# Smartphone-based digital phenotyping to detect high-risk affect states in body dysmorphic disorder

> **NIH NIH K23** · MASSACHUSETTS GENERAL HOSPITAL · 2020 · $183,304

## Abstract

Body dysmorphic disorder (BDD) is associated with extremely high risk for suicide attempts (22-28%) and
substance use disorders (49%), underscoring the critical importance of risk detection in BDD. Negative affect
states - particularly anxiety and shame - are well-documented risk factors for suicide and substance use in
BDD, offering clear targets for risk detection and intervention. This K23 aims to develop and validate
unobtrusive, time-sensitive, and ecologically valid measures of anxiety, shame, and general negative affect
states in BDD, using smartphone-based digital phenotyping. Passive (i.e., unobtrusive) smartphone
measurement of negative affect states will be based on GPS, accelerometer, and communication logs, used to
detect behavioral features of anxiety (avoidance, rituals), shame (social withdrawal, isolation), and general
negative affect (aggregated avoidance, rituals, withdrawal, and isolation features). We will collect passive and
active (i.e., ecological momentary assessment [EMA]) smartphone data in 85 adults with BDD and will use
EMA ratings of negative affect as outcomes, to build and validate predictive statistical models from passive
data. We will also test the hypotheses that passive smartphone measures of negative affect states can
significantly predict next-day suicidal ideation and substance use in BDD, above and beyond common clinical
indices of risk. This project synthesizes the Candidate’s expertise in emotion-based risk for suicide in BDD with
her experience conducting smartphone research. Building from this foundation, this K23 will provide critical
new training in key areas to launch the Candidate’s independent research career: (1) digital phenotyping,
including statistical learning and longitudinal analysis; (2) EMA methods; (3) assessment of suicide and
substance use; (4) career development, including R01 writing; and (5) ethics of technology-based suicide and
substance use research. Training goals will be accomplished with stellar mentorship and institutional support at
Massachusetts General Hospital and Harvard Medical School. Dr. Sabine Wilhelm, a leader in BDD and
clinical research, will serve as the primary mentor. Dr. Jukka-Pekka Onnela, an expert in digital phenotyping
and its statistical approaches, and Dr. Michael Armey, an expert in EMA research of emotions and suicide, will
serve as co-mentors. Complementary guidance in EMA and substance use will be provided by the advisory
team: Drs. Bettina Hoeppner and A. Eden Evins. In line with NIMH Strategic Objective 2, this K23 will yield
scalable, unobtrusive tools to detect acute, modifiable risk factors for suicide and substance use in a high-risk
population. Moreover, negative affect states are transdiagnostic risk factors. As a next step to this proof-of-
concept K23, the Candidate will apply for an R01 to further validate passive mobile detection of negative affect
states and their ability to predict risk transdiagnostically. This program of research can ...

## Key facts

- **NIH application ID:** 10018106
- **Project number:** 5K23MH119372-02
- **Recipient organization:** MASSACHUSETTS GENERAL HOSPITAL
- **Principal Investigator:** Hilary Weingarden
- **Activity code:** K23 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2020
- **Award amount:** $183,304
- **Award type:** 5
- **Project period:** 2019-09-13 → 2024-08-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10018106

## Citation

> US National Institutes of Health, RePORTER application 10018106, Smartphone-based digital phenotyping to detect high-risk affect states in body dysmorphic disorder (5K23MH119372-02). Retrieved via AI Analytics 2026-05-25 from https://api.ai-analytics.org/grant/nih/10018106. Licensed CC0.

---

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