A in the Journal of Clinical Medicine suggests that the agency's Fukuda empirical criteria for diagnosing Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) identified the highest proportion of patients with ME/CFS (83%) versus only 58% for the commonly employed . Readers could conclude that using the Fukuda empirical criteria is the best method for identifying patients with ME/CFS. However, the CDC continues to promote a case definition that does not require of ME/CFS, including post-exertional malaise, cognitive impairment, and unrefreshing sleep.
I'd like to call into question the continued use of these Fukuda empirical criteria. I believe they are too broad and may be leading researchers to categorize people as having ME/CFS who in reality just have CFS-related symptoms. This has consequences for determining accurate prevalence rates and for the continued study of the illness.
For context, in 1994, the CDC's Fukuda criteria were developed and used in research by investigators. In 2005, epidemiologist William Reeves, MD, MSc, and colleagues operationalized these criteria in what has been called the "Fukuda empirical criteria," for use not just in research but for actually diagnosing patients. Since then, the CDC has used these Fukuda empirical criteria in its ME/CFS work.
Now let's take a closer look at the new CDC paper. The researchers focused on only well-identified patients with ME/CFS from specialty clinics or practices who already had exclusionary conditions removed. In other words, if investigators use broad criteria, and utilize it on a well-defined group of patients with ME/CFS, it will capture most of those patients with ME/CFS. However, using the same wide criteria with a more heterogenous sample will result in the diagnosis of many who do not have ME/CFS -- and this involves major problems with specificity.
As an example, had the investigators in their study used the , which only requires 6 months of chronic fatigue for a diagnosis, it would have identified an even higher percentage of patients with ME/CFS, but when these criteria are applied to the general population, it would also have identified many who do not have this diagnosis. No credible investigators are willing to endorse the Oxford criteria for identifying patients with ME/CFS. Yet, I believe the Fukuda empirical criteria are also too broad.
The Fukuda empirical criteria use a specific set of assessment tools to measure fatigue, symptoms, and disability. When using these empirical criteria in Georgia, found ME/CFS prevalence rates were 6 to 10 times higher than previous . One explanation for this significant increase in the CDC's ME/CFS prevalence is that the studies using the empirical criteria included many more people who have other underlying (non-CFS) causes of their fatigue and related symptoms. Therefore, dramatic prevalence changes occurred when the Fukuda empirical criteria were used.
In another study, which I co-authored in 2009, we found that these Fukuda empirical criteria selected some individuals with solely affective disorders, like major depressive disorder. Findings indicated that 38% of those with a sole diagnosis of a major depressive disorder were misclassified as meeting the Fukuda empirical case definition.
Within the , the authors reviewed a study by that used the Fukuda empirical criteria to examine the role of childhood trauma in ME/CFS. The IOM suggested that the use of these Fukuda empirical criteria resulted in a biased sample in Heim's study, with an overrepresentation of individuals with depression and posttraumatic stress disorder. The IOM report concluded that the high proportion of patients with serious psychiatric problems likely explains the Heim study's finding of a link between meeting the Fukuda empirical criteria and having experienced adversity in childhood.
In today's society, many individuals are exhausted from multiple job and family responsibilities and can experience cognitive and fatigue problems when not getting enough sleep. Some people experiencing fatigue for these reasons are unwell, and if the evaluators are not extremely careful, these individuals could be incorrectly classified as meeting the Fukuda empirical criteria. In other words, a variety of other conditions can lead people to experience CFS-related symptoms, including depression, anxiety, over-exertion, medications, poor sleep hygiene, unhealthy weight and poor diet, deconditioning, and inactivity.
The recent CDC-funded study leaves open the likelihood that some researchers will continue to use the Fukuda empirical criteria in their ME/CFS studies even though it does not require the core symptoms, and it appears likely that CDC will use some of the methods and approaches developed for the empirical criteria to identify cases.
If there are ambiguities with the case definition, samples of patients will differ on fundamental aspects of an illness. This results in difficulty replicating findings across different labs, estimating the prevalence of the illness (as is evident in the changing rates of ME/CFS prevalence studies), identifying biomarkers with consistency, and determining which treatments help patients (as evident with with the PACE trial).
Ideally, researchers should determine a unified, and more appropriate, criteria so that we can study this illness accurately and to the greatest benefit of those suffering.
is a professor of psychology and director of the Center for Community Research at DePaul University. He recently co-edited the book,. He serves as ME/CFS expert for ILLInet RECOVER.