Digital Phenotyping and Behaviour Change in Addiction Recovery:

Towards Personalized Physical Activity Interventions

Pedro Morouço

, Tânia Caetano

and Eduardo Ramadas

Unit for Development, Research and Training, VillaRamadas, Leiria, Portugal

Keywords: Digital Phenotyping, Addiction Recovery, Physical Activity Promotion, Wearable Sensors, Personalized

Intervention.

Abstract: This position paper explores the integration of digital phenotyping into addiction recovery through a 3-in-1

mobile application that combines adaptive physical activity, psychological support, and mindfulness. Framed

within the clinically applied Change & Grow® Therapeutic Model, the proposal aims to transform real-time

behavioural and physiological data into personalized, responsive interventions for individuals recovering

from substance use disorders. The primary research objective is to assess whether an intelligent digital

companion—driven by wearable data, ecological momentary assessment, and adaptive feedback—can

support relapse prevention and emotional self-regulation in high-vulnerability populations. We describe the

conceptual design of the Digital-PA Loop, a system that interprets users’ daily patterns and delivers tailored

interventions aligned with therapeutic goals. Ethical considerations, clinical integration pathways, and

implementation strategies are discussed in detail. While the app remains under development, a pilot study

involving patients at VillaRamadas is planned to assess feasibility, usability, and early signals of effectiveness.

This proposal seeks to foster interdisciplinary collaboration at the intersection of sport science, digital health,

and psychotherapy, and sets the stage for a data-driven evolution in addiction rehabilitation.

1 INTRODUCTION

Substance use disorders (SUD) continue to pose a

significant challenge to public health, with high

relapse rates undermining long-term recovery (Stokes

et al., 2018). Among non-pharmacological strategies,

physical activity (PA) has consistently demonstrated

benefits across physiological, psychological, and

social domains (Zupet et al., 2020; Cabral et al.,

2024). Regular exercise can alleviate withdrawal

symptoms, improve mood regulation, and reinforce

self-agency (Boswell et al., 2022; Giménez-

Meseguer et al., 2020). Despite this, PA remains

poorly integrated into most treatment programs, often

lacking personalization and long-term adherence

strategies.

The recent evolution of digital phenotyping offers

a novel opportunity to bridge this gap (Ebner-Priemer

& Trull, 2009). By capturing real-time behavioural

and physiological data through smartphones and

https://orcid.org/0000-0002-5956-9790

https://orcid.org/0000-0002-9335-0602

https://orcid.org/0000-0002-4857-6768

wearables, it enables the detection of patterns related

to mood, sleep, activity levels, and relapse

vulnerability. When combined with ecological

momentary assessment (EMA) and adaptive

feedback systems, this technology can support

dynamic, personalized interventions that evolve with

everyone’s recovery process.

This position paper presents a conceptual model

that integrates digital phenotyping with personalized

PA promotion, embedded within the Change &

Grow® Therapeutic Model — a structured five-phase

intervention framework applied clinically at

VillaRamadas. While PA is encouraged within this

model, its digital personalization remains largely

unexplored.

We propose the development of a “3-in-1” mobile

health platform that simultaneously supports: (i)

adaptive physical activity routines; (ii) motivational

and psychological reinforcement; and (iii)

mindfulness-based strategies for emotional regulation.

Morouço, P., Caetano, T. and Ramadas, E.

Digital Phenotyping and Behaviour Change in Addiction Recovery: Towards Personalized Physical Activity Interventions.

DOI: 10.5220/0013739900003988

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 13th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2025), pages 223-229

ISBN: 978-989-758-771-9; ISSN: 2184-3201

223

Rather than a static solution, the envisioned app

functions as a responsive therapeutic companion,

adapting its content and intensity based on users’

physiological signals and self-reported states.

Recent editorial work has emphasized the need

for cross-disciplinary approaches to understanding

and promoting physical activity motivation (e.g.,

Morouço et al., 2024, 2025). Thus, the central

research question guiding this work is: Can a digitally

augmented, real-time behavioural intervention

improve physical activity adherence and emotional

regulation during residential addiction treatment?

To explore this, we aim to: (i) conceptualize the

system architecture using digital phenotyping; (ii)

align it with the Change & Grow® therapeutic

process; and (iii) define a roadmap for clinical

implementation and future feasibility testing.

2 BACKGROUND AND

RATIONALE

SUD are complex conditions that often involve

recurring cycles of relapse, difficulties in self-

regulation, and diminished motivation to change

(Ishino et al., 2020; Colder et al., 2016). Although

conventional therapies can initiate abstinence, they

often struggle to sustain engagement and prevent

relapse over the long term.

In recent years, physical activity (PA) has gained

recognition as a valuable non-pharmacological

strategy to support recovery. Research points to its

benefits for mood stabilization (Mikkelsen et al.,

2017; Muskens et al., 2024), anxiety reduction

(Zhang et al., 2022; Vancampfort et al., 2018),

neuroplasticity (El-Sayes et al., 2019; Dergaa et al.,

2025), and the reinforcement of self-efficacy and

well-being (Wang et al., 2022; Peng et al., 2025).

Despite this evidence, PA remains underutilized in

rehabilitation settings. Common barriers include

fluctuating emotional states, low motivation, and a

lack of personalization—both in design and delivery.

This is where digital phenotyping presents a

compelling opportunity. By collecting continuous

behavioural and physiological data via smartphones

and wearables—such as movement, sleep, heart rate

variability (HRV), and mood—digital phenotyping

enables a real-time understanding of an individual’s

recovery journey. When combined with ecological

momentary assessment (EMA), these insights can

guide timely, tailored interventions that respond to

the user’s daily needs.

Evidence from mental health research shows that

such tools can predict anxiety spikes or

disengagement and support adaptive behavioural

strategies. Still, their potential in addiction

recovery—particularly in reinforcing PA habits—

remains largely untapped.

This position paper proposes leveraging digital

phenotyping not for surveillance, but for

empowerment. The idea is to help individuals

reconnect with their bodies, recognize patterns, and

receive context-sensitive feedback that promotes

motivation and self-regulation. When embedded in a

coherent therapeutic model, the potential for

meaningful behavioural change increases

substantially.

The Change & Grow® Therapeutic Model,

implemented at VillaRamadas, offers such a

structure. This five-phase approach emphasizes

personal development, emotional insight, and

behavioural accountability. Although PA is already

encouraged within the model, it has not yet been

enhanced through technology.

We envision a 3-in-1 digital platform that

integrates adaptive PA routines, motivational content

aligned with therapeutic goals, and mindfulness-

based interventions. By aligning digital feedback with

the user’s physiological signals and emotional states,

the system becomes a responsive companion in the

recovery process.

Unlike existing tools like reSET-O or Just-in-

Time Adaptive Interventions (JITAIs), our proposal

places physical activity at the core of the intervention

loop, fully embedded within a residential therapeutic

framework. This unique combination supports a more

holistic, dynamic, and person-centred path to

recovery.

3 CONCEPTUAL

FRAMEWORKS: THE

DIGITAL-PA LOOP IN

ADDICTION RECOVERY

The core concept underlying this proposal is the

Digital-PA Loop: a closed, adaptive feedback system

that integrates digital phenotyping data with

personalized interventions across three synergistic

pillars — physical activity, motivational-

psychological support, and mindfulness — to support

sustained recovery from addiction. This loop is not

meant to replace the therapist or the therapeutic

process, but to augment it through real-time data,

meaningful insights, and behaviourally adaptive

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

224

prompts that reinforce engagement, autonomy, and

emotional regulation.

3.1 Functional Components of the

Digital-PA Loop

The system comprises four interconnected

components:

(i) Data Acquisition Layer

This layer collects both passive and active data

streams from the user, including:

 Passive sensors:

 Step count, GPS-derived mobility patterns,

heart rate (HR), heart rate variability (HRV),

sleep quality metrics from wearables.

 Active self-reporting:

 EMA-style prompts for mood, cravings,

fatigue, motivation levels, or pain/discomfort.

 Short qualitative check-ins reflecting the user’s

cognitive or emotional state.

 App usage behaviour:

 Frequency of interaction with each module

(exercise, mindfulness, motivation), time of

use, and abandonment patterns.

These inputs form the real-time digital phenotype

of the user — a behavioural and physiological

snapshot that continuously evolves.

(ii) Analysis and Interpretation Layer

Using simple rule-based logic or, in future versions,

machine learning algorithms, the system analyses

incoming data to detect:

 Reduced physical engagement (e.g., <4000

steps for 3 consecutive days)

 Low HRV and poor sleep (as stress proxies)

 Negative mood trends or craving spikes (via

EMA)

 Disengagement with the app (reduced

interaction, skipped exercises)

This analysis triggers a personalized

interpretation of the user’s current state, identifying

moments of vulnerability, opportunity, or need for

reinforcement.

(iii) Adaptive Intervention Engine

Based on the interpreted profile, the system delivers

personalized prompts across the 3-in-1 pillars:

1. Physical Activity:

 Adjusted intensity and volume of PA sessions

 Automatic suggestion of low-intensity

alternatives when fatigue or negative mood is

detected

 “Streak” rewards and milestone notifications

aligned with progress

2. Psychological/Motivational Support:

 Short, value-based reflections aligned with the

Change & Grow® phase the user is in

 Affirmations or therapeutic micro-tasks based

on themes (e.g., responsibility, self-worth,

resilience)

 Nudges tied to personalized goals or past effort

(“You walked more this week than last – keep

building!”)

3. Mindfulness Content:

 Suggested meditation based on time of day or

physiological state (e.g., HRV drop triggers 5-

minute breathing exercise)

 Body scan reminders when self-reported

tension is high

 Grounding techniques when location patterns

suggest social isolation

(iv) Feedback and Reflection Layer

At the end of each day (or week), users receive a

simple visual feedback panel:

 Trends (steps, mood, sleep, app use)

 Positive reinforcements (“You completed 4

mindfulness sessions this week”)

 Reflection prompts (“What did you notice in

your body this week during PA?”)

 Therapist dashboard view (in clinical settings)

for tailoring sessions

The personalization of physical activity routines

can be informed by social-cognitive models such as

Self-Determination Theory or the Transtheoretical

Model, which have demonstrated relevance in

explaining behaviour change in PA contexts

(Rodrigues et al., 2023). Additionally, will be tailored

by certified exercise professionals, based on

individual assessments of baseline physical fitness,

fatigue levels, and health status. Parameters such as

perceived exertion, heart rate variability and

motivation will guide progression or regression

within the program.

3.2 Integration with the Change &

Grow® Therapeutic Model

Each of the five phases of the Change & Grow®

model provides a distinct psychological and

Digital Phenotyping and Behaviour Change in Addiction Recovery: Towards Personalized Physical Activity Interventions

225

emotional landscape. The app dynamically aligns

with these transitions:

Truth & Acceptance:

 Focus on stabilizing routines and normalizing

activity

 Encouraging PA not for performance, but for

reconnection and rhythm

Gratitude:

 Encouraging reflection on positive aspects of

life and recognition of personal achievements

 EMA questions help the user track the

relationship between physical and emotional

states, fostering an attitude of gratitude and

emotional resilience

Love:

 Promoting self-care, compassion, and

strengthening interpersonal relationships

 PA metrics focused on emotional well-being

and affect regulation

Responsibility:

 Users begin to craft personalized routines

 PA metrics tied to self-regulation and progress

ownership

 Introduction of commitment tracking

This modular adaptation ensures that the user

journey is matched by the app’s behavioural logic,

rather than imposing a uniform experience.

3.3 Flowchart of the Digital-PA Loop

[Data Collection]

↓

[Real-Time Analysis] ← (Personal

History + Context)

↓

[Adaptive Intervention Selection]

↓

[3-in-1 Response]

→ Exercise Plan (Dynamic)

→ Mindfulness Practice (Targeted)

→ Motivation Prompt (Contextual)

↓

[User Interaction + Feedback]

↓

[Data Collection (loop restarts)]

This loop represents a continuous learning system,

where the user’s physiological and emotional states

inform the system’s responses, and those responses,

in turn, shape future behaviour.

3.4 Clinical Application: From

Concept to Practice

Within the clinical setting (e.g., VillaRamadas), the

app can be embedded in the daily therapeutic routine:

 Used during morning check-ins to guide PA

plans

 Integrated into one-on-one therapy as a self-

monitoring aid

 Applied during transitional phases (e.g., pre-

discharge planning) to scaffold independent

routines

The app thus acts as a bridge between supervised care

and autonomous health management, ensuring that

therapeutic gains are sustained beyond the residential

setting.

3.5 Future Potential: Towards

Predictive Models

As data accumulates across users, the system could

evolve into a predictive engine, identifying

personalized risk signatures for relapse or

disengagement. For example:

 A combination of low HRV, poor sleep, and

skipped PA over 5 days may trigger a clinical

alert

 Machine learning models could refine

thresholds based on longitudinal trends and

outcomes

Such insights would enable both early intervention

and tailored long-term prevention strategies.

4 CLINICAL AND ETHICAL

CONSIDERATIONS

The integration of digital phenotyping into addiction

recovery offers a promising path toward more

personalized, adaptive, and empowering care. The

proposed 3-in-1 mobile application—combining

physical activity, psychological support, and

mindfulness—has the potential to strengthen the

therapeutic alliance, foster autonomy, and extend the

reach of treatment beyond the walls of the clinic. Still,

its application in vulnerable populations requires a

careful balance between innovation and ethical

responsibility.

Clinically, one of the most relevant strengths of

this model is its capacity to tailor physical activity—

an intervention often delivered generically—

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

226

according to real-time physiological and emotional

data. In many rehabilitation settings, exercise

programs are not adjusted to mood, fatigue, or

motivation. A digital system that recognizes these

fluctuations allows both clinicians and users to co-

create routines that are more relevant, achievable, and

engaging. It can also reinforce therapeutic insights

between sessions, helping individuals carry their

progress into daily life, especially during periods of

transition or vulnerability.

Yet, such benefits must be weighed against

potential risks. Excessive notifications or overly

intrusive feedback could lead to digital fatigue,

avoidance, or emotional reactivity—particularly in

those with a history of surveillance or trauma. For this

reason, the app must adopt a minimalist and

compassionate design philosophy, avoiding language

that suggests failure or non-compliance. All feedback

should be framed to encourage reconnection, not

reinforce shame.

There is also a need to avoid the over-

medicalization of everyday life. While real-time data

can highlight relevant trends, not every variation is

pathological. Outputs must always be interpreted in

clinical dialogue, not treated as standalone

diagnostics. A core principle of this proposal is that

technology should augment, not replace, human care.

Therapists remain central in interpreting digital

patterns and contextualizing them within each

person’s recovery journey.

Ethically, implementation must begin with

transparent and ongoing informed consent. Users

must understand what data is collected, how it is used,

and when it might trigger alerts or recommendations.

Data protection must follow strict privacy protocols,

and therapist access to dashboards must be limited,

consensual, and guided by institutional policy.

Additional attention must be given to digital

equity. The system should work on low-cost devices,

be usable with minimal connectivity, and respect

cultural and linguistic diversity. To prevent

dependency on the tool itself, the app will include off-

ramping features that encourage progressive

autonomy post-discharge.

When thoughtfully designed and ethically

applied, this model transforms digital data into a tool

for reflection, engagement, and self-regulation—

supporting not only clinical progress but a more

embodied and sustainable recovery.

5 IMPLEMENTATION ROADMAP

AND FUTURE PERSPECTIVES

The successful implementation of the proposed

digital system—an adaptive, real-time intervention

platform integrating physical activity, psychological

support, and mindfulness—requires a phased,

interdisciplinary approach. While the conceptual

foundation is robust, translating it into clinical

practice demands careful coordination between

technological development, therapeutic alignment,

and user-centred design.

5.1 Phase 1: Prototype Development

and Feasibility Testing

The initial step involves the co-design of a functional

prototype in close collaboration with clinicians,

psychologists, sports scientists, and individuals in

recovery. This ensures that the technological solution

is not only technically sound but also clinically

meaningful. The prototype should include:

 A wearable-compatible interface to collect

basic physiological and activity data (e.g., HR,

HRV, step count, sleep).

 A mobile application with three core modules:

physical activity, mindfulness, and

motivational support.

 A basic decision engine capable of adapting

prompts based on predefined thresholds (e.g.,

sustained inactivity or negative mood reports).

 A therapist dashboard for supervised

environments like VillaRamadas.

While the present article is conceptual, the system

will be evaluated through a pilot study involving 15

participants in a controlled clinical setting, measuring

feasibility, user experience, and impact on

engagement with physical activity and therapeutic

content.

5.2 Phase 2: Clinical Pilot and

Integration

Following technical validation, the system should be

embedded within the therapeutic workflow of a

structured program such as the Change & Grow®

model at VillaRamadas. This pilot phase will

examine:

 How therapists incorporate app-generated data

into sessions.

 Patterns of user engagement and dropout.

Digital Phenotyping and Behaviour Change in Addiction Recovery: Towards Personalized Physical Activity Interventions

227

 Concordance between physiological data and

therapeutic indicators (e.g., mood, motivation,

relapse risk).

 Perceived usefulness of the system by patients

and staff.

This phase may involve 30–50 participants over 6–12

weeks and can form the basis for a preliminary impact

assessment, particularly regarding physical activity

adherence and emotional self-regulation.

5.3 Technical and Organizational

Requirements

From an infrastructural perspective, implementation

requires:

 Basic wearable integration (e.g., with PLUX or

equivalent bio signal platforms).

 A secure cloud-based backend to process and

store data.

 Data privacy compliance mechanisms (e.g.,

encryption, anonymization, GDPR-aligned

protocols).

 Interoperability with existing clinical records

or digital platforms.

At the organizational level, it is essential to train

clinical teams in the ethical interpretation of digital

data and establish protocols for responding to alerts

or disengagement flags. These workflows must be

sensitive to patient autonomy and avoid clinical

overreaction to system signals. Addressing emotional

regulation and promoting adaptive coping strategies

through digital tools can significantly impact

satisfaction with life, even in working populations

(Rodrigues et al., 2023).

5.4 Vision for Future Development

Beyond the initial pilot, the vision is to evolve the

system into a predictive and adaptive platform,

capable of learning from longitudinal user patterns.

Future iterations may include:

 Machine learning algorithms to personalize

intervention intensity and anticipate relapse

risk.

 Natural language processing (NLP) to interpret

open-ended EMA responses or voice notes.

 Integration of gamification mechanics to

sustain motivation and reinforce therapeutic

goals.

 Use of digital twin models to simulate potential

future scenarios based on behavioural trends.

Such evolution would transform the platform from a

rule-based companion into a dynamic, learning

system—one that evolves with the user and supports

increasingly autonomous recovery.

5.5 Scalability and Replication

The model, while conceived within the clinical

context of VillaRamadas, has potential for broader

application in other residential and outpatient

settings, particularly those that already integrate PA

and mindfulness into care. By preserving modularity

and interoperability, the system can be scaled to other

clinical populations (e.g., depression, anxiety,

burnout) or adapted to community-based health

promotion contexts.

International dissemination may be facilitated

through partnerships with academic institutions,

digital health startups, and therapeutic networks.

Furthermore, its integration into public health

strategies for relapse prevention could be explored,

particularly in systems where digital therapeutics are

gaining regulatory recognition.

6 CONCLUSION

Addiction recovery is a multifaceted process that

requires both structure and flexibility, emotional

support and behavioural reinforcement. In this

landscape, physical activity holds unique potential—

not only for its physiological and psychological

benefits, but as a behavioural anchor that promotes

identity reconstruction, emotional regulation, and

resilience. Yet its impact depends on how

meaningfully it is integrated into everyday

therapeutic practice.

This position paper proposed a digital

intervention model designed to support that

integration through a 3-in-1 mobile health platform,

combining physical activity, mindfulness, and

psychological reinforcement. Grounded in digital

phenotyping and aligned with the Change & Grow®

Therapeutic Model implemented at VillaRamadas,

the proposed system aims to deliver real-time,

personalized interventions that evolve with the user’s

physiological and emotional state.

Rather than a theoretical abstraction, this model is

being prepared for real-world application. The next

phase will involve clinical piloting within a

residential setting, with initial feasibility studies

assessing usability, acceptability, and short-term

effects on activity engagement and emotional

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

228

regulation. These results will guide iterative

refinement and inform a future clinical trial.

Ethical and human-centred design will remain

central throughout this process. The goal is not to

replace care, but to extend it—bridging the gap

between therapy sessions and lived experience. This

vision depends not only on technology, but on the

values that shape its use: empathy, autonomy, and

dignity.

In sum, the Digital-PA Loop represents a

pragmatic, ethically conscious step toward a more

personalized and embodied model of recovery. Its

success will not lie in technological sophistication

alone, but in how well it responds to the complexity—

and humanity—of those it seeks to support.

REFERENCES

Boswell, T., Prochnow, T., Amo, C., Spadine, M., Heinrich,

K., Russell, A., Francis, A., & Patterson, M. (2022).

Exercise in the Treatment of Addiction: A Systematic

Literature Review. Health Education & Behavior, 49,

801-819. https://doi.org/10.1177/10901981221090155.

Cabral, D., Tavares, V., Schuch, F., & Wang, D. (2024).

The role of physical exercise on the brain and cognitive

functions of patients in recovery from substance use

disorder: A narrative review and recommendations for

researchers and practitioners. Mental Health and

Physical Activity. https://doi.org/10.1016/j.mhpa.

2024.100594.

Colder, C., Chassin, L., Sher, K., & Hussong, A. (2016).

Substance Use and Substance Use Disorders.

Development and Psychopathology, 1-65.

https://doi.org/10.1002/9781119125556.DEVPSY319.

Dergaa, I., Dhahbi, W., Omri, A., Ceylan, H., Saad, B.,

Stefanica, V., Ezzdine, L., Chamari, K., & Guelmami,

N. (2025). Physical activity and neuroplasticity in

neurodegenerative disorders: a comprehensive review

of exercise interventions, cognitive training, and AI

applications. Frontiers in Neuroscience, 19.

https://doi.org/10.3389/fnins.2025.1502417.

Ebner-Priemer, U. W., & Trull, T. J. (2009). Ecological

momentary assessment of mood disorders and mood

dysregulation. Psychological Assessment, 21(4), 463–

475. https://doi.org/10.1037/a0017075

El-Sayes, J., Harasym, D., Turco, C., Locke, M., & Nelson,

A. (2019). Exercise-Induced Neuroplasticity: A

Mechanistic Model and Prospects for Promoting

Plasticity. The Neuroscientist, 25, 65 - 85.

https://doi.org/10.1177/1073858418771538.

Ishino, F., McNab, P., Gilreath, T., Williams, F., &

Salmeron, B. (2020). A comprehensive multivariate

model of biopsychosocial factors associated with

opioid misuse and use disorder in a 2017–2018 United

States national survey. BMC Public Health, 20.

https://doi.org/10.1186/s12889-020-09856-2.

Mikkelsen, K., Stojanovska, L., Polenakovic, M., Bosevski,

M., & Apostolopoulos, V. (2017). Exercise and mental

health. Maturitas, 106, 48-56.

https://doi.org/10.1016/j.maturitas.2017.09.003.

Morouço, P., Rogowska, A.M., & Behzadnia, B. (2024)

Editorial: Motivation for physical activity, volume II.

Front. Psychol. 15:1533943. doi:

10.3389/fpsyg.2024.1533943

Morouço, P., Rogowska, A.M., & Pavlova, I. (2025)

Editorial: Motivations for physical activity—Volume

III. Front. Psychol. 16:1666422. doi:

10.3389/fpsyg.2025.1666422

Muskens, L., Kop, W., Weinstein, A., Van Aert, R., &

Donovan, K. (2024). Affective Responses to Acute

Exercise: A Meta-Analysis of the Potential Beneficial

Effects of a Single Bout of Exercise on General Mood,

Anxiety, and Depressive Symptoms. Psychosomatic

Medicine, 86, 486 - 497. https://doi.org/10.

1097/PSY.0000000000001321.

Peng, Y., Liu, H., & Zhao, L. (2025). Physical activity

adherence in behavioral recovery: The role of feedback,

personalization, and digital nudges. Journal of

Behavioral Health Technology, 12(1), 25–38.

Rodrigues, F., Figueiredo, N., Jacinto, M., Monteiro, D., &

Morouço, P. (2023). Social-cognitive theories to

explain physical activity. Education Sciences, 13(2),

122. https://doi.org/10.3390/educsci13020122.

Stokes, M., Schultz, P., & Alpaslan, A. (2018). Narrating

the journey of sustained recovery from substance use

disorder. Substance Abuse Treatment, Prevention, and

Policy, 13, 1-12.

Vancampfort, D., Rebar, A., Firth, J., Kandola, A., Stubbs,

B., Herring, M., & Hallgren, M. (2018). Moving to Beat

Anxiety: Epidemiology and Therapeutic Issues with

Physical Activity for Anxiety. Current Psychiatry

Reports, 20. https://doi.org/10.1007/s11920-018-0923-x.

Wang, K., Li, Y., Zhang, T., & Luo, J. (2022). The

Relationship among College Students’ Physical Exercise,

Self-Efficacy, Emotional Intelligence, and Subjective

Well-Being. International Journal of Environmental

Research and Public Health, 19.

https://doi.org/10.3390/ijerph191811596.

Zhang, L., Jia, Y., Wang, S., Yang, F., So, K., Wang, M.,

Liu, X., Wei, J., Cheng, T., & Yan, L. (2022). Physical

Exercise Prevented Stress‐Induced Anxiety via

Improving Brain RNA Methylation. Advanced Science,

9. https://doi.org/10.1002/advs.202105731.

Zupet, P., Debruyne, A., Loellgen, H., & Bachl, N. (2020).

Physical Activity, Exercise Prescription for Health and

Home-Based Rehabilitation. Sustainability, 12, 10230.

https://doi.org/10.3390/su122410230.

Digital Phenotyping and Behaviour Change in Addiction Recovery: Towards Personalized Physical Activity Interventions

229