nts were asked to select all places from which they borrowed money. Responses on the 4-point scales were converted into 1 to 4 points, with “Disagree” and “Never” being 1 point. Responses on the multiple answer question on borrowing money were converted into 0 to 7 points. The score was 0 if the respondents selected “Never borrowed,” but otherwise the number of places from which the respondents borrowed money became the points for that question. The sum of the 4-point scale points and the points for the question on borrowing money were used in the analysis as the Pachinko-pachislot Playing Disorder Scale score (PPDS score). The obtained PPDS score ranged from 26 to 111.
A request for survey participation was sent by mail to 9000 individuals chosen by two-stage stratified random sampling. More specifically, postcards requesting survey participation were first sent to them and a package of survey materials followed. The questionnaire was collected either manually by investigators, via mail, or through a web-based form. Those who submitted responses received a 1000 yen-worth reward either directly or via mail.
3. Result and Discussion
3.1. Data Used in Analysis
Using the cut-off value proposed by Akiyama et al.  , respondents whose PPDS score was 61 or higher were categorized as “pachinko/pachislot players suspected to have a gambling disorder” (Disorder Group). Their responses were compared with those of respondents who played pachinko/pachislot in the past year and had a PPDS score of 60 or lower (Non-Disorder Group). Among all who submitted survey responses (5060), 582 played pachinko/pachislot in the past year. Among these 582 individuals, 24 submitted invalid PPDS responses and thus were excluded from analysis. As a result, data obtained from 558 PPDS respondents were analyzed.
3.2. Disorder Score Distribution
Table 1 shows the average and standard deviation of the score for each PPDS item, sub-scale score, and total score. Among the 558 PPDS respondents, 3.8% (21 individuals) had a “suspected gambling disorder.” These 21 individuals accounted for 0.4% of all 5,060 respondents (95% confidence interval: 0.2% - 0.6%). The DSM-5 (APA, 2013) reports that the past-year prevalence rate of gambling disorder is about 0.2% - 0.3% in the general population, and our study resulted in approximately the same appearance ratio.
3.3. Experience of Playing Other Gambling Games
There were no significant differences in the past-year gambling experience (excluding pachinko/pachislot) between those who had gambling disorder and those who did not. Engagement in non-pachinko/pachislot gambling games did not indicate any possibility of resulting in higher likeliness of developing Pachinko-pachislot Playing Disorder (χ2(1) = 1.04, n.s., V = .04; see Table 2).
3.4. Differences Attributable to Demographic Variables and SES
Table 2 shows the frequency and breakdown of responses to the demographic and SES items by the Disorder and Non-Disorder Groups.
Table 2. Demographic variables and experience of other types of gambling games.
3.4.1. Demographic Variables
No significant gender difference was found between the Disorder and Non-Disorder Groups. In both groups, male pachinko/pachislot players accounted for more than the female counterparts (χ2(1) = 2.61, n.s., V = .07). No significant age differences were found between the two groups either. There were no particular age groups having high percentages of individuals with gambling disorder (χ2(2) = 0.68, n.s., V = .03). There was also no significant difference in distribution of areas of living (χ2(3) = 3.93, n.s., V = .08). Although some studies reported that the prevalence of gambling disorder was higher for men or younger age groups  , there were also study results showing the possibility of gender and age influencing the prevalence at different levels depending on how they interact with other variables. In the case of electronic gambling machines (EGMs), which are similar to pachinko/pachislot machines, no sexual difference was found in the prevalence when the frequency of play was controlled . In a previous gambling disorder study in Japan, Nishimura and Ishida  used data from individuals who called a telephone consulting agency to talk about their pachinko addiction to show that men scored significantly higher on the short form of the SOGS for addicted gamblers. When the same callers were divided into problem gamblers and non-problem gamblers using the cut-off value, however, there was no significant sexual difference. In pachinko/pachislot, gender strongly influences the rate of participation, but its influence on the level of disorder seems marginal.
Individuals with gambling disorder and those without showed no differences in the distributions of the two socio-economic status indicators, “Highest educational attainment” and “Annual household income”. (Highest education attainment: χ2(5) = 3.08, n.s., V = .08; Annual household income U = 2380.50, n.s., r = −.07). On the other hand, a significant difference was found in the distribution of “Amount of savings.” The Disorder Group had a higher percentage of individuals who reported they had no savings (χ2(2) = 23.98, p < .001, V = .21). SES, like gender, is considered to be a risk factor for gambling disorder. However, the influence of SES could change with other variables as reported in a previous study where its relationship with gambling disorder became apparent only after other risk factors such as race and marital history were controlled .
3.5. Divorce Experience
Table 2 shows the frequency distributions of “Divorce experience” for both the Disorder and Non-Disorder Groups. The Disorder Group had a higher percentage of individuals who responded that they had divorce experience (χ2(2) = 8.61, p < .01, V = .12). This agreed with previous studies that reported a higher divorce rate for problem gamblers than for non-problem gamblers  .
The relationship between problem gambling and negative emotions, such as depression, anxiety, and stress, has been pointed out in gambling disorder research. Some studies concluded that gambling was an escape-based coping strategy and was a maladaptive coping style  , and other studies suggested higher psychological stress resulted in a heightened desire to gamble . Both the present and previous studies showed a higher divorce rate for problem gamblers than non-problem gamblers   pointing out that a higher percentage of problem gamblers experienced negative life events than non-problem gamblers. Therefore, psychological stress caused by negative life events may make individuals rely more heavily on pachinko/pachislot as a method of releasing stress, resulting in higher likeliness of developing Pachinko-pachislot Playing Disorder.
The present study indicated that approximately 0.4% of the survey respondents had Pachinko-pachislot Playing Disorder. The study also showed that individuals with a suspected disorder were more likely than those without it to have been divorced, and have no savings. There were not enough levels of difference in gender, annual household income, or the highest education attainment between the two groups to consider them as characteristics of either group. Based on these findings, an important risk factor for Pachinko-pachislot Playing Disorder may be the personality that easily accepts pachinko/pachislot as a form of entertainment that provides enjoyment or allows an escape from uncomfortable situations rather than the social background of the individual. Note, however, the present study was a one-point correlational study. Examination of the influences or causal relationships of these variables will require detailed investigation through longitudinal studies involving surveys conducted twice or more.
All authors designed the study, had advises during the study and reviewed the completed manuscript. YH performed the statistical analysis and wrote the manuscript. HI monitored data collection. KA and AS performed the statistical analysis. All authors read and approved the final manuscript.
Kikunori Shinohara is on the board of trustees of Nichiyukyo (Japan Pachinko Pachi-Slot industry association).
Ethical Approval and Consent to Participate
This study was approved by the Ethical Review Board for Research in the Humanities at Ochanomizu University (Approval No. 2016-98).
This study is wholly funded by the Nikkoso Research Foundation for a Safe Society.