Shunsuke NonakapersName, Motohiro SakaipersName (2023date). “The suitability of outing frequency as a definition of hikikomori (prolonged social withdrawal)..” Frontiers in Psychiatrytitle·j.
Object of study — outing frequencyterm and the quality of outings as conditions defining hikikomoriterm·dom
Research question — whether the outing-frequency criteria (1 or 4 days a week) and qualitative indicators of outingsterm adequately predict hikikomoriterm·dom
Contribution — evidence that outings with interpersonal interactionterm·key predict hikikomoriterm·dom while outings without it do not
phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | prolonged social withdrawal (hikikomori) — “Hikikomori, or prolonged social withdrawal, has attracted research attention in Japan since the 1990s and, more recently, in many other countries and regions, including America ( 1 ), Finland ( 2 ), Hong Kong ( 3 – 5 ), Italy ( 6 ), and Spain ( 7 ).” | |
| claim-type | existence — “The hikikomori group was associated with a lower frequency of outings (Low vs. High, Medium vs. High), lower outings with interpersonal interaction, and higher subjective social impairment compared to the recovered and control groups ( Table 4 ).” | |
| method — evidence | — “The data included three datasets: online self-rating, offline self-rating, and parent-rated samples.” | |
| object of study — evidence | — “This study aimed to clarify the relationship between hikikomori tendencies and the frequency and quality of outings to bridge this gap in research.” | |
| research question — evidence | — “Furthermore, we examined whether the criteria of outing frequency (1 or 4 days per week) and qualitative indicators adequately predict hikikomori.” | |
| contribution — evidence | — “An essential finding of this study is the difference that outings with interpersonal interaction predicted hikikomori, but those without interaction did not.” | |
| mechanism | research accessibility How easily people with hikikomori can be reached by research may shape the outing frequencies each sample shows. — “These trends may reflect research accessibility, similar to treatment accessibility.” | mechanisms[0] |
| replicates | Sakai et al. — “This difference supports Sakai et al.’s study ( 22 ) and another report ( 33 ) wherein interpersonal relationships were strongly associated with hikikomori.” | claim_relations[0] |
| extends | Kato et al. — “Furthermore, this study provided findings that partially support the recently proposed outing frequency condition for hikikomori ( 14 , 15 ) but that the association between outing behavior and hikikomori differed depending on whether it involved interpersonal interaction or not.” | claim_relations[1] |
| sense | hikikomori (Ministry of Health, Labour and Welfare definition) - withdrawal that still counts outings without interpersonal interaction as hikikomori — “This differs from the Ministry of Health, Labor, and Welfare’s definition, which includes outings without interpersonal interaction as hikikomori ( 13 ).” | senses[0] |
| sense | hikikomori (the recently proposed criteria) - severity graded by the number of days a week spent outside the home — “In these studies, the severity of hikikomori was linked to outing frequency, here referring to as one’s frequency of going out of the home, as follows: outings on 2–3 days/week implied mild hikikomori.” | senses[1] |
| sense | hikikomori (the authors' conclusion) - defined by low outing frequency together with outings that lack interpersonal interaction — “For integration with previous findings, the results showed that in defining hikikomori and determining its severity, in addition to outing frequency, which is a quantitative indicator, it is necessary to identify whether or not the outing involves interpersonal interaction, which is a qualitative indicator.” | senses[2] |
| relation | hikikomori —same-as→ prolonged social withdrawal — Recent research has employed outing frequency, referred to in this study as one’s frequency of going out of the home, as a condition to define and determine the severity of hikikomori (prolonged social withdrawal). — “Recent research has employed outing frequency, referred to in this study as one’s frequency of going out of the home, as a condition to define and determine the severity of hikikomori (prolonged social withdrawal).” | concept_edges[0] |
| weight | hikikomori —same-as→ prolonged social withdrawal — 0.85 — “Recent research has employed outing frequency, referred to in this study as one’s frequency of going out of the home, as a condition to define and determine the severity of hikikomori (prolonged social withdrawal).” | concept_edges[0] |
| relation | AUC —same-as→ Area Under the Curve — The Area Under the Curve (AUC) was 0.874 between the hikikomori and control groups and 0.797 between the hikikomori and recovered groups. — “The Area Under the Curve (AUC) was 0.874 between the hikikomori and control groups and 0.797 between the hikikomori and recovered groups.” | concept_edges[1] |
| weight | AUC —same-as→ Area Under the Curve — 0.9 — “The Area Under the Curve (AUC) was 0.874 between the hikikomori and control groups and 0.797 between the hikikomori and recovered groups.” | concept_edges[1] |
| relation | VIF —same-as→ Variance Inflation Factor — The Variance Inflation Factor (VIF) was calculated to check for multicollinearity. — “The Variance Inflation Factor (VIF) was calculated to check for multicollinearity.” | concept_edges[2] |
| weight | VIF —same-as→ Variance Inflation Factor — 0.9 — “The Variance Inflation Factor (VIF) was calculated to check for multicollinearity.” | concept_edges[2] |
| replication | original | |
| limitation | Outing frequency was measured cross-sectionally and may be biased by memory. — “This study used a cross-sectional design to determine outing frequency, and findings may have been biased by memory.” | limitations[0] |
| limitation | No reliable and validated instrument exists for the qualitative aspects of outings. — “Even though we used items that have been implemented in several previous studies, the lack of an appropriately reliable and validated instrument is a severe limitation of this study.” | limitations[1] |
| limitation | The offline self-rated sample was too small, and none of its recovered members went out one day a week or less. — “Although data from multiple collection methods were analyzed to reduce sample bias, the sample size was insufficient in the self-reported offline sample, and no recovered individuals indicated an outing frequency of 1 day or less per week.” | limitations[2] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | The outing-frequency condition excluded about 14.5-20.6% of those previously considered to have hikikomori. — “The results revealed that the outing frequency condition excluded about 14.5–20.6% of those previously considered to have hikikomori.” | content_tags.findings[0] |
| finding | Going out freely and outings without interpersonal interaction did not predict hikikomori. — “In contrast, going out freely and outings without interpersonal interaction did not predict whether one was a hikikomori.” | content_tags.findings[1] |
| finding | The cut-off for identifying hikikomori was 4.0-4.7 days outside the home per week. — “This means that the cut-off point for identifying a hikikomori was 4.0–4.7 days outside the home per week.” | content_tags.findings[2] |
| quantity | 397; 72; 784 — — “Data included 397 self-rated online samples, 72 self-rated offline samples, and 784 parent-rated samples.” | content_tags.quantities[0] |
| key-terms | hikikomori, prolonged social withdrawal, outing frequency, interpersonal interaction, subjective social functioning impairment, social isolation, logistic regression, recovered group |
| concept | hikikomori — prolonged social withdrawal — “Hikikomori, or prolonged social withdrawal, has attracted research attention in Japan since the 1990s and, more recently, in many other countries and regions, including America ( 1 ), Finland ( 2 ), Hong Kong ( 3 – 5 ), Italy ( 6 ), and Spain ( 7 ).” | concepts[0] |
| concept | prolonged social withdrawal — not defined in this paper's own words; long-lasting withdrawal from social participation, used here as the gloss of hikikomori — “Recent research has employed outing frequency, referred to in this study as one’s frequency of going out of the home, as a condition to define and determine the severity of hikikomori (prolonged social withdrawal).” | concepts[1] |
| concept | outing frequency — one’s frequency of going out of the home — “Recent research has employed outing frequency, referred to in this study as one’s frequency of going out of the home, as a condition to define and determine the severity of hikikomori (prolonged social withdrawal).” | concepts[2] |
| concept | qualitative indicators of outings — outings with or without interpersonal interaction — “However, many previous studies have focused not only on quantitative indicators of outings (outing frequency, which means leaving their homes) but also on qualitative indicators of outings (outings with or without interpersonal interaction) as a condition for having hikikomori.” | concepts[3] |
| concept | quantitative indicators of outings — outing frequency, which means leaving their homes — “However, many previous studies have focused not only on quantitative indicators of outings (outing frequency, which means leaving their homes) but also on qualitative indicators of outings (outings with or without interpersonal interaction) as a condition for having hikikomori.” | concepts[4] |
| concept | subjective social functioning impairment — not defined in this paper's own words; a person's self-rated difficulty in social participation, scored on a 1-10 scale — “The results showed a weak or no correlation between outing frequency and subjective social functioning impairment.” | concepts[5] |
| concept | hikikomori group — those who currently met the condition for over 6 months — “According to their self-reports on the questions, participants were divided into three groups: the hikikomori group for those who currently met the condition for over 6 months; the recovered group for those who did not currently meet the condition for over 6 months but did so previously; and control group for those who had never met the condition.” | concepts[6] |
| concept | Area Under the Curve — not defined in this paper's own words; the area under a receiver operating characteristic curve, a measure of how well a score separates two groups — “The Area Under the Curve (AUC) was 0.874 between the hikikomori and control groups and 0.797 between the hikikomori and recovered groups.” | concepts[7] |
| concept | AUC — Area Under the Curve — “The Area Under the Curve (AUC) was 0.874 between the hikikomori and control groups and 0.797 between the hikikomori and recovered groups.” | concepts[8] |
| concept | Variance Inflation Factor — not defined in this paper's own words; a statistic calculated to check for multicollinearity among predictors — “The Variance Inflation Factor (VIF) was calculated to check for multicollinearity.” | concepts[9] |
| concept | VIF — Variance Inflation Factor — “The Variance Inflation Factor (VIF) was calculated to check for multicollinearity.” | concepts[10] |
| Area Under the Curve AUC | as written: AUC — “The Area Under the Curve (AUC) was 0.874 between the hikikomori and control groups and 0.797 between the hikikomori and recovered groups.” | content_tags.statistics[0] |
Full text — tagged
The reference list is not part of the text we hold for this paper. Our copy was extracted from the publisher's page and the bibliography did not survive that extraction, so the article below ends before its references. Nothing has been removed by us.
The suitability of outing frequencyterm as a definition of hikikomoriterm·dom (prolonged social withdrawalterm·dom).
Abstract Recent research has employed outing frequencyterm, referred to in this study as one’s frequency of going out of the home, as a condition to define and determine the severity of hikikomoriterm·dom (prolonged social withdrawalterm·dom). However, there is little definitive evidence on this topic. Furthermore, compared to the previous definition, it is unclear how the scope of hikikomoriterm·dom included in the proposed condition differs. This study aimed to clarify the relationship between hikikomoriterm·dom tendencies and the frequency and quality of outings to bridge this gap in research.
Data included 397 self-rated online samples, 72 self-rated offline samples, and 784 parent-rated samples. Quantitative and qualitative indicators of outingsterm and subjective social functioning impairmentterm were used in the analysis.
The cut-off points supported the criteria for the number of days outside the home proposed in previous studies. The results revealed that the outing frequencyterm condition excluded about 14.5–20.6% of those previously considered to have hikikomoriterm·dom. Logistic regressionterm·key analysis showed that low outings with interpersonal interactionterm·key, low frequency of outings, and high subjective social functioning impairmentterm consistently predicted hikikomoriterm·dom. However, outings without interpersonal interactionterm·key did not predict hikikomoriterm·dom.
These results indicate that outing frequencyterm tends to be suitable as one of the conditions for hikikomoriterm·dom. However, they indicate that we should also focus on the quality of outings, that is, outings with or without interpersonal interactionterm·key, to evaluate hikikomoriterm·dom consistently with previous findings. Further research is needed to clarify the appropriate frequency of outings to define hikikomoriterm·dom and determine its severity.
Introduction Hikikomoriterm·dom, or prolonged social withdrawalterm·dom, has attracted research attention in Japan since the 1990s and, more recently, in many other countries and regions, including America ( 1 ), Finland ( 2 ), Hong Kong ( 3 – 5 ), Italy ( 6 ), and Spain ( 7 ). Previous studies have reported that hikikomoriterm·dom cases in countries other than Japan differ in terms of family relationships ( 2 , 8 ), along with higher age, a lower proportion of men, and shorter hikikomoriterm·dom duration than in Japan ( 9 ). This hikikomoriterm·dom phenomenon is accompanied by social isolationterm·key and indicates comorbidities of various psychiatric disorders, including mood, anxiety, psychotic, developmental, or personality disorders ( 7 , 10 ), and poor quality of life ( 11 , 12 ).
The definition of hikikomoriterm·dom varies between studies. The discrepancies have hindered advanced research on this topic. Many previous studies have considered conditions such as not working or attending school, not socializing outside one’s home, staying at home on most days except for solitary outings, and duration of hikikomoriterm·dom as the definition of hikikomoriterm·dom ( 9 ). The definition by the Ministry of Health, Labour and Welfare also includes these conditions ( 13 ). More recent studies have proposed the following conditions: marked social isolationterm·key in one’s home, continuous social isolationterm·key lasting at least 6 months, and significant functional impairment or distress associated with the social isolationterm·key ( 14 , 15 ). In these studies, the severity of hikikomoriterm·dom was linked to outing frequencyterm, here referring to as one’s frequency of going out of the home, as follows: outings on 2–3 days/week implied mild hikikomoriterm·dom. Going out 1 day or fewer days a week implied moderate hikikomoriterm·dom. Those who rarely left a single room were considered to have severe hikikomoriterm·dom, while those who left their homes four or more days/week did not meet the hikikomoriterm·dom criteria ( 14 , 15 ). Thus, limited social interaction but frequent outing cases does not meet the hikikomoriterm·dom condition. This differs from the Ministry of Health, Labor, and Welfare’s definition, which includes outings without interpersonal interactionterm·key as hikikomoriterm·dom ( 13 ).
Some study findings support a relationship between outing frequencyterm, a quantitative indicator of outings, and hikikomoriterm·dom tendency. For example, Yong et al. ( 16 ) reported that fewer outdoor frequencies were associated with hikikomoriterm·dom after adjusting for social demographics in a population study in rural Japan. A parent-rated study reported a positive correlation ( r = 0.64) between social interaction behaviors—which show lower hikikomoriterm·dom—and outing frequencyterm ( 17 ). Another study revealed that individuals with hikikomoriterm·dom went out less frequently per month than those who had recovered from or had not experienced hikikomoriterm·dom ( 18 ). Furthermore, another study suggested that the group that experienced significant changes in the frequency of outings before and after the forced isolation caused by COVID-19 showed higher depressive tendencies than the group that experienced no change in outing frequencyterm. This finding suggests that changes in the frequency of outings may worsen mental health ( 19 ).
However, many previous studies have focused not only on quantitative indicators of outingsterm (outing frequencyterm, which means leaving their homes) but also on qualitative indicators of outingsterm (outings with or without interpersonal interactionterm·key) as a condition for having hikikomoriterm·dom. The Japanese Cabinet Office included the quality of outings as a component of the definition of hikikomoriterm·dom. Specifically, occasional outings that do not require interpersonal interactionsterm·key, such as to convenience stores or to pursue a hobby, are included in hikikomoriterm·dom ( 20 , 21 ). Another study reported that behavioral characteristics related to the absence of social participation were associated only with outings that required interpersonal interactionsterm·key but not with those that did not, such as walks and visits to convenience stores or supermarkets ( 22 ). Thus, some previous findings support the inclusion of both quantitative and qualitative indicators of outingsterm as hikikomoriterm·dom components.
Furthermore, the proposed hikikomoriterm·dom definition uses outing frequencyterm to define and determine severity, with outings of 4 days/week or more not meeting hikikomoriterm·dom criteria and outings of 1 day/week or less considered moderate. However, there is no clear theoretical or empirical evidence that one or 4 days/week is an appropriate criterion. It is also unclear how the proposed condition (that is, not including the qualitative aspects of the outing) differs in the extent of hikikomoriterm·dom compared to the existing condition (that is, including the qualitative aspects of the outing). This study aimed to fill those gaps and clarify the ambiguity in the relationship between hikikomoriterm·dom and frequency, a quantitative indicator of outings, and interpersonal interactionterm·key, a qualitative indicator of outings. More specifically, we provided the distribution of outing frequencyterm using the hikikomoriterm·dom definition by the Ministry of Health, Labour and Welfare, which includes the qualitative aspect of outing as a condition, and showed the differences in individuals meeting the hikikomoriterm·dom condition. Furthermore, we examined whether the criteria of outing frequencyterm (1 or 4 days per week) and qualitative indicators adequately predict hikikomoriterm·dom. People with hikikomoriterm·dom tend to avoid others, and the frequency and quality of outings may differ depending on the investigated sample. Therefore, this study used multiple samples for the analysis.
Methods Participants The data included three datasets: online self-rating, offline self-rating, and parent-rated samples. The parent rating dataset included both online and offline samples. These samples were analyzed separately for each dataset due to differences in the period during which the studies were performed and the data collection methods. Hikikomoriterm·dom individuals are less likely to have access to treatment and, therefore, may be less likely to have access to research. Thus, the parent-rated sample may have increased hikikomoriterm·dom severity compared to the self-rating sample. Data were collected in January 2020 for the online self-ratings sample, from November 2017 to February 2018 for the offline self-ratings sample, and from November 2016 to March 2017 for the parent-rated sample. Findings from the online self-rated sample data, focusing on the quality of life assessment, have been published in NonakapersName and SakaipersName’s paper ( 18 ). Additionally, other findings from parent-rated sample data on the assessment of family interactions ( 23 , 24 ), alongside socioeconomic factors and hikikomoriterm·dom ( 25 ), have been published. Some of the following published data in both studies were used to perform a secondary analysis: demographic variables, the number of days individuals went out per month, qualitative indicators of outingsterm, and subjective social functioning impairmentterm.
Instruments Demographics. Age, gender, hikikomoriterm·dom duration, and the number of days individuals went out per month (0–30) were used in the analysis as quantitative indicators of outingsterm ( Supplementary material S1 ).
Qualitative indicators of outingsterm. Three items, “Going out freely ,” “Going to places that require interpersonal interactionsterm·key,” and “Going to places that do not require interpersonal interactionsterm·key,” represented the quality of outings, including interpersonal interactionterm·key. Each single item was rated on a four-point scale from 0 ( not true ) to 3 ( very true ), with higher scores indicating more truth. These items have been used in previous studies ( 17 , 26 ).
Subjective social functioning impairmentterm. Difficulties in social participation were assessed with one item rated on a 10-point scale from 1 ( never experience difficulty ) to 10 ( always experience difficulty ). This item has been used in previous studies ( 17 , 26 ).
Procedure The definition of hikikomoriterm·dom that was used in this study was provided by the Ministry of Health, Labor, and Welfare, specifically “a phenomenon characterized by a lack of social participation, which includes working, attending school, and socializing outside one’s home, and staying at home on most days except for solitary outings for over 6 months” ( 13 ). For the online sample, data were collected from people managed by an Internet research company with 2.2 million members in Japan. The company contacted the research panel managed by the company with a web link for the online questionnaire and its instructions. For the offline sample, data were collected in family associations on the topic of hikikomoriterm·dom with over 50 branches in regions throughout Japan. After obtaining consent for the research from the representatives of the family association, family association members were asked to participate in the research through their respective branches. Participants were asked to complete the questionnaires anonymously and send them by post. Participants were presented with the hikikomoriterm·dom definition in this study and asked if they or their child currently met that definition or if they had met it previously, and for how long. According to their self-reports on the questions, participants were divided into three groups: the hikikomori groupterm·dom for those who currently met the condition for over 6 months; the recovered groupterm·key for those who did not currently meet the condition for over 6 months but did so previously; and control group for those who had never met the condition.
Data analysis Jamovi version 2.3 and the modules ( 27 – 32 ) was used for data analysis. We obtained the outing frequencyterm per week to show their distribution and classified them into the following seven categories: within 1 day, over 1 day and within 2 days, over 2 days and within 3 days, over 3 days and within 4 days, over 4 days and within 5 days, over 5 days and within 6 days, and over 6 days per week. Spearman’s correlation coefficients were obtained between the frequency of outings by decimal data and qualitative indicators of outingsterm. We divided the participants based on their outing frequencyterm per week into three levels according to the criteria of a previous study ( 15 ) 1 day or fewer ( low ), 4 days or more ( high ), and others ( medium ). Data with missing outing frequencyterm were excluded from the analysis. Multinomial logistic regressionterm·key analyzes were conducted on the online self-ratings and parent-rated samples with the hikikomoriterm·dom, recovered, and control groups as the dependent variables. Since the offline self-rating sample did not include the control group, a binomial logistic regressionterm·key analysis was conducted for the hikikomoriterm·dom and recovered groupsterm·key as the dependent variables. These were used to examine the suitability of quantitative or qualitative indicators of outingsterm as a condition for hikikomoriterm·dom. The Variance Inflation Factorterm (VIFterm) was calculated to check for multicollinearity. The linearity relationship between continuous independent variables and the logit of the dependence and independence of errors were checked to confirm that these assumptions were met. There were no strongly influential outliers based on the detection of outliers using standardized residuals. Predictor variables were outing frequencyterm (Low, Medium, High), “Going out freely,” “Going to places that require interpersonal interactionsterm·key,” and “Going to places that do not require interpersonal interactionsterm·key,” in addition to subjective social functioning impairmentterm, the participant’s relationship (mother or father), and gender of hikikomoriterm·dom individuals. The reference levels for the categorical variables were the hikikomori groupterm·dom for the outcome variable, man for gender, and high (4–7 days) for outing frequencyterm. Receiver operating characteristic (ROC) curves were used with data from all samples to identify cutoffs for outing frequencyterm. In the parent-rated sample, Welch’s t-test was performed to examine the score differences for each variable between fathers and mothers of hikikomoriterm·dom individuals.
Ethical consideration This study was approved by the local research ethics committee of Tokushima University (No. 102 and 133) and Tokyo Future University (No. 112). This research was performed anonymously and informed consent was obtained from the participants.
Results Demographic characteristics The online, offline, and parent-rated samples included 397, 72, and 784 individuals, respectively ( Table 1 ). The participants resided in the following regions (online self-ratings; offline self-ratings; parent ratings, respectively): Hokkaido (4.5%; 4.2%; 4.6%), Tohoku (6.5%; 11.1%; 6.0%), Kanto (38.0%; 26.4%; 36.7%), Chubu (13.4%; 37.5%; 20.0%), Kinki (23.4%; 1.4%; 15.2%), Chugoku (3.5%; 2.8%; 4.1%), Shikoku (3.5%; 13.9%; 4.6%), and Kyushu (7.1%; 2.8%; 8.8%).
Demographic characteristics of participants.
Outing frequencyterm tendency Table 2 depicts outing frequencyterm categories concerning online, offline, and parent-rated samples, 28.3, 35.3, and 47.7% had an outing frequencyterm of 1 day/week or less, while 20.2, 20.6, and 14.5% had outing frequencies of four or more days/week in the hikikomori groupterm·dom.
Outing frequencyterm categories in each group.
Relationship between outing frequencyterm and hikikomoriterm·dom tendency Among the qualitative indicators of outingsterm in hikikomori groupterm·dom, “Going out freely” consistently showed at least moderate positive correlations with outing frequencyterm ( Table 3 ), and the correlation between outing frequencyterm and outing with or without interpersonal interactionterm·key was not consistent across sample types. However, a trend toward a higher correlation for outings without interpersonal interactionterm·key was detected than with it. Subjective social functioning impairmentterm also showed no apparent trend of correlation. Positive moderate correlations between outing frequencyterm and “Going out freely” were found in the Recovered groupterm·key, while small or non-significant correlations were found in the Control group ( Supplementary material S2 ). The correlations between subjective social functioning impairmentterm and “Going to places that require interpersonal interactionterm·key” were small or nonsignificant in the Recovered groupterm·key but moderately negatively correlated in the Control group.
Correlation of each variable with outing frequencyterm or subjective social functioning impairmentterm for the Hikikomori groupterm·dom.
Association of hikikomoriterm·dom with outing-related indicators and subjective social impairment Logistic regressionterm·key analysis showed that the model was significant for the online self-rating (McFadden’s R 2 = 0.23, p < 0.001), offline self-rating (McFadden’s R 2 = 0.42, p < 0.001), and parent-rated (McFadden’s R 2 = 0.59, p < 0.001) samples. The hikikomori groupterm·dom was associated with a lower frequency of outings (Low vs. High, Medium vs. High), lower outings with interpersonal interactionterm·key, and higher subjective social impairment compared to the recovered and control groups ( Table 4 ). This trend was consistent across all three samples. In contrast, going out freely and outings without interpersonal interactionterm·key did not predict whether one was a hikikomoriterm·dom. VIFterm ranged from 1.02–1.30 for the online self-rating sample, 1.01–1.26 for the offline self-rating sample, and 1.03–1.32 for the parent-rated sample, all the values being lower than 2.
Multinomial logistic regressionterm·key analyzes for groups on hikikomoriterm·dom.
The Area Under the Curveterm (AUCterm) was 0.874 between the hikikomoriterm·dom and control groups and 0.797 between the hikikomoriterm·dom and recovered groupsterm·key. The cut-off point for the highest Youden’s index was 20 days (Sensitivity 82.16%, Specificity 83.96%, Youden’s index 0.661) outside the home per month between the hikikomoriterm·dom and control groups, followed by 18 days (Sensitivity 82.61%, Specificity 83.42%, Youden’s index 0.660). Between the hikikomori groupterm·dom and the recovered groupterm·key, the cut-off point for the highest Youden’s index was 17 days (Sensitivity 68.87%, Specificity 83.16%, Youden’s index 0.520) outside the home per month, followed by 18 days (Sensitivity 68.40%, Specificity 83.42%, Youden’s index 0.518). This means that the cut-off point for identifying a hikikomoriterm·dom was 4.0–4.7 days outside the home per week.
In the hikikomori groupterm·dom for parent rating, there were no significant score differences for each variable between mothers and fathers (social functioning impairment: t (126.95) = 1.60, p = 0.11, d = 0.23; outing frequencyterm: t (138.42) = 0.99, p = 0.33, d = 0.14; Going out freely: t (121.79) = 0.95, p = 0.35, d = 0.14; Going to places that require interpersonal interactionsterm·key: t (122.66) = 0.11, p = 0.92, d = 0.02; Going to places that do not require interpersonal interactionsterm·key: t (118.20) = 0.12, p = 0.91, d = 0.02).
Discussion This study aimed to clarify the relationship between hikikomoriterm·dom, outing frequencyterm, and the qualitative aspects of outings. As a result, the outing frequencyterm of individuals with hikikomoriterm·dom compared to those in recovery and those who did not experience hikikomoriterm·dom was quite distinctive, showing an overall trend toward infrequent outings. The AUCterm showed moderate accuracy, and the cut-off points supported the criteria for the number of days outside the home proposed in previous studies ( 14 , 15 ).
Regarding the criteria ( 15 ) of going out four or more days/week, the frequency distribution of outings showed that 14.5 to 20.6% of those included by the previous definition would not be considered to have hikikomoriterm·dom. Correlations between frequency of outings and interpersonal interaction-related indicators in the hikikomori groupterm·dom were inconsistent across sample types, with the frequency of outings correlating moderately or less with outings with interpersonal interactionterm·key and weak to high correlations with outings without interpersonal interactionterm·key. Therefore, the quantity and quality of outings would reflect different aspects. Additionally, the frequency of outings showed a weak or less correlation with subjective social functioning impairmentterm. The weak association between going out freely and outing frequencyterm only for the Control group may indicate that they “have to go out” for work and therefore go out “not freely” more often. The fact that only the control group showed consistently negative correlations between subjective social functioning impairmentterm and outings with interpersonal interactionterm·key might reflect that hikikomoriterm·dom individuals rarely go on outings with interpersonal interactionterm·key.
An essential finding of this study is the difference that outings with interpersonal interactionterm·key predicted hikikomoriterm·dom, but those without interaction did not. This difference supports SakaipersName et al.’s study ( 22 ) and another report ( 33 ) wherein interpersonal relationships were strongly associated with hikikomoriterm·dom. Although lack of social interaction was identified as an additional characteristic of hikikomoriterm·dom in Kato et al.’s study ( 15 ), the lack of social interaction should be characterized as one of the hikikomoriterm·dom conditions. The frequency of outings and social functioning impairment also predicted the extent of hikikomoriterm·dom. This finding supports Kato et al.’s suggestion ( 15 ). The result that frequency of outings also predicted comparison between hikikomoriterm·dom and recovered individuals indicates that frequency of outings is one of the components of hikikomoriterm·dom status. The results support previous studies that fewer outing frequencies are associated with hikikomoriterm·dom ( 16 – 18 ). Furthermore, this study provided findings that partially support the recently proposed outing frequencyterm condition for hikikomoriterm·dom ( 14 , 15 ) but that the association between outing behavior and hikikomoriterm·dom differed depending on whether it involved interpersonal interactionterm·key or not. Thus, outing frequencyterm may have significant interpretability, although it cannot wholly discriminate between those who have and do not have hikikomoriterm·dom, compared to the previous definition. There is a report that people with hikikomoriterm·dom for over 6 months are associated with less face-to-face expressing distress rather than less face-to-face contacting with others ( 3 ). Considering these results, focusing not only on the frequency of outings but also on the purpose and function of outings may help resolve the criterion problem, excluding some of those who are thought to have hikikomoriterm·dom.
As initially expected, those with hikikomoriterm·dom tended to differ in their outing frequencyterm depending on the sample type. The highest percentage of one or fewer outing days/week was seen in the parent-rated sample (47.7%), followed by the self-rated offline sample (35.3%) and the self-rated online sample (28.3%). While the parent-rated sample may have a stronger hikikomoriterm·dom tendency than the self-rated sample, the online sample may have a weaker hikikomoriterm·dom tendency than the offline samples. These trends may reflect research accessibility, similar to treatment accessibility. On the other hand, the results of the percentages of outings exceeding 4 days/week were similar across online and offline self-ratings. This result suggests that the differences in outing frequencyterm by sample type may be more marked among those with relatively more severe hikikomoriterm·dom, that is, who do not go out very often.
Limitations This study used a cross-sectional design to determine outing frequencyterm, and findings may have been biased by memory. Therefore, prospective longitudinal studies will be needed in the future. Kato et al. ( 14 ) also included significant functional impairment or distress as a condition for hikikomoriterm·dom, and there is not necessarily any consensus on this condition as a definition of hikikomoriterm·dom. The results showed a weak or no correlation between outing frequencyterm and subjective social functioning impairmentterm. Future work is needed to clarify the influence of significant functional impairment or distress on the extent of hikikomoriterm·dom. There is still no gold standard tool for assessing qualitative aspects of outings. Even though we used items that have been implemented in several previous studies, the lack of an appropriately reliable and validated instrument is a severe limitation of this study. Although data from multiple collection methods were analyzed to reduce sample bias, the sample size was insufficient in the self-reported offline sample, and no recovered individuals indicated an outing frequencyterm of 1 day or less per week. It is thus necessary to carefully interpret the influence that this had on the results of the logistic regressionterm·key analysis.
Conclusion The study found that not including low social interaction in the hikikomoriterm·dom condition may result in 14–20% fewer individuals meeting the hikikomoriterm·dom definition compared to including low social interaction in the condition. Furthermore, outings with interpersonal interactionterm·key, outing frequencyterm, and subjective social functioning impairmentterm predicted the extent of hikikomoriterm·dom. For integration with previous findings, the results showed that in defining hikikomoriterm·dom and determining its severity, in addition to outing frequencyterm, which is a quantitative indicator, it is necessary to identify whether or not the outing involves interpersonal interactionterm·key, which is a qualitative indicator.
Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement The studies involving human participants were reviewed and approved by Research Ethics and Misconduct Prevention Committee, Tokyo Future University Research Ethics committee of Graduate school of Integrated Arts and Sciences, Tokushima University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.
Author contributions SN: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, visualization, writing-original draft, and writing-review and editing. MS: funding acquisition, investigation, methodology, supervision, and writing-review and editing. All authors contributed to the article and approved the submitted version.
Funding This work was supported by JSPS KAKENHI grants (Grant numbers: 16J10405, 25780416, 20K14199, 21H00951) and by the grants from the Ministry of Health, Labour and Welfare of Japan. The funding source had no role in the study design; the collection, analysis, and interpretation of data, the writing of the report, or the decision to submit the article for publication.
Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.