This case–control study was conducted at the Behavioral Medicine Department of Sultan Qaboos University Hospital (SQUH), a tertiary care teaching hospital in Muscat, Oman. The study was nested within a larger cross-sectional investigation that aimed to identify Hikikomori-like idiom of distress (HLID) among adults attending the department. Initial screening was carried out using the Hikikomori Questionnaire-25 (HQ-25), alongside the collection of socio-demographic information. Participants were subsequently invited to complete a structured diagnostic interview using the HiDE-I to allow for accurate case classification. Following diagnostic assessment, a subset of participants was selected by convenience sampling to undergo neuropsychological testing using the Cambridge Neuropsychological Test Automated Battery (CANTAB), which assessed executive function, working memory, attention, and response inhibition. Hikikomori case status in the present study was determined using the clinician-administered HiDE-I rather than HQ-25 screening scores alone.

The initial cross-sectional sample consisted of 1,200 Omani nationals aged between 18 and 59 years, recruited from outpatients attending the Behavioral Medicine Department at SQUH and their accompanying attendees. Of these, 454 participants completed the HiDE-I structured diagnostic interview. From this group, 150 participants were selected to complete CANTAB testing, including 85 patients and 65 attendees. Cognitive performance outcomes were examined across multiple domains. The importance of each cognitive domain was evaluated using standardized β coefficients derived from adjusted regression models. ( Figure 1).

Hikikomori cases were identified using the Hikikomori Diagnostic Evaluation Interview (HiDE-I), a structured clinical assessment tool that aligns with international diagnostic criteria for prolonged social withdrawal ( Teo et al., 2023).

For the purposes of this neurocognitive case-control study, a broad HiDE-I case definition was used. This definition classified participants as Hikikomori cases if they met criteria for any HiDE-I category: pathological Hikikomori, at-risk Hikikomori, or Hikikomori resemblance. Participants who did not meet criteria for any of these three categories were classified as non-cases. This broader definition was deliberately adopted because cognitive differences associated with social withdrawal may not be limited to fully pathological Hikikomori but may also be present in earlier or subthreshold stages of withdrawal.

Pathological Hikikomori cases were those who had been socially withdrawn for six months or more, with significant psychological distress or functional impairment. At-risk Hikikomori referred to individuals showing early signs of social withdrawal but not meeting full diagnostic criteria, while Hikikomori resemblance applied to those exhibiting some withdrawal behaviors but lacking clear impairment.

A total of 54 participants were identified as Hikikomori cases and agreed to participate in the study, yielding a response rate of 93%. All cases were evaluated by two trained mental health professionals, each experienced in assessing social withdrawal disorders. Inter-rater reliability was assessed using percent agreement between two independent clinicians who conducted the HiDE-I interviews. Initial agreement on Hikikomori classification across the three subtypes was 96%. Discrepant cases were subsequently reviewed through a structured consensus adjudication process involving joint discussion of interview data and clinical criteria, after which full agreement (100%) was reached. This consensus-based approach was used to ensure diagnostic accuracy rather than to inflate reliability estimates.

Control participants were recruited from the Behavioral Medicine Department at SQUH and confirmed to be Hikikomori-free through the Hikikomori Diagnostic Evaluation Interview (HiDE-I). To be included, participants were required to have no history of prolonged social withdrawal and not meet criteria for any HiDE-I subcategory, including pathological, at-risk, or resemblance Hikikomori. A total of 96 control participants consented to participate, yielding a response rate of 87%.

Inter-rater reliability for control classification was assessed using percent agreement between two independent clinicians who conducted the HiDE-I interviews. Initial agreement on control status was 95%. Discrepant cases were subsequently reviewed through a structured consensus adjudication process involving joint discussion of interview data and clinical criteria, after which full agreement (100%) was reached. This consensus approach was applied to ensure accurate exclusion of Hikikomori rather than to inflate reliability estimates.

Hikikomori cases and controls were assessed using standardized screening and diagnostic tools. The Arabic validated version of the Hikikomori Questionnaire (HQ-25), which has demonstrated excellent internal consistency (Cronbach’s alpha = 0.91), was used to assess social withdrawal ( Al-Sibani et al., 2023). Sociodemographic data were also collected using a structured questionnaire administered in the previous cross-sectional study, providing initial data on social withdrawal patterns and participant characteristics.

In addition to Hikikomori-specific assessments, participants underwent cognitive testing using the Cambridge Neuropsychological Test Automated Battery (CANTAB) ( J. Fray et al., 1996). CANTAB is a computerized neuropsychological assessment tool widely used in psychiatric research to measure executive function, attention, working memory, and processing speed. Testing was conducted via touchscreen devices, with responses recorded automatically. A trained examiner provided standardized instructions to ensure comprehension. The testing was conducted in a quiet environment, and headphones were provided when necessary to minimize auditory distractions.

CANTAB assessed multiple cognitive domains relevant to this study. Executive function was measured using the Intra/Extra Dimensional Set Shift (IED), which evaluates cognitive flexibility, and the One Touch Stockings of Cambridge (OTS), which assesses planning and problem-solving skills. Working memory was tested using the Spatial Working Memory (SWM) and Spatial Span (SSP) tasks. Attention and processing speed were assessed through the Rapid Visual Information Processing (RVP) task, while Pattern Recognition Memory (PRM) was used to evaluate visual memory recall. The CANTAB tasks and corresponding outcome measures used in this study are summarized in Table 1.

Table 2 presents the socio-demographic characteristics of participants who underwent CANTAB testing. Of the 150 participants included, 54 were classified as Hikikomori cases and 96 as non-cases based on the HiDE-I broad definition. Patients constituted 56.7% of the sample (n = 85), while attendees accounted for 43.3% (n = 65).

Among Hikikomori cases, the majority were female (70.4%) and aged 18–29 years (77.8%), compared with 62.5% and 67.7%, respectively, among non-cases. A higher proportion of cases were unemployed (85.2%) compared to non-cases (67.7%). Differences were also observed across family income categories, with cases more frequently represented in lower income brackets.

No substantial differences were observed between cases and non-cases with respect to marital status, parental status, educational attainment, or governorate of residence. Distributions across governorates reflected the clinic-based sample, with Muscat and Al Batinah accounting for the largest proportions in both groups.

Table 3 presents comparisons of cognitive performance between Hikikomori cases and non-cases. Across the full sample, Hikikomori cases demonstrated significantly poorer performance in visual processing accuracy, response latency, immediate recall, delayed recall, and planning efficiency (p < 0.05). No significant group differences were observed for planning latency, spatial working memory errors, or strategy use.

In subgroup analyses, patients with Hikikomori showed significantly poorer performance across multiple domains, including visual processing accuracy, memory recall, and planning efficiency. Among attendees, significant differences were observed for visual processing accuracy and planning efficiency; however, differences in immediate and delayed recall were not statistically significant. Supplementary nonparametric analyses (Mann–Whitney U tests) produced comparable results, supporting the robustness of these findings.

Table 4 presents the crude and adjusted regression models assessing the association between Hikikomori status and cognitive performance across all participants, patients, and attendees. Among all participants (N = 150), Hikikomori status was significantly associated with lower visual processing accuracy, slower response latency, poorer immediate and delayed recall, and reduced planning efficiency after adjusting for sociodemographic factors. Among patients, significant associations were found for greater spatial working memory errors, poorer memory recall, and reduced planning efficiency. Among attendees, lower visual processing accuracy and poorer memory recall remained significant. Non-significant domains included planning latency, strategy use, and spatial memory errors.

Table 5 presents the standardized β coefficients from the adjusted regression model ranked by relative importance. The strongest associations with Hikikomori status were observed for visual processing accuracy (RVPA) and planning efficiency (OTSPSFC), which showed comparable standardized effect sizes (β ≈ −0.45, p = 0.001 for both). These were followed by response latency (RVPMDL) (β = 0.266, p = 0.001), delayed recall (PRMPCD) (β = −0.255, p = 0.003), and immediate recall (PRMPCI) (β = −0.201, p = 0.019). All remaining cognitive domains showed non-significant association.

Our results indicate that Hikikomori individuals experience executive function deficits, particularly in planning efficiency and response inhibition, which suggests difficulties in cognitive flexibility, decision-making, and problem-solving. These impairments are well-documented in psychiatric conditions such as schizophrenia, bipolar disorder, and anxiety disorders ( Bourke et al., 2012; Gau & Shang, 2010), which research has found are comorbid in individuals with Hikikomori ( Cai et al., 2023; Chauliac et al., 2017). CANTAB has been established as a reliable tool for assessing executive function ( Luciana, 2003). Prior research indicates that deficits in response inhibition and attentional shifting, commonly seen in social anxiety disorder, may contribute to social withdrawal ( Pupi et al., 2025; Rhodes et al., 2005). Given the high prevalence of social anxiety in Hikikomori ( Kato et al., 2024), these cognitive difficulties may reinforce avoidance behaviors ( Cai et al., 2023; Lo et al., 2023). In addition to deficits in planning efficiency, Hikikomori cases also exhibited slower response times, suggesting broader impairments in processing speed and cognitive efficiency that may further exacerbate withdrawal tendencies.

In this study, planning efficiency and visual processing accuracy showed the strongest associations with Hikikomori, reflecting substantial impairments in goal-directed behavior and in the interpretation of environmental cues necessary for social interaction. These findings are consistent with evidence showing that executive dysfunction and attentional processing difficulties are linked to reduced social functioning and withdrawal in clinical populations ( Bourke et al., 2012; Kim et al., 2014; Luciana & Nelson, 2002). The minimal associations observed for spatial working memory and strategy use suggest that Hikikomori is characterized by selective rather than global cognitive impairment, a pattern also reported in recent neuropsychological research on Hikikomori ( Santona et al., 2023) and in studies demonstrating that only specific executive domains may be affected rather than overall cognitive capacity ( Rhodes et al., 2005). This selective vulnerability highlights the importance of interventions that target planning, cognitive flexibility, and visual–attentional processing, rather than assuming a broad cognitive decline among Hikikomori individuals.

Our study showed that Hikikomori individuals exhibited impairments in immediate and delayed memory recall, suggesting difficulties in processing and retrieving information, potentially affecting academic performance, work responsibilities, and social interactions. Memory deficits have been observed in individuals with social anxiety disorder ( Airaksinen et al., 2005), and dysfunction in the medial temporal lobe and prefrontal cortex has been linked to social cognition impairments ( Luciana & Nelson, 2002). Given that Hikikomori individuals often display heightened sensitivity to social judgment ( Sakamoto et al., 2005), difficulties in retaining and recalling social information may reinforce withdrawal tendencies ( Kubo et al., 2022).

We found that Hikikomori individuals demonstrated significant deficits in visual processing accuracy, after adjusting for socio-demographic variables, suggesting a specific vulnerability in interpreting environmental cues critical for social engagement. In contrast, no significant differences were observed in spatial working memory errors, planning latency, or strategy use, indicating that Hikikomori does not reflect global cognitive dysfunction. These findings align with research showing that psychiatric conditions often affect distinct cognitive domains ( Bourke et al., 2012; Kim et al., 2014; Luciana & Nelson, 2002).

Given that social anxiety disorder (SAD) is one of the most common comorbidities in Hikikomori ( Lin et al., 2022; Pupi et al., 2025; Tateno et al., 2012), it is plausible that attentional and cognitive flexibility impairments—key features of SAD—may contribute to persistent withdrawal behaviors. SAD has been linked to difficulties in processing social cues and regulating attention and emotion ( Bourke et al., 2012), which may explain the heightened social disengagement observed in Hikikomori cases ( Bowker et al., 2019; Chauliac et al., 2017).

Furthermore, neuroimaging studies have shown that chronic social withdrawal is associated with changes in brain regions responsible for visual processing and attention, including the hippocampus, ventral striatum, and prefrontal cortex ( Zovetti et al., 2021). These findings align with recent work on COVID-19-related isolation, which underscores the broader neurological and psychological consequences of prolonged social disengagement ( Roza et al., 2020).

This study demonstrated that Hikikomori individuals showed deficits in executive function, memory recall, and attention, while their spatial working memory remained intact. This suggests that not all cognitive domains are equally affected ( Luciana & Nelson, 2002). While spatial memory was preserved, impairments in attention shifting and planning may still interfere with social and occupational functioning ( Rhodes et al., 2005). These findings highlight the importance of targeted interventions addressing executive function deficits, rather than assuming global cognitive impairments in Hikikomori individuals ( Santona et al., 2023). Furthermore, subgroup analyses among patients and attendees revealed consistent impairments in visual processing accuracy and planning efficiency, highlighting the robustness of these findings across different clinical and non-clinical populations ( Bowker et al., 2019; Siedlecki et al., 2019) .

While previous research has established gender-based cognitive differences in the general population ( Camarata & Woodcock, 2006; Herlitz & Rehnman, 2008), our study did not find significant gender differences in memory recall or executive function among Hikikomori individuals. This suggests that social withdrawal may affect both genders similarly. Previous neuropsychological research also indicates that cognitive impairments in psychiatric conditions are not necessarily gender-dependent ( Grissom & Reyes, 2019). Although our study did not observe normative gender trends, previous research in the general population has shown that females often perform better in tasks involving attentional control and inhibition, while males tend to show advantages in spatial processing speed ( Giofre et al., 2024; Siedlecki et al., 2019). These normative trends may serve as a reference point for future studies exploring whether prolonged social withdrawal disrupts or attenuates gender-based cognitive patterns.

Our analysis revealed that Hikikomori individuals were more likely to be unemployed, consistent with research linking long-term social withdrawal to reduced workforce participation ( Nonaka & Sakai, 2021). While marital status and income level were not statistically different, Hikikomori cases tended to report lower income, suggesting that economic stressors may contribute to withdrawal behaviors. The role of socioeconomic barriers in Hikikomori highlights the need for interventions that address both cognitive and environmental factors. Cognitive disparities may also intersect with broader sociocultural influences, such as educational and occupational pathways shaped by gender-role identification and self-efficacy beliefs ( Nakajima et al., 2019; Riley et al., 2016). These factors could influence the cognitive reserves of individuals at risk for social withdrawal, compounding the effects of isolation and further limiting re-engagement with work or education.

Given the overlap between Hikikomori and social anxiety disorder, interventions focusing on cognitive remediation may also help reduce avoidance behaviors and improve social engagement ( Bourke et al., 2012; Kato et al., 2024). Given known cognitive gender differences in verbal memory, inhibition, and spatial ability in the general population ( Goel & Tripathi, 2024; Riley et al., 2016), future research should examine whether targeted interventions can be tailored to individual cognitive strengths and weaknesses across genders within Hikikomori populations.

Several limitations should be considered. First, the case–control design precludes causal inference; cognitive differences may precede social withdrawal or emerge because of prolonged disengagement. Longitudinal studies are needed to clarify directionality. Second, multiple cognitive domains were examined, raising the possibility of type I error due to multiple comparisons. Although results showed a coherent pattern across key domains, findings should be interpreted cautiously. Third, participants were recruited from a hospital-based behavioral medicine clinic, introducing potential selection bias. While inclusion of non-patient attendees broadened representation, results may not fully generalize to community samples. Fourth, although assessors administering CANTAB were blinded to case status, psychiatric comorbidities—particularly depression and anxiety—were not statistically controlled in the cognitive analyses. As such, observed differences may reflect comorbidity-related cognitive effects rather than Hikikomori-specific mechanisms. Finally, medication status was not systematically assessed. Psychotropic medications may influence attention, processing speed, and memory, representing a potential confound.

Future research should integrate neuroimaging techniques to identify neural correlates of Hikikomori-related cognitive dysfunction, examine cultural variations across populations ( Kubo et al., 2022), and distinguish between primary Hikikomori and cases with psychiatric comorbidities. Given the high rates of co-occurring conditions such as social anxiety disorder, depression, and personality disorders among Hikikomori cases ( Teo & Gaw, 2010), it remains unclear whether the observed cognitive deficits are specific to Hikikomori itself.