Pathological and non-pathological hikikomori: social media use, digital engagement, and therapeutic implications.

Introduction: Hikikomori is traditionally defined as a form of pathological social withdrawal marked by extreme social isolation in one's home, leading to significant functional impairment or distress. However, shifts in working and study habits since COVID-19 have introduced the concept of 'non-pathological hikikomori' to describe individuals who are isolated in their homes but do not experience functional impairment or distress. Hikikomori are frequent users of the internet and social media, which raises interesting questions regarding the relationship between social withdrawal and physical withdrawal. This study examined whether social media use differs by hikikomori status (pathological vs. non-pathological) and phase (early [<3 months], pre- [3-6 months], full [6+ months]). Method: A cross-sectional study recruited 1,420 self-identified frequent internet users (aged 18-25) via Prolific, who completed a questionnaire on their social media activity (time spent; type of communication), hikikomori status (pathological/non-pathological), and phase (early/pre/full). Of these, 1,235 identified as hikikomori (Mage = 21.5, SD = 2.2; females = 661, males = 572, undisclosed = 2). Within this group, 455 were classified as pathological hikikomori (early = 113, pre = 151, full = 191), while 780 were non-pathological (early = 179, pre = 201, full = 400). Results: Pathological hikikomori used significantly more social media platforms than non-pathological hikikomori (4.16 vs 3.84: F(1,1224)=20.05, p<.001, ηp²=.016). In terms of phase, full hikikomori (3.82) used fewer social media platforms than early (4.01) and pre (4.13) hikikomori, (F(2,1224)=7.19, p<.001, ηp²=.012: early and pre hikikomori did not differ from each other). The interaction between pathological status and phase was not significant [F(2,1224)=1.28, p=.278, ηp²=.002]. Social media platforms were not used for more time by pathological compared to non-pathological hikikomori, but there were differences in how the social media platforms were used. Regarding communication style, across all phases, pathological hikikomori consistently engaged with others via TikTok and YouTube significantly more than non-pathological hikikomori. Using TikTok and YouTube, Pathological hikikomori sent more messages [F(1,958)=8.77, p=.003, ηp²=.009; F(1,1161)=21.50, p<.001, ηp²=.018; respectively] and received more messages [F(1,958)=13.15, p<.001, ηp²=.014; F(1,1161)=21.37, p<.001, ηp²=.018; respectively], had more targeted messages [F(1,958)=8.49, p=.004, ηp²=.009; F(1,1161)=20.77, p<.001, ηp²=.018; respectively], had more stylised messages [F(1,958)=13.60, p<.001, ηp²=.014; F(1,1161)=24.13, p<.001, ηp²=.020; respectively] and had more broadcast messages [F(1,958)=7.58, p=.006, ηp²=.008; F(1,1161)=13.22, p<.001, ηp²=.011; respectively] than non-pathological hikikomori. Conclusion: These findings suggest a complex relationship between social media use and social withdrawal. Future research should explore whether communication through YouTube and TikTok is linked to social isolation and whether these platforms could serve as intervention tools to support pathological hikikomori. Importantly, as the sample consisted of self-identified frequent internet users, generalizability to the broader population of hikikomori should be treated with caution.