(See end for  a short synopsis)

“Collectively, these data indicate that ME/CFS participants were unable to reproduce most physiological measures at both maximal and ventilatory threshold intensities during test 2, despite exercising to maximal effort during both tests” Keller

tired woman on bicycle

This study suggests the cause of post-exetional relapse is pretty clear – poor energy production!

The Keller paper starts off proposing that if a bout of exercise makes you weaker rather than stronger, something is probably wrong with your ability to produce (aerobic) energy. That makes sense, and three studies have now demonstrated that in ME/CFS, but assessing one’s aerobic energy capacity is not a test one associates with an ME/CFS diagnosis – at least not yet.

Inability of myalgic encephalomyelitis/chronic fatigue syndrome patients to reproduce VO₂peak indicates functional impairment. Keller BA, Pryor JL, Giloteaux L. J Transl Med. 2014 Apr 23;12:104. doi: 10.1186/1479-5876-12-104.

This study – the fourth to suggest that metabolic issues are keeping people with Chronic Fatigue Syndrome from hitting the gym (or even the streets, sometimes) – simply validates what a critical area of research exercise is, and in the most exercise-challenged illness of them all.

The Study

In this study twenty-two people with Chronic Fatigue Syndrome (ME/CFS) underwent a two-exercise challenge: two exercise tests to exhaustion within 24 hours of each other. Keller examined multiple variables during the tests.

The VO2 max at peak effort of the ME/CFS patients was only 77% of expected normal in the first test, and things got worse from there.

Peaking Early: Across the Board Declines in Metabolic Functioning at Peak Effort

The broad declines in almost all the measures on the second test suggested that, yes, something went very wrong with the ME/CFS patients’ energy production system. The breadth of the declines was notable. Most successful research studies show some statistically significant abnormalities, but in this study virtually every measure of cardiovascular/metabolic functioning declined significantly and, in most cases, statistically very significantly on the second exercise test. That suggested the energy production system got whacked in just about every way possible.

The probability figures denote the probability that the results occurred by chance. A p< .05, the cutoff for a ‘significant’ or meaningful result, indicates a 5% probability that the result occurred by chance. That’s good enough for most researchers to call their study a success.

declining graph

Declines in parameters of energy production in ME/CFS were broad and consistent.

We know about the problems with VO2 max (14% decline, p<.001), but Keller’s study also showed that at peak exercise the ME/CFS patient’s peak heart rates declined by about 16% (p<.001), ventilation was down by 15% (p<.003), carbon dioxide output was down by 16% (p<.001), and cardiac output (oxygen delivery) dropped by 9% (p<.003).

The ME/CFS patients’ heart functioning, breathing, rate of CO2 removal, and the amount of oxygen delivery as well as their oxygen consumption (VO2 max) declined dramatically.

The same pattern prevailed when they reached the point of greatest lung workload, i.e., when their lungs were exchanging the most gas with the atmosphere. The peak oxygen consumption dropped by 16% (p<.003), their work output dropped by 22% (p<.03), their carbon dioxide output dropped by 11% (p<.014), their cardiac output (oxygen delivery) dropped by 13% (p<.003), and their ventilation dropped by 8% (p<.035). Their heart rates did not drop significantly.

Everyone Flunks

When she put aside the group data and looked at the 22 patients individually, she found that everyone tested flunked the VO2 max test in some way. Most patients had reduced VO2 max at peak exercise, but those who did not had reduced VO2 max at their ventilatory threshold.


Toxic by-products that gather in the muscles during anaerobic energy production in ME/CFS cause fatigue, pain and other symptoms

Our anaerobic energy production system gets engaged when we participate in short bursts of rapid exertion such as running up the stairs to get to a meeting. Toxic byproducts accumulate in the muscles and blood when this occurs, but a return to the slower-paced aerobic energy production system usually quickly removes them. The reduced aerobic energy production found in ME/CFS patients suggests they can’t engage in many normal activities without entering into anaerobic metabolism. A wimpy aerobic energy production system that is less effective at cleaning up the toxins results in increased recovery times.

Interestingly, Keller said ME/CFS patients who are able to successfully reproduce their VO2 max/VT levels on the second test often have problems with blood pressure/heart rate regulation; i.e., something else is wrong.

Next, Keller compared the past 2-day ME/CFS CPET (maximal exercise) studies. Note that what we have here is what three exercise physiology groups found in three different locations looking at three different sets of patients. This is what you want in medical research: validation by different laboratories looking at different patient groups. Let’s see what’s happened.

Comparing Studies or Will the Real Metabolic Dysfunction Please Stand Up?

The results of this study confirm previous work that demonstrated an abnormal
response to exercise in fatigued ME/CFS patients. Keller

Keller’s comparison of her results with the results from prior two day maximal (CPET) exercise studies revealed some similarities and some significant differences. The upshot is that people with ME/CFS appear to be pretty adept at flunking a second exercise test; they just sometimes do it in different ways.

V02 Max

The first study by Snell et al containing just six ME/CFS patients and six controls found a 22% drop in VO2 max. A second larger study from the same lab containing 61 participants did not find significant differences in VO2 max between the first and second exercise tests, but it did find significantly reduced work at ventilatory threshold. Using VO2, VT VO2 as well as peak workload and VT workload the Snell group was able to diagnose correctly 95% of the ME/CFS patients.

Vermoulen also did not find significant reductions in VO2 max in his first study. The 6.3% drop in VO2 max found in ME/CFS patients in the Vermoulen 30-person study (15 ME/CFS, 15 controls) was essentially inconsequential statistically since it was within range of normal test-retest variability. Keller noted, however, that Vermoulen did not thoroughly explain his testing parameters in that study.

With Keller’s finding of reduced VO2 max, we have two studies finding reduced VO2 max and two studies not finding it.

Ventilatory Threshold – the New VO2 Max?

broken chain

Ventilatory threshold – which indicates the point at which the aerobic energy system wears out – may be a better measure for ME/CFS than VO2 max.

Ventilatory threshold indicates the point at which increased levels of CO2 and lactic acid cause your breathing rate to markedly increase. It’s a sign that anaerobic metabolism has come to dominate your energy production, and as such may be a better indicator for ME/CFS. Another way of thinking about this is to see ventilatory threshold as the point at which your ventilation or breathing has little impact on energy production because your aerobic system is tapped out, and you’re now running on anaerobic metabolism. VO2 max at ventilatory threshold is the point of maximum oxygen consumption during this time.

We were fixated on VO2 max at peak workload, but V02 max at ventilatory threshold may be the more common problem in ME/CFS. Keller, Van Ness, and Snell all found reduced VO2 max at ventilatory threshold on the second exercise test. With the fifteen percent or so decreases in VO2 max at ventilatory threshold, ME/CFS patients are far higher than the 1.5% changes in healthy controls or athletes.

The amount of work measured at ventilatory threshold was reduced in the Keller and Snell study, but not in the Vermoulen study.

The reduced ventilatory threshold during the second test indicates that something happening during the first exercise session impaired the ME/CFS patient’s ability to produce energy aerobically in the second test. That reduced aerobic production resulted in their entering into anaerobic metabolism much more quickly than normal. As Dr. Keller put it, exercise knocks people with ME/CFS into ‘premature anaerobiosis’.


Keller’s and Snell’s study found significant decreases in work, and she found reductions in heart rate while Vermoulen found neither. Both Keller and Vermoulen (in his latest study) found evidence of reduced oxygen delivery. The other studies did not look for it.

Not Equipment Failure!

This drop in aerobic capacity from one exercise test to the other is so unusual that other researchers have assumed something must be wrong with the equipment, so Keller did something in her paper that I’ve never seen before: she took care of that elephant in the room by bluntly stating there was nothing the matter with her equipment, because repeat CPET tests during the same time period in healthy controls were normal.

A Unique Group of Patients

ME/CFS patients currently represent a unique class of ill patients who do not reproduce maximal CPET measures, unlike individuals with cardiovascular disease [27,30], lung disease [28], end-stage renal disease [26], pulmonary arterial hypertension [25], and cystic fibrosis [29]. Keller


ME/CFS is the only disorder thus far identified with this strange reduction in energy production after exercise.

Calling ME/CFS patients a ‘unique class of ill patients’ she outlined how unusual these findings are by noting that people with cardiovascular disease, lung disease, end-stage renal disease, pulmonary hypertension and cystic fibrosis can all do repeat exercise tests with no drop in aerobic capacity.

VO2 max at Peak Effort vs VO2 max at Ventilatory Threshold – Does it matter?

I asked both Staci Stevens of Workwell and Dr. Keller if the two measures are showing something very different. A reduced VO2 max shows an inability to generate normal amounts of energy. A reduced VO2 max at anaerobic threshold shows that ME/CFS patients enter into anaerobic energy production more quickly.  (You can measure VO2 max  at while resting, at peak workload, at AT…). Staci stated both measures show metabolic dysfunction, and that’s what’s important.

It was significant that Snell used four metabolic measures (VO2, VT VO2, peak workload and VT workload) to diagnose correctly 95% of the ME/CFS patients in their study correctly.  The individual measures weren’t as important diagnostically as the fact that some type of metabolic dysfunction was present in the ME/CFS patients and was not present in the healthy controls.

Both exercise physiologists use anaerobic threshold to identify the heart rates ME/CFS patients should stay below in order to use their clean-burning aerobic energy system and avoid more pain and fatigue.

Dr. Keller suggested that biggerstudies could correlate specific metabolic abnormalities with different symptom clusters.

The Big (Big) Study

Both believe large-scale studies will be necessary to tease out the subsets present, search for biomarkers, and fully document the exercise challenges present in ME/CFS. We’re taking about a several hundred patient study costing from a half million to several million dollars. Both stressed these studies are completely doable; it’s just a matter of funding. Given studies of that scope Dr. Keller believes it’s quite possible they would find a biomarker that shows up during the first exercise test that would allow them to dump the second one.

When asked about the responses she’s had thus far to the study (which was published about a month ago) Dr. Keller said she’s gotten positive response from the ME/CFS community but has not heard from exercise physiologists who are not engaged in ME/CFS work. This didn’t surprise her.

“I’ve not heard from other exercise physiologists (other than Workwell), but that’s not unusual. Frankly, I don’t see others running to study ME/CFS until there’s an indication of funding availability for important studies and grant reviewers who are supportive of others to come into the field.”


  • Dr. Keller finds across the board  declines in almost all measures of energy production in ME/CFS patients in a second exercise test
  • This suggests damage occurring as a result of the first exercise period inhibits energy production afterwards.
  • This is a highly unusual  – perhaps unique finding – that has not been found in other any  disorders to date
  • Dr. Keller’s comparison of the three other studies with similar findings suggests  that metabolic dysfunction is common in ME/CFS but that variations may occur
  • A large study able to definitively identify the subsets in ME/CFS and perhaps identify biomarkers  that would allow only  one exercise test to be done would probably  cost  from $500,000 to several million dollars.

Health Rising Really Does Rely on Your Donations! 🙂

A BIG thanks to the 320 people who helped make Health Rising's fundraising drive a success!

Stay Up to Date with ME/CFS, Long COVID and Fibromyalgia News

Get Health Rising's free blogs featuring the latest findings and treatment options for the ME/CFS, long COVID, fibromyalgia and complex chronic disease communities. 

Thanks for subscribing! :)

Pin It on Pinterest

Share This