Associations Between Smartphone Keystroke Metadata and Mental Health Symptoms in Adolescents: Findings From the Future Proofing Study.
adolescents
anxiety
depression
digital phenotype
keystroke dynamics
keystroke metadata
smartphone
students
Journal
JMIR mental health
ISSN: 2368-7959
Titre abrégé: JMIR Ment Health
Pays: Canada
ID NLM: 101658926
Informations de publication
Date de publication:
15 May 2023
15 May 2023
Historique:
received:
13
12
2022
accepted:
22
03
2023
revised:
24
02
2023
medline:
15
5
2023
pubmed:
15
5
2023
entrez:
15
5
2023
Statut:
epublish
Résumé
Mental disorders are prevalent during adolescence. Among the digital phenotypes currently being developed to monitor mental health symptoms, typing behavior is one promising candidate. However, few studies have directly assessed associations between typing behavior and mental health symptom severity, and whether these relationships differs between genders. In a cross-sectional analysis of a large cohort, we tested whether various features of typing behavior derived from keystroke metadata were associated with mental health symptoms and whether these relationships differed between genders. A total of 934 adolescents from the Future Proofing study undertook 2 typing tasks on their smartphones through the Future Proofing app. Common keystroke timing and frequency features were extracted across tasks. Mental health symptoms were assessed using the Patient Health Questionnaire-Adolescent version, the Children's Anxiety Scale-Short Form, the Distress Questionnaire 5, and the Insomnia Severity Index. Bivariate correlations were used to test whether keystroke features were associated with mental health symptoms. The false discovery rates of P values were adjusted to q values. Machine learning models were trained and tested using independent samples (ie, 80% train 20% test) to identify whether keystroke features could be combined to predict mental health symptoms. Keystroke timing features showed a weak negative association with mental health symptoms across participants. When split by gender, females showed weak negative relationships between keystroke timing features and mental health symptoms, and weak positive relationships between keystroke frequency features and mental health symptoms. The opposite relationships were found for males (except for dwell). Machine learning models using keystroke features alone did not predict mental health symptoms. Increased mental health symptoms are weakly associated with faster typing, with important gender differences. Keystroke metadata should be collected longitudinally and combined with other digital phenotypes to enhance their clinical relevance. Australian and New Zealand Clinical Trial Registry, ACTRN12619000855123; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=377664&isReview=true.
Sections du résumé
BACKGROUND
BACKGROUND
Mental disorders are prevalent during adolescence. Among the digital phenotypes currently being developed to monitor mental health symptoms, typing behavior is one promising candidate. However, few studies have directly assessed associations between typing behavior and mental health symptom severity, and whether these relationships differs between genders.
OBJECTIVE
OBJECTIVE
In a cross-sectional analysis of a large cohort, we tested whether various features of typing behavior derived from keystroke metadata were associated with mental health symptoms and whether these relationships differed between genders.
METHODS
METHODS
A total of 934 adolescents from the Future Proofing study undertook 2 typing tasks on their smartphones through the Future Proofing app. Common keystroke timing and frequency features were extracted across tasks. Mental health symptoms were assessed using the Patient Health Questionnaire-Adolescent version, the Children's Anxiety Scale-Short Form, the Distress Questionnaire 5, and the Insomnia Severity Index. Bivariate correlations were used to test whether keystroke features were associated with mental health symptoms. The false discovery rates of P values were adjusted to q values. Machine learning models were trained and tested using independent samples (ie, 80% train 20% test) to identify whether keystroke features could be combined to predict mental health symptoms.
RESULTS
RESULTS
Keystroke timing features showed a weak negative association with mental health symptoms across participants. When split by gender, females showed weak negative relationships between keystroke timing features and mental health symptoms, and weak positive relationships between keystroke frequency features and mental health symptoms. The opposite relationships were found for males (except for dwell). Machine learning models using keystroke features alone did not predict mental health symptoms.
CONCLUSIONS
CONCLUSIONS
Increased mental health symptoms are weakly associated with faster typing, with important gender differences. Keystroke metadata should be collected longitudinally and combined with other digital phenotypes to enhance their clinical relevance.
TRIAL REGISTRATION
BACKGROUND
Australian and New Zealand Clinical Trial Registry, ACTRN12619000855123; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=377664&isReview=true.
Identifiants
pubmed: 37184904
pii: v10i1e44986
doi: 10.2196/44986
pmc: PMC10227695
doi:
Types de publication
Journal Article
Langues
eng
Pagination
e44986Informations de copyright
©Taylor A Braund, Bridianne O’Dea, Debopriyo Bal, Kate Maston, Mark Larsen, Aliza Werner-Seidler, Gabriel Tillman, Helen Christensen. Originally published in JMIR Mental Health (https://mental.jmir.org), 15.05.2023.
Références
Trials. 2020 Jan 2;21(1):2
pubmed: 31898512
Behav Res Ther. 2015 May;68:1-12
pubmed: 25768982
Sci Rep. 2022 May 11;12(1):7690
pubmed: 35546606
PLoS One. 2019 Nov 7;14(11):e0224365
pubmed: 31697686
J Am Med Inform Assoc. 2020 Jul 1;27(7):1007-1018
pubmed: 32467973
JMIR Ment Health. 2022 May 4;9(5):e35549
pubmed: 35507385
Biol Psychiatry Cogn Neurosci Neuroimaging. 2018 Mar;3(3):223-230
pubmed: 29486863
J Clin Epidemiol. 2016 Mar;71:35-42
pubmed: 26464194
Sleep Med. 2011 May;12(5):463-70
pubmed: 21493134
Bipolar Disord. 2018 May;20(3):285-288
pubmed: 29516666
BMJ. 2006 May 6;332(7549):1080
pubmed: 16675816
Biol Psychiatry Cogn Neurosci Neuroimaging. 2022 Mar;7(3):276-284
pubmed: 34363999
J Med Internet Res. 2018 Jul 20;20(7):e241
pubmed: 30030209
J Anxiety Disord. 2003;17(6):605-25
pubmed: 14624814
Sleep Med. 2014 Aug;15(8):934-41
pubmed: 24958244
NPJ Digit Med. 2019 Sep 6;2:88
pubmed: 31508498
Sensors (Basel). 2020 Nov 08;20(21):
pubmed: 33171646
Compr Psychiatry. 2015 Jan;56:29-34
pubmed: 25266848
Behav Res Ther. 1998 May;36(5):545-66
pubmed: 9648330
Arch Gen Psychiatry. 2011 Mar;68(3):263-9
pubmed: 21041606
Int J Methods Psychiatr Res. 2022 Nov 28;:e1954
pubmed: 36444163
Front Psychiatry. 2021 Dec 22;12:739022
pubmed: 35002792
JAMA. 2017 Oct 3;318(13):1215-1216
pubmed: 28973224
Sci Rep. 2019 Sep 16;9(1):13414
pubmed: 31527640
J Adolesc Health. 2002 Mar;30(3):196-204
pubmed: 11869927
Br J Clin Psychol. 2022 Jun;61(2):287-305
pubmed: 34569066
Aust N Z J Psychiatry. 2019 Aug;53(8):782-793
pubmed: 30880405
Sleep Med. 2001 Jul;2(4):297-307
pubmed: 11438246
Mol Psychiatry. 2022 Jan;27(1):281-295
pubmed: 34079068
Proc Natl Acad Sci U S A. 2018 Jul 3;115(27):E6106-E6115
pubmed: 29915059