Background Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder characterized by developmentally inappropriate inattention and/or hyperactivity-impulsivity. Diagnosis requires a comprehensive, multi-informant clinical assessment; performance on a single attention task is neither necessary nor sufficient. This cross-sectional study examined the preliminary convergent validity, reliability, and group discrimination of Attention Robots, a serious game designed to assess attention-related performance. Methods Participants were 117 children and adolescents aged 6–18 years: 47 with a clinician-established ADHD diagnosis and 70 community comparison participants without a recorded ADHD diagnosis. Attention Robots scores were examined alongside CARAS-R and CogniFit measures using correlations, age- and sex-adjusted group comparisons, split-half reliability, and exploratory ROC analyses. Results Attention Robots scores were associated with CARAS-R inhibitory-control and several CogniFit measures. After adjustment for age and sex and correction for multiple comparisons, participants with ADHD showed lower Attention Robots concentration and working-speed scores than community comparison participants, whereas error percentage did not differ between groups. Among the CogniFit outcomes, only visual perception remained significantly different between groups after Holm correction. The game-only ROC model showed limited discrimination (AUC = 0.61; sensitivity = 0.28; specificity = 0.80). A model that also included sex yielded an AUC of 0.73, with sensitivity of 0.57 and specificity of 0.76; however, the higher AUC cannot be attributed to the game because sex was unevenly distributed between groups. Conclusions Attention Robots may provide complementary information about attention-related performance, but the present two-gate, single-site findings do not support stand-alone screening or diagnosis. Independent validation with consistently verified, sex-balanced, clinically heterogeneous samples is required.
Individual cognitive behavioural therapy (CBT) for social anxiety disorder is effective, but CBT has not become widespread enough. Since young people with social anxiety also experience above a certain level distress and functional impairment, the development of readily available CBT-based self-help interventions is urgently needed. The objective was to evaluate the effectiveness of two five-week, unguided CBT-based self-help interventions in reducing gaze-related anxiety and social anxiety symptoms among university students with elevated gaze-related anxiety. A three-arm, randomised controlled trial was conducted in Japan from April to December 2025. University students with gaze-related anxiety were randomly assigned to bibliotherapy (Bib) or unguided digital self-help interventions (DSI), or to a six-month waitlist control. Social anxiety symptoms were evaluated using several self-report scales, and assessments were conducted at baseline, post-intervention, and at two follow-ups. Of 252 university students enrolled, 244 were included in the modified intention-to-treat analyses, which used linear mixed-effects models. For the primary outcome, the adjusted mean difference in the Gaze Anxiety Rating Scale at post-intervention was −10.09 (95% CI: −14.61 to −5.56, Hedges' g = −0.66) for Bib and − 6.95 (95% CI: −12.11 to −1.79, Hedges' g = −0.47) for DSI, both showing a significantly greater reduction compared to the waitlist control. Similarly, for a secondary outcome, the adjusted mean difference in the Liebowitz Social Anxiety Scale at post-intervention was −13.57 (95% CI: −20.86 to −6.29, Hedges' g = −0.55) for Bib and − 12.47 (95% CI: −20.78 to −4.17, Hedges' g = −0.53) for DSI, which were also significantly higher than the waitlist control. The effectiveness of these two interventions in reducing social anxiety symptoms compared to the waitlist control was maintained at the 3-month and 6-month follow-ups. These findings support brief unguided CBT-based self-help as a potentially accessible low-intensity option for university students with elevated social anxiety symptoms.
While the increasing use of Large Language Models (LLMs) as eHealth tools facilitates rapid information retrieval, it also carries the risk of credibility inflation for medically inaccurate content. The aim of the study was to examine the effects of the form of presentation of health information (LLM response vs. online news article vs. online forum) and its truthfulness on perceived credibility judgments. The study (N = 405) employed a mixed-design experiment, with presentation format as a between-subjects factor and true vs false health information as a within-subjects factor. The results showed that although participants were generally able to distinguish between true and false information, the form of presentation significantly moderated this process with a medium effect size (η2p = 0.07). Specifically, both LLM and online forum elevated credibility rating of false medical information. The findings have important implications for understanding cognitive mechanisms underlying credibility judgments in digital environments and for the design of safe and responsible eHealth tools.
Background Frequent episodic and chronic migraine are associated with disability, psychological distress, and reduced quality of life. This study evaluated the long-term effects of I AM, an individually tailored, therapist-guided, acceptance-based, Internet-delivered cognitive behavioral therapy (ICBT) programme combined with treatment as usual (TAU) in Swedish primary care. Objective The aim was to evaluate the effects of acceptance-based Internet-delivered cognitive behavioral therapy + TAU compared with TAU alone on migraine frequency, pain intensity, migraine-related disability, and mental health and well-being. Methods In this randomized controlled study, 154 participants with frequent episodic or chronic migraine were randomized to I AM + TAU or TAU alone. Outcomes were assessed at baseline, post-treatment, and at 6- and 12-month follow-ups. Intervention effects were evaluated using time × condition interactions, with baseline as the reference timepoint. Results For migraine-related disability (HIT-6), the time × condition interaction was significant at 12 months (β = −3.406, 95% CI [−5.970, −0.841], p = .010), indicating a greater reduction from baseline in the I AM + TAU group compared with TAU alone. For mental health and well-being (MHC-SF), significant time × condition interactions were observed at post-treatment (β = 5.472, 95% CI [0.071, 10.873], p = .047) and 12 months (β = 5.563, 95% CI [0.176, 10.950], p = .043), indicating greater improvements in the I AM + TAU group. The interaction at 6 months was not significant for either HIT-6 or MHC-SF. No significant time × condition interactions were observed for number of monthly migraine days or pain intensity at post-treatment, 6 months, or 12 months. Conclusions I AM combined with TAU resulted in greater reduction in migraine-related disability at 12 months and in mental health and well-being at post-treatment and 12 months compared with TAU alone. No differences between the groups were observed on number of monthly migraine days or pain intensity. These findings suggest that acceptance-based ICBT may provide benefits primarily in reducing the functional burden of migraine and supporting psychological well-being rather than in reducing migraine frequency or pain intensity.
Purpose Digital interventions for cannabis use typically produce small short-term effects that fade over time. This three-arm randomized controlled trial examined whether integrating mindfulness into an established digital cognitive behavioural therapy (CBT) self-help intervention improves the durability of outcomes. Methods We randomized 301 frequent adult cannabis users to Mindfulness CANreduce (mindfulness-enhanced CBT), Standard CANreduce (CBT only), or a waitlist control group with continued access to care as usual. The primary outcome was self-reported cannabis use days in the past 30 days, assessed at 6 weeks, 3 months, and 6 months, and analysed on an intention-to-treat basis using zero-inflated negative binomial mixed models with multiple imputation. Main results Retention was 51.8% at 6 weeks and 31.6% at 6 months, and 6-month attrition did not differ across groups. At 6 weeks, both active interventions reduced cannabis use days relative to the control group (IRR = 0.76, 95% CI [0.64, 0.90]; IRR = 0.80, 95% CI [0.67, 0.95]). At 6 months, the Mindfulness CANreduce group reported fewer use days than both the Standard CANreduce group (IRR = 0.52, 95% CI [0.43, 0.64]) and the control group (IRR = 0.46, 95% CI [0.38, 0.56]), with greater improvements in depressive and anxiety symptoms. Complete-case and tipping-point sensitivity analyses supported these findings, which remain conditional on assumptions about the missing data. Conclusion Integrating mindfulness into a fully automated digital intervention was associated with more durable reductions in cannabis use. Given the low retention and engagement, these findings should be interpreted cautiously.
