Theranica's AI Achieves 91% Precision on Next-Day Migraine Likelihood in Largest Study of Its Kind

Peer-reviewed study of 53,000 Nerivio users finds a patient's own headache pattern over the past 30 days, not the prodromal symptoms clinicians have watched for decades, is the strongest warning sign of tomorrow's migraine attack

BRIDGEWATER, N.J. and CAESAREA, Israel, Sept. 23, 2026 /PRNewswire/ -- Theranica, a neuromodulation therapeutics company redefining migraine care through its proprietary Remote Electrical Neuromodulation (REN) technology, today announced peer-reviewed study results validating Your Day Ahead, the company's artificial intelligence (AI)-powered feature, already incorporated in the Nerivio® app, which forecasts a patient's next-day migraine risk.

The peer-reviewed study, Machine Learning Assessment of Next-Day Migraine Likelihood Using Data From 53,000 App Users Living With Migraine, found that the machine learning (ML) model developed by the company forecasted next-day migraine at 91.2% precision, with area under the curve (AUC, a common metric for evaluating prediction algorithms) of 0.893,  and that the strongest indicator was a patient's own headache pattern over the preceding 30 days, not the prodromal symptoms clinicians have relied on for decades.

Your Day Ahead is powered by ML, the branch of AI that finds patterns in data, trained on real-world, patient-reported information from the Nerivio app. The results, published in Neurology® Open Access, are drawn from the largest study of its kind: 53,065 real-world users of the FDA-cleared Nerivio wearable.

Your Day Ahead is intended to help patients plan for an anticipated migraine attack; it does not diagnose migraine, nor does it recommend any changes in prescribed treatment.

Migraine, a chronic neurological disorder, is among the world's most disabling diseases. According to the World Health Organization, migraine is the third leading cause of years lived with disability worldwide. The burden is not limited to the attack itself: between one attack and the next, people with migraine often experience stress and anxiety in anticipation of the next one, and anxious behaviors such as avoidance and catastrophizing occur at high rates, increasing the risk that migraine attacks become more frequent and contributing to higher disability and treatment costs. Anticipating those attacks has been studied for decades, but a recent systematic review, prior to the Theranica study, concluded that existing models still required improvement before they could deliver high forecasting accuracy.

The study set out to determine whether the risk of a next-day migraine attack could be identified from the kind of data patients already record. Analyzing 770,473 daily reports from 53,065 people who used the Nerivio app between January 2020 and July 2025 – the largest prospective dataset of patients with migraine assembled to date, and the longest pre-attack time frame examined at 30 days – researchers trained seven machine learning algorithms on diary entries, pre-treatment questionnaires, demographic data and location-based weather conditions. The best-performing model, a customized XGBoost classifier, achieved 91.2% precision, 81.0% accuracy, 80.0% sensitivity, 83.0% specificity and an AUC of 0.893, with performance holding across female and male patients (p = 0.42) and across adults and adolescents (p = 0.28). That performance surpasses any diary-based model the authors reviewed, leading them to conclude that accurate next-day assessment of migraine risk is achievable at a level not previously reported. The authors note that the forecasted outcome included mild as well as moderate and severe headache pain.

Knowing that migraine risk may be elevated can help a person prepare. The study's authors state that forecasting at this resolution could reduce anticipatory anxiety and interictal burden — the toll migraine takes between attacks — and allow patients to take actionable lifestyle strategies such as sleep, nutrition and schedule adjustments, and improved adherence to their prescribed treatment. Because the model's relatively low false-positive rate supports confidence in its next-day forecasts, the authors note that it may help patients plan and prepare, while the feature itself does not direct treatment decisions or recommend preemptive medication use.

The most useful warning signal turned out not to be the one the headache field has watched for. Prodromal symptoms, which are the premonitory cues relied on for decades, contributed just 11.1% of the model's performance. The strongest signal by far was the patient's own rolling average headache severity over the preceding 30 days, and the relative contribution of features calculated over the 30 days preceding an attack was 56.3%, leading the authors to conclude that long-term headache patterns may warrant greater emphasis than prodromal cues in clinical care.

"For decades the field has looked to the day or the hours immediately before an attack, the premonitory (prodrome) phase, for warning of what is coming. What this analysis suggests, in the largest dataset reported to date, is that the strongest predictive signal isn't in that narrow pre-attack window. Instead, a patient's own recorded pattern of headache severity over the preceding month turns out to be the most informative signal in the model," said Chia-Chun Chiang, MD, Associate Professor of Neurology, Headache Specialist, Mayo Clinic, Rochester, Minnesota. "That's a meaningful shift in how we think about forecasting migraine risk, and it matters for patients: consistent, longitudinal patient-reported data isn't just a record of what's happened — it may be a window into what's likely to come next. That said, this cohort came from users of a single therapeutic device and skewed toward frequent, treatment-engaged patients, so the findings need to be confirmed in broader migraine populations."

For Theranica, the finding completes a broader picture. Nerivio is already indicated for acute and/or preventive treatment of migraine with or without aura in patients 8 years of age or older. Your Day Ahead adds next-day migraine likelihood information within the Nerivio app, without requiring an additional drug or adding cost for patients.

"These results demonstrate how AI can use longitudinal patient-reported data to provide information about next-day migraine likelihood. By incorporating Your Day Ahead into the app, we aim to give patients greater visibility into their own patterns and help them plan for the day ahead while continuing to follow their prescribed treatment plan," said Ronen Jashek, Co-Founder and Chief Business Officer of Theranica and a co-author of the study. "This is what redefining the standard of care looks like in practice. Every treatment and every diary entry deepen the science, and Nerivio keeps getting better after it's prescribed. A pill cannot do that."

Nerivio is covered by major health insurers reaching more than 158 million lives across the United States and has been used in more than 2 million treatments, supported by more than 35 original peer-reviewed clinical studies. Backed by leading healthcare investors including New Rhein Healthcare, Vivo Capital, aMoon and Lightspeed Venture Partners, Theranica is advancing REN beyond migraine as a scalable, mechanism-driven therapeutics platform for a broader range of chronic pain conditions.

About Nerivio®

Nerivio® is the first and only FDA-cleared, prescription Remote Electrical Neuromodulation (REN) therapy indicated for both acute and preventive treatment of migraine, establishing a new standard of care. Supported by more than 35 original peer-reviewed published clinical studies, including multiple randomized controlled trials as well as large-scale real-world evidence studies, Nerivio has demonstrated consistent efficacy, a favorable safety profile, and effectiveness in clinical practice.

Worn on the upper arm, Nerivio activates the body's endogenous pain-inhibitory pathways through Conditioned Pain Modulation (CPM), delivering targeted, drug-free relief. Cleared for patients aged 8 and older, it delivers patented electrical pulses without needles, systemic side effects, or drug interactions. Controlled through a smartphone app, Nerivio enables targeted, on-demand treatment aligned with how migraine occurs in real life — providing relief during attacks and reducing migraine frequency with preventive use.

About Theranica

Theranica is a growth-stage neuromodulation therapeutics company redefining the standard of care for migraine and other complex chronic neurological disorders through its proprietary Remote Electrical Neuromodulation (REN) technology. Grounded in the science of Conditioned Pain Modulation (CPM), REN selectively engages the body's endogenous descending pain-inhibitory pathways to deliver clinically validated, drug-free relief.

Theranica is advancing REN beyond migraine as a scalable mechanism of action, developing a pipeline of effective, reimbursed, non-pharmacologic treatment options for patients across a broader range of chronic pain conditions.

Media Contact
Kris Barr
Executive Director, Public Relations & Investor Relations
Theranica
[email protected]
+1 410.919.7889

SOURCE Theranica