Tweet
Dear Maxim
Does your study make the hypothesis that the treatment improves the
condition of the OA and therfore the dogs are able to walk faster and
produce higher PVF?
If the body weight was the only change then normalising the output as you
have done would reveal the same N/kg ratio output. If the dog also walk
faster, ie higher velocity of CoM, then the ratio should change. So for
instance if at week 1 the dog weighed 30kg and had a PVF of 100N on the
front right leg, this =3.33N/kg. If it then at week 4 weighed 32kg and at
the same velocity the PVF was 106.6N, then the ratio is still 3.33N/kg. If
the dog now also walks faster because of less pain then the PVF might be say
112N and the normalised ratio is now 3.5N/kg.
I assume that the weight gain is due to increased muscle weight, due to
ability to increase exercise, and not fat. If it were fat then there might
be a problem then that the dogs walk more slowly because of increased weight
and also increased joint forces cause the dog to walk more slowly, if the
treatment was not effective, and PVT normalised values might not be very
different even if the pain and function had improved. To find a true
dimensionless normalisation value to reliably indicated how treatment
improved performance might be difficult.
I think you may get a more reliable results if you were able to calculate
the joint forces and if the dogs were able to tolerate higher joint forces
then, regardles of weight and velocity, this might indicate an improved
function and reduction in pain. This approach might be difficult since you
would require kinematic data, which I would imagine is dificult with
animals.
Perhaps it would be safe to assume that higher PVF = higher joint forces and
so regardless of changes in weight or velocity it may also be safe to assume
that higher joint forces = less pain and therfore improved outcome. This may
not be reliable though since changes in gait style, that may occur with
improved function, may change moment arm lengths but have the same PVF for a
higher or lower joint force.
Its a difficult problem ( at my level anyway)
Cheers Dave Smith
----- Original Message -----
From: "maxim moreau"
To:
Sent: Monday, August 21, 2006 3:27 PM
Subject: [BIOMCH-L] ForcePlate problem
> Dear Experts,
>
> I do clinical research in the veterinary field using a ForcePlate set up
> to acquire Ground Reaction Force particularly in dogs afflicted by
> osteoarthritis. In my latest clinical trials, I encountered a problem.
> Let me explain: A group of dogs was followed over time (baseline, Week-4
> and Week-12) to pinpoint the effect of a therapeutic modality. Dogs
> significantly increased their body weight between Week-4 and Week-12,
> while Week-4 and baseline's body weight were similar. As recommended I
> analysed the Peak Vertical Force (primary study outcome) expressed in
> percentage of body weight. This gain in body weight could have biased the
> study and consequently, mask a possible beneficial effect. Does a
> mathematical or statistical approach exist to limit the effect of weight
> increase on PVF values?
>
> Your help and opinion are obviously welcome
> Cordially
> Maxim Moreau
> Université de Montréal, Canada
>
> ---------------------------------------------------------------
> Information about BIOMCH-L: http://www.Biomch-L.org
> Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
> ---------------------------------------------------------------
Does your study make the hypothesis that the treatment improves the
condition of the OA and therfore the dogs are able to walk faster and
produce higher PVF?
If the body weight was the only change then normalising the output as you
have done would reveal the same N/kg ratio output. If the dog also walk
faster, ie higher velocity of CoM, then the ratio should change. So for
instance if at week 1 the dog weighed 30kg and had a PVF of 100N on the
front right leg, this =3.33N/kg. If it then at week 4 weighed 32kg and at
the same velocity the PVF was 106.6N, then the ratio is still 3.33N/kg. If
the dog now also walks faster because of less pain then the PVF might be say
112N and the normalised ratio is now 3.5N/kg.
I assume that the weight gain is due to increased muscle weight, due to
ability to increase exercise, and not fat. If it were fat then there might
be a problem then that the dogs walk more slowly because of increased weight
and also increased joint forces cause the dog to walk more slowly, if the
treatment was not effective, and PVT normalised values might not be very
different even if the pain and function had improved. To find a true
dimensionless normalisation value to reliably indicated how treatment
improved performance might be difficult.
I think you may get a more reliable results if you were able to calculate
the joint forces and if the dogs were able to tolerate higher joint forces
then, regardles of weight and velocity, this might indicate an improved
function and reduction in pain. This approach might be difficult since you
would require kinematic data, which I would imagine is dificult with
animals.
Perhaps it would be safe to assume that higher PVF = higher joint forces and
so regardless of changes in weight or velocity it may also be safe to assume
that higher joint forces = less pain and therfore improved outcome. This may
not be reliable though since changes in gait style, that may occur with
improved function, may change moment arm lengths but have the same PVF for a
higher or lower joint force.
Its a difficult problem ( at my level anyway)
Cheers Dave Smith
----- Original Message -----
From: "maxim moreau"
To:
Sent: Monday, August 21, 2006 3:27 PM
Subject: [BIOMCH-L] ForcePlate problem
> Dear Experts,
>
> I do clinical research in the veterinary field using a ForcePlate set up
> to acquire Ground Reaction Force particularly in dogs afflicted by
> osteoarthritis. In my latest clinical trials, I encountered a problem.
> Let me explain: A group of dogs was followed over time (baseline, Week-4
> and Week-12) to pinpoint the effect of a therapeutic modality. Dogs
> significantly increased their body weight between Week-4 and Week-12,
> while Week-4 and baseline's body weight were similar. As recommended I
> analysed the Peak Vertical Force (primary study outcome) expressed in
> percentage of body weight. This gain in body weight could have biased the
> study and consequently, mask a possible beneficial effect. Does a
> mathematical or statistical approach exist to limit the effect of weight
> increase on PVF values?
>
> Your help and opinion are obviously welcome
> Cordially
> Maxim Moreau
> Université de Montréal, Canada
>
> ---------------------------------------------------------------
> Information about BIOMCH-L: http://www.Biomch-L.org
> Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
> ---------------------------------------------------------------