.94 to  0.34)

The authors conclude that BMD is almost invariably low in newly

diagnosed celiac patients in childhood, but adherence to a strigt gluten-free diet promoted a significant increase in BMD.

Choudhary, 2017

Type of study:

Cross sectional cohort study

Setting and country:

Hospital, India

Funding and conflicts of interest:

Not reported

Inclusion criteria:

Patients with CD diagnosed according

to the criteria of the European Society for Pediatric

Gastroenterology (ESPGHAN 2012) ] by increased serum

levels of tissue transglutaminase (tTG) followed by intestinal

biopsy (demonstration of crypt hyperplasia and villous atrophy

in the duodenal mucosa) in each patient

Exclusion criteria:

Children with presence of other diseases known to affect the BMD like renal disease, steroid therapy

etc.

N total at baseline:

Total: 72

Group A (newly diagnosed): 36

Group B (already on a gluten-free diet): 36

Important prognostic factors²:

Group A:

Age, years (SD):

7.25 (2.78)

BMD z-score (SD)

-2.02 + 1.01

Group B:

Age, years (SD):

7.83 (2.66)

BMD z-score (SD)

-0.65 + 0.92

Describe intervention:

On a gluten-free diet for over a year

Describe  control:

Recently diagnosed and starting on a gluten free diet

Length of follow-up:

No follow-up.

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size

BMD z-score ; mean difference (95%CI)

-1.37 (-1.82 to -0.92)

The authors conclude that children with CD are at risk for reduced BMD.

Strict gluten-free diet significantly improves BMD. Early diagnosis and treatment of celiac disease during childhood may protect CD patients from osteoporosis.

Szakacs, 2017

Yes

Yes

Yes

Yes

Not applicable

Yes

Yes

Yes

Unclear1

Lasa, 2014

Yes

Yes

Yes

Yes

Not applicable

Yes

Yes

Yes

Unclear1

Nikniaz, 2019

Yes

Yes

Yes

Yes

Not applicable

Yes

Yes

Yes

Unclear1

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Studies in children

Chellan, 2019

Type of study:

Prospective follow-up study

Setting and country:

Tertiary care center in North India from June 2014 to November 2015

Funding and conflicts of interest: Declaration of conflict of interest: None  

Inclusion criteria:

• Children aged 2-12 years

(CD according to ESPGHAN criteria)

Exclusion criteria:

• Patients with chronic illnesses (chronic kidney disease, cirrhosis, and congestive cardiac failure)
• Parents who denied consent for endoscopy in their child

N total at baseline: 60

Important prognostic factors²:

Mean ± SD age:

6.03 ± 0.42  years

Sex:

32% girls

Initiation of GFD

Before initiation of GFD

Length of follow-up:

6 months

Loss-to-follow-up:
16

N =3 because endoscopic and histopathological features were not consistent with CD

N= 5 because of non-compliance to GFD

N=8 lost to follow up

QoL as assessed by the pediatric symptom checklist (PSC) questionnaire

At baseline:

11.5 (2–35)

At 6 months follow-up:

2.5 (0–34)

P < 0.001

Nordyke, 2013

Type of study:

prospective nested case-referent study

Setting and country:

5 regions in Sweden. The screening took place in their schools in collaboration

with school health systems and regional pediatric

departments

Funding and conflicts of interest: This study was funded by: the Swedish Research Council (grants 521-2004-7093 and 521-2007-2953), the Swedish Research Council for Environment, Agricultural Sciences, and

Spatial Planning (grants 222-2004-1918 and 222-2007-1394), the Swedish

Council for Working Life and Social Research (grant 2005–0802), and by the European Union supported project (FP6-2005-FOOD-4B-36383-PREVENTCD),

and was undertaken within the Umeå Centre of Global Health Research at

Umeå University. The authors declare that they have no competing interests.

Inclusion criteria:

adolescents who had participated in a CD screening study and were positive for CD.

N total at baseline: 144

Important prognostic factors²:

Mean ± SD age:

Around 12 years

Sex:

51% girls

Initiation of GFD

Before initiation of GFD

Length of follow-up:

1 year

Loss-to-follow-up:
41 (28.5%)

QoL as assessed by the EQ-ED5

Mobility

Baseline: 3/103 (2.9%)

Follow-up: 1/103 (1.0%)

Self care

Baseline: 0/103

Follow-up: 0/103

Activity

Baseline: 3/103 (2.9%)

Follow-up: 2/103 (1.9%)

Pain

Baseline: 21/203 (20.4%)

Follow-up: 13/103 (12.6%)

Anxiety/depression

Baseline: 13/103 (12.6%)

Follow-up : 15/103 (14.6%)

Studies in adults

Choung, 2018

Type of study:

longitudinal study

Setting and country: Tertiary care center; Mayo Clinic, Rochester, Minnesota, USA

Funding and conflicts of interest:  Supported by a grant from the National Institutes of Health (R01DK057892), by a grant from the National Institute on Aging of the National Institutes of Health (Rochester Epidemiology Project Award Number R01AG034676), and by Mayo Foundation for Medical Education and Research.

Inclusion criteria:

• biopsy proven CD
• responders of a validated self-report Bowel Disease Questionnaire (BDQ) and the 36-Item Short Form Survey Instrument (SF-36) at the time of diagnosis.

To classify the classical and nonclassical phenotypes of CD, clinical evaluations and the self-report bowel disease questionnaire were used; Classical phenotype was defined by patients who had watery diarrhea, steatorrhea, or weight loss at the time of diagnosis. Those who did not have these classical symptoms were categorized as having nonclassical CD.

N total at baseline: 130

Important prognostic factors²:

Mean ± SD age:

Total: 45.0 ± 14.9

Classical: 44.7 ± 15.5

Nonclassical 45.5 ± 14.4

Sex:

67% female

Initiation of GFD

Before initiation of GFD

Length of follow-up:

mean duration of GFD was 6.7 ± 3.7 years

QoL as assessed by the

Short Form-36

Classical group

PCS

Baseline: 43.6 ± 10.8

Follow-up: 48.0 ± 9.8

MCS

Baseline: 44.3 ± 11.2

Follow-up: 48.2 ± 12.1

Nonclassical group

PCS

Baseline: 46.5 ± 11.1

Follow-up: 48.0 ± 10.1

MCS

Baseline: 49.8 ± 9.4

Follow-up: 53.3 ± 8.5

Deepak, 2018

Type of study:

prospective follow-up study

Setting and country: gastroenterology

clinic of a tertiary-care referral hospital at the Postgraduate

Institute of Medical Education and Research,

Chandigarh, India

Funding and conflicts of interest: No specific funding has been received for this project. No author has any conflict of interest with the work reported or the conclusions, implications, or opinions stated.

Inclusion criteria:

• age > 12 years
• having history suggestive of malabsorption (chronic diarrhea, anemia, and short stature) with a positive IgA tissue transglutaminase antibody (IgA antitTG)

Exclusion criteria:

• chronic illnesses (chronic kidney disease, cirrhosis, and congestive cardiac failure
• pregnant/lactating
• unwilling to go for endoscopy

N total at baseline: 60

Important prognostic factors²:

Mean ± SD age:

28.85  ±  12.43 year

Sex:

44% female

Initiation of GFD

Before initiation of GFD

Length of follow-up:

6 months

Qol as assessed by the SF-12 questionnaire

PCS

Baseline: 37.2 ± 11.1

Follow-up: 50.3 ± 9.9

MCS

Baseline: 42.9 ± 8.4

Follow-up: 50.2 ± 9.0

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Burger, 2017

[individual study characteristics deduced from Burger, 2017]

PS., study characteristics and results are extracted from the SR (unless stated otherwise)

Systematic review and meta-analysis

Literature search January

1960 to 30 September 2015

A: Johnston, 2004

B: Kurppa, 2010

C: Mustalahti, 2002

D: Nachman, 2009

E: Nachman, 2010

F: Tontini, 2010

G: Ukkola, 2011

H: Vilppula, 2011

Setting and Country:

A: United Kingdom

B: Finland

C: Finland

D: Argentina

E: Argentina

F: Italy

G: Finland

H: Finland

Source of funding and conflicts of interest:

None

Inclusion criteria SR:

1. HRQoL assessed using validated questionnaires

2. participants were above 16 years of

Age

3. coeliac disease was confirmed by biopsy

4. comparing HRQoL in patients and non-coeliac controls, before and after

gluten-free diet initiation or between subgroups of treated patients

Exclusion criteria SR:

1. adapted versions of validated questionnaires were used

2. an intervention other than a gluten-free diet was evaluated

3. only an abstract was provided

8 studies were included, comparing HRQoL in patients before and after gluten-free diet initiation

Important patient characteristics at baseline:

Number of patients:

A: 31

B: 73

C: 40

D: 35

E: 53

F: 33

G: 698

H: 35

Age

A: mean 48.5 years

B: range 16 – 70 years

C: median 49 years

D: median 38 years

E: mean 38 years

F: range 22 – 65 years

G: median 52 years

H: median 61 years

HRQoL assessed by

A: SF-36

B: PGWB

C: PGWB

D: SF-36

E: SF-36

F: SF-36

G: PGWB

H: PGWB

Before gluten-free diet initiation

After gluten-free diet initiation

End-point of follow-up:

At least 1 year

Study reference

(first author, year of publication)

Bias due to a non-representative or ill-defined sample of patients?¹

(unlikely/likely/unclear)

Bias due to  insufficiently long, or   incomplete follow-up, or differences in follow-up between treatment groups?²

(unlikely/likely/unclear)

Bias due to ill-defined or inadequately measured outcome ?³

(unlikely/likely/unclear)

Bias due to inadequate adjustment for all  important prognostic factors?⁴

(unlikely/likely/unclear)

Studies in children

Chellan, 2019

Unlikely

Likely

44/60 patients had data at follow-up

Unlikely

Unclear

Nordyke, 2013

Unlikely

Likely

103/144 patients had data at follow-up

Unlikely

Unclear

Studies in adults

Choung, 2018

Unlikely

Likely

122/130 patients had data at follow-up

Unlikely

Unclear

Deepak, 2018

Unlikely

Unlikely

1/60 patient was excluded as his endoscopic and histopathological features were not consistent with CD

Unlikely

Unclear

Burger, 2017

Yes

Yes

No, not for excluded studies

Yes

Not applicable

Yes

Yes

Yes

Unclear

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Studies in children

Deora, 2017

Type of study:

Cohort study – medical records were examined

Setting and country: Celiac Outpatient Clinic at the Children’s Hospital, Health Sciences Center, Winnipeg, Manitoba,

Canada

Funding and conflicts of interest: The authors report no conflicts of interest.

Inclusion criteria:

Children with CD seen in the Celiac Outpatient Clinic

N total at baseline:

140

Important prognostic factors²:

Mean age ± SD:

7.8 ± 4.01 years

Sex:

62% girls

Modified Marsh Classification:

Number of children

I: 5

II: 12

IIIa: 27

IIIb: 36

IIIc: 60

Serology

Mean anti-TtG IgA antibodies at diagnosis:

111.3 ± 73.5

Mean anti-TtG IgA antibodies at 6 months on GFD:

44.3 ±  54.2

Mean anti-TtG IgA antibodies at 18 months on GFD:

17 ± 32.7

18 months of GFD

Before GFD initiation

Length of follow-up:

18 months

Forcielli, 2020

Type of study:

Prospective cohort study

Setting and country:

Paediatrics, University of Bologna, Italy

Funding and conflicts of interest:

Financial disclosure: none to be declared. The corresponding author a
rms that this manuscript is an honest, accurate and transparent

account of the study being reported.

Inclusion criteria:

Age between 1 and 18 years, diagnosis of CD, and the availability of a set of measurements taken at two points in time, one before diagnosis and another 12 months after initiation of GFD.

N total at baseline:

79

Important prognostic factors²:

Mean age ± SD:

7.9 ± 3.8 years

Sex:

66% girls

1 year of GFD

Before initiation of GFD

Length of follow-up:

1 year

Wessels, 2016

Type of study:

Prospective cohort study

Setting and country: Leiden University Medical Center, the Netherlands

Funding and conflicts of interest: The authors declare no conflicts of interest.

Inclusion criteria:

Children with CD, who had medical checks between January 2009 and January 2015 at the Leiden University Medical Center.

N total at baseline:

182, of which 119 were diagnosed during the study period

Important prognostic factors²:

Mean age ± SD at diagnosis:

6.3 ± 4.3 years

Sex:

65% girls

Marsh Classification (%):

Marsh 2:     4

Marsh 3a: 25

Marsh 3b: 49

Marsh 3c: 21

CD specific antibodies at diagnosis  (%):

EMA and/or TG2A positive: 97

EMA and TG2A negative: 1

EMA and TG2A unknown: 2

Initiation of GFD

Before initiation of GFD

Length of follow-up:

Mean ± SD follow-up: 3.1 ± SD 3.1 years

On average, 17% of results per year were

missing because of incomplete blood investigations

Studies in adults

Capristo, 2000

Type of study:

Prospective cohort study

Setting and country: Department of Internal Medicine at

the Catholic University in Rome, Italy

Funding and conflicts of interest: Not reported

Inclusion criteria:

adult patients with the classic form of celiac disease

Exclusion criteria:

• secondary causes of intestinal atrophy
• endocrine disorders
• consumption of drugs able to influence data collection evidence of intestinal bacterial overgrowth (assessed by a lactulose breath-hydrogen test)
• impairment of glucose and lipid metabolism, hepatic or renal disease
• fever
• pregnancy, smoking >10 cigarettes daily
• intensive physical activity

N total at baseline:

43

Important prognostic factors²:

Mean ± SD age:

29.9 ± 7.6 years

Sex:

59% female

Initiation of GFD

Before initiation of GFD

Length of follow-up:

12 months

Four patients did not complete the study protocol, 3 because

of low compliance with the GFD diet and 1 because of relocation

to another country.

Dahele, 2001

Type of study:

Prospective cohort study

Setting and country: Gastrointestinal Unit, Department of Medical Sciences, University of Edinburgh,

Western General Hospital, Edinburgh, Scotland

Funding and conflicts of interest: This study was supported by a grant from The Chief Scientist’s

Office, Scottish Executive, Edinburgh.

Inclusion criteria:

untreated patients  with biopsy proven celiac disease diagnosed between September 1997 to

February 1999.

N total at baseline:

39

Important prognostic factors²:

Median (range) age:

48 (22 – 77) years

Sex:

82% female

Initiation of GFD

Before initiation of GFD

Length of follow-up:

Median (range) follow-up: 4 (2 – 17) months

Mitchell, 2002

Type of study:

Setting and country:

Funding and conflicts of interest:

Inclusion criteria:

patients with newly diagnosed celiac disease confirmed by villous atrophy on duodenal histology were recruited.

N total at baseline:

51

Basic characteristics (e.g. age and sex) of the population were not described.

Initiation of GFD

Before initiation of GFD

Length of follow-up:

12 – 18 months

Five patients were excluded (one died shortly after diagnosis, four were lost to follow-up).

Shepherd, 2013

Type of study:

Prospective cohort study

Setting and country:

a single dietetic service, from gastroenterologists and general practitioners, or the Coeliac Clinic at Box Hill Hospital, Melbourne, Victoria, Australia

Funding and conflicts of interest: SJS was supported by a Dora Lush Scholarship from the NHMRC of Australia and received a Grant in Aid from the Australian and New Zealand Coeliac Research Fund. All authors critically reviewed the manuscript and approved the final version submitted for publication.

Inclusion criteria:

newly-diagnosed, untreated patients who were consecutively referred

to a single dietetic service, from gastroenterologists and

general practitioners, or the Coeliac Clinic at Box Hill Hospital

N total at baseline:

57

Important prognostic factors²:

Median (range) age:

44 (range 18–71) years

Sex:

71% female

Initiation of GFD

Before initiation of GFD

Length of follow-up:

12 months

Only 50 patients had complete data and were included in the analysis.

Study reference

(first author, year of publication)

Bias due to a non-representative or ill-defined sample of patients?¹

(unlikely/likely/unclear)

Bias due to  insufficiently long, or   incomplete follow-up, or differences in follow-up between treatment groups?²

(unlikely/likely/unclear)

Bias due to ill-defined or inadequately measured outcome ?³

(unlikely/likely/unclear)

Bias due to inadequate adjustment for all  important prognostic factors?⁴

(unlikely/likely/unclear)

Studies in children

Deora, 2017

Unlikely

Likely

Sample differs for each outcomes and follow-up point

Unlikely

Unclear

Forcielli, 2020

Unlikely

Unlikely

Likely

Unclear

Wessels, 2016

Unlikely

Likely

Sample differs for each outcomes and follow-up point

Unlikely

Unclear

Studies in adults

Capristo, 2000

Unlikely

Likely

39/43 patients had data at follow-up, not all outcomes were reported

Unlikely

Unclear

Dahele, 2001

Unlikely

Likely

Not all outcomes were reported

Unlikely

Unclear

Mitchell, 2002

Unlikely

Likely

46/51 patients had data at follow-up

Unlikely

Unclear

Shepherd, 2013

Unlikely

Likely

50/57  patients had data at follow-up, not all outcomes were reported

Likely

Unclear

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Lionetti, 2017

Type of study:

Systematic review of RCTs

A: Biesierskie, 2011

B: Carroccio, 2012

C: Biesierskie, 2013

D: Brotveit, 2013

E: Shahbazkhani, 2015

F: Di Sabatino, 2015

G: Zanini, 2015
H: Capannolo, 2015

I: Elli, 2016

J: Picarelli, 2016

K: Rosinach, 2016

Study design:

RCT (A-K)

Setting and country:

Not reported for individual studies.

Funding and conflicts of interest:

No commercial funding or conflicts of interest.

Inclusion criteria SR:

clinical trials performing a gluten re-challenge with or

without a placebo control in patients with a suspected

diagnosis of NCGS were considered for inclusion in this meta-analysis.

Exclusion criteria SR:

Studies on patients

suffering from CD or other gluten-related disorders (glutenataxia,

autism, etc.) or WA were excluded

Important patient characteristics at baseline:

n; gluten:

A: 34; 16g

B: 920; 20g

C: 37; 16g

D: 22; 4.4g

E: 72; 52g

F: 59; 4.4g

G: 35;7.9g

H: 364; not stated

I: 97; 5.6g

J: 26; 10g

K: 18; 16.2g

Groups comparable at baseline?

Yes

Describe intervention:

double-blind

provocation

with gluten

Describe  control:

double-blind

provocation

with placebo

End-point of follow-up

A: 12 weeks

B: 6 weeks

C: 7 weeks

D: 1 week

E: 12 weeks

F: 12 months

G: 7 months

H: 7 months

I: 4 weeks

J: 1 week

K: 18 months

For how many participants were no complete outcome data available?

(intervention/control)

A-K: not described in SR

Outcome measures and effect size

Relapsing after gluten-challenge; RR (95%CI)

0.4 (−0.15 to −0.9)

Relapsing after gluten-challenge according to salerno criteria; RR (95%CI)

2.8 (1.5 to 5.5)

The authors conclude that the prevalence of NCGS after gluten re-challenge is low, and the percentage of relapse after a gluten or a placebo challenge is similar

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Barone, 2020

Type of study:

Randomized Double-Blind

Placebo-Controlled Crossover Trial

Setting and country:

Home, Italy

Funding and conflicts of interest:

No specific funding received

No conflict of interest.

Inclusion criteria:

diagnosis of IBS based on Rome IV criteria, absence of alarm symptoms, and no use of medications for the treatment of bowel habit abnormalities in the

previous 3 months

Exclusion criteria:

administration of a gluten-free diet in the previous six months, presence

of coeliac disease or wheat allergy, chronic intestinal inflammatory diseases, psychiatric disorders,

major abdominal surgery (in particular intestinal resections), diabetes mellitus, previous anaphylactic

episodes and pregnancy

N total at baseline:

Total: 40

Important prognostic factors²:

male/female
 

10/30

Age+SD (years)

40.5+12.5

BMI

24.7+4.1

Abdominal pain (%)

100

Abdominal distention (%) 50

Constipation/diarrhea (%) 38

Fatigue (%)

29

Describe intervention:

consumption of only cereals or pseudo-cereals

that are gluten-free, such as rice, buckwheat, corn, millet and quinoa (all naturally gluten-free foods with low FODMAP content

Describe  control:

Placebo (Rice starch) exposure

Length of follow-up:

9 weeks

Incomplete outcome data:

14 were lost to follow-up due to limited improvement in symptoms. 2 discontinued the study

Outcome measures and effect size

VAS score, MD (95%CI)

5.20 (4.71 to 5.69)

The authors conclude that FODMAP intolerance could hide the response to a challenge test with gluten for the identification of NCGS in IBS patients. A low FODMAP-GFD followed by gluten/placebo challenge is able to identify patients

with NCGS better

Barmeyer, 2017

Type of study:

Prospective cohort study

Setting and country:

Home, Germany

Funding and conflicts of interest:

Commercial funding from Dr. Schär AG

All authors report conflicts of interests referring to grants for their work.

Inclusion criteria:

Consecutive non-constipated IBS patients with weekly symptoms,

above 18 years old, male and non-pregnant female, who were able and willing to follow a 4-month GFD, were included. IBS patients had to fulfill the Rome III criteria for IBS-D or IBS-M based on a diagnostic questionnaire published by the Rome Foundation

Exclusion criteria:

other medical conditions presenting with similar symptoms as IBS were excluded on the basis of

medical history, physical examination, blood tests

N total at baseline:

Total: 35

Important prognostic factors²:

Age, mean (range)

48.9 (27–66)

Female, n(%)

26 (74%)

Symptom severity (mean ± SD)

IBS-QOL: 50.7±18.9

IBS-SSS: 290.7±74.8

EQ-5D VAS (mm): 49.7 ± 27.2

Describe intervention:

Gluten-free diet for 4 months

Describe  control:

Baseline prior to starting a gluten-free diet.

Length of follow-up:

4 months

Incomplete outcome data:

None reported

Outcome measures and effect size

Responders vs Non-responders

Baseline IBS-SSS score ,MD (95% CI)

-5.00 (-40.18 to 30.18)

IBS-SSS score after Gluten-free diet,MD (95% CI)

-60.00 (-93.51 to -29.49)

Baseline IBS-SSS score ,MD (95% CI)

2.00 (-13.60 to 17.60)

IBS-SSS score after Gluten-free diet,MD (95% CI)

5.00 (-11.58 to 21.58)

Baseline EQ-5D VAS score ,MD (95% CI)

-20.00 (-41.07 to 1.07)

EQ-5D VAS score after Gluten-free diet,MD (95% CI)

10.00 (-9.22 to 29.22)

The authors conclude that Using strict criteria as recommended for IBS

studies, about one third of patients with IBS-D or IBS-M are wheat sensitive.

Dale 2018

Type of study:

Randomized Double-Blind

Placebo-Controlled Crossover Trial

Setting and country:

Home, Norway

Funding and conflicts of interest:

No funding reported

No conflicts of interest

Inclusion criteria:

patients aged over 18 who experienced relief of symptoms while eating a gluten-free diet and who completed gastroscopy with duodenal biopsies and blood tests within the last year. CD was excluded by duodenal biopsies, according to Marsh-Oberhuber classification.

Exclusion criteria:

patients diagnosed

with CD, WA, inflammatory bowel disease, IBS and lactose intolerance,

ongoing infection, use of immune suppressive drugs and pregnant

or lactating women

N total at baseline:

Total: 20

Important prognostic factors²:

male/female
 

14/6

Mean age (range), years 36.4 (21-62)

Mean BMI (range)

25.1 (19.5-40.8)

Severity of symptoms, mean (SD)

IBS-SSS sum 170 (104.1)

IBS-SQ sum 10.7 (11.9)

Describe intervention:

four periods (of 4 days) of double-blind

provocation

with gluten

Describe  control:

four periods (of 4 days) of double-blind

provocation

with placebo

Length of follow-up:

26 days weeks

Incomplete outcome data:

None reported

Outcome measures and effect size

Placebo vs Exposure

IBS-SSS score ,MD (95% CI)

44.40 (-18.68 to 107.48)

IBS-SQ score, MD (95% CI)

-0.40 (-15.92 to 15.12)

The authors conclude that the majority of patients

with suspected NCGS are not able to identify when challenged with gluten in a

double-blind

placebo-controlled

food challenge, indicating that gluten is not the

cause of their symptoms.

Lionetti, 2017

Yes

Yes

Yes

Yes

Not applicable

Yes

Yes

Yes

Unclear1

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Barone, 2020

The low FODMAP-GFD responders were randomized based on a computer-generated

randomization list to take gluten or placebo for 7 days

Likely

No concealment during allocation.

Unlikely

Sachets containing gluten had the same shape, dimension, indication and appearance as those containing placebo and were marked with a serial number. To ensure blindness of the investigators and patients, only the manufacturers were in a position to associate each sachet number with its content.

Unlikely

Sachets containing gluten had the same shape, dimension, indication and appearance as those containing placebo and were marked with a serial number. To ensure blindness of the investigators and patients, only the manufacturers were in a position to associate each sachet number with its content.

Unclear

No mention of blinding of outcome assessors

Unlikely

The trial was registered at ClinicalTrials.gov under identifier NCT04017585

Likely

16/42 (38%) were lost to follow-up.

Likely

No mention of ITT-analysis and 38% loss to follow-up.

Barmeyer, 2017

Not described

Unclear

Not described

Unclear

Not described

Unclear

Not described

Unclear

Not described

Unclear

Not described

Unclear

Not described

Unclear

Not described

Dale, 2018

Not described

Unlikely

The patients randomly received two packages with gluten and two packages with placebo, in an order blinded for both patients and

the researcher.

Unlikely

The patients randomly received two packages with gluten and two packages with placebo, in an order blinded for both patients and

the researcher.

Unlikely

The patients randomly received two packages with gluten and two packages with placebo, in an order blinded for both patients and

the researcher.

Unclear

Not described

Unclear

Not described

Unlikely

No loss to follow-up mentioned

Unclear

Not described

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Francavilla, 2018

Type of study:

Randomized Double-Blind

Placebo-Controlled Crossover Trial

Setting and country:

Home, Italy

Funding and conflicts of interest:

No specific funding received

No conflict of interest.

Inclusion criteria:

Pediatric patients with chronic functional gastrointestinal symptoms associated with gluten ingestion

Exclusion criteria:

Patients without self-reported gluten

related symptoms

N total at baseline:

Total: 28

Important prognostic factors²:

Male; n (%)

11 (39%)

Age+SD (years)

11.4+4.3

BMI+SD (centiles)

53+30.7

Describe intervention:

Exposure to gluten for 14 days containing 80% protein; the non-protein part was mainly made of starch (14%), fiber (2%), fat (1.5%), and ash

(0.75%)

Describe  control:

Exposure to a placebo for 14 days

Length of follow-up:

9 weeks

Incomplete outcome data:

3 due to lack of improvement

Outcome measures and effect size

Placebo vs Exposure – All participants

Global VAS score, MD (95% CI)

0.40 (95% CI -1.04 to 1.84)

IBS-SS score, MD (95% CI)

-2.00 (-59.69 to 55.69)

Placebo vs Exposure – NCGS diagnosed participants

Global VAS score, MD (95% CI)

4.30 (2.85 to 5.75)

IBS-SS score, MD (95% CI)

72.50 (95% CI -13.92 to 158.92)

The authors conclude that NCGS may be overestimated in children

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Francavilla, 2018

children from both arms started the

wash-out period of one week and subsequently started the final

2-week period on placebo or gluten sachets, depending on a computer-

generated randomization list.

Unclear

No mention of concealment of allocation

Unclear

No mention of blinding of outcome patients

Unclear

No mention of blinding of care providers

Unclear

No mention of blinding of outcome assessors

Unlikely

The trial was registered at ClinicalTrials.gov under identifier NCT04017585

Unlikely

3 persons were lost to follow-up.

Likely

No mention of ITT-analysis