Compulsive Buying in Adolescents. The Impact of Gender and Alcohol and Cannabis Use

: The purpose of this study to analyse whether compulsive buying in teenagers is related to gender and alcohol and cannabis use in a sample of 573 students aged 14–17 from secondary education schools in Burgos (Spain) (M = 15.65; SD = 1.04). Random cluster sampling was performed to select the sample. The Compulsive Buying Questionnaire was used together with two extra promts: ‘Indicate how much alcohol you consume’ and ‘Indicate how much cannabis you take’. Descriptive statistics were used in data analysis, while MANOVA was used to study gender differences in alcohol and cannabis use, compulsive buying and their interaction. The results show higher scores for female compulsive buyers than for men, higher scores for alcohol and cannabis users’ compulsive buying than for non-users, respectively, and higher scores for female users than for male users. A certain interaction was also observed between alcohol and cannabis use. A higher alcohol consumption entailed a higher score in compulsive buying, with cannabis users who did not consume alcohol obtaining the highest scores. Thus, prevention programmes should consider teenagers’ gender and the risk of taking toxic substances.


Introduction
Addiction has historically been associated with drug use in the population's collective imagination. The concept has now been redefined to no longer be limited to behaviour generated by uncontrollable substance use but to also cover apparently harmless behaviour habits which may, in certain circumstances, become addictive and seriously affect the addicted individuals' daily life [1]. In other words, some behaviours commonly considered normal, or even healthy, may become troublesome depending on the frequency, the intensity and the time and/or money invested in them [2]. Any regular activity which turns out to be pleasurable for a given individual may turn into a harmful behaviour when the latter loses self-control over such an activity and continues to do it despite the wide range of adverse implications involved, thus giving rise to an obsession-a constant and unavoidable desire that becomes the main interest or activity of the subject, who eventually puts the other activities aside [3].
Therefore, beyond any disorders related to substance abuse, the concept of addiction must be widened to cover any addiction disorders not related to substance abuse [4,5]. Moreover, there is an increasing interest in studying behavioural addictions, as increasing numbers of people have become concerned by such problematic attitudes, given their variety and general prevalence [6].
One of the regular activities which may become addictive is buying [7]. There has been an increasing focus on buying as subject of study in the last years, and the scientific literature suggests that compulsive buying may be object of addiction [8][9][10][11]. This behavioural addiction may be caused by the consumer society itself, where shopping is becoming an increasingly popular recreational activity and a way to escape daily routine, rather than a merely transactional activity to meet material needs [12]. With the aim of clarifying this concept, it can be said that compulsive buying is based on an irresistible and dominant desire to purchase goods beyond one's needs and resources. Compulsive buyers show concerns, impulses and/or unadaptive and uncontrollable behaviour and can only find relief in excessive buying. However, the satisfaction obtained becomes quickly replaced by a feeling of guilt and, with time, this behaviour becomes automatic, activated by emotions and impulses, making it difficult to control cognitively, and the activity is accompanied by a great deal of negative self-criticism [13,14]. The situation becomes worse as the person loses control over their own behaviour and continues the compulsive buying habits, seeking immediate rewards and not thinking of the adverse effects on their personal, familial, social and financial situation. Buying itself then becomes a pathology [2,9,[13][14][15].
Some studies exploring the prevalence of compulsive buying in different countries have revealed an increase in the number of compulsive shoppers over the last years [16,17]. Several studies show that 1-11.3% of the general population are compulsive buyers [12,17,18]; this value is 7.1% in the particular case of Spain [19]. The increasing rate of this type of buyers and the major psychological, familial and economic implications associated with this activity, among other factors, have eventually turned compulsive buying into one of the most serious behavioural addictions, thus triggering a large amount of research activity in the field [17].
In this context, it can be suggested that the risk for citizens, and especially teenagers, to develop pathological buying behaviour may be higher in increasingly consumerist societies than in those which foster austere lifestyles [15]. The findings seem to point out that compulsive buying starts in adolescence-a developmental stage where different psychological changes take place, in which values, attitudes and preferences are adopted and behaviours related to individual consumption are shaped [20]. Moreover, some authors state that buying itself as a strategy to face stress is now remarkably widespread amongst both young boys and girls [21,22].
As regards gender, most studies reveal that compulsive buying is a mostly female pathology [23][24][25] which mainly affects people aged 18-30 [26]. On the contrary, Muller et al. [27] detected that men are increasingly showing this type of behaviour.
On the other hand, behavioural addictions are frequently combined with the abuse of one or more types of substances [1]. The results of some research conducted on adolescents suggest that shopping addiction is related to other excessive behaviours, such as abusing alcohol and other psychoactive substances [11]. However, other authors do not find such a relation [28].
It is widely known that alcohol and cannabis use is frequent amongst teenagers. The former is the most used substance by people aged 14-18 in Europe, where 76.9% of the population admit to ever having drunk alcohol, compared to 19.4% who admit to ever having used cannabis. According to gender, alcohol consumption is slightly more spread amongst females (78.2%) than amongst males (75.7%), which is not the case with cannabis, as males (16.3%) are more often abusers of this substance than females (9.8%), and this abuse tends to increase with age [29]. Furthermore, some studies indicate that cannabis is the most popular psychoactive substance amongst alcohol users [30,31].
In recent years, compulsive buying has been conceptualised as a behavioural addiction. However, little attention has been paid to the study of its co-occurrence with addiction to certain substances, and even less so in teenagers, who are especially vulnerable not only because they are currently considered major users but also due to the harmful effects of substance abuse in their development [32]. The goal of the present study is to fill this gap by detecting the prevalence of adolescents' compulsive buying depending on gender and age and its concurrence with alcohol and cannabis abuse in order to determine any possible interaction among these variables.

Sample
The study was performed on a sample made up of 573 teenagers attending eight fully or partly state-funded high schools in the city of Burgos (Spain). Of the total, 50.7% were female and 49.3% were male. Subjects were aged 14-17 (M = 15.65; SD = 1.04). In total, 24% were enrolled in the third year of compulsory secondary education, 27.4% in the fourth year, 28.4% in the first year of post-compulsory secondary education and 20.1% were enrolled in the second and last years. In relation to cannabis use, 29.8% of the sample admitted to using cannabis occasionally or frequently, while 80.2% alleged that they had not ever used it. As regards alcohol consumption, only 26.7% reported that they would never consume alcohol.
Another variable analysed was the respondents' family environment and, to this end, we asked them who they usually lived with. The vast majority of the adolescents surveyed lived with their mother and father, 77% shared a home with their siblings and a much smaller number said they lived with other relatives. The socioeconomic levels were diverse, as an attempt was made to ensure that all neighborhoods in the city were represented.
To maintain the representativeness of the study, the sample was weighted by sex, school ownership and educational level.

Instrument
A seven-question Compulsive Buying Questionnaire with five Likert-scale responses (from 1 = never to 5 = always) was used to assess buying habits. There have been no studies that have analysed the psychometric properties of such a questionnaire, so an exploratory factor analysis was performed, resulting in one sole factor, comprising six items, making up 50.42% of the variance. In one of the items, a saturation of 0.379 was obtained ("Do you think that any friend of yours has problems with shopping?"), and it was disregarded in the calculation of the total compulsive buying score. The Cronbach's α reliability of the questionnaire in this study with the six final items was 0.848.
To assess substance use, two prompts were included: "Indicate how much alcohol you consume" and "Indicate how much cannabis you take", the responses of which were sorted by use or non-use.

Procedure
First, a meeting was held with officials of the Provincial Education Directorate of Burgos (Spain) to inform them about the study, its objectives and purpose. It was agreed that we would only include only schools located in the city of Burgos for the study. Subsequently, the sample was selected using random cluster sampling, with the sampling unit being the educational centers. The Provincial Directorate informed the school principals about the study, and they were responsible for obtaining the families' permissions. It was ensured that the number of public or private/subsidised schools in the study was representative of the reality observed (five public and three subsidised schools). The only inclusion criterion for an adolescent to participate in the study was to be enrolled in one of the grades included in the study. No exclusion criteria were established.
Subsequently, informative letters were sent to the directors of the selected centers, and in a face-to-face meeting, the director was informed of the objectives and steps of the research. Next, informed consent was requested from the students of the schools. In the case of underage students, the school management requested informed consent from the parents or legal guardians of the students. All the students in the selected schools agreed to participate in the study.
The research team, together with the group tutor, provided the booklet with the measurement instruments to the participants. The approximate duration of the application of the tests was 20 min. These tests were applied collectively in the respective classrooms, and the same order was always followed in the application of the instruments. To control social desirability in the responses, the students were informed of the confidentiality of the Sustainability 2021, 13, 3982 4 of 10 data and that the responses were anonymous. Likewise, once the students began to fill out the questionnaires, the tutor of each class was asked to leave the classroom.

Data Analysis
Data analysis was performed with the SPSS-25 statistical package, using MANOVA to assess any differences in compulsive buying according to gender, age, alcohol and cannabis use and to determine any possible interaction among these factors which could influence compulsive buying.
In case the main effects of the factors of three or more levels or the interactions among them were significant, individual ANOVA and post hoc analyses were performed later. To determine the size of the effect, the partial eta squared approach was used while following Richardson's [33] interpretation suggestions; thus, an effect size of 0.0099 was considered small, 0.0588 was considered medium and 0.1397 was considered large.

Results
Multivariate analysis was performed to determine the influence of gender, age, alcohol and cannabis use and the interaction among them on the surveyed teenagers' compulsive buying habits. As shown in Table 1, significant differences were observed in gender and cannabis use, as well as interactions between gender and alcohol consumption, alcohol and cannabis use and lastly between gender, alcohol consumption and cannabis use. No significant differences were observed in age or alcohol consumption factors, nor in the rest of the interactions (age and substance use, gender and age or the interaction among gender, age and substance use). As regards gender, females (M = 12.89; SD = 0.29) were reported to obtain a significantly higher score for compulsive buying than males (M = 8.94; SD = 0.22), as shown in Figure 1. The effect size of these differences was moderate, 0.17, but it was the highest size effect obtained (see Table 1). With respect to the second factor. which showed significant differences for cannabis use, Figure 2 shows that those teenagers who usually took cannabis (M = 10.64; SD = 0.21) obtained a higher score for compulsive buying than those who did not (M = 12.16; SD = 0.5.1). These differences had a moderate effect size of 0.014 (see Table 1). As regards the interaction between gender and alcohol consumption, according to Figure 3, male alcohol consumers hardly showed differences in compulsive buying scores (M = 9.23, SD = 3.75) when compared to male non-drinkers (M = 8.72, SD = 3.53). However, female alcohol consumers showed a significantly higher score for compulsive buying (M = 19.93, SD = 4.84) than female non-drinkers (M = 11.40, SD = 4.70). The effect size of this interaction was moderate but high, with a partial eta squared value of 0.16 as indicated in Table 1. With respect to the second factor. which showed significant differences for cannabis use, Figure 2 shows that those teenagers who usually took cannabis (M = 10.64; SD = 0.21) obtained a higher score for compulsive buying than those who did not (M = 12.16; SD = 0.5.1). These differences had a moderate effect size of 0.014 (see Table 1). With respect to the second factor. which showed significant differences for cannabis use, Figure 2 shows that those teenagers who usually took cannabis (M = 10.64; SD = 0.21) obtained a higher score for compulsive buying than those who did not (M = 12.16; SD = 0.5.1). These differences had a moderate effect size of 0.014 (see Table 1). As regards the interaction between gender and alcohol consumption, according to Figure 3, male alcohol consumers hardly showed differences in compulsive buying scores (M = 9.23, SD = 3.75) when compared to male non-drinkers (M = 8.72, SD = 3.53). However, female alcohol consumers showed a significantly higher score for compulsive buying (M = 19.93, SD = 4.84) than female non-drinkers (M = 11.40, SD = 4.70). The effect size of this interaction was moderate but high, with a partial eta squared value of 0.16 as indicated in Table 1.  As regards the interaction between gender and alcohol consumption, according to Figure 3, male alcohol consumers hardly showed differences in compulsive buying scores (M = 9.23, SD = 3.75) when compared to male non-drinkers (M = 8.72, SD = 3.53). However, female alcohol consumers showed a significantly higher score for compulsive buying (M = 19.93, SD = 4.84) than female non-drinkers (M = 11.40, SD = 4.70). The effect size of this interaction was moderate but high, with a partial eta squared value of 0.16 as indicated in Table 1.  There was also an interaction effect between alcohol and cannabis use on the compulsive buying habits of the analysed sample. In this case, the subjects scoring highest in compulsive buying were those who did not consume alcohol but who did use cannabis (M = 13.33, SD = 5.07) as represented in Figure 4. The scores were quite similar and even lower than those in the previous case for those subjects who only consumed alcohol (M = 11.70, SD = 4.61) or who used both substances (M = 12.10, SD = 5.59). The non-use of any of the above indicated substances was related to the lowest scores for compulsive buying (M = 9.73, SD = 4.20). As shown in Table 1, this interaction had the smallest size effect, which can be considered moderate-low (partial eta squared of 0.007). Lastly, the results showed an interaction between three variables-gender, alcohol consumption and cannabis use-with a moderate-low size effect (partial eta squared of 0.009) as indicated in Table 1. The results indicated (see Figure 5) that the highest scores for compulsive buying in males were obtained when they only used cannabis (M = 14.33, SD = 5.50). However, there were similarities between the scores obtained by males who  There was also an interaction effect between alcohol and cannabis use on the compulsive buying habits of the analysed sample. In this case, the subjects scoring highest in compulsive buying were those who did not consume alcohol but who did use cannabis (M = 13.33, SD = 5.07) as represented in Figure 4. The scores were quite similar and even lower than those in the previous case for those subjects who only consumed alcohol (M = 11.70, SD = 4.61) or who used both substances (M = 12.10, SD = 5.59). The non-use of any of the above indicated substances was related to the lowest scores for compulsive buying (M = 9.73, SD = 4.20). As shown in Table 1, this interaction had the smallest size effect, which can be considered moderate-low (partial eta squared of 0.007).  There was also an interaction effect between alcohol and cannabis use on the compulsive buying habits of the analysed sample. In this case, the subjects scoring highest in compulsive buying were those who did not consume alcohol but who did use cannabis (M = 13.33, SD = 5.07) as represented in Figure 4. The scores were quite similar and even lower than those in the previous case for those subjects who only consumed alcohol (M = 11.70, SD = 4.61) or who used both substances (M = 12.10, SD = 5.59). The non-use of any of the above indicated substances was related to the lowest scores for compulsive buying (M = 9.73, SD = 4.20). As shown in Table 1, this interaction had the smallest size effect, which can be considered moderate-low (partial eta squared of 0.007). Lastly, the results showed an interaction between three variables-gender, alcohol consumption and cannabis use-with a moderate-low size effect (partial eta squared of 0.009) as indicated in Table 1. The results indicated (see Figure 5) that the highest scores for compulsive buying in males were obtained when they only used cannabis (M = 14.33, SD = 5.50). However, there were similarities between the scores obtained by males who  Lastly, the results showed an interaction between three variables-gender, alcohol consumption and cannabis use-with a moderate-low size effect (partial eta squared of 0.009) as indicated in Table 1. The results indicated (see Figure 5) that the highest scores for compulsive buying in males were obtained when they only used cannabis (M = 14.33, SD = 5.50). However, there were similarities between the scores obtained by males who did did not use any substance (M = 8.62, SD = 3.43), males who only consumed alcohol (M = 9.09, SD = 3.27) and males who used both substances (M = 9.41, SD = 4.35).
The results for females were different. As indicated above, females generally obtained a higher score for compulsive buying than males. However, these compulsive buying scores progressively increased (see Figure 5) when no substances were used (M = 11.33, SD = 4.69), when only cannabis was used (M = 12.83, SD = 5.30), when only alcohol was consumed (M = 13.39, SD = 4.58) and when both substances were used (M = 15.32, SD = 5.24).

Discussion
The goal of the present study was to analyse the prevalence of adolescents' compulsive buying depending on gender, age and concurrence with alcohol and cannabis abuse in order to determine any possible interaction between these variables.
As in other studies [23][24][25], females showed compulsive buying behaviour more often than males. In this study, the highest prevalence can be seen in female users. Furthermore, it has been proved that adolescents who often use toxic substances also experience an irresistible desire to purchase unnecessary products with a greater frequency without evaluating how suitable those products are, probably because alcohol and cannabis users are more impulsive than non-users. These results are supported by the findings of Pérez de Albéniz-Garrote, Rubio and Medina-Gómez [34], which indicate that toxic substance use amongst teenagers is associated with impulsivity. On the other hand, González-López and Lemos [24] found female buyers to be more impulsive than male buyers.
Moreover, this research sought to relate substance use to compulsive buying. Given that this study found that the non-use of toxic substances was related to a lower score in compulsive buying, substance abuse could be considered a risk factor or vulnerability for adolescents to develop co-occurring disorders in the future. Furthermore, an interaction between regular alcohol and cannabis users and compulsive buying has been observed, as higher scores were obtained when the two substances were used or only when cannabis was used, unlike the effect shown when consuming only alcohol. Probably, in increasingly consumerist societies, buying itself has turned into a leisure activity which creates a shortterm feeling of wellness, but which can immediately provoke a feeling of guilt. Compulsive buyers lose control over their behaviour, which leads to use of other substances in an attempt to reduce these emotions. Other studies have also found a strong relationship between buying addiction and substance addiction [1,11]. The results for females were different. As indicated above, females generally obtained a higher score for compulsive buying than males. However, these compulsive buying scores progressively increased (see Figure 5

Discussion
The goal of the present study was to analyse the prevalence of adolescents' compulsive buying depending on gender, age and concurrence with alcohol and cannabis abuse in order to determine any possible interaction between these variables.
As in other studies [23][24][25], females showed compulsive buying behaviour more often than males. In this study, the highest prevalence can be seen in female users. Furthermore, it has been proved that adolescents who often use toxic substances also experience an irresistible desire to purchase unnecessary products with a greater frequency without evaluating how suitable those products are, probably because alcohol and cannabis users are more impulsive than non-users. These results are supported by the findings of Pérez de Albéniz-Garrote, Rubio and Medina-Gómez [34], which indicate that toxic substance use amongst teenagers is associated with impulsivity. On the other hand, González-López and Lemos [24] found female buyers to be more impulsive than male buyers.
Moreover, this research sought to relate substance use to compulsive buying. Given that this study found that the non-use of toxic substances was related to a lower score in compulsive buying, substance abuse could be considered a risk factor or vulnerability for adolescents to develop co-occurring disorders in the future. Furthermore, an interaction between regular alcohol and cannabis users and compulsive buying has been observed, as higher scores were obtained when the two substances were used or only when cannabis was used, unlike the effect shown when consuming only alcohol. Probably, in increasingly consumerist societies, buying itself has turned into a leisure activity which creates a shortterm feeling of wellness, but which can immediately provoke a feeling of guilt. Compulsive buyers lose control over their behaviour, which leads to use of other substances in an attempt to reduce these emotions. Other studies have also found a strong relationship between buying addiction and substance addiction [1,11]. There results may possibly be related to other personal, familial and social factors that have not been taken into account in this study but which may provoke a greater vulnerability and concurrence of these addictive behaviours; therefore, there is a need for future studies to address the underlying vulnerabilities and mechanisms for more effective prevention programmes in a society that is becoming increasingly consumerist. The results reflect the importance of the early detection of both toxic substance use and compulsive buying habits in adolescents, given their interrelation, especially amongst females.
This study has some shortcomings which must be highlighted. Firstly, it would be of great interest to use a longitudinal design rather than a cross-sectional one, since the former is more appropriate for analysing long-term variables. On the other hand, we must also take into account the restrictions of using questionnaires to collect this information, since we are assessing the adolescents' perception of their behavior or consumption.
Likewise, asking adolescents themselves about their alcohol or cannabis consumption can lead to a bias in the responses due to social desirability. An attempt has been made to alleviate this limitation by guaranteeing the anonymity of the responses, but it would be more rigorous to have an instrument that assesses this variable.

Conclusions
Compulsive buying is a pathology that has its onset in late adolescence or early adulthood and is usually featured by a lack of self-control and an impulsive behaviour. This might explain why teenage compulsive buying entails a greater risk of alcohol and cannabis abuse. Most researchers suggest that there are underlying neurobiological mechanisms similar to those of substance addiction disorders.
Given the serious implications of this problem on individuals' lives, early prevention becomes fundamental. At this stage, major mental changes occur that determine future behaviour. Thus, a clear and accurate identification of risk behaviours is a priority when scheduling preventive programmes addressed to this population.
Educational interventions have also proved effective when addressing such problems. In this sense, it would be desirable that professionals of the Departments of Guidance designed programmes to take these results into consideration. The early detection of potential at-risk subjects is a way to reduce the negative consequences of addiction.
This study contributes to sustainability issues by introducing a psychological perspective and implementing a sustainable culture among adolescents as responsible consumers. This research will help in the design of more effective educational programs to detect the most vulnerable students.