Unmasking concealed depression through the digital leakage effect: A narrative review and theoretical framework with...

A persistent paradox haunts modern psychiatric epidemiology: despite decades of investment in mental health screening, suicide rates have not declined appreciably, with up to 60% of suicide decedents denying suicidal ideation during their final clinical encounter. Individuals with concealed depression-particularly in East and Southeast Asian populations where cultural norms of face (mianzi), emotional restraint, and somatization discourage disclosure-systematically evade detection by standard self-report instruments such as the PHQ-9.

This narrative review introduces the Digital Leakage Effect (DLE), a three-stage theoretical framework linking daytime psychosocial masking to nocturnal digital behavioral leakage through passive smartphone sensor data, with systematic mapping to DSM-5 criteria and cultural calibration for Asian populations.

Through targeted synthesis of literature across cognitive psychology, affective neuroscience, and digital phenotyping, this review integrates ego depletion theory, cognitive load theory, and allostatic load frameworks to construct a testable model for detecting concealed depression via passive sensing.

The DLE framework comprises: Stage I (Daytime)-active psychosocial masking depletes prefrontal executive resources; Stage II (Evening)-ego depletion and circadian decline diminish inhibitory control; Stage III (Nocturnal)-subconscious behavioral leakage manifests through sleep fragmentation, keystroke kinematics, and semantic-temporal divergence. All nine DSM-5 depression criteria are mapped to corresponding passive digital biomarkers. Cultural calibration reveals that WEIRD-trained models risk misinterpreting Asian stoicism as pathological masking, while cultural amplification of daytime suppression paradoxically strengthens nocturnal signals. An ethical architecture grounded in federated learning, dynamic informed consent, and independent oversight is proposed, with linkage to Just-in-Time Adaptive Interventions (JITAI) for suicide prevention.

The DLE framework offers a theoretically grounded, culturally calibrated approach to detecting concealed depression through passive digital phenotyping, with potential to reduce false-negative identification in Asian populations and transform suicide prevention from reactive inquiry to proactive safety net.