# Predicting Recurrences in Bipolar Illness (Prompt-BD)

> **NIH NIH R01** · UNIVERSITY OF PITTSBURGH AT PITTSBURGH · 2022 · $745,546

## Abstract

With each recurrence, the prognosis for Bipolar Disorder (BD) worsens and the risk for suicidality and substance
abuse increases, indicating the need to identify and manage the factors associated with recurrence risk. Factors
associated with recurrence risk for the group as a whole have been identified, however, the question of how to
predict recurrence risk for an individual with BD remains unanswered. Since the course of BD is heterogenous,
the ability to predict recurrence risk at the person level would allow treatment to be tailored to the specific
individual. As in other fields of medicine, we developed a risk calculator (RC) that predicts with approximately
80% discrimination the 1-5 year risk of any mood recurrence and the polarity of the recurrence in youths/young
adults with BD. Given that the RC must be externally validated before it can reliably be used in clinical practice,
the Predicting Recurrence of Mood in Patients with BD (PROMPT-BD) study proposes to externally validate the
RC in BD youth/young adults. The existing RC (termed “distal” RC) is valuable, but only predicts long-term (1-5
year) recurrence risk and not proximal risk (i.e.1-4 weeks). Predicting proximal risk could enable prompt
intervention upon detection of “warning” signs, potentially improving the course of BD. To do this, factors
associated with impending recurrences in BD such as sleep-activity rhythms, mobility, and digital social
interactions will be measured continuously in real-time using smartphone passive sensing and subsequently
combined with clinical predictors from the distal RC to further enhance prediction. In contrast to self-reports,
passive sensing can assess continuous changes in behavior over extended periods without requiring participant
input. While actigraphy is the “gold standard” for assessment of sleep-activity rhythms, it is not feasible for
participants to wear a research-grade activity monitor for the duration necessary for risk prediction of mood
recurrences. Thus, we will examine how passive sensing-based assessments of sleep-activity data map onto
actigraphy by collecting actigraphy data for five 2-week periods over follow-up (intake and at each follow-up).
We will also explore if development and sex influence passive sensing measures of interest. To carry out this
proposal, 120 BDI/II individuals (14-25 years old) who are currently in remission will be recruited. Clinical
assessments will be administered at intake and 6,12,18, and 24-months. Sleep-activity rhythms (e.g. phone
usage, accelerometer), mobility (e.g., GPS), and digital social interaction (e.g. phone communication log), will
be continuously evaluated via smartphone passive sensing, yielding over 300 days of data per participant
(>30,000 days combined across the sample). Finally, to more precisely identify the timing of recurrence and to
temporally anchor retrospective mood assessments, participants will complete a brief weekly questionnaire for
mood prompted via text messag...

## Key facts

- **NIH application ID:** 10474530
- **Project number:** 5R01MH126991-02
- **Recipient organization:** UNIVERSITY OF PITTSBURGH AT PITTSBURGH
- **Principal Investigator:** BORIS BIRMAHER
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2022
- **Award amount:** $745,546
- **Award type:** 5
- **Project period:** 2021-09-01 → 2026-06-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10474530

## Citation

> US National Institutes of Health, RePORTER application 10474530, Predicting Recurrences in Bipolar Illness (Prompt-BD) (5R01MH126991-02). Retrieved via AI Analytics 2026-05-21 from https://api.ai-analytics.org/grant/nih/10474530. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*