Altered Duodenal Microbiota Composition in Celiac Disease Patients Suffering From Persistent Symptoms on a Long-term Gluten-free Diet

Pirjo Wacklin PhD; Pilvi Laurikka; Katri Lindfors PhD; Pekka Collin MD; Teea Salmi MD; Marja-Leena Lähdeaho MD; Päivi Saavalainen PhD; Markku Mäki MD; Jaana Mättö PhD; Kalle Kurppa MD; Katri Kaukinen MD


Am J Gastroenterol. 2014;109(12):1933-1941. 

In This Article


Aberrations in intestinal microbiota have been associated with a variety of gastrointestinal diseases such as IBS,[24,25] inflammatory bowel disease,[26] and celiac disease.[5–7] These associations demonstrate the importance of intestinal homeostasis for well-being and health. In this study, we assessed the duodenal microbiota of the treated celiac disease patients and found that patients with persistent symptoms had altered, Proteobacteria-rich intestinal microbiota composition and lower microbial richness compared with patients whose symptoms disappeared on GFD. All patients, including those with persistent symptoms, had been committed to a strict GFD and showed restored small bowel mucosa and negative celiac disease autoantibodies, demonstrating good histological and serological response to the treatment with no signs of inadvertent gluten consumption. Normal villous morphology also rules out refractory celiac disease as a cause of the symptoms, as this condition is defined by persistent mucosal atrophy despite strict GFD.[27] It must also be emphasized that all patients of this study were voluntarily participating into a health survey, and did not visit a clinic for their symptoms. This indicates that many treated celiac disease patients are forced to accept persistent symptoms as an inevitable part of their daily life. Our findings suggest that altered duodenal microbiota composition and reduced microbial richness are associated with the persistent gastrointestinal symptoms occurring in some treated celiac disease patients. Previous results have shown that persistent gastrointestinal symptoms are common among adult celiac disease patients, even while on a strict GFD.[4] Furthermore, we have shown that the health-related quality of life is lower in celiac disease patients with persistent symptoms.[2] Understanding the causes behind the persistent symptoms would provide a basis for the development of applications to alleviate the symptoms, subsequently also improving the quality of life of these patients.

Intestinal microbiota composition in healthy adults is relatively stable and can tolerate normal stress in the intestine caused by, e.g., daily changes in diet. However, after strong perturbation such as antibiotics therapy,[28] microbiota may undergo an incomplete recovery, stabilizing into an altered state. This reformed but stable and possibly dysbiotic state of intestinal microbiota has been suggested to contribute to chronic diarrhea or inflammation.[29] We propose that untreated celiac disease may disrupt a stable intestinal microbiota community that, in some patients, could then reform in a dysbiotic state, subsequently contributing to the persistent gastrointestinal symptoms. In line with our hypothesis, treated celiac disease patients have been shown to differ in their intestinal microbiota from healthy subjects in pediatric studies,[5–7] indicating that GFD may not completely restore microbiota in patients with major dysbiosis. Furthermore, in this study, duodenal microbiota in the Persistent symptom group showed reduced microbial richness and Proteobacteria-rich microbiota, both of which have been associated with intestinal dysbiosis, for example, in inflammatory bowel disease.[30–32] In healthy subjects, Proteobacteria, a phylum containing many pathobionts and pathogens, usually comprise a minor fraction of small bowel microbiota.[33,34] In the treated patients in this study with persistent symptoms, the Proteobacteria-dominating microbiota and the Bacteroidetes- and Firmicutes-poor microbiota actually resemble the microbiota of untreated celiac patients with gastrointestinal symptoms,[8] even if GFD seems to diminish the dysbiosis. The reduced richness and enrichment of Proteobacteria supports the hypothesis that microbiota in patients with persistent symptoms on GFD has not fully recovered, although the causality of this phenomenon in relation to the symptoms remains to be shown. Interestingly, we have recently shown that delayed diagnosis is associated with the occurrence of persistent symptoms in celiac disease.[2] Similarly, a long duration of gastroenteritis is a major risk factor for persistent intestinal symptoms in IBS. It has been suggested that prolonged initial illness causes more serious mucosal injury and inflammation, resulting in persistent changes to immunological cells and intestinal microbiota.[35,36] Accordingly, it is possible that long-time untreated celiac disease also enhances intestinal dysbiosis, thus predisposing to persistent symptoms even on GFD. Another interesting issue possibly affecting intestinal microbiota composition is the so-called host genetic gardening.[37] Indeed, based on our earlier results that showed association of FUT2 with intestinal microbiota,[38,39] the effect of host genes would have been interesting to investigate here also. However, the number of subjects was too small for reliable analyses[39] and this remains an issue for further studies.

Small intestinal bacterial overgrowth (SIBO) is one of the factors suggested as a cause of persistent symptoms in treated celiac disease patients.[40,41] SIBO is defined as ≥1 × 103 colony-forming units per ml of proximal jejunal aspiration[42] and characterized by an increase in colonic bacteria in the small bowel in general, rather than overgrowth of single bacterial strain.[43,44] Thus, SIBO could also be considered as a specific dysbiotic state of the small bowel. We did not perform diagnostics targeted at SIBO, such as culturing of jejunal aspirates or hydrogen breath test, and this may be considered as a drawback.[41] On the other hand, the interpretation of SIBO tests is very challenging and there is no consensus on the best diagnostic method.[40,45] The participants had no medical conditions predisposing to SIBO such as previous abdominal surgery, diabetes, or neurological disorder.[45] The subjects with persistent symptoms were colonized with microbiota typical for the small bowel, that is Streptococcus and Veillonella spp.,[9,46] and the seven genera significantly differing between the groups (Table 2) did not include the colonic bacteria commonly detected in SIBO, e.g., Escherichia, Enterococcus, Klebsiella, and Proteus spp..[40] It is also noteworthy that in previous studies exploring the role of SIBO in nonresponsive celiac disease, mostly patients with villous atrophy had been investigated, whereas all subjects here had normal mucosal morphology on GFD.[41] Therefore, it is unlikely that SIBO plays a major role in causing persistent symptoms in our study.

The celiac disease patients with persistent symptoms on GFD had previous symptom-based diagnoses of other mild gastrointestinal diseases, such as gastroesophageal reflux and lactose intolerance, more often than patients with no symptoms on GFD. Nevertheless, there were no differences between the study groups in the small bowel mucosal morphology or degree of inflammation. Moreover, the separate microbiota analysis, when repeated without patients with additional gastrointestinal diseases, gave in most part similar differences in microbiota composition and richness. A diagnosis of IBS could also be considered in subjects with persistent symptoms here. It can be difficult to distinguish symptoms of IBS from celiac disease as there are no pathognomonic signs or definite test for IBS. However, the patients attending this study were considering themselves healthy and only one patient had previous diagnosis for IBS and had fulfilled the Rome III criteria. We consider it unlikely that a significant part of the persistent symptoms would be explained by undiagnosed IBS. Another clinical explanation for the persistent symptoms might be disorders in gastrointestinal motility. Although this was not systematically investigated, none of the participants had signs of severe motility disorder or predisposing conditions such as diabetes. In theory, temporary acid suppression with PPI may cause false negative results in H. pylori testing, but we did not observe any differences in the use of PPI medication between the groups. Furthermore, the relationship between uncomplicated H. pylori infection and recurrent abdominal symptoms is controversial.[47] Some of the patients had thyroidal disease that, in theory, could cause intestinal symptoms, but the prevalence of the condition did not differ between the groups and all of these patients received treatment for the disease. Thus, it is unlikely that the concomitant gastrointestinal maladies or other disorders would explain the detected differences in symptoms between the study groups.

The median age was significantly lower in the treated patients with persistent symptoms. Although ageng has been suggested to alter intestinal microbiota,[48,49] the microbiota is relatively stable throughout adulthood. A recent study showed that intestinal microbiota was rather similar in young adults and the aged, and in fact the major changes of microbiota in adulthood do not occur until after the age of 75–80 years.[50] The difference in the median ages in our study (54 vs. 63 years) was caused mostly by the presence of two young patients in the Persistent symptoms group. The microbiota richness of these subjects varied from low (17 genera) to one of the highest (36 genera) among patients with persistent symptoms. In addition, the two young patients did not differ from the other patients in multidimensional scaling clustering (Figure 2). Thus, the age difference between the study groups is unlikely to have had a major effect on the findings. Long-term dietary habits and medications may also have an influence on intestinal microbiota.[51] In this study, all patients were on a strict GFD, and there was no difference between the groups in the consumption of oats, starch, or fiber. In addition, the use of PPI and nonsteroid anti-inflammatory drugs was comparable, and none of the patients had medications affecting bowel function. Nevertheless, we had no long-term information on the history of antibiotic use, and their effect on the results cannot be fully excluded. According to our study design, only subjects with no symptoms and most symptomatic were investigated, and thus it is possible that the results would have been different if patients with subtle symptoms had been included. Furthermore, to minimize confounding factors, only patients with gastrointestinal symptoms were included, as we have previously shown that microbial composition varies between celiac disease phenotypes.[8] It is possible that symptoms fluctuate by time, but the present symptomatic patients had a history of long-lasting complaints. It is also true that this study was not able to show causality or distinguish the effects of different bacteria to the persistent symptoms, and further studies are needed in this area.

Our hypothesis that dysbiotic microbiota is associated with persistent symptoms in treated celiac disease patients raises an intriguing question on their treatment. One possibility could be the intestinal microbiota modulation by probiotics. Although probiotic intervention studies have reported symptom alleviation in IBS, inflammatory bowel disease, and celiac disease, it still remains to be shown whether probiotics modify dysbiotic microbiota toward a stable and disease-free state.[52] Additional possibility for the modification of intestinal microbiota is fecal transplantation that has been successful in treating recurrent Clostridium difficile infection.[53] The pioneering studies on the use of fecal transplantation as treatment for a variety of gastroenterological diseases, including Crohn's disease and IBS, have been encouraging.[54]

In conclusion, our results showed that celiac disease patients suffering from persistent symptoms on GFD had altered Proteobacteria-rich duodenal microbiota and reduced richness of bacteria, indicating intestinal dysbiosis. We propose that this altered microbiota is associated with persistent symptoms in celiac patients with strict GFD and small bowel mucosal recovery. Further studies are warranted to confirm our results and possibly to find applications for alleviating the symptoms of this specific patient subgroup by intestinal microbiota modulation.