Human-Edited Generative AI-Assisted Multiple-Choice Questions in Postgraduate Family Medicine: Blinded Cross-Sectional Comparative Psychometric Study.

Loneliness has increasingly been recognized as an important social determinant of health and well-being , affecting a substantial proportion of the population in many countries . Previous studies, conducted mainly in Western countries, have reported associations between loneliness and increased risk of all-cause mortality, including deaths from cancer and cardiovascular disease, as well as a range of adverse mental health outcomes such as depression and anxiety . These associations may be mediated through multiple, interrelated pathways, including biological processes (such as neuroendocrine and immune dysregulation), behavioral changes (such as reduced engagement in health-promoting behaviors), and psychological stress responses . In light of these broad health implications, the World Health Organization has identified loneliness and social disconnection as emerging global public health concerns and has called for comprehensive actions to strengthen social connections . Despite the growing interest in the health effects of loneliness, several limitations remain in the existing epidemiological literature. Much of the prior research has focused on objective endpoints such as mortality and cardiovascular events , whereas associations with more proximal indicators of health burden in daily life―including patient-reported health and symptom burden, as well as the broader prevalence of chronic conditions―have been less thoroughly examined . In addition, individuals experiencing loneliness may be less likely to engage in healthcare-seeking behaviors, such as visiting medical institutions or utilizing preventive health services, even when they are aware of health problems . These issues may be underestimated in studies that rely solely on diagnosis-based outcomes, without considering patient perspectives and healthcare-seeking behaviors. Furthermore, much of the existing evidence on the association between loneliness and health has been derived from Western countries, and its generalizability to East Asian populations, which differ in social structure and cultural context , remains uncertain. In Asian countries, prior studies have often relied on measures of depressive symptoms as proxy indicators of loneliness, whereas large-scale epidemiological studies using standardized instruments that directly assess loneliness are limited . Therefore, there is a clear knowledge gap regarding the comprehensive evaluation of the associations between loneliness and a broad range of health-related outcomes in non-Western settings. Japan provides a unique context in which to examine these associations. Under a universal health coverage system with a free-access model, healthcare barriers at the system level are relatively low . However, even in this setting of high healthcare accessibility, it remains unclear how loneliness relates to multiple dimensions of health and healthcare engagement, beyond specific outcomes such as coronavirus disease 2019 (COVID-19) infection and hospitalization . Therefore, in this study, we used the University of California, Los Angeles (UCLA) 3-Item Loneliness Scale to assess loneliness in a large, nationwide, internet-based survey of the general adult population in Japan. We examined the associations between loneliness and a range of health-related outcomes, including subjective health status, symptom burden, chronic disease prevalence, and healthcare-seeking behaviors.