Answers are grounded in this paper and cite their sources.
Virtual reality (VR) is increasingly permeating medical practice, facilitating new health interventions and diagnostic tools to delay or alleviate cognitive aging. However, it remains unclear whether fully automated virtual memory tasks match the efficacy and physiological safety of traditional interventions administered by human providers. To address this, we evaluated an odour-triggered autobiographical memory task across four discrete conditions. Selected as strategic points along the latent „virtuality continuum”, these conditions model the step-by-step translation of clinical memory tasks—from reality human-facilitated real-world care (VR0), augmented care using multisensory VR (VR1) to fully autonomous, text-guided VR (VR2) and virtual agent-mediated interactions (VR3). Subjective sense of presence was equivalent across all VR environments (VR1–VR3), confirming a controlled baseline of visual immersion. Subjects recalled comparable numbers of autobiographical memories in all conditions, but the virtual social environment (VR3) triggered the longest recall durations. All VR environments elicited memories with more positive emotional valence than real-person delivery. Critically, salivary levels of the social hormone oxytocin were significantly decreased after tasks performed in automated environments (VR2 and VR3) compared to the human-guided VR1. These findings confirm VR’s high potential for memory training while highlighting a possible reduction in physiological social activation when replacing human providers with virtual agents. As virtual agents expand in daily healthcare, our data underscore the importance of assessing physiological social responses, particularly for prolonged interventions.
No takes yet. Share an insight, caveat, or question.
Papasteri et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: