New Study Aims to Determine Whether iPhone and Apple Watch Can Detect Early Signs of Dementia

MacRumors

macrumors bot
Apr 12, 2001
7,187
8,363
0
19
www.macrumors.com



Apple has partnered with pharmaceutical company Eli Lilly and health startup Evidation to determine whether data collected from the iPhone and Apple Watch can be used to detect early signs of dementia.

A research paper published this week and shared by CNBC lists researchers from Eli Lilly, Apple, and Evidation Health. The paper, called "Developing Measures of Cognitive Impairment in the Real World from Consumer-Grade Multimodal Sensor Streams," explores whether sensor data and activity info from smart watch devices can be mined for "physiological and behavior signatures of cognitive impairment."

The ubiquity and remarkable technological progress of wearable consumer devices and mobile-computing platforms (smart phone, smart watch, tablet), along with the multitude of sensor modalities available, have enabled continuous monitoring of patients and their daily activities. Such rich, longitudinal information can be mined for physiological and behavioral signatures of cognitive impairment and provide new avenues for detecting MCI in a timely and cost-effective manner.

In this work, we present a platform for remote and unobtrusive monitoring of symptoms related to cognitive impairment using several consumer-grade smart devices. We demonstrate how the platform has been used to collect a total of 16TB of data during the Lilly Exploratory Digital Assessment Study, a 12-week feasibility study which monitored 31 people with cognitive impairment and 82 without cognitive impairment in free living conditions. We describe how careful data unification, time-alignment, and imputation techniques can handle missing data rates inherent in real-world settings and ultimately show utility of these disparate data in differentiating symptomatics from healthy controls based on features computed purely from device data.
According to the abstract, 31 people with cognitive impairment and 82 without cognitive impairment were monitored over a 12-week period, with 16TB of data collected. The study claims that the data was able to be used to differentiate people with early signs of cognitive impairment from those who were healthy.

People who had symptoms of cognitive decline typed more slowly, typed less regularly, relied more heavily on support apps, and sent fewer text messages. The study did not reach long-term conclusions as more analysis is needed.

In a statement to CNBC, Evidation co-founder Christine Lemke said that data collected from the iPhone, Apple Watch, and Beddit sleep monitors was used for the study. Apple acquired the company behind the Beddit sleep monitor back in 2017.
"With this research, we looked at how everyday behavior data, such as those captured by iPhones, Apple Watches, and Beddit sleep monitors, may be effective in differentiating between individuals with mild cognitive impairment and early Alzheimer's disease, and those without symptoms."
Early detection of dementia is important because an early diagnosis can allow for better management of symptoms and quality of life improvements even though the progression of the disease can't be stopped. According to the World Health Organization, 50 million people around the world have dementia, with close to 10 million new cases surfacing every year.

Article Link: New Study Aims to Determine Whether iPhone and Apple Watch Can Detect Early Signs of Dementia
 

Relentless Power

macrumors Penryn
Jul 12, 2016
29,956
30,249
0
Crazy. I mean, to see the advancements that technology is making in the wearables sector is incredible (Which also explains it’s growth). EKG, fall detection, wheelchair capabilities on the Apple Watch and possibly detecting dementia, this is what’s making the future for wearables a ‘must have’ versus what once started primarily as a notification device.
 

macduke

macrumors G4
Jun 27, 2007
10,549
14,011
0
Central U.S.
Did they control for age? Because if not, congrats, you determined that those people are old, and like 99% of people with dementia are old.

Edit: Ok, this part from the paper gives me pause:

Even so, due to difficulties with recruiting symptomatic participants, the symptomatic cohort was an average of 3 years older than the healthy control cohort.
Ok, it looks like they did match ages, but not genders, and the sample size is kinda small.

In order to verify that the device-derived features were not de- tecting differences in behavior due to normal aging, we selected the nearest age-matched control within the healthy control cohort for every participant in the symptomatic cohort. Doing so produced two age-matched cohorts of 31 symptomatics and 31 healthy controls with an average age difference of less than six months2. We re-ran the full analysis on these age-matched cohorts and report the results in Table 3. After controlling for age via matching, there was a large drop in performance for the demographics only models (AUROC=0.519, AUPRC=0.536). However, device-derived features still show moderate performance, with AUROC decreasing from 0.771 to 0.726 on the full cohort, and AUPRC increasing from 0.628 to 0.758.
Seems like it still correlates, but not as strongly, when controlling for age? Hmmm. Don't have time to read the full thing in-depth but would love to hear from someone who did.
 
Last edited:

mikeheenan

macrumors 6502
Aug 8, 2007
296
4
0
I wonder how accurate this is. The watch is a horrible calculator of steps. I work in retail and it records almost every single arm movement as a step. Heck, I can get 25 steps driving my car at times.
 

now i see it

macrumors 68040
Jan 2, 2002
3,905
7,865
0
They've lost their mind.
(I hope I didn't type that too slowly).
Any old person wearing an A-Watch now will now be under suspicion that they're concerned they're developing dementia.
"Oh he has that dementia surveillance watch on his wrist".
 
Last edited:

Doc C

macrumors regular
Nov 5, 2013
187
107
0
Did they control for age? Because if not, congrats, you determined that those people are old, and like 99% of people with dementia are old.

Edit: Ok, this part from the paper gives me pause:

...

Ok, it looks like they did match ages, but not genders, and the sample size is kinda small.

...

Seems like it still correlates, but not as strongly, when controlling for age? Hmmm. Don't have time to read the full thing in-depth but would love to hear from someone who did.

Unfortunately, this is a problem with a lot of (medical) studies of this size - it can be difficult to build appropriately matched cohorts without having very large sample sizes (thus having random sampling variation minimize some of the variability). Even in larger trials, it isn't unheard of for minor (statistically non-significant) variations between cohorts to account for the difference found in the study, when all of them are considered on an aggregate basis.
[doublepost=1565218040][/doublepost]
I know it already knows I can't spell worth a damn.....
No no - that's a "feature"

like when my son texted that he lost his pen. And autocorrect thought he forgot the "is" at the end of the word.
 

The Cappy

macrumors regular
Nov 9, 2015
189
273
0
Dunwich Fish Market
One way to detect dementia is by counting the number of times an owner uses the 'Find My Phone/Watch' feature every week :)
Crazy. I mean, to see the advancements that technology is making in the wearables sector is incredible (Which also explains it’s growth). EKG, fall detection, wheelchair capabilities on the Apple Watch and possibly detecting dementia, this is what’s making the future for wearables a ‘must have’ versus what once started primarily as a notification device.
Because...? If it said you were going to develop Alzheimer's, we've just ruined your life for no good reason. It's not like we have preventative medicines. The only thing you might do is what you're supposed to do anyway which is to be as healthy as possible in other respects so you are more resilient when the dementia gets its teeth into you.

But what will happen instead is that, armed with this cognitive death sentence, far more people will greet the early dementia with fatalism and surrender. This is why we shouldn't (and generally don't) do DNA tests on family members of Alzheimer's patients to see if they're at high risk.
 

PickUrPoison

macrumors 601
Sep 12, 2017
4,326
4,588
0
Sunnyvale, CA
Boy. Talk about a slippy slope. Ring the right algorithm bell and be in front of the death panel the next day.
Yikes what country do you live in? You might want to consider moving, for your own safety. But I think this will be very useful, especially for those who have a family history of dementia or early-onset dementia. There are also other forms of cognitive impairment that can mimic dementia, which this could possibly detect and lead to effective treatment.
 
Last edited:
  • Like
Reactions: CarlJ and Edsel

robjulo

macrumors 65816
Jul 16, 2010
1,224
1,878
0
You know of a “cure” for dementia. Let the NE Journal of Medicine know. Perhaps you’ll win a Nobel prize.

This would be amazing. Dementia is a horrible horrible disease that takes every last bit of dignity from you. The earlier the detection, the earlier the planning can begin or even cure it.
 

Edsel

macrumors 6502
Mar 18, 2010
340
342
0
Over There
My mother had dementia and it is a weird disease. They can be perfectly normal one minute then think my wife has of 30 years was my girlfriend, my mother thought I was my Father. She stormed away angry that I had a “girlfriend”.

So, as an oldster now, I worry that I might get dementia. There are moments when I find myself thinking about the past. Will I get dementia? I will never know but my family will. Godspeed to the scientists and medical professionals for trying to root out this affliction that effects so many.
 

Suttree

macrumors regular
Jul 21, 2018
231
311
0
I know nothing about medicine and have zero medical training and knowledge, yet even I know this study is bunk. My mom asked me why am I still living at home, depriving this world of my sage knowledge and skillset when I can be goodly remunerated and leave a positive impact rather than merely armchair quarterbacking. But it is obvious this study is plain bunk with only 31 subjects. Common sense. Simple as that.
 

robjulo

macrumors 65816
Jul 16, 2010
1,224
1,878
0
You don’t have to have knowledge in anything to be an expert on macrumors. I’ve seen at least a couple dozen posters on here who are experts in everything from semi conductor engineering to bank finance to marketing to healthcare. Especially so when their master needs to be defended.


I know nothing about medicine and have zero medical training and knowledge, yet even I know this study is bunk. My mom asked me why am I still living at home, depriving this world of my sage knowledge and skillset when I can be goodly remunerated and leave a positive impact rather than merely armchair quarterbacking. But it is obvious this study is plain bunk with only 31 subjects. Common sense. Simple as that.
 
  • Like
Reactions: Shirasaki

steve23094

macrumors 68030
Apr 23, 2013
2,532
1,259
0
I know nothing about medicine and have zero medical training and knowledge
It’s not whether the Apple Watch can detect dementia per se, of course the study is way too limited to show that. Buts it’s an interesting proof of concept. Can a device be leveraged to make a correlation between it’s normal day to day use to show signs of a specific illness? The answer here is ‘maybe’, and it’s worth investigating further.
 
  • Like
Reactions: ipedro

FourDegrees

macrumors member
Jan 25, 2017
65
95
0
"31 people with cognitive impairment and 82 without cognitive impairment were monitored"

"The study did not reach long-term conclusions as more analysis is needed."

it's a start. There's nothing definitive in these results. Rather more of a suggestion of how useful these wrist-worn devices might become.
 

spinedoc77

macrumors G3
Jun 11, 2009
9,338
3,371
0
Fairly amusing that what I think will be found as one of the major causes of dementia is being used to detect dementia. No there isn't any hard research to back that up, just my personal opinion but it is something that doctors talk about and is being researched. I don't mean to single out smartphones though, they fall into a lot of technology which doesn't necessarily have to be new such as television and computers. Smartphones just really seem to have accelerated the process of dopamine overload and addiction as they are always with you.