Increased risk of chronic fatigue syndrome in patients with inflammatory bowel disease: a population-based retrospective cohort study (2019) Tsai et a

Increased risk of chronic fatigue syndrome in patients with inflammatory bowel disease: a population-based retrospective cohort study

https://translational-medicine.biomedcentral.com/articles/10.1186/s12967-019-1797-3
(Open access)

Abstract

Background
Similarities in the symptoms of chronic fatigue syndrome (CFS) and inflammatory bowel disease (IBD) have been observed as follows: severe disease activity in IBD correlates with severe fatigue, major psychiatric signs, the common use of medication, and bacterial translocation. One of several hypotheses for explaining the mechanisms underlying CFS suggests a similarity to the impaired intestinal mucosa of IBD. “This study investigated the risk of incident CFS among patients with IBD”.

Methods
We conducted a population-based retrospective cohort study by using Taiwan’s National Health Insurance Research Database to evaluate the subsequent risk of CFS in patients with IBD, according to demographic characteristics and comorbidities. The exposure cohort comprised 2163 patients with new diagnoses of IBD. Each patient was randomly selected and frequency matching according to gender and age with four participants from the general population who had no history of CFS at the index date (control cohort). Cox proportional hazards regression analysis was conducted to estimate the relationship between IBD and the subsequent risk of CFS.

Results
The exposure cohort had a significantly higher overall risk of subsequent CFS than that of the control group [adjusted hazard ratio (Christophi in Inflamm Bowel Dis 18(12):2342–2356, 2012) = 2.25, 95%, confidence interval (Aaron and Buchwald in Ann Intern Med 134(9 Pt 2):868–881, 2001; Farraye et al. in Am J Gastroenterol 112:241, 2017) 1.70–2.99]. Further analysis indicated a significantly higher risk of CFS in patients who were male (HR = 3.23, 95% CI 2.12–4.91), were older than 35 years, and had IBD but without comorbidity status, e.g. Cancers, diabetes, obesity, depression, anxiety, sleep disorder, renal disease (HR = 2.50, 95% CI 1.63–3.84) after adjustment.

Conclusion
The findings from this population-based retrospective cohort study suggest that IBD, especially Crohn’s disease, is associated with an increased risk of subsequent CFS

 

CFS in this paper = Fukuda criteria, so PEM is only mentioned as one of eight possible symptoms, with only six required to make the diagnosis.

 

Before reading i assumed that they only considered ‘chronic fatigue’ as criteria, and i was actually pleasantly surprised that they included various symptoms. Still not the strongest criteria, but one needs to remember this is a retrospective cohort and while PEM is well known within our community, it may not be known at all in countries where language barriers are significant. It makes one wonder, are there links that have not been recognized before? Will it stand the test of science? Honestly, it would serve us well to compare other diseases to ours according to the common symptoms.

 

Important detail. So what they mean is, 'People with IBD are also likely to have chronic fatigue as part of their illness.'

 

Not really. Fukuda criteria are not equivalent to chronic fatigue. Many of these patients could have met more stringent criteria.

 

Indeed and in fact Dr Natelson is still using Fukuda according to his scientific commentary published this month. We need biomarkers already! Come on, researchers, you can do it!

 

In order to meet Fukuda you specifically cannot have an illness that would explain the symptom complex. IBD should be an exclusionary illness.

These patients were also not required to have the cardinal symptom of ME.

So I don't think these patients can be said to have CFS--IBD alone might be enough to cause fatigue and five other symptoms.

 

Whether or not these patients have ME or chronic fatigue - this bit was still interessting:

"Further analysis indicated a significantly higher risk of CFS in patients who were male."

 

For what its worth, I can offer my own N=1 experience having had CFS diagnosis (strict criteria) for decades and now, much more recently, the onset of what appears to be a spondyloarthritis with IBD overlap.

The fatigue from an IBD flare feels quite similar to PEM, in terms of making any sort of concentration or exertion very difficult. What is quite different is that, for me at least, it comes on very rapidly (usually alongside a bad bowel movement) and is generally shorter lived than PEM. I would say that there is more somnolence with an IBD/arthritis flare and more clear symptoms of mild fever.

Both PEM and IBD/arthritis flares will screw with my autonomic function (subjective & objectively noted), so I wouldn't be surprised if much of the overlap is linked to that. I recall SFN is found in IBD quite frequently too.

 

We have lots of biomarkers. We need validated diagnostic biomarkers, preferably with defined sensitivity and specificity. In other words, using the markers we have, we need more research into how to best use them. OMF is doing just that on a subset right now.

 

I am sorry, where have I been in the last decade? Please enlighten me.

 

A biomarker is something that is biologically different and statistically significant. It seems to be that many researchers use two standard deviations different from normal as the cut-off. By my count we currently have about 2500 biomarkers, but when I did that based on some studies, the details of which elude me as it was a while ago, I did nothing to ensure that there were no overlaps in markers. It was an simple summation of findings in cytokines, metabolomics, genes, etc. I suspect about 20% will be false positives, mostly because they were things found in patients in the study but not about ME. That is a guesstimate, and might be wrong. Of course there are probably more markers we have not found yet.

Another study just came out showing urinary abnormalities before and after exercise. Yet another study on red blood cell deformability was just published as well, by Ron Davis et. al. This one in particular might explain much of the lack of energy, fatigue, and even pain.

Other findings include CPET data, neurological data, immunological data, metabolic, genomic, urinary profiles, etcetera. I am particularly concerned about the CPET data, given that the lowest anaerobic threshold found in ME patients still able to exercise a bit was 68. In addition, the highest loss in capacity recorded after one brief bout of exercise was 44% in one day. That is a massive loss of capacity in one single day.

The issue here is that a diagnostic biomarker has to undergo much further testing than a regular biomarker, which is just otherwise just an observation of physiological difference. The data also has to be replicated in other studies. In particular we need to know sensitivity, which is the ability to detect CFS. We also need to know specificity, which relates to how likely this is in other conditions.

There is, for example, a small study from New Zealand in which they show that MS patients do not have PEM, and if I recall correctly this was another CPET study.

We now have multiple biomarker combinations that seem to reach 100% sensitivity, and more at 95%. What we do not know is how prevalent these markers are in other diseases. For example, the miRNA (?) data for ME also matches African Sleeping Sickness. Its unlikely we have that. The metabolomic profile matches sepsis. Indeed one researcher is pursuing the question as to whether or not ME is chronic sepsis.

However I suspect that using a strict ME definition, such as ICC, or possibly even in cases with observed PEM , highly sensitive biomarkers could be used clinically right now. Biomarkers plus meeting good diagnostic criteria could be enough. Clinically this would be useful, but the problem seems to be in medical politics as much as science.

So technically we only have candidate diagnostic biomarkers, or tentative ones, and at my last count there were about fourteen studies looking at different biomarkers or groups of biomarkers.

This issue of no confirmed diagnostic biomarkers is often conflated with the issue of no biomarkers, in my view.

Biomarkers confirm the reality of the disease ME and its physiological properties. They need much more work, and expensive and resource demanding work, to become diagnostic. Other kinds of biomarkers include outcome biomarkers, which can be used to determine the efficacy of a clinical treatment.

PS Some people might prefer to talk of candidate biomarkers, with diagnostic biomarkers being called just biomarkers. I personally think this confuses things too much, and is where a lot of the poor interpretations arise. These findings were significant biological markers in patients, and should be considered as such.

 

The study isn't bad, but there is a potential red flag, namely these authors are fishing in the data retrospectively for associations... Now if the authors preregistered their hypotheses and mentioned which hypotheses were tested and found to have null results, their results could be interpreted more accurately.


https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4603016/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6078132/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6282352/
https://www.ncbi.nlm.nih.gov/pubmed/24619129
https://www.ncbi.nlm.nih.gov/pubmed/24715153

and related eg Fibromyalgia:
https://www.ncbi.nlm.nih.gov/pubmed/26825913