Post-exertional symptoms distinguish Myalgic Encephalomyelitis/Chronic Fatigue Syndrome subjects from healthy controls

2.1Participants

Sick subjects from different parts of the United States were referred by their health care providers to the testing facility for evaluation of functional capacity. Fifteen of the 49 were formally recruited for the study and met the Fukuda (1994) criteria. The remaining 34 were physician diagnosed. Ten sedentary, but otherwise healthy, control subjects from the community were also recruited for the study. Control subjects did not participate in an exercise program or perform more than 30 minutes of moderate physical activity on a weekly basis. All subjects were required to undergo and complete two days of exercise testing for inclusion in the study.

All participants signed an informed consent document prior to testing. The study was reviewed and approved by the Institutional Review Board at University of the Pacific.

2.2Two-day cardiopulmonary exercise testing

Subjects were asked to refrain from all forms of physical activity for at least 24 hours before exercise testing. All protocols followed the procedures outlined by Stevens et al. [21] and described elsewhere [2, 19].

2.3Questionnaires

Subjects were given an open-ended questionnaire to complete over the two days of testing and into the week following both tests. Subjects evaluated how they felt immediately after the first exercise test, before and immediately after the second exercise test, 24 hours after the second test and during the seven days after both tests. They were also asked to specify how long it took them to recover from the two-day CPET. For this, they chose from a range of answers (e.g., less than 1 day, 2 days, 7 days, still not recovered, etc.). Completed questionnaires were either mailed, faxed or emailed to the testing facility seven days after the second exercise test.

2.4Categorization of responses

Fifty-nine surveys were digitized and labeled by study identification numbers. A coding schema, based on a review of literature and the authors’ clinical and research experiences, was developed prior to the evaluation of the questionnaires. For example, the symptom category, ‘Cognitive Dysfunction,’ was designated when responses were related to memory, concentration, confusion and/or problems with communication, reading or information recall. Common jargon utilized by patients, such as “brain fog”, “trouble finding words”, or “feeling [mentally] out of it”, were also considered acceptable descriptors for Cognitive Dysfunction. Next, researchers who were blinded to demographic information and subject diagnoses (LJM, LC) tested the schema on a small, random subset of surveys. The schema was then refined reiteratively to assure that it thoroughly and accurately reflected the experiences of subjects.

A total of nineteen codes, or symptom categories, were established. The symptom categories were Cardiopulmonary, Cognitive Dysfunction, Cold Limbs, Decrease in Function, Fatigue, Flu-like Symptoms, Gastrointestinal, Headache, Increase in Sensitivity, Light-headedness, Mood, Muscle/Joint Pain, Neurologic, Pain (for general pain or pain unfit for the other categories), Sleep Disturbances, Temperature Control, Tingling, Positive Feelings and Weakness. Excluding Positive Feelings, the remaining eighteen symptom categories were considered adverse. Representative keywords and phrases followed each category to classify responses. Two of the researchers (LJM and LC) independently reviewed each subjects’ questionnaire using Excel spreadsheets which listed the nineteen symptom categories at each of the five time points (i.e., immediately after the first exercise test, 24 hours after the first exercise test, immediately after the second test, 24 hours after the second test and a week after the second test). Subjects’ open-ended, written responses were interpreted as much as possible at face value, without effort to uncover the meaning behind phrases beyond the words written (i.e., reviewers focused on manifest rather than latent content). Spreadsheets were completed in a binary fashion. A ‘1’ signified that the subject had experienced the symptom category at that time point and a ‘0’ signified that the subject had not.

Upon completion of coding for a select number of questionnaires, the reviewers’ results were compared to identify any differences. Where there were discrepancies, the reviewers referred to the original classification schema and arrived at a consensus for how to designate a response. The average discrepancy rate was 1.7% which equates to an average of roughly two differences within each survey. Four rounds of coding were necessary to review the fifty-nine questionnaires (i.e., reviewers conducted the coding process for 16 questionnaires, 14 questionnaires, 13 questionnaires then 16).

2.5Data analysis

All data were exported and organized into a master Microsoft Excel spreadsheet. The master spreadsheet included each subject’s identification number, demographic and diagnostic data, the symptom categories at each time point, the final versions of the questionnaire interpretations and a column that determined whether the subject experienced the symptom category at any point. Total number, means, and standard deviations of symptom categories and duration of symptoms for both groups were calculated by a third researcher (JMV). Student’s t-test was used to compare continuous data from each group, while Fisher’s exact test compared categorical data (https://www.graphpad.com/quickcalcs/). Bonferroni’s correction adjusted for multiple comparisons. A two-tailed p-value equal to or less than 0.0025 was considered significant.

3.1Demographic analysis

A total of 59 subjects participated in the study (n = 59). There were no statistical differences in age or anthropometric indices between the ME/CFS group and the control group. Seventy percent of the control and 80% of the ME/CFS subjects were female. The average age for the ME/CFS patients was 41.3±7.0 years, while the average age for the control subjects was 35.3±12.2 years. The average height and weight were 1.6±0.6 m and 65.5±10.3 kg for the ME/CFS group and 1.6±0.5 m and 68.7±9.5 kg for the control group.

3.2Quantitative analysis

The post-exertional symptoms reported by the ME/CFS patients were more diverse than those reported by the control subjects. All symptom categories were reported by at least one subject from the ME/CFS group, except for the symptom category, Cold Limbs. Subjects in the control group did not report any symptoms corresponding to Cognitive Dysfunction, Decrease in Function, Flu-like Symptoms, Cold Limbs or Cardiopulmonary. On average, the ME/CFS subjects’ responses accounted for more symptom categories than the control subjects’ (7.63±3.30 vs. 1.90±1.52, p < 0.001) (Table 1, column 2). Over the days of testing, the average number of symptoms reported by the ME/CFS group was greater than that reported by the control group (14.35±7.30 vs. 3.50±2.59, p < 0.002) (Table 1, column 3). Adverse responses in the week following the two-day CPET were seen only in the ME/CFS group (4.14±2.65), with none of the control subjects reporting any symptoms (Table 1, column 4).

Of the symptom categories that the control subjects experienced, the ME/CFS group reported almost all of them at greater frequencies at each of the five time points. Figure 1 displays eleven of the nineteen categories and the percentages at which the ME/CFS and control subjects reported them. The symptom categories are shown on the x-axis, while the associated percentages at each time points are shown on the y-axis. The symptom categories that were reported with the greatest frequency by the ME/CFS group are shown at the top of Fig. 1. These include Fatigue, Muscle/Joint Pain, Cognitive Dysfunction, Decrease in Function and Headache. For comparison, the symptom categories and their percentages for the control group are seen at the bottom of Fig. 1. Similarly, Sleep Disturbances, Pain, Weakness, Cardiopulmonary, Light-headedness and Flu-like Symptoms are displayed for the ME/CFS and healthy groups at the top and bottom of Fig. 2, respectively. The only symptom categories with similar percentages between the ME/CFS subjects and the control subjects were ‘Pain’ at time point 1 (immediately after the first test, 16.3% vs. 10.0%, respectively) and ‘Weakness’ at time points 2 (24 hours after the first test, 12.2% vs. 10.0%) and 3 (immediately after the second test, 20.4% vs. 20.0%). Table 2 shows symptom expression over time for further comparison. For each symptom category at each time point, counts and percentages are given for each group.

Fig. 1

Group comparisons for symptoms: Fatigue, Muscle/Joint Pain, Cognitive Function, Decrease in Function, and Headache.

Fig. 2

Group comparisons for symptoms: Sleep Disturbances, Pain, Weakness, Cardiopulmonary, Light-headedness, and Flu-like Symptoms.

To account for subjects who might have reported the same symptoms during multiple time points (e.g., a single subject reporting Pain over three time periods vs. three different subjects each reporting Pain once during different time periods), Table 3 exhibits the number of subjects in each group who endorsed a symptom category during any of the five time periods. Except for Positive Feelings, all symptom categories were endorsed by more ME/CFS subjects than control subjects. Significantly more ME/CFS subjects experienced Cognitive Dysfunction and Decrease in Function than control subjects (61% more subjects for both categories, p < 0.0025). Conversely, 52% more control subjects communicated Positive Feelings than ME/CFS subjects (p < 0.0025).

The time to recover from the bouts of exercise was also dramatically different between the groups. For 24 of the 49 ME/CFS subjects, the average time to return to their baseline of health was 4.5 days (Table 1, Column 5). The remaining 25 ME/CFS subjects reported that they had not recovered by the seventh day after the second exercise test. For the control group, recovery was rapid. All but one control subject returned to baseline in less than one day of the CPET (Table 1, Column 5).

3.3Qualitative analysis

The descriptions of symptoms were qualitatively different between the control and ME/CFS subjects. Control subjects used terse descriptions such as “fatigued,” “tired,” “headache,” “sore,” “hot” and “muscles hurt” to explain how they felt after the exercise tests. In contrast, ME/CFS subjects provided detailed accounts of their symptoms. One subject wrote:

“[I was] too fatigued to do any basic activities of daily living. I would have been in bed all day, but I was too weak/fatigued to walk back upstairs to get there, laid on couch all day instead. I have been on a downward spiral since the test. I have done nothing but rest, and I continue to feel worse and worse. The level of fatigue is indescribable. I’ve been too fatigued to shower for the last 4 days, to wash my hair since before the test, to make food, to answer phone, understand what I’m watching on TV, couldn’t get my taxes done. This is the most extreme my fatigue has been in a long time.”

Other typical statements from ME/CFS subjects about their experiences included:

Many of the control subjects expressed positive feelings after exercise compared to ME/CFS subjects. For example, two control subjects mentioned increased energy immediately afterwards which extended to the next day:

Collectively, the narrative of the post-exertional experience differs between an ME/CFS patient and a control subject.

4.1Limitations

There are several limitations in the study. First, the average age of the study sample is younger than in other ME/CFS studies. Second, the questionnaire did not quantify length of recovery beyond seven days, decreases in functionality, and symptom magnitude. There were options for subjects who recovered in less than one day and in 2 days, but not for subjects who recovered in 1 day. Subjects also could not disclose how many days past a week it took them to return to their baseline level of health. Open-ended responses were restricted to a one-week timeframe following the CPET. Future questionnaires might include a standardized instrument that allows subjects to describe decrements in the ability to perform basic or instrumental activities of daily living. Furthermore, a study should be performed to assess and quantify the severity of reported symptoms. If post-exertional symptoms had been reported both quantitatively (i.e., numerical representations for symptom magnitude) and qualitatively (i.e., descriptions written by subjects) within the current survey, the data might have shown that the symptom categories reported by both subject groups were experienced at disparate intensities. Third, the coding schema was constructed based on our ideas about PEM. The independent reviewers used it to categorize responses at face value. Latent content analysis and in-depth questioning might reveal further details or lead to different interpretations. Finally, our findings are based on subjects’ responses after maximal effort and physical exertion. Submaximal or other types of stressors may produce different results. Despite these limitations, these results agree with and support prior studies. Future studies may consider designing a study without these limitations and including objective CPET results within the analysis.