Psychiatry · Medical Literature
Association Between Loneliness and Hikikomori Symptoms. A Systematic Review and Meta-Analysis.
Abstract
A systematic review and meta-analysis were conducted to investigate the association between loneliness and hikikomori symptoms. We searched three electronic databases (plus hand search). Our review followed the PRISMA guidelines, and a PROSPERO registration was made. We included observational studies determining the association between loneliness and hikikomori symptoms. Eight cross-sectional studies were finally included. Our psychometric meta-analysis revealed a mean corrected correlation of 0.68 (95% CI: 0.62 to 0.74). The mean corrected correlations of loneliness with the subscales socialization, isolation, and emotional support were 0.54 (95% CI: 0.50 to 0.58), 0.62 (95% CI: 0.60 to 0.64), and 0.71 (95% CI: 0.67 to 0.75), respectively. The mean corrected correlation for the association between loneliness and hikikomori symptoms was 0.66 (95% CI: 0.62 to 0.70) among European countries, whereas it was 0.90 (95% CI: 0.81 to 0.99) among Asian countries. Five of the studies were of good quality, and three were of fair quality. Our work identified a strong, positive association between loneliness and hikikomori symptoms. Upcoming longitudinal studies are required to clarify the directionality between such factors. There is also a clear need for research in regions that have been neglected so far.
Clinical relevance
Is loneliness significantly and meaningfully associated with hikikomori symptoms (pathological social withdrawal) across observational studies, and does the strength of this association differ by geographic region or symptom subscale? __SNAPSHOT__ {"studyDesign":"Systematic review with psychometric meta-analysis (PRISMA-compliant; PROSPERO-registered)","setting":"International; included studies from European and Asian countries; institutional settings not reported","sampleSize":"8 cross-sectional studies included (individual study-level n values not reported in the abstract)","population":"General adult and/or community populations in whom both loneliness and hikikomori symptoms were measured with validated instruments; specific inclusion criteria beyond observational design and measurement of the target constructs not detailed in the abstract","intervention":"Exposure: loneliness (assessed via validated self-report measures)","comparator":"Not applicable (observational correlational design; no intervention or active comparator)","primaryOutcome":"Association between loneliness and hikikomori symptoms, operationalized as a corrected correlation coefficient (psychometric meta-analysis); subscale-level correlations for socialization, isolation, and emotional support also examined","followUp":"Not applicable (all included studies were cross-sectional)","fundingSource":"Not reported"}
Key findings
The psychometric meta-analysis—using the Hunter-Schmidt artifact-correction framework, which adjusts observed correlations for measurement unreliability and sampling error to yield estimated true-score correlations—yielded a mean corrected correlation of r = 0.68 (95% CI: 0.62–0.74) for the overall association between loneliness and hikikomori symptoms. This magnitude falls in the 'large' range by conventional benchmarks (Cohen's r ≥ 0.50). At the subscale level, mean corrected correlations were: socialization r = 0.54 (95% CI: 0.50–0.58), isolation r = 0.62 (95% CI: 0.60–0.64), and emotional support r = 0.71 (95% CI: 0.67–0.75). A notable geographic moderation effect was observed: the pooled corrected correlation was r = 0.66 (95% CI: 0.62–0.70) in European countries versus r = 0.90 (95% CI: 0.81–0.99) in Asian countries—a difference of approximately 0.24 correlation units (calculated), though formal between-group significance testing for this moderator is not reported in the abstract. Heterogeneity statistics (I², τ²) and formal p-values are not reported in the abstract. NNT/NNH is not applicable to correlation-based analyses. Adverse events are not applicable to this observational design. Study quality: five studies rated good quality, three rated fair quality; no studies were rated poor quality.
Practice implications
This meta-analysis establishes a strong positive cross-sectional association between loneliness and hikikomori symptoms, with the corrected correlation of r = 0.68 indicating that loneliness accounts for approximately 46% of the variance in hikikomori symptom severity (r² = 0.46, calculated). The emotional support subscale showed the strongest association (r = 0.71), suggesting that perceived deficits in emotional connectedness may be particularly central to the hikikomori construct. The substantially higher corrected correlation observed in Asian versus European samples (r = 0.90 vs. 0.66) raises important questions about cultural mediation—potentially reflecting the deeper sociocultural significance of social role fulfillment and shame in Japanese and other East Asian contexts—but causal inference is not possible. CRITICAL LIMITATIONS ON INFERENCE: (1) All 8 included studies were cross-sectional, meaning the direction of causality cannot be established. Loneliness may drive social withdrawal, social withdrawal may intensify loneliness, or both may share common upstream causes (e.g., depression, social anxiety, trauma). (2) The psychometric meta-analysis reports 'corrected' (disattenuated) correlations, which adjust for measurement unreliability. While methodologically appropriate for estimating true construct-level relationships, these corrected values are inherently higher than observed correlations and may overstate the magnitude of the observable clinical association in practice. (3) Only 8 studies were included, and aggregate n is not reported, limiting precision and generalizability. (4) Hikikomori is not yet a formal DSM-5-TR or ICD-11 diagnosis, and measurement heterogeneity across studies (e.g., use of HQ-25 versus other tools) may introduce construct variability. (5) The Asian subsample correlation of r = 0.90 approaches the theoretical maximum for disattenuated correlations and should be interpreted with caution given the small number of contributing studies and possible publication bias. This summary is based solely on the published abstract; the full text was not available for review, which limits assessment of heterogeneity statistics, individual study characteristics, and moderator analysis methods. __LIMITATIONS__ All 8 included studies used cross-sectional designs, precluding any causal or temporal inference about the loneliness–hikikomori relationship.,The total pooled sample size is not reported in the abstract, limiting assessment of statistical power and precision.,The psychometric (Hunter-Schmidt) meta-analysis yields corrected (disattenuated) correlations that adjust for measurement error; these are systematically higher than raw observed correlations and represent an idealized construct-level estimate rather than what a clinician would observe with typical instruments.,Only 8 studies were included, raising concerns about the robustness of subgroup analyses (e.g., Asian vs. European samples); formal moderation significance testing is not reported.,Publication bias assessment results are not described in the abstract.,Heterogeneity statistics (I², Q, τ²) are not reported in the abstract.,Hikikomori lacks standardized DSM/ICD diagnostic criteria; variation in operationalization across studies introduces construct heterogeneity.,Funding source is not reported, precluding assessment of potential conflicts of interest.,The Asian subsample corrected correlation of r = 0.90 (95% CI: 0.81–0.99) is unusually high and may reflect sparse data, a small number of contributing studies, range restriction, or publication bias.,The studies may not adequately represent older adults, younger adolescents, or individuals with severe psychiatric comorbidities who are often underrepresented in general population online surveys.,Loneliness measurement instruments were not standardized across studies, introducing heterogeneity in the exposure construct. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ This PRISMA-compliant psychometric meta-analysis of 8 cross-sectional studies identifies a strong positive association between loneliness and hikikomori symptoms (corrected r = 0.68, 95% CI: 0.62–0.74), with the emotional support subscale showing the highest correlation and the association appearing substantially stronger in Asian versus European populations. Causality cannot be established from cross-sectional data, and the corrected correlations represent disattenuated true-score estimates rather than directly observed values. Clinicians should routinely screen for loneliness in patients presenting with social withdrawal, but targeted longitudinal studies are necessary before this association can inform specific preventive or therapeutic algorithms for hikikomori.