Smartphone-Based Digital Phenotyping Across Health Conditions: Scoping Review

Background: Smartphone-based digital phenotyping uses built-in sensors and usage patterns to passively capture behavioral and environmental data relevant to health and has been applied extensively in mental health and chronic disease contexts. Objective: This review synthesizes studies that use smartphone-based digital phenotyping, defined as approaches that rely exclusively on onboard smartphone sensors to characterize specific health conditions. To our knowledge, this work provides the most comprehensive cross-condition synthesis of smartphone-based digital phenotyping to date, spanning mental health, physical health, and substance use disorders (SUDs), and highlighting common practices, gaps, and opportunities for future research. Methods: We conducted a scoping review of English-language, peer-reviewed papers published between 2012 and 2025 in Google Scholar, IEEE Xplore, ACM Digital Library, and PubMed using terms such as “mobile sensing” and “digital phenotyping.” Eligible papers used onboard smartphone sensors to assess health and went beyond self-report. Studies that did not rely on smartphone auxiliary sensing modalities or digital phenotyping were excluded. Results: We performed a descriptive synthesis of study characteristics, sensors, and health domains. Of 111 papers identified, 65 met inclusion criteria. Most studies were observational and relied on passive sensing. Sample sizes ranged from fewer than 10 to over 18,000 participants, with a median of 52 (IQR=26‐126). Mental health conditions were most frequently examined, including depression (n=16), bipolar disorder (n=11), stress or anxiety (n=10), and schizophrenia (n=8). Less commonly studied conditions included SUDs (n=7), Parkinson disease (n=4), and sleep apnea (n=2). Sensor streams varied widely and included diverse passive smartphone data sources capturing mobility, communication, device usage, environmental context, and user interaction patterns. Ground-truth measurements most commonly relied on validated clinical scales (eg, Patient Health Questionnaire-9, Young Mania Rating Scale [YMRS], and Pittsburgh Sleep Quality Index; n=41), followed by ecological momentary assessments (n=18), clinician-confirmed diagnoses (n=9), and physiological measures such as polysomnography (n=3). Across studies, recurring methodological limitations included incomplete or inconsistent sensor descriptions, limited reporting of data quality (eg, sampling rates and missingness), and heterogeneous validation practices. These issues limit comparability and reproducibility and underscore the need for clearer reporting standards and greater data availability. Conclusions: This scoping review provides the first comprehensive synthesis of smartphone-only digital phenotyping studies spanning mental health, physical health, and SUDs. Unlike prior reviews, this work maps behavioral associations derived exclusively from smartphone sensors across a broad range of health domains. The primary contribution of this review lies in its consolidation of behavioral associations observed across studies, enabling researchers to correlate new findings to the existing evidence base and identify opportunities for replication, extension, or clinical translation. Collectively, these findings highlight both the promise of smartphone-based digital phenotyping in real-world settings and the need for improved standardization to support translation into clinical and public health applications.