Abstract
INTRODUCTION: Dementia comprise a broad spectrum of cognitive declines affecting 47 million people worldwide, with numbers projected to reach 131 million by 2050. Predominantly associated with older adults, dementia can also impact younger individuals, having a significant impact on daily functioning of the affected patients, relatives, caregivers and the socioeconomic system. Recent research underscores the utility of social media, particularly X (previously designed as Twitter), in understanding public perceptions and sentiments related to neurological disorders. Despite some initial studies have explored social perceptions of dementia in X, broader and deeper analysis of this condition is still warranted. MATERIALS AND METHODS: In this retrospective study, we collected and examined all tweets posted in English or Spanish from 2007 to 2023 that mentioned dementia and compare the information with other highly representative neurological disorders like migraines, epilepsy, multiple sclerosis, spinal cord injury, or Parkinson's disease. We developed a codebook to analyze tweets, classifying them by themes such as trivialization, treatment perceptions, and etiopathogenesis. Manually categorized tweets trained machine learning models, BERTWEET for English and BETO for Spanish, which then classified larger datasets with high accuracy. Statistical analysis, including ANOVA, Kruskal-Wallis, and chi-square tests, was conducted to explore linguistic and cultural differences in perceptions of neurological disorders, with results visualized. RESULTS: Our study reveals that dementia is by far the most frequently discussed neurological disorder on X. Likewise, this condition appears to be the most trivialized neurological disorder in Spanish tweets and the second most trivialized in English tweets, with notable differences in geolocation data. Additionally, we found significant differences in perceptions of dementia treatment and associated sentiments between Spanish and English tweets. Furthermore, our study identified varying perceptions of medical content (etiology) and non-medical content (positive/negative experiences and aid requests) related to dementia and other neurological disorders, unveiling a complex landscape of these topics on X. CONCLUSIONS: This study explores the importance of X as a social platform for addressing various critical issues related to dementia, comparing it with other neurological disorders in English and Spanish tweets. Future research could further investigate the valuable role of social media in understanding public perceptions and needs regarding dementia and neurological disorders among X users.