Psychosocial Determinants of Substance Use Among University Students: Implications for Social-Emotional Learning and Wellbeing

Abstract

Substance use among university students is a complex phenomenon influenced by sociodemographic, contextual, and psychosocial factors, with direct implications for wellbeing and social-emotional learning. This study examined the relationships between sociodemographic variables, substance use, and psychosocial dimensions, using a combined descriptive and correlational analysis. The sample comprised 924 students from different fields of study at the University of Córdoba (Spain): Social and Legal Sciences (65.59%), Engineering and Architecture (3.35%), Arts and Humanities (12.88%), Health Sciences (8.98%), and Sciences (9.09%). Variables assessed included substance use, frequency of use, leisure activities, social function, perceptions and prejudices, intrinsic factors, and behaviour. Results showed significant associations between substance use and leisure activities and intrinsic motivation, as well as interrelationships among psychosocial dimensions. The results also indicated that sociodemographic variables have a limited explanatory role; sex and family context emerged as relevant predictors for certain dimensions. These findings highlight the need to address substance use from a multidimensional perspective that integrates developmental, motivational, and contextual factors, with implications for the design of educational strategies and prevention programmes focused on wellbeing and social-emotional learning in higher education.

1. Introduction

The transition to university represents a critical developmental period in which students face new academic, social, and emotional demands that may increase their vulnerability to substance use. In this context, alcohol and other psychoactive substances are often integrated into leisure activities and peer dynamics, frequently becoming normalised as part of university culture.

This scenario poses a significant challenge both for public health and educational institutions, given its potential impact on academic performance, social relationships, mental health, and overall wellbeing (Ardenghi et al., 2023; Cabanach et al., 2017; Guil et al., 2021). The move to higher education involves exposure to novel social environments, greater autonomy, academic pressures, and new emotional demands, which may heighten students’ vulnerability to risk behaviours, including alcohol and drug use (Panadero et al., 2022; Colomer-Pérez et al., 2019).

Previous research (Boateng et al., 2022; Al Ansari et al., 2022; Merdassa, 2025) indicates that substance use among university students is shaped by a complex interaction of psychosocial factors rather than solely individual choice. Among the most influential determinants are social support networks, coping strategies, perceived stress, and emotional competencies. Students with solid social support and effective coping mechanisms tend to exhibit lower levels of substance use, whereas those experiencing high stress or difficulties in emotional regulation face a greater risk (Neddermann-Carrillo et al., 2024; Ibáñez-Tejedor & Cauli, 2025; Sampaio et al., 2024). These findings underscore the importance of considering both risk and protective factors to understand student behaviour. Although numerous studies have explored individual psychosocial determinants, few have examined them simultaneously within a multidimensional framework. An integrated approach that incorporates multiple factors and their interactions may provide a clearer picture of how these variables collectively influence substance use patterns.

Scientific evidence indicates that these consumption patterns are not explained solely by individual decisions, but by psychosocial determinants such as sex, family socioeconomic level, and parental educational attainment, which influence the frequency, motivations, and social meaning of consumption. In this context, previous studies (Sampaio et al., 2025; Al Ansari et al., 2022; Ibáñez-Tejedor & Cauli, 2025) have documented the relevance of sex as a risk factor, with higher prevalence of intense consumption among men, as well as the association between low family income and substance-related problems. Likewise, parental education is linked to attitudes, perceptions, and underlying motivations for consumption through family socialisation processes (Kollath-Cattano et al., 2020). However, there remains a gap in integrating these sociodemographic variables with the psychosocial dimensions of consumption in the university population, and their explicit connection to social-emotional learning and wellbeing—key competencies for preventing risk behaviours and promoting healthy trajectories.

Understanding these interactions is essential for identifying at-risk groups and tailoring effective interventions that address both the psychological and social dimensions of student life. The educational implications are profound. Higher education institutions can implement evidence-based strategies that promote mental health, reduce risk behaviours, and improve academic outcomes. Programmes that strengthen emotional regulation, enhance stress management, and foster peer support networks have shown promising results in preventing substance use and promoting student wellbeing (Merdassa, 2025; Li et al., 2025; Noar et al., 2020; Llorent-Bedmar et al., 2023). These initiatives not only benefit individual students but also contribute to creating safer, healthier, and more productive learning environments.

Scientific evidence consistently shows that substance use patterns differ by sex, with men generally reporting higher levels of consumption, frequency of use, and greater involvement in risk-related contexts. These differences have been explained by a combination of biological, psychological, and sociocultural factors, including gender norms and coping strategies. In contrast, underlying motivations for substance use tend to be more similar across sexes, suggesting that differences are more pronounced at the behavioural and contextual levels than at the intrinsic level (Erol & Karpyak, 2015).

Socioeconomic status has also been identified as a relevant determinant of substance use, although its influence is complex and sometimes contradictory. Some studies indicate that individuals from higher socioeconomic backgrounds may show increased engagement in alcohol and drug use due to greater access to leisure contexts and social opportunities (Humensky, 2010; Patrick et al., 2012). Conversely, lower socioeconomic status has been associated with higher vulnerability to problematic substance use and related psychosocial stressors, suggesting differentiated patterns depending on the dimension of consumption analysed (Melotti et al., 2013).

Overall, these findings highlight that substance use cannot be understood as a homogeneous behaviour but rather as a multidimensional phenomenon shaped by the interaction of sociodemographic and psychosocial factors. In this regard, variables such as sex and socioeconomic status are not only associated with the frequency and intensity of consumption but also with its social meaning and contextual expression, reinforcing the need for integrative approaches in the study of substance use among university populations.

To contribute to this line of research, the present study adopts an explicitly integrated conceptual framework in which sociodemographic variables (sex, family income, and parental education) operate as distal contextual factors that shape psychosocial dimensions (e.g., intrinsic motivation, social function, and leisure-related consumption), which in turn influence substance use behaviours. These behavioural patterns are further linked to social-emotional learning (SEL) competencies, particularly emotional regulation, self-management, and responsible decision-making.

Within this framework, SEL is conceptualised not merely as an outcome but as a potential mediating and protective mechanism. Furthermore, this study is positioned as a contextual and confirmatory contribution within a Southern European university setting.

In this context, a more detailed theoretical framework is required to explain the relationships between sociodemographic and psychosocial variables and to support the formulation of the study hypotheses.

2. Theoretical Framework

Substance use among university students is widely recognised as a multifactorial phenomenon shaped by the interaction of individual, social, and contextual determinants. From a socioecological perspective, variables such as gender, socioeconomic status, and parental educational level play a key role in shaping patterns of consumption through processes of socialisation, normative influence, and access to resources. Previous research has consistently shown that gender differences in substance use are associated with social norms, risk perception, and culturally embedded expectations regarding behaviour (Erol & Karpyak, 2015; Kuntsche et al., 2017). Similarly, socioeconomic status has been linked to differential patterns of substance use, with both resource availability and psychosocial stressors influencing consumption behaviours (Patrick et al., 2012; Melotti et al., 2013). In addition, parental educational level has been identified as a relevant factor in the intergenerational transmission of attitudes, beliefs, and behavioural norms related to substance use, particularly through family socialisation processes and value internalisation (Humensky, 2010). These theoretical perspectives provide a robust framework for understanding the multidimensional nature of substance use and support the formulation of the following hypotheses.

Given the above, the following objectives were established:

General objective:

To analyse the psychosocial determinants of substance use among university students, considering sex, family income, and parental educational level, and their implications for social-emotional learning and wellbeing.

Specific objectives:

• To examine differences in substance use dimensions (frequency, leisure, social function, perceptions and prejudices, intrinsic factors, behaviour, and overall consumption) according to sex (women vs. men).
• To evaluate the relationship between family income level and psychosocial dimensions of substance use.
• To analyse the influence of the mothers’ educational level on substance use dimensions.
• To analyse the influence of the father’s educational level on substance use dimensions, particularly intrinsic factors.

Based on this theoretical framework, the following hypotheses are proposed:

Hypothesis 1. 

Men will show significantly higher scores than women in overall consumption, frequency of use, leisure, social function, and perceptions and prejudices, with no differences in intrinsic factors.

Gender differences in substance use are well documented, with men generally reporting higher consumption levels, frequency of use, and greater involvement in social and leisure-related drinking contexts. These differences have been associated with gender norms and risk-related behaviours, whereas intrinsic motivations tend to show less consistent variation by sex.

Hypothesis 2. 

Students from low-income families will show higher scores in psychoactive substances, while leisure-related consumption will increase with income level; the rest of the variables will not differ significantly.

Socioeconomic status is associated with substance use patterns, with lower-income groups often showing higher vulnerability to substance use, while higher income is linked to increased participation in leisure contexts where consumption may occur.

Hypothesis 3. 

Mother’s educational level will not be significantly associated with most consumption dimensions, except for social function.

Maternal educational level has shown inconsistent associations with substance use, with some evidence suggesting limited or indirect effects, particularly related to social dimensions rather than overall consumption.

Hypothesis 4. 

Father’s educational level will be significantly related only to intrinsic factors of consumption, with no differences in overall consumption or frequency.

Paternal educational level has been associated more with cognitive and motivational aspects of behaviour, showing selective relationships with internal or intrinsic factors rather than general consumption patterns.

3. Materials and Methods

3.1. Participants

The target population comprised undergraduate students at the University of Córdoba (Spain) (

Table 1. Sample characteristics by sociodemographic variables.

Degree Programme	Bachelor’s Degree in Primary Education	362	39.2%
	Bachelor’s Degree in Early Childhood Education	191	20.7%
	Bachelor’s Degree in Psychology	38	4.1%
	Bachelor’s Degree in Social Education	99	10.7%
	Double Bachelor’s Degree in English Studies/Primary Education	31	3.4%
	Bachelor’s Degree in Hispanic Philology	21	2.3%
	Double Bachelor’s Degree in Translation and Interpreting/English Studies	5	0.5%
	Double Bachelor’s Degree in Translation and Interpreting (French)/Hispanic Philology	7	0.8%
	Bachelor’s Degree in Translation and Interpreting	9	1.0%
	Bachelor’s Degree in Film and Culture	6	0.6%
	Double Bachelor’s Degree in History/History of Art	1	0.1%
	Double Bachelor’s Degree in Tourism/Translation and Interpreting (English)	3	0.3%
	Bachelor’s Degree in Cultural Management	1	0.1%
	Bachelor’s Degree in Business Administration and Management	1	0.1%
	Bachelor’s Degree in Environmental Sciences	12	1.3%
	Bachelor’s Degree in Biology	40	4.3%
	Bachelor’s Degree in Agri-Food and Rural Environment Engineering	27	2.9%
	Double Bachelor’s Degree in Agri-Food Engineering/Forestry Engineering	4	0.4%
	Bachelor’s Degree in Oenology	1	0.1%
	Double Bachelor’s Degree in Oenology/Agri-Food and Rural Environment Engineering	2	0.2%
	Bachelor’s Degree in Nursing	34	3.7%
	Bachelor’s Degree in Medicine	28	3.0%
	Bachelor’s Degree in Chemistry	1	0.1%
Sex	Woman	717	77.6%
	Man	201	21.8%
Household income level	Low	125	13.5%
	Middle	769	83.2%
	High	30	3.2%
Maternal education level	No formal qualifications	68	7.4%
	GCSEs or equivalent	288	31.2%
	Level 2 Vocational Qualification (NVQ/BTEC)	93	10.1%
	A-Levels or equivalent	138	14.9%
	Level 4 Vocational Qualification (HND/HNC)	68	7.4%
	Degree level or above	252	27.3%
Paternal education level	No formal qualifications	103	11.%
	GCSEs or equivalent	318	34.4%
	Level 2 Vocational Qualification (NVQ/BTEC)	92	10.0%
	A-Levels or equivalent	109	11.8%
	Level 4 Vocational Qualification (HND/HNC)	67	7.3%
	Degree level or above	196	21.2%

). The sample consisted of a total of 924 students, of whom 77.6% were women, 21.8% were men, and 0.6% preferred not to disclose their gender. The distribution of the sample across faculties and centres was as follows: 73.4% from the Faculty of Education Sciences and Psychology, 10.5% from the Faculty of Philosophy and Arts, 5.8% from the Faculty of Sciences, 3.6% from the Higher Technical School of Engineering, and 6.7% from the Faculty of Medicine and Nursing.

Regarding fields of knowledge, 65.59% belonged to Social and Legal Sciences, 3.35% to Engineering and Architecture, 12.88% to Arts and Humanities, 8.98% to Health Sciences, and 9.09% to Sciences. In terms of academic year, 31.4% were first-year students, 25.4% were in their second year, 28.6% were in their third year, and 14.6% were in their fourth year.

The characteristics of the sample were further detailed to provide a clearer understanding of its composition. The predominance of female students and the overrepresentation of participants from the field of Social and Legal Sciences can be partly explained by accessibility and recruitment context, as these disciplines typically show higher female enrollment rates and were more readily accessible during the data collection process. This distribution is consistent with broader trends in higher education, where women are more highly represented in these academic areas. These features should be considered when interpreting the results, as they may influence the observed patterns of substance use and associated psychosocial variables, and may limit the generalizability of the findings to other fields with different gender distributions.

3.2. Instruments

A questionnaire specifically designed for this research was employed, based on an exhaustive literature review and previous studies related to the topic (Zárate et al., 2006; Ortega-Pérez et al., 2011; Cooke et al., 2014; Rodríguez-García et al., 2022). The questionnaire is a structured scale comprising several dimensions. It includes sociodemographic variables (age, sex, academic year, and degree course), as well as variables related to the family context, such as the educational attainment and employment status of the parents, and household income. Furthermore, it incorporates psychosocial variables associated with leisure, social function, perception and prejudice, behaviour, and intrinsic motivational factors linked to the consumption of psychoactive substances. The scale consists of 49 items distributed across eight factors, using a four-point Likert-type response format to measure the participants’ level of agreement with each statement.

To strengthen the validity of the instrument beyond internal consistency, a multi-step validation strategy was implemented. First, content validity was ensured through an extensive literature review and alignment with previously validated instruments addressing psychosocial determinants of substance use.

Second, structural validity was examined through exploratory factor analysis (EFA) using methods appropriate for ordinal data, which supported the eight-factor structure of the scale and confirmed the dimensional coherence of the instrument. Where applicable, additional model fit indicators were considered to assess the adequacy of the factor solution.

Third, construct validity was reinforced by explicitly linking each dimension to established psychosocial frameworks. In this regard, intrinsic factors refer to internal motivational processes such as emotional regulation and sensation-seeking; social function captures the interpersonal and relational role of substance use within peer contexts; and perceptions and prejudices reflect normative beliefs and attitudes toward consumption.

Finally, the internal consistency of the scale was assessed using Cronbach’s alpha coefficient for each dimension. The results showed acceptable levels of reliability in most factors, including psychoactive substances (α = 0.768), frequency of consumption (α = 0.761), and leisure (α = 0.751). Moderate values were observed for behaviour (α = 0.675), social function (α = 0.647), perceptions and prejudices (α = 0.636), consumption (α = 0.625), and intrinsic factor (α = 0.618). Overall, these findings indicate an acceptable level of internal consistency, particularly considering the multidimensional nature of the instrument.

The exploratory factor analysis conducted using the Principal Component Analysis method allowed for the identification of the underlying structure of the variables. Prior to factor extraction, the suitability of the data for this type of analysis was assessed using the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity. The results showed a KMO index of 0.844, which is considered high and adequate, indicating good correlations among variables and an appropriate sample for factor analysis. Likewise, Bartlett’s test was statistically significant (χ² = 11,187.174; df = 820; p < 0.001), confirming that the correlation matrix is not an identity matrix and that sufficient relationships exist among the analysed variables.

Regarding the total variance explained, eight components with eigenvalues greater than 1 were identified, following Kaiser’s criterion. These eight factors jointly explained 50.339% of the total cumulative variance, which is considered acceptable in studies within the social and behavioural sciences. The first component accounted for 14.357% of the variance, while the second and third components explained 9.339% and 7.400%, respectively. Subsequently, a Varimax rotation with Kaiser normalisation was applied, achieving convergence after nine iterations, which facilitated a clearer interpretation of the extracted factors.

3.3. Procedure and Data Analysis

Data collection was carried out by first contacting each university department to provide information and obtain consent to participate in the research. Once consent had been obtained, participants were asked to complete the questionnaires, which took approximately 30 min. All participants, who were of legal age, provided written informed consent and were informed that participation was voluntary and that the data supplied would be treated anonymously and confidentially. Any questions that arose were addressed, and data collection proceeded without incident. The process followed the ethical principles set out in the Declaration of Helsinki and received approval under registration number CEIH-25-77 from the Human Research Ethics Committee (CEIH) of the University of Córdoba.

Data were analysed using IBM SPSS Statistics (v29). First, a descriptive exploration of sociodemographic, family, and psychosocial variables was performed, including means, standard deviations, and frequencies, in order to characterise the sample (Ferrando & Lorenzo-Seva, 2017; Flora & Curran, 2004).

Second, to ensure the robustness of the findings, non-parametric contrast analyses were conducted after verifying, through the Kolmogorov–Smirnov test, that the score distribution did not meet normality criteria (p < 0.05) for the majority of the scale dimensions. Consequently, the Mann–Whitney U test was employed to compare differences by sex, while the Kruskal–Wallis test was used to assess variations based on household income and parental educational attainment. Finally, the reliability of the scale was verified using Cronbach’s Alpha, yielding satisfactory internal consistency coefficients for all eight dimensions analysed.

Thirdly, to identify associations between psychosocial dimensions, a Pearson correlation matrix (r) was calculated, evaluating the strength and direction of linear relationships.

4. Results

Regarding sex, statistically significant differences were observed between women and men in most variables analysed (

Table 2. Differences in psychosocial dimensions of substance use by sex.

Variable	Sex				p	Cohen’s d
	Woman		Man
	M	SD	M	SD
Psychoactive substances	1.05	0.16	1.06	0.16	0.025 *	−0.06
Frequency of consumption	1.89	0.46	2.00	0.50	0.031 *	−0.23
Leisure	2.36	0.48	2.26	0.50	0.030 *	0.20
Perceptions and prejudices	2.39	0.29	2.43	0.32	0.025 *	0.13
Social function	3.29	0.50	3.24	0.56	0.048 *	0.10
Intrinsic factor	1.92	0.42	1.91	0.43	0.831	0.02
Behaviour	2.60	0.48	2.53	0.51	0.054	0.14
Consumption	2.31	0.42	2.44	0.51	0.000 *	−0.77

Note: * p < 0.05 = significant.

). Women showed lower mean scores in psychoactive substances (M = 1.05 vs. M = 1.06), frequency of consumption (M = 1.89 vs. M = 2.00), leisure (M = 2.36 vs. M = 2.26), perceptions and prejudices (M = 2.39 vs. M = 2.43), social function (M = 3.29 vs. M = 3.24) and overall consumption (M = 2.31 vs. M = 2.44), with statistical significance in all these dimensions (p < 0.05). The intrinsic factor showed no significant differences (p = 0.831), while behaviour exhibited a trend towards significance (p = 0.054). These results indicate that men consistently display patterns of higher consumption and greater psychosocial involvement across the evaluated dimensions.

Effect sizes for sex differences were generally small across most psychosocial dimensions. Specifically, frequency of consumption showed a small effect (d ≈ −0.23), indicating that men reported slightly higher consumption frequency than women. Similarly, leisure (d ≈ 0.20), perceptions and prejudices (d ≈ −0.13), and social function (d ≈ 0.10) showed small effects. The largest difference was observed in consumption (d ≈ −0.77), indicating a moderate to large effect, with men reporting higher levels than women. The remaining variables showed negligible or trivial effects (|d| < 0.10) (

Table 3. Differences in psychosocial dimensions of substance use by Household income level.

Variable	Household Income Level						p	Cohen’s d
	Low		Middle		High
	M	SD	M	SD	M	SD
Psychoactive substances	1.11	0.34	1.04	0.13	1.02	0.05	0.001 *	0.06
Frequency of consumption	1.94	0.49	1.91	0.47	1.88	0.40	0.850	0.00
Leisure	2.26	0.50	2.34	0.48	2.44	0.52	0.038 *	0.02
Perceptions and prejudices	2.37	0.27	2.41	0.31	2.38	0.32	0.368	0.01
Social function	3.29	0.4	3.28	0.50	3.31	0.64	0.722	0.00
Intrinsic factor	2.00	0.52	1.91	0.40	1.78	0.35	0.102	0.03
Behaviour	2.50	0.47	2.60	0.49	2.63	0.60	0.140	0.02
Consumption	2.36	0.46	2.33	0.44	2.38	0.39	0.623	0.00

Note: * p < 0.05 = significant.

).

As regards family income level, significant differences were found in psychoactive substance scores (p = 0.001) and leisure (p = 0.038), whereas the remaining variables showed no relevant variations between groups (low, medium, high). Participants from low-income backgrounds presented higher scores in psychoactive substances, whilst leisure scores increased progressively with income level. Frequency of consumption, perceptions and prejudices, social function, intrinsic factor, behaviour, and overall consumption did not differ significantly between income levels (Table 3).

Overall, although some statistically significant differences were identified, the magnitude of these differences was generally small, indicating a limited practical impact of household income on the psychosocial dimensions of substance use (Table 3).

With respect to mother’s educational level in

Table 4. Comparison of psychosocial dimensions of substance use across parental educational levels using the Kruskal–Wallis test.

		No Formal Qualifications		GCSEs or Equivalent		Vocational Qualification (NVQ/BTEC)		A-Levels or Equivalent		Vocational Qualification (HND/HNC)		Degree Level or Above		KW
Variable	P	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	p
Psychoactive substances	F	1.06	0.21	1.04	0.11	1.04	0.17	1.05	0.17	1.04	0.12	1.05	0.13	0.082
	M	1.10	0.41	1.04	0.14	1.05	0.13	1.04	0.13	1.07	0.17	1.05	0.13	0.795
Frequency of consumption	F	1.88	0.50	1.94	0.45	1.93	0.48	1.89	0.50	1.85	0.51	1.93	0.47	0.459
	M	1.98	0.48	1.91	0.48	1.86	0.46	1.89	0.49	2.00	0.52	1.92	0.46	0.532
Leisure	F	2.29	0.49	2.37	0.49	2.39	0.49	2.37	0.48	2.23	0.48	2.32	0.47	0.115
	M	2.33	0.45	2.39	0.50	2.32	0.54	2.32	0.48	2.37	0.43	2.28	0.48	0.087
Perceptions and prejudices	F	2.43	0.31	2.40	0.30	2.44	0.34	2.39	0.31	2.44	0.30	2.39	0.29	0.836
	M	2.41	0.32	2.39	0.30	2.40	0.31	2.42	0.31	2.39	0.32	2.41	0.29	0.936
Social function	F	3.34	0.50	3.25	0.52	3.34	0.54	3.25	0.45	3.26	0.61	3.30	0.49	0.531
	M	3.43	0.51	3.25	0.54	3.27	0.52	3.23	0.49	3.40	0.47	3.31	0.48	0.002 *
Intrinsic factor	F	1.88	0.46	1.91	0.42	1.95	0.41	1.83	0.39	1.84	0.45	2.01	0.42	0.003 *
	M	1.93	0.46	1.88	0.40	1.91	0.43	1.88	0.44	1.97	0.48	1.97	0.41	0.084
Behaviour	F	2.59	0.47	2.57	0.45	2.57	0.53	2.63	0.51	2.60	0.60	2.61	0.50	0.498
	M	2.65	0.45	2.58	0.44	2.59	0.54	2.57	0.53	2.54	0.51	2.61	0.50	0.643
Consumption	F	2.43	0.45	2.35	0.44	2.24	0.45	2.34	0.42	2.28	0.42	2.31	0.44	0.051
	M	2.33	0.45	2.37	0.45	2.29	0.46	2.34	0.43	2.38	0.40	2.30	0.45	0.459

Note: * p < 0.05 = significant.

, the results showed stability across most variables, with no significant differences in psychoactive substances (p = 0.795), frequency of consumption (p = 0.532), leisure (p = 0.087), perceptions and prejudices (p = 0.936), intrinsic factor (p = 0.084), behaviour (p = 0.643) or overall consumption (p = 0.459). Only social function exhibited significant differences (p = 0.002), suggesting a specific impact of maternal education on this dimension.

Finally, regarding the father’s educational level, a few significant differences were observed (Table 4). The variables psychoactive substances (p = 0.082), frequency of consumption (p = 0.459), leisure (p = 0.115), perceptions and prejudices (p = 0.836), social function (p = 0.531), behaviour (p = 0.498), and overall consumption (p = 0.051) did not reach statistical significance. Only the intrinsic factor showed significant differences (p = 0.003), indicating a specific influence of paternal education on this dimension.

Table 5. Pearson correlation matrix of psychosocial factors and consumption.

		Frequency of Consumption	Leisure	Perceptions and Prejudices	Social Function	Intrinsic Factor	Behaviour	Consumption
Psychoactive substances	r	0.393	0.134	0.002	0.019	0.214	−0.074	−0.004
	p	0.000	0.000	0.948	0.567	0.000	0.024	0.898
Frequency of consumption	r	1.000	0.409	0.030	0.051	0.288	−0.065	−0.057
	p	.	0.000	0.368	0.120	0.000	0.049	0.086
Leisure	r	0.409	1.000	−0.018	−0.076	0.122	−0.061	−0.087
	p	0.000	.	0.593	0.022	0.000	0.064	0.009
Perceptions and prejudices	r	0.030	−0.018	1.000	−0.057	0.107	0.107	0.107
	p	0.368	0.593	.	0.083	0.001	0.001	0.001
Social function	r	0.051	−0.076	−0.057	1.000	0.146	0.256	−0.087
	p	0.120	0.022	0.083	.	0.000	0.000	0.008
Intrinsic factor	r	0.288	0.122	0.107	0.146	1.000	0.090	−0.059
	p	0.000	0.000	0.001	0.000	.	0.006	0.071
Behaviour	r	−0.065	−0.061	0.107	0.256	0.090	1.000	−0.095
	p	0.049	0.064	0.001	0.000	0.006	.	0.004

presents a Pearson correlation matrix (r), a fundamental statistical analysis used to identify the strength and direction of the linear relationships between the psychosocial and consumption variables of the study.

The variables leisure, social function, intrinsic factor, and behaviour exhibit a coherent pattern of correlations. Of particular note is the positive relationship between social function and behaviour (r = 0.256; p < 0.001), as well as between social function and Yesintrinsic factor (r = 0.146; p < 0.001), suggesting that greater social functionality is associated with stronger internal motivations and specific behavioural patterns.

Furthermore, perceptions and prejudices show significant positive correlations with both the intrinsic factor and behaviour (r = 0.107; p = 0.001 in both cases); although the magnitude is low, it indicates the existence of definite, albeit weak, relationships.

However, it is important to emphasise that, although several of these associations reach statistical significance, most correlation coefficients are small in magnitude (r < 0.30), indicating weak relationships in practical terms. Therefore, these findings should be interpreted with caution, as they reflect modest links between psychosocial variables rather than strong or determining associations. This pattern suggests that substance use is influenced by a complex interplay of factors, in which each dimension contributes in a limited but complementary way.

Finally, overall consumption maintains significant negative correlations with leisure, social function, and behaviour, which may reflect that higher levels of consumption are associated with poorer adjustment across these dimensions.

5. Discussion

Before interpreting the findings, several boundary conditions must be considered. First, the cross-sectional design limits causal inference. Second, the predominance of female participants and Education students may reflect specific cultural and disciplinary norms. Therefore, findings should be interpreted as context-specific. Additionally, the use of self-reported data may introduce social desirability bias.

The findings relating to sex confirm systematic differences between women and men, with men exhibiting a higher frequency of consumption, greater involvement in substance-related leisure, and higher overall consumption scores (Appel et al., 2023; Merdassa, 2025). These results are consistent with the literature identifying male sex as a risk factor for more intense consumption patterns among university students, possibly due to social norms of masculinity, greater exposure to risk contexts, and differences in emotional coping styles. The absence of differences in the intrinsic factor suggests that underlying internal motivations for consumption are similar across genders. In contrast, differences emerge in frequency, social context, and associated attitudes (Panadero et al., 2022; Colomer-Pérez et al., 2019).

Regarding family income level, the greater presence of substance-related problems in the low-income group replicates patterns documented in epidemiological studies, where socioeconomic disadvantage is associated with heightened vulnerability to risk behaviours. The increase in leisure scores at higher income levels may reflect greater integration of consumption into structured recreational activities, in contrast to more uncontrolled patterns in disadvantaged contexts (Ibáñez-Tejedor & Cauli, 2025; Morgan et al., 2008). The stability across most psychosocial variables indicates that the impact of socioeconomic status primarily concerns problem intensity and the role of leisure, rather than deep-seated attitudes or motivations (Neddermann-Carrillo et al., 2024; Sampaio et al., 2025).

Mother’s educational level exerts a limited but specific effect on the social function of consumption, which may be mediated by family socialisation models and intergenerational transmission of norms regarding the role of consumption in social contexts (Rada & Lungu, 2023; Holstege et al., 2022). The absence of differences in overall consumption or frequency suggests that maternal influence operates more on the meaning attributed to consumption than on its quantity.

For its part, father’s educational level relates exclusively to intrinsic factors, pointing to a possible influence on internal motivations such as sensation-seeking or emotional regulation. This may be explained by paternal coping models or transmitted expectations that modulate underlying motives for consumption without necessarily altering observable behaviour (Ardenghi et al., 2023; Cabanach et al., 2017; Guil et al., 2021).

Overall, these results highlight that psychosocial determinants of substance use operate differentially according to the dimension evaluated and the sociodemographic variable considered (Boateng et al., 2022; Cabanach et al., 2017; Kollath-Cattano et al., 2020). From the perspective of social-emotional learning and wellbeing, the findings suggest the need for segmented interventions: targeted programmes for men addressing gender norms and coping skills; support for low-income students focused on healthy recreational alternatives; and family-oriented actions reinforcing protective educational models (Ardenghi et al., 2023; Cabanach et al., 2017; Guil et al., 2021). These strategies can mitigate the identified psychosocial vulnerability, promoting healthier university trajectories. The association between intrinsic factors and consumption indicates that strengthening emotional competencies may reduce risk behaviours. Thus, SEL should be understood not only as an educational outcome but as a preventive mechanism embedded within psychosocial processes.

In line with the validation framework of the instrument, the interpretation of the findings supports the multidimensional and theoretically grounded nature of the measured constructs. The observed pattern of associations among psychosocial dimensions—particularly between intrinsic factors, social function, leisure, and behaviour—suggests that substance use cannot be understood as a set of isolated variables, but rather as an interconnected system of motivational, social, and behavioural processes.

Specifically, the relationships identified between intrinsic motivational factors and behavioural outcomes reinforce the conceptualization of these constructs as reflecting underlying emotional regulation and decision-making processes. Similarly, the role of social function and leisure contexts highlights the importance of interpersonal dynamics and peer-related environments in shaping substance use behaviours. These findings provide empirical support for the construct coherence of the instrument, as the dimensions operate in a manner consistent with their theoretical definitions.

Overall, the results strengthen the validity of the scale by demonstrating that the measured dimensions not only show internal consistency but also meaningful and theoretically aligned interrelations, thereby supporting a multidimensional interpretation of substance use within the university context.

In line with the correlation analysis, it is important to note that most observed associations between psychosocial dimensions are small in magnitude. Although statistically significant, these relationships should be interpreted as modest, indicating limited practical strength. This reinforces the understanding of substance use as a multifactorial phenomenon, where multiple variables interact with relatively small individual contributions rather than strong, direct effects.

The findings of this study should be interpreted within the framework of a cross-sectional correlational design; therefore, the observed relationships do not imply causality, but rather indicate that the variables examined are associated with, are related to, or are linked to different patterns of substance use. In this regard, although statistically significant differences were observed in several psychosocial dimensions according to sex and household income level, the interpretation of these findings requires careful consideration of the magnitude of the effect sizes. Overall, the results show that most associations present small effect sizes, suggesting that, although differences exist, their practical relevance is limited.

Specifically, sex differences showed small effects in dimensions such as frequency of consumption, leisure, and perceptions, indicating that these variables are weakly associated with sex. However, the consumption dimension showed a larger effect size (moderate), suggesting that this variable is more strongly linked to sex compared to the other dimensions analysed.

Regarding household income level, most differences showed small or negligible effect sizes, indicating that these variables are weakly related to psychosocial dimensions of substance use. Even in cases where statistical significance was reached, the magnitude of the effect suggests that these differences should be interpreted with caution.

Overall, these findings highlight the importance of complementing statistical significance with effect size analysis, particularly in large samples, where statistically significant differences are more likely to emerge despite having limited practical significance.

6. Conclusions

This study elucidates the complex interplay between sociodemographic variables and the psychosocial dimensions underlying substance use within the university environment. Through detailed analysis, it has been established that the consumer profile is not uniform but is mediated by gender, socioeconomic status, and, in a subtler yet significant manner, by parental educational capital.

Firstly, the findings reinforce the thesis of the male sex as a predominant risk factor regarding the intensity and frequency of consumption. However, the discovery that intrinsic motivations do not differ between sexes suggests that while behavioural manifestations vary—likely modulated by constructs of masculinity and social exposure—the underlying psychological impulses are transversal. This implies that campus interventions must not only be gender-sensitive in their forms (leisure contexts) but also universal in their emotional and motivational approach.

Furthermore, household income level emerges as a critical determinant of vulnerability. The data indicate that students from lower-income households present a higher prevalence of consumption-related problems, pointing to an equity gap in mental health and coping resources. In contrast, higher income levels appear to shift consumption towards a more structured leisure sphere, suggesting that socioeconomic status does not eliminate risk but rather alters its social functionality.

One of the most novel contributions of this research lies in the differentiated influence of the parents’ educational attainment. The analysis of Table 3 reveals that the mother’s education level significantly influences the social function of consumption (p = 0.002). This relationship suggests that maternal cultural capital acts as a socialisation vector that defines the meaning and role the young adult assigns to consumption in peer interactions. Meanwhile, data from Table 4 show that the father’s educational attainment has an exclusive impact on the intrinsic factor (p = 0.003). This dichotomy is fundamental: whereas maternal influence appears to shape relational and external behaviour, the paternal figure seems to affect the individual’s deeper motivational dimensions, such as sensation-seeking or internal emotional regulation.

From an educational intervention perspective, these conclusions demand a paradigm shift. It is insufficient to implement generalist information programmes; it is imperative to design segmented strategies. For male students, programmes should focus on deconstructing risk norms associated with masculinity. For socioeconomically disadvantaged groups, the priority must be the democratisation of healthy leisure alternatives. Finally, the involvement of the family core is vital: strengthening communication models and parental norms can act as a protective factor that mitigates the academic environment’s influence on psychoactive substance use.

In summary, the university student’s trajectory is conditioned by an ecosystem of factors extending beyond the classroom. Understanding that parental educational attainment and the household’s economic context shape not only how much is consumed, but the why and the what for, allows for a more humane and effective approach towards the integral wellbeing of the student body. Importantly, these findings support a multidimensional understanding of substance use, where psychosocial determinants and SEL competencies are dynamically interconnected. Interventions should prioritise emotional and social competence development over demographic profiling.

The findings of this study indicate that several sociodemographic variables are associated with psychosocial dimensions of substance use among university students. However, most of these associations show small effect sizes, suggesting a limited practical influence. In particular, sex is related to some dimensions of substance use, showing a more consistent association with overall consumption, whereas household income level is weakly linked to the variables analysed.

These results reinforce the importance of interpreting findings beyond statistical significance by incorporating effect size indicators that allow for assessing the real magnitude of the relationships observed. They also suggest that substance use behaviour is likely influenced by multiple factors not included in the present model, opening avenues for future research.

Finally, given the cross-sectional nature of the study, the findings should be interpreted in terms of associations rather than causal relationships, and future longitudinal studies are needed to examine the directionality of these relationships. Likewise, although several statistically significant associations were identified, most of them are small in magnitude, which limits their practical relevance. This highlights the need to avoid overstating the strength of the observed relationships and to interpret substance use as a multidimensional phenomenon influenced by multiple factors with modest effects.

Limitations and Future Research Directions

Despite the significant contributions of this study, certain limitations must be acknowledged when interpreting the results. Firstly, the cross-sectional nature of the methodological design precludes the establishment of definitive causal relationships between sociodemographic variables and substance use. While clear associations have been identified, particularly regarding the influence of parental educational attainment, longitudinal studies are required to observe how these patterns evolve throughout the university trajectory. Secondly, the use of self-report questionnaires, although based on an exhaustive literature review, may be subject to social desirability bias, especially concerning sensitive topics such as psychoactive substance use.

Therefore, the findings should be interpreted in terms of associations rather than causation. Additionally, the observed relationships may be influenced by unmeasured or confounding variables not included in the analysis, and the directionality of these associations cannot be determined. Consequently, the results should be interpreted with caution, and future research using longitudinal or experimental designs is needed to better understand the nature and direction of these relationships.

This study also presents limitations related to external validity. Specifically, the composition of the sample—characterised by a predominance of female students and an overrepresentation of Social and Legal Sciences—may restrict the generalizability of the findings to more diverse academic populations. In addition, the use of a single-institution design limits the extent to which the results can be extrapolated to other university contexts, which may differ in structural, academic, or demographic characteristics. Furthermore, the cultural specificity of the sample, situated within a Southern European context, should be considered when interpreting the findings, as patterns of substance use and associated psychosocial factors may vary across cultural settings.

Regarding future research directions, the findings open new avenues of analysis. Given that the mother’s educational attainment specifically impacts the social function of consumption and the father’s educational attainment affects the intrinsic factor, it would be of great interest to delve deeper into parenting styles and attachment models. Investigating how family communication and parental supervision, mediated by education levels, act as specific protective mechanisms could provide more precise tools for prevention. Furthermore, it is recommended to expand the sample to other academic institutions to test whether regional socioeconomic and cultural differences alter the influence of family educational capital identified at the University of Cordoba.