Smartphone Data May Predict Rheumatoid Arthritis Flares, reports study
Usually, rheumatoid arthritis (RA) patients may recognize flares only after symptoms significantly worsen. A study of nearly 300 patients has revealed that early changes in mobility and texting behavior, captured via smartphones, were strongly associated with impending flares. These subtle behavioral shifts may allow earlier detection and intervention, potentially improving flare management before severe symptoms develop. The study was published in the journal ACR Open Rheumatology by Elizabeth M. and colleagues.
RA is a condition that has periods of active and inactive phases with frequent episodes of flare-ups that exacerbate symptoms of pain, function, and quality of life. The early detection of flare-ups in RA patients is a problem that needs innovative patient-centered monitoring strategies. The current study had three major aims. The first aim was to determine and analyze flare-ups and their severity in RA patients over a period of time. The second was to investigate whether changes in patient-reported outcomes could predict flare-ups in RA patients. The third was to analyze inactive patient smartphone use as an indicator of RA flare-ups.
This study also involved 292 RA adult participants who were retrieved from the FORWARD Databank. The study involved PRO evaluations in two separate phases. In the conditional phase, PRO flare queries were administered depending on predetermined changes in PRO scores. On the other hand, flare evaluations were carried out at fixed bi-weekly time intervals in the fixed phase. Passive Smartphone data were also consistently collected in association with the PRO scores. These parameters included the radius of movement, reaction time, and text message details. RA flares were also evaluated as binary variables. Statistical analyses used logistic generalized estimating equation models, multivariate backward selection, and k statistics for agreement.
Key findings
The trial enrolled 292 RA patients. In the conditional phase, during 441 days, patients experienced at least one flare in 71% of cases, averaging 2.9 flares per patient.
In the fixed phase, with a total of 172 days, the flare was experienced in 76% of patients, averaging 3.7 flares per patient.
The mobility radius was a factor against flare during the conditional phase (OR=0.88).
In the fixed phase, high patient global assessment scores (OR=1.25) and high scores for pain assessment were predictive of flare episodes (OR=1.30).
Higher mobility, longer text, and longer reaction time were related to a few flare episodes, whereas short text was related to high flare episodes.
There were fewer flare episodes during the summer months.
The combination of patient-reported data with passive smartphone measures has identified new and important correlations between these measures and flare events in patients with rheumatoid arthritis. Such findings support and suggest that, with further investigation, proactive and more effective management of rheumatoid arthritis could be made possible via personalized digital tools.
Reference:
Mollard, E., Pedro, S., & Michaud, K. (2026). Associations between smartphone-derived behavioral data and rheumatoid arthritis flares. ACR Open Rheumatology, 8(1), e70162. https://doi.org/10.1002/acr2.70162
