Nature.com report that Alter paper reviewers want "additional studies"

[alex3619]

Hi Tania,

The cynic in me has responded to this on another thread - I think if we ask for the samples to be retested we will find the samples were used up. So sad, limited supply, yada yada yada

However, if I am wrong and there is enough left, then half should go to the WPI, and half to an independent lab that all parties are happy with.

Bye
Alex

Interesting.. there if there is actually no XMRV to be found in those CDC studies (cause they killed the virus or whatever).. would this then mean that these samples will get added to the testing the NIH already did.. hence adding the CDC samples to there own and hence then lessening the percentage of the all up outcome?

 

[dean]

4. The following items are about activities you might do during a typical day. Does your health now limit you in these activities? If so, how much?
(check one on each line) ACTIVITIES

Yes, Limited A Lot

Yes, Limited A Little

No, Not Limited At All

TYPICAL DAY ????? Any study of CFS that asks patients questions like this ignores the periodicity of the disease. This can't be stressed enough.....You can be good one day, and bedridden the next. Any analysis based on these questions that try to get at exhaustion levels will be worthless....or worse.... The waxing and waning of the disease, has to be taken to account in any research that looks at activity levels. Whoever is writing these questions doesnt seems to have talked to many patients with CFS...... It is something to ask about every study that purports to study CFS.

 

[Levi]

Lets see, how do I do this politely . . .?

TGOP,
Please look at my link from above:
http://www.newswise.com/articles/view/536946

Note the methodology and results.

For their study, the researchers screened 19,381 residents of Georgia, selecting 292 people who had CFS, 268 who were considered chronically unwell, and 163 who were considered well to participate. The researchers then measured free cortisol concentrations in saliva, which was collected on regular workdays, immediately upon awaking and 30 minutes and 60 minutes after awakening.

Now look at the source of CFS patients in the new CDC article:
http://www.retrovirology.com/content/pdf/1742-4690-7-57.pdf

Look under the section of Methods/Study population and specimen preparation on page 17 of the PDF. Note reference 2 which is found on page 28 of the PDF:

2. Reeves WC, Jones JF, Maloney E, Heim C, Hoaglin DC, Boneva RS,
Morrissey M, Devlin R: Prevalence of chronic fatigue syndrome in
metropolitan, urban, and rural Georgia. Popul Health Metr 2007, 8:5.

This study is found here:
http://www.cdc.gov/cfs/publications/surveillance_6.htm

Note the methology and results.

Methods: Based on a random-digit dialing survey we ascertained CFS cases and controls to estimate the prevalence of CFS in metropolitan, urban, and rural populations of Georgia. This report focuses on the 5,623 of 19,381 respondents ages 18 to 59 years old. Fatigued (2,438), randomly selected unwell not fatigued (1,429) and randomly selected well (1,756) respondents completed telephone questionnaires concerning fatigue, other symptoms, and medical history. Subsets of those identified by interview as having CFS-like illness (292)

I submit to you that these separate studies used the the same CFS study population of 292 patients from Georgia gleaned by the CDC in 2007/2008. At least they represent a significant part (22 of 51 CFS patients) of the study population in the new CDC article. Or is this just a weird coincidence? If this IS the same 292 people, then they have been tested for cortisol levels and were found to have low cortisol. Mikovits has stated that cortisol is up-regulated in CFS patients, not down-regulated. My source for that is a still-embargoed presentation she gave at UOP last fall. I was there and heard it first hand.

So, when it comes to the bio-marker cortisol, we have apples and oranges vis-a-vis the WPI study population and the CDC study population. That should come as no surprise. . .

Thanks for the observation Levi, but I have to disagree with your assumption that the CDC cohort had low cortisol.
From what I can see there is no evidence to support that assertion at all.

As far as Dr. Alter's cohort is concerned, we have no information and so IMO, it's pointless (unhelpful) speculating / starting unfounded rumours.


TGOP

 

[Eric Johnson from I&I]

> stated that cortisol is up-regulated in CFS patients, not down-regulated

I believe that point may be debatable

 

[Levi]

Eric,

Right, to clarify, I am not saying Mikovits was right about that, but the Science Article study population very well may have been selected using cortisol levels as one of many bio-markers. The discussion of the selection criteria for the Science article is pretty vague about if and how bio-markers were employed. However, I do believe that WPI used bio-markers heavily in the selection of thier patient study population.

> stated that cortisol is up-regulated in CFS patients, not down-regulated

I believe that point may be debatable

 

[Angela Kennedy]

I think it would be best to access a copy of the published JCEM paper so that full methodology and results can be seen.

A whole load of questions arise about the claims made in this 'newswise' article.

But does any of this necessarily mean 'CFS' is diagnosed by the Reeves/CDC et al team by cortisol as a 'biomarker'?

I can't see evidence to support that, to be honest.

 

[Levi]

Angela,

I could not readily find a link to the JCEM paper. Apparently, you could not either or you would have given the link in your post. Look, these are all CFS patients selected by William Reeves, a psychiatrist who is on record maintaining that CFS is not an organic illness, rather it is an unwellness belief of stressed out psychiatric patients. Who were probably sexually abused as childen, to boot (yes, he has published a study).

The one CFS bio-marker that W. Reeves has shown any interest in is cortisol, and he has used the Georgia dial-up study population repeatedly to study associations between CFS and cortisol. So they have been tested. If the markers do not line up, then the study populations will probably not either.

http://www.endfatigue.com/health_articles_r-s/Sleep-cfs_low_morning_adrenal_cortisol.html

A bio-marker is just that, a marker. Most markers are just a small part of any diagnostic process. My point here is that as far as the CDC and WPI go, we are talking about a marker that would seem to correlate inversely to each organizations paradigms about CFS.

Why is any of this important? I think the big question fulminating from all the conflicting XMRV research is whether the research conclusion discrepancies lie in the PCR testing procedures, or with the selection of the various study populations. I submit that its not the tubes, its the patients used in each particular study.

I think it would be best to access a copy of the published JCEM paper so that full methodology and results can be seen.

A whole load of questions arise about the claims made in this 'newswise' article.

But does any of this necessarily mean 'CFS' is diagnosed by the Reeves/CDC et al team by cortisol as a 'biomarker'?

I can't see evidence to support that, to be honest.

 

[redo]

This morning the Patient Advocate blog said that the NIH research people are being asked to test the CDC samples (the ones designed to NOT have XMRV in them - with ridiculous cohort, wrong test tubes, and whatever else).
http://cfspatientadvocate.blogspot.com/
I don't know what his source is for this, but it does worry me. I fear the CDC using these shenanigans to try to dilute or weaken the powerful findings of the NIH/FDA paper.

I got a bad feeling about this too. Chances are many of the CDC samples aren't CFS samples. This was one of the worst moves I could see.

Testing the WPI's samples would be much better.

If you got thew enregy to do it, I'd suggest you'd write them.

 

[Angela Kennedy]

Levi,

My problem is with the 'cortisol as biomarker of CFS' is around the lack of unequivocal findings about 'cortisol' in 'CFS' (it all seems VERY equivocal), methodology problems (cohorts, defining 'CFS', lack of cortisol level testing of more specific controls - for example other illness sufferers), weak correlations at best - not actually fulfilling Bradford Hill Causation Criteria, the weak but trumpeted association of these already weak correlations around 'low' cortisol in 'CFS' with child 'abuse' or 'trauma' and 'stress' (and all the problems in how THOSE are defined), retrospective reporting of same (!), the possible association of high cortisol with 'depression', etc. etc. which makes the possible comment by Mikovits about HIGH cortisol in XMRV confusing to say the least. etc. etc. etc.

The confounders are compounding alright.

I don't think we are in any position to be confident in the strength of cortisol as a 'biomarker' for either of these cohorts. But i do agree with you that the cohorts are different. But this appears to be in the CDC ability to exclude physical signs of illness that would be found in 'Canadian' ME/CFS (and in this respect the Reeves 'empirical' definition is just like Oxford, even Fukuda and London with a little wriggling), and the amazing use of the 'Role Emotional Scale' to select 'CFS' contenders.

ALL the negative studies have very similar problems. They're either deliberately selecting out people who comply with Canadian criteria (which they can do under the auspices of the Oxford etc. criteria), or, in the case of the Kerr study, not taking enough steps to select in 'Canadians', so that their cohort is unknowable, leading to speculation as to whether people with the neurological and other system problems found in Canadian defined sufferers can be in there. The people who believe they are only going on trust- and bearing in mind the Kerr 'sub-grouping' and his psycho-social 'stress' promotion as an explanation for 'CFS', that cannot safely be done either.

What a mess. That 's even without the testing methodology problem.

It would be funny, if it wasn't such a tragedy for real people with massive impairment causing physical devastation also having to contend with being constructed as malingers/hypochondriacs/wimps who can't handle 'stress'/odd personalities and the social and material exclusion caused by all of that.

 

[Levi]

Angela,

I probably should not have used cortisol for my bio-marker example, since it is apparently a loaded issue here. Like I said to Eric, I make no claims about its worth in actually diagnosing CFS/ME. It would seem that both CDC and WPI have looked at it though, and drawn different conclusions.

I see you are from across the pond where ME patients are hunted for their pelts by GPs and subjected to indignities and humiliation that I personally have no experience with and can only fathom in the abstract. You folks have a lot riding on the outcome of all this weirdness which would make me prickly too.

You also have Simon Wessely, who puts William Reeves to shame as the Rasputin of CFS/ME research, since he has actually managed to single-handedly eradicate medical recognition of ME, which never actually gained much of a foothold as a disease entity in the USA.

When all the smoke clears, everyone will still have to deal with the distinction between spontaneous and epidemic forms of the disease(s). We can all look forward to that.

Levi,

(snipped)But i do agree with you that the cohorts are different.

(re-snip)
ALL the negative studies have very similar problems. What a mess. That 's even without the testing methodology problem.

(More snipping)
It would be funny, if it wasn't such a tragedy for real people with massive impairment causing physical devastation also having to contend with being constructed as malingers/hypochondriacs/wimps who can't handle 'stress'/odd personalities and the social and material exclusion caused by all of that.

 

[anciendaze]

Cortisol and biomarkers

Cortisol can be used diagnostically, but not with the naive "too high, too low" interpretations. The endocrine system is a very complicated dynamical system with a great deal of feedback. A patient can be very low on cortisol at some time and then go unusually high at other times in the same day. There is a defect in the response to stress which has measurable properties.

For comparison of levels versus dynamics, you might consider the traditional idea of taking your pulse and deciding it is too high or too low, versus the array of tests a cardiologist now uses.

Orthostatic intolerance is another example where CFS/ME patients don't fit in the too high/too low paradigm. I have literally been checked out as fine w.r.t. blood pressure, then become dizzy when I stood up after examination. In one case a suspected neurological problem was reduced to "unexplained loss of consciousness". (If you pay more, you get a diagnosis of "idiopathic syncope".)

There are similar issues with many other biomarkers. Part of the trouble in settling on one is the difficulty many people in medicine have in understanding anything more sophisticated than a simple comparison of numbers.

 

[Hope123]

I agree with the comments about cortisol. Cortisol can fluctuate a lot during the day, can be affected by many factors, and must be tested carefully. Also, there have been conflicting studies about cortisol in CFS in the past if I remember right.

Also, nowhere in the current Retrovirology papers is cortisol mentioned as a criteria for how they selected their subjects in any of the cohorts (Wichita/ Georgia community/ Georgia Bibb registry) used in the study. Instead, the Reeves Empiric Criteria (which doesn't include any labs at all) is used.

Not everyone in the Georgia community study had low cortisol either.

Also, when the WPI paper came out, scientists were asking why the WPI didn't test all 101 CFS subjects nor all the controls with all the tests available. E.g. why were only the 33 PCR (-) tested using antibody/ Western Blot, etc.? The same question could be asked of the CDC:

There were 292 CFS patients identified by Georgia community and 65 CFS in Wichita. But the CDC only used 33/292 from Georgia and 11/65 from Wichita in the current study? Why were these specific patients selected? Did they fit the Reeves Empiric? (in which case it's a small percentage from each study) Were these the only people who had blood stored? Were they selected randomly? Inquiring minds want to know.

 

[Cort]

I agree with the comments about cortisol. Cortisol can fluctuate a lot during the day, can be affected by many factors, and must be tested carefully. Also, there have been conflicting studies about cortisol in CFS in the past if I remember right.

Also, nowhere in the current Retrovirology papers is cortisol mentioned as a criteria for how they selected their subjects in any of the cohorts (Wichita/ Georgia community/ Georgia Bibb registry) used in the study. Instead, the Reeves Empiric Criteria (which doesn't include any labs at all) is used.

Not everyone in the Georgia community study had low cortisol either.

Also, when the WPI paper came out, scientists were asking why the WPI didn't test all 101 CFS subjects nor all the controls with all the tests available. E.g. why were only the 33 PCR (-) tested using antibody/ Western Blot, etc.? The same question could be asked of the CDC:

There were 292 CFS patients identified by Georgia community and 65 CFS in Wichita. But the CDC only used 33/292 from Georgia and 11/65 from Wichita in the current study? Why were these specific patients selected? Did they fit the Reeves Empiric? (in which case it's a small percentage from each study) Were these the only people who had blood stored? Were they selected randomly? Inquiring minds want to know.

I imagine that the CDC used patients that fit both the Empirical and Fukuda criteria and this cut down the number substantially..but there must've been more filtering involved I would think.

My recollection of the cortisol studies, and there have been so many of them in CFS, is that about 40% of them are negative - do not show abnormal cortisol. In the notorious sexual abuse study cortisol I believe was low in patients with a history of sexual abuse but not in the significant proportion of patients that did not have a history of sexual abuse. In any case, a significant proportion of patients did not have low cortisol readings.

 

[taniaaust1]

Hi Tania,

There is a variant to 3 that is part of the optimist view. The DHHS have finally realized how problematic the CDC research is, wanted them to publish, and are now setting things up to discredit the CDC prior to investigating their problems.

Bye
Alex

oh lol. i didnt think of that but thou i do believe in possible conspiratory, its very hard for me to believe that one gov agency would purposely try to discrete another government agency.

 

[taniaaust1]

Thanks for the observation Levi, but I have to disagree with your assumption that the CDC cohort had low cortisol.
From what I can see there is no evidence to support that assertion at all.

As far as Dr. Alter's cohort is concerned, we have no information and so IMO, it's pointless (unhelpful) speculating / starting unfounded rumours.


TGOP

Maybe Levi has made that assumption due to the CDC cortisol findings in a study a while ago. CDC found that one third of those with CFS had lower levels of cortisol.. i cant now thou remember what group of CFS patients the CDC picked in that study, it thou probably did it's usual thing.

 

[taniaaust1]

Originally Posted by glenp

"4. The following items are about activities you might do during a typical day. Does your health now limit you in these activities? If so, how much?
(check one on each line) ACTIVITIES

Yes, Limited A Lot

Yes, Limited A Little

No, Not Limited At All"
.......................
"TYPICAL DAY ????? Any study of CFS that asks patients questions like this ignores the periodicity of the disease. This can't be stressed enough.....You can be good one day, and bedridden the next. Any analysis based on these questions that try to get at exhaustion levels will be worthless....or worse.... The waxing and waning of the disease, has to be taken to account in any research that looks at activity levels. Whoever is writing these questions doesnt seems to have talked to many patients with CFS...... It is something to ask about every study that purports to study CFS. "
.................

I actually thought that question was a good one, as it didnt ask about how you've been just this week, or this fortnight (as you know some may crash for a fortnight or have a better couple of weeks, only to crash again, those in push/crash cycles.. but it said "typical day". I'd take that to do an average of my condition over the past couple of mths. What i think they did wrong is not having a cut off point for CFS with that question.. as we all know real CFS (and not just fatigue) limits "A Lot". That should of been a defining feature to get into a CFS study.

(i personally dont think they should be using anyone in CFS studies who isnt on an average kind of day, limited by at least 50% of what they could do if well).

If they are thou using just symptoms rather than how limited one is, to get into studies. That may keep some with quite severe CFS esp the long timers out of some studies as we've learnt to stay within strict limits of what is done to avoid symptoms. My symptoms nowdays are post exertional (i can nowdays prevent symptoms for weeks by doing practically nothing.. im over 60% limited thou).

 

[taniaaust1]

TGOP,
Please look at my link from above:
http://www.newswise.com/articles/view/536946

Note the methodology and results.



Now look at the source of CFS patients in the new CDC article:
http://www.retrovirology.com/content/pdf/1742-4690-7-57.pdf

Look under the section of Methods/Study population and specimen preparation on page 17 of the PDF. Note reference 2 which is found on page 28 of the PDF:


This study is found here:
http://www.cdc.gov/cfs/publications/surveillance_6.htm

Note the methology and results.


I submit to you that these separate studies used the the same CFS study population of 292 patients from Georgia gleaned by the CDC in 2007/2008. At least they represent a significant part (22 of 51 CFS patients) of the study population in the new CDC article. Or is this just a weird coincidence? If this IS the same 292 people, then they have been tested for cortisol levels and were found to have low cortisol. Mikovits has stated that cortisol is up-regulated in CFS patients, not down-regulated. My source for that is a still-embargoed presentation she gave at UOP last fall. I was there and heard it first hand.

So, when it comes to the bio-marker cortisol, we have apples and oranges vis-a-vis the WPI study population and the CDC study population. That should come as no surprise. . .

Interesting that Mikovits has found it to be upregulated. From as far as i can work out, it seems to be upregulated in those who havent had CFS for a long time (early in CFS).. but is often low in those who have had CFS for a long time. ive come to that conclusion due to studies done and throu knowing lots with CFS over the years and talking to them about their cortisol results. My own cortisol is just out of normal range of results and low.. mine varies in tests from 48-51 (normal is 50-350) (ive had CFS/ME for 13 yrs).

i think that is another thing researchers need to be looking at.. at how CFS can be different in the early years of it to in the later years of it. It's not just the cortisol results which seem to change and having those who have only had CFS for a couple of years in the same studies as those who have had it a decade, with results at opposite ends it may end up making a study appear to have average results by the time you average everything out.

 

[taniaaust1]

Cortisol can be used diagnostically, but not with the naive "too high, too low" interpretations. The endocrine system is a very complicated dynamical system with a great deal of feedback. A patient can be very low on cortisol at some time and then go unusually high at other times in the same day. There is a defect in the response to stress which has measurable properties.

.

nods.. i know one who clearly had proper ME and her cortisol was both abnormally high and low (it came back different in tests.. her doctors were struggling to make sense of this) .. so like most things in CFS .. obviously something was dysfunctional or dysregulated.

 

[Levi]

Study populations; CDC vs. WPI

Cort,

Lets assume for the sake of hypothetical argument here that the CDC, having all the cortisol level information on these patients, decided to filter the new XMRV study population to maximize patients who had been sexually abused. Maybe because they are more likely to have fatiguing illness that is non-viral in nature based on Reeves clinical experiences. Maybe there was even a question in the dial-up survey that targeted a prior history of sexual abuse (do we HAVE those phone survey questions?)

Folks here are saying the CDC would never do something sneaky like that , and even if they did, they would surely mention it in the report (not). The PCR operators would never know anything about this. They would just blindly test away at the blood Reeves was giving them. It would be simple to find this out. Just have one of our intrepid CFS reporters put the question directly to the CDC spokesperson about the cortisol levels of the cohort. You will always get straight talk answers from the CDC . . .

In the still embargoed UOP presentation I mentioned, Dr. Mikovits mentioned that the Science Article study population was chosen using 27 different bio-markers (including cortisol levels), together with predictive software using Random Forest Algorithms to discover the patients most likely to be XMRV positive. Or something similar, my notes are a bit fuzzy.

I believe this is why that Mikovits talk is still embargoed to this day. This was before it was obvious that they should have mentioned these methods in the Science Paper, together with the need for declaring possible conflicts of interests like her being the vice president of VIPdx, a commercial XMRV testing lab. Water under the bridge.

Anyway, WPI/Mikovits would have probably eliminated the low cortisol patients for study, because their predictive methods of using bio-markers were for maximizing the representation of XMRV positives in the group in order to slam-dunk the results. Or at least exclude other non-XMRV CFS sub-sets. I am coming to the belief that however well-intentioned anyone has been here, that impartiality may have taken a back seat to advocacy on both sides of the XMRV mess. Its hard to explain otherwise.

In the notorious sexual abuse study cortisol I believe was low in patients with a history of sexual abuse but not in the significant proportion of patients that did not have a history of sexual abuse.

 

[Anika]

All available specimens were used in study, per CDC paper; "empirical" criteria used

I imagine that the CDC used patients that fit both the Empirical and Fukuda criteria and this cut down the number substantially..but there must've been more filtering involved I would think.

Apparently, there aren't very many suitable specimens archived from the CDC's studies in Wichita and Georgia, including the provider Registry. According to the paper, they tested ALL the specimens that were available from these studies - at least those that had enough "volume" to do their tests. This is what they say at pages 19-20 of the provisional pdf at http://www.retrovirology.com/content/pdf/1742-4690-7-57.pdf (emphasis is mine):

Specimens were available from 89 persons (33 CFS and 56 well controls) from

the population-based case-control studies and 18 CFS persons from the Registry study

described above. Subjects were included based on availability of specimens, and

comprised 11 of 43 persons with CFS and 26 of 53 healthy controls from Wichita, KS

and 22 of 32 persons with CFS and 30 of 51 healthy controls from Georgia. ...Clinical and demographic characteristics of subjects with specimens available

for this study did not differ from those persons who did not have ample specimen

volumes and case-control matching was maintained.​

And, remember, there is no distinction between "empiric / Reeves" and "Fukuda" in this or other current CDC papers. "Pre-Reeves Fukuda" has no independent meaning to Reeves or CDC any more.

If I remember correctly, the "Reeves / empirical" criteria were developed from the Wichita studies. Dolphin has some good posts on the "empirical" or "Reeves" criteria I should look back at, but I think the gist of one of his posts was speculating that the Wichita study came up with so embarrassingly few CFS cases under the "unfiltered" Fukuda criteria, that they came up with the "empirical / Reeves" criteria to get a larger number of CFS cases out of the study, or they wouldn't have had anything worth publishing about (please feel free to jump in and correct me or clarify here). I don't think there's any verified explanation for how they came up with the criteria.

So, when Reeves and CDC now talk about Fukuda criteria in current publications, they mean the Reeves / empirical criteria, which they rationalize as simply "operationalizing" Fukuda - though they have done absolutely no studies to validate that assertion.

Look at around pages 19 - 20 of the paper where they describe the study population, and you'll see them using the "Reeves/empirical" version of the criteria. That is what allows them to include as CFS (as severely impaired) at least one person who scored 100 out of 100 on the physical functioning scale of SF-36. (emphasis mine):

Subjects with CFS had been ill on average 13.9 years (median 11.15 yrs, range 3 - 40 yrs), were

severely fatigued (MFI General Fatigue 16.5, range 10 - 20; MFI Reduced Activity 12.8,

range 4 20) and severely impaired (SF-36 physical functioning 65.5, range 10 -100);

SF-36 bodily pain 48.8, range 12 84), and 3/33 (9%) reported sudden onset to their

illness.​

Sorry for the long post.