An interpretable machine learning approach to multimodal stress detection in a simulated office environment.

BACKGROUND AND OBJECTIVE: Work-related stress affects a large part of today's workforce and is known to have detrimental effects on physical and mental health. Continuous and unobtrusive stress detection may help prevent and reduce stress by providing personalised feedback and allowing for the development of just-in-time adaptive health interventions for stress management. Previous studies on stress detection in work environments have often struggled to adequately reflect real-world conditions in controlled laboratory experiments. To close this gap, in this paper, we present a machine learning methodology for stress detection based on multimodal data collected from unobtrusive sources in an experiment simulating a realistic group office environment (N=90). METHODS: We derive mouse, keyboard and heart rate variability features to detect three levels of perceived stress, valence and arousal with support vector machines, random forests and gradient boosting models using 10-fold cross-validation. We interpret the contributions of features to the model predictions with SHapley Additive exPlanations (SHAP) value plots. RESULTS: The gradient boosting models based on mouse and keyboard features obtained the highest average F1 scores of 0.625, 0.631 and 0.775 for the multiclass prediction of perceived stress, arousal and valence, respectively. Our results indicate that the combination of mouse and keyboard features may be better suited to detect stress in office environments than heart rate variability, despite physiological signal-based stress detection being more established in theory and research. The analysis of SHAP value plots shows that specific mouse movement and typing behaviours may characterise different levels of stress. CONCLUSIONS: Our study fills different methodological gaps in the research on the automated detection of stress in office environments, such as approximating real-life conditions in a laboratory and combining physiological and behavioural data sources. Implications for field studies on personalised, interpretable ML-based systems for the real-time detection of stress in real office environments are also discussed.

Exploring the potential of mobile health interventions to address behavioural risk factors for the prevention of non-communicable diseases in Asian populations: a qualitative study.

BACKGROUND: Changing lifestyle patterns over the last decades have seen growing numbers of people in Asia affected by non-communicable diseases and common mental health disorders, including diabetes, cancer, and/or depression. Interventions targeting healthy lifestyle behaviours through mobile technologies, including new approaches such as chatbots, may be an effective, low-cost approach to prevent these conditions. To ensure uptake and engagement with mobile health interventions, however, it is essential to understand the end-users' perspectives on using such interventions. The aim of this study was to explore perceptions, barriers, and facilitators to the use of mobile health interventions for lifestyle behaviour change in Singapore. METHODS: Six virtual focus group discussions were conducted with a total of 34 participants (mean ± SD; aged 45 ± 3.6 years; 64.7% females). Focus group recordings were transcribed verbatim and analysed using an inductive thematic analysis approach, followed by deductive mapping according to perceptions, barriers, facilitators, mixed factors, or strategies. RESULTS: Five themes were identified: (i) holistic wellbeing is central to healthy living (i.e., the importance of both physical and mental health); (ii) encouraging uptake of a mobile health intervention is influenced by factors such as incentives and government backing; (iii) trying out a mobile health intervention is one thing, sticking to it long term is another and there are key factors, such as personalisation and ease of use that influence sustained engagement with mobile health interventions; (iv) perceptions of chatbots as a tool to support healthy lifestyle behaviour are influenced by previous negative experiences with chatbots, which might hamper uptake; and (v) sharing health-related data is OK, but with conditions such as clarity on who will have access to the data, how it will be stored, and for what purpose it will be used. CONCLUSIONS: Findings highlight several factors that are relevant for the development and implementation of mobile health interventions in Singapore and other Asian countries. Recommendations include: (i) targeting holistic wellbeing, (ii) tailoring content to address environment-specific barriers, (iii) partnering with government and/or local (non-profit) institutions in the development and/or promotion of mobile health interventions, (iv) managing expectations regarding the use of incentives, and (iv) identifying potential alternatives or complementary approaches to the use of chatbots, particularly for mental health.

Digital Behavior Change Interventions for the Prevention and Management of Type 2 Diabetes: Systematic Market Analysis.

BACKGROUND: Advancements in technology offer new opportunities for the prevention and management of type 2 diabetes. Venture capital companies have been investing in digital diabetes companies that offer digital behavior change interventions (DBCIs). However, little is known about the scientific evidence underpinning such interventions or the degree to which these interventions leverage novel technology-driven automated developments such as conversational agents (CAs) or just-in-time adaptive intervention (JITAI) approaches. OBJECTIVE: Our objectives were to identify the top-funded companies offering DBCIs for type 2 diabetes management and prevention, review the level of scientific evidence underpinning the DBCIs, identify which DBCIs are recognized as evidence-based programs by quality assurance authorities, and examine the degree to which these DBCIs include novel automated approaches such as CAs and JITAI mechanisms. METHODS: A systematic search was conducted using 2 venture capital databases (Crunchbase Pro and Pitchbook) to identify the top-funded companies offering interventions for type 2 diabetes prevention and management. Scientific publications relating to the identified DBCIs were identified via PubMed, Google Scholar, and the DBCIs' websites, and data regarding intervention effectiveness were extracted. The Diabetes Prevention Recognition Program (DPRP) of the Center for Disease Control and Prevention in the United States was used to identify the recognition status. The DBCIs' publications, websites, and mobile apps were reviewed with regard to the intervention characteristics. RESULTS: The 16 top-funded companies offering DBCIs for type 2 diabetes received a total funding of US $2.4 billion as of June 15, 2021. Only 4 out of the 50 identified publications associated with these DBCIs were fully powered randomized controlled trials (RCTs). Further, 1 of those 4 RCTs showed a significant difference in glycated hemoglobin A1c (HbA1c) outcomes between the intervention and control groups. However, all the studies reported HbA1c improvements ranging from 0.2% to 1.9% over the course of 12 months. In addition, 6 interventions were fully recognized by the DPRP to deliver evidence-based programs, and 2 interventions had a pending recognition status. Health professionals were included in the majority of DBCIs (13/16, 81%,), whereas only 10% (1/10) of accessible apps involved a CA as part of the intervention delivery. Self-reports represented most of the data sources (74/119, 62%) that could be used to tailor JITAIs. CONCLUSIONS: Our findings suggest that the level of funding received by companies offering DBCIs for type 2 diabetes prevention and management does not coincide with the level of evidence on the intervention effectiveness. There is considerable variation in the level of evidence underpinning the different DBCIs and an overall need for more rigorous effectiveness trials and transparent reporting by quality assurance authorities. Currently, very few DBCIs use automated approaches such as CAs and JITAIs, limiting the scalability and reach of these solutions.

Self-guided Cognitive Behavioral Therapy Apps for Depression: Systematic Assessment of Features, Functionality, and Congruence With Evidence.

BACKGROUND: Mental health disorders affect 1 in 10 people globally, of whom approximately 300 million are affected by depression. At least half of the people affected by depression remain untreated. Although cognitive behavioral therapy (CBT) is an effective treatment, access to mental health specialists, habitually challenging, has worsened because of the COVID-19 pandemic. Internet-based CBT is an effective and feasible strategy to increase access to treatment for people with depression. Mental health apps may further assist in facilitating self-management for people affected by depression; however, accessing the correct app may be cumbersome given the large number and wide variety of apps offered by public app marketplaces. OBJECTIVE: This study aims to systematically assess the features, functionality, data security, and congruence with evidence of self-guided CBT-based apps targeting users affected by depression that are available in major app stores. METHODS: We conducted a systematic assessment of self-guided CBT-based apps available in Google Play and the Apple App Store. Apps launched or updated since August 2018 were identified through a systematic search in the 42matters database using CBT-related terms. Apps meeting the inclusion criteria were downloaded and assessed using a Samsung Galaxy J7 Pro (Android 9) and iPhone 7 (iOS 13.3.1). Apps were appraised using a 182-question checklist developed by the research team, assessing their general characteristics, technical aspects and quality assurance, and CBT-related features, including 6 evidence-based CBT techniques (ie, psychoeducation, behavioral activation, cognitive restructuring, problem solving, relaxation, and exposure for comorbid anxiety) as informed by a CBT manual, CBT competence framework, and a literature review of internet-based CBT clinical trial protocols. The results were reported as a narrative review using descriptive statistics. RESULTS: The initial search yielded 3006 apps, of which 98 met the inclusion criteria and were systematically assessed. There were 20 well-being apps; 65 mental health apps, targeting two or more common mental health disorders, including depression; and 13 depression apps. A total of 28 apps offered at least four evidence-based CBT techniques, particularly depression apps. Cognitive restructuring was the most common technique, offered by 79% (77/98) of the apps. Only one-third of the apps offered suicide risk management resources, whereas 17% (17/98) of the apps offered COVID-19-related information. Although most apps included a privacy policy, only a third of the apps presented it before account creation. In total, 82% (74/90) of privacy policies stated sharing data with third-party service providers. Half of the app development teams included academic institutions or health care providers. CONCLUSIONS: Only a few self-guided CBT-based apps offer comprehensive CBT programs or suicide risk management resources. Sharing of users' data is widespread, highlighting shortcomings in health app market governance. To fulfill their potential, self-guided CBT-based apps should follow evidence-based clinical guidelines, be patient centered, and enhance users' data security.

Can digital health researchers make a difference during the pandemic? Results of the single-arm, chatbot-led Elena+: Care for COVID-19 interventional study.

Background: The current paper details findings from Elena+: Care for COVID-19, an app developed to tackle the collateral damage of lockdowns and social distancing, by offering pandemic lifestyle coaching across seven health areas: anxiety, loneliness, mental resources, sleep, diet and nutrition, physical activity, and COVID-19 information. Methods: The Elena+ app functions as a single-arm interventional study, with participants recruited predominantly via social media. We used paired samples T-tests and within subjects ANOVA to examine changes in health outcome assessments and user experience evaluations over time. To investigate the mediating role of behavioral activation (i.e., users setting behavioral intentions and reporting actual behaviors) we use mixed-effect regression models. Free-text entries were analyzed qualitatively. Results: Results show strong demand for publicly available lifestyle coaching during the pandemic, with total downloads (N = 7'135) and 55.8% of downloaders opening the app (n = 3,928) with 9.8% completing at least one subtopic (n = 698). Greatest areas of health vulnerability as assessed with screening measures were physical activity with 62% (n = 1,000) and anxiety with 46.5% (n = 760). The app was effective in the treatment of mental health; with a significant decrease in depression between first (14 days), second (28 days), and third (42 days) assessments: F2,38 = 7.01, p = 0.003, with a large effect size (η2G = 0.14), and anxiety between first and second assessments: t54 = 3.7, p = <0.001 with a medium effect size (Cohen d = 0.499). Those that followed the coaching program increased in net promoter score between the first and second assessment: t36 = 2.08, p = 0.045 with a small to medium effect size (Cohen d = 0.342). Mediation analyses showed that while increasing number of subtopics completed increased behavioral activation (i.e., match between behavioral intentions and self-reported actual behaviors), behavioral activation did not mediate the relationship to improvements in health outcome assessments. Conclusions: Findings show that: (i) there is public demand for chatbot led digital coaching, (ii) such tools can be effective in delivering treatment success, and (iii) they are highly valued by their long-term user base. As the current intervention was developed at rapid speed to meet the emergency pandemic context, the future looks bright for other public health focused chatbot-led digital health interventions.

Development of "LvL UP 1.0": a smartphone-based, conversational agent-delivered holistic lifestyle intervention for the prevention of non-communicable diseases and common mental disorders.

Background: Non-communicable diseases (NCDs) and common mental disorders (CMDs) are the leading causes of death and disability worldwide. Lifestyle interventions via mobile apps and conversational agents present themselves as low-cost, scalable solutions to prevent these conditions. This paper describes the rationale for, and development of, "LvL UP 1.0″, a smartphone-based lifestyle intervention aimed at preventing NCDs and CMDs. Materials and Methods: A multidisciplinary team led the intervention design process of LvL UP 1.0, involving four phases: (i) preliminary research (stakeholder consultations, systematic market reviews), (ii) selecting intervention components and developing the conceptual model, (iii) whiteboarding and prototype design, and (iv) testing and refinement. The Multiphase Optimization Strategy and the UK Medical Research Council framework for developing and evaluating complex interventions were used to guide the intervention development. Results: Preliminary research highlighted the importance of targeting holistic wellbeing (i.e., both physical and mental health). Accordingly, the first version of LvL UP features a scalable, smartphone-based, and conversational agent-delivered holistic lifestyle intervention built around three pillars: Move More (physical activity), Eat Well (nutrition and healthy eating), and Stress Less (emotional regulation and wellbeing). Intervention components include health literacy and psychoeducational coaching sessions, daily "Life Hacks" (healthy activity suggestions), breathing exercises, and journaling. In addition to the intervention components, formative research also stressed the need to introduce engagement-specific components to maximise uptake and long-term use. LvL UP includes a motivational interviewing and storytelling approach to deliver the coaching sessions, as well as progress feedback and gamification. Offline materials are also offered to allow users access to essential intervention content without needing a mobile device. Conclusions: The development process of LvL UP 1.0 led to an evidence-based and user-informed smartphone-based intervention aimed at preventing NCDs and CMDs. LvL UP is designed to be a scalable, engaging, prevention-oriented, holistic intervention for adults at risk of NCDs and CMDs. A feasibility study, and subsequent optimisation and randomised-controlled trials are planned to further refine the intervention and establish effectiveness. The development process described here may prove helpful to other intervention developers.

The Terms of "You(s)": How the Term of Address Used by Conversational Agents Influences User Evaluations in French and German Linguaculture.

Background: Conversational agents (CAs) are a novel approach to delivering digital health interventions. In human interactions, terms of address often change depending on the context or relationship between interlocutors. In many languages, this encompasses T/V distinction-formal and informal forms of the second-person pronoun "You"-that conveys different levels of familiarity. Yet, few research articles have examined whether CAs' use of T/V distinction across language contexts affects users' evaluations of digital health applications. Methods: In an online experiment (N = 284), we manipulated a public health CA prototype to use either informal or formal T/V distinction forms in French ("tu" vs. "vous") and German ("du" vs. "Sie") language settings. A MANCOVA and post-hoc tests were performed to examine the effects of the independent variables (i.e., T/V distinction and Language) and the moderating role of users' demographic profile (i.e., Age and Gender) on eleven user evaluation variables. These were related to four themes: (i) Sociability, (ii) CA-User Collaboration, (iii) Service Evaluation, and (iv) Behavioral Intentions. Results: Results showed a four-way interaction between T/V Distinction, Language, Age, and Gender, influencing user evaluations across all outcome themes. For French speakers, when the informal "T form" ("Tu") was used, higher user evaluation scores were generated for younger women and older men (e.g., the CA felt more humanlike or individuals were more likely to recommend the CA), whereas when the formal "V form" ("Vous") was used, higher user evaluation scores were generated for younger men and older women. For German speakers, when the informal T form ("Du") was used, younger users' evaluations were comparable regardless of Gender, however, as individuals' Age increased, the use of "Du" resulted in lower user evaluation scores, with this effect more pronounced in men. When using the formal V form ("Sie"), user evaluation scores were relatively stable, regardless of Gender, and only increasing slightly with Age. Conclusions: Results highlight how user CA evaluations vary based on the T/V distinction used and language setting, however, that even within a culturally homogenous language group, evaluations vary based on user demographics, thus highlighting the importance of personalizing CA language.

Elena+ Care for COVID-19, a Pandemic Lifestyle Care Intervention: Intervention Design and Study Protocol.

Background: The current COVID-19 coronavirus pandemic is an emergency on a global scale, with huge swathes of the population required to remain indoors for prolonged periods to tackle the virus. In this new context, individuals' health-promoting routines are under greater strain, contributing to poorer mental and physical health. Additionally, individuals are required to keep up to date with latest health guidelines about the virus, which may be confusing in an age of social-media disinformation and shifting guidelines. To tackle these factors, we developed Elena+, a smartphone-based and conversational agent (CA) delivered pandemic lifestyle care intervention. Methods: Elena+ utilizes varied intervention components to deliver a psychoeducation-focused coaching program on the topics of: COVID-19 information, physical activity, mental health (anxiety, loneliness, mental resources), sleep and diet and nutrition. Over 43 subtopics, a CA guides individuals through content and tracks progress over time, such as changes in health outcome assessments per topic, alongside user-set behavioral intentions and user-reported actual behaviors. Ratings of the usage experience, social demographics and the user profile are also captured. Elena+ is available for public download on iOS and Android devices in English, European Spanish and Latin American Spanish with future languages and launch countries planned, and no limits on planned recruitment. Panel data methods will be used to track user progress over time in subsequent analyses. The Elena+ intervention is open-source under the Apache 2 license (MobileCoach software) and the Creative Commons 4.0 license CC BY-NC-SA (intervention logic and content), allowing future collaborations; such as cultural adaptions, integration of new sensor-related features or the development of new topics. Discussion: Digital health applications offer a low-cost and scalable route to meet challenges to public health. As Elena+ was developed by an international and interdisciplinary team in a short time frame to meet the COVID-19 pandemic, empirical data are required to discern how effective such solutions can be in meeting real world, emergent health crises. Additionally, clustering Elena+ users based on characteristics and usage behaviors could help public health practitioners understand how population-level digital health interventions can reach at-risk and sub-populations.

Corrigendum: Elena+ Care for COVID-19, a Pandemic Lifestyle Care Intervention: Intervention Design and Study Protocol.

[This corrects the article DOI: 10.3389/fpubh.2021.625640.].

The effects of acute work stress and appraisal on psychobiological stress responses in a group office environment.

BACKGROUND: The high prevalence of office stress and its detrimental health consequences are of concern to individuals, employers and society at large. Laboratory studies investigating office stress have mostly relied on data from participants that were tested individually on abstract tasks. In this study, we examined the effect of psychosocial office stress and work interruptions on the psychobiological stress response in a realistic but controlled group office environment. We also explored the role of cognitive stress appraisal as an underlying mechanism mediating the relationship between work stressors and the stress response. METHODS AND MATERIALS: Ninety participants (44 female; mean age 23.11 ± 3.80) were randomly assigned to either a control condition or one of two experimental conditions in which they were exposed to psychosocial stress with or without prior work interruptions in a realistic multi-participant laboratory setting. To induce psychosocial stress, we adapted the Trier Social Stress Test for Groups to an office environment. Throughout the experiment, we continuously monitored heart rate and heart rate variability. Participants repeatedly reported on their current mood, calmness, wakefulness and perceived stress and gave saliva samples to assess changes in salivary cortisol and salivary alpha-amylase. Additionally, cognitive appraisal of the psychosocial stress test was evaluated. RESULTS: Our analyses revealed significant group differences for most outcomes during or immediately after the stress test (i.e., mood, calmness, perceived stress, salivary cortisol, heart rate, heart rate variability) and during recovery (i.e., salivary cortisol and heart rate). Interestingly, the condition that experienced work interruptions showed a higher increase of cortisol levels but appraised the stress test as less threatening than individuals that experienced only psychosocial stress. Exploratory mediation analyses revealed a blunted response in subjective measures of stress, which was partially explained by the differences in threat appraisal. DISCUSSION: The results showed that experimentally induced work stress led to significant responses of subjective measures of stress, the hypothalamic-pituitary-adrenal axis and the autonomic nervous system. However, there appears to be a discrepancy between the psychological and biological responses to preceding work interruptions. Appraising psychosocial stress as less threatening but still as challenging could be an adaptive way of coping and reflect a state of engagement and eustress.