Progenitores, compañeros y colegios: ¿qué impacto tienen sobre la salud y los comportamientos de riesgo de los adolescentes?

Doctoral thesis entitled “Health and risk behaviors among children and adolescents in Spain: examining the influence of individual, family and contextual factors”, defended on March 18, 2024 at the University of Las Palmas de Gran Canaria, obtaining the grade of outstanding cum praise. Thesis directed by doctors Juan Oliva Moreno and Laura Vallejo Torres. Court: Rosa Urbanos Garrido (Complutense University of Madrid), Toni Mora Corral (International University of Catalonia) and Migue Ángel Negrín Hernández (University of Las Palmas de Gran Canaria).

summary

This doctoral thesis focuses on the well-being of youth, as an element of the future of our society, because it is not only important to have healthy young people in the present, but also healthy adults in the future. Understanding the connections between health and social context is a crucial step in defining appropriate policies.

The general objective was to explore the self-perceived health and risk behaviors of children and adolescents in Spain, taking into account their family environment, their friends and their school. To carry out the three substudies that include this work, multilevel regression models were used, nested by school, using microdata from the survey. Health Behavior in School-Age Children (HBSC) in Spain.

first study (Zozaya N and Vallejo L, 2020)

The objective of the first study of the thesis was to investigate the effect of the economic crisis experienced in Spain between 2008 and 2013 on the self-perceived health of children and adolescents, taking into account that previous literature had focused mainly on the adult population, and not so much in the impact on young people.

Data from four waves (2002, 2006, 2010 and 2014) of the HBSC survey (n=77,651) were used to perform different multilevel regression models on three variables related to self-perceived health (good self-reported health; no health problems; and life satisfaction scale). The annual variation in regional unemployment rates is used as a proxy for the economic crisis. A growing set of control variables is included, consisting of individual, socioeconomic, and family and peer relationship indicators. They were estimated median odds ratios to quantify the variability between regions and between schools.

Our results confirm that economic recession is associated with worse self-rated child health in the simplest models. However, the effect was no longer statistically significant when the indicators of family relationships and peer relationships were included, suggesting the existence of a protective effect against the impact of the crisis. Our results also showed that schools had a greater effect on health than the regions of residence of minors.

second study (Zozaya N et al., 2022)

Obesity is an important risk factor for children’s health, the prevalence of which has been increasing. The literature shows contradictory results on the possible effect of the increasing labor participation of mothers on childhood obesity, due to the change it may entail in family decisions related to the education and care of children. The objective of the second study was to contribute to the understanding of this complex relationship.

We investigated whether childhood obesity in Spain was related to parental employment, separately considering situations in which: a) both parents have a job; b) only the father has a job (the mother is the main caregiver); c) only the mother has a job (the father is the main caregiver); d) neither parent has a job.

The 2010 and 2014 HBSC surveys were analyzed (n=32,694), applying multilevel models to determine the association between parents’ employment and their children’s Body Mass Index (BMI), taking into account the school effect and controlling for socioeconomic factors. To identify the possible mechanisms of this relationship, we will also explore whether dietary habits (for example, low consumption of fruits and consumption of sweets) and unhealthy leisure activities (for example, sedentary lifestyle or excessive use of screens) among adolescents could be associated with the employment situation of their parents.

Maternal employment is not associated with higher childhood BMI or unhealthy behaviors when both parents were employed. However, in some subgroups, such as girls under 15 years of age, mothers’ sole labor force participation was associated with obesity and unhealthy lifestyles. These families are likely to be of limited economic resources, suggesting that unobserved socioeconomic factors could also play a role in the observed association. On the other hand, the school effect was similar to family wealth. Our results identify certain asymmetries that can lead to an inequitable distribution of childcare or the results of said distributions.

third study

Bullying is a global concern with long-lasting negative implications. However, studies on harassment with a multilevel approach. The objective of the third study was to characterize the victimization and perpetration of the harassment -individually and simultaneously-; and explore what bullying factors could be related to the decrease in schoolchildren experienced in Spain during the last decade.

The 2014 and 2018 HBSC surveys were used (n=64,886). The analysis focused on different variables of harassment perceived (being harassed; harassing others; being simultaneously a harasser and a victim) and four types of observed bullying (physical, verbal, relational, and cyberbullying). A descriptive analysis was carried out to compare the situation in both years; and multilevel logistic regressions were fitted for each variable of harassmentwith schools as the second level, controlling for individual, family and contextual characteristics of the students.

Our results revealed that bullying experienced a consistent decline across all types of bullying definitions. harassment, although policies could have focused more on visible and more recent harassment. The variable most significantly associated with being bullied was having bullied other students, and vice versa. Men were more prone to physical harassment, and women to relational harassment and cyberbullying. The probability of reporting having been bullied decreased as students grew older, and increased for grade repeaters, immigrants, and overweight/obese children. Perceived support from friends and family showed a sustained protective effect on the harassment, while its prevalence increased throughout the period analyzed. The unexplained variation between schools was as relevant as the individual characteristics of the students to understand the prevalence of bullying.

Conclusions

A contribution of our work is the sustained protective effect of the social context on the health and risk behaviors of adolescents. Therefore, it is crucial to enhance good communication and relationships with family members and colleagues. Our findings also suggest that the school has an influence that may be even greater than that of the individual characteristics of the student, highlighting the need to promote a positive school climate and reduce inequalities between schools.

Other implications are related to the importance of maintaining solid active employment policies, as well as social protection policies focused on single-parent families and less well-off households. We also suggest the need to distribute the burden of care more equitably between both parents, and to achieve working conditions better designed to promote conciliation between family and professional life.

The work has certain limitations that must be noted. The cross-sectional design of the HBSC study limits the ability to perform longitudinal analyzes and establish causal relationships, which is why it suggests to public authorities a change in its design. On the other hand, the use of self-reported data could show problems of perception and precision, and the lack of homogeneous questions between the different waves prevented a broader comparison between years for some variables. Finally, the lack of availability of additional detailed variables also contributed to limitations.

It is important to design timely and holistic policies for the well-being of present and future populations. In this sense, we hope that this work can serve as a basis for health promotion and health education policies, which should include knowledge of the school environment and the social protective factors that surround our young people.