Holiday Internet Usage Time and the Risk of Internet Addiction Tendency among Working Adults in their 30s in Japan

Abstract

Introduction: A positive association between Internet usage time and Internet addiction among adolescents and adults has been frequently reported; however, studies of working adults focusing on weekdays and holidays are limited. Therefore, this study aimed to elucidate the association between Internet usage time and psychometric tests among working adults in their 30s, focusing on weekdays and holidays. Methods: A total of 129 workers aged 30–39 years participated in this study. Participants completed a questionnaire and interview regarding psychometric tests, including Internet usage time, Internet addiction tendency, smartphone addiction tendency, depression tendency, and personality traits. A correlation analysis focusing on differences between weekdays and holidays was conducted. Results: The scores on Internet addiction scales are weakly positively correlated with holiday Internet usage time. The scores of smartphone addiction scales are also weakly positively correlated with the holiday Internet time. No correlation was found between weekdays Internet usage time and scores on Internet addiction scales. Conclusions: Internet usage time during holidays is associated with Internet addiction tendency among the working adult samples. Holiday Internet usage time could be a useful indicator of risk of Internet addiction. Our pilot findings provide clues to the mental health affected by the Internet, especially among adults.

1. Introduction

The Internet has become an important communication tool in modern society. The Internet enables people to interact with each other without any restrictions, and Internet-based social networking services (SNS) and online games have spread widely. In 1998, addictive behavior toward the Internet was coined “Internet addiction”, based on the concept of behavioral addiction, such as pathological gambling disease [1]. Internet addiction has been reported in several countries worldwide and requires the establishment of diagnostic criteria and treatment methods [2,3]. Internet gaming disorder is classified as a “Condition for Further Study” under the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) [4].

A survey of adolescents in Japan revealed a positive correlation between problematic Internet use and the amount of time spent on the Internet, especially during holidays rather than weekdays [5]. On the other hand, a survey in the United Kingdom demonstrated that an increase in screen time on computers and smartphones disturbed the mental well-being of adolescents [6]. Differences in working status have been noted in the association between Internet usage time and compulsive Internet use among adults [7]. A recent survey of working adults in Japan has indicated that groups at risk for Internet addiction tended to spend more time using the Internet [8].

Internet addiction, smartphone addiction, and depression are closely related [9]. A high prevalence of smartphone addiction has been reported in various countries, particularly among teenagers [10,11]. The use of SNS has also been linked to mental illnesses, such as depression and eating disorders [12]. In terms of gender, females tend to be more dependent on SNS, whereas males tend to be more dependent on games [13]. Research on Facebook’s problematic use has indicated that women’s borderline personality traits potentially predict the problematic use of Facebook [14]. Furthermore, Internet addiction among university students in Taiwan is reportedly related to personality traits, especially borderline personality [15]. A relationship between time of using information devices and depression has also been noted. Multiple studies have reported that sedentary time using the Internet on a computer is positively correlated with depressive symptoms and anxiety [16].

The above studies strongly suggest a positive association between Internet usage and mental health among adolescents, especially students [13]. Previous analyses of Internet addiction among adults have primarily focused on adults in their 20s. Moreover, most analyses of smartphone addiction have focused on adults in their 20s [17,18]; however, studies focusing on working adults over 30s are very limited.

We hypothesized that there is an association between Internet usage time and mental health parameters in working adults in their 30s. We also hypothesized that Internet usage time on holidays, rather than on weekdays, would be more indicative of the psychological and psychosocial effects of Internet use.

Therefore, this study aimed to elucidate the association between Internet usage time and psychometric tests among working adults in their 30s with a particular focus on weekdays and holidays.

2. Materials and Methods

This study was conducted in accordance with the latest version of the Declaration of Helsinki and approved by the Ethics Committee of Kyushu University (29–600). All participants provided informed consent prior to assessment.

2.1. Participants

A total of 156 workers participated in this study according to a convenience sampling method.

Full-time and part-time workers aged 30–39 years were recruited from companies near Fukuoka City using posters to complete in-person assessments. The inclusion criterion is “relatively healthy working adults”, and the exclusion criterion is “those who are not working”.

All participants were requested to complete a questionnaire regarding psychometric tests, including Internet usage, and interviewed by a psychiatrist or psychologist. Twenty-seven participants who did not respond to the questionnaire on Internet usage were excluded. Hence, data from 129 participants were finally analyzed.

2.2. Measures

2.2.1. Internet Usage Time

Participants completed a self-administered questionnaire regarding the hours per day they spent on the Internet and purpose of using the Internet. The questionnaire was divided into weekdays and holidays. Internet usage time during holidays was used in the analysis because Internet usage time on weekdays varied greatly due to occupation. Internet usage time tended to be extremely long when using the Internet for work, and we determined that Internet usage time during weekdays was not an appropriate method for evaluating free-will Internet usage.

2.2.2. Depression

Three psychometric scales were used to assess depression. The nine-item Patient Health Questionnaire (PHQ-9) and Beck Depression Inventory II (BDI-II) were used to assess the severity of depressive tendencies [19,20,21]. The Hamilton Rating Scale for Depression (HRDS) was also used [22,23].

2.2.3. Hikikomori and Modern-Type Depression

The Hikikomori Questionnaire (HQ-25) includes 25 self-rated items containing three subscales representing socialization, isolation, and emotional support [24]. All HQ-25 items were rated from 0 (strongly disagree) to 4 (strongly agree) and summed to obtain a total score ranging from 0 to 100. The HQ-25 cutoff score was set at 42 (out of 100), which was reported to have a sensitivity of 94% and specificity of 61% in a previous study.

The 22-item Tarumi Modern-Type Depression Trait Scale (TACS-22) is a 22-item self-rated questionnaire that is classified into three subscales: avoidance of social roles, complaint, and low self-esteem [25]. Responses were rated on a 5-point scale from 0 (disagree) to 4 (agree). The total scores ranged from 0 to 88. A previous study reported a sensitivity of 63.1% and specificity of 82.9%.

2.2.4. Internet and Smartphone Addiction

Young’s Internet Addiction Test (IAT) comprises 20 items on Internet overuse to measure the severity of Internet addiction. Responses were rated on a five-point scale: 5 (always), 4 (often), 3 (frequently), 2 (occasionally), and 1 (rarely). The total IAT score ranged from 20 to 100 [26,27]. A previous study investigated the reliability and validity of the Japanese version of the IAT [28].

The standard Smartphone Addiction Scale (SAS) includes 33 questions, and it is rated from 1 (strongly disagree) to 6 (strongly agree). The short version of the SAS (SAS-SV) was reduced to 10 questions from the standard SAS for purposes of screening [29]. In a previous study, the reliability and validity of the Japanese version of the SAS-SV were investigated [30].

2.2.5. Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, (DSM-IV) Personality Disorders Personality Questionnaire (SCID-II-PQ)

The SCID-II-PQ is a 119-item self-administered questionnaire on the Structured Clinical Interview for the DSM–IV (SCID-II). The response options for all items are “yes” (1 point) and “no” (0 points). The SCID-II-PQ is divided into 12 subscales: avoidant, dependent, obsessive-compulsive, passive-aggressive, depressive, paranoid, schizotypal, schizoid, histrionic, narcissistic, borderline, and antisocial personalities. Each item was summed to obtain a total score, defined as the index of each personality trait [31].

2.3. Statistical Analysis

Statistical analysis was conducted using SPSS (IBM^® SPSS^® Statistics ver. 26).

The initial analysis assessed the correlation between each psychometric tests score and Internet usage time during holidays. Weekdays usage hours were influenced by the nature of the occupation (e.g., systems engineer). First, we assessed distribution normality using the Shapiro–Wilk test and performed a two-sample t-test or the Mann–Whitney U test to reveal gender differences in the demographic data. Second, we assessed the correlation between Internet usage hours during holidays/weekdays and the score of each mental rating scale. Furthermore, the data were classified by gender, and the purpose of using the Internet was predominantly for SNS (SNS +/−). Correlation estimates (r) were calculated by Pearson’s product moment correlation or Spearman’s rank-order correlation. As a second subgroup analysis, we further stratified the correlation analysis with Internet usage time during holidays by the Internet usage type (SNS +/−) and gender. Internet usage data by hour were stratified by gender, Internet usage type (SNS +/−) and SNS type mainly used, and personality type over cutoff values for the screening questions (Table S4). We set the level of significance at p < 0.05 (two-tailed).

3. Results

The results of the analysis using the data of 129 participants were as follows.

Sample population shows no gender differences in Internet usage (weekdays and holidays). Women in the sample had a higher depressive tendency than the men in the sample. Men in the sample had higher tendency for borderline, antisocial, and narcissistic traits than the women in our sample (

Table 1. Demographic characteristics: gender comparison of the average measures.

		Total n = 129 (SE)	Females n = 71 (SE)	Males n = 58 (SE)	p-Value
	Age, years	34.81 (0.25)	34.82 (0.32)	34.79 (0.41)	0.975
Internet Usage Time	On weekdays, hours	3.001 (0.20)	2.981 (0.30)	3.026 (0.27)	0.336
	On holidays, hours	2.858 (0.18)	2.981 (0.21)	2.707 (0.32)	0.112
Depressive Tendency	BDI-II	6.62 (0.65)	7.5 (0.94)	5.51 (0.87)	0.214
	PHQ-9	3.29 (0.28)	3.73 (0.38)	2.76 (0.42)	0.022
	HRSD	2.18 (0.26)	2.55 (0.37)	1.74 (0.37)	0.048
Modern-Type Depression and Hikikomori	TACS-22 #	33.91 (0.87)	33.59 (1.15)	34.32 (1.35)	0.681
	HQ-25	26.19 (1.49)	26.42 (1.94)	25.9 (2.33)	0.718
Internet and Smartphone Addiction	IAT #	27.53 (1.43)	25.32 (1.65)	30.33 (2.46)	0.084
	SAS-SV	24.26 (0.75)	24.65 (1.01)	23.77 (1.14)	0.717
SCID-II-PQ	Avoidant	2.03 (0.18)	2.17 (0.23)	1.85 (0.27)	0.3
	Dependent	1.36 (0.14)	1.54 (0.22)	1.13 (0.18)	0.298
	Obsessive-compulsive	2.97 (0.16)	3.05 (0.21)	2.88 (0.25)	0.713
	Passive-aggressive	1.46 (0.15)	1.31 (0.20)	1.65 (0.23)	0.194
	Depressive	1.96 (0.19)	2.11 (0.27)	1.77 (0.28)	0.495
	Paranoid	1.21 (0.14)	1.26 (0.21)	1.15 (0.18)	0.656
	Schizotypal	1.09 (0.12)	1.28 (0.18)	0.85 (0.16)	0.074
	Schizoid	1.26 (0.12)	1.31 (0.16)	1.21 (0.18)	0.615
	Histrionic	1.57 (0.14)	1.35 (0.17)	1.85 (0.23)	0.099
	Narcissistic	1.9 (0.17)	1.55 (0.21)	2.33 (0.27)	0.027
	Borderline	1.72 (0.21)	2.14 (0.30)	1.19 (0.25)	0.021
	Antisocial	1.1 (0.15)	0.71 (0.13)	1.61 (0.30)	0.023

SE: standard error; #: these items were normally distributed; BDI-II: the Beck Depression Inventory II; PHQ-9: Patient Health Questionaire-9; HRSD: the Hamilton Rating Scale for Depression; TACS-22: the 22-item Tarumi’s Modern-Type Depression Trait Scale; HQ-25: the Hikikomori Questionnaire has self-rated 25 items; IAT: Young’s Internet Addiction Test; SAS-SV: Smartphone Addiction Scale (Short Version); SCID-II-PQ: the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) Personality Disorders Personality Questionnaire.

). In addition, females have higher rates of part-time employment and lower annual incomes compared to males (Table S1). The significances confirmed in the present samples are not always confirmed in other population [32].

No significant correlations were identified between Internet usage time during holidays and depressive symptoms or many personality tendencies. (

Table 2. Correlation analysis between Internet usage time during holidays and each mental rating scale.

Internet Time Using on Holidays
	Total Participants n = 129		SNS– n = 85		SNS+ n = 44		Females n = 71		Males n = 58
	r	p	r	p	r	p	r	p	r	p
BDI-II	0.119	0.206	0.026	0.824	0.250	0.119	0.084	0.512	0.125	0.384
PHQ-9	0.182	0.039	0.134	0.222	0.251	0.101	0.136	0.260	0.193	0.146
HRSD	0.173	0.053	0.217	0.048	0.062	0.695	0.187	0.125	0.112	0.409
TACS 22 #	0.085	0.344	0.056	0.615	0.149	0.333	0.022	0.859	0.144	0.285
HQ 25	0.095	0.313	0.049	0.680	0.145	0.365	0.069	0.587	0.112	0.433
IAT #	0.309	0.001	0.338	0.003	0.310	0.048	0.295	0.017	0.341	0.014
SAS	0.341	<0.001	0.398	<0.001	0.262	0.085	0.408	<0.001	0.257	0.054
Avoidant	0.110	0.240	0.098	0.400	0.064	0.689	0.112	0.375	0.053	0.708
Dependent	0.180	0.053	0.165	0.154	0.143	0.372	0.208	0.096	0.125	0.377
Obsessive- compulsive	0.078	0.405	0.050	0.666	0.051	0.751	0.186	0.137	−0.060	0.674
Passive- aggressive	0.065	0.484	0.045	0.700	0.072	0.656	0.112	0.373	0.023	0.872
Depressive	0.157	0.091	0.083	0.477	0.253	0.110	0.134	0.289	0.136	0.335
Paranoid	0.012	0.899	−0.006	0.958	−0.017	0.915	−0.043	0.734	0.141	0.317
Schizotypal	0.169	0.069	0.117	0.312	0.276	0.081	0.233	0.062	0.033	0.815
Schizoid	0.132	0.157	0.065	0.578	0.227	0.153	0.131	0.299	0.107	0.451
Histrionic	−0.075	0.424	−0.160	0.168	0.081	0.613	0.052	0.682	−0.188	0.182
Narcissistic	0.046	0.621	−0.013	0.909	0.141	0.384	0.157	0.216	−0.017	0.905
Borderline	0.229	0.013	0.098	0.401	0.430	0.005	0.216	0.085	0.165	0.243
Antisocial	−0.063	0.504	−0.084	0.474	−0.008	0.961	−0.093	0.462	0.024	0.867

SNS+ indicates those who used the Internet primarily for SNS, and SNS– indicates those who used it for other purposes. #: these items were normally distributed; r: Spearman’s or Pearson’s correlation estimates, capitalized if r > 0.2 and p-value < 0.05; BDI-II: the Beck Depression Inventory II; PHQ-9: Patient Health Questionaire-9; HRSD: the Hamilton Rating Scale for Depression; TACS-22: the 22-item Tarumi’s Modern-Type Depression Trait Scale; HQ-25: the Hikikomori Questionnaire has self-rated 25 items; IAT: Young’s Internet Addiction Test; SAS-SV: Smartphone Addiction Scale (Short Version); SCID-II-PQ: the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM–IV) Personality Disorders Personality Questionnaire.

) There is a weak positive correlation between holiday Internet use and higher scores on Internet addiction scales (r = 0.309, p = 0.001). This result is valid regardless of gender or type of SNS when the IAT is used (Females: r = 0.295. p = 0.017; Males: r = 0.341, p = 0.014; SNS-: r = 0.338, p = 0.003; SNS +: r = 0.310, p = 0.048). The SAS is also weakly positively correlated among women or those who do not primarily use SNS (Females: r = 0.408, p ≤ 0.001; Males: r = 0.257, p = 0.054; SNS -: r = 0.398, p ≤ 0.001; SNS +: r = 0.262, p = 0.085). Personality traits are weakly positively correlated with holiday SNS + Internet use (r = 0.43, p = 0.005).

The correlation with weekdays Internet usage time is shown (Table S2). No correlation was found between Weekdays Internet usage time and IAT, even when categorized by SNS use or gender.

4. Discussion

Few associations were reported between Internet use time and depressive symptoms or personality traits among 129 working adult samples in their 30s. Internet usage time during holidays was associated with a tendency toward Internet addiction evaluated by the IAT.

We suggest that holiday Internet usage time could be a weak surrogate marker for higher risk for Internet addiction. Holiday Internet usage time may be a weak risk factor for smartphone dependence among females, but not among males. Borderline personality traits are associated with higher holiday social media Internet usage.

Mihara et al. [5] conducted a detailed analysis of the prevalence and trends of harmful Internet use (PIU) among Japanese teenagers. They reported that the greater the amount of time spent on the Internet on holidays, the more harmful Internet use is as defined by the Diagnostic Questionnaire developed by Young (YDQ). In addition, female students in that study indicated that social networking use was a risk factor for PIU. Some similarities were observed in our present study of working adults in their 30s. Nakayama et al. reported that both weekday and holiday Internet usage time were risk factors for Internet addiction among Japanese students. The present study of the adult samples indicated that Internet usage time during holidays may indicate more of a risk for Internet addiction due to differences in lifestyle depending on the work situation [7]. Tsumura et al. warned of the risk of Internet addiction among adult school personnel. This report on working adults was analyzed separately for holidays and weekdays and showed that individuals at risk for Internet addiction tended to spend significantly more time on the Internet on both weekdays and holidays. Tsumura et al.’s report confirms our current study, which targeted a more diverse range of occupations. Furthermore, our report is unique in that it discusses differences between weekdays and holidays [8].

Tateno et al. reported gender differences wherein female university students tended to depend on SNS via smartphones and were less enthusiastic about online games than their male counterparts [7]. Women with borderline personality traits are potentially more likely to depend on the Internet for SNS purposes. On the other hand, Tateno and Kato reported that smartphone addiction tendencies among female vocational school students in Japan was positively related to modern-type depression traits [33]; thus, future studies should seek to clarify the association of Internet usage time with modern-type depression and borderline personality traits among different populations.

The pandemic of COVID-19 infection has increased the risk of harmful Internet use in many regions of the world [34,35]. A Japanese study including adults reported a 1.5 times increase in problematic Internet use (PIU). The increase in Internet usage time after the pandemic was shown to be a possible risk factor for PIU [36]. Online work is gradually increasing due to the diversity of work patterns, and Internet-related addictive behaviors will become more important as a working place mental health. The results of our study may indicate that monitoring internet usage on holidays is important for the maintenance of mental health in the workplace.

Limitations

The present pilot study had several limitations. First, the sample size was small. Second, Internet usage time exhibits inferior accuracy when assessed via self-administered questionnaires compared with other methods, such as those assessing screen time. Third, our study was conducted among adult workers, and we did not clinically diagnose Internet addiction. Fourth, the assessment of personality tendencies (SCID-II-PQ) is only a screening self-administered questionnaire, and false positives are common. Fifth, there are many variables to be compared, which raises a multiple testing problem. Since this is a pilot study, a post hoc analysis of significance was not conducted.

5. Conclusions

This study demonstrated that Internet usage time during holidays was associated with a high score on IAT among the 129 working adult samples. There were no significant gender differences or personality traits in Internet dependency. Holiday Internet usage time could be a useful minor indicator of risk of Internet addiction as measured by IAT. This result may lead to an enhanced understanding of how mental health is affected by Internet and smartphone addiction in the future. The next task is to expand the target population of the survey by including patients with Internet addiction, borderline personality disorders, and depression.