Challenges in Using mHealth Data From Smartphones and Wearable Devices to Predict Depression Symptom Severity: Retrospective Analysis.
behavioral patterns
depression
digital phenotypes
missing data
mobile health
mobile phone
smartphones
wearable devices
Journal
Journal of medical Internet research
ISSN: 1438-8871
Titre abrégé: J Med Internet Res
Pays: Canada
ID NLM: 100959882
Informations de publication
Date de publication:
14 08 2023
14 08 2023
Historique:
received:
21
12
2022
accepted:
23
04
2023
revised:
11
04
2023
medline:
15
8
2023
pubmed:
14
8
2023
entrez:
14
8
2023
Statut:
epublish
Résumé
Major depressive disorder (MDD) affects millions of people worldwide, but timely treatment is not often received owing in part to inaccurate subjective recall and variability in the symptom course. Objective and frequent MDD monitoring can improve subjective recall and help to guide treatment selection. Attempts have been made, with varying degrees of success, to explore the relationship between the measures of depression and passive digital phenotypes (features) extracted from smartphones and wearables devices to remotely and continuously monitor changes in symptomatology. However, a number of challenges exist for the analysis of these data. These include maintaining participant engagement over extended time periods and therefore understanding what constitutes an acceptable threshold of missing data; distinguishing between the cross-sectional and longitudinal relationships for different features to determine their utility in tracking within-individual longitudinal variation or screening individuals at high risk; and understanding the heterogeneity with which depression manifests itself in behavioral patterns quantified by the passive features. We aimed to address these 3 challenges to inform future work in stratified analyses. Using smartphone and wearable data collected from 479 participants with MDD, we extracted 21 features capturing mobility, sleep, and smartphone use. We investigated the impact of the number of days of available data on feature quality using the intraclass correlation coefficient and Bland-Altman analysis. We then examined the nature of the correlation between the 8-item Patient Health Questionnaire (PHQ-8) depression scale (measured every 14 days) and the features using the individual-mean correlation, repeated measures correlation, and linear mixed effects model. Furthermore, we stratified the participants based on their behavioral difference, quantified by the features, between periods of high (depression) and low (no depression) PHQ-8 scores using the Gaussian mixture model. We demonstrated that at least 8 (range 2-12) days were needed for reliable calculation of most of the features in the 14-day time window. We observed that features such as sleep onset time correlated better with PHQ-8 scores cross-sectionally than longitudinally, whereas features such as wakefulness after sleep onset correlated well with PHQ-8 longitudinally but worse cross-sectionally. Finally, we found that participants could be separated into 3 distinct clusters according to their behavioral difference between periods of depression and periods of no depression. This work contributes to our understanding of how these mobile health-derived features are associated with depression symptom severity to inform future work in stratified analyses.
Sections du résumé
BACKGROUND
Major depressive disorder (MDD) affects millions of people worldwide, but timely treatment is not often received owing in part to inaccurate subjective recall and variability in the symptom course. Objective and frequent MDD monitoring can improve subjective recall and help to guide treatment selection. Attempts have been made, with varying degrees of success, to explore the relationship between the measures of depression and passive digital phenotypes (features) extracted from smartphones and wearables devices to remotely and continuously monitor changes in symptomatology. However, a number of challenges exist for the analysis of these data. These include maintaining participant engagement over extended time periods and therefore understanding what constitutes an acceptable threshold of missing data; distinguishing between the cross-sectional and longitudinal relationships for different features to determine their utility in tracking within-individual longitudinal variation or screening individuals at high risk; and understanding the heterogeneity with which depression manifests itself in behavioral patterns quantified by the passive features.
OBJECTIVE
We aimed to address these 3 challenges to inform future work in stratified analyses.
METHODS
Using smartphone and wearable data collected from 479 participants with MDD, we extracted 21 features capturing mobility, sleep, and smartphone use. We investigated the impact of the number of days of available data on feature quality using the intraclass correlation coefficient and Bland-Altman analysis. We then examined the nature of the correlation between the 8-item Patient Health Questionnaire (PHQ-8) depression scale (measured every 14 days) and the features using the individual-mean correlation, repeated measures correlation, and linear mixed effects model. Furthermore, we stratified the participants based on their behavioral difference, quantified by the features, between periods of high (depression) and low (no depression) PHQ-8 scores using the Gaussian mixture model.
RESULTS
We demonstrated that at least 8 (range 2-12) days were needed for reliable calculation of most of the features in the 14-day time window. We observed that features such as sleep onset time correlated better with PHQ-8 scores cross-sectionally than longitudinally, whereas features such as wakefulness after sleep onset correlated well with PHQ-8 longitudinally but worse cross-sectionally. Finally, we found that participants could be separated into 3 distinct clusters according to their behavioral difference between periods of depression and periods of no depression.
CONCLUSIONS
This work contributes to our understanding of how these mobile health-derived features are associated with depression symptom severity to inform future work in stratified analyses.
Identifiants
pubmed: 37578823
pii: v25i1e45233
doi: 10.2196/45233
pmc: PMC10463088
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
e45233Subventions
Organisme : British Heart Foundation
Pays : United Kingdom
Organisme : Department of Health
Pays : United Kingdom
Organisme : Medical Research Council
ID : MR/N013700/1
Pays : United Kingdom
Organisme : Wellcome Trust
Pays : United Kingdom
Organisme : Chief Scientist Office
Pays : United Kingdom
Informations de copyright
©Shaoxiong Sun, Amos A Folarin, Yuezhou Zhang, Nicholas Cummins, Rafael Garcia-Dias, Callum Stewart, Yatharth Ranjan, Zulqarnain Rashid, Pauline Conde, Petroula Laiou, Heet Sankesara, Faith Matcham, Daniel Leightley, Katie M White, Carolin Oetzmann, Alina Ivan, Femke Lamers, Sara Siddi, Sara Simblett, Raluca Nica, Aki Rintala, David C Mohr, Inez Myin-Germeys, Til Wykes, Josep Maria Haro, Brenda W J H Penninx, Srinivasan Vairavan, Vaibhav A Narayan, Peter Annas, Matthew Hotopf, Richard J B Dobson, RADAR-CNS Consortium. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 14.08.2023.
Références
Int J Methods Psychiatr Res. 2003;12(1):22-33
pubmed: 12830307
NPJ Digit Med. 2022 Jan 11;5(1):3
pubmed: 35017634
World Psychiatry. 2013 Jun;12(2):92-8
pubmed: 23737408
Psychol Med. 2004 Aug;34(6):1001-11
pubmed: 15554571
JMIR Mhealth Uhealth. 2019 Aug 01;7(8):e11734
pubmed: 31373275
Gen Hosp Psychiatry. 2009 Jul-Aug;31(4):306-15
pubmed: 19555789
JMIR Mhealth Uhealth. 2021 Nov 19;9(11):e27674
pubmed: 34806993
JMIR Mhealth Uhealth. 2021 Jul 30;9(7):e29840
pubmed: 34328441
JMIR Mhealth Uhealth. 2022 Oct 4;10(10):e40667
pubmed: 36194451
J Affect Disord. 2020 Mar 15;265:314-324
pubmed: 32090755
J Med Internet Res. 2019 Nov 28;21(11):e16273
pubmed: 31778122
Nat Rev Dis Primers. 2016 Sep 15;2:16065
pubmed: 27629598
BMC Med. 2017 Dec 12;15(1):215
pubmed: 29228943
J Neural Transm (Vienna). 2009 Jun;116(6):777-84
pubmed: 18726137
Biol Psychiatry. 2002 Sep 15;52(6):655-75
pubmed: 12361673
JAMA Psychiatry. 2019 Apr 1;76(4):399-408
pubmed: 30673066
NPJ Digit Med. 2020 Feb 17;3:21
pubmed: 32128451
Comput Methods Programs Biomed. 2022 Dec;227:107204
pubmed: 36371974
J Chiropr Med. 2016 Jun;15(2):155-63
pubmed: 27330520
BMC Psychiatry. 2019 Feb 18;19(1):72
pubmed: 30777041
Eur J Epidemiol. 2012 Jul;27(7):537-46
pubmed: 22623145
Front Psychiatry. 2019 May 21;10:315
pubmed: 31178764
Chronobiol Int. 2018 Apr;35(4):465-476
pubmed: 29235907
J Med Internet Res. 2015 Jul 15;17(7):e175
pubmed: 26180009
NPJ Digit Med. 2023 Feb 17;6(1):25
pubmed: 36806317
Comput Math Methods Med. 2020 Mar 26;2020:7398324
pubmed: 32300374
Soc Psychiatry Psychiatr Epidemiol. 1996 Jan;31(1):3-9
pubmed: 8821918
JMIR Mhealth Uhealth. 2022 Jan 28;10(1):e28095
pubmed: 35089148
J Clin Med. 2022 Dec 01;11(23):
pubmed: 36498739
BMJ Open Sport Exerc Med. 2015 Jul 8;1(1):e000013
pubmed: 27900119
Physiol Meas. 2016 Dec;37(12):2154-2169
pubmed: 27841157
Psychiatr Serv. 2004 Dec;55(12):1379-85
pubmed: 15572565
Arch Gen Psychiatry. 2005 Jun;62(6):593-602
pubmed: 15939837
Educ Psychol Meas. 2016 Feb;76(1):64-87
pubmed: 29795857
J Clin Epidemiol. 1991;44(1):5-13
pubmed: 1986057
BMC Psychiatry. 2022 Feb 21;22(1):136
pubmed: 35189842
J Affect Disord. 2017 Jan 15;208:191-197
pubmed: 27792962
JMIR Ment Health. 2022 Mar 11;9(3):e34898
pubmed: 35275087
BMC Med. 2015 Apr 06;13:72
pubmed: 25879936
J Clin Monit Comput. 2019 Feb;33(1):65-75
pubmed: 29644558
JMIR Mhealth Uhealth. 2018 Aug 13;6(8):e165
pubmed: 30104184
J Adolesc Health. 2012 May;50(5):503-8
pubmed: 22525115
Anesth Analg. 2018 May;126(5):1763-1768
pubmed: 29481436
Front Psychol. 2017 Apr 07;8:456
pubmed: 28439244
J Affect Disord. 2009 Apr;114(1-3):163-73
pubmed: 18752852
Gait Posture. 2017 Mar;53:201-206
pubmed: 28199925
JAMA. 2003 Jun 18;289(23):3095-105
pubmed: 12813115
JMIR Mhealth Uhealth. 2021 Apr 12;9(4):e24604
pubmed: 33843591
Adapt Phys Activ Q. 2007 Jul;24(3):245-53
pubmed: 17916920
JMIR Mhealth Uhealth. 2022 Mar 25;10(3):e34148
pubmed: 35333186
J Med Internet Res. 2020 Sep 25;22(9):e19992
pubmed: 32877352