Economic Costs of Gluten Free Diet

A recent study (AR Lee et al. Nutrients; 2019, 11, 399). Open access: Persistent Economic Burden of the Gluten Free Diet) quantifies the additional costs of a gluten free diet (GFD) in the U.S. Thanks to Kipp Ellsworth for this reference.

The authors conducted a “market basket” study to establish the cost of a GFD. “A market basket is a group of products that are purchased by consumers …for this study, the market basket was food that would necessitate a GF substitute, including staple foods, snack foods, and commonly used ready-made or convenience meals.”

Key findings:

  • GF products were more expensive, overall the increase was 183%.  This is an improvement from a 2006 study which found the increase overall at 240% (adjusted for inflation).
  • Mass-market products were 139%  more expensive than wheat-based versions


  • Cost is identified as a frequent reason for nonadherence with diet, cited by 33% in one study
  • Overall, the burden of GFD is more frequently related to the restrictive nature of the diet which leads to a negative impact on quality of life. According to the authors, in one study (Am J Gastroenterol 2014; 109: 1304-11), treatment burden for celiac was ranked higher than for diabetes hypertension, and congestive heart failure

My take: This study shows the significant economic burden of a GFD.  In Italy, the  “government offers celiac patients vouchers to buy gluten-free food — up to 140 euros per month.” (NPR: Italy, Land of Pizza and Pasta)

Related blog posts:



More Sensitive Detection of Gluten Free Diet Adherence

A recent cross-sectional study (K Gerasimidis et al. JPGN 2018; 67: 356-60) examined the use of fecal gluten immunogenic peptide (GIP) to assess for adherence with gluten free diet (GFD) in biopsy-proven celiac disease (CD).

GIP reflects recent gluten consumption.  There is a commercially-available kit available (Ivydal GIP Testing) –though I am uncertain about how its reliability compares to the GIP measured in this study.

In the study, the authors note that GIP positivity can occur with as little as 100 mg of gluten/day ingestion.  GIP is a 33-mer peptide from α2-gliadin that is stable against breakdown by gastric, pancreatic, and intestinal brush border enzymes.

Key findings of this study:

  • GIP was detectable in 16% of patients with previous CD diagnosis (N=67)
  • GIP was detectable in 95% of newly-diagnosed CD patients (n=19) and was detectable in 27% at 1 year afterwards.
  • When compared with traditional indicators of GFD adherence (eg. TTG levels, Biagi score, clinical assessment), 4 out of 5 children with detectable GIP were missed

My take: Fecal GIP for celiac disease adherence has similar potential as a biomarker as calprotectin has for IBD.  A normal GIP appears to be much more sensitive at detecting gluten ingestion.

Related blog posts:

Celiac Disease: How Long Is a Gluten Free Diet Needed? (30 Year Data)

A recent study (L Norsa et al. JPGN 2018; 67: 361-6) examines data from 197 patients with celiac disease (CD) (out of a cohort of 337) who had a diagnosis established before 1985. The authors examined three groups: lifelong strict GFD (n=133), discontinued GFD (n=29), and no GFD (22).  A total of 63 had follow-up endoscopy data available, with 29 in lifelong GFD, 20 in discontinued GFD, and 14 in no GFD.

Key findings:

  • In those with followup endoscopy, in those with lifelong GFD 27 of 29 (93%) had no atrophy (Marsh 0-1-2) on histology, in those with discontinued GFD 12 of 20 (60%) had no atrophy on histology, and in those with no GFD 8 of 14 (57%) had no atrophy on histology.
  • Thus, among the group with long-term poor adherence to gluten-free diet, almost two-thirds showed no recurrence of villous atrophy on duodenal biopsies.
  • In the entire cohort of 197, there were no apparent differences in autoimmune diseases between those receiving lifelong GFD (26%) compared to the other two groups, 17% and 23% respectively.


  • retrospective design.
  • initial diagnosis was more than 30 years ago & there are significant differences in the diagnostic approach currently
  • sample size

My take: This study indicates that some individuals who have been diagnosed with celiac disease may be OK with ongoing gluten consumption. Those who maintained a GFD were much more likely  to have no villous atrophy on duodenal biopsies.

Related blog posts:


Reliability of High Serology in Asymptomatic Celiac Disease

Another study (SP Paul et al. JPGN 2018; 66: 641-44) has shown that high anti-TTG IgA levels are reliable in establishing the diagnosis of celiac disease in asymptomatic children from high-risk groups. In this study with prospectively-collected data from 2007-2017, 84 of 157 children had anti-TTG titers >10x ULN.  75 of these 84 were from high-risk groups, mainly type 1 diabetes (36), and first degree relatives (24)

Key finding:

  • All 75 with high titers from high-risk groups had histologic evidence of celiac disease.

Related blog posts:

Related studyR Mandile et al. JPGN 2018; 66: 654-56. This prospective study showed that 19 of 35 (54%) patients with potential celiac disease had a complete clinical response on a gluten-free diet to symptoms like abdominal pain and diarrhea.  Thus, in many patients with potential celiac disease, a gluten-free diet will not be effective.

Related blog posts:


Briefly Noted: Celiac Serology Normalization, Inflammatory Markers in Crohn’s Disease, Nutrition in Neurologically-Impaired

  • DM Isaac et al. JPGN 2017; 65: 195-99.

This retrospective study of 487 pediatric patients shows that it takes a long time to normalize celiac serology/anti-tissue transglutaminase antibody (TTG). The median time was 407 days for the 80.5% of patients that normalized their serology in the study time frame.  The time was 364 days for those who were considered adherent to a gluten-free diet.  Patients with type 1 diabetes were less likely to normalize their TTG levels. Faster normalization occurred in those with lower titers at baseline.

Related blog posts:

  • A Alper et al. JPGN 2017; 65: e25-e27

In this chart review, among 135 children, normal ESR and CRP were observed in 28% of children with Crohn disease and 42% of children with ulcerative colitis.

Related blog post: Do you really need both a ESR and CRP?

  • C Romano et al. JPGN 2017; 65: 242-64

This guideline paper details 31 recommendations (some with multiple parts) for the evaluation and management of children with neurologic impairment.  The recommendations include detailed evaluations including knee heights, skinfold thickness measures, DXA scan, routine micronutrient bloodwork, along with a low threshold for oropharyngeal dysphagia assessment.  The paper has recommendations for evaluations of reflux, constipation, and dental problems.  The authors suggest “considering use of enteral feeding if total oral feeding time exceeds 3 hours per day.”

Related blog post: Surgery for reflux works best for those who need it the least

Celiac Hepatopathies 2016

A recent study (GJ Lee et al. JPGN 2016; 63: 340-3) adds a little bit more information regarding hypertransaminasemia in newly diagnosed celiac disease.  Some previous information was summarized in a previous blog: Celiac Hepatopathies (2013)

In this retrospective, single center study, 185 children had transaminases obtained at the time of celiac diagnosis (185/388 = 47.7%).

Key findings:

  • Among this group, 28 (15.1%) had elevated transaminases, with an average of ALT 2.52 x ULN and AST 1.87 x ULN.
  • Patients with elevated liver transaminases tended to be younger (mean 6.3 yrs compared with 11.0 without elevation). Among those who had followup blood testing available, 15/21 (71.4%) normalized their values over an average of 210 days.
  • For the 6 who had persistent elevation of transaminases, 3 were suspected to have poor adherence, 1 was thought to have a fatty liver, 1 had only mild elevation, and 1 remained unexplained.

My take: This study indicates that elevated transaminases are common in children with celiac, particularly younger children.  As with other studies, the majority resolve on a gluten-free diet.  As there is a recognized association with autoimmune hepatitis, in those with elevated ALT, followup after institution of a gluten free diet seems prudent.

Iceberg Lake, Glacier Natl Park

Iceberg Lake, Glacier Natl Park

Microscopic, Lymphocytic and Collagenous Colitis

Microscopic Colitis (MC) is a rare pediatric problems and occurs when chronic diarrhea occurs in the presence of a normal-appearing endoscopic exam but with abnormal histology.  In adult populations, microscopic colitis is seen more frequently and can be confused with irritable bowel syndrome.  The two subtypes:

  • Lymphocytic Colitis (LC):  >20 intraepithelial lymphocytes/100 colonocytes
  • Collagenous Colitis (CC): thickened subeptihelial collagen band in addition to changes seen with LC

In a recent study (JPGN 2013; 57: 557-61), 27 MC cases were identified from a pathology database between 1995-2011.  5 were excluded due to an enteric infection.  Among the 22 other cases, 19 had LC and 3 had CC.  Association with celiac disease was evident in 4 patients and many had preceding drug exposures.

Treatment included steroids, melamine, an bismuth.

Additional references:

  • -JPGN 2011; 53: 579. lymphocytic colitis case report
  • -Clincal Gastro & Hep 2011; 9: 13.  Celaic with persistent symptoms: consider poor adherence**, SBBO*, pancreatic insufficiency*, refractory celiac (rare), PLE, giardia, malignancy, lactose intolerance, functional d/o*, microscopic colitis, Crohn’s*, NSAIDs
  • -Gut 2009; 58: 68-72. Collagenous colitis: Budesonide at 6mg/day maintained remission in ~25%.
  • Gastro 2008; 135: 1510.  Budesonide effective for collagenous colitis; n=48, 9mg/day.
  • -Gastro 2011; 140: 1155. Review of microscopic colitis/collagenous colitis.
  • -Am J Gastroenterol 2010; 105: 859-865.  n=466 & 451 controls.  Microscopic colitis present in 1.5% of IBS patients.  IBS pts with lower incidence of adenomas (7.7.% vs 26%).  9% had diverticulosis (lower).
  • -Clin Gastro & hep 2009; 7: 1210. 4.3% of pts w microscopic colitis had celiac. 44/1009.

Gluten avoidance -quite common

In a pediatric Boston cohort of 579 patients presenting for celiac evaluation but no previous diagnosis, 7.4% had previous gluten avoidance (J Pediatr 2012; 161: 471-5). In this cohort, the mean age at presentation was 8.7 years.

Independent predictors of gluten avoidance and the odds ratio (OR) included the following:

  • Irritability OR 3.2
  • Family history of celiac OR 2.2
  • Diarrhea OR 2.5
  • Pervasive developmental disorder OR 5.3

From my perspective, many families and sometimes referring physicians are convinced that their child has celiac disease before subspecialty evaluation. While a gluten-free diet may reduce some gastrointestinal symptoms even in the absence of celiac disease, it is probably helpful for families to complete diagnostic testing and to obtain dietary counseling prior to implementing a gluten-free diet.

Previous related blog entries: