Recently, further attention is given to the concept of quality of life (QoL), for example, in the disciplines of medical, economic, and social sciences (Broekaert et al., 2017). Quality of life refers to the goodness of various aspects of life that go beyond mere subsistence (Broekaert et al., 2017; Ben-Arieh, & Frones, 2011). Importantly two different yet complementary perspectives can be distinguished when considering individuals' QoL. The objective perspective focuses on objective measures of QoL, such as income, health, education, and other resources. The subjective perspective focuses on individuals' satisfaction with and views on their lives. QoL is understood as a multidimensional construct that includes the interaction of the environment and personal aspects (objective and subjective indicators). QoL could be improved by the person's sense of belonging, determination, material and social resources, and purposes in life (Broekaert et al., 2017). Before the last two decades, the majority of research in the field of QoL adopted an objective perspective (Casas, 2011). However, within the light of changes in society (i.e., more community-based support, increasing importance of people's perceptions and values, and the rise of person-centered and empowerment approaches), during the last 20 years, there is a shift towards the subjective perspective (Casas et al., 2007; Huebner et al., 2004; Schalock et al., 2002). Additionally, it is advised to measure objective and subjective indicators separately, as many of these indicators do not correlate (Broekaert et al., 2017; Cummins, 2010).
The concept of well-being is based on the assessment of QoL, health, and living conditions (Ben-Arieh & Frones, 2011). Its evaluation from the person's subjective perspective is one of the essential micro-aspects of QoL. There are two perspectives of well-being: (1) the hedonic perspective, which emphasizes the person's satisfaction in terms of happiness and pleasure, and (2) the eudemonic perspective, which emphasizes the development of the personal potential and the achievement of a meaningful life (Broekaert et al., 2017).
Diener, Lucas and Scollon (2006) define the subjective well-being (SWB) as a general evaluation of the person's life. Subjective well-being is defined as the people's own evaluation and satisfaction with different domains of their life (The International Well-Being Group, 2013). The theoretical construct of SWB includes cognitive and affective components (Montserrat et al., 2015).
In comparison to studies on SWB in adults, those on adolescents' SWB are less extended (González-Carrasco et al., 2017). Studying SWB in this period of life is highly relevant, as during adolescence there are diverse developmental transitions and changes (García et al., 2017). SWB'S homeostatic theory states that there is an innate neurological tendency in each person that maintains the SWB level around set points. However, the SWB homeostatic system may be influenced by life disturbances or regulatory adjustment processes, which may be particularly prevalent during the turbulent developmental phase of adolescence (Cummins et al., 2014). Throughout the lifespan, different domains can become more or less important, satisfaction with particular domains can increase or decrease, and domain-specific determinants of SWB may change. The study of Casas (2011) pointed out that the interpersonal relationships domain is important within the developmental period of adolescence. This study suggested that other domains may be equally important for adolescents, such as their perception and satisfaction about their rights, their opinions about their families, their satisfaction with their conversations with adults, and the use of technology (Casas, 2011).
As mentioned above, research on SWB in adolescence is still limited. Previous works have focused on the effects of critical changes in life and gender (González-Carrasco, 2017; Montserrat et al., 2015), the relationship between self-concept and SWB (Ramos-Díaz, Rodríguez-Fernández, & Agirre, 2017; Rodríguez-Fernández et al., 2016), as well as the effects of age (González-Carrasco, 2017; Brann et al., 2017), culture (García et al., 2017; Casas et al., 2012; Castella-Sarriera et al., 2012), and substance use (Arria et al., 2016; Bogart et al., 2007; Zullig et al., 2001) on adolescents' SWB. However, the majority of these studies adopt a global instead of a multidimensional approach, thereby hampering a more detailed evaluation. Assessing adolescents' SWB from a multidimensional approach is relevant for determining their relative satisfaction with each specific life domain as well as for exploring domain-specific determinants of SWB in this period of life. The limited number of studies that did adopt a multidimensional approach indeed indicates its relevance. For example, the study of Brann et al. (2017) compares specific SWB domains between adolescents and young adults in Sweden. One of the aims of this study was to assess how well-being (in all domains) was affected by environment, societal changes, and physical conditions. They found significant differences between the two groups for the stress balance domain (defined as feeling calm, unconcerned, relaxed), while for the other SWB domains (e.g., mood, physical condition, energy, self-esteem) no significant differences were found. Also, in both groups, girls reported significant less well-being than boys in all dimensions (Brann et al., 2017). Another example is the study of Casas et al. (2015), using the Personal Well-being Index (PWI) and the Brief Multidimensional Student's Life Satisfaction Scale, in a sample of adolescents from Spain, Chile, and Brazil. Their results show that overall, adolescents are most satisfied with the groups of people they belong to (community-connectedness) and least satisfied with their religion practice (Casas et al., 2015).
Previous work on SWB (Huebner et al., 2004; Park, 2004) in adolescents has shown that healthy development highly relates to a high level of SWB. Park (2004) considers the SWB as an essential factor in the healthy development of adolescents. The most relevant variables related to a high level of SWB in youth are good physical and mental health, positive interpersonal relationships, and high academic performance (Park, 2004). Additionally, prosocial aspects, personality, and temperament factors (e.g., extraversion, and low degree of neuroticism), a high self-perception of control, high self-esteem, and optimism are indicated as positive correlates of SWB (Cummins, 2010). On the other hand, unhealthy development in this age period is linked to a low SWB level. Adolescents who present social and psychological problems are likely to report low levels of SWB. The most common problems that previous studies reported in relation to low levels of SWB are depressive symptoms, dissocial behavior, suicidal intent and ideation, low self-esteem, and family and peer relationship problems (Alfaro et al., 2016; Suldo & Huebner 2004; Zullig et al., 2001). In addition, low levels of SWB in adolescents have been found to be associated with substance use and other risky behaviors, such as aggression to others and sexual harassment (Proctor, Lineley, & Maltby 2009). Noteworthy, most of the above studies focus only on the overall level of SWB, not providing details about variations on specific predictors of each domain of SWB.
Aims of the Study
Overall, the majority of research on SWB has been conducted among adults, while research in the transitional period of adolescence is still scarce (García et al., 2017). In addition, the majority of previous studies focus on global SWB, while a multidimensional approach may yield more detailed findings. One of the benefits of taking a multidimensional approach is that it makes possible a systematically understanding of SWB. Moreover, studies in Latin America on this topic are still incipient (Castella-Sarriera et al., 2012), specifically in Uruguay. This study was designed to fill this gap by studying domain-specific predictors of each domain of the PWI in a sample of school-going adolescents in Uruguay. We expect to find differences in the determinants for the specific SWB domains, in the variables of psychopathology, resilience, and substance use. The selection of independent variables was based on prior indications that resilience is associated with higher levels of SWB (Satici, 2016; Rodríguez-Fernández et al., 2016), while substance use and psychopathology are associated with reduced levels of SWB (Park, 2004; Cummins, 2010). The selection of socio-demo-graphic covariates was based on prior indications that SES, age, and gender (Fernández et al., 2017) are also likely to influence young people's SWB.
Method
Sample
The sample (non-probabilistic) recruitment took place between May and June 2016 in one secondary school. This school is located in a city on the outskirts of Montevideo and is governed by a catholic board. Secondary school is compulsory in Uruguay until 18 years old. The sample consisted of a total of325 adolescents; 172 girls (53.2 %) and 153 boys (46.8 %) from 12 to 18 years old (M age= 14.67; 5D=1.62). The socioeconomic status was low for 8 (3 %) adolescents, medium for 156 (48 %), and high for 159 (49 %) adolescents. In table 1, we present the sample distribution of ages and school grades.
Procedure
Before starting data collection, a verbal explanation was given to the students about the content of the study and the ethical issues regarding their participation. The head of the institution, tutors, parents, and students signed the active, informed consent. The self-report scales (all in their electronic version) were applied in the school computer labs in Spanish during the class period. On average, the administration took around an hour. An information technology teacher was present during the administration for assistance on any problems with the computers and the software used. Additionally, the first author, as well as a research assistant, were present during administration in case students had any further questions. The Ethical Committee at the University of Montevideo approved the current study.
Instruments
Socio-demographic characteristics. A socio-demographic survey was used, consisting of 41 items regarding individual, family, and school characteristics. This survey is part of the Adolescent self-report (Autoinforme de Adolescentes, ADA) (Daset et al., 2015). In the current study, we used the variable of SES, subdivided into low, medium, and high. SES is calculated by classifying households according to their consumption or expenditure capacity. We also included gender and age, with the latter as a continuous variable.
Psychopathology and Resilience. The adolescent self-report (Daset et al., 2015) was used to assess adolescents' psychological problems. This screening instrument consists of 82 items and is scored using a 5-point Likert scale (0-4). Psycho-pathological symptoms refer to emotional and behavioral problems (e.g., "I feel sad and fed-up most of the time"). The instrument also includes some items referring to resilience (e.g., "When I have a problem, I think there is a solution"), also, positive development, including strengths, life planning, coping skills, and social desirability (e.g., "I have self-confidence"). Cronbach's alpha ranges from .70 to .90 (Daset et al., 2015). This screening test is based on the empirical taxonomies and studies of Achenbach and Edelbrock (1978), Lemos et al. (2002), and López-Soler et al. (1998). It has six cluster dimensions: Factor1 Depression-anxiety, Factor 2 Dissocial behavior, substance use, and negative emotionality, Factor 3 Disrupted and dys-regulated behavior, Factor 4 Social anxiety, Factor 5 Resilience and pro-sociality, and Factor 6 Obsessive-compulsive symptoms. In the current sample, the Cronbach's alpha for each factor ranges between .84 to .94.
Alcohol use. To assess adolescents' alcohol use, a survey (JND, 2011) was conducted. The survey assessed (1) the lifetime prevalence of alcohol use (yes/no); (2) the prevalence of alcohol use during the last 12 months (yes/no); (3) the prevalence of alcohol use during the last 30 days (yes/no), and (4) the age of first alcohol use.
Marihuana use. We used the Cannabis Abuse Screening Test (CAST) (Legleye et al., 2013) in its Spanish version validated by the Junta Nacional de Drogas, Uruguay (2011), to measure marihuana use and associated problems. This scale considers possible problems in relation to the consumption of marihuana in the last 12 months and is based on the criteria for substance abuse of the DSM-IV. It consists of 6 items and has a Likert scale of 4 points (1-5), with higher scores indicating more severe marihuana use. In the current study, Cronbach's alpha was .73. We dichotomized the variables alcohol and marihuana use into use/never use (including, once in a lifetime, past year, and last month) in order to have more cases and be able to include the variables in the model.
Subjective well-being. The PWI is one of the most widely used international instruments that measures individuals' global level of life satisfaction, as well as their satisfaction with seven specific domains of life -standard of living, health, achievements, relationships with others (peers and family) safety, community-connectedness, and future security-, developed to measure subjective well-being across 50 countries and different cultures. The instrument showed good psychometric properties and cultural stability (The International Well-Being Group, 2013). We used the PWI (Cummins et al., 2003) in its Spanish version, validated for Chile (Alfaro et al., 2016). The scale consists of 7 items that evaluate the areas of the standard of living, health, achievements, relationships with others (peers and family), safety, community-connectedness, and future security. The items asses how "satisfied are you with..." this particular area of life. It has a Likert scale of 0-10, being 0 (completely dissatisfied) and 10 (completely satisfied). The index is calculated summing up all the items and transforming the scores into a 0-100 scale. We use an overall score of SWB and the mean of each area/item of SWB. Cronbach's alphas of the original test, in international studies, are between .70 and .85. The version used had internal reliability of .77 and was validated for teenagers. For the current study, Cronbach's alpha was .87.
Statistical Analysis
First, we calculated descriptive statistics for the total sample, as well as for boys and girls separately. Second, we compared the means of each SWB domain and gender differences of each domain using independent t-tests and ANOVAS for continuous variables. Third, we ran linear regression models for each specific domain of the PWI: standard of living, health, achievements, relationships with others (peers and family), safety, community-connectedness, and future security as dependent variables. The six ADA psychopathology factors, alcohol use, and marihuana use were included as independent variables. Age, gender, and SES were included as socio-demographic covariates. We performed a correlation analysis and found that dissocial behavior and social anxiety are highly correlated (0.9). They appear in the estimations (first performed linear regression model with all variables) with almost the same coefficient but of opposite sign, probably indicating that in fact they are not significant if included alone because they are canceling each other out (same coefficient of opposite sign). We decided to perform separate regression analysis including each variable (dissocial behavior and social anxiety) separate. In the annexes of this article we included the table of the correlation analysis between all empirical syndromes. The R squared is stated for each model, indicating the variation in SWB that is accounted for by the predictors in the model. For the statistical analysis, we used the statistical package SPSS Statistics 21 (IBM Corporation, 2012). A p-value of 0.05 was used in all analyses as the standard for statistical significance.
Results
Table 1 shows the age and school grade distribution, with the frequencies and percentages for each one.
Ages | Frequency | % | School cycle | Frequency | % |
---|---|---|---|---|---|
12 | 24 | 7.4 | 1 | 63 | 19.4 |
13 | 74 | 22.8 | 2 | 70 | 21.5 |
14 | 57 | 17.5 | 3 | 52 | 16 |
15 | 63 | 19.4 | 4 | 50 | 15.4 |
16 | 56 | 17.2 | 5 | 59 | 18.2 |
17 | 41 | 12.6 | 6 | 31 | 9.5 |
18 | 10 | 3.1 |
Table 2 presents the mean scores of each SWB domain with the standard deviation for the total sample, as well as for girls and boys separately. Adolescents in this sample are most satisfied with how connected they feel with their community (M = 89; SD = 1.45) and least satisfied with safety (M = 78; SD = 1.99) and future security (M= 79; SD= 1.97); the mean of the domain safety was significantly lower for girls, compared to boys. In other life domains, no significant gender differences in the mean scores were observed.
Total sample (N=325) | Girls (N=173) | Boys (N=152) Boys vs. Girls | |||||
---|---|---|---|---|---|---|---|
PWI domains | Mean | SD | Mean | SD | Mean | SD | t (df) |
Standard of living | 85 | 1.56 | 84 | 1.56 | 87 | 1.47 | -1.645 (323) |
Health | 83 | 1.7 | 81 | 1.78 | 84 | 1.59 | -1.635 (323) |
Achievements | 82 | 1.72 | 81 | 1.89 | 84 | 1.51 | -1.438 (323) |
Relationships with others | 84 | 1.64 | 83 | 1.76 | 85 | 1.49 | -1.225 (323) |
Safety | 78 | 1.99 | 74 | 2.17 | 83 | 1.65 | -4.082 (316)* |
Community-connectedness | 89 | 1.45 | 88 | 1.48 | 89 | 1.41 | -0.690 (323) |
Future security | 79 | 1.97 | 77 | 2.15 | 81 | 1.71 | -1.754 (320) |
* p< 0.05
Table 3 presents the linear regression models for the different domains of the PWI. We only present the variables that were significant for the specific PWI domain of interest. The models indicate the existence of diverse predictors for the different domains of SWB. The safety domain was most influenced by the selected independent variables (R 2 = .530; R 2 = .525), while those had a rather limited impact on the health domain (R 2=.242, R 2=.240).
PWI domains | Variables | B | SE | Beta | t | p | R 2 | F(df) |
---|---|---|---|---|---|---|---|---|
Standard of living 1/ | Depression-anxiety | -0.040 | 0.010 | -0.325 | -3.897 | 0.000* | 0.255 | 10.716 (10) |
Resilience | 0.037 | 0.019 | 0.186 | 1.926 | 0.055 | |||
Marihuana | -0.566 | 0.249 | -0.123 | -2.270 | 0.024* | |||
SES | 0.019 | 0.007 | 0.144 | 2.921 | 0.004* | |||
Standard of living 2/ | Depression-anxiety | -0.043 | 0.010 | -0.344 | -4.157 | 0.000* | 0.255 | 10.715 (10) |
Resilience | 0.038 | 0.020 | 0.190 | 1.959 | 0.051 | |||
Marihuana | -0.555 | 0.249 | -0.120 | -2.232 | 0.026* | |||
SES | 0.019 | 0.007 | 0.144 | 2.917 | 0.004* | |||
Health 1/ | Depression-anxiety | -0.030 | 0.011 | -0.224 | -2.661 | 0.008* | 0.242 | 9.989 (10) |
Resilience | 0.062 | 0.021 | 0.283 | 2.911 | 0.004* | |||
Marihuana | -0.629 | 0.273 | -0.125 | -2.303 | 0.022* | |||
Health 2/ | Depression-anxiety | -0.034 | 0.011 | -0.253 | -3.034 | 0.003* | 0.240 | 9.910 (10) |
Resilience | 0.063 | 0.021 | 0.287 | 2.928 | 0.004* | |||
Marihuana | -0.608 | 0.273 | -0.121 | -2.229 | 0.027* | |||
Achievements 1/ | Depression-anxiety | -0.051 | 0.010 | -0.372 | -5.198 | 0.000* | 0.454 | 25.989 (10) |
Resilience | 0.073 | 0.018 | 0.328 | 3.966 | 0.000* | |||
Marihuana | -0.475 | 0.235 | -0.093 | -2.018 | 0.044* | |||
Alcohol | -0.471 | 0.179 | -0.134 | -2.635 | 0.009* | |||
Age | -0.124 | 0.053 | -0.116 | -2.315 | 0.021* | |||
Achievements 2/ | Depression-anxiety | -0.057 | 0.010 | -0.417 | -5.873 | 0.000* | 0.453 | 25.869 (10) |
Resilience | 0.075 | 0.018 | 0.337 | 4.048 | 0.000* | |||
Marihuana | -0.445 | 0.235 | -0.087 | -1.893 | 0.059 | |||
Alcohol | -0.441 | 0.179 | -0.125 | -2.464 | 0.014* | |||
Age | -0.131 | 0.053 | -0.123 | -2.458 | 0.014* | |||
Relationships 1/ | Depression-anxiety | -0.034 | 0.009 | -0.266 | -3.636 | 0.000 | 0.430 | 23.593 (10) |
Social anxiety | -0.059 | 0.016 | -0.257 | -3.715 | 0.000 | |||
Resilience | 0.065 | 0.018 | 0.306 | 3.621 | 0.000 | |||
Age | -0.200 | 0.052 | -0.197 | -3.846 | 0.000 | |||
Relationships 2/ | Depression-anxiety | -0.036 | 0.009 | -0.275 | -3.792 | 0.000* | 0.428 | 23.446 (10) |
Dissocial behaviour | -0.060 | 0.017 | -0.342 | -3.594 | 0.000* | |||
Resilience | 0.061 | 0.018 | 0.289 | 3.394 | 0.001* | |||
OCD | 0.056 | 0.026 | 0.159 | 2.170 | 0.031* | |||
Age | -0.206 | 0.052 | -0.202 | -3.958 | 0.000* | |||
Safety 1/ | Depression-anxiety | -0.067 | 0.010 | -0.429 | -6.460 | 0.000* | 0.530 | 35.261 (10) |
Social anxiety | -0.044 | 0.018 | -0.157 | -2.503 | 0.013* | |||
Disrupted dysregulated | 0.043 | 0.016 | 0.207 | 2.734 | 0.007* | |||
Resilience | 0.034 | 0.020 | 0.132 | 1.724 | 0.086* | |||
Safety 2/ | Depression-anxiety | -0.073 | 0.010 | -0.463 | -7.007 | 0.000* | 0.525 | 34.564 (10) |
Dissocial behavior | -0.032 | 0.018 | -0.148 | -1.709 | 0.088 | |||
Disrupted dysregulated | 0.043 | 0.016 | 0.208 | 2.729 | 0.007* | |||
Resilience | 0.034 | 0.020 | 0.131 | 1.688 | 0.092 | |||
Age | -0.189 | 0.057 | -0.154 | -3.300 | 0.001* | |||
Community-connectedness 1/ | Depression-anxiety | -0.016 | 0.010 | -0.140 | -1.690 | 0.092 | 0.262 | 11.107 (10) |
Social anxiety | -0.058 | 0.016 | -0.285 | -3.626 | 0.000* | |||
Resilience | 0.047 | 0.018 | 0.252 | 2.620 | 0.009* | |||
Age | -0.111 | 0.052 | -0.124 | -2.121 | 0.035* | |||
Community-connectedness 2/ | Depression-anxiety | -0.017 | 0.009 | -0.144 | -1.753 | 0.081 | 0.262 | 11.118 (10) |
Dissocial behavior | -0.061 | 0.017 | -0.393 | -3.638 | 0.000* | |||
Resilience | 0.043 | 0.018 | 0.231 | 2.391 | 0.017* | |||
OCD | 0.084 | 0.026 | 0.269 | 3.231 | 0.001* | |||
Age | -0.115 | 0.052 | -0.129 | -2.216 | 0.027* | |||
Future security 1/ | Depression-anxiety | -0.056 | 0.012 | -0.362 | -4.709 | 0.000* | 0.369 | 18.281 (10) |
Resilience | 0.051 | 0.022 | 0.204 | 2.292 | 0.023* | |||
Age | -0.243 | 0.065 | -0.200 | -3.717 | 0.000* | |||
Future security 2/ | Depression-anxiety | -0.061 | 0.012 | -0.394 | -5.157 | 0.000* | 0.366 | 18.063 (10) |
Resilience | 0.052 | 0.023 | 0.206 | 2.294 | 0.022* | |||
Age | -0.251 | 0.065 | -0.206 | -3.830 | 0.000* |
Notes: * p< 0.05, 1/ Not including Dissocial behavior. 2/ Not including Social anxiety.
OCD=obsessive compulsive disorder
Depression-anxiety was the most robust predictor for SWB in all domains in the negative direction. While resilience also appears closely related to a high SWB in all of them (with a positive association). The other variable which appears in most of the SWB domains with a negative association is age (apart from the safety domains (model 1), health (models 1+2), and standard of living (models 1+2).
Regarding substance use, marihuana appears with a negative association to SWB in the standard of living, health, and achievement domains. Alcohol use is negatively associated with SWB in the achievements in the life domain, yet it does not seem to have an impact on the other SWB domains.
Discussion
In this study, we aimed to assess the domain-specific predictors in a sample of 325 school-going adolescents in Uruguay. Our results suggest that adolescents are most satisfied with community-connectedness and least satisfied with safety. These differences support the importance of a multidimensional approach when studying individuals' SWB. Adolescents' relatively high level of satisfaction with community-connectedness may be explained by the characteristics of the local environment (it's a small city near the capital where people tend to know each other), characteristics of the school (as a catholic school they have many extracurricular activities, as well as close supervision and involvement with students). Their relatively low level of satisfaction with safety may be due to the general context of the country: recently, there is a major concern among the population of Uruguay regarding rising concerns about crime and insecurity (Aboal et al., 2015). Further studies are needed to explore possible explanations and to confirm these findings.
The regression models of the current study indicate that different psychopathology factors are significantly associated with different SWB domains. Depression-anxiety and resilience were most frequently identified as significant predictors. First of all, depression-anxiety seems to be the most significant predictor for low SWB in all the domains. In line with prior work, our results suggest that it is important to consider the prevention and treatment of depression and anxiety in adolescents, in order to safeguard or improve their well-being (Nitkowskiet al., 2017). In this respect, early screening of psychopathology allows working on prevention and early intervention by detecting early emerging symptoms and doing early referrals to professionals. In addition, it could be useful to teach adolescents strategies on how to deal with anxiety and to help them to develop social skills in order to avoid social isolation (Ramos-Díaz et al., 2017; Rodríguez-Fernández et al., 2016).
Second, resilience displayed a positive association to SWB in all domains. In line with previous research, our study points to the importance of resilience, confirming that this construct is tightly related to one's SWB (Satici, 2016; Rodríguez-Fernández et al., 2016). Resilience is understood as a dynamic adaptation process in an adversity context (Luthar, Ciccehtti & Becker, 2000). In relation to clinical implications, this means that enhancing resilience (for example, teaching problem solving, coping, and social skills to adolescents) is also a way to foster adolescents' SWB.
The factor 'disrupted and dysregulated behavior' appears with a positive association in the SWB domain of safety. It is likely that adolescents who display high levels of dissocial behavior, impulsivity, and difficulties in regulating emotions lack empathy with others and insight about their behavior and its consequences. This is not uncommon within the developmental phase of adolescence, in which youngsters become more self-centered, as illustrated by the emergence of aspects of omnipotence, invulnerability, and personal uniqueness (Aalsma, Lapsley, & Flannery, 2006; Berk, 2006; Seagrave, & Grisso, 2002).
Dissocial behavior appears with a negative association with SWB in the domains of relationships with others, community-connectedness, and safety. This factor involves refusal and confrontation with norms and rules in a persistent way and restricted prosocial emotions (Molinuevo, 2014). Therefore, adolescents who display high dissocial behavior might have difficulties in their social relationships, mainly with authority figures such as parents and teachers. Social anxiety had a negative association with SWB in the domains of safety and community-connectedness. Adolescents that present these psychopathological symptoms usually ruminate and worry about their future, safety, and the negative judgement of other people (Boileau, 2011).
In relation to substance use, marihuana had a significant negative association with SWB in the standard of living, health, and achievement domains, and seems to have no significance in the other SWB domains. This is in line with the longitudinal study of Arria et al. (2016), indicating that even occasional, non-problematic use of marihuana can have an impact on the mental health and global SWB of young people. Alcohol use was negatively associated to SWB only in the achievement domain. The longitudinal study of Bogart et al. (2007) also found the detrimental effect of substance use (cigarettes, alcohol, marihuana, and hard drugs) on adolescents' global SWB over time. The above-mentioned results provide more detail regarding specific SWB domains. Regarding the use of marihuana and alcohol, our findings suggest that preventive work is needed as the use of these substances (especially marihuana) affects adolescents' SWB. A way of preventing early use of alcohol and marihuana in adolescents is to design specific school and community-based interventions that increase healthy peer relationships, assertive-ness, and greater social participation (Bogart et al., 2007). Our results offer detailed information regarding specific domains of SWB in adolescents that could be particularly useful for teachers, youth workers, and educational psychologists, helping to tailor programs to the developmental requirements of adolescents.
In this study, age has been observed to have a significant negative association with SWB in the domains of achievements in life, safety (model 2), relationships with others, community-connectedness, and future security. This means that among older adolescents, the level of SWB decreased. This goes in line with previous work that indicates the decrease of SWB satisfaction in older adolescents (Brann et al., 2017; Castella-Sarriera et al., 2012). However, in our findings, age does not seem to have an impact on the adolescents' satisfaction with their subjective evaluation of their safety (model 1), their standard of living, and their health. A possible explanation, following Cummins' homeostatic theory, is that the other SWB domains (achievements, safety, community-connectedness, relationships with others, and future safety) could imply more life uncertainty and be more challenging areas as adolescents move into adulthood. Nevertheless, more studies with a multidimensional approach are needed to confirm this finding. In relation to clinical implications, our results indicate the importance of having age-appropriate interventions (Montserrat et al., 2015), taking into account the specific life domains that play a role in this period, and that SWB tends to decrease in older adolescents. For future studies, we recommend doing additional moderator analyses regarding age and gender for each domain of SWB.
Research Limitations and Recommendations
The results of the current study need to be considered in light of some limitations. First, the cross-sectional nature of our study makes it not possible to draw any causal conclusions. Future research should adopt a longitudinal design to address the direction of the associations between SWB specific domains and psychopathology, substance use and sociodemographic characteristics, as well as the development of these different concepts throughout the developmental period of adolescence (González-Carrasco et al., 2017; Montserrat et al., 2015). Second, the results can only be considered for this sample, since data was collected from one school that has some specific characteristics (e.g., catholic, located in the metropolitan are outside the country capital). Moreover, some of the instruments used were not validated for the Uruguayan population. Future studies would benefit from obtaining a national profile of SWB and to do comparative studies with other countries of the region, in order to gain more insight into possible cultural differences (Casas et al., 2015; Casas, 2011). Third, we only focused on school-going adolescents. It would be relevant for further studies to include adolescents at risk, no longer connected to the school system. Measuring SWB in different social realities can provide useful indicators for politicians and social programs aiming to improve the living conditions of all adolescents, including the most vulnerable ones (Casas, 2011). Additionally, future studies on the topic should include other positive determinants for SWB as independent variables, such as a reliable measure of academic performance.