Participants

We evaluated a convenience sample that consisted of 503 students who regularly attended the educational system from 9th to 12th grade of a public school, in the northern region of Portugal, whose ages ranged from 13 to 19 (M = 16; SD = 1.2), most of them female (63%).

Materials

Health Perception

Assessed with the item General Health Perception included in the SF-36 survey (first item of the questionnaire), developed by ^{Ware, Snow, Kosinski e Gandek (1993}) and validated for Portugal by ^{Ferreira e Santana, (2003}), which inquires: “In general, how would you say your health is?”, with a five position Likert scale answer option that ranges from “1= excellent” to “5 = poor”. Low scores indicate better health perception. There are various reasons for considering the use of a single item for general health perception: the main one, in line with the research carried out by ^{Benyamini (2008}, ²⁰¹¹a, ²⁰¹¹b). Its usefulness is apparent insofar as the subjective perceptions of individuals about their own health capture physical, psychological and social factors, in accordance with a holistic view of health, besides providing comprehensive outlines of the set of factors that people consider as being part of their own health concept. Global health assessment is a measure that is broadly accepted in epidemiology (^{Benyamini, 2008}; ²⁰¹¹a; ²⁰¹¹b; ^{Idler & Benyamini, 1997}). Since the end of the last century, European and world statistics use this indicator (^{Eurostat, 2020}). It is also an indicator that has been integrated to health assessment questionnaires for specific diseases such as, for cancer - QLQ-C30 (^{Aaronson et al., 1993}), for Multiple Sclerosis - MSQOL 54 (^{Vickrey et al., 1995}), for epilepsy - ESI 55 (^{Vickrey et al., 1992}).

Developmental Assets

Profile of Student Life: Attitudes and Behaviors ® (A&B) - Questionnaire developed by the Search Institute (^{Benson et al., 1998}; ^{Leffert et al., 1998}; ^{Search Institute, 2019}) and validated for European-Portuguese by ^{Soares, Pais-Ribeiro and Silva (2018}a; ²⁰¹⁸c). It is addressed to adolescents whose ages range from 12 to 18 years in the school context, or, more specifically, to adolescents who are currently attending to the 2nd and 3rd cycle of basic education, to secondary education in the Portuguese education system (in the original, middle school-high school, corresponding to the educational system in the United States of America) (^{Leffert et al., 1998}).

The assessment of the experience of the 40 individual assets comprehends 92 items, mainly consisting of statements in relation to which the answer options correspond to a five position Likert scale that range from, for instance, 1 = strongly agree” to 5 = strongly disagree, 1 = not important to 5 = extremely important, 1 = not at all like me to 5 = very much like me.

There are two approaches to describe the experience of Assets: 1) Data can be analyzed assuming that it is a Likert scale, thus the value obtained for each of the assets is interpreted as follows: the higher the score, the greater the experience of the Asset; 2) On the other hand, for communicational purposes, the score for each Asset can be used as a dichotomous variable, i. e, the adolescent has or doesn’t have the Asset. This score corresponding to each 40 Assets yields an Index for which the punctuation ranges from 0-40 Assets, which describes the average number of Assets that each adolescent has. This approach also affords the analysis of the concept of Asset Levels, which reflects the broader continuum of healthy development and defines four Asset Levels based on the Asset Index experienced by the adolescent. It represents a developmental continuum ranging from “At-risk” to “Optimal”, as well as the Asset Building axiom “the more assets, the better”: Level 1 - At-risk Development: 0 to 10 Assets; Level 2 - Vulnerable Development: 11 to 20 Assets; Level 3 - Adequate Development: 21 to 30 Assets; and Level 4 - Optimal Development: 31 to 40 Assets (^{Benson et al., 1998}; ^{Leffert et al., 1998}).

The Portuguese version of the A&B questionnaire has similar validity characteristics to the original version (^{Soares et al., 2018}c). Twenty-one of the 40 Assets are measured by at least 3 items. Most of them show internal consistency values that are considered acceptable, ranging from .60 to .80 (^{Cronbach, 1951}). Thirteen Assets are measured using individual items, therefore such an evaluation cannot be applied. The evaluation of internal consistency of Categories: Support α = .81; Empowerment α = .76; Limits an Expectations α = .65; Constructive use of time α = .39 (multidimensional category); Commitment to Learning α = .67; Positive Values α = .78; Social Competences is α = .72; Positive Identity α = .83. Dimensional Analysis of: External Assets α = .84; Internal Assets α = .84.

Procedure

To carry out this research, consent was obtained from the Directorate-General for Education. This study was conducted in accordance with the regulations of professional ethics, as per the declarations of the ^{Order of Portuguese Psychologists (2016)}. The National Commission for the Protection of Data confirmed the anonymity of the participants during data collection by giving a favorable review. For data collection, authorizations were requested to the General Council and the Director of School Grouping, and Professors. The parents/legal guardians of all students who participated in the study signed the free and informed consent form for data collection, use and publication of results. Data collection was carried out in classrooms with standardized instructions. The students who obtained parental/legal guardian consent for participation were invited to respond to the questionnaires and to put them in an envelope. Anonymity and confidentiality were assured, as well as the use of the data for research purposes exclusively.

Health Perception

Table 2 shows mean scores (and standard deviations between brackets) of SRH by gender, age, years of schooling, and total sample. Data indicate that adolescents’ mean score for Self-rated Health is 2.25 (SD = 1.02). Taking gender into consideration, we observed statistically significant differences in mean scores, which indicate that boys (M= 2.02; SD = 0.92) report a better Health Perception in comparison with girls (M = 2.39; SD = 1.04), t (495) = 3.96, p < .0001. It was confirmed that there tends to be no statistically significant correlation between age and Health Perception, r (501) = .07, p = .13. Results also suggest there is no statistically significant difference in mean scores of SRH among the years of schooling, F (3.496) = 1.57, p = .19.

Results from this study indicate that the highest percentage of adolescents (88% of them) report that their Health Perception ranks between Good, Very Good and Excellent: 28.7% (n = 144) report their Health Perception as being Excellent; 29.8% (n = 150) report Very Good; 29.4% (n = 148) report Good. The rest, 10.9% (n = 55), report their Health Perception as begin Fair and 0.8% (n = 4) report Poor.

Table 2: Mean Scores (and Standard Deviation between Brackets) of Self-Rated Health according to Gender, Age, Schooling and Total Sample 

Effects of Cumulative Experience of Developmental Assets in Health Perception

Table 3 shows mean scores (and standard deviations between brackets) of SRH of the total sample of adolescents and by the Asset Level that the adolescents report - i. e, the four Levels of Assets defined as 0-10, 11-20, 21-30 and 31-40 Assets. In order to analyze the cumulative impact of the experience of developmental assets in SRH, an analysis of variance was conducted (one-way ANOVA, Post-Hoc Bonferroni). Results point to the existence of statistically significant differences according to the number of assets experienced by adolescents in different groups, F (3.463) = 15.78, p < .0001 (Post-Hoc Bonferroni test indicates that all groups differ significantly among each other). It is possible to observe that adolescents with higher asset levels, i. e, who experience a greater number of assets, tend to report a better SRH.

Table 3: Mean Scores (and Standard Deviation between Brackets) of Self-Rated Health according to Asset Leve 

Developmental Assets Predictors of Health

In order to determine which developmental assets and which demographic variables produce results in SRH, a stepwise multiple regression analysis (Table 4). Two sets of predictive variables were included. The first one included demographic measures as predictive variables (gender, age and schooling of the adolescent) and the second, in addition, the 40 developmental assets. Results suggest that gender (masculine) is an independent significant predictor of SRH in adolescents; this factor represents 3% of the variance in the results. An addition of the asset variables revealed that Self-esteem Internal Asset contributed the most in the prediction of SRH, and that Positive view of Personal Future Asset, Planning and Decision Making, experience of Safety, participation in Youth Programs and the experience of Youth as Resources also produced significant predictability. The final model accounted for 23% of the variance in the results for SRH of the adolescents: R2=0.24; R2(Adj.) = 0.23, F (7.462) = 20.5, p < .0001.

Table 4: Predictors of Health Perception of Health (Stepwise Regression Analysis) 

Discussion

The results of this study indicate that the largest percentage of adolescents (88% of the adolescents) state that their Health Perception is located within the positive interval, i.e., between Good, Very good and excellent. It can be observed that the adolescents’ SRH is within the interval corresponding to Very Good. Following the studies’ estimates, the results indicate that a better SRH is associated to a greater experience of Developmental Assets. These suggest that the higher the number of Developmental Assets the adolescent reports to experience, the larger the probability of their also disclosing a better SRH, therefore these data support the assumption of the additive nature of Assets in relation to their impact on positive developmental outcomes for adolescents, thus expanding the specific knowledge to the results in relation to SRH. In turn, these results are in line with ^{Benyamini’s (2011}a) the suggestion that SRH results reflect the resources of a person.

If we consider demographic variables, the data from the present study suggest that only the gender variable contributes to the variance of results in SRH. Previous studies with adults suggest the existence of gender differences in SRH, which in comparison becomes a stronger health predictor among men than among women (^{Benyamini et al., 2000}). Research on gender and age differences in adolescents’ SRH led to non-consensual results. On the one hand, in line with the present study results, most studies indicate that, boys rate their health at comparatively higher levels (^{Benyamini, 2009}; ^{Benyamini et al., 2000}; ^{Craig et al., 2018}; ^{Currie et al., 2012}; ^{Jardén et al., 2011}; ^{Vingilis et al., 2002}), even though some studies indicate that the causal effect decreases with the inclusion of other personal and contextual factors (^{Breidablik, Meland, & Lydersen, 2009}; ^{Vingilis et al., 2002}). On the other hand, no significant differences were found in relation to age and schooling. We also found that the age variable did not contribute to the variance of results for SRH. These results could suggest theoretically that this study focused on a homogenous group of adolescents that were exposed to a context that involves similar experiences.

Previous studies reports are inconsistent regarding the trajectories of SRH considering gender and age through adolescence. ^{Jardén et al. (2011)} observed that during adolescence, girls described their health and reported lower levels than boys, and that this gender difference increased over the years. In turn, in a study conducted by ^{Potrebny et al. (2019}), a lower proportion of girls, in comparison with boys, classified their health as excellent, however, it was observed that, as time went by, the number of boys that evaluated their health as being excellent decreased, while the evaluations carried out by girls improved. In a similar vein, in a study by ^{Sokol et al. (2017}) conducted with the aim of observing SRH trajectories from 13 to 31 years of age, results indicated that men obtained better initial SRH results, in comparison to women, however, they also experience larger declines in SRH over time. The authors suggested that this could be attributed to the emergences of hegemonic masculinity ideas, associated in previous studies to risk health-related behaviours (^{Heath et al., 2017}; ^{Kreager & Staff, 2009}; ^{Wells et al., 2014}). Theoretically, this construal of gender corresponds to a constructionist approach, according to which, gender-based demands create different kinds of femininity and masculinity conceptions (^{Connell, 2014}).

^{Breidablik et al. (2009}) observed that over 4 years’ time girls reported a deterioration of their SRH with higher frequency than boys, nevertheless, this difference among genders, as well as the differences among age groups, became non-significant in multivariate models, leading the authors to conclude that SRH is a relative stable construct during adolescence, and that it deteriorates steadily when there is a lack of general wellbeing and of the frequency of healthcare services, disability, and lifestyle factors that compromise health.

Biological and social differences between women and men, as well as their interaction, lead to gender differences in the perception of health and disease. These factors can also interact, in such a way that different symptoms or diseases, experienced in different social contexts, which generate gender-based expectations, can lead to larger differences than those observed in relation to biological and social factors separately (^{Benyamini, 2011}a).

Thus, the results of this study suggest that interventions aimed at improving adolescents’ general health can be more effective, if gender specific actions are considered. In turn, we emphasize that the effect of gender and age on SRH can depend on the specific sociocultural context itself (^{Benyamini, 2011}a; ^{Potrebny et al., 2019}), and the need of conducting further investigations in order to better analyze this issue is suggested. From the perspective of public health, it is also important to acknowledge possible macro-level influences in health, and to foster public policies that promote health. As ^{Benyamin (2011}a) explains, is seems impossible to understand SRH without taking into consideration the cultural context within which they are formulated.

In turn, data suggest the impact of a subset of Developmental Assets on adolescent health prediction, which indicates that both, Internal and External Assets, would contribute to the variance in the results, with Internal Assets being the stronger predictors. Self-esteem Internal Asset, which belongs to the category of Positive Identity, constitutes the stronger predictor of SRH. These results are congruent with a research that suggests that the most important component of health/disease perception is identity (^{Benyamini, 2011}a; ^{Rees, Fry, Cull, & Sutton, 2004}). Additionally, Positive View of Personal Future, also belonging to the category of Positive Identity, which indicates to what extent an adolescent is optimistic in relation to his/her personal future, also contributed to health prediction. ^{Benyamini et al. (2011)} argue that health perception is expected to be related to worldview and generalized expectations for the future, since they involve perceptions and expectations for the present and the future that are specific to health conditions. Also related to internal resources, the results highlight the important of Social Competences involving the personal capacity for Planning and Decision Making, which points to the adolescent’s ability to plan in advance and to make choices related to the prediction of health. In line with these results, ^{Benyamini et al. (2011)} suggest that the preservation of a sense of mastery and control, even in spite of very poor health, could constitute an indicator or a causal factor for survival. Likewise, External Assets, which provide contextual or environmental support conditions to optimize health, also could produce significant predictability. Results highlight the importance of the experience of Safety, i. e, of adolescents feeling safe at home, in school, in their neighborhood and their communities. Adolescents who express feelings of safety have a larger probability of feeling healthy than those who perceive their safety as being threatened (^{Benson, 2006}; ^{Simonsen et al., 2017}). The experience of participating in Youth Programs, related to the category of Constructive Use of Time, which conveys the importance of the participation in sports, clubs and school/community events, also produced significant predictability. These results match previous studies that suggest that physical activity is an important factor that produces changes in the perception of health and predicts lower mortality rates (^{Benyamini et al., 2011}). In turn, in line with previous studies (^{Simonsen et al., 2017}) results also highlight the experience of Youth as Resources, which focuses on the capacity of the community to provide opportunities that allow adolescents to contribute significantly to society, and they also indicate to what extent the community encourages youth undertaking an active role, which is also an important predictor of health.

This study’s results should be interpreted by taking into consideration the following limitations we now describe. On account of the transversal design of the study, causal relations cannot be inferred between SRH results and the experience of Developmental Assets. The development of longitudinal studies is needed to determine whether SRH results are a consequence or a determinant of the experience of Developmental Assets.