Kentaro Katahira

Although psychiatric disorders have conventionally been treated categorically, recent research indicates a continuous and hierarchical structure among psychiatric symptoms, with a general psychopathology factor ( p -factor) at the top and several specific factors at lower levels. However, the understanding of the psychometric properties of these factors in the general population remains limited. In this study, by analyzing responses from approximately 1000 non-Western individuals across three datasets to various diagnostic questionnaires, we extracted general and specific factors using diverse hierarchical structures and factor numbers. Comparative analysis revealed distinct properties of the general and specific factors, regardless of model assumptions. Specifically, the general factor explained a greater proportion of shared variance in the data and exhibited higher internal consistency than the specific factors. Consistent with the concept of the p -factor, the general factor was associated with a history of diagnoses across various diagnostic categories. Conversely, the content of the specific factors varied depending on the factor numbers, and some corresponded to traditional diagnostic categories, indicating that specific factors may be advantageous for describing individual differences. Our findings, we believe, provide a scaffold for disentangling the complex structure among psychiatric symptoms.

BACKGROUND Recent mobile-health interventions leverage social relationships, offering opportunities for users to collaborate with, or compete against, other users to improve physical activity (PA). Studies have shown, however, large heterogeneity in the efficacy of social interventions. Also, social components in interventions were often combined with other factors like reward and punishment, which were typically provided to teams that achieved or failed to meet predetermined PA goals. It was therefore unclear whether social components work alone and how the heterogeneity could be explained. OBJECTIVE This 3-arm randomized controlled trial investigated the efficacy of collaboration and competition in improving PA. We also explored whether pre-intervention prosocial attitudes (e.g., tendency of helping others) would explain individual differences in participants’ responses to the interventions. METHODS Inactive adults (ages 20–64; 95 women) participated in the study as a pair of families or friends. Each pair was randomized to receive a control (to self-monitor their own step counts on an activity tracker, n = 50), collaboration (to achieve a shared step goal within a pair, n = 50), or competition manipulation (to surpass the pair’s step count, n = 50). No reward was provided even when a goal was achieved. Following a 1-week baseline period, participants received relevant interventions for three weeks. Step count was monitored throughout the study period and questionnaires were administered at the pre- and post-intervention timing. RESULTS The collaboration group showed significantly larger step counts than the control group over the 3-week intervention period (786.42 steps, 95%CI [29.16, 1543.68], d = 0.21, P = 0.04), whereas the competition group did not differ from the control (601.26 steps, 95%CI [-145.46, 1347.98], d = 0.16, P = 0.11). Baseline prosocial attitudes were associated with increases in self-reported levels of PA (r = 0.40, P = 0.02) but not in step count (r = 0.00, P = 0.99) among individuals allocated to the collaboration group. CONCLUSIONS Results indicate that collaboration within a pair is an effective add-on to self-monitoring even without external reward. Prosocial attitudes explain (part of) individual differences in participants’ responses to the collaboration manipulation, which can be an important tailoring variable when prescribing a social intervention. CLINICALTRIAL https://osf.io/2abg9/?view_only=15cbd6283d174ccb9f0c569b7fed494e

Mobile health technology plays an important role in improving physical activity (PA). However, commercial healthcare applications for smartphones (apps) have poor retention, and understanding how people adopt and integrate app use in daily life is critical. We investigated the use patterns of PA apps and explored the usage styles that are predictive of continuation and discontinuation of use and of changes in PA levels over time. We analyzed 2-wave longitudinal survey data concerning commercial PA-app use, which included 4465 respondents (mean age = 50.7; 1932 women) identified as PA-app users at baseline. The participants completed a questionnaire regarding how and for what purpose they used the apps. A 6-month follow-up survey was administered that asked participants about their current app use and PA levels. At baseline, 2737 were identified as long-term users of a PA app (ie, use for more than 6 months). Long-term users reported appreciating the lifestyle management aspects (eg, constant accessibility to health information, tracking, and monitoring PA), whereas short-term users indicated that they appreciated their app’s distraction and building relationships (eg, finding like-minded peers) aspects. Prospective analyses (logistic regressions) demonstrated that lifestyle management was associated with continuing to use the app and increased PA levels at the 6-month follow-up, whereas distraction predicted discontinuing the use of the app and decreased PA levels. These findings suggest that on-purpose use (ie, using a PA app to improve one’s PA) is the key feature of being in an appropriation state, whereas off-purpose use may hinder app use, leading to less active lifestyles. The implications of appropriation theory and practice are also discussed.

Emotional vulnerability arises from the interplay of genetic variation, cortical network capacity, and subjective processing of adverse experiences, yet these components are rarely examined together in population-scale datasets. Monoaminergic signaling offers an opportunity for understanding such interactions because variation in related genes shapes sensitivity to environmental inputs. Among them, the vesicular monoamine transporter 1 ( VMAT1 ) Thr136Ile variant is notable for its functional effects and its long-term maintenance at intermediate frequencies, a pattern consistent with context-dependent selection. These properties make it an informative marker for investigating how genetic sensitivity interacts with neurobiological and experiential factors to shape affective functioning. Using data from up to 9,625 participants in the Tohoku Medical Megabank Project, a cohort established after the 2011 Great East Japan Earthquake, we investigated how monoaminergic variation, cortical morphology, and traumatic memory shape affective functioning. We found associations between the Thr136Ile variant and negative affectivity and depressive symptoms, consistent with prior reports of heightened emotional reactivity associated with the 136Thr allele. However, magnitude of these effects was substantially amplified in specific experiential and neurobiological contexts. Subjective trauma sensitivity, capturing the discrepancy between earthquake disruption and current traumatic memory, displayed a genotype-dependent association with depressive symptoms, and this relationship was strongest when parietal or insular cortical surface area was smaller. Together, these results identify a multilevel pathway through which monoaminergic variation contributes to emotional vulnerability by interacting with both cortical network capacity and trauma processing. The context-dependent influence of monoaminergic variation may further contribute to maintenance of affective diversity in human populations.

In psychology and neuroscience, computational models representing competing hypotheses are often fitted to behavioural data and compared to evaluate their plausibility. In a recent article, Piray proposed a general framework for estimating the statistical power of such model comparisons, focusing on random-effects Bayesian model selection, and suggested that many previous studies may be insufficiently powered. Piray further noted that a large number of studies rely on fixed-effects Bayesian model selection, assuming that the same model applies to all participants, and argued that this practice can lead to extremely high false-positive rates. These conclusions raise concerns about the validity of findings based on computational model selection. In this commentary, however, we argue that Piray’s analyses depart from realistic model-selection scenarios. In particular, his power analysis involves assumptions that are, in different respects, overly pessimistic or overly optimistic, potentially biasing power estimates and similar issues may arise in the simulations used to evaluate fixed-effects Bayesian model selection.

BACKGROUND In mobile healthcare, text messages play an important role in improving physical activity. Recent studies have developed message banks based on theories, including the behaviour change technique (BCT) taxonomy. However, little evidence is available for individual differences (i.e. who responds to what BCTs recommended in messages), which are crucial for optimizing message delivery. OBJECTIVE We investigated how individuals perceive messages supporting physical activity and what clusters of individuals are identified by their responses. METHODS Japanese-speaking adults (N = 2859; mean age = 54.5, SD = 17.5; 1486 women) were presented with messages conveying different BCTs, and rated how motivational each message was. The motivation ratings were subjected to latent class analysis to identify clusters of individuals per motivation rating. RESULTS Seven clusters were identified through analysis. One cluster gave the highest ratings to any messages, and two others showed high ratings, but some preferences for specific BCTs (e.g. information about social and environmental consequences). Another cluster showed variable ratings across messages (with salience of consequences rated as the most motivational), and the remaining three gave overall low ratings and typically included inactive individuals in the pre-action stages. CONCLUSIONS These results highlight individual differences in perceived motivations across BCTs, informing what BCTs should be prioritized in delivery. The practical implications for message tailoring are also discussed.

Background: Scales for measuring health literacy and numeracy have been broadly classified into performance-based (objective) and self-reported (subjective) scales. Both types of scales have been widely used in research and practice; however, they are not always consistent and may assess different latent constructs. Furthermore, an increasing number of objective measures have been developed, and it is unclear how many latent factors should be assumed. Objective: This study aimed to examine the psychometric properties and factor structure of items assessing objective health literacy across multiple scales and to clarify which aspects of objective health literacy would be correlated with subjective measures, as well as health behaviors and lifestyles. Methods: A total of 5 objective scales (72 items in total) were administered to Japanese-speaking adults (N=16,097; women: 7722/16,097, 48%; mean age 54.89, SD 16.46 years). The analyzed scales included items assessing the numeracy, comprehension, and application of health information, some of which were contextualized for specific diseases, such as diabetes and cancer. Participants’ responses were submitted to exploratory factor analysis, and individual factor scores were calculated to test correlations with subjective health literacy, health behavior, and lifestyle. Results: Exploratory factor analysis identified 3 factors, which were interpreted as conceptual knowledge, numeracy, and synthesis. The conceptual knowledge factor consisted of items about medical word comprehension. All numeracy items loaded onto the same factor, even when contextualized for different diseases. The synthesis factor was characterized by items assessing the ability to read and understand health-related information and make judgments on it using one’s own knowledge. The identified factors showed high interfactor correlations ( r values 0.53‐0.64) and small-to-moderate correlations with subjective health literacy ( r values 0.14‐0.45). Additionally, each factor indicated small positive correlations with healthy diet and nutrition and lower substance use ( r values 0.17‐0.26). Conclusions: Our findings suggest that scales of objective health literacy have at least three latent constructs (ie, conceptual knowledge, numeracy, and synthesis) and that disease specificity is not psychometrically prominent. Each factor has some overlap with subjective health literacy, but overall, subjective and objective health literacy should be interpreted as independent constructs, given the small-to-modest correlations.

Background: The ecological momentary intervention (EMI) is one of the most promising digital—primarily mobile—interventions to enhance physical activity (PA) and other health behaviors. It is a combination of ecological momentary assessment (EMA), in which participants are prompted to indicate their momentary states and ongoing behaviors in daily life, and the just-in-time delivery of interventions tailored to the EMA responses. The EMI has typically been implemented in message-based interventions (eg, activity recommendations tailored to users’ physical locations), but its efficacy and feasibility have not been sufficiently established because of the variability in design and implementation. Objective: This pilot, two-arm, parallel-group randomized controlled trial aimed to be an exemplar of EMI for improving PA and establishing efficacy and feasibility among adults without a habit of PA. Methods: A total of 40 participants (23 women; mean age 45.40, SD 10.50 years) were recruited from among community dwellers in northeast Japan and randomly allocated to the EMI or control group. Each participant wore an activity tracker to monitor their daily step count and heart rate (HR) for 4 weeks (fully automated). Simultaneously, they responded to EMA questions about the current weather, location, and social context 3 times during the daytime and an additional evening question about motivations for PA each day. Only the EMI group received messages tailored to their responses to EMA, recommending more active alternative behaviors suited to EMA-reported contexts. Results: Participants wore a Fitbit device for 90.3% (21.66/24 hours per day) of the study period (mean 90.3, SD 10.0), and no dropouts were observed. The EMI group showed no significant improvement in the self-reported amount of PA ( P =.44), step count ( P =.24), or motivation for PA (from P =.11 to P =.91) compared with the control group. However, the EMI group showed a significantly larger increase in the minutes of 40% HR reserve, a measure of moderate or high intensity of PA (mean 16.03, 95% CI 3.76-28.29; Cohen d =0.20‐0.41; P =.02 for the follow-up weeks). The intervention was rated as marginally useful and satisfactory, and approximately half of the participants expressed a willingness to continue the intervention. The timing of the EMA prompting was considered inappropriate. Conclusions: These findings suggest that the EMI with messages tailored to EMA-reported contexts was not effective in increasing the amount or motivation for PA but may increase the intensity as assessed by the HR. The intervention aimed to help individuals implement small but slightly more active behaviors in their daily routine, which may not accompany prominent body movements but may be reflected in the increased HR. Marginal feasibility indicates that the intervention has sufficient scope for improvement, particularly in terms of prompt timing.

Balancing efficiency and flexibility in collective decision-making is increasingly critical in modern societies characterized by rapid sociocultural and technological change. Recent research in cognitive neuroscience has proposed two contrasting computational algorithms for social learning: value shaping (VS) and decision biasing (DB). VS posits that others’ choices serve as “pseudo-rewards” that directly shape an observer’s valuations, leading them to prefer popular options even in the absence of outcome feedback. In contrast, DB confines the influence of social information to behavior—observers may imitate popular actions, but they update their valuations solely through personal experience. Although both algorithms facilitate individual adaptation under uncertainty, their interactive dynamics and group-level consequences remain largely unexplored. To address this gap, we developed computational models of VS and DB within a reinforcement learning framework and conducted agent-based simulations to examine collective performance in uncertain and dynamically changing environments. The results reveal a trade-off: VS enables rapid convergence and high efficiency in stable contexts, whereas DB promotes greater adaptability under environmental volatility. These differences are amplified in larger groups, particularly under strong majority influence. Importantly, evolutionary analyses indicate that both learning types can coexist stably, allowing their complementary strengths to enhance group performance. Together, our findings provide a computational and evolutionary account of how social learning can both enhance and impair collective intelligence—and suggest design principles for fostering resilient collective decision systems in human and AI societies facing rapid change.