Drug use intentions among young adults in the Republic of Korea: a cross-sectional study applying the extended theory of planned behavior with emphasis on impulsive behavior and sensation seeking

Abstract

Objectives

The increasing prevalence of drug use in the Republic of Korea has emerged as a significant social concern. This study applied the extended theory of planned behavior to investigate the factors influencing intentions to use drugs among young adults (aged 20–30 years) in the Republic of Korea. The study integrated personal traits—specifically, impulsivity, sensation seeking, and self-efficacy—into 2 core theory of planned behavior constructs: attitudes and subjective norms. The principal aim was to improve the prediction of drug use intentions by incorporating these variables.

Methods

Data were obtained from the 2023 National Survey of Drug Harm Perception, which sampled 1,500 individuals aged 19 to 39 years. Hierarchical regression analysis was employed to assess the influence of psychological and social factors on intentions to use drugs.

Results

In the primary model, age emerged as a significant predictor of drug use intentions (R²=0.01). The secondary model showed that positive attitudes toward drugs, subjective norms, and lower self-efficacy significantly increased drug use intentions (R²=0.23). In the final tertiary model, the addition of sensation seeking and impulsivity further amplified these intentions (R²=0.25).

Conclusion

The findings underscore the pivotal roles of attitudes, subjective norms, and self-efficacy in shaping intentions to use drugs. Sensation seeking and impulsivity were found to further elevate vulnerability. Effective prevention efforts must address both psychological traits and social influences. Future research should examine the long-term behavioral outcomes associated with these factors.

Graphical abstract

Introduction

The Republic of Korea has witnessed a substantial rise in illicit drug use since the coronavirus disease 2019 pandemic, encompassing narcotics, psychotropic drugs, and cannabis [1]. The Psychotropic Drug Control Act classifies these substances as highly dangerous and strictly regulates their use. Despite rigorous enforcement, both the incidence of drug-related offenses and the prevalence of drug-related criminal activities continue to climb [2].

Recent data from the National Police Agency and the Supreme Prosecutors’ Office [3] indicate a marked increase in arrests for drug-related crimes. Specifically, the number of drug offenders has more than tripled, rising from fewer than 9,000 in 2010 to over 27,000 in 2023. Individuals aged 20 to 30 years accounted for more than half of these cases, with 8,368 offenders (30.3%) in their 20s and 6,683 (24.2%) in their 30s in 2023 [4]. This highlights the urgent need for targeted prevention strategies for this high-risk demographic. Furthermore, the National Bureau of Investigation reported an alarming 172.7% increase in drug-related deaths over a decade, with fatalities climbing from 205 in 2011 to 559 in 2021. These figures starkly illustrate the severe and growing public health crisis posed by drug abuse in the Republic of Korea. Additionally, drug distribution has shifted increasingly from offline to online platforms, facilitated by the Internet, social media, and cryptocurrency transactions [5].

It is evident that intensifying regulatory measures alone is inadequate for combating drug abuse in the Republic of Korea. The development of prevention strategies that consider both social and psychological factors is imperative. Currently, there is a lack of research analyzing the psychological and social determinants of drug use among Korean users. Most existing studies focus on strengthening legal regulations and imposing harsher penalties. Moreover, the stigmatization of drug users complicates prevention efforts by reducing their willingness to participate in research, thereby impeding the collection of reliable data. There is thus a pressing need for research that thoroughly examines the psychological and social influences on drug use in the Republic of Korea, enabling the design of more effective and targeted prevention strategies.

The theory of planned behavior (TPB) provides a valuable theoretical framework for understanding drug use intentions. It posits that attitudes, subjective norms, and perceived behavioral control shape intentions to engage in behaviors such as drug use. The extended TPB (ETPB) builds upon this model by incorporating personal traits—including impulsivity, sensation seeking, and self-efficacy—which enhance the accuracy of predictions regarding drug use intentions [6–9]. Individuals with high impulsivity may be more vulnerable to drug use due to diminished behavioral control. In contrast, those with elevated sensation-seeking tendencies may be more likely to experiment with drugs in pursuit of novel experiences [10–14]. Individuals with higher self-efficacy are better equipped to regulate their drug use. These individual differences underscore the necessity of an extended model to provide a more comprehensive account of drug use intentions than the standard TPB.

This study aimed to apply the ETPB to identify factors influencing drug use intentions among young adults (aged 20–30 years) in the Republic of Korea. Personal traits—specifically impulsivity, sensation seeking, and self-efficacy—were incorporated into the core TPB components of attitudes and subjective norms. The primary objective was to enhance the prediction of drug use intentions by including these factors. The findings are expected to inform the development of practical, effective prevention and intervention strategies. By taking into account individual psychological and behavioral characteristics, these approaches will address the limitations of current prevention efforts that focus predominantly on legal deterrents.

Materials and Methods

Study Design and Participants

This study was conducted in 2023, utilizing data from a total of 3,000 participants aged 19 to 59 years [4]. To better understand drug use intentions among younger adults, a subset of 1,500 participants aged 19 to 39 years was selected for secondary analysis. The demographic and other characteristics of these 1,500 participants are presented in Table 1. The original 2023 survey employed a cross-sectional design with self-reported data, using stratified sampling based on gender, age, and geographical region.

Table 1.

Demographic characteristics of participants

Variable	Total (n=1,500)	Man (n=750)	Woman (n=750)
Age (y)
19–29	750 (50.0)	375 (50.0)	375 (50.0)
30–39	750 (50.0)	375 (50.0)	375 (50.0)
Education level
Less than a high school diploma	324 (21.6)	186 (24.8)	138 (18.4)
High school diploma or higher	1176 (78.4)	564 (75.2)	612 (81.6)
Occupation
Professional/business/management	757 (50.5)	358 (47.7)	399 (53.2)
Labor/production/transportation/self-employed/sales	207 (13.8)	137 (18.3)	70 (9.3)
Student/unemployed/other	536 (35.7)	255 (34.0)	281 (37.5)
Monthly household income (million KRW)
Less than 2	175 (11.7)	93 (12.4)	82 (10.9)
2–4	546 (36.4)	282 (37.6)	264 (35.2)
4–6	328 (21.9)	177 (23.6)	151 (20.1)
More than 6	451 (30.1)	198 (26.4)	253 (33.7)

Data are presented as n (%).

KRW, Korean won.

Variables

This study employed the ETPB to examine associations between psychological and social factors and drug use intentions. To ensure reliability and validity, all study variables were measured using scales previously validated in the literature. All items were scored on a 5-point Likert scale (1=not at all, 5=very much). For each variable, scores were calculated by summing the relevant item scores, with higher values indicating greater expression of the given psychological or social factor.

Demographic characteristics

Demographic variables served as control variables: gender (male/female), age in years (continuous variable), education level (2-level categorical variable), monthly household income (4 income quintiles), and occupation (3-level categorical variable). Gender was included to capture differences in drug use patterns between men and women [14], while age was included to account for varying risks of experimental drug use across the life span [15]. Education level and monthly household income were considered due to their established impact on drug accessibility and availability [10,11]. Occupation was categorized into 3 groups: (1) professional/business/management; (2) labor/production/transportation/self-employed/sales; and (3) student/unemployed/other. Including this variable enabled assessment of potential differences in social structure, economic stability, and daily routine. Prior research indicates that employment status significantly influences substance use and risk perception. Individuals who are unemployed or working irregular jobs may experience greater psychological stress, social isolation, or exposure to high-risk environments—factors associated with heightened vulnerability to drug use [16].

Dependent variable

The dependent variable, drug use intention (i.e., likelihood of future drug use), was assessed by summing scores from 4 items measuring the likelihood of drug use in the following year. The reliability coefficient (Cronbach’s α) was 0.66.

Independent variables

TPB constructs

Attitude variables (positive drug attitudes, negative drug attitudes) were measured using a scale evaluating both positive and negative attitudes toward drug use. Positive drug attitudes were measured by summing 3 items that assessed respondents' beliefs regarding whether drug use would help them recover from fatigue, lose weight, or feel more alert (Cronbach’s α=0.77). Negative drug attitudes were measured by summing 3 items that gauged perceptions of the probability of adverse consequences from drug use, such as sleep disturbances, digestive problems, or physical alterations like damaged teeth (Cronbach’s α=0.89). Higher scores reflected stronger positive or negative attitudes, respectively.

Subjective norms were evaluated with 2 subscales: public subjective norms and peer subjective norms. Public subjective norms assessed perceptions of societal approval of drug use, while peer subjective norms measured perceived expectations of close friends regarding drug use. Each subscale consisted of 3 items. Both subscales demonstrated acceptable internal consistency (Cronbach’s α=0.87 for peer subjective norms; α=0.81 for public subjective norms). Higher scores indicated greater perceived social pressure to use drugs. Because respondents might be uncertain about how drug use is perceived by the general public or their peers, a “don’t know” option was included for all subjective norms items. These responses were excluded from mean score calculations. This option was not included for other variables.

Self-efficacy was measured by summing 3 items assessing an individual's confidence in their ability to regulate drug use (Cronbach’s α=0.79), with higher scores indicating greater self-efficacy.

Propensity

Impulsivity was measured with 4 items from the Barratt impulsiveness scale (BIS-11), assessing the tendency to act impulsively (Cronbach’s α=0.63). Higher scores indicated greater impulsivity [11].

Sensation seeking was measured using 3 items from Zuckerman [17]’s sensation seeking scale, which assesses the tendency to seek novel or risky stimuli (Cronbach’s α=0.76). Higher scores reflected greater sensation-seeking propensity.

Statistical Methods

All analyses were conducted using IBM SPSS Statistics for Windows ver. 25.0 (IBM Corp.). Descriptive statistics were calculated to summarize participant characteristics. Pearson correlation coefficients were used to examine the relationships between independent variables (e.g., positive and negative drug attitudes, subjective norms) and the dependent variable (drug use intentions). Cronbach’s α coefficients were calculated to evaluate the internal consistency of the study’s scales. Multiple linear regression was used to examine the influence of independent variables on drug use intention, which was analyzed as a continuous variable. The analysis included assessments for multicollinearity using variance inflation factors, and checked the assumptions of independence.

Ethics Statement

This study, which was conducted in 2023, received approval from the Institutional Review Board of Sahmyook University (approval number: SYU 2023-05-008-001).

Results

Participants

Among the total sample of respondents (n=1,500), gender distribution was balanced, with 750 men and 750 women. Regarding age, 50% (n=750) of participants were between 19 and 29 years old, while 25% (n=375) were aged 30 to 39 years. In terms of education, 78.4% (n=1,176) held a high school diploma or higher, comprising 75.2% (n=564) of men and 81.6% (n=612) of women. The remaining 21.6% (n=324) reported having less than a high school diploma.

Professional, business, and management positions constituted the largest proportion of occupations (50.5%), followed by the student/unemployed/other category (35.7%, n=536). The labor/production/transportation/self-employed/sales category accounted for a relatively small proportion (13.8%, n=207) of the total sample.

Monthly household income most commonly fell within the 2 to 4 million Korean won (KRW) range (36.4%), with little difference between men (37.6%) and women (35.2%). Many respondents reported a monthly household income exceeding 6 million KRW (30.1%). Incomes between 4 to 6 million KRW were slightly more prevalent among men (23.6%) than women (20.1%). A minority of participants (11.7%) reported monthly household incomes below 2 million KRW.

Descriptive Statistics and Normality Analysis

Descriptive statistics are summarized in Table 2. The mean score for positive drug attitudes was 2.09±0.96 out of 5, indicating that respondents generally did not hold favorable perceptions of drugs. In contrast, the mean score for negative drug attitudes was high (3.97±1.03), reflecting a strong recognition of the harmful effects of drug use.

Table 2.

Descriptive statistics and normality analysis

Variable	Name	Min–max	Mean±SD	Skewness
Theory of planned behavior	Positive drug attitudes	1–5	2.09±0.96	0.57
	Negative drug attitudes	1–5	3.97±1.03	–1.26
	Peer subjective norms	1.3–5	4.71±0.55	–1.99
	Public subjective norms	1.4–5	4.44±0.72	–1.36
	Self-efficacy	1–5	4.38±0.74	–1.47
Propensity	Sensation seeking	1–4	1.67±0.62	0.94
	Impulsivity	1–4	2.22±0.51	–0.13
	Log drug use intention	0–0.51	0.03±0.08	2.94

Min, minimum; Max, maximum; SD, standard deviation.

For subjective norms, respondents reported perceiving robust social norms against drug use. The mean perceived peer disapproval of drug use was 4.71±0.55 out of 5, while the mean response for public disapproval was 4.44±0.72.

Young adults in the sample showed a greater propensity for impulsivity than for sensation seeking. The mean sensation seeking score was 1.67±0.62, and the mean impulsivity score was 2.22±0.51 [11].

Most participants indicated no intention to use drugs, as reflected by a mean drug use intentions score of 0.03±0.08, ranging from a minimum of 0 to a maximum of 0.51. This resulted in a right-skewed distribution with a pronounced long tail for drug use intentions, suggesting that the normality assumption may not have been met; log transformation was thus performed to achieve normality.

Skewness and kurtosis were analyzed to assess the normality of all variables. Skewness measures distribution asymmetry, and kurtosis assesses the peakedness or concentration. Normality is considered violated when the absolute value of skewness exceeds 3 or kurtosis exceeds 10 [18]. For all study variables, skewness and kurtosis were within the established thresholds, as shown in Table 2, confirming that normality assumptions were met.

Correlation Analysis

Table 3 presents the results of the correlation analysis for all respondents. The Pearson correlation coefficient between positive drug attitudes and drug use intentions was r=0.20 (p<0.01), indicating that stronger positive attitudes toward drugs corresponded with higher intentions to use drugs. The correlation coefficient between negative drug attitudes and drug use intentions was r=–0.08 (p<0.01), suggesting that stronger negative perceptions of drugs tended to decrease drug use intentions, though the effect was relatively weak. The correlation coefficients between peer subjective norms and drug use intentions, and between public subjective norms and drug use intentions, were r=–0.45 (p<0.01) and r=–0.33 (p<0.01), respectively. These findings indicate that stronger subjective perceptions of social norms against drug use—both from peers and the general public—were associated with lower intentions to use drugs. Thus, individuals who perceive negative attitudes toward drug use from their peers and the broader society are less likely to intend to use drugs. The correlation coefficient between self-efficacy and drug use intentions was r=–0.27 (p<0.01), indicating that higher self-efficacy tended to be associated with lower intentions to use drugs. This underscores the pivotal role of self-control, defined as the capacity to refuse drug use, in inhibiting drug use intentions. Furthermore, the analysis showed a correlation coefficient of r=0.27 (p<0.01) for the association between sensation seeking and drug use intentions. A similarly significant correlation was observed between impulsivity and drug use intentions (r=0.20; p<0.01), suggesting that impulsivity also influences the intention to use drugs.

Table 3.

Correlation analysis of study variables for all participants (n=1,500)

Variable	1	2	3	4	5	6	7	8
1. Positive drug attitudes	-
2. Negative drug attitudes	–0.02	-
3. Peer subjective norms	–0.23***	0.13***	-
4. Public subjective norms	–0.20***	0.07*	0.53***	-
5. Self-efficacy	–0.24***	0.26***	0.32***	0.18***	-
6. Sensation seeking	0.22***	–0.14***	–0.32***	–0.19***	–0.33***	-
7. Impulsivity	0.19***	–0.05	–0.15***	–0.17***	–0.27***	0.27***	-
8. Drug use intentions	0.20***	–0.08**	–0.45***	–0.33***	–0.27***	0.27***	0.20***	-

^*p<0.05,

^**p<0.01,

^***p<0.001.

Factors Influencing Drug Use Intentions

Table 4 presents the results of the hierarchical regression analysis, which was conducted to identify the main factors influencing the intention to use drugs. Peer subjective norms had the highest standardized beta value (β) across all analyses, indicating that this variable exerted the greatest impact on drug use intentions in this study. The secondary analysis coefficient (β=–0.34, p<0.001) and the tertiary analysis coefficient (β=–0.32, p<0.001) were both significant, demonstrating that stronger peer subjective perceptions of social norms against drug use were associated with lower intentions to use drugs. These findings suggest that peer social norms and pressures may deter drug use.

Table 4.

Hierarchical regression analysis of factors influencing drug use intentions

	Primary			Secondary			Tertiary
	β	SE	p	β	SE	p	β	SE	p
Gender	0.00	0.00	0.920	0.03	0.00	0.269	0.05	0.00	0.030
Age	–0.06	0.00	0.039	–0.04	0.00	0.149	–0.03	0.00	0.191
Education level	–0.02	0.00	0.548	–0.02	0.00	0.391	–0.02	0.00	0.379
Occupation	0.00	0.00	0.880	0.01	0.00	0.568	0.01	0.00	0.690
Monthly household income	–0.05	0.00	0.059	–0.03	0.00	0.135	–0.03	0.00	0.148
Positive drug attitudes				0.08	0.00	0.001	0.05	0.00	0.023
Negative drug attitudes				0.00	0.00	0.963	0.00	0.00	0.843
Peer subjective norms				–0.34	0.00	0.000	–0.32	0.00	0.000
Public subjective norms				–0.10	0.00	0.000	–0.09	0.00	0.001
Self-efficacy				–0.12	0.00	0.000	–0.08	0.00	0.002
Sensation seeking							0.11	0.00	0.000
Impulsivity							0.07	0.00	0.004
${\mathrm{R}}_{\mathrm{adj}}^2$	0.01			0.23			0.25
Change in ${\mathrm{R}}_{\mathrm{adj}}^2$				0.22			0.02

SE, standard error.

Secondary and tertiary analyses revealed a negative relationship between self-efficacy and drug use intentions (β=–0.12 and β=–0.08, respectively; p<0.001), indicating that higher self-efficacy was associated with a reduced intention to use drugs. Positive drug attitudes showed a significant positive coefficient in both the secondary (β=0.08, p<0.01) and tertiary analyses (β=0.05, p<0.05), confirming that positive attitudes toward drugs were associated with an increased intention to use drugs. Additionally, the tertiary analysis indicated that impulsivity (β=0.11, p<0.001) and sensation seeking (β=0.07, p<0.01) had significant effects on drug use intentions.

Furthermore, gender was found to have a significant effect on drug use intentions in the tertiary analysis only (β=0.05, p<0.05). This suggests that gender differences may influence intentions to use drugs, but their impact is relatively modest compared to other key variables. Age, education level, occupation, monthly household income, and negative drug attitudes did not have significant effects on drug use intentions. In the primary analysis, the age coefficient was significant (β=–0.06, p<0.05), but it became nonsignificant in subsequent analyses. Education level, occupation, and monthly household income were also nonsignificant, with standardized coefficients close to zero and p-values exceeding 0.1 in all analyses.

The model's explanatory power (R²) was 0.01 in the primary analysis, indicating that demographic control variables contributed minimally to explaining drug use intentions. In the secondary analysis, R² increased to 0.23, suggesting that the TPB variables (positive drug attitudes, negative drug attitudes, subjective norms, and self-efficacy) played a substantial role in predicting drug use intentions. In the tertiary analysis, R² further increased to 0.25, indicating that the addition of propensity variables (sensation seeking and impulsivity) provided a modest improvement in the model’s explanatory power.

Discussion

In recent years, increasing drug use among young adults has become a prominent social issue in the Republic of Korea, raising widespread concern about its broader implications. This phenomenon can be attributed to a complex interplay between psychological (individual) and social factors. Greater access to drugs and evolving social attitudes toward drug use have shaped young adults’ intentions to use drugs. Against this backdrop, the present study empirically identified and analyzed key determinants of young adults’ drug use intentions to provide a multidimensional understanding of this issue. Hierarchical regression analysis revealed that subjective norms, self-efficacy, sensation seeking, positive attitudes toward drugs, and impulsivity were all significant predictors of drug use intentions. In contrast, demographic factors—including age, education, occupation, and monthly income—did not significantly influence drug use intentions. These findings indicate that individual psychological and behavioral characteristics, along with the social environment, play a more substantial role in shaping drug use intentions than demographic characteristics [19].

Our findings demonstrate that positive attitudes toward drug use significantly increase intentions to use drugs, whereas negative attitudes exerted no significant effect. This is consistent with previous research by McMillan and Conner [20], who found that positive attitudes are a major predictor of drug use intentions, but negative attitudes do not necessarily reduce intentions. Young people are therefore more likely to be influenced by the perceived benefits of drug use, such as stimulation, weight loss, and stress relief. As a result, prevention strategies that simply reinforce negative perceptions are unlikely to be sufficiently effective [21,22]. One plausible explanation is that young adults may cognitively discount or normalize potential harms, especially when the perceived benefits are immediate or personally relevant. This asymmetry in influence suggests that health communication strategies focusing only on negative consequences are inadequate. Instead, effective prevention efforts should challenge and reframe perceived benefits of drug use, and offer realistic alternatives that fulfill similar psychological or emotional needs (e.g., stress management or energy enhancement). Interventions should therefore prioritize reducing positive expectations and emphasize long-term negative consequences to maximize deterrence.

Subjective norms emerged as the most robust predictor of drug use intentions in this study. Stronger perceptions of social norms against drug use were associated with significantly lower intentions to use drugs, consistent with the TPB, which posits that social norms and others’ expectations moderate behavioral intentions [6]. Kam et al. [9] similarly found that subjective norms reduce intentions to use drugs, and that social norms can indirectly influence intentions by interacting with attitudes and perceived behavioral control. When young adults perceive drug use as socially unacceptable, their intentions to use drugs are diminished. Accordingly, interventions that reinforce social norms—such as media campaigns or peer-based programs—may be effective in reducing drug use intentions.

Self-efficacy was also identified as an important inhibitor of drug use intentions. This study introduced and analyzed self-efficacy as a concept closely related to perceived behavioral control in the TPB framework. Higher self-efficacy was associated with a lower likelihood of intending to use drugs. This finding aligns with Bandura’s theory, which holds that individuals with greater self-efficacy are better able to regulate their behavior and resist external pressures [23]. Enhancing self-efficacy may thus be a powerful component of prevention education, enabling young adults to exercise greater control over their actions despite environmental influences [24].

Sensation seeking was shown to be a significant predictor of intentions to use drugs. Individuals with high sensation-seeking tendencies are more inclined to seek out novel stimuli, which may, in turn, increase their interest in drug use [17]. Several studies have similarly reported that young adults with high sensation-seeking tendencies are more likely to engage in substance use and exhibit a greater propensity for drug use [15,25]. These individuals may also develop more positive attitudes toward drug use, as the pursuit of stimulating experiences can increase their intentions to use drugs. Consequently, it is essential to develop prevention strategies that target and reduce positive attitudes toward drug use among young adults. This may be accomplished by providing alternative activities—such as extreme sports or creative hobbies—that can satisfy sensation-seeking tendencies in healthier ways .

This study also identified impulsivity as a significant predictor of drug use intentions. Moshier and Otto [13] found that impulsivity can widen the gap between intentions to use drugs and actual drug use behavior. Highly impulsive young adults may lack cognitive inhibition, thereby increasing the likelihood of acting on their intentions, such as using drugs. This highlights impulse control training as a potentially effective prevention strategy, particularly for young adults. By helping young adults improve impulse regulation and self-control, it may be possible to reduce the likelihood of drug use.

Demographic characteristics—including age, education level, occupation, and monthly household income—did not have a significant impact on drug use intentions in this study. While age was significant in the primary analysis, its effect became nonsignificant in later analyses. Thus, drug use intentions appear to be influenced more strongly by psychological and social factors than by basic demographic characteristics.

Variables derived from the TPB were robust predictors of drug use intentions, underscoring the importance of a multifaceted approach that incorporates both individual characteristics and social influences. Our findings suggest that interventions designed to reduce drug use would be more effective if tailored to psychological and social factors, rather than demographic variables alone. This study therefore emphasizes the need to design comprehensive prevention education and intervention programs that reflect the diverse characteristics and social environments of young people.

Conclusion

The findings of this study underscore the necessity of a multifaceted approach to reducing drug use among young adults. Specifically, there is a clear need for interventions that address a range of factors, including subjective norms, self-efficacy, sensation seeking, and impulsivity. The results indicate that prevention strategies focusing on reinforcing social norms against drug use, enhancing self-efficacy, and providing alternative activities to manage sensation-seeking and impulsive tendencies may be more effective than approaches relying solely on stricter regulation or harsher punishment. Future research should consider the impact of varying cultural contexts and environmental factors on drug use intentions to further refine prevention strategies.

HIGHLIGHTS

A cross-sectional study applying the extended theory of planned behavior analyzed a nationally representative sample of individuals aged 19–39 to explore psychological and social factors influencing drug use intentions.

Hierarchical regression analysis revealed that peer subjective norms and self-efficacy were strongly associated with reduced intentions, while sensation seeking and impulsivity were significantly linked to increased intentions. The findings highlight the need for multidimensional prevention strategies that address both individual traits and perceived social influences.