What ME research has used actimeters/Fitbits?

[Keela Too]

One complication with the Fitbit might be that if I were a study subject, then I could see my own data as it happens, and could potentially alter my behaviour to help it show what I thought the study wanted....

Of course just knowing it was being measured could mean a subject could still alter their behaviour, but they would have less knowledge of what they were actually clocking up.

I know I keep an eye on my Fitbit during the day a bit, and if I think I've been a bit busy I look and see if I really have and will moderate what I do based on that knowledge. Of course that is how I am using it... and it helps. I really do have fewer proper crashes now...

 

[user9876]

There seem to be a variety of devices that seem to do slightly more than a fit bit in particular there are ones that measure sleep patterns (don't know how) and blood oxygen levels and heart rate. The combinations of different variables may be more robust. One I saw claimed a battery life of 1 year and that it is waterproof.

A couple of thoughts in terms of analysis:
Overall activity levels - In which case what time period would you average activity over (day, week, month). We could look for periodic behavior in smaller time period data. Trends on a windowed time series may then reflect major health trends. I'm sure there will be a number of trend analysis algorithms for processing time series data looking for either gradual improvements (e.g. could simply fit a regression line over a larger window) or looking for sudden dips (I think some sort of auto-correlation function may help here but its a long time since I did signal processing). Any trend data would need to take account of periodic activity patterns. The question is what would you want to get out I assume something like long term variability and frequency of crashes/sudden improvements.

Response to additional activity (i.e. after a peak of activity how long do various readings take to settle down back to normal I'm assuming things like sleep patterns may be interesting here). Activity could simply be defined in terms of a raise in the normal activity based on a sliding time window and above a certain threshold. In a trial it might be interesting to look at response patterns to activity but the frequency of such responses would not be a good indicator since PWME will often pace to avoid such patterns. But where there is data the profile and any change in profile may be interesting if it suggests something about PEM and the course of the disease.
If several variables are measured then looking for cross correlation between them may tell us something and is quite easy to do. It may be necessary to do this on a time delay basis if one lags another.

I tend to think that to work out an analysis plan involves looking at some data.

 

[Min]

From PACE questions and answers page FAQ2:

Actigraphy is a measure of physical activity, measured by a wrist watch sized accelerometer, worn around the ankle continuously for a week. Before we started the trial, we were advised that the number and scope of the outcome measures were too great and that it might reduce the proportion of participants making it through to the end of the trial. Actigraphy was the obvious measure to reject because of its burden in time and effort required by participants. The patient charity advising us agreed that this would be sensible.

The only charity involved in PACE was Action for ME.

 

[Keela Too]

Fitbit can measure sleep patterns based on movements at night. I used it for a while, but no longer monitor that angle.

I understand a new Fitbit due out soon will also monitor HR.

 

[MeSci]

"Push and crash" for the Brits, but I think the Dutch psychobabblers would call you "relatively active" if you're doing too much and crashing. Once you start pacing, you become "passive"

I've always heard it called 'boom and bust'.

 

[bertiedog]

I had 2 Fitbits but managed to loose both of them. What I found was that it encouraged me to do more at first and wasn't really good for me because I was overdoing it.

Now I have a Garmin Viofit and find that on good days I can just about do 7000 steps but when I am bad I can only manage 4000 or a lot less. This happens when I get a virus which I have at a moment and the number of steps in a day really drops consistently. I feel that I learned from my bad experiences with the Fitbit where I kept pushing myself and for the past 2 months using the Garmin wristband I don't do this, I am much better at accepting how things are on any given day.

I don't both monitoring my sleep with this device but use a programme that is on my phone but they definitely aren't always right as sometimes it will show my sleep quality at only 67% but in fact I have slept all night.

I do think it would be very helpful if lots of sufferers with ME wore activity monitors as they would be giving out lots of data that couldn't be argued with. With the Viofit it shows a green line for the strength of the activity being "very active" but I have never achieved that in 2 months".

Pam

 

[A.B.]

I think the belief behind boom and bust is that the patient is healthy, but does so much more than a normal person that it causes health problems, and lacks insight into their self-harming behaviour. Then eventually the patients get so scared of exercise they do nothing at all. At least that is the impression one gets from that CFS treatment training video for UK general practitioners which portraits the typical ME/CFS patient as someone who hits the pub every night and complains about fatigue as result.

This has nothing to do with ME/CFS and I think we shouldn't use that expression because it might give the impression that patients are agreeing with this misanthropic view.

 

[lansbergen]

When I saw they found that pattern I thought : Och they can do something right. But as always the conclusions were ridiculous.

Their ridiculous conclusion are no reason for me to deny what I did.

I think the belief behind boom and bust is that the patient is healthy, but does so much more than a normal person that it causes health problems, and lacks insight into their self-harming behaviour. Then eventually the patients get so scared of exercise they do nothing at all. At least that is the impression one gets from that CFS treatment training video for UK general practitioners which portraits the typical ME/CFS patient as someone who hits the pub every night and complains about fatigue as result.

This has nothing to do with ME/CFS and I think we shouldn't use that expression because it might give the impression that patients are agreeing with this misanthropic view.

 

[Keela Too]

Psychobabblers can portray boom and bust as they like, but it can be a very real feature of trying to do what we can... Boom doesn't need to mean pushing oneself or mad partying, only doing more than your body's tolerance... which is very easy if it is something fun as a little bit of adrenalin kicks in.

The payback - or bust comes a couple of days later... I've been a little opinionated on that too:
http://www.sallyjustme.blogspot.co.uk/2014/09/play-up-and-lay-up.html

 

[Keela Too]

@bertiedog Why do you think the Fitbit encouraged you to do more but a different step counter did not?

Also your numbers demonstrate how individual each person's numbers are.... I'm operating on almost exactly 1/10th of your range... I know there are others who would be on about 1/10th of my range too.

I think it takes nearly a month of averaging daily steps to get a fair account of what level an individual is operating on.

 

[Jonathan Edwards]

From PACE questions and answers page FAQ2:
'Actigraphy is a measure of physical activity, measured by a wrist watch sized accelerometer, worn around the ankle continuously for a week. Before we started the trial, we were advised that the number and scope of the outcome measures were too great and that it might reduce the proportion of participants making it through to the end of the trial. Actigraphy was the obvious measure to reject because of its burden in time and effort required by participants. The patient charity advising us agreed that this would be sensible.'

The only charity involved in PACE was Action for ME.

This is a very peculiar statement - to suggest that the decision was justified by a patient charity 'advising us'. I could understand a patient charity 'funding us' agreeing, but why would a charity have a useful input in terms of advice about trial logitics and scientific merit? They might do, but as it stands it doe s not make much sense. Most of the patients on this list seem to think wearing an actometer would not be a problem.

 

[Jonathan Edwards]

One complication with the Fitbit might be that if I were a study subject, then I could see my own data as it happens, and could potentially alter my behaviour to help it show what I thought the study wanted....

Of course just knowing it was being measured could mean a subject could still alter their behaviour, but they would have less knowledge of what they were actually clocking up.

I know I keep an eye on my Fitbit during the day a bit, and if I think I've been a bit busy I look and see if I really have and will moderate what I do based on that knowledge. Of course that is how I am using it... and it helps. I really do have fewer proper crashes now...

I think that is a very important observation Keela Too. It seems that as actometer measurements went down you felt that you were happier with your situation because of not inducing crashes. There might be lots of different ways of looking at that but it highlights the fact that what the actometer measures is not actually what matters to the patient.

 

[Sasha]

I think that is a very important observation Keela Too. It seems that as actometer measurements went down you felt that you were happier with your situation because of not inducing crashes. There might be lots of different ways of looking at that but it highlights the fact that what the actometer measures is not actually what matters to the patient.

Maybe this is an argument for blinding patients to the data produced by the actimeter.

 

[Sasha]

Incidentally, one issue around the Fitbit specifically is that it's marketed as a fitness-training device and I imagine that the instructions and the way it works reflect that. Any actimeter for an ME study would need all that stuff to be absent - otherwise you'd be accidentally priming people with the idea that they're supposed to be grading their activity up and getting fit.

 

[Keela Too]

Yes - in my view the patient would need to be blinded to the data.
Although that too presents problems - for example I noted that riding my mobility scooter across rough ground whilst wearing the Fitbit clocked up a huge number of steps... well at least in my context of only walking 500 - 600 steps most days, the fact that a scoot round the fields could add another 800 is rather significant. Yet I know this is relaxing for me not stressful, as my heart rate is nice and relaxed.

On an interesting note I did once try to very gradually increase my steps by adding 2% to my total each month.... it didn't work. My ceiling is not for budging up no matter how slowly I attempt an increase. I stay cautious for very good reason!

 

[Keela Too]

Another area I clocked up about 50 steps was in the car - I finally worked out it was a rumble strip coming up to a round about.... but only in hubby's car, in another car the same strip had no effect! LOL

 

[Sasha]

Yes - in my view the patient would need to be blinded to the data.
Although that too presents problems - for example I noted that riding my mobility scooter across rough ground whilst wearing the Fitbit clocked up a huge number of steps... well at least in my context of only walking 500 - 600 steps most days, the fact that a scoot round the fields could add another 800 is rather significant. Yet I know this is relaxing for me not stressful, as my heart rate is nice and relaxed.

That's useful info if anyone is designing a study. Actimeters are basically measuring the acceleration and deceleration of the bit of the body that they're attached to (I think it's done by a little ball that can move in its housing). So it confuses steps with other movements that rattle the ball inside the device.

 

[bertiedog]

Why do you think the Fitbit encouraged you to do more but a different step counter did not?

I tried to learn from what I did wrong last time and possibly because the Viofit is just a watch it doesn't seem to influence me the way the Fitbit did. That might not make a lot of sense but it has worked for me.

I am sure there would be a really large range of steps done in a day and how active each person is. Mine is pretty consistent until these viruses come along which unfortunately this time of the year they regularly do! (Have had the Garmin Viofit since early September).

Pam

 

[Keela Too]

I take the Fitbit off and clip it to the gate when I go out on the scooter .... of course THEN I'm clocking up less "active time" - which can be just moving about, not necessarily walking I think....

 

[Keela Too]

Indeed @bertiedog and the important thing (for us as patients, rather than as researchers trying to collate data) is to find something that "works for us" !!