Key points
- How a disease can differ from person to person (disease heterogeneity) is thought to be one of the biggest challenges in ME/CFS research.
- The current diagnostic process for ME/CFS – and differences in practice between clinicians – may be one explanation for differences in disease characteristics between people with the disease.
- A paper published in the Journal of Clinical Medicine investigated whether heterogeneity could be seen between people with ME/CFS diagnosed at different clinics, and therefore by different clinicians.
- The findings suggest the clinicians in the study were recognising the same disease – albeit one that differs from person to person.
- Future ME/CFS research must account for disease heterogeneity both in study design and during analysis.
Background
At the 1st International Conference on Clinical and Scientific Advances in ME/CFS and Long COVID earlier this year, former ME Research UK grant award-holder Dr Nuno Sepúlveda highlighted that disease heterogeneity – how a disease can present differently from person to person – is one of the biggest challenges in ME/CFS research.
In ME/CFS, this diversity in disease presentation is thought to relate to differences in factors such as severity and type of symptoms, duration of illness, comorbidities, and disease triggers.
One potential reason ME/CFS may differ from person to person relates to the diagnostic process for the disease. As there is no validated diagnostic biomarker, ME/CFS is diagnosed based on patients’ self-reported symptoms, clinical opinion, and by excluding other conditions.
There are several criteria which can be used to diagnose ME/CFS, such as:
| Criteria | Year | Disease | Purpose |
|---|---|---|---|
| Fukuda criteria | 1994 | CFS | Research |
| International Consensus Criteria (ICC) | 2011 | ME | Research and Clinical |
| Canadian Consensus Criteria (CCC) and Revised CCC | 2003 | ME/CFS | Research and Clinical |
| Institute of Medicine criteria(IoM; now National Academy of Medicine) | 2015 | ME/CFS (recommended a name change to “systemic exertion intolerance disease”) | Clinical |
| NICE ME/CFS Guideline | 2021 | ME/CFS | Clinical |
Notably, the different criteria require varying symptoms to be present for a diagnosis of ME/CFS, ME or CFS to be made. For example, the Fukuda criteria do not require post exertional malaise (PEM) – often referred to as the main symptom of ME/CFS – for a diagnosis, while the NICE (2021), CCC, IoM and ICC criteria all do.
These differences limit the comparability of individuals diagnosed using the different criteria, and may mean that it is not appropriate to group participants diagnosed using the different methods together in research – regrettably, this is a common limitation of many ME/CFS studies.
In addition, it possible that different clinicians may take alternate approaches to the same criteria – something recognised by Professor Leonard Jason when he developed the DePaul Symptom Questionnaire (DSQ). This means that even those diagnosed with ME/CFS using the same tool – or through clinical opinion alone – may not be comparable where they have been diagnosed by different doctors.
Therefore, in a paper published in the Journal of Clinical Medicine, Dr Elizabeth Unger and colleagues investigated whether clinicians at seven clinics in the USA were diagnosing people who had significantly different symptoms with ME/CFS – including those who may not meet ME/CFS diagnostic criteria.
What did the study do?
Information was collected on 465 participants aged between 18 and 70 years who had either received a diagnosis of “ME”, “CFS” or “post-infectious fatigue” from a clinician at one of seven clinics in the USA, or were a “patient” at one of these clinics and receiving the same care as others with ME/CFS.
Although not described in detail in the current paper, a previous article by the research team stated that the eligibility of people with ME/CFS to take part in the study was determined by the clinicians who had “specialized expertise in and experience with diagnosis and management of ME/CFS” – importantly, participants were not required to fit a specific case definition for the disease.
Once participants had been recruited, standardised questionnaires were used to measure:
- General characteristics of the participants, such as age and body mass index (BMI; a measurement of weight and height).
- Symptoms relating to ME/CFS, including functional impairment, PEM, fatigue, sleep, neurocognitive/autonomic symptoms, pain, and “other symptoms” which included inflammation, gastrointestinal symptoms, and emotional or behavioural symptoms.
Responses to these questionnaires were then used to:
- Evaluate whether patient characteristics differed by clinic.
- Assess whether participants met one of three ME/CFS case definitions: Fukuda (CFS), the CCC (ME/CFS) and the IoM (ME/CFS).
- Determine whether the proportions of participants meeting each diagnostic criteria differed by clinic.
When considering differences by clinic, the research team looked at two types of significance:
- Statistical significance: where a relationship or difference observed using the data in a study is unlikely to be due to chance.
- Clinical significance: where a relationship observed in a study is unlikely to make a difference in clinical practice. Here a relationship may be statistically significant, but not clinically significant – this is often seen where a dataset is large and has the statistical power to detect very small associations or differences which would not be considered significant outside of statistical testing.
Results
Significant differences – both statistical and clinical – were identified between clinics in relation to age at diagnosis, mode of disease onset, duration of fatigue, and BMI.
Despite this, the researchers found few statistically significant and no clinically significant differences between clinics in measures relating to any of the following:
- Functional impairment,
- PEM,
- Fatigue,
- Sleep,
- Neurocognitive/autonomic symptoms,
- Pain,
- Other symptoms (including inflammation, gastrointestinal symptoms, and emotional or behavioural symptoms).
Although a lack of variation between clinics was identified, the participants within each clinic – and in the overall sample – showed a wide distribution in scores and measures, indicating overall disease heterogeneity.
When assessing whether participants met diagnostic criteria for ME/CFS, the results showed that:
- 83.4% of participants met the Fukuda criteria for CFS,
- 50.1% met the CCC criteria for ME/CFS,
- 57.6% met the IoM criteria for ME,
Overall, there were 352 participants (70.2%) who met at least two criteria, and 41 participants (9.1%) who did not meet any.
Notably, 74 participants (16.5%) met the Fukuda criteria only, which do not require PEM for a diagnosis of CFS to be made – this suggests that some participants were being given a diagnosis of “ME/CFS” even though they only met criteria for “CFS” and were not experiencing what is said to be the cardinal symptom of ME/CFS, namely PEM.
Interestingly, the findings showed that the proportions of participants meeting the different diagnostic criteria did not vary significantly between clinics.
Strengths of the study
- The authors state that the main strength of the study was that participants were identified by ME/CFS specialists.
- Participants were assessed against three ME/CFS diagnostic criteria in the study – although not all were for ME/CFS (the Fukuda criteria for CFS were used).
Limitations of the study
- It is unlikely that the participants in the study are representative of all people with ME/CFS in the USA.
- The clinics often required payment from patients rather than relying on health insurance, meaning that those who could not afford care, or sought primary care instead, were not included in the study.
- Those whose illness severity prevented travel to the clinics were not included in the study.
- The participants were mainly white, highly educated and insured.
- The participants were assessed against ME/CFS diagnostic criteria using a computer algorithm, and the results from their responses to self-reported questionnaires.
- The computer algorithm may not give the same diagnosis as a clinician or ME/CFS specialist.
- Self-reported information also has potential limitations, for example:
- Participants may interpret questions differently.
- Some participants may give answers that they think the researchers wants to hear, or that are more socially acceptable. An example of this is where someone who smokes reports that they do not because they feel that not smoking is perceived to be more socially acceptable.
Conclusions
The findings of this study indicated that the clinicians involved were generally recognising the same disease – albeit one that differs from person to person.
However, it was also shown that in these specialist clinics, 74 participants (16.5% of all those included in the study) were given a diagnosis of “ME/CFS”, when in fact they were not experiencing PEM.
There were also 41 participants (9.1%) who clinicians identified as having ME/CFS, but when assessed by the researchers in this study did not meet diagnostic criteria for the disease.
As the participants in the study are not representative of all people with ME/CFS in the USA, it would be important to consider the impact of increasing diversity on the findings – although the authors of the study state that the results would be unlikely to change.
ME Research UK notes that it is important to consider that, while the results may not change overall as suggested by the authors, an increasing population diversity would enable researchers to consider the accuracy of diagnostic procedures in subgroups of the population, such as for males and females, and for different ethnic groups.
Future ME/CFS research needs to account for heterogeneity both in study design and during analysis, and the potential impact of heterogeneity should also be considered when interpreting research findings.
