Efficacy of a mobile-based approach-avoidance task training (PROTECTapp) for problematic usage of the internet in young adults: A randomized controlled trial
Clinical Snapshot
PICO Framework
| P — Population | University students aged approximately 18–30 years (mean age 22.00 years, 69.6% women) with elevated levels of problematic internet use (PUI) |
| I — Intervention | 3-week mobile-based approach-avoidance task (AAT) training via PROTECTapp, comprising multiple training sessions targeting automatic approach biases toward internet-related stimuli |
| C — Comparator | Waitlist control group (no active intervention during the 3-week study period) |
| O — Outcomes | Primary: PUI severity and internet-related craving. Secondary: motivation to change, psychopathological symptoms, academic functioning. Follow-up at 3 and 12 weeks post-intervention (intervention group only) |
Bottom Line
This RCT provides preliminary evidence that PROTECTapp, a 3-week mobile-based approach-avoidance task training, produces a large reduction in problematic internet use severity (d = -0.80) in university students compared to a waitlist control. The intervention is brief, scalable, and demonstrates an acceptable safety profile with infrequent adverse events. Importantly, however, the trial carries significant methodological limitations: the absence of blinding means self-reported primary outcomes are vulnerable to expectancy and demand effects; the waitlist comparator does not control for non-specific engagement effects; and the predominantly female, single-institution German sample constrains generalisability. The theoretically central outcome of internet-related craving showed no significant improvement, raising questions about the proposed mechanism of action. Secondary psychopathological and academic outcomes were also non-significant. For Australian clinicians, PROTECTapp represents a promising but not yet practice-ready adjunctive tool. Replication in larger, more diverse samples with active sham controls, objective behavioural outcomes, and blinded assessment is needed before confident clinical recommendation. In the interim, it may be considered as a low-risk, low-burden adjunct within university student wellbeing programmes.
Key Findings
P Value: p = .003 for primary PUI outcome; motivation to change (ambivalence p = .020; taking steps p = .002); craving and psychopathological outcomes non-significant (ps > .05)
Effect Size: Cohen's d = -0.80 (large effect by conventional benchmarks)
Primary Outcome: PUI severity significantly reduced in the PROTECTapp group compared to waitlist control at post-intervention assessment
Nnt Or Sensitivity: NNT not calculable from abstract data; mean session completion = 35.52 sessions over 3 weeks; adverse event rate = 7.1%; user satisfaction M = 18.32/32
Confidence Interval: 95% CI [-1.23, -0.38] — entirely favouring intervention, does not cross zero
Clinical Application
High feasibility for deployment as a low-threshold, self-administered digital intervention. The mobile-based format requires only a smartphone and eliminates need for clinician contact, making it suitable for stepped-care models. The 3-week duration and average 35.52 sessions suggest moderate but achievable engagement. Integration into university health services or student wellbeing platforms is plausible. Moderate satisfaction scores indicate room for optimisation of user experience. Problematic internet use is not currently listed as a formal diagnostic category in DSM-5-TR or ICD-11 (internet gaming disorder is included in ICD-11 as a condition for further study), and there is no specific PBS-listed pharmacotherapy or TGA-approved digital therapeutic for PUI in Australia. RACGP guidelines on digital health and mental health in young people support the use of evidence-based digital interventions as adjuncts to care. PROTECTapp could complement existing Australian university mental health services (e.g., headspace on-campus, university counselling) as a low-threshold, scalable tool. However, the app would require cultural and linguistic adaptation, Australian regulatory consideration under the TGA's Software as a Medical Device (SaMD) framework, and replication in Australian samples before routine clinical recommendation. Young adults (approximately 18–30 years) with elevated problematic internet use, particularly university students. Caution is warranted in applying findings to males, older adults, non-university populations, or those with severe comorbid psychiatric conditions, as these groups were underrepresented or not specifically characterised.
Abstract
BACKGROUND AND AIMS: Problematic usage of the internet (PUI) has been linked to impaired mental health and academic functioning in young adults. This randomized controlled trial evaluated the efficacy of a 3-week mobile-based approach-avoidance task (AAT) training (PROTECTapp) for reducing PUI in university students. METHODS: Ninety-two participants (Mage = 22.00 years, 69.6% women) with elevated levels of PUI were randomized to the PROTECTapp intervention (n = 45) or a waitlist control group (n = 47). Primary outcomes were PUI severity and internet-related craving. Secondary outcomes included motivation to change, psychopathological symptoms and academic functioning. Participants were assessed at baseline and postintervention; the intervention group completed additional 3- and 12-week follow-ups. RESULTS: Intention-to-treat analyses indicated greater reductions in PUI following the PROTECTapp intervention compared to the waitlist (p = .003; d = -0.80, 95% CI [-1.23, -0.38]). No significant effects emerged for craving or broader psychological outcomes (ps > .05), though favorable effects were observed on motivation to change (ambivalence: p = .020; d = -0.24, 95% CI [-0.65, 0.17]; taking steps: p = .002; d = 0.45 [0.04, 0.87]). Satisfaction with the intervention was moderate (M = 18.32 of 32), and participants completed on average 35.52 training sessions. Adverse events were reported infrequent (7.1%). DISCUSSION AND CONCLUSIONS: PROTECTapp is a promising mobile-based intervention to reduce PUI and enhance motivation to change in young adults. Its brevity, scalability, and safety profile highlight its potential as a low-threshold preventive or adjunctive intervention for young individuals at-risk.
References
- 1.Huth, D., Brand, S., & Lindenberg, K. (2026). Efficacy of a mobile-based approach-avoidance task training (PROTECTapp) for problematic usage of the internet in young adults: A randomized controlled trial. Journal of Behavioral Addictions. https://doi.org/10.1556/2006.2026.00041
Related Research
Cognitive, affective & behavioral neuroscience
Emotion detection unveiled: A cognitive-computational synthesis of physiological models, machine learning, and datasets.
3 Aug 2026
Cognitive, affective & behavioral neuroscience
A review of current capabilities and future directions in machine-based emotion recognition
2 Aug 2026
JMIR mHealth and uHealth
Machine Learning Frameworks for Wearable-Based Stress Modeling in Naturalistic Settings: Scoping Review
1 Aug 2026
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service