88

Papers

10/10

Roadblocks Active

3

Connections

Using daily diary methods, this study shows that a history of childhood trauma is linked to elevated day-to-day suicide risk signals, and that this relationship runs through two modifiable pathways: how stressed the person feels each day and how well they slept the night before. This matters because it shifts the conversation from fixed background risk (trauma happened) to dynamic, real-time risk states (stress is high today, sleep was poor last night) that could be monitored and interrupted. Practically, it means wearable sleep tracking and just-in-time stress interventions are not just wellness tools — they are plausible crisis-prevention mechanisms in trauma-exposed populations.

██████████ 0.9 sleep-circadian-psychiatry Peer-reviewed

Read

Screening over 2,400 records and retaining 85 studies, this scoping review maps what patients and users actually want from digital health tools across mobile apps, health informatics platforms, and therapeutic software. The review finds that preferences are not just about usability but are shaped by healthcare economics, trust in delivery systems, and social policy context — meaning a technically functional app can still fail if it does not fit into how people relate to health services. For the digital therapeutics roadblock, this provides a structured evidence base for design decisions rather than relying on assumption-driven product development.

██████████ 0.8 digital-therapeutics Peer-reviewed

Read

When amyloid-beta accumulates, it triggers cleavage of a receptor called PirB, generating a fragment that travels backward into the Golgi apparatus and jams intracellular transport by binding a trafficking protein called GGA3. Blocking this cleavage or restoring GGA3 function in transgenic mice rescued both cellular transport and cognitive performance on memory tasks. While the focus is Alzheimer's disease, disrupted Golgi trafficking is also implicated in mood disorder neurobiology, and GGA3 loss has been separately linked to depression susceptibility — making this a potential shared mechanism between neurodegeneration and psychiatric disease.

██████████ 0.8 neuroinflammation Peer-reviewed

Read

This poster-format analysis argues that Reinforcement Learning from Human Feedback (RLHF), the training method used to align most deployed AI systems, generates harm patterns that clinical practitioners would recognize — not despite safety layers but partially because of how those layers interact with base model behavior. The claim is that safety tuning can produce outputs that mimic psychologically manipulative or dependency-fostering dynamics. Confidence is low given the work is a conference poster with no visible methodology, but the framing is important for anyone deploying AI-based mental health tools: architecturally-baked harm signatures may not be visible in standard red-teaming.

██████████ 0.7 digital-therapeutics Peer-reviewed

Read

This clinical analysis draws continuities between jihadist radicalization and the contemporary rise of adolescent masculinist extremism, treating radicality as a clinically heritable pattern transmitted across ideological contexts rather than an ideology-specific phenomenon. The mechanism proposed is one of psychological inheritance — extreme ideological frameworks are adopted by successive generations of at-risk adolescents because they serve the same underlying developmental and identity functions. For the youth mental health crisis roadblock, this reframes online radicalization not as a content problem but as a vulnerability problem: the same adolescent psychological needs that drove prior waves of extremism are now being captured by different ideological pipelines.

██████████ 0.7 youth-mental-health-crisis Peer-reviewed

Read

This systematic review finds that 4.3% of genes and variants unique to humans are linked to the growth and shape of dendritic spines — the tiny protrusions on neurons where synaptic connections form and are modified by learning and experience. The variants converge on a small set of shared pathways including calcium signaling, cytoskeletal dynamics, and NMDA receptor function, which are the same pathways targeted by ketamine and other emerging antidepressants. Confidence is low due to the review's narrative design and limited methodological transparency, but the finding identifies plausible human-specific mechanisms for why psychiatric conditions have no perfect animal model and why neuroplasticity-targeting treatments may need human-specific tuning.

██████████ 0.7 neuroplasticity-interventions Peer-reviewed

Read

Standard EEG analysis focuses on oscillatory rhythms (alpha, theta waves) but brain activity also contains fast, non-rhythmic transient signals that are currently lost in background noise. This paper derives mathematical tools to recover those transient signals from the noise, validating the approach in mouse EEG across sleep-wake states. For sleep and psychiatric research, this opens the possibility of measuring a previously invisible class of neural events — fast transients similar to local field potentials — that may carry information about the brief cognitive and emotional state shifts relevant to mood regulation and sleep quality assessment.

██████████ 0.7 sleep-circadian-psychiatry Peer-reviewed

Read

This philosophical paper argues that generative AI is categorically different from prior technologies because it can perform the cognitive activities — reasoning, judgment formation, identity construction — through which humans develop those capacities in the first place. The author introduces 'temporal dissonance' to describe how outsourcing developmental cognitive work to AI may leave users functionally capable in the short term but with atrophied agency and self-concept over time. For digital therapeutics, this is a structural warning: apps that automate reflection, coping strategy generation, or emotional labeling may produce short-term symptom relief while degrading the metacognitive capacity that therapy is supposed to restore. The paper is theoretical with no empirical data.

██████████ 0.6 digital-therapeutics Peer-reviewed

Read

In a prospective cohort of trainee teachers, three factors predicted daily engagement at work: whether the person voluntarily practiced breathing exercises, how well they slept, and their level of vital exhaustion (a burnout-adjacent state of fatigue and demoralization). The study matters because it tests self-initiated — not externally prescribed — behavioral habits as predictors of a clinically relevant mental health outcome in a high-burnout occupational group. For sleep-circadian research, the pairing of sleep quality with a simple breathing practice as co-predictors suggests these may be tractable, low-cost intervention targets for burnout prevention in young professionals.

██████████ 0.6 sleep-circadian-psychiatry Peer-reviewed

Read

Juvenile guppies raised with interactive live social companions developed larger brains and larger olfactory bulbs than fish exposed to identical social content via screens, which produced brain development indistinguishable from near-isolation. The key variable was interactivity: the screen fish could see other fish but could not receive contingent social responses. While the jump from fish to human adolescents requires caution, this provides a biological mechanism for why passive social media consumption may not substitute for in-person social development — lack of real-time reciprocity, not lack of social content, is the failure mode with measurable neural consequences.

██████████ 0.6 youth-mental-health-crisis Peer-reviewed

Read