PDMeasure: collecting our own data
 

Do you ever wonder whether turmeric slows down the progress of your PD? Or whether exercise has an impact? Or whether something about your environment gave you PD in the first place and whether it is now affecting the rate of your deterioration? Or wonder what type of music will best relieve, if only briefly, some of your symptoms? Or wonder whether you're in a cluster of PwP?

I'd like to know the answers to those and a thousand other questions. Big pharma and big science provide some of the answers and they will provide more answers over time. But I'd prefer not to wait: seven years into PD, I don't have the time left to wait 10 years for something to get to market and even then possibly to find it either too expensive to use or that it only gives a limited marginal benefit.

So what do we do? We let big pharma and big science do "big" things. We wish them well. They have a lot of good, bright people. We should support them if we can. But, in my opinion, we shouldn't stop there and wait for the cure to be given to us. We should do our own science, "small" science. "Small" science can be big if enough people do it.

We, PwP, could be the biggest resource to finding therapies for PD.

At its simplest "small" science involves measuring, aggregating and analysing our symptoms. We would benefit greatly if, rather then saying "this therapy makes me less stiff", we could say "it made me walk 11% faster". And, rather than having just a few anocdotal reports, we could say "this therapy led to an average 27% improvement on a group of 500".

How do we accelerate progress? We need the infrastructure to collect, store and analyse data.

As a very small step to this end, I am building a computer based system, PDMeasure, that integrates PD survey data with tests that measure PD symptoms. The data can be aggregated for each person, and aggregated over time. All data is stored anonymously. PDMeasure will be freely available online.

Data on its own has little use. We want people to analyse it. To make this as easy as possible, the data is open to anyone to access.

A prototype is ready. I am looking for people to test it, before going live in, perhaps, 2 weeks. If you have 10 minutes to spare to test the system, please PM me.

Thanks.

John

First, my thanks go to the testers: your comments have led to a better program.

If you wish to take part in the survey (both PwP and non-PwP are welcome) go to:

http://www.parkinsonsmeasurement.org/PDMeasure

There you will see "Getting started in just 5 minutes" and an explanation of how the project runs.

Going beyond that you can repeat the test as often as you want.

If you have any problems, please get in touch.

What are the objectives of this survey? To measure the extent to which the tap test results:
- are affected by learning, time since last medication, disease duration and gender;
- distinguish between PwP and healthy people.

Next week I'll begin a series of surveys that will look at the role of the environment in PD.

If there's any subject you would like polling, please get in touch.

John

john,
biggest problem i see is that unless you are checking ip addresses, anyone can post multiple times and skew the results. and you can't even verify if someone has pd, the severity of their pd nor rule out the placebo affect. can't verify anyone took tumeric, they might be taking something else or a much weaker form, totally unintentionally.
as far as exercise, i always feel better after exercise, but i've also taken more sinemet so i can exercise better, so what is the affect from?

you can rail against big pharma all you want but i see a lot of small organizations doing small trials on exercise at least and some supplements and MJFF is funding a lot of small projects. big pharma certainly isn't looking at forced exercise but the researcher who initially did the phase1 is doing a phase2 with $500,000 of funding and i think is looking at the slowing progression aspect.

secondly, there's no way to prove you accurately collected/analyzed your results. as far as slowing progression, how can you even think of trying to prove that just based on a honor system of reporting?

statistics from an anonymous population are very hard to verify. i don't know about anyone else, but my pd symptoms sometimes change 50% from day to day, hour to hour, i wouldn't even attempt to report that a supplement or exercise or any other activity made me feel better until after at least a month of taking it.

Patient Centered Research
 

John,

MJFF is working on this type of project. Can you PM me to discuss further?

Debi

It's great to hear from Debi Brooks that MJFF are working on this type of project. I am an admirer of their work. The advantage that they have over me is that they are likely to get far more people involved than I can. So, I've offered both to collaborate with them and to let them use my software.

I'll be talking to one of their people on Monday.

The crucial questions are:

When are they going live?

What role will PwP play? We're a great resource. We know our needs. Our involvement needs to be more than just providing data.

How open is the data? To encourage as many people as possible to provide data it needs to be anonymous. To encourage as many people as possible to research the data, it needs to be open. For instance, every student of statistics in the world should have access to it. Who knows what they might find? One can envisage a time when, with enough data and sufficiently complex models, it is worth distributing the analysis, SETI-like, across the web.

(Incidentally, any software I write on www.parkinsonsmeasurement.org, of which PDMeasure is part, is open. Anyone can use it for free.)

John

... Parkinson's medications survey
 

Do you ever wonder how your meds compare with other people's?

I've written an online survey on Parkinson's medications that allows us to answer exactly that question. It takes just 5 minutes to complete. If you want to take part, please go to:

http://www.parkinsonsmeasurement.org/PDMeasure

If you have not already done so, register first. Otherwise, just follow the link.

Where is all this going?

We now have two surveys and a test. More will be added. The infrastructure will be extended.

The data is in a common format so we can combine the results from one survey with another, and another ..., and we can combine the results from one test with another, or with a survey. As the number of surveys and tests grow the number of possible links grows exponentially. Who knows what we may find by data mining this data?

But, to have any statistical significance we need a large sample size. We need people to participate. We need YOU!

John

First results: the learning effect
 

What's been happening at PDMeasure?

Since it was launched last October, PDMeasure has gradually been receiving data from a wide range of people.

Attachment 7600

Currently:
96 people have completed the basic data survey;
51 people have completed the medications survey;
340 completions of the side-to-side tap test have been recorded.

My thanks go to everyone who has donated data, and especially to those 5 or 6 people who have provided a long sequence of test data. (Everything is anonymous, so I can't thank them personally.)

Most of the data came as a result of PDMeasure featuring In a MJFF blog. My thanks go to them. They will be launching something similar later this year. The quality of their brand name is such that I expect them to get far more data.

In the meantime, PDMeasure requires more data. A pressing need is for people to repeat the side-to-side tap test. To join PDMeasure, or to donate more data or to access the data, go to:

http://www.parkinsonsmeasurement.org/PDMeasure/

Attachment 7601

A key requirement is for PDMeasure to have an online, objective symptom measurement test. We use the side-to-side tap test. This needs to be validated.

Attachment 7602

We have insufficient data to provide statistical significance, but the data so far shows, as we would expect:

- older people have, on average, lower scores than younger people,

- people with more years from diagnosis have, on average, lower scores than the newly diagnosed,

- it will vary from person to person, and according to the drugs they take, but after a person goes beyond a time, of, say, 100 min after their last medication, its effect will wear off, leading to lower scores.

It appears that people are learning to get higher scores in the test. One person reported that they could improve their score by moving their body as well as their arms and hands as they go to and thro between the q and p keys.

Combined left and right hand averages for doing the test for the 1st through 5th time are 38.1, 40.7, 38.4, 39.2, 39.6.

Why does this matter?

The improvement due to the "learning effect" could be credited wrongly either to the efficacy of the therapy under test or to the placebo effect.

To avoid this, any use of such tools should work through the learning period before their use in trials.

Finally, you can do more than just donating data, though if that's all you want to do, I'm still very grateful. You can help analyse the data. The computer code is open. Use it as you want. The data is open: you can use either the raw data or, simpler, use the joined data at the end of:
http://www.parkinsonsmeasurement.org...Statistics.php
It's still under construction, but what is there is good.

The data is in csv form, and can be put directly into Excel or, if you have some programming experience, I recommend R. Mine it.

MORE DATA + MORE ANALYSIS => MORE PROGRESS

John

Latest results from PDMeasure
 

My thanks goes to the people who have provided data to PDMeasure.

Some stats:
109 people have completed the basic data survey;
58 people have completed the medications survey;
546 runs of the side-to-side tap test have been completed;

New registrations are rare. The numbers taking the side-to-side tap test appear healthy, but in fact the majority of test completions come from just two people.

New registrations are welcome. Returning participants are welcome. I would be particularly grateful if people would repeat the test frequently, so that for each person we could estimate progression.

Here's a couple of interesting graphs showing data from me.

Attachment 8009

The first shows the side to side tap scores changes as the number of minutes since the last dose increases. As one would expect immediately after a dose scores are low, they rapidly build up as, in my case, the Stalevo takes effect; before declining, until they settle down at a constant level provided by endogenous dopamine, and the long acting rasagiline and Requip XL that I take.

Attachment 8010

The second graph shows the variation in my side-to-side tap test score over the course of a year, or so. The trend lines show that performance has improved over that time. As reported in a previous post, for people new to the test there is a definite learning effect, certainly over the course of the first 10 times that a person takes the test. But, having done the test so many times over the years, I think the learning effect is low for me. My hope is that the test, in itself, is mildly therapeutic. Far fetched? Possibly. But, by analogy, consider a fitness regime being measured by weekly timed 1000m runs.

MJFF plan to start a PwP data collection scheme later this year. I think it will be a great success. As I understand it, they have no immediate plans for running objective online tests, such as we use here. So, there will still be a need for PDMeasure.

To register on PDMeasure, to do the test or to browse the data go to:

http://www.parkinsonsmeasurement.org...sure/index.php

John

Don't worry, john"."."
 

You WILL live to get a cell implant that WILL almost eliminate pd symptoms. It WILL be available before you are like many of us who are too "long in the tooth" with this disease. I had similar hopes 17 years ago, but it was not to be, but you, and all those who are at the beginning of this terrible journey, as YPWP, you guys and gals have REAL reason for hope, and soon I might add. How soon? I give it within the proverbial "5 more years" when a working safe cell implant strategy is initiated, on the market.

An update on PDMeasure. It's still chugging away:
135 registrations;
70 medication responses;
780 tap tests taken.

Latest results from my own data are:

Attachment 9265

Attachment 9266

My side-to-side tap test results are gradually improving. I put this down to long-term learning. This raises an interesting question: can you beat Parkinson's, if only locally, by exercising based on making rapid, repetitive movements?

John

Reading this John I immediately thought of the recent post regarding boxing as a therapy which also involves rapid repetitive movements.

Nigel

Can you beat PD by exercising
 

using repetitive movements, I believe the answer is yes. I believe you can force the brain to adapt and create new pathways by repetition.

I believe this applies more for younger rather than older PWP and is the route of the different types of PD suffered by younger and older PWP.

I myself recently completed "Insanity". This is a very hard exercise program, lots of repetitive movements, I was improving every day.

Glad someone raised this point, I will speak to my Neuro soon and ask his opinion.

Neil.

I continue to take the tests reported above.

The graph below shows the tap test scores for the 89 tests done by me at least 10 hours after the last medication was taken. This time should be sufficient to eliminate most of the "noise" caused by the difference, from test to test, of the time since I last took my drugs.

Attachment 9467

The results described in this thread relies on measuring Parkinson's using a computer keyboard. Giancardo et al. [1] have taken the work further, making it more accurate. Their system "measures the key hold times (the time required to press and release a key) during the normal use of a computer without any change in hardware and converts it to a PD motor index ... This new approach discriminated early PD groups from controls with an AUC = 0.81 (n = 42/43; mean age = 59.0/60.1; women = 43%/60%;PD/controls). The performance was comparable or better than two other quantitative motor performance tests used clinically: alternating finger tapping (AUC = 0.75) and single key tapping (AUC = 0.61)."

Their "alternating finger tapping test" is very similar to the side-to-side (QP) tap test that can be run directly from my web site. There is a stand-alone version that does not store the results at:
Parkinson's Disease Measurement: PwP, surveys, trials, analysis
And a version that stores the results in a database, which you need to register with first, at:
Parkinson's Measurement > PDMeasure: home page

Reference

[1] Giancardo, L. et al. Computer keyboard interaction as an indicator of early Parkinson's disease. Sci. Rep. 6, 34468; doi: 10.1038/srep34468 (2016).
http://www.nature.com/articles/srep34468

John

Two years later and I'm still measuring my PD.

Attachment 10184

The generally improving trend of the period 7.5 to 12 years after diagnosis seems to have gone in reverse in the last 6 months.

In my view, the side by side tap test is a good way to measure short term changes, such as monitoring the immediate effects of a change of drug regimen.

However, regarding the use of the side by side tap test to measure long term changes in PD, I'm not so sure. My subjective view is that my PD has worsened more than these tests suggest. What seems to happen is that a new symptom develops (e.g. drooling), which gets worse for a while, before stabilizing, only to be followed by a yet another new symptom (e.g. dystonia).

You can run these tests or access the data online, just go to my website:

Parkinson's Disease Measurement: PwP, surveys, trials, analysis

John

John,
My route seems similar 11 years after dx in which I have been reasonably stable. I have been hit by the extra saliva over the past 3 months and then it levels off. Also, in the mornings my legs tremor/shake rapidly and my toe dystonia is more uncomfortable as well. I had two botox sessions 18 months ago for my legs/feet/toes which helped, but I decided not to put that poison in my body anymore. Who knows how safe that stuff will turn out to be in time. Hang in folks...

Best,
Eric

An update of my side-to-side tap test results. They now cover the period from just after 7 years to almost 14 years post diagnosis. The regression line is still going upwards. This suggests that I am improving. However, this is probably due to a training effect. My last few readings have been below average, possibly showing a progression in my PD. This is mirrored in my subjective assessment of my symptoms.

Attachment 10240

The decision only to show data which have been collected 600 minutes after the last dose, which in my regimen is Stalevo, L/C/E, was chosen to be almost certainly indicative of an "off" state. There will also be some benefit from the 8mg dose of ropinirole CR and the 1mg dose of rasagilene taken approximately 1200 minutes prior to the test.

John

john this may interest you as UPDRS is in need of a rebirth.
Regards