What a Parkinson’s Researcher Learned From Tracking Her Own Disease

Sara Riggare has been living with Parkinson’s symptoms for over decades. At 13, she started experiencing stiffness and rigidity in her wrists and ankles, problems with balance, and difficulties with making fine motor movements to move her feet to follow the beat at a concert. 

While Parkinson’s usually affects older individuals, a small number, 0.25 percent, are diagnosed with a rare juvenile form before age 20. But despite having these symptoms, Riggare didn’t receive her diagnosis until about two decades later. 

She pivoted, using the skills she learned from training as an engineer to study health informatics and later completed a PhD focused on her own experience tracking and managing her symptoms and medications. Now, a researcher in the Participatory e-Health and Health Data research group at Uppsala University, she continues to study patient-led research, self-tracking, and self-care in Parkinson’s. 

Now with artificial intelligence (AI) chatbots like Claude, ChatGPT, and Gemini, always within reach, she’s researched how people with Parkinson’s are using these tools. “You can’t put the genie back in the bottle,” Riggare told Being Patient. “It’s here, and we have to live with it.”    

Riggare and her research group are hosting a virtual seminar on Aug.t 26 about how people can use AI to help manage Parkinson’s.

Tracking symptoms before AI 

After her diagnosis, Riggare used a physical calendar to track how medications affected her symptoms. Parkinson’s is caused by the death of neurons in the brain that produce the neurotransmitter dopamine, which is important for refining motor movements. Many of the symptomatic treatments focus on activating these dopamine pathways, but their effectiveness and side effects depend on medication timing and other factors. 

To optimize her treatment, Riggare brought the information she was tracking back to her neurologist. When she was taking medications, she noticed the effects wore off a little by the afternoon. She went to her doctor and asked him to change the dosing from one pill in the morning to two pills with smaller doses split across the day. “I did get a more even effect throughout the day as a result,” she said. 

Though many people are using wearable watches, fitness trackers, and rings to track their health, Riggare doesn’t, calling them “mostly useless” for Parkinson’s. However, she found a use for her Apple Watch. Her phone ringing could freeze her gait, her feet immobile, . Now she uses her watch to alert her of notifications on her wrist, silencing her phone and reducing the risk of falls. 

For Riggare, tracking is goal-oriented rather than constant. She tracks symptoms or dosage, for example, to find the optimal timing or dose for her medication, but once that’s set she doesn’t see a purpose in continuing to do it. “I don’t want to use my time tracking,” she said. I want to use my time living my life.” 

How people with Parkinson’s are using AI

People with Parkinson’s aren’t necessarily using AI to track their symptoms. In her recent survey of 149 participants, people used these tools in three different ways. 

Some respondents used them for informational purposes, for example, to learn more about certain Parkinson’s medications or side effects. 

Others used it to help interpret information. Some people asked whether the symptom they were experiencing was Parkinson’s or something else.

They also fed the chatbot information about their medication and symptom timing, to organize it ahead of a doctor’s appointment. 

Not everyone in the survey used AI. While some people didn’t know how, others were concerned about accuracy or privacy, and some wanted to abstain from it altogether. 

Riggare advised people not to use AI chatbots to make decisions but rather to translate information. Ask it to explain studies or side effects in plain language but “never take it as the truth,” she said. Since AI chatbots can be inaccurate and tend to agree with their users, she prompts the chatbot to find errors in its responses and challenges it.

Do they help? “The short answer is we don’t know whether it’s helpful or not,” she said. More research is needed to determine whether these tools are helpful and in what context. 

Regardless, today, doctors — among whom 80 percent are using some form of chatbot in their regular medical practice — need to adapt to the reality that their patients are using these tools too. Riggare said, “They can’t keep pretending that this doesn’t exist.”