In this longitudinal study of UC patients, we have demonstrated that the microbiota composition is constant over time with no alterations before, during, or after a flare of the disease or during remission, both at the time of diagnosis and in established UC. Our results further show that the microbial dissimilarity is higher between than within patients, irrespective of disease activity or time since diagnosis, supporting the notion of the long-term stability of an individual’s gut microbiota.
During the last decade several studies have reported differences in fecal microbial composition between active and inactive IBD[5, 10–12]. However, most previous studies have had a cross-sectional design comparing fecal microbiota from patients with active disease and patients in remission. Our results clearly show that the inter-individual differences in fecal microbiota composition are larger than the intra-individual differences over time, confirming previous reports on the matter[7, 8]. Thus, differences in fecal microbiota composition related to disease activity demonstrated using a cross sectional set-up most likely reflect individual differences in microbiota composition rather than disease activity.
More recently, data from long-term investigations have suggested that gut microbiota composition within an individual is relatively constant over time, although microbiota of IBD patients has been proposed to fluctuate more than those of healthy subjects[5, 7–9]. Potential instability of gut microbiota of IBD patients may be linked to inflammation and our approach was therefore to follow patients before, during and after flare of disease. Our study demonstrates a high microbial stability over time in UC, both at the time of diagnosis and with established disease, with no changes in the overall microbial composition related to disease activity. Moreover, the high stability of the microbial profile during the three years after time of diagnosis indicates that the deviation from a healthy microbiota was firmly established before or during early development of disease. Thus, by including several consecutive samples obtained before, during, and after a flare of the disease linked to thorough clinical follow-up we confirm the stability of the microbiota community both at the time of diagnosis and in established UC.
Similar to the lack of influence by disease activity, increased dose of 5ASA or addition of corticosteroids did not affect the microbial community in the present study, neither when being introduced to patients with newly diagnosed disease, nor when being intensified in patients with established disease. Comparable results have recently been reported for both CD and UC with no impact of medication on microbial stability[8, 13]. Nevertheless, microbial diversity was improved in pediatric IBD patients responding to anti-TNF therapy[14]. The inconsistent reports of potential impact of medical regimens on gut microbiota raises the question whether the different outcomes is related to the type of drug or that gut microbiota of adolescent patients are more volatile than that of adult patients. It might also be considered if the stability of the microbiota profile in adult IBD patients may be a potential explanation for fecal transplantation only partly meeting the high expectations as a therapeutic option for IBD patients[15, 16]. Furthermore, the usefulness of autologous fecal transplantation, i.e. one’s own microbiota, for treatment in IBD patients could also be disputed[17]. Concerning diets and the microbiota, UC patients on vegetarian or gluten free diets show altered gut microbiota profiles in comparison to omnivores but without any clinical benefits[18]. For diet interventions, a study employing a low fermentable oligosaccharides, disaccharides, monosaccharides and polyols (FODMAP) diet for 4 weeks in IBD patients reported higher relief of gut symptoms vs. normal diet and showed lower abundance of Bifidobacterium adolescentis, Bifidobacterium longum and Feacalibacterium prausnitzii but no alterations in phyla distribution or α− and β−diversity[19]. A 4-week diet of low fat, high fiber was evaluated in UC patients in comparison to an improved standard American diet and resulted in decreased markers of inflammation and reduced dysbiosis.[20] However, long term persistence of the altered microbiota is still unknown and considering the high stability of the microbiota such studies are warranted. Altogether this further emphasizes the need of longitudinal studies evaluating the dynamics of gut microbiota related to different treatment regimens including medical therapies, diets and fecal transplantation.
The main strength of our study is the longitudinal study design, which allows for assessment of associations between microbiota composition and disease activity within patients over time; this is particularly important considering the inter-individual differences in microbiota composition and medication use. Compared to previous studies with similar research focus[7, 8], we have included more samples per individual, for the established phase encompassing time-points both before and after the flare to evaluate possible changes preceding or following a flare, and for the newly diagnosed encompassing both short- and long-term sampling. We also used a different mode of analyzing the microbiota, as the GA-map™ technology is based on a pre-selected set of bacteria known to discriminate IBD patients from healthy subjects. The use of the GA-map™ Dysbiosis Test has limitations since it only determines defined bacterial sequences, ruling out the possibility to find new strains or to deliver full taxonomic details and α and β diversity values. On the other hand, strengths of the method include the generation of absolute values instead of relative abundance, enabling direct comparisons between samples, and the ease of data handling as compared to deep or shotgun sequencing.
A limitation of our study is that the limited number of patients, however, the standardized treatment regimens, the large number of samples per individual and the use of two different cohorts strengthen the findings of the study. Another limitation is that we have not ruled out the presence of an infectious agent, such as Clostridium difficile, as a cause for the flares in the established cohort. However, symptoms were consistent with UC exacerbation and the patients responded well to an increased dose of 5-ASA with or without corticosteroids.
In summary, we have shown that the gut microbiota in UC is highly stable irrespective of stage of the disease, disease activity or treatment escalation. This suggests that previous studies showing differences in microbiota composition between patients in remission and patients experiencing a flare reflect individual differences, rather than disease activity. It also suggests that nutritional modulation of the gut microbiota as a strategy to control the disease requires long-term commitment.