Abnormal blood lactate accumulation during repeated exercise testing in ME/CFS, 2019, Lien et al

Merged thread

Abnormal blood lactate accumulation during repeated exercise testing in myalgic encephalomyelitis/chronic fatigue syndrome. Katarina Lien1,2 , Bjørn Johansen3, Marit B. Veierød4, Annicke S. Haslestad1, Siv K. Bøhn1, Morten N. Melsom5, Kristin R. Kardel1 & Per O. Iversen1,6. 1 Department of Nutrition, Institute of Basic Medical Sciences, University of Oslo, Oslo, Physiol Rep, 7 (11), 2019, e14138, https://doi.org/10.14814/phy2.14138


Cort discusses this here:
Norwegian 2-Day ME/CFS Exercise Study Adds Crucial Factor to Exercise Intolerance Findings
https://www.healthrising.org/blog/2019/10/11/norwegian-two-day-exercise-chronic-fatigue-lactate/

the crucial factor is lactate.
18 women with CCC diagnosis versus 18 matched controls
looks like further PEM validation. However is the study size still too small ?

 

Can anyone explain to me why the p values I’ve highlighted the graphs on pages 6 & 7 are so low? Surely those p values should not be so low? Or am I mis-interpreting something?

 

Sorry - should also have said that this is a very interesting paper, showing clear differences in how ME peeps react to exertion.

 

In conclusion, previous exercise deteriorates physical performance and increases [Laa] during exercise in patients with ME/CFS while it lowers [Laa] in healthy subjects.

That is clear then. The normal rules do not apply to patients.

When will Wessely and co understand that?

 

I think that the Answer is partly when our side funds larger studies that can’t be ignored, possibly...

 

I hope someone is going to flag this up to NICE?

 

@Simon M , I think your post above went a bit wonky. Unless it's in invisible ink you don't seem to have written anything.

 

The method used was paired t-test, which would compare the performance of the same participant on both days. If there is a consistent trend for most/all participants between both days then the p-value can indeed be that low.

The test that was not used was the two-sample t-test which simply compares the group means, which probably wouldn't have such a low P-value.

 

But the comparisons between MECFS & controls (eg two before samples) would not be paired. So surely there should be annotations to explain which were paired tests and which not??

 

Well the choice of the statistical test is mutually exclusive, but it could be stated in the figure description.

The frustrating part for me is the mean results and SDs were not provided (where the data was not skewed), along with a few other bits and bobs like the log10 tests, and respiratory exchange rate nor the heart rate at GET. I just hate "data not shown"...

 

Lien 2019: some analysis and thoughts​

The day 2 differences for patients in lactic production, and power at the gas exchange threshold (GET), look pretty interesting – though the differences aren't huge. However, the big differences between patients and controls seen on for almost every outcome at CPET1 and CPET2 are harder to interpret, because it looks like the controls are poorly matched (they are probably not really sedentary). The study made a large number of comparisons and did not correct for this statistically, though the most interesting findings would survive correction. It is a small study and the findings need replication, of course.

Poorly matched controls?

Controls' VO2 peak results were dramatically higher than patients' (37 versus 24 ML/KG/min), a much bigger difference than I remember seeing in previous studies. Conntrols "exercise less than twice a week on a regular basis ", but the results indicate they were quite fit (see Fig 2b in a new large European study of female healthy non-smokers with a similar BMI). This study took place in Norway and I believe that studies have shown that Norwegians are fitter than Americans who make up participants in most studies to date. So these controls may not be particularly sedentary.

Also, controls have a significantly low BMI than patients (22.0 versus 25.2) and are younger (age 34 versus 37) and both these factors (esp BMI) would lead to a higher VO2 peak.

The authors say that lower levels of blood lactate in trained athletes is due to faster lactate clearance, rather than fast at lactate production. And it is possible that greater fitness and (possible) faster lactate clearance explains the why controls comfortably outperformed patients in most measures in this study. Previous (single CPET) studies have inconsistent results on lactate vs conntrols, as the paper notes, so this is one more study adding to the uncertainty. The new study uses CCC criteria vs Fukuda in older lactate studies, but typically 90% of Fukuda patients have PEM so that might not be a significant factor.


Day one versus day 2

But poorly matched controls would not explain the findings of differences between day one and day two, and I think these are the meaningful findings from this study.

Patients producing significantly more lactate per watt output might be down to controls being much fitter and poorly matched. More interesting is that there is a small but significant decrease in lactate for controls on day two compared with a similarly small increase in patients. These changes, and the difference between patients and controls has a P value of <0.001 and so would survive statistical correction for the 50+ comparisons made in the study.

​

The authors say that regular endurance training improves lactate clearance and they propose that the day one exercise test leads (rapidly!) to fast at lactate parents and so lower lactate levels on day two. This is Interesting, though it is a pity there isn't other data to support this, such as lactate levels in controls in other 2-day maximal exercise tests.

Other lactate results also show high lactate levels in patients, but not controls, on day two. There is also higher lactate at gas exchange threshold (p = 0.03, which would not survive correction).

Higher lactate is also seen at the at the lactate turn point, LT, (effectively the switch to more anaerobic respiration, and usually seen as the underlying process measured by gas exchange threshold). The difference is also seen in power at onset of blood lactate accumulation, OBLA, which is the power output at 4 nmol per litre of lactate.

​

Replication of reduced power at GET on day-2

​

P = 0.02 would certainly not survive statistical correction in this study, but it seems fair to include it as it is really a comparison with other studies on a single measure.


No difference in delta peak VO2

Another study fails to find a drop in day to VO2 peak for patients. I agree with @Snow Leopard that non-athletes are unlikely to exercise to total exhaustion. But the initial finding from Workwell were held to be important because healthy controls and patients with cardiopulmonary diseases did not show a drop in VO2 peak on day two. I am not sure there is an obvious reason to assume a difference in motivation between mecfs patients, other patients and sedentary controls.

Fig 2B CPET1 vs 2 differences for mecfs and controls​

Note that there is actually a lot of variability in controls (suggesting that replication isn't always so reliable on day two) and a serious overlap with patients data. (I don't remember other studies showing so many individual data points, which is particularly helpful, especially in these rather small studies.)

The paper said that participants were given "strong vocal encouragement during the test". Heart rate and the RER both indicate high and similar level of effort in patients and controls in the study at VO2 peak. The authors also said that a 10% drop from day one was needed to be considered significant, and it doesn't look like we're even getting close to that.

What does it all mean?

I have to say, I am surprised we don't see a much bigger difference on day two between patients and controls in the studies. PEM is such a dramatic symptom and exercises such a big issue for us.

And why do lactate levels increase on day two in patients when they decrease in controls? What is the significance of reduced power at the gas exchange threshold, but not at VO2 peak?

Beats me.

However, I think this is important:

There is a lot of variation between patients in how long it takes for full PEM to cut in. But the point is very important. Given that these type of studies aim to probe what's happening with PEM, it is surprising that they don't routinely measure PEM in patients before the day 2 test. Perhaps having good PEM data would help to unpick the slightly confusing results from these two days maximal exercise studies.

 

@Simon M

Regarding the VO2Peak and peak power findings, it is notable that patients exercised to a higher heart rate and slightly higher RER on the second day, compared to the first, whereas the controls were the reverse. This suggests that the encouragement worked well for patients, but not the control participants.

But VO2Peak isn't very interesting unless you are an athlete. VO2Peak doesn't tell you how effectively the oxygen is being used for performance. (oxidative phosphorylation vs fatty acid oxidation, higher basal metabolic rate in the rest of the body, and buffering)

VO2Max is effectively just a measure of the blood being pumped by the heart. We know this due to the effects of blood doping on VO2Peak, along with phlebotomy eliminating the benefit of increased blood volume on VO2Peak.
https://www.ncbi.nlm.nih.gov/pubmed/24622974

Reproducing the same VO2Max suggests that there is insignificant differences in blood volume between the two days. Dehydration for example can lead to a loss of blood volume and lower VO2Max. The fact that VO2Peak is reproducible in patients (yet lower performance per oxygen consumption past the gas exchange threshold on the second day (Ventilatory threshold, VT1), provides stronger evidence that patients are suffering from increased fatiguability.

Conditioning the body to increase stroke volume requires bursts of intense cardio activity, typically exceeding 80% of peak heart rate and certainly exceeding the point of onset of blood lactate accumulation.
Neilly Buckalew commented on Cort's article on Health Rising, saying Lien et al. is flawed because it did not control for lean muscle mass and that lower lean muscle mass may explain the findings at VT1. I believe this is irrelevant. The main reason is that the finding is invariant of VO2Peak - controls with low VO2Peak are able to reproduce their performance at VT1, whereas patients with above average VO2Peak are not. Some ME/CFS patients may actually have higher lean muscle mass per VO2Peak, compared to the controls given that patients tend to avoid intense activity.
Nevertheless, Neilly Buckalew was suggesting to think along the lines of what is happening at the muscle fibre level and suggested using EMG, which I have been hinting at for some time. They will discover, on the contrary to hypotheses of "increased perception of effort", there is increased central effort and increased quantities of muscle fibres being activated earlier on the second CPET. (all this leads to my theory of fatigue perception, but alas I don't yet have enough data)

 

Thanks, @Snow Leopard, interesting.

VO2max is certainly affected by the volume of blood pumped, and also by the quantity of red blood cells (hence the effectiveness of erythropoietin in anaemia and in doping). But it must also be related to actual oxygen consumption by the body. If tissues are not consuming oxygen, then blood returning to the lungs won't be fully deoxygenated and so will take up less oxygen as a result. VO2max does reflect oxygen consumption by the body.

This is the heart of the matter: how effectively is the oxygen turned into energy?

This is where it gets tricky. Because not only was there no difference in delta oxygen consumption for patients vs controls, there was no difference in delta power either.

​

And that's what I find so puzzling, particularly given that power at GET is reduced on day two. Though it's worth noting that power at the lactate turn point is not different. The authors comment
"One explanation might be that some patients probably reached their LT during unloaded pedaling on CPET2, at which point we did not have lactate measures every 30th second. This would affect the computed regression lines. Nonetheless, [Laa] showed the same pattern, with similar [Laa] on CPET1 and significant differences in [Laa] between the two study groups on CPET2. A left shift of the lactate curve in ME/CFS". But the finding adds some uncertainty.

We do know that there is more lactate accumulation on day two for patients compared with controls. The paper says that differences in lactate accumulation in healthy people are due to differences in clearance, not production. And they suggest the same could be true here
"PBMCs from ME/CFS patients show upregulated expression of PDK (Fluge et al. 2016), proposing a disturbed PDK regulation which could limit the pyruvate flux with the potential to affect the clearance of lactate through oxidation.".

That's why I wrote my previous post: the different findings don't really fit together comfortably. I guess that science in the raw.

I think using electromyography during these tests is a smart idea and could reveal a lot.

Presumably "they will discover" is a prediction rather than something you already know? And it would make sense if there was a problem with energy production. But if muscle fibres have to be recruited sooner, wouldn't that then lead to a lower VO2 peak, which we don't see?

It is quite possible that are small differences in power output and VO2 at peak activity are masked by the uncertainty around peak measures. But that means they must be quite small, which still surprises me. Maybe using electromyography will provide a way forward.

 

This does give me some pause for thought. But peak power doesn't necessarily happen at the same time as VO2Peak. The graph shows there was in fact a drop in most patients, but that some controls also had a drop. A percentage drop would also be more relevant than just watts, but I'm not sure how that would affect the sensitivity of the comparison.

Neither measure will have much sensitivity unless we can guarantee all patients are reaching a true exertion peak.

​

The build up of slightly more lactate is merely a consequence, not a cause. Difference in power output per amount of lactate at various points would be the interesting measure.

Yes.

No, burst power during maximal voluntary contraction (eg. within first five seconds of Wingate anaerobic test) can be at a far higher power output, as much as twice or more than is measured on a maximal CPET. Likewise, this occurs at a much higher rate of muscle fibre recruitment, but is still not 100%. Which is why also twitch interpolation can be a bit misleading if the problem is fatigue in muscle fibres which are primarily utilising oxidative phosphorylation for energy. As I said, (given constant gas exchange dynamics) VO2Max is primarily a measure of the oxygen carrying capacity of the cardiovascular system, rather than a measure of maximal energy converted using oxidative phosphorylation.

I'm still not convinced that patients and controls are consistently reaching VO2Max and peak CPET wattage, given my personal experience of the 2 day CPET. I realise my experiences may be different to others given differences in fitness and ramping protcols, but on the first day I reached VO2Peak (probably a true VO2Max) of the test around eight minutes after reaching VT1/GET and reached peak power (10w higher, 40-50 seconds later) when I had to stop after feeling very dizzy. I still had more leg power available, but that power was primarily anerobic. I'm sure you can appreciate how this leads to variability in results that detracts from what we would like to or expect to see. Faster ramping wouldn't necessarily solve the problem (eg. in my case I may have stopped earlier due to being unable to sustain performance with the power ramp step).

 

Does anyone have access to the Ethical application and/or research protocol for Liens study? Or knows how one might get access?

 

Hi Anna, if you email the main author and ask kindly you may obtain that information that you want.

 

Sorry about the very late reply to your post. This was a point I wanted to make, but never got around to.

I think we need to keep in mind that her statement was aimed at the Norwegian ME-debate. She has been activly engaging in the debate here, and knows her proponents arguments. One of them have been accusing, very loudly, ME-doctors advising pacing and avoiding PEM of making their patients illness worse by endorsing inactivity.

She is usually very good at also making a point of just how little activity can be sustained, and uses real world examples as how draining a shower can be in watts (?) vs how low energy production is in even mild and moderate patients.

Also, we don’t really have that focus on activity programs in Norway, at least not for adults.

Our main problem is the focus on ‘instant heal’ and how LP have been a synonym for ‘ME-treatment’ :-/

 

I do not understand much about the details of the biology or statistics in these studies, but I do wonder if some of the results are not as dramatic as they might be because people with ME, have had to find compensatory mechanisms to be able to function at all.

Even away from the cell biology side people with ME are desperate to show they are not ill because they are lazy so are likely to put more effort into the exercise. Then it is hard to find matched controls because no one would chose to do as little exertion as we do. The laziest of the healthy population still walks to get a glass of water when they are thirsty.

At the cellular level, we do not understand ME well enough to rule out that we utilise some unusual mechanisms to find the energy we need.

I think what I am trying to say is that the fact there are differences may be more important for knowing where to look next than the actual size of these differences.

For instance, if someone gets a rash when they touch latex it is important that it happens at all. The size and extent of the blisters is only of secondary interest.

 

Thank you @Milo! Asking for a friend, I will let her know .

 

It should also be possible to obtain from the regional ethics commitee (REK) I belive.

Link to English page.

https://helseforskning.etikkom.no/r...iginformasjon?p_dim=35030&_ikbLanguageCode=us