The relationship between psychological distress and dependence on addictive substances and A survey of public attitudes toward addiction

Authors: Takako Nagata, Jun Konishi

Abstract

Aim

To examine the self‐medication hypothesis, which posits that individuals experiencing greater psychological distress tend to engage in addictive substances and behaviors (ASBs).

Methods

We analyzed data from a public survey on addiction awareness. Psychological distress was measured using the K6 scale, and its association with the frequency of use of the nine ASBs was assessed. We also observed a tendency for the usage frequency to vary by age group across ASBs. Furthermore, a general linear model (GLM) was employed with usage frequency as the dependent variable, K6 scores and type of ASBs as independent variables, and age group as a covariate. Post‐hoc analyses were conducted for each ASB to explore specific associations.

Results

Higher K6 scores were associated with increased usage frequency across all addictive targets, including shopping and Internet use, which are not formally recognized as medical disorders. Internet use, shopping, and prescription/over‐the‐counter drug use appeared to show stronger associations with distress than alcohol use, suggesting that they may be more sensitive indicators of psychological burden. Age‐related patterns varied; younger individuals tended to engage more in gaming and internet/Social Network Service (SNS) use, while older individuals showed higher frequencies of gambling‐related behaviors.

Conclusion

These findings highlight the need for addiction policies that address underlying psychological distress and incorporate age‐appropriate education and access regulations.

INTRODUCTION

There is growing global concern regarding addiction. In 2010, member states of the World Health Organization (WHO) adopted a Global Strategy to Reduce the Harmful Use of Alcohol, signaling a coordinated international effort to address substance‐related harms. ^1^ In alignment with this global initiative, Japan enacted the Basic Act on Measures Against Alcohol‐Related Health Problems in 2013 and the Basic Act on Measures Against Gambling Addiction in 2018. Concurrently, the Japanese government significantly increased its budget for addiction countermeasures, from 110 million yen in 2016 to 810 million yen in 2018, while advancing the Integrated Resort initiative, which includes the development of casino facilities. National efforts have encompassed the establishment of specialized medical institutions, the training of addiction counselors, and the implementation of public education and awareness campaigns. ^2^

Despite the allocation of substantial financial resources, the effectiveness of these interventions remains unclear. A significant treatment gap has been identified in alcohol addiction. According to a national survey conducted in 2017, approximately 1.07 million individuals in Japan, representing 0.87% of the population, were estimated to have alcohol dependence. ^3^ However, fewer than 60,000 individuals have receiving treatment. ^4^ This disparity may be attributed to limited public awareness of alcohol dependence as a medical condition as well as the pervasive stigma associated with seeking treatment for substance use disorders.

The worldwide prevalence of suspected gambling addiction ranges from 0.2% to 5.3%. ^5^ However, in Japan, the rate is notably higher among men, with 6.7% of men and 0.6% of women affected, resulting in an overall prevalence of 3.6%. ^6^ Approximately 80% of these cases are believed to be associated with addictions to pachinko and slot machines. ^6^ Pachinko is a unique form of entertainment developed in Japan. Although not illegal, it includes gambling‐like features as players can trade the balls or coins, they win for prizes that are often exchanged for cash. ^7^ As of 2024, there were still 6706 pachinko parlors in Japan. ^8^ The government operates four types of public gambling—horses, bicycles, boats, and motorcycle racing (known domestically as “auto race”)— to generate public revenue. Despite being legal, addiction issues persist, and the rise of online platforms has exacerbated this problem. Although online casinos are illegal in Japan, many individuals do not recognize their illegality. In response, the Japanese government amended the Basic Act on Measures against Gambling Addiction in June 2025 to prohibit the establishment of online casino websites and dissemination of related advertisements via the Internet.

In addition to alcohol, illegal drugs, and gambling, there has been a growing concern over behavioral addictions and substance abuse among young people, including gaming disorders and abuse or dependence on prescription and over‐the‐counter (OTC) medications. Regarding gaming disorder, a 2019 survey conducted among the general population aged 10–29 years reported that 7.6% of men, 2.5% of women, and 5.1% overall were suspected of having a gaming disorder. ^9^ As a countermeasure against OTC drug abuse and dependence among minors, regulations have been implemented to prevent the bulk purchase of cough medicines containing ingredients such as codeine and ephedrine. These include verifying the purchaser's name and age, and limiting the quantity sold. Nevertheless, according to a nationwide survey conducted in 2021, 1.5% of high school students (1.2% men and 1.7% women) reported using OTC medications in excess of the recommended dosage or frequency for the purpose of altering their mood rather than for treatment, equivalent to approximately one in every 60 students. ^10^ Abuse and dependence on OTC medications among youth remains a latent and ongoing issue.

In the current context, effective addiction prevention requires more than merely restricting access to addictive substances or behaviors; it is essential to understand the psychological and behavioral mechanisms that lead to addiction and develop appropriate interventions. The self‐medication hypothesis is a psychological mechanism that has attracted attention for a long time. ^11^ , ^12^ , ^13^ Khantzian posits that psychological pain is the core issue underlying addiction and addictive behaviors and that individuals with vulnerabilities come to rely on addictive substances or behaviors because they discover that they provide temporary relief or comfort that cannot be obtained elsewhere. He extended this argument to encompass both substance dependence and behavioral addictions. ^13^ To reduce addiction, it is recommended to address the circumstances that compel individuals to resort to addictive substances or behaviors as a means of escaping psychological distress. However, despite the appeal of this hypothesis, quantitative supporting data are lacking.

Many studies have reported associations between mood and anxiety disorders and substance use. ^14^ , ^15^ , ^16^ , ^17^ For example, alcohol use disorder and major depressive disorder are frequently co‐occurring conditions, with each disorder reportedly increasing the adjusted odds of the other by approximately 2.0 to 2.09 times. ^14^ Similarly, anxiety disorders and substance use have been reported to have a mutual influence. ^17^ While these findings are of considerable importance, they predominantly address the relationship between clinically diagnosed psychiatric disorders, such as depression and anxiety, and addictive behaviors. However, in the context of public health policy development, particularly regarding prevention and early intervention strategies for addiction within the general population, it is critical to examine the link between subclinical psychological distress and engagement in addictive behaviors. Furthermore, if individuals use addictive substances or behaviors as a coping mechanism for psychological distress, it is necessary to examine not only alcohol, illicit drugs, and gambling, but also more accessible and addiction‐prone behaviors such as Internet use gaming, and shopping, which are not formally recognized as “addiction” in psychiatric diagnosis criteria such as the DSM‐5 or ICD‐10. To clarify terminology, in this paper, we use “addiction” as an inclusive descriptive term referring to substance use or behaviors that persist compulsively despite harmful consequences without the intention of conferring a diagnosis.

AIM

This study aimed to clarify the relationship between the degree of psychological distress and the frequency of engagement in addictive substances and behaviors (ASBs) and to examine the extent to which this relationship varies across different types of ASBs among the general public.

METHODS

Resource

The survey on public attitudes toward addiction was conducted as part of an initiative led by Yokohama City. This survey targeted people aged 16 years and older who were selected through random sampling from the Basic Resident Register. Yokohama City, with an estimated population of approximately 3.8 million, is the most populous city in Japan. The survey was conducted using self‐administered questionnaires. The questionnaire was distributed by mail, and responses were collected either by mail or online. Data collection was conducted between September 7 and October 6, 2024. The questionnaire included items on age, gender, living situation, mental health status (including the K6 score and tendency to rely on addictive behaviors under stress), knowledge and attitudes toward addiction, and addiction‐related issues concerning the respondent or their family members. The question regarding dependency tendencies under stress was phrased as “To relieve dissatisfaction, worries, hardships, or stress in your daily life, have you ever relied on any of the alcohol, gaming, internet/Social Networking Services (SNS), shopping, pachinko/slot machines, public gambling (refers to horse racing, bicycle racing, motorboat racing, and motorcycle racing), over‐the‐counter or prescription drugs, illegal drugs, or sexual behavior?” For each of the nine ASBs, respondents were asked to indicate their frequency of use on a 4‐point Likert never, rarely, occasionally, and always.

We conducted analyses using the following items from the aforementioned age, gender, cohabitation status, occupation, income, K6 scale score, and frequency of engagement in ASBs under stressful conditions.

Analysis

Psychological distress was assessed by using the K6 scale. The K6 is a self‐report instrument designed to assess psychological distress experienced over the past 30 days. ^18^ The validity of the Japanese version of the K6 has been confirmed. ^19^ The scale consists of six items asking respondents how frequently they have experienced the following nervousness, hopelessness, restlessness or fidgetiness, depression, feelings that everything is an effort, and feelings of worthlessness. Each item is rated on a 5‐point Likert scale ranging from 0 (none of the time) to 4 (all of the time), with total scores ranging from 0 to 24. Previous studies have frequently considered a K6 score of 5 or higher to indicate psychological distress, and a score of 13 or higher to reflect serious psychological distress. ^19^ , ^20^ In accordance with these studies, the present study classified participants into three those scoring 4 or below (“low”), those scoring between 5 and 12 (“moderate”), and those scoring 13 or above (“high”).

Demographic variables, including gender, age, cohabitation status, employment, and income, are summarized. Subsequently, a chi‐squared test was conducted to examine whether there were differences in the frequency of engagement in ASBs across the three K6 groups. Similarly, for each of the nine ASBs—alcohol, gaming, Internet use, shopping, pachinko/slot, public gambling, prescription/over‐the‐counter drugs, illegal drugs, and sexual behavior—the association between the K6 group and ASB frequency was analyzed. The frequency of ASBs across the age groups was calculated and presented graphically for each ASB.

A general linear model (GLM) analysis was conducted with usage frequency as the dependent variable, the K6 score and type of ASBs as independent variables, and age as the covariate. Post hoc analyses were performed for each ASB. The inclusion of the K6 score and ASBs as independent variables was intended to examine how the degree of dependency varies depending on the level of psychological distress and type of ASBs. Age was included as a covariate because some ASBs, such as gaming and social networking services, are expected to be strongly influenced by age, and it was necessary to control for the potential confounding effect of age on psychological distress.

Statistical analyses were performed using R software (http://cran.r-project.org). Statistical significance was set at p < 0.05.

Ethics

A survey on public attitudes toward addiction was publicly conducted by Yokohama City between September 7 and October 6, 2024 as part of a municipal public health initiative. Participation in the survey was voluntary, and no personally identifiable information, such as names or addresses, was collected. After data collection, responses were managed using newly assigned identification numbers that were anonymized and not linked to individual participants.

The present study constitutes a secondary analysis of the de‐identified dataset. No direct identifiers were available to the authors, no re‐contact of participants occurred, and the Ethics Review Committee of the Yokohama City Institute of Public Health reviewed and approved the retrospective analysis protocol in July 2025 (approval No. 2025‐03). The committee determined that additional consent for secondary analysis was not required, given the minimal risk and the anonymized nature of the data.

RESULTS

Out of 5000 individuals invited to participate, responses were obtained from 1830 participants (702 via the Internet and 1,128 via post), resulting in a response rate of 36.6%. The sample included 804 men, 991 women, 17 individuals identified as other, and 18 individuals with no gender response. The age distribution showed a peak in the 50–59 years age group (n = 345) and 70–79 years age group (n = 323), forming a bell‐shaped pattern. Regarding cohabitation status (multiple responses allowed), the most common category was living with a spouse, partner, or significant other (n = 1122; 61.3%), whereas 235 participants (12.8%) reported living alone. In terms of employment, the largest group was comprised of non‐managerial and non‐executive employees (n = 413; 22.6%), followed by unemployed individuals (n = 356; 19.5%). The most frequently reported income range was between 2.5‐ and 5‐million‐yen, accounting for 25.3% of the respondents (see Table 1).

Table 2 presents the number and proportion of respondents categorized into the three groups based on their K6 scores according to the frequency of ASBs during periods of stress. The proportion of each K6 group was calculated for each frequency category to investigate the distribution of the severity of psychological distress within specific frequencies. Subsequently, the characteristics of the three K6 groups were examined for each ASB type.

Among individuals who reported engaging in any ASB, the “Never” category was most prevalent in the low K6 group (84.6%). In the moderate K6 group, the proportions of “Occasionally” (35.8%) and “Always” (38.9%) had the highest proportions, while in the high K6 group, “Always” showed the highest proportion (20.3%) (p < 0.01).

When analyzed by specific ASB, the following patterns were Alcohol: “Never” showed the highest proportion in the low K6 group (69.1%), “Occasionally” in the moderate K6 group (40.5%), and “Always” in the high K6 group (23.2%) (p < 0.01). Gaming, prescription/OTC drugs, and sexual “Never” showed the highest proportion in the low K6 group, “Occasionally” in the moderate K6 group, and “Always” in the high K6 group (p < 0.01). Internet use, shopping, and pachinko/slot “Never” was most prevalent in the low K6 group, while “Always” was most prevalent in both the moderate and high K6 groups (pachinko/slot p < 0.05; internet and p < 0.01). Public “Never” showed the highest proportion in the low K6 group (64.7%), “Always” in the moderate K6 group (53.8%), and “Rarely” in the high K6 group (12.0%) (p < 0.01). Illegal “Never” showed the highest proportion in the low K6 group (63.6%), “Rarely” in the moderate K6 group (33.3%), and both “Occasionally” and “Always” were equally most prevalent (50.0%) in the high K6 group (p < 0.05).

Figures 1 and 2 show the data listed in Table 2. Figure 1 shows the frequency of ASB during stress in the trichotomized K6 groups. Figure 2 shows the frequencies of each of the nine ASBs. In both figures and Table 2, the proportion of respondents selecting “Never” was highest in the low K6 group, while the proportions of “Rarely,” “Occasionally,” and “Always” increased with higher K6 scores.

Table 3 presents the results of the general linear model examining the extent to which K6 scores influence the frequency of engagement in each ASB. For all nine categories of the ASBs, increases in K6 scores were significantly associated with a higher frequency. The standardized regression coefficient (estimate) for alcohol consumption was 0.035, indicating that among adolescents, a one‐point increase in the K6 score was related to a 0.035 increase in alcohol use frequency. The strongest association was observed for Internet use (estimate = 0.063), followed by shopping (0.049), and prescription/OTC drug use (0.042). The weakest association was observed between illegal drug use (0.0029) and pachinko/slot machines (0.0087).

Figure 3 illustrates the proportion of stress‐related ASBs by age group. The proportion of individuals who reported using alcohol at least once under stress (i.e., the combined percentages of “Rarely,” “Occasionally,” and “Always”) was highest among those aged 40–49, and alcohol use was observed across all age groups except for adolescents, for whom alcohol consumption is legally prohibited. The use of gaming and Internet‐related behaviors decreased with age, showing the highest level of engagement among teenagers. By contrast, engagement in pachinko/slot machines and public gambling tended to increase with age. Shopping and sexual behavior were most frequently reported among individuals in their 20 s. The use of OTC drugs and prescription medications under stress was more common among older adults, although similar levels of use were also observed among individuals in their 20 s, comparable to those in their 30s–70 s.

DISCUSSION

This study demonstrated that among the general population, individuals with higher K6 scores, indicating greater psychological distress, tended to engage more frequently in various ASBs. The present findings are consistent with the self‐medication hypothesis, which suggests that individuals resort to ASBs to alleviate their psychological distress. This association was observed not only in cases of severe depression and anxiety but also across a continuum of mental states, including those unlikely to meet the diagnostic criteria for psychiatric disorders. Furthermore, the association was evident not only for traditionally recognized addictions such as alcohol, illegal drugs, and gambling, but also for behaviors such as Internet/SNS use and shopping, which are not widely recognized as medical conditions or “addictions” in Japan.

A significant association was observed between psychological distress and the frequency of engagement in each of the nine ASBs. Even after controlling for age as a covariate, increases in K6 scores significantly influenced usage frequency, although the strength of this association varied depending on the specific addictive behavior. Although alcohol is a widely recognized addictive substance, the association between psychological distress and usage frequency was stronger for Internet (estimate = 0.063), shopping (0.049), and prescription/OTC drug use (0.042) than for alcohol use (0.035) in the present study. One possible explanation is that Internet use and shopping are common everyday activities that are easily accessible to many individuals. Higher psychological distress is often associated with reduced activity levels; however, behaviors such as gaming and Internet use can be performed at home without the need to go out. Unlike alcohol, they do not typically involve physical burden, making them more accessible coping mechanisms. Internet use, shopping, and prescription/OTC drug use may serve as more sensitive indicators of psychological distress than alcohol consumption.

However, the associations between psychological distress and usage frequency were relatively weaker for illegal drugs (estimate = 0.0029), pachinko/slot machines (0.0087), public gambling (0.0106), and sexual behavior (0.0163) than for alcohol (0.035). The social penalties associated with the use and possession of illegal drugs, as well as the financial cost and need for specific social connections to obtain them, are likely deterrents. Although 8.7% and 7.5% of respondents reported having engaged in pachinko/slot machines and public gambling at least once, respectively, these activities often involve financial losses and require traveling to venues such as parlors or racetracks. Therefore, the behavioral threshold for engaging in these activities during periods of psychological distress may be relatively high, resulting in a lower frequency.

However, with the growing accessibility of online platforms for public gambling activities such as horse and motorboat racing, home‐based participation has become increasingly convenient, indicating the potential for an increase in engagement in the future. Controlling access to addictive behaviors can be an effective strategy for preventing addiction. Nevertheless, if psychological distress remains unresolved, individuals may attempt to circumvent such restrictions or shift their dependency to other accessible behaviors. Thus, in addition to access control, it is essential to implement measures that connect individuals to support services, where they can seek help when experiencing distress.

The relationship between age and usage frequency varies according to the type of ASBs used. In this study, younger individuals tended to engage relatively more frequently in gaming and the Internet/SNS, whereas the frequency of engagement in pachinko/slot machines and public gambling increased with age. Public gambling in Japan became available shortly after World War II and continues today, whereas the Internet became widely accessible in the 1990s. Individuals under the age of 30 are considered “digital natives,” having been familiar with the internet and SNS since early childhood. As the proportion of digital natives increases in future generations, the number of individuals who develop a dependency on the Internet and SNS may also increase. With regard to the use of prescriptions and OTC medications, the highest frequency was observed in individuals in their 80 s. This may be attributed to the higher prevalence of illnesses in older adults, which often necessitate medication use. However, it is noteworthy that individuals in their 20s and 30s—who were presumed to have lower rates of physical illness, exhibited frequencies comparable to those in their 70s. In cases of prescription drug dependence, substances such as ephedrine and codeine, which are commonly found in OTC medications for cold and cough, pose a risk. Although measures such as limiting the number of purchases, providing verbal warnings to minors, and restricting sales quantities have been implemented, challenges remain because these products can still be easily purchased online.

As demonstrated by the results, engagement in ASBs tends to increase with greater psychological distress, and the frequency of engagement with each ASB varied across age groups. Therefore, to implement addiction prevention strategies effectively, it is necessary to raise public awareness that psychological distress may underlie addictive behaviors and to establish community‐based support services through which individuals may seek help for both psychological issues and addiction‐related concerns simultaneously. Additionally, public outreach methods should be tailored to the age demographics most likely to engage in each behavior. For example, awareness campaigns targeting younger individuals who are more familiar with SNS can be disseminated through SNS platforms in collaboration with influencers from the same age group. Support programs should be designed to reflect age‐specific stressors—such as those found in schools, workplaces, homes, or elderly communities—and to communicate through age‐appropriate channels.

This study has several limitations. The overall response rate was 36.6%, which may introduce non‐response bias. Individuals with poorer mental health or those heavily engaged in alcohol use, drug use, gambling, or other ASBs may have been less likely to participate or may have felt reluctant to report such behaviors—possibly due to factors such as self‐stigmatization. This could lead to their underrepresentation. Consequently, prevalence estimates and some associations should be interpreted with caution. The mode of response (postal vs. online) may have influenced certain self‐reports, particularly those related to internet use. Although age group was included as a covariate—and thus partly accounts for the strong correlation between age and response mode—residual mode‐related differences (e.g., accessibility, privacy, social desirability) cannot be ruled out. Future studies should explicitly examine mode effects (e.g., by including response mode as a covariate or via stratified analyses). Previous research has identified risk factors for gambling addiction, including being a single young man, being married for less than five years, living alone, having low educational attainment, and experiencing financial hardships ^21^ . Gambling addiction is more prevalent among men in Japan, whereas shopping addiction is generally considered more common among women. These findings indicate that it may be beneficial for future studies to take gender and living conditions into account when examining specific types of ASBs.

In this study, ASBs were assessed using a four‐point Likert scale, which was treated as a continuous variable based on the assumption of equal psychological intervals, and a generalized linear model (GLM) was employed for analysis. Although age, gender, occupation, and income were initially considered as covariates, only age was retained in the final model due to minimal impact on the association with K6 scores. Ordinal regression was also considered as an alternative analytical approach. However, due to convergence issues and unstable parameter estimates arising from skewed category distributions in certain ASBs, it was deemed less suitable for the study's objectives. GLM was ultimately selected for its greater estimation stability and ease of interpretation.

In addition to methodological considerations, this study employed a cross‐sectional design, which allowed for the identification of correlations between psychological distress and the frequency of ASBs, but not causal relationships. Longitudinal research is required to determine whether individuals engage in ASBs to alleviate distress or whether such ASB exacerbate depressive symptoms. Furthermore, to explore the specific nature of psychological distress and its relationship to different types of ASBs, qualitative studies focusing on individual experiences are necessary.

Our findings suggest that a high frequency of ASBs may reflect an underlying psychological distress. Despite those limitations, this study makes a valuable contribution by supporting the self‐medication hypothesis and indicating the potential efficacy of interventions that address psychological distress. It is essential to develop targeted awareness and prevention efforts tailored to the population most likely to access each type of ASBs. Such efforts may contribute to the prevention of psychiatric diseases and addiction.

CONCLUSION

Higher K6 scores were associated with increased usage frequency across addictive targets. Frequency of engagement with each ASB varied across age groups. These findings highlight the need for addiction policies that address underlying psychological distress and incorporate age‐appropriate education and access regulations.

AUTHOR CONTRIBUTIONS

Takako Nagata and Jun Konishi conceived the study, analyzed the data, interpreted the results, and wrote and revised the paper.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS APPROVAL STATEMENT

This study was approved by the Ethics Review Committee of Yokohama City (Approval No. 2025‐3).

PATIENT CONSENT STATEMENT

N/A (Participation in the civic awareness survey was voluntary.).

CLINICAL TRIAL REGISTRATION

N/A.