By Peter C. Rowe, M.D.
Professor of Pediatrics
Johns Hopkins Children’s Center
Note: Included in this post are slides and their corresponding text from Dr. Peter Rowe‘s presentation regarding his work on pediatric CFS from the June 13th meeting of the federal CFS Advisory Committee. Click here to see Dr. Rowe‘s latest research, funded under the Association’s Research Institute Without Walls. Click on a slide to magnify it. You can dowload slides here: Slides
I have titled my talk “Gains and Gaps” because I want to emphasize some of the gains in understanding pediatric CFS that have been accomplished through careful clinical research in the field. As usually happens in science, those gains also help clarify where there are gaps in our knowledge.
In the 20 minutes I’ve been given, I want to begin with an overview that summarizes concepts that are generally accepted in the field, and then move on to discuss symptoms, examination findings, the impact of CFS, drawing from a prospective cohort study we have been conducting in the Pediatric CFS clinic at Johns Hopkins. At the end I want to emphasize the sobering fact that while we have the potential to improve outcomes, we have persistent gaps in the provision of service to adolescents with CFS that warrant urgent attention.
Let’s start with some comments that are not very controversial. By definition, CFS affects previously active individuals. Both Clinic-based and population-based studies confirm that pediatric CFS is much more common after age 10. It can be seen at a younger age, but it is particularly important for clinicians to hunt for other causes of symptoms in young children who meet the CFS diagnostic criteria. As in adults, females are affected more than males. The female to male ratio in various studies ranges from 2:1 to 5:1, for reasons that are as yet unclear.
CFS affects somewhere between 100-300 of every 100,000 adolescents. There are methodological reasons to believe that these prevalence estimates are incomplete. (I have listed some of the pediatric studies on the prevalence of CFS, and some of the methodological limitations in the supplemental slides in your handout).
The emerging consensus is that there are many ways that adolescents develop CFS, and many factors that can contribute to perpetuation of symptoms. For those of us who are clinicians, this heterogeneity necessitates individualized treatments rather than formulaic algorithms; for those of us who are researchers, this heterogeneity creates a lot of background noise, and requires large sample sizes in treatment studies, as well as careful sub-grouping to identify valid findings in descriptive studies.
We have known for some time that CFS can appear in approximately 10% following EBV infection, but EBV is not the only infectious precipitant of symptoms, and there is much we do not understand about other infectious triggers. In addition, a gradual onset of symptoms occurs, in some studies at rates of up to 68%.
A number of physiological conditions can converge to create CFS symptoms, and several groups are beginning to quantify these co-morbid problems, notably Esther Crawley and her colleagues in Britain.
With that as a brief overview, what are the main symptoms in pediatric CFS?
All the individuals in the study had fatigue for 6 or more months, but this slide shows the relative frequency of the 8 other symptoms in the Fukuda definition, 4 or which are required for the diagnosis of CFS. The most common were unrefreshing sleep in 84%, post-exertional worsening of malaise for greater than 24 hours in 80%, memory and concentration problems in 79%, headaches in 78%, muscle pain in 59%, and joint pain in 48%. Note that sore throat and tender glands were the least common symptoms reported, in accordance with clinical experience in this country as well. But, if we focus just on the Fukuda criteria, we can miss a great deal that is important in pediatric CFS.
Here, for example, are the other symptoms identified on the Symptom Checklist 90 survey in the Johns Hopkins CFS cohort study we are conducting: the SCL-90 questionnaire is a validated survey that measures the degree of distress caused by 90 different symptoms. This slide lists the frequency of symptoms that were reported as being of at least moderate severity. Dizziness was reported by 70%, nausea by 56%, temperature fluctuations in 48%, numbness and tingling in 48%, along with heart racing, shortness of breath, chest pain, and reduced appetite—all indicating that adolescents with CFS have a number of common manifestations that go well beyond the list composed by Fukuda and colleagues in 1994.
The proposed Pediatric CFS case definition attempts to acknowledge this fact. It requires that, in addition to the familiar categories of post-exertional malaise, unrefreshing sleep, pain, and cognitive problems, those meeting the case definition have at least 1 symptom from 2 of 3 categories including autonomic manifestations like lightheadedness, palpitations, and shortness of breath, neuroendocrine problems such as heat intolerance, sub-normal temperature, temperature fluctuations, and impaired appetite, and immune manifestations, by which they mean sore throat, tender glands, and new sensitivities to foods.
Some of those symptoms foreshadow what I want to say about associated examination findings. There were papers in the early 1990s that emphasized the low yield of physical examinations in patients with CFS, the implication that this was a nothing disease that was largely psychosomatic in origin. The data do not support this assertion, and that is especially the case in adolescents.
One of the most readily apparent abnormalities on the examination of those with pediatric CFS is that resting heart rate is often elevated, and that HR and BP abnormalities with prolonged upright posture are common. I have listed some of the pediatric studies on the prevalence of orthostatic intolerance (OI) in the supplemental slides [provided to the committee].
This slide shows examples of what happens with the common forms of OI. In response to standing or upright tilt table testing, people with postural tachycardia syndrome (POTS) develop an excessive increase in heart rate in the first 10 minutes, associated with orthostatic symptoms of increased fatigue, lightheadedness, mental fog, or headache. The adult cut-off for excessive heart rate is a 30 bpm increase, whereas for adolescents we now require a 40 bpm change. One also needs to see a reproduction of symptoms, not just a HR change. Some develop a more profound drop in BP, as is seen in this panel, consistent with neurally mediated or vaso-vagal hypotension. Upright posture in these individuals is also associated with symptoms. Despite what some literature suggests, these two conditions are not mutually exclusive, and the treatments are almost identical.
One striking physical finding among those with CFS and orthostatic intolerance is shown here, namely a purple discoloration in the dependent limbs termed acrocyanosis. On the left is the hand of a young woman who had been standing for just a couple of minutes, with my hand shown for purposes of comparison. The right hand panel shows that after compressing my fingers against her lower leg, I had time to step back, pick up a camera, focus it, and take a picture, and she still had no capillary refill, reflecting an impressive abnormality in the circulation of blood.
In addition to the overlap between orthostatic intolerance and CFS, a little over a decade ago we started to notice that we were seeing a number of patients in the CFS clinic with a genetic disorder of connective tissue known as Ehlers-Danlos syndrome (EDS)—certainly more than would have been expected by chance. Those with EDS have stretchy skin, very loose, hypermobile joints that often dislocate easily, along with delayed wound healing, and a tendency to very early onset of varicose veins. The latter is due to the blood vessels being more compliant and stretchy than in unaffected individuals. Those with EDS have chronic fatigue and widespread pain of uncertain cause, which of course overlaps with the central features of CFS.
Here are some examples of joint hypermobility in our CFS patients. This slide shows items that comprise the 9 point Beighton score of JH: one point is assigned for the ability to dorsiflex each 5th finger past 90 degrees, one point on each side for bringing the thumb to the forearm, one for more than 10 degrees of hypermobility at the elbows, and knees, and the 9th point is for the ability to place the palms on the floor bending over at the waist.
Now, if CFS patients were anything like the healthy population, we would expect them to be less flexible after several months of reduced activity rather than more flexible. In contrast, here’s what we found when we examined things systematically:
Comparing 58 consecutive new patients with CFS to 58 adolescents without fatigue, note the completely different distribution of their results on the 9-point Beighton score. 60% with CFS compared to 24% of healthy children met criteria for hypermobility, identifying this trait as another exam abnormality that is present from birth, and constitutes a risk factor for developing CFS.
The synthesis of the pediatric CFS examination studies is that all studies in which the response to upright posture is measured in adolescents with CFS report higher rates of OI. Upright posture consistently aggravates CFS. All studies that examine HRV in adolescents report a sympathetic predominance. Dependent acrocyanosis is a common sign. JH is a risk factor for both CFS and OI. So, in contrast to the papers from the early 1990s, it turns out there are a number of objective abnormalities on the examination of adolescents with CFS. As Yogi Berra said, you can observe a lot just by looking.
Let’s shift the focus now to the impact of pediatric CFS.
As mentioned earlier, CFS is heterogeneous, and how it affects an individual will differ depending on the developmental circumstances of the child, the duration and severity of the illness, the number of physiologic disturbances that converge to create CFS symptoms, all modified by the quality of support from family, friends, and the medical profession. In adult CFS, several studies by Tony Komaroff and others in the 1990s showed that adults with CFS could have lower scores on quality of life (QOL) measures than those with multiple sclerosis or congestive heart failure.
In children, the comparisons of health-related QOL have been aided by the development of age-appropriate valid instruments like the PedsQL survey. This slide, generated from work done at the Cincinnati Children’s Hospital, compares the total Peds QL score in healthy children at the right, to those with sickle cell disease, diabetes, EGID, and CF. Using this Peds QL survey in our Pediatric CFS Cohort study, we were able to add in our data, showing that adolescents with CFS have lower QOL than children with any of these other conditions.
When you look at the Peds QL subscales in our Pediatric CFS Cohort study, physical function correlates most closely with school attendance, as many other pediatric studies have shown. Emotional, social, and psychosocial subscale scores are higher than the physical function subscale score.
Crawley in England has emphasized this point about school attendance. She enrolled 211 with CFS from her CFS specialist clinic in UK. Many studies have shown that CFS is one of the most common causes of prolonged school absence, and in this group 56.9% attending school 20% of the time or less. As would be expected, those with better physical function were more likely to attend school (OR 1.70, 95% CI 1.36-2.13). There was no association between attendance rates and anxiety, gender, age, or family history of ME/CFS. The dominant influence on attendance was the severity of illness.
Another way of measuring impact is to use the Functional Disability Inventory, a well validated 15-item measure of how well a child does with daily tasks like walking up stairs, doing something with a friend, being up all day without a nap, being at school all day, doing homework, or going shopping. Note the striking differences between CFS patients and controls, statistically significant at the .001 level.
We use the Wood Mental Fatigue Inventory in our studies as a valid measure of cognitive symptoms, higher scores meaning worse difficulties with concentration and memory. The mean score for those with CFS was 12, and for controls it was 3, with a median score for controls of 0.
In the last couple of minutes, I want to briefly mention what is possible with individualized, multi-modal treatment. Here are the preliminary wellness scores (ranging from zero, meaning “dying” to 100 meaning as well as one could imagine feeling) from our 2 year pediatric CFS cohort study that is nearing completion. Large gains in overall QOL were seen in the first 6 months, with further refinements in the mean wellness score over time.
But this kind of care is not available to many people due to gaps in medical service for CFS
When I spoke here on pediatric CFS in 2005, I used this slide. Frankly, despite the gains in knowledge about the illness, not much has changed in 7 years. We continue to have a striking mismatch of the patients and providers. Because insurers in this country do not reimburse in a manner commensurate with complexity of CFS, there are very few hospital/university clinics dedicated to CFS. We have no CFS training grants or federally funded research or treatment centers to help advance and translate knowledge to the primary care physicians.
Now, against this context of a service gap, I am very worried about an unintended consequence of changing the definition of Pediatric CFS to require 3 months of fatigue rather than 6. As a pediatrician and an advocate for the needs of those with CFS, I am sensitive to what I believe to be the intent of the proposal, namely to encourage earlier treatment of those with fatigue after 3 months. But, why restrict this to children? I think this suggestion should apply to adults as well. And, for that matter, we should be attempting to understand the cause of fatigue and to intervene at some level when the symptoms first arise. Why wait 3 months?
Let’s look at one of the problems that might arise for care of our sickest patients if we change the case definition. Here’s a hypothetical estimate of the increase in referral numbers after acute EBV infection, based on data from the large study conducted by Katz and colleagues in Chicago. Fatigue affects close to 100% at the onset of mono, but by 6 months, the Katz data show that 13% of patients meet criteria for CFS.
If approximately 50% still have fatigue at 3 months, then we risk swelling the CFS rolls with patients who will improve spontaneously. Even if just a fraction of the 37% of the total group who are going to improve spontaneously get in to see CFS specialists, the unintended consequence is that this will delay the evaluation of those who truly meet the criteria for CFS after 6 months or more, and who have more severe illness.
Now, this is a problem only because we are currently so deficient in meeting the needs of those with CFS at all ages, and that is a social and political choice we continue to make, to the detriment of young people with this illness. I think we can do so much better.
I want to end by thanking some of those who have made our work possible.
“Joint Hypermobility and CFS,” by Dr. Alan Pocinki, MD, of George Washington University Hospital. Posted: October 31, 2011.
“The Outs and Ins of Orthostatic Intolerance,” by Kim McCleary. Posted: June 19, 2011.
“Neuromuscular Strain and CFS,” from our SolveCFS publication. Posted: June 13, 2012.