Karim Zahed
Research Subjects
Research Interests
Karim Zahed has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Farzan Sasangohar
Houston Methodist
14 shared publications - Ranjana K. Mehta
University of Wisconsin-Madison, USA
7 shared publications - Carl Markert
Wm Michael Barnes '64 Department of Industrial & Systems Engineering, Texas A&M University, College Station, TX, USA
6 shared publications - Khalid Qaraqe
College of Science and Engineering, Hamad bin Khalifa University, Doha, Qatar
6 shared publications - Madhav Erraguntla
Department of Industrial and Systems Engineering, Texas A&M University, College Station, TX, USA
6 shared publications - Changwon Son
Department of Industrial Engineering, College of Engineering, Texas Tech University, Lubbock, TX 79409, USA
2 shared publications
Latest Publications
- Open Access
- 1 May 2025
- Published by SAGE Publications in Digital Health
- Vol. 11
Utility of wearable technology in predicting panic attacks: A scoping review
Objective: This scoping review reports on findings of recent studies which assess the methodologies, significant features, and machine learning (ML) models employed in wearable-based panic attack (PA) research. Background: PAs affect a significant percentage of people worldwide and can induce rapid heartbeat, sweating, and trembling in affected individuals. They are often unpredictable and can seriously affect day-to-day activities and functioning. The integration of wearable technology with advanced ML methods has been successfully used in identifying and managing several health conditions. Despite significant advancements, the predictive capabilities of these tools to detect PAs remain less understood. Method: A comprehensive search of databases including PubMed, PsycINFO, Embase, and Google Scholar identified seven studies focusing on PA prediction using wearable devices. Results: These studies employed a range of ML models, such as supervised anomaly detection, deep learning (e.g. LSTM, RNN), random forests, and mixed regression models. The studies analyzed physiological metrics like heart rate (HR) variability, respiratory rate, and activity levels. Accuracy rates varied, with models achieving between 67.4% and 94.8% predictive accuracy. Conclusion: Findings show the utility of combining psychological, physiological, and environmental data for improved predictions, and highlight the key data features, such as resting HR, heart rate variability, and certain sleep metrics that may help predict the onset of PAs. However, most of these studies have impractical prediction time frames of PAs with limited evidence of successful near-real-time prediction, highlighting the need for further research to predict the onset of PAs in real time.
- 1 May 2025
- Published by SAGE Publications in Digital Health
- Vol. 11
Open AccessUtility of wearable technology in predicting panic attacks: A scoping review
Objective: This scoping review reports on findings of recent studies which assess the methodologies, significant features, and machine learning (ML) models employed in wearable-based panic attack (PA) research. Background: PAs affect a significant percentage of people worldwide and can induce rapid heartbeat, sweating, and trembling in affected individuals. They are often unpredictable and can seriously affect day-to-day activities and functioning. The integration of wearable technology with advanced ML methods has been successfully used in identifying and managing several health conditions. Despite significant advancements, the predictive capabilities of these tools to detect PAs remain less understood. Method: A comprehensive search of databases including PubMed, PsycINFO, Embase, and Google Scholar identified seven studies focusing on PA prediction using wearable devices. Results: These studies employed a range of ML models, such as supervised anomaly detection, deep learning (e.g. LSTM, RNN), random forests, and mixed regression models. The studies analyzed physiological metrics like heart rate (HR) variability, respiratory rate, and activity levels. Accuracy rates varied, with models achieving between 67.4% and 94.8% predictive accuracy. Conclusion: Findings show the utility of combining psychological, physiological, and environmental data for improved predictions, and highlight the key data features, such as resting HR, heart rate variability, and certain sleep metrics that may help predict the onset of PAs. However, most of these studies have impractical prediction time frames of PAs with limited evidence of successful near-real-time prediction, highlighting the need for further research to predict the onset of PAs in real time.
- 27 February 2025
- Published by Elsevier in Applied Ergonomics
- Vol. 126, 104485
Understanding the role of beliefs on intentions and actual usage of a tool for self-management of mental health among college students
- 27 February 2025
- Published by Elsevier in Applied Ergonomics
- Vol. 126, 104485
Understanding the role of beliefs on intentions and actual usage of a tool for self-management of mental health among college students
- Open Access
- 3 May 2023
- Published by JMIR Publications Inc. in JMIR Diabetes
- Vol. 8 (1) , e41501
Understanding Patient Beliefs in Using Technology to Manage Diabetes: Path Analysis Model From a National Web-Based Sample
Background: With 425 million individuals globally living with diabetes, it is critical to support the self-management of this life-threatening condition. However, adherence and engagement with existing technologies are inadequate and need further research. Objective: The objective of our study was to develop an integrated belief model that helps identify the significant constructs in predicting intention to use a diabetes self-management device for the detection of hypoglycemia. Methods: Adults with type 1 diabetes living in the United States were recruited through Qualtrics to take a web-based questionnaire that assessed their preferences for a device that monitors their tremors and alerts them of the onset of hypoglycemia. As part of this questionnaire, a section focused on eliciting their response to behavioral constructs from the Health Belief Model, Technology Acceptance Model, and others. Results: A total of 212 eligible participants responded to the Qualtrics survey. Intention to use a device for the self-management of diabetes was well predicted (R2=0.65; F12,199=27.19; P<.001) by 4 main constructs. The most significant constructs were perceived usefulness (β=.33; P<.001) and perceived health threat (β=.55; P<.001) followed by cues to action (β=.17; P<.001) and a negative effect from resistance to change (β=−.19; P<.001). Older age (β=.025; P<.001) led to an increase in their perceived health threat. Conclusions: For individuals to use such a device, they need to perceive it as useful, perceive diabetes as life-threatening, regularly remember to perform actions to manage their condition, and exhibit less resistance to change. The model predicted the intention to use a diabetes self-management device as well, with several constructs found to be significant. This mental modeling approach can be complemented in future work by field-testing with physical prototype devices and assessing their interaction with the device longitudinally.
- 3 May 2023
- Published by JMIR Publications Inc. in JMIR Diabetes
- Vol. 8 (1) , e41501
Open AccessUnderstanding Patient Beliefs in Using Technology to Manage Diabetes: Path Analysis Model From a National Web-Based Sample
Background: With 425 million individuals globally living with diabetes, it is critical to support the self-management of this life-threatening condition. However, adherence and engagement with existing technologies are inadequate and need further research. Objective: The objective of our study was to develop an integrated belief model that helps identify the significant constructs in predicting intention to use a diabetes self-management device for the detection of hypoglycemia. Methods: Adults with type 1 diabetes living in the United States were recruited through Qualtrics to take a web-based questionnaire that assessed their preferences for a device that monitors their tremors and alerts them of the onset of hypoglycemia. As part of this questionnaire, a section focused on eliciting their response to behavioral constructs from the Health Belief Model, Technology Acceptance Model, and others. Results: A total of 212 eligible participants responded to the Qualtrics survey. Intention to use a device for the self-management of diabetes was well predicted (R2=0.65; F12,199=27.19; P<.001) by 4 main constructs. The most significant constructs were perceived usefulness (β=.33; P<.001) and perceived health threat (β=.55; P<.001) followed by cues to action (β=.17; P<.001) and a negative effect from resistance to change (β=−.19; P<.001). Older age (β=.025; P<.001) led to an increase in their perceived health threat. Conclusions: For individuals to use such a device, they need to perceive it as useful, perceive diabetes as life-threatening, regularly remember to perform actions to manage their condition, and exhibit less resistance to change. The model predicted the intention to use a diabetes self-management device as well, with several constructs found to be significant. This mental modeling approach can be complemented in future work by field-testing with physical prototype devices and assessing their interaction with the device longitudinally.
- Open Access
- 19 April 2023
- Published by JMIR Publications Inc. in JMIR Diabetes
- Vol. 8 (1) , e40990
Hypoglycemia Detection Using Hand Tremors: Home Study of Patients With Type 1 Diabetes
Background: Diabetes affects millions of people worldwide and is steadily increasing. A serious condition associated with diabetes is low glucose levels (hypoglycemia). Monitoring blood glucose is usually performed by invasive methods or intrusive devices, and these devices are currently not available to all patients with diabetes. Hand tremor is a significant symptom of hypoglycemia, as nerves and muscles are powered by blood sugar. However, to our knowledge, no validated tools or algorithms exist to monitor and detect hypoglycemic events via hand tremors. Objective: In this paper, we propose a noninvasive method to detect hypoglycemic events based on hand tremors using accelerometer data. Methods: We analyzed triaxial accelerometer data from a smart watch recorded from 33 patients with type 1 diabetes for 1 month. Time and frequency domain features were extracted from acceleration signals to explore different machine learning models to classify and differentiate between hypoglycemic and nonhypoglycemic states. Results: The mean duration of the hypoglycemic state was 27.31 (SD 5.15) minutes per day for each patient. On average, patients had 1.06 (SD 0.77) hypoglycemic events per day. The ensemble learning model based on random forest, support vector machines, and k-nearest neighbors had the best performance, with a precision of 81.5% and a recall of 78.6%. The results were validated using continuous glucose monitor readings as ground truth. Conclusions: Our results indicate that the proposed approach can be a potential tool to detect hypoglycemia and can serve as a proactive, nonintrusive alert mechanism for hypoglycemic events.
- 19 April 2023
- Published by JMIR Publications Inc. in JMIR Diabetes
- Vol. 8 (1) , e40990
Open AccessHypoglycemia Detection Using Hand Tremors: Home Study of Patients With Type 1 Diabetes
Background: Diabetes affects millions of people worldwide and is steadily increasing. A serious condition associated with diabetes is low glucose levels (hypoglycemia). Monitoring blood glucose is usually performed by invasive methods or intrusive devices, and these devices are currently not available to all patients with diabetes. Hand tremor is a significant symptom of hypoglycemia, as nerves and muscles are powered by blood sugar. However, to our knowledge, no validated tools or algorithms exist to monitor and detect hypoglycemic events via hand tremors. Objective: In this paper, we propose a noninvasive method to detect hypoglycemic events based on hand tremors using accelerometer data. Methods: We analyzed triaxial accelerometer data from a smart watch recorded from 33 patients with type 1 diabetes for 1 month. Time and frequency domain features were extracted from acceleration signals to explore different machine learning models to classify and differentiate between hypoglycemic and nonhypoglycemic states. Results: The mean duration of the hypoglycemic state was 27.31 (SD 5.15) minutes per day for each patient. On average, patients had 1.06 (SD 0.77) hypoglycemic events per day. The ensemble learning model based on random forest, support vector machines, and k-nearest neighbors had the best performance, with a precision of 81.5% and a recall of 78.6%. The results were validated using continuous glucose monitor readings as ground truth. Conclusions: Our results indicate that the proposed approach can be a potential tool to detect hypoglycemia and can serve as a proactive, nonintrusive alert mechanism for hypoglycemic events.
- Open Access
- 12 April 2023
- Published by JMIR Publications Inc. in JMIR Formative Research
- Vol. 7 (1) , e41018
Use of a Mobile Biofeedback App to Provide Health Coaching for Stress Self-management: Pilot Quasi-Experiment
Background: Mental health is an increasing concern among vulnerable populations, including college students and veterans. Objective: The purpose of this study was to determine if mobile health technology combined with health coaching can better enable a user to self-manage their mental health. Methods: This study evaluated the mobile app “Biofeedback” that provided health coaching on stress self-management for college student veterans’ mental health concerns. Twenty-four college student veterans were recruited from a large public university in Texas during the spring 2020 semester, impacted by COVID-19. Ten participants were assigned to the intervention group where they used the mobile Biofeedback app on their smartphones and smartwatches, and 14 were assigned to the control group without the app; assignment was based on mobile phone compatibility. Both groups participated in one initial lab session where they learned a deep-breathing exercise technique. The intervention group was then asked to use the mobile Biofeedback app during their daily lives and a smartwatch, and the control group was asked to perform the breathing exercises on their own. Both groups filled out Patient Health Questionnaire (PHQ-9) and Generalized Anxiety Disorder (GAD-7) self-assessments at 2-week intervals. At the end of the semester, both groups were given an exit interview to provide user experience and perceived benefits of health coaching via the mobile biofeedback app. Results: The deep-breathing exercise in the initial lab session reduced stress in both groups. Over the course of the study, the app recorded 565 coached breathing exercises with a significant decrease (approximately 3 beats per minute) in participants’ heart rate during the 6-minute time period immediately after conducting the breathing exercises (Spearman rank correlation coefficient –0.61, P<.001; S=9,816,176). There was no significant difference between the two groups for PHQ-9 and GAD-7 scores over the course of the semester. Exit interview responses indicated that participants perceived that the mobile Biofeedback app improved their health and helped them address stress challenges. All participants reported that the intervention helped them manage their stress better and expressed that health coaching via a mobile device would improve their overall health. Conclusions: Participants reported a positive perception of the app for their mental health self-management during a stressful semester. Future work should examine long-term effects of the app with a larger sample size balanced between male and female participants, randomized participant allocation, real-time detection of mental health symptoms, and additional features of the app.
- 12 April 2023
- Published by JMIR Publications Inc. in JMIR Formative Research
- Vol. 7 (1) , e41018
Open AccessUse of a Mobile Biofeedback App to Provide Health Coaching for Stress Self-management: Pilot Quasi-Experiment
Background: Mental health is an increasing concern among vulnerable populations, including college students and veterans. Objective: The purpose of this study was to determine if mobile health technology combined with health coaching can better enable a user to self-manage their mental health. Methods: This study evaluated the mobile app “Biofeedback” that provided health coaching on stress self-management for college student veterans’ mental health concerns. Twenty-four college student veterans were recruited from a large public university in Texas during the spring 2020 semester, impacted by COVID-19. Ten participants were assigned to the intervention group where they used the mobile Biofeedback app on their smartphones and smartwatches, and 14 were assigned to the control group without the app; assignment was based on mobile phone compatibility. Both groups participated in one initial lab session where they learned a deep-breathing exercise technique. The intervention group was then asked to use the mobile Biofeedback app during their daily lives and a smartwatch, and the control group was asked to perform the breathing exercises on their own. Both groups filled out Patient Health Questionnaire (PHQ-9) and Generalized Anxiety Disorder (GAD-7) self-assessments at 2-week intervals. At the end of the semester, both groups were given an exit interview to provide user experience and perceived benefits of health coaching via the mobile biofeedback app. Results: The deep-breathing exercise in the initial lab session reduced stress in both groups. Over the course of the study, the app recorded 565 coached breathing exercises with a significant decrease (approximately 3 beats per minute) in participants’ heart rate during the 6-minute time period immediately after conducting the breathing exercises (Spearman rank correlation coefficient –0.61, P<.001; S=9,816,176). There was no significant difference between the two groups for PHQ-9 and GAD-7 scores over the course of the semester. Exit interview responses indicated that participants perceived that the mobile Biofeedback app improved their health and helped them address stress challenges. All participants reported that the intervention helped them manage their stress better and expressed that health coaching via a mobile device would improve their overall health. Conclusions: Participants reported a positive perception of the app for their mental health self-management during a stressful semester. Future work should examine long-term effects of the app with a larger sample size balanced between male and female participants, randomized participant allocation, real-time detection of mental health symptoms, and additional features of the app.
- Open Access
- 1 January 2023
- Published by SAGE Publications in Digital Health
- Vol. 9
Investigating the effect of an mHealth coaching intervention on health beliefs, adherence and blood pressure of patients with hypertension: A longitudinal single group pilot study
Background: Mobile health technologies have shown promise as delivery platforms for digital health coaching for chronic conditions. However, the impacts of such strategies on users’ health beliefs, intentions and ultimately clinical outcomes are understudied. Objective: This study sought (1) to evaluate the effects of a digital health coaching intervention on participants’ belief constructs; and (2) to assess relationships between these belief constructs and intentions to utilize the technological intervention, actual adherence metrics and clinical outcomes related to hypertension. Methods: Thirty-four participants with hypertension were recruited from a university community from January to May 2021. They self-measured weight and blood pressure (BP) for 30 days followed by digital coaching delivered via a mobile application for 30 days. Surveys assessed constructs from the Health Belief Model and Technology Acceptance Model, compared to intention, health belief, BP self-monitoring adherence and BP outcomes. A path analysis model was used to assess the relationships between constructs and intention, adherence metrics and clinical outcomes. A Kruskal–Wallis test was used to identify changes in beliefs. Results: Participant health beliefs significantly improved after coaching, including self-efficacy ( H(1) = 15.12, p < 0.001), cues to action ( H(1) = 5.33, p = 0.02), attitude ( H(1) = 10.35, p = 0.002), perceived usefulness ( H(1) = 15.02, p < 0.001) and decreased resistance to change ( H(1) = 4.05, p = 0.04). Adherence to BP measurements positively correlated with perceived health threat ( β = .033, p = 0.007) and perceived ease of use ( β = .0277, p < 0.001). Self-efficacy ( β = −2.92, p = 0.02) and perceived usefulness ( β = −3.75, p = 0.01) were linked with a decrease in diastolic BP. Conclusions: A mobile health coaching intervention may help participants improve beliefs regarding hypertension self-management.
- 1 January 2023
- Published by SAGE Publications in Digital Health
- Vol. 9
Open AccessInvestigating the effect of an mHealth coaching intervention on health beliefs, adherence and blood pressure of patients with hypertension: A longitudinal single group pilot study
Background: Mobile health technologies have shown promise as delivery platforms for digital health coaching for chronic conditions. However, the impacts of such strategies on users’ health beliefs, intentions and ultimately clinical outcomes are understudied. Objective: This study sought (1) to evaluate the effects of a digital health coaching intervention on participants’ belief constructs; and (2) to assess relationships between these belief constructs and intentions to utilize the technological intervention, actual adherence metrics and clinical outcomes related to hypertension. Methods: Thirty-four participants with hypertension were recruited from a university community from January to May 2021. They self-measured weight and blood pressure (BP) for 30 days followed by digital coaching delivered via a mobile application for 30 days. Surveys assessed constructs from the Health Belief Model and Technology Acceptance Model, compared to intention, health belief, BP self-monitoring adherence and BP outcomes. A path analysis model was used to assess the relationships between constructs and intention, adherence metrics and clinical outcomes. A Kruskal–Wallis test was used to identify changes in beliefs. Results: Participant health beliefs significantly improved after coaching, including self-efficacy ( H(1) = 15.12, p < 0.001), cues to action ( H(1) = 5.33, p = 0.02), attitude ( H(1) = 10.35, p = 0.002), perceived usefulness ( H(1) = 15.02, p < 0.001) and decreased resistance to change ( H(1) = 4.05, p = 0.04). Adherence to BP measurements positively correlated with perceived health threat ( β = .033, p = 0.007) and perceived ease of use ( β = .0277, p < 0.001). Self-efficacy ( β = −2.92, p = 0.02) and perceived usefulness ( β = −3.75, p = 0.01) were linked with a decrease in diastolic BP. Conclusions: A mobile health coaching intervention may help participants improve beliefs regarding hypertension self-management.
- Open Access
- 6 October 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Using mHealth to Support Mental Health Coaching for College Students: A Mixed Design Study (Preprint)
BACKGROUND Student mental health is an increasing concern among college student populations. OBJECTIVE To determine if college students could better self-manage their anxiety, stress, and depression symptomology by using automated mHealth to support health coaching for managing mental health. METHODS Students were provided with a mobile health self-management app called mHELP. The intervention group students had on-demand access to interactive mental health management tools on the app and completed weekly self-assessments of anxiety (GAD-7), perceived stress (PSS-10), and depression (PHQ-8) with immediate feedback of self-reported measurements during the first phase (five weeks) of the study. During the second five-week phase, students were additionally provided daily educational material in the form of videos, brochures, and articles related to the self-management of mental health issues. The educational material was developed in collaboration with the university’s Counseling and Psychological Services Department. Weekly self-reported measures of anxiety, perceived stress, and depression were compared in a pre-post analysis for each of the two phases as well as an analysis of variance of the changes in the outcome values for each phase to check for significance (within- and between-group). The control group did not receive or have access to any digital health coaching information from the mHELP app, but they did complete weekly assessments of self-reported measures of anxiety, perceived stress, and depression with immediate feedback on the results. RESULTS A total of 87 participants, 72.4% assigned randomly to the intervention group, completed the study with a mean age of 22.1 years (SD = 4.35, Range: 18-37). No significant improvements in outcomes over time were noted for the intervention group during the first phase of the study; however, during the second phase of the study, significant improvements were noted for anxiety (p = .002, d = .428) and perceived stress (p = .033, d = .306) over time. Over the duration of both phases of the study, the control and intervention groups both showed improvements in mean GAD-7, PSS-10 and PHQ-8 scores. The intervention group exhibited a statistically significant improvement in mean GAD-7 (M = 3.143, SE = 0.638, p < .001, d = .621), PSS-10 (M = 3.476, SE = 0.968, p = .001, d = .452) and PHQ-8 (M = 1.619, SE = 0.648, p = .041, d = .315). The control group only exhibited a statistically significant improvement in their GAD-7 score (M = 3.167, SE = 1.251, p = .019, d = .517). An ANOVA of the differences in improvements between groups did not find a significant effect of the mHELP intervention for GAD-7, PSS-10 or PHQ-8 scores. CONCLUSIONS Our study demonstrated that using mHealth to support health coaching may assist in the self-management of anxiety and stress among college students, while a self-measurement routine may also provide benefits for improving anxiety over time, perhaps due to improved self-awareness of mental health conditions.
- 6 October 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessUsing mHealth to Support Mental Health Coaching for College Students: A Mixed Design Study (Preprint)
BACKGROUND Student mental health is an increasing concern among college student populations. OBJECTIVE To determine if college students could better self-manage their anxiety, stress, and depression symptomology by using automated mHealth to support health coaching for managing mental health. METHODS Students were provided with a mobile health self-management app called mHELP. The intervention group students had on-demand access to interactive mental health management tools on the app and completed weekly self-assessments of anxiety (GAD-7), perceived stress (PSS-10), and depression (PHQ-8) with immediate feedback of self-reported measurements during the first phase (five weeks) of the study. During the second five-week phase, students were additionally provided daily educational material in the form of videos, brochures, and articles related to the self-management of mental health issues. The educational material was developed in collaboration with the university’s Counseling and Psychological Services Department. Weekly self-reported measures of anxiety, perceived stress, and depression were compared in a pre-post analysis for each of the two phases as well as an analysis of variance of the changes in the outcome values for each phase to check for significance (within- and between-group). The control group did not receive or have access to any digital health coaching information from the mHELP app, but they did complete weekly assessments of self-reported measures of anxiety, perceived stress, and depression with immediate feedback on the results. RESULTS A total of 87 participants, 72.4% assigned randomly to the intervention group, completed the study with a mean age of 22.1 years (SD = 4.35, Range: 18-37). No significant improvements in outcomes over time were noted for the intervention group during the first phase of the study; however, during the second phase of the study, significant improvements were noted for anxiety (p = .002, d = .428) and perceived stress (p = .033, d = .306) over time. Over the duration of both phases of the study, the control and intervention groups both showed improvements in mean GAD-7, PSS-10 and PHQ-8 scores. The intervention group exhibited a statistically significant improvement in mean GAD-7 (M = 3.143, SE = 0.638, p < .001, d = .621), PSS-10 (M = 3.476, SE = 0.968, p = .001, d = .452) and PHQ-8 (M = 1.619, SE = 0.648, p = .041, d = .315). The control group only exhibited a statistically significant improvement in their GAD-7 score (M = 3.167, SE = 1.251, p = .019, d = .517). An ANOVA of the differences in improvements between groups did not find a significant effect of the mHELP intervention for GAD-7, PSS-10 or PHQ-8 scores. CONCLUSIONS Our study demonstrated that using mHealth to support health coaching may assist in the self-management of anxiety and stress among college students, while a self-measurement routine may also provide benefits for improving anxiety over time, perhaps due to improved self-awareness of mental health conditions.
- Open Access
- 27 July 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Understanding Patient Beliefs in Using Technology to Manage Diabetes: Path Analysis Model From a National Web-Based Sample (Preprint)
BACKGROUND With 425 million individuals globally living with diabetes, it is critical to support the self-management of this life-threatening condition. However, adherence and engagement with existing technologies are inadequate and need further research. OBJECTIVE The objective of our study was to develop an integrated belief model that helps identify the significant constructs in predicting intention to use a diabetes self-management device for the detection of hypoglycemia. METHODS Adults with type 1 diabetes living in the United States were recruited through Qualtrics to take a web-based questionnaire that assessed their preferences for a device that monitors their tremors and alerts them of the onset of hypoglycemia. As part of this questionnaire, a section focused on eliciting their response to behavioral constructs from the Health Belief Model, Technology Acceptance Model, and others. RESULTS A total of 212 eligible participants responded to the Qualtrics survey. Intention to use a device for the self-management of diabetes was well predicted (R2=0.65; F12,199=27.19; P<.001) by 4 main constructs. The most significant constructs were perceived usefulness (β=.33; P<.001) and perceived health threat (β=.55; P<.001) followed by cues to action (β=.17; P<.001) and a negative effect from resistance to change (β=−.19; P<.001). Older age (β=.025; P<.001) led to an increase in their perceived health threat. CONCLUSIONS For individuals to use such a device, they need to perceive it as useful, perceive diabetes as life-threatening, regularly remember to perform actions to manage their condition, and exhibit less resistance to change. The model predicted the intention to use a diabetes self-management device as well, with several constructs found to be significant. This mental modeling approach can be complemented in future work by field-testing with physical prototype devices and assessing their interaction with the device longitudinally.
- 27 July 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessUnderstanding Patient Beliefs in Using Technology to Manage Diabetes: Path Analysis Model From a National Web-Based Sample (Preprint)
BACKGROUND With 425 million individuals globally living with diabetes, it is critical to support the self-management of this life-threatening condition. However, adherence and engagement with existing technologies are inadequate and need further research. OBJECTIVE The objective of our study was to develop an integrated belief model that helps identify the significant constructs in predicting intention to use a diabetes self-management device for the detection of hypoglycemia. METHODS Adults with type 1 diabetes living in the United States were recruited through Qualtrics to take a web-based questionnaire that assessed their preferences for a device that monitors their tremors and alerts them of the onset of hypoglycemia. As part of this questionnaire, a section focused on eliciting their response to behavioral constructs from the Health Belief Model, Technology Acceptance Model, and others. RESULTS A total of 212 eligible participants responded to the Qualtrics survey. Intention to use a device for the self-management of diabetes was well predicted (R2=0.65; F12,199=27.19; P<.001) by 4 main constructs. The most significant constructs were perceived usefulness (β=.33; P<.001) and perceived health threat (β=.55; P<.001) followed by cues to action (β=.17; P<.001) and a negative effect from resistance to change (β=−.19; P<.001). Older age (β=.025; P<.001) led to an increase in their perceived health threat. CONCLUSIONS For individuals to use such a device, they need to perceive it as useful, perceive diabetes as life-threatening, regularly remember to perform actions to manage their condition, and exhibit less resistance to change. The model predicted the intention to use a diabetes self-management device as well, with several constructs found to be significant. This mental modeling approach can be complemented in future work by field-testing with physical prototype devices and assessing their interaction with the device longitudinally.
- Open Access
- 12 July 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Use of a Mobile Biofeedback App to Provide Health Coaching for Stress Self-management: Pilot Quasi-Experiment (Preprint)
BACKGROUND Mental health is an increasing concern among vulnerable populations, including college students and veterans. OBJECTIVE The purpose of this study was to determine if mobile health technology combined with health coaching can better enable a user to self-manage their mental health. METHODS This study evaluated the mobile app “Biofeedback” that provided health coaching on stress self-management for college student veterans’ mental health concerns. Twenty-four college student veterans were recruited from a large public university in Texas during the spring 2020 semester, impacted by COVID-19. Ten participants were assigned to the intervention group where they used the mobile Biofeedback app on their smartphones and smartwatches, and 14 were assigned to the control group without the app; assignment was based on mobile phone compatibility. Both groups participated in one initial lab session where they learned a deep-breathing exercise technique. The intervention group was then asked to use the mobile Biofeedback app during their daily lives and a smartwatch, and the control group was asked to perform the breathing exercises on their own. Both groups filled out Patient Health Questionnaire (PHQ-9) and Generalized Anxiety Disorder (GAD-7) self-assessments at 2-week intervals. At the end of the semester, both groups were given an exit interview to provide user experience and perceived benefits of health coaching via the mobile biofeedback app. RESULTS The deep-breathing exercise in the initial lab session reduced stress in both groups. Over the course of the study, the app recorded 565 coached breathing exercises with a significant decrease (approximately 3 beats per minute) in participants’ heart rate during the 6-minute time period immediately after conducting the breathing exercises (Spearman rank correlation coefficient –0.61, P<.001; S=9,816,176). There was no significant difference between the two groups for PHQ-9 and GAD-7 scores over the course of the semester. Exit interview responses indicated that participants perceived that the mobile Biofeedback app improved their health and helped them address stress challenges. All participants reported that the intervention helped them manage their stress better and expressed that health coaching via a mobile device would improve their overall health. CONCLUSIONS Participants reported a positive perception of the app for their mental health self-management during a stressful semester. Future work should examine long-term effects of the app with a larger sample size balanced between male and female participants, randomized participant allocation, real-time detection of mental health symptoms, and additional features of the app.
- 12 July 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessUse of a Mobile Biofeedback App to Provide Health Coaching for Stress Self-management: Pilot Quasi-Experiment (Preprint)
BACKGROUND Mental health is an increasing concern among vulnerable populations, including college students and veterans. OBJECTIVE The purpose of this study was to determine if mobile health technology combined with health coaching can better enable a user to self-manage their mental health. METHODS This study evaluated the mobile app “Biofeedback” that provided health coaching on stress self-management for college student veterans’ mental health concerns. Twenty-four college student veterans were recruited from a large public university in Texas during the spring 2020 semester, impacted by COVID-19. Ten participants were assigned to the intervention group where they used the mobile Biofeedback app on their smartphones and smartwatches, and 14 were assigned to the control group without the app; assignment was based on mobile phone compatibility. Both groups participated in one initial lab session where they learned a deep-breathing exercise technique. The intervention group was then asked to use the mobile Biofeedback app during their daily lives and a smartwatch, and the control group was asked to perform the breathing exercises on their own. Both groups filled out Patient Health Questionnaire (PHQ-9) and Generalized Anxiety Disorder (GAD-7) self-assessments at 2-week intervals. At the end of the semester, both groups were given an exit interview to provide user experience and perceived benefits of health coaching via the mobile biofeedback app. RESULTS The deep-breathing exercise in the initial lab session reduced stress in both groups. Over the course of the study, the app recorded 565 coached breathing exercises with a significant decrease (approximately 3 beats per minute) in participants’ heart rate during the 6-minute time period immediately after conducting the breathing exercises (Spearman rank correlation coefficient –0.61, P<.001; S=9,816,176). There was no significant difference between the two groups for PHQ-9 and GAD-7 scores over the course of the semester. Exit interview responses indicated that participants perceived that the mobile Biofeedback app improved their health and helped them address stress challenges. All participants reported that the intervention helped them manage their stress better and expressed that health coaching via a mobile device would improve their overall health. CONCLUSIONS Participants reported a positive perception of the app for their mental health self-management during a stressful semester. Future work should examine long-term effects of the app with a larger sample size balanced between male and female participants, randomized participant allocation, real-time detection of mental health symptoms, and additional features of the app.
- Open Access
- 11 July 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Hypoglycemia Detection Using Hand Tremors: Home Study of Patients With Type 1 Diabetes (Preprint)
BACKGROUND Diabetes affects millions of people worldwide and is steadily increasing. A serious condition associated with diabetes is low glucose levels (hypoglycemia). Monitoring blood glucose is usually performed by invasive methods or intrusive devices, and these devices are currently not available to all patients with diabetes. Hand tremor is a significant symptom of hypoglycemia, as nerves and muscles are powered by blood sugar. However, to our knowledge, no validated tools or algorithms exist to monitor and detect hypoglycemic events via hand tremors. OBJECTIVE In this paper, we propose a noninvasive method to detect hypoglycemic events based on hand tremors using accelerometer data. METHODS We analyzed triaxial accelerometer data from a smart watch recorded from 33 patients with type 1 diabetes for 1 month. Time and frequency domain features were extracted from acceleration signals to explore different machine learning models to classify and differentiate between hypoglycemic and nonhypoglycemic states. RESULTS The mean duration of the hypoglycemic state was 27.31 (SD 5.15) minutes per day for each patient. On average, patients had 1.06 (SD 0.77) hypoglycemic events per day. The ensemble learning model based on random forest, support vector machines, and k-nearest neighbors had the best performance, with a precision of 81.5% and a recall of 78.6%. The results were validated using continuous glucose monitor readings as ground truth. CONCLUSIONS Our results indicate that the proposed approach can be a potential tool to detect hypoglycemia and can serve as a proactive, nonintrusive alert mechanism for hypoglycemic events.
- 11 July 2022
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessHypoglycemia Detection Using Hand Tremors: Home Study of Patients With Type 1 Diabetes (Preprint)
BACKGROUND Diabetes affects millions of people worldwide and is steadily increasing. A serious condition associated with diabetes is low glucose levels (hypoglycemia). Monitoring blood glucose is usually performed by invasive methods or intrusive devices, and these devices are currently not available to all patients with diabetes. Hand tremor is a significant symptom of hypoglycemia, as nerves and muscles are powered by blood sugar. However, to our knowledge, no validated tools or algorithms exist to monitor and detect hypoglycemic events via hand tremors. OBJECTIVE In this paper, we propose a noninvasive method to detect hypoglycemic events based on hand tremors using accelerometer data. METHODS We analyzed triaxial accelerometer data from a smart watch recorded from 33 patients with type 1 diabetes for 1 month. Time and frequency domain features were extracted from acceleration signals to explore different machine learning models to classify and differentiate between hypoglycemic and nonhypoglycemic states. RESULTS The mean duration of the hypoglycemic state was 27.31 (SD 5.15) minutes per day for each patient. On average, patients had 1.06 (SD 0.77) hypoglycemic events per day. The ensemble learning model based on random forest, support vector machines, and k-nearest neighbors had the best performance, with a precision of 81.5% and a recall of 78.6%. The results were validated using continuous glucose monitor readings as ground truth. CONCLUSIONS Our results indicate that the proposed approach can be a potential tool to detect hypoglycemia and can serve as a proactive, nonintrusive alert mechanism for hypoglycemic events.