Is there research evidence to support this?
-
New Forum Software Installed (Still a few issues but reopening the forum) More details.
Experience with LDN? low dose naltrexone
- Thread starter Jonathan Edwards
- Start date
-
- Tags
- ldn
Very good point regarding testing itself. I struggle with texts and would never do a form or something where getting it wrong would be important when under the weather - which sometimes is an issue if said thing needs to be done within the week but can often be worked around.
Having tests where the timing is vital is a problem in itself - even taking meds at the right time can be hard if your alarm goes off but you are that poorly at that moment.
Expecting all the functions that would be involved with opening, looking at, holding and selecting buttons on a phone and being time-specific I'm not even sure mildest in a crash or just after something has 'overdone them' (which could be as simple as a flashing light or something else 'inruding' ie out of their control noise-wise such as building work, or someone dropping by for a visit or call) that is an appropriate expectation for the condition.
I am pretty confident from having met enough of them over the years that product designers would have the temperament and personal qualities to get this sort of thing correct and get frustrated that aspects like this aren't briefed to that sort of expertise (ie researcher says 'this is the measure we aim to get', and it is the product designer working with a properly representative group of patients to the sample being used for research that works out how that can actually be operationalised accurately). So I find it frustrating more studies don't appear to be using that and instead assuming they can do it themselves when that often isn't the case for various reasons.
Most studies will require proxies where you think of the exact measure you want then have to sit down and see what data is available or you can get, and then have to be really good at working out what actually proxies what and it isn't as straight a line as many think if you don't use the expert in that domain to work out what is underneath each dataset and underneath how whatever measure will be used. I've been there myself.
I can't get past thinking that the representative sample of patients is important (and then them probably having other experience or knowledge) being a vital aspect of that and how I never hear of any panel being developed of said expertise in order that is 'to hand', along with developing a list of good product designers - it seems fundmanetal a competency/bank to move a lot of possible research forward (and doesn't need to be 'fresh' each time, which will come with other issues about 'are they qualified or genuinely representative'). That is how normal market research works in the exploratory stage and get shocked at how that is often missing.
DigitalDrifter
Senior Member (Voting Rights)
LDN never made a difference for me. I think it's probably an over hyped treatment just like vitamin b12 and other supplements were popular back in 2011 - 2013. The problem with patients testing these treatments is that many ME patients have fluctuating symptoms which can co-incidentally change the same time as starting treatment giving the illusion that they are effective. It is best to wait for blinded placebo controlled trials before accepting anecdotes as facts.
I did plenty of experimenting with LDN. The effects on my pain were reliable. I forgot doses reasonably frequently, and it was the re-emerging pain that reminded me that I'd forgotten a dose. The pain went away for a few years, and then returned recently, and LDN (several years past its expiry date!) worked effectively the way it had before. I admit that this isn't a well-designed double-blinded test, but I'm convinced that it works well for me, and probably for other people. It doesn't work for all PWME, because this disease varies so much.
Right, neither can be clinically measured or even clearly defined. I don't put much weight on studies based on reported pain or fatigue, because those can't be measured reliably.
Jonathan Edwards
Senior Member (Voting Rights)
This is what one keeps hearing but it is interesting to consider how the physicians ever came to know this was the right thing to do. It is easy to think physicians have an eye for such things - 'clinical judgment' - and can be skilled at juggling doses.
My experience of the reality is that one doesn't;t really have a clue as to what is going on and what is the right dose. I asked PubMed to show trials wit dose titration and the first hit I looked at was on AHD trials where there is also a go low go slow policy.
The conclusion was: 'no discernable scientific justification for any particular dose was given.'
Quite a few resources on this here https://ldnresearchtrust.org/how-low-dose-naltrexone-works
I'm not sure if it's been flagged that Luis Nacul is principal investigator of a double blind, randomized, placebo controlled, LDN trial for 'post-covid 19 fatigue syndrome' (PCFS) with 160 participants. The titrations are given in the arms and interventions table. Whilst pwme are technically not included, the trial registration page has the following on PCFS as a subset of Long Covid:
https://clinicaltrials.gov/ct2/show/NCT05430152
Thread here.
https://clinicaltrials.gov/ct2/show/NCT05430152
Thread here.
Jonathan Edwards
Senior Member (Voting Rights)
I had a look at that and was not impressed that there was anything other than speculation involved.
If we do not even know its a drug works it is premature to claim to know how it works!
There seems to be a suggestion that the effect might be due to a dextroisomer, rather than the levoisomer binding to endorphin receptors. If so, I would have thought it was worth finding something that was not quite so muddled as a racemic mixture of a tiny dose of a drug that has no very obvious reason to work (either isomer) in either FM or ME - since neither is know to involve the pathways mentioned.
Mister Person
Established Member (Voting Rights)
Hehehe this is why I didnt listen to any dosing strategies, though I would think just logically go low go slow is a good bet for ppl w who r sensitive
Yeah, but the titration is also the perfect way to get people to believe that a drug works. Someone starts with 0.5- no result. A month on and they try 1.0. Still no result. Another month on, they try 1.5, still no result. Another month on, they try 2.0. Fantastic, they feel a bit better! They report 'the drug helps, but it's important to work out your individual dose'.
But is it the drug, or is it a natural fluctuation upwards? A few months later, they try 2.5 - they feel a bit worse. Is it a side effect, is it because they have moved away from their perfect dose? Or is it just a natural fluctuation downwards, back to their normal state of ill-health? So, they go back to 2.0, and they feel a bit better than how they were at the 2.5 dose. Or, they give up on that drug and move on to another one - but their report of how great the drug was hangs around on the internet or in anecdotes, uncorrected by later events.
I mean no disrespect to people who believe LDN or any drug helps their ME/CFS. Maybe it does. But personal experimentation is not a good basis for determining if something really works. It should not be up to patients to work out what works - clinicians who prescribe these drugs, and especially drugs with significant side effects like Abilify, are irresponsible if they don't do what they can to get well-designed clinical trials underway.
But is it the drug, or is it a natural fluctuation upwards? A few months later, they try 2.5 - they feel a bit worse. Is it a side effect, is it because they have moved away from their perfect dose? Or is it just a natural fluctuation downwards, back to their normal state of ill-health? So, they go back to 2.0, and they feel a bit better than how they were at the 2.5 dose. Or, they give up on that drug and move on to another one - but their report of how great the drug was hangs around on the internet or in anecdotes, uncorrected by later events.
I mean no disrespect to people who believe LDN or any drug helps their ME/CFS. Maybe it does. But personal experimentation is not a good basis for determining if something really works. It should not be up to patients to work out what works - clinicians who prescribe these drugs, and especially drugs with significant side effects like Abilify, are irresponsible if they don't do what they can to get well-designed clinical trials underway.
Jonathan Edwards
Senior Member (Voting Rights)
I agree that there are various aspects of the LDN story that have an alternative medicine ring about them and might be convenient if wanting to argue against negative results from trials.
1. 'Go low, go slow' is a standard mantra for unproven remedies used off label. Dose schedules for drugs that have a good evidence base have been worked out.
2. The suggestion that fillers may be responsible for side effects that may make it look as if the drug is not helpful when it is if taken without these fillers.
3. The suggestion that LDN is not working through its normal pharmacology (which sounds psychiatric) but through some speculated action on immunity (which sounds better but nobody has shown it to be relevant).
4. The suggestion that the benefit of LDN is due to a different isomer. Until we actually know there is a benefit this must also be speculation.
On the other hand features of this sort can also go with drugs that don't work terribly well and have lots of side effects (hence the need to go low and slow) but do have proven benefits in formal trials (even if maybe overegged and without enough emphasis on side effects).
I think methotrexate (MTX) for RA is a good example. You are supposed to start low and increase to maximum safe or tolerated level. Side effects have been linked to oral usage and some physicians have claimed it should be used IM to get higher doses in. Nobody knows how MTX works in RA and the dose is lower than in cancer. It may be something to do with a specific effect on adenosine pathways in inflammation but I don't think it has ever been sorted.
MTX produces a demonstrable improvement in a good proportion of RA patients but rheumatologists often forget that the remission rate is something like 5%, whereas the previous generation of drugs - gold and penicillamine - probably produced remission in as much as 20% of cases. For gold this could be permanent. The reason for gold and penicillamine going out of fashion was that their side effects were life threatening and unpredictable.
I am cutting this into two posts.
1. 'Go low, go slow' is a standard mantra for unproven remedies used off label. Dose schedules for drugs that have a good evidence base have been worked out.
2. The suggestion that fillers may be responsible for side effects that may make it look as if the drug is not helpful when it is if taken without these fillers.
3. The suggestion that LDN is not working through its normal pharmacology (which sounds psychiatric) but through some speculated action on immunity (which sounds better but nobody has shown it to be relevant).
4. The suggestion that the benefit of LDN is due to a different isomer. Until we actually know there is a benefit this must also be speculation.
On the other hand features of this sort can also go with drugs that don't work terribly well and have lots of side effects (hence the need to go low and slow) but do have proven benefits in formal trials (even if maybe overegged and without enough emphasis on side effects).
I think methotrexate (MTX) for RA is a good example. You are supposed to start low and increase to maximum safe or tolerated level. Side effects have been linked to oral usage and some physicians have claimed it should be used IM to get higher doses in. Nobody knows how MTX works in RA and the dose is lower than in cancer. It may be something to do with a specific effect on adenosine pathways in inflammation but I don't think it has ever been sorted.
MTX produces a demonstrable improvement in a good proportion of RA patients but rheumatologists often forget that the remission rate is something like 5%, whereas the previous generation of drugs - gold and penicillamine - probably produced remission in as much as 20% of cases. For gold this could be permanent. The reason for gold and penicillamine going out of fashion was that their side effects were life threatening and unpredictable.
I am cutting this into two posts.
Jonathan Edwards
Senior Member (Voting Rights)
I have come around to thinking that a high quality LDN trial for ME must be done. I don't feel I should give a running commentary on the Working Group discussions but I get the feeling this is widely agreed. In fact it may be a fait accompli in that physicians are already planning LDN studies (including Dr Younger and Dr Nacul in Long Covid). The remit of the working group is to try to get proposals like this past the final hurdles - especially funding. I think that might work.
What concerns me a bit is that the existence of MTX in RA led to what seemed to me to be a degree of complacency amongst physicians 'we have a treatment, so why try to cure it?'. (Has a familiar BPS ring?) If LDN is approved for ME will it be handed out and patients discharged? Probably not, but there is little doubt that it only rarely has (if it does) a big impact.
What attracts me is the idea that the justification for doing a good trial of LDN in ME provides an opportunity to confront the logistic problems of trials in ME and produce a model that can be repeated for new things coming along. Fluge and Mella managed to do a good phase III trial of rituximab using fatigue scores because rituximab can be reasonably reliably blinded. Drugs that need titrating because of common side effects, like LDN, pose a much bigger problem because blinding will be very tricky. The Younger FM study reported that patients could not tell if they were on test or placebo but the adverse event rate on test was way higher (which confirms that this is a real problem). Hard to interpret that I think. (Any thoughts on that - I may have misinterpreted?)
So we need an objective outcome. I think that has to be some form of actimetry over a long period, but the actimeter output has to be fed into some validated algorithm that extracts a single index of relevant activity and I don't think we are there yet.
What concerns me a bit is that the existence of MTX in RA led to what seemed to me to be a degree of complacency amongst physicians 'we have a treatment, so why try to cure it?'. (Has a familiar BPS ring?) If LDN is approved for ME will it be handed out and patients discharged? Probably not, but there is little doubt that it only rarely has (if it does) a big impact.
What attracts me is the idea that the justification for doing a good trial of LDN in ME provides an opportunity to confront the logistic problems of trials in ME and produce a model that can be repeated for new things coming along. Fluge and Mella managed to do a good phase III trial of rituximab using fatigue scores because rituximab can be reasonably reliably blinded. Drugs that need titrating because of common side effects, like LDN, pose a much bigger problem because blinding will be very tricky. The Younger FM study reported that patients could not tell if they were on test or placebo but the adverse event rate on test was way higher (which confirms that this is a real problem). Hard to interpret that I think. (Any thoughts on that - I may have misinterpreted?)
So we need an objective outcome. I think that has to be some form of actimetry over a long period, but the actimeter output has to be fed into some validated algorithm that extracts a single index of relevant activity and I don't think we are there yet.
Would doing a crossover trial help, with 6 months on LDN and 6 months on placebo? It could be blinded, I think, but I agree there needs to be an objective outcome measure if there are likely to be side effects messing up the blinding.
Given that most accounts I have read of people claiming some improvement with LDN cite reduced pain and/or increased cognitive function, I'm not sure actimetry on its own is adequate.
How about looking at a single figure extracted from actimetry by averaging daily steps each month, and seeing whether there is a consistent increase of, say at least 50% in average step count for the months while on LDN once the optimal dose is reached, compared with the months on placebo.
And, for each patient depending whether they report pain or Cognitive dysfunction as troubling them most, combining that with either:
a predetermined improvement in subjective pain score recorded weekly on a simple 0 to 10 visual analogue scale (say, mean 3 points improvement being clinically significant)
and/or a congitive test score from a regular weekly cognitive test or subjective congnitive function visual analogue scale.
Or do all 3 with everyone, and report both separately and as a combined outcome.
I think with something like ME/CFS that fluctuates significantly for at least some patients, and the problems with blinding, we need really noticeable differences that are sustained over a period of months before they are classed as clinically significant.
The advantage of recording steps daily throughout the trial rather than just at the beginning and end is that it captures PEM eplsodes as a major drop in step count, so if someone is pushing themselves more because they hope the treatment is working, over 6 months that is likely to lead to crashes, which will lower the average score.
Given that most accounts I have read of people claiming some improvement with LDN cite reduced pain and/or increased cognitive function, I'm not sure actimetry on its own is adequate.
How about looking at a single figure extracted from actimetry by averaging daily steps each month, and seeing whether there is a consistent increase of, say at least 50% in average step count for the months while on LDN once the optimal dose is reached, compared with the months on placebo.
And, for each patient depending whether they report pain or Cognitive dysfunction as troubling them most, combining that with either:
a predetermined improvement in subjective pain score recorded weekly on a simple 0 to 10 visual analogue scale (say, mean 3 points improvement being clinically significant)
and/or a congitive test score from a regular weekly cognitive test or subjective congnitive function visual analogue scale.
Or do all 3 with everyone, and report both separately and as a combined outcome.
I think with something like ME/CFS that fluctuates significantly for at least some patients, and the problems with blinding, we need really noticeable differences that are sustained over a period of months before they are classed as clinically significant.
The advantage of recording steps daily throughout the trial rather than just at the beginning and end is that it captures PEM eplsodes as a major drop in step count, so if someone is pushing themselves more because they hope the treatment is working, over 6 months that is likely to lead to crashes, which will lower the average score.
JemPD
Senior Member (Voting Rights)
daily step count is good.
But the trouble with it is that life isnt consistent. So weeks where my step count is low may be because i had to do a lot more cognitive work, so for example a wk where i had to fill in forms/sort out banking/read or type more than usual/spend more time online,time talking on phone, will mean i have no strength left to walk around the house.
Days i have to spend more mental energy i have to sacrifice physical activity, & vice versa. For example if things are happening & i need to be talking on the phone, or reading messages, i will have to do it all lying down & greatly minimise trips to kitchen etc, so my steps are hugely reduced.
On the other hand it if things are very calm - no problems with the house/benefits/relationships/phone calls with doctors/nothing i need to research in order to make a sensible purchase or decision about something, then I will be able to devote more of my energy to physical things, like having a bath or leaving the house & therefore step count will go up.
Some of that will be mitigated by a long term recording, but not that much.
For example the last 3 mnths i have had to spend the majority of my time on cognitive tasks, the 2mnths before I had to spend more of my energy budget on physical activity & sacrifice things like reading/talking etc, the activity was fuelled by adrenaline so those months would show me doing a lot more steps, but my symptoms were actually worse.
So many confounding factors. I think participants whose lives allow no routine whatsoever (like me) would make poor candidates tbh. People whose lives will allow them to control their activities to consistently avoid PEM may be better candidates.
In addition
All that waffle is to say that a huge hurdle to objective measures of activity with actometers is that they cant measure when you're concentrating/expending mental energy. Because the expenditure of mental enegry has a massive impact on how much phycical activity/steps can be done, which would have nothing to do with any drug you were taking but could easily appear as if it did.
And another problem is trying to find an objective measure of cognitive function that's results arent wrecked by when its performed. My carer has said its like working with 2 different people... someone educated/intelligent & intellectually bright, and also someone who has either significant dementia/learning disability or is drugged/stoned out of their mind and cant understand a sentence asking if they want coffee or water to drink.... Often going from one to the other in space of 20mins. That would confound any objective test depending on when i did it.
The only way i can think of getting around that is maybe a very short task such as number sequence recall, writing 3 simple words or some other thing, which is done every day on an app at the same time of day. Not the same time by the clock, but eg as soon as you wake up, or last thing before sleep, i think that may give a more consistent/ reliable result because then although it will fluctuate, you wont get people doing this 'oh i'm feeling not too bad i'll do that cognitive test now' or 'i want to show how bad it is so i'll wait till i'm utterly exhusted/in PEM', & then mixing that up wk by wk.
But the trouble with it is that life isnt consistent. So weeks where my step count is low may be because i had to do a lot more cognitive work, so for example a wk where i had to fill in forms/sort out banking/read or type more than usual/spend more time online,time talking on phone, will mean i have no strength left to walk around the house.
Days i have to spend more mental energy i have to sacrifice physical activity, & vice versa. For example if things are happening & i need to be talking on the phone, or reading messages, i will have to do it all lying down & greatly minimise trips to kitchen etc, so my steps are hugely reduced.
On the other hand it if things are very calm - no problems with the house/benefits/relationships/phone calls with doctors/nothing i need to research in order to make a sensible purchase or decision about something, then I will be able to devote more of my energy to physical things, like having a bath or leaving the house & therefore step count will go up.
Some of that will be mitigated by a long term recording, but not that much.
For example the last 3 mnths i have had to spend the majority of my time on cognitive tasks, the 2mnths before I had to spend more of my energy budget on physical activity & sacrifice things like reading/talking etc, the activity was fuelled by adrenaline so those months would show me doing a lot more steps, but my symptoms were actually worse.
So many confounding factors. I think participants whose lives allow no routine whatsoever (like me) would make poor candidates tbh. People whose lives will allow them to control their activities to consistently avoid PEM may be better candidates.
In addition
I think i mentioned earlier the weekly test score worries me, becaue it would be radically altered by when the person chooses to do the test. And then if you say do it every wk at a specific time, performance will depend on how much i did the day before, and in the hours previous, either mental or physical energy expenditure. I dont see how an objective measure would be much more reliable than a subjective one, but for very different reasons.
All that waffle is to say that a huge hurdle to objective measures of activity with actometers is that they cant measure when you're concentrating/expending mental energy. Because the expenditure of mental enegry has a massive impact on how much phycical activity/steps can be done, which would have nothing to do with any drug you were taking but could easily appear as if it did.
And another problem is trying to find an objective measure of cognitive function that's results arent wrecked by when its performed. My carer has said its like working with 2 different people... someone educated/intelligent & intellectually bright, and also someone who has either significant dementia/learning disability or is drugged/stoned out of their mind and cant understand a sentence asking if they want coffee or water to drink.... Often going from one to the other in space of 20mins. That would confound any objective test depending on when i did it.
The only way i can think of getting around that is maybe a very short task such as number sequence recall, writing 3 simple words or some other thing, which is done every day on an app at the same time of day. Not the same time by the clock, but eg as soon as you wake up, or last thing before sleep, i think that may give a more consistent/ reliable result because then although it will fluctuate, you wont get people doing this 'oh i'm feeling not too bad i'll do that cognitive test now' or 'i want to show how bad it is so i'll wait till i'm utterly exhusted/in PEM', & then mixing that up wk by wk.
Last edited:
Jonathan Edwards
Senior Member (Voting Rights)
Good points, Trish.
I have never been terribly attracted to crossover trials and tried to think why. (Presumably they may give more data from fewer subjects - each giving outcome for two or more trial arms.)
I think the problem is in knowing whether being on LDN first arm and placebo second is the same as vice versa. What if 6 months of LDN sets you up for a good run for a year? I guess my objection to crossover is that it introduces more unknowns.
The point about pain is crucial. If we want to see if LDN reduces pain we need a strictly blinded trial with a pain scale outcome. Cut and dried. Pain could also be included as a secondary measure in a trial with actimetry as primary outcome measure of 'total impact on ME'. But the pain result would have to be repeated if the primary measure came out negative and the secondary measure positive (at least if only without Bonferroni). This is where being picky about power is critical. Planning for more than one trial may be the right approach.
What I had not said is that actimetry looks the best option if the trial is designed to test the impact of LDN on ME as a whole. There seems a good consensus that the primary aim for patients for an ME drug is being able to do more of actitivites that make life worth living. Whether through helping pain or cognition or whatever, that means that a validated actimetry index should reflect the aim. Not everyone may be able to do more, even if feeling better, but with trials you allow for that. If a drug is capable of reaching the primary aim for some it should show statistically. Outcome measures do not need to pick up every case of improvement. It is more important that they are simple and robust. Actimetry is criticised for missing things or being misleading but if, as you say, it is done over extended periods, and validated against diaries in a calibration study, then it ought to pick up a drug benefit. As we have discussed, scepticism about actimetry may partly reflect the fact that so far nothing has shown a useful benefit on it - simply because nothing so far has been good enough.
The bit about optimal dose of LDN being reached begs the question of how you do that without losing blinding - I think we have covered that. It makes the case for an objective measure.
I have never been terribly attracted to crossover trials and tried to think why. (Presumably they may give more data from fewer subjects - each giving outcome for two or more trial arms.)
I think the problem is in knowing whether being on LDN first arm and placebo second is the same as vice versa. What if 6 months of LDN sets you up for a good run for a year? I guess my objection to crossover is that it introduces more unknowns.
The point about pain is crucial. If we want to see if LDN reduces pain we need a strictly blinded trial with a pain scale outcome. Cut and dried. Pain could also be included as a secondary measure in a trial with actimetry as primary outcome measure of 'total impact on ME'. But the pain result would have to be repeated if the primary measure came out negative and the secondary measure positive (at least if only without Bonferroni). This is where being picky about power is critical. Planning for more than one trial may be the right approach.
What I had not said is that actimetry looks the best option if the trial is designed to test the impact of LDN on ME as a whole. There seems a good consensus that the primary aim for patients for an ME drug is being able to do more of actitivites that make life worth living. Whether through helping pain or cognition or whatever, that means that a validated actimetry index should reflect the aim. Not everyone may be able to do more, even if feeling better, but with trials you allow for that. If a drug is capable of reaching the primary aim for some it should show statistically. Outcome measures do not need to pick up every case of improvement. It is more important that they are simple and robust. Actimetry is criticised for missing things or being misleading but if, as you say, it is done over extended periods, and validated against diaries in a calibration study, then it ought to pick up a drug benefit. As we have discussed, scepticism about actimetry may partly reflect the fact that so far nothing has shown a useful benefit on it - simply because nothing so far has been good enough.
The bit about optimal dose of LDN being reached begs the question of how you do that without losing blinding - I think we have covered that. It makes the case for an objective measure.
Jonathan Edwards
Senior Member (Voting Rights)
Thanks JemPD.
I think I would enlarge on the point made to Trish, that inconsistency needn't be seen as an insuperable problem. In most biomedical science we assume massive variation within what we are measuring. If an intervention is useful you still expect to see a shift. A smear of results from 2 to 7 becomes a smear from 3 to 12. As long as the statistics say that there is no reasonable chance that in fact nothing happened you have a result. That is how we know there is global warming. You couldn't possibly tell that just looking at one day each year, but on average days are hotter and that will still apply to the earth as a whole even if the Gulf Stream reverses and England becomes like Patagonia.
I think that's the point, Jem, over a period of 6 months or more for a largish group of patients, any natural fluctuations should not be a problem, as it's likely there will be such fluctuations in the placebo group as well as the treatment group.
I take your point about the combined effect on activity capacity of both physical and cognitive activity, but I think if a drug really worked in any meaningful way for ME/CFS it should raise the overall activity capacity, so even if we had greater demands on us to do cognitive activity for a while, that shouldn't knock back the physical capacity so much. So over 6 months there should still be an upward trend in physical activity if the drug is getting to the core of ME/CFS limitation on overall activity capacity.
I take your point about the combined effect on activity capacity of both physical and cognitive activity, but I think if a drug really worked in any meaningful way for ME/CFS it should raise the overall activity capacity, so even if we had greater demands on us to do cognitive activity for a while, that shouldn't knock back the physical capacity so much. So over 6 months there should still be an upward trend in physical activity if the drug is getting to the core of ME/CFS limitation on overall activity capacity.
Jonathan Edwards
Senior Member (Voting Rights)
As a rule yes, JemPD. If 20% of subjects leave their Fitbits in a drawer for the last 3 months it doesn't matter as long as they really don't know what treatment they were on and so on average it is the same for both groups.