Dear Anna,

I found some papers which deal with this subject. In addition, I found some informations about the ideal place to put an accelerometer using a GT3X. Throughout the activity protocol two GT3X+ units were worn, one on the right wrist (W, using a Velcro wrist strap positioned over the dorsal aspect of the wrist midway between the radial and ulnar Styloid processes) and one on the right upper arm (UA, using a small elastic belt positioned on the lateral surface of the arm midway between the acromion process and lateral epicondyle of the humerus). Two GENEActiv accelerometers were also worn; one distal to the GT3X+ on the right W and one on the posterior aspect of the midpoint on the right UA, securely fixed to the skin over the triceps brachii using a 10 cm² patch of tape (Hyperfix self-adhesive dressing retention tape, Smith & Nephew Healthcare Ltd., UK).

I hope this helps.

http://www.nature.com/sc/journal/v54/n2/full/sc201533a.html

Dear Marcelo thank you so much for the answer. 

So you would recommend putting two accelerometers at one time? We want to measure everyday activity so people will be wearing the device for 7 days. I am only wondering how comfortable it will be if you put two of them. 

Thank you one more time and those articles will for sure be very helpful!

Hi

Marcelo has provided plenty of useful reading but I can provide one more, where they used to activPAL

http://www.nature.com/sc/journal/v49/n3/full/sc2010126a.html

Hope its useful

Thank you very much. I was also wondering about dominant and non-dominant arm. In some studies they used non-dominant in some they just put it on right arm without saying if it was dominant or non-dominant. Do you have any experience with that?

Hi Anna

We are currently writing a review on activity monitor placement to measure wheelchair physical activity, their outcomes and validity. There is quite a lot of literature now available with devices positioned on the wrist(s), arm and on the wheelchair. Validity of the actigraph GT3X+ has been published a number of times using different protocols. It had poor validity during slower propulsion speeds and during other tasks (activities of daily living) when positioned on the wrist and the upper arm. This is a similar issue for all body worn devices. There is also a major issue with using energy expenditure as an outcome and I wouldnt advise this. I would advise, however, duration of movement / PA which is more accurate, for both body and wheelchair mounted devices, distance and speed are also much more accurately measured using a number of devices such as the tri-axial activPAL, VitaMove and gyroscope. Or if appropriate in your sample population you may find it suitable to position a device both on the wheelchair and the body?

Hope this helps!

Thank you very much Elaine. You are right there is a lot of literature. Most of the studies used the devices to measure particular movements or tasks. What we want to do is to measure 7 days and see what was the activity of the person. That is why we are looking for the most comfortable option. We also did the literature search but did not find the satysfying answer to the question. 

All those things you are writing are for sure very helpful! Thank you one more time!

That would be really interesting Anna. We have some unpublished work on physical activity in people with spinal cord injury. We found on average people propelled their wheelchair in the community for approximately 1 hour per day, travelling an average of 1.5km/day. The majority of PA events were of short duration at  slow speeds. 

Good luck with your planned work, I look forward to hearing about it!

Thank you Elaine. We already did some pilot study with 15 people. We put the accelerometer on the dominant wrist. But we now have some problems with analyzing the data. Are you also using Actigraph? 

Good luck with the review. I look forward to reading that!

Is it a power chair? If it's not, then why not use a cycle computer? Your outcome would then be distance.

Maia Phillips Smith manual wheelchair. We are more interested in the activity of the person than the distance. There are many activities that are performed when the wheelchair is not moving

Is it the type of activity that an accelerometer would capture well? Accelerometry doesn't capture weightlifting very well (see https://www.nature.com/articles/s41598-018-33288-z ) or other low-acceleration sports.

We do not want to measure sport activities but everyday activities