EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Digital phenotyping for non-communicable diseases: Emerging evidence, challenges, and public health implications

Jayita Pal, Sanjoy Kumar Sadhukhan, Shilpa Mondal

International Journal of Noncommunicable Diseases · 13.11.2026

DOI: 10.4103/jncd.jncd_109_26 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Noncommunicable Diseases, online erschienen am 13.11.2026. Verfasst von Jayita Pal und 2 weiteren Autor:innen. Die Arbeit nennt 64 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Digital phenotyping refers to the longitudinal quantification of behavioral and physiological states using data generated by personal digital devices, particularly smartphones and wearable sensors. Its relevance to noncommunicable diseases lies primarily in capturing real-world behavioral and physiological variation between conventional clinical encounters rather than replacing established diagnostic methods. This narrative review synthesizes evidence on digital phenotyping across mental health disorders, diabetes and metabolic disorders, and cardiovascular diseases, while distinguishing it from digital biomarkers, remote monitoring, digital diagnostics, and mobile health interventions. A structured narrative search was undertaken in PubMed/MEDLINE, Scopus, Web of Science, and Google Scholar for English-language literature published from January 2010 through March 31, 2026. Evidence was synthesized thematically according to digital signal, application, study design, outcome, validation, and translational maturity. Mental health research provides substantial evidence of associations between passively sensed behavioral features and symptoms, with emerging evidence for longitudinal prediction but limited external validation and clinical utility. Diabetes and metabolic applications encompass behavioral and physiological phenotyping, self-monitoring, and multimodal glucose behavioral approaches, but validated individual-level prediction remains limited. Cardiovascular research includes relatively advanced diagnostic applications, particularly wearable-based atrial fibrillation detection, although these findings should not be generalized to broader cardiovascular prediction. Across domains, major challenges include missing and context-dependent data, within-person versus between-person inference, sensor and platform heterogeneity, overfitting, limited temporal and external validation, model drift, privacy, governance, and digital inequity. Overall, digital phenotyping is currently better supported for longitudinal behavioral and physiological characterization than as a validated standalone approach for disease prediction, diagnosis, prognosis, or clinical decision support.

Zitieren

APA

Pal, J., Sadhukhan, S. K., & Mondal, S. (2026). Digital phenotyping for non-communicable diseases: Emerging evidence, challenges, and public health implications. International Journal of Noncommunicable Diseases. https://doi.org/10.4103/jncd.jncd_109_26

Vancouver

Pal J, Sadhukhan SK, Mondal S. Digital phenotyping for non-communicable diseases: Emerging evidence, challenges, and public health implications. International Journal of Noncommunicable Diseases. 2026. doi:10.4103/jncd.jncd_109_26

BibTeX
@article{Pal2026,
  author = {Pal, Jayita and Sadhukhan, Sanjoy Kumar and Mondal, Shilpa},
  title = {Digital phenotyping for non-communicable diseases: Emerging evidence, challenges, and public health implications},
  journal = {International Journal of Noncommunicable Diseases},
  year = {2026},
  doi = {10.4103/jncd.jncd_109_26}
}
Kostenlos starten →