Another interesting read from Peter Buzzacott and friends!
http://journal.frontiersin.org/article/10.3389/fphys.2017.00181/abstract
http://journal.frontiersin.org/article/10.3389/fphys.2017.00181/abstract
Cutting Edge Bubble study indicates possible damage to circulatory system
- Thread starter Randy Thornton
- Start date
It's not my study, nor do I claim to be a "bubble expert" but the way I read this report is that potentially all bubbles have a potential to damage the circulatory system.
Not sure if you are channeling Ross or joking... But a positive dose response relationship with a plausible mechanism of injury (within the scope of one study) is evidence of causality.
AND we needed a study to confirm that - ???
regards Baz[/QUOTE]
Baz, I guess the most interesting thing to me is to more fully understand exactly how the bubbles are affecting the circulatory system.
uwxplorer
rEvo carrier
I am essentially paraphrasing the paper:
"Recently, we found that endothelial dysfunction correlated positively and linearly with bubble amounts in a rat DCS model, indicating a close relationship between bubble formation and endothelial injury (Zhang et al., 2016b)"
or:
"Whether bubbles are the cause or not, endothelial dysfunction is detectable and obvious following most diving exposures"
and later:
"The present results confirm our previous finding that endothelial dysfunction correlates well with bubble formation (Zhang et al., 2016b)."
and:
"The results also reconfirm that bubble amounts are an objective and suitable parameter to predict endothelial dysfunction".
Yes and it seems pretty obvious those lead to the 2nd to last sentence in their abstract.
"The results further verify that bubbles are the causative agents of decompression induced endothelial damage and bubble amounts are an objective and suitable parameter to predict endothelial dysfunction"
That's the sentence I thought you were commenting (negatively) on. Seems beyond mere correlation and pretty causal to me.
You’re right, of course.
However, in a proper experiment, eliminating spurious correlations is the whole point. That’s why they used 68 rats and not just one.
Let’s consider the specifics. They found a correlation between total bubble count and levels of three molecules.
Timing is relevant.
Peak bubble score was 20 minutes after the dive.
Peak ET-1 and ICAM-1 levels were 20 hours after the dive.
Clearly ET-1 and ECAM-1 are not causing the bubbles.
The nature of the molecules is relevant. From what I could find...
High levels ET-1 have been correlated with heart disease. ET-1 is also known to promote angiogenesis (creation of new blood vessels). Generally speaking, to me it seems like the body is signalling that there’s something wrong with the circulatory system.
ICAM-1 binds endothelial cells (blood vessels) and leukocytes (white blood cells) and facilitates the crossing of the later into tissues.
I’m not clear on the role and significance of MDA. Maybe somebody can explain?
So, after the dive, first we have bubbles, then we have the body signalling damage to blood vessels and foreign invaders, and there's a correlation between bubble count and the amount of signalling.
Rather strong evidence that bubbles are damaging the blood vessels. Either by bursting their way out of the tissues, or possibly by being blocked there.
Cheers,
Matthieu
Last edited:
uwxplorer
rEvo carrier
This sentence should probably have been corrected by reviewers, as this paper does not offer proof that bubbles cause damage. It observes that bubbles are produced, that symptoms of cellular damage are detectable at the molecular level, and that these observations are consistent with a common hypothesis that the existence of bubbles create the damage.
The authors do not say anything else, except in this unfortunate sentence.
Note that I am not arguing with the hypothesis itself (that bubble create damage). I am just trying to clarify how scientific arguments work.
The sentence by Randy:
is what triggered my comment that correlation is not causation, in the sense that, for instance, potentially only some bubbles could be causing the damage. In fact, bubbling has been observed in asymptomatic divers in the past, which by itself is indication that not all bubbles cause damage.
Note that the authors themselves recognize that another phenomenon (as of yet undetected?), could be the source of the problem, and happens to be accompanied by the very visible creation of bubbles: "No matter whether the injury was directly or indirectly from bubbles, bubbles are the most likely initial causative agents of endothelial dysfunction following diving decompression". They can't prove it, but they favor the hypothesis. Fair enough.
I will note in conclusion that this study doesn't bring much in terms of how to prevent DCI. It would have been nice to try to treat those poor rats in a chamber and see whether those markers are prevented from shooting up. Or maybe give them an experimental anti-DCS pill before the dive... Now that I think of it, what about half-an-hour on a vibrating mattress?
BTW, the authors seem to be looking for divers to "explore the etiology, prevention and treatment of decompression sickness".
Any volunteers?
Surely you’re aware that in science things are only ever proven wrong. Never right. That everything remains falsifiable is fundamental to the scientific method. Therefore proof is impossible.
Going back to the problem at hand...
To use an analogy, you’re painting this up as if all they established is that the suspect had a gun, and the victim was found with holes in his body, and that's consistent with the suspect shooting the victim. You’re ignoring that there’s three empty rounds in the revolver and three holes in the victim, that CCTV in the corridor shows the victim alive through the door when the suspect entered the room and dead when he left, and that the police thought he would try to kill the victim but couldn’t nab him in time.
Not proof, you’re absolutely right - that’s not possible, never mind expected.
Bit different, though.
And that's fine, but that's not "correlation isn't causation". "Correlation isn't causation" does not address the magnitude of an acknowledged causal link. It denies the existence of any causal link in first place. That is, it could be a coincidence. Or there may be something else causing both.
In any case, all bubbles have a potential to damage the circulatory system versus only some bubbles are actually causing damage, I don't see the contradiction.
It could be that only bubbles that eventually reach a certain size cause damage.
It could be that the observed bubbles are not causing damage because the ones that do are the one that never made it out of the capillaries.
It could be any number of things.
Or it could be that all detectable bubbles do indeed cause damage, but just not necessarily enough damage, or not enough localised damage, for the diver to notice. Endothelial cells die after a couple of months to a couple of years. The body is well able to handle some of that. Which fits in nicely, incidentally, with the bubble grade vs DCS correlation.
Now that is "correlation isn't causation" - common cause. That's not what they're saying, though. What they are saying is that whether bubbles cause damage, or bubbles cause some other thing that causes damage is irrelevant... Nothing about some third factor that causes both.
They don't have proof so it's down to personal preference?
No. The hypothesis was tested. It could have failed. It didn't. Any number of alternatives, stated or not, did fail. Any alternative must agree with theirs in the tested scenario. That's not them "favoring".
The important thing about science is not that positive proof is impossible, it's that it's not necessary.
First, as the body of evidence that has failed to disproved the hypothesis grows, the room for alternatives gets smaller. Which means that the confidence that the hypothesis will in fact accurately describe a new problem increases.
Second, unless there was mistake, involuntary or otherwise, which is always possible, new data does not erase previous data. So when a hypothesis is proven "wrong" by new data, it still remains valid for a range of cases that includes those that do have supporting data. And any better hypothesis will have to say the same thing in those cases.
In other words, the whole thing is always moving forwards.
What does this mean here?
Let's imagine that some mechanism is discovered that both create bubbles and damage the endothelium. Bubble don’t damage the endothelium! Yay!
Except that that alternative mechanism will still have to show a linear relation between amount of bubbling and amount of damage. Then what? Boo…
Until that happens, though, until there’s some suggestion of what that might be, or some contradictory evidence emerges, it's fair to just say that bubbles are “the causative agents of decompression induced endothelial damage”. Would you rather have a long list of disclaimers all the way down to quantum physics? Worse case, bubbles are not in fact the causative agent, but damage will still be proportional to bubbling, so anyone who assumed that it was will still be good, so there's no problem.
I'm a bit confused by this post, I have to say. There's a number of things that seem worth saying about the paper...
If you said that this is just one experiment, and that pending some corroboration, or a more detailed explanation of the mechanism, some scepticism is in order, I’d be right with you.
If you have contradictory evidence, that’s even better.
If you said that because the rats were sacrificed, the sample size at each sampling time is rather small (which they note), fair enough.
If you said that the excessive dive profile may trigger effects that do not happen on “normal” dives, I’d point out that some rats did show low bubble levels anyway, and we still have a linear relation, but fair enough again (because of the point above).
If you pointed out that there’s an oddity at low bubble levels because the rats that have few bubbles still show elevated marker levels (which they don’t discuss), that’s certainly seems interesting to me (if I was to hazard a wild guess, I’d say the relationship is not linear but an exponential of some kind).
But all this “proof” stuff? I just can't get there. Maybe the wall of text is a hint of that... Sorry
Cheers,
Matthieu
The fact that reviewers (at least one peer, potentially three peers plus an editor) did not "catch" this terminology "error" suggests that its not actually the mistake you think it is. But that point aside, if this study (in your mind) doesn't offer proof of causality, then what DOES it take consistent with the current evidence to prove that bubbles cause endothelial damage? Fantastic Voyage perhaps? (1966 film where a tiny spacecraft navigates the human circulatory system)...
Simon Mitchell in post on another forum made the following statement: "It is generally accepted that decompression and bubble formation initiate inflammatory responses in the body. What is less clear is the extent to which these responses are responsible for the symptoms of decompression sickness. To be clear, we don't fully understand the relative contributions of the physical effects of bubbles and harmful effects of inflammation. Related questions will define one of the controversial issues surrounding this paper: viz, what do the measured differences in inflammatory markers really mean? I think the best way to interpret it is that the inflammatory markers are a plausible measure of decompression stress and (perhaps) a poorly characterized indicator of risk of DCS."
Not sure if the endothelial damage would be considered an inflammatory response or not, but obviously bubbles are not our friends!
Not sure if the endothelial damage would be considered an inflammatory response or not, but obviously bubbles are not our friends!
uwxplorer
rEvo carrier
Not sure if the endothelial damage would be considered an inflammatory response or not, but obviously bubbles are not our friends!
...or not. Many have argued that bubbles are good, since that means we are off-gasing.
Bubbles in tendons, cartilage or other poorly vascularized places, not so much.
As far as reviewers missing points of details, or obvious mistakes, this wouldn't be the first time that happens and certainly not the last one.
Anyway, enough metaphysics and epistemology as far as I am concerned
Thanks for pointing the paper to us.
Last edited:
Not sure if the endothelial damage would be considered an inflammatory response or not, but obviously bubbles are not our friends!
I'm not sure what is and isn't defined as an inflammatory response but endothelial damage could quite plausibly be related to cardiac events by loosening or dislodging pre-existing plaques or through other stress related changes. This has been an active body of research and testing for quite some time https://www.cedars-sinai.edu/Patien...er/Services/Endothelial-Function-Testing.aspx
kwinter
PITA
Maybe I'm an idiot, but I didn't think this was news. Most people don't have a PFO or fistula or AV malformation or shunt of any kind. But we still listen for bubbles with a Doppler and we are listening to the VENOUS side. So the concern has always been for venous bubbles. Sure, there is a greater pathway for damage if they cross over to the arterial side. But we never thought venous bubbles were harmless. I guess the thing that is interesting is the mechanism of injury?
That is what is interesting about this study Ken. (At least to me!) I think there has always been some level of discussion going on about exactly how the bubbles were destructive and incredulous as it may sound, there seems to be some people who still feel that there is no direct correlation between bubbles and DCS. For me, the big question that remains is whether the damage caused by bubbles at a cellular level is permanent in anyway.
The argument (particularly prominent in the Ross-Simon deep stop threads) has been that venous bubbles are natural, "good", and a harmless expression of off-gassing even at high bubble loads - as long as they get filtered out by the lungs before the blood enters the arterial side. This study demonstrates that venous bubbles cause damage - to veins themselves. How much damage is problematic is still to be worked out