Apple’s quest to put a smartwatch on every wrist continues with its latest health research: a new study showing the Apple Watch can be used to monitor symptoms of Parkinson’s disease.
Researchers at Apple, working with specialists who treat Parkinson’s, designed a system that uses the Apple Watch to detect the motor symptoms that are a hallmark of the neurological disease. By monitoring resting tremors and other involuntary movements, the researchers were able to identify the characteristic “on” and “off” patterns of medication’s effects. Their findings were published Wednesday in Science Translational Medicine.
The research could be a boon to both clinical trials and care for the millions living globally with Parkinson’s. If further developed, the researchers’ system could be used to capture round-the-clock objective measurements of symptoms with the Apple Watch. Specialists often rely on infrequent clinical visits and self-reporting to monitor the disease’s progression and the impacts of medicine. While there are specialized devices in the market that can do such monitoring, there are advantages to using a gadget people recognize and feel comfortable around.
“Having the ability to take a commonly available device that’s already out there like an Apple Watch … and be able to do this type of monitoring is really nice because you’ll be able to give the clinicians who are caring for these people in their home a much clearer idea of what’s going on throughout the cycle of the day,” said Michael Okun, executive director at the Fixel Institute for Neurological Diseases at the University of Florida and national medical advisor for the Parkinson’s Foundation. Okun was not involved in the new research.
The new system, called the Motor Fluctuations Monitor for Parkinson’s disease (MM4PD), uses the Apple Watch’s accelerometer and gyroscope data to detect the presence of resting tremor or dyskinesia. Resting tremors, which can affect the hands, legs, and other parts of the body, are a common symptom of Parkinson’s. Dyskinesia, another type of involuntary movement, is a frequent side effect of medication used to treat the disease.
The algorithms underlying the model were developed using data from a pilot study with 118 people in which researchers matched subject’s smartwatch data to a scoring system called MDS-UPDRS Part III, the gold standard by which motor symptoms of Parkinson’s are measured.
Three movement disorder specialists rated video recordings of subjects, which were time-aligned with the smartwatch data, against the standard scoring. The pilot study was extended for one week of regular use to assess how the algorithm worked in real-world scenarios.
Following the pilot study, the researchers tracked 225 people with Parkinson’s disease for up to six months. Then, they used that measurement data to create symptom profiles. Those were evaluated by clinicians to see if MM4PD could be used to identify symptom response to treatment changes and more broadly whether the system might be used as a decision support tool.
The authors say that the measurements helped spot symptoms missed in regular care and identified changes after subjects underwent surgery for deep brain stimulation. The paper also suggests the tool helped pinpoint people who slipped on medication adherence, as well as cases in which a person might benefit from a modified medication regimen.
The researchers also looked at data from a control group of 171 older subjects without Parkinson’s, who were monitored for up to 12 months. In the pilot study, 36 participants were women and 82 were men. In the longitudinal study, 69 participants were women and 156 were men. The researchers did not report on the race or ethnicity of the subjects.
Doctors tasked with treating people with Parkinson’s are confronted with both many variables and imperfect information. Symptoms of the disease vary from person to person and can change over time as the disease progresses. Treatment plans are complex and highly personalized, and medical appointments provide only narrow snapshots of symptoms.
The researchers say using a device like the Apple Watch could help fill in some of the gaps in observation so that specialists can better tailor treatments to symptoms and outside experts agree.
“The hope is that some of this technological innovation will help in the areas of better tracking medications, better tracking symptoms, and having more real-time control over this because it’s complicated,” said Okun. “And you’re hoping that these systems are going to make it easier both for persons with diseases like Parkinson’s, but also for practitioners managing them.”
Continuous monitoring of symptoms could be useful for clinical trials as well, with the added benefit that it could help enroll more subjects who may have trouble with frequent visits to clinics.
Claudia Revilla, who was diagnosed with Parkinson’s in 2010 and serves on the Michael J. Fox Foundation’s Patient Council, told STAT that based on her experience with doctors and clinical trials, having objective measurements from a smartwatch could be “a big advantage.” She lives in Peoria, Ill., and drives three hours a few times a year to Chicago see her movement disorders specialist.
“We as patients can forget or can avoid discussing certain things,” she said. When she gets to appointments, she sometimes initially responds to doctor’s questions by saying that everything’s fine. “I don’t have a smartwatch for that purpose, but I have my husband,” He’s quick to clarify the situation: “Oh, no, that’s not true. You should see her in the morning,” he’ll say.
“The smartwatch is going to be… an additional witness, more evidence of what’s really going on with you,” she said.
Apple, which declined to make its researchers available for an interview, hasn’t expressed plans to create any kind of Parkinson’s monitoring system within the Apple Watch as it has for certain heart conditions. The paper’s authors also noted that FDA clearance may be necessary before their system can be used for clinical trials.
If it is to be used more broadly, though, the Apple researchers caution that there are limitations to MM4PD, including that the system focuses on only two motor symptoms. The watch also focuses on the wrist as a single observation point, which means it may not capture symptoms elsewhere on the body.
On May 11-12, STAT will bring you an inside look at the convergence of biology and technology. From Silicon Valley giants to upstarts emerging from stealth, you’ll hear from the voices changing the future of health. Get your early-bird ticket now.
K. Ray Chaudhuri, a professor of neurology who treats people with Parkinson’s at King’s College, said that while it’s notable that the study addresses the “on” and “off” fluctuations experienced by people with Parkinson’s, not everyone experiences such fluctuations, and among those that do, it doesn’t always impact movement.
“So you have motor fluctuations, you have non-motor fluctuations, and the two usually go together but the pattern is very different,” he told STAT. A small percentage of people experience only non-motor fluctuations, according to Chaudhuri. “So what this device or this technique is validating is really the amount of fluctuation when this is defined by tremor when off [and] dyskinesia when on,” he added.
The paper acknowledges that MM4PD cannot monitor non-motor symptoms that can impact quality of life.
Chaudhuri’s clinic uses the PKG watch from Global Kinetics to monitor a patient’s motor symptoms, which also allows them to track sleep trends. And he’s studied using digital sensors to monitor bowel sounds, gait, and eating behaviors in people with Parkinson’s.
While devices like the Apple Watch may be widely distributed in the United States, Chaudhuri noted that Parkinson’s is a global problem with rates expected to increase significantly in coming years. “The largest impact of this will be felt in many countries where the standard of living is not very high. So that applicability and availability of a smartwatch or smartphone is very restricted and limited,” he said.
It’s also important to design digital systems so that they can be used by people with tricky symptoms. “You have to be conscious about how much you’re asking them to do [in terms of] monitoring, and particularly when they’re on and off, you know, even [using] a smartphone may not be very easy.”