Suicidal Behaviour in Youth in Luxembourg - Findings from the HBSC 2014 Luxembourg Study

Report (2020)

Suicide is one of the leading causes of death among young people worldwide. In order to prevent suicides, early identification of groups at risk is needed. In the Luxembourgish HBSC study, data on ... [more ▼]

Suicide is one of the leading causes of death among young people worldwide. In order to prevent suicides, early identification of groups at risk is needed. In the Luxembourgish HBSC study, data on suicidal behaviours among adolescents were collected in 2006, 2010 and 2014. These can be used to identify suicide risk factors and to develop comprehensive suicide prevention programs. In Luxembourg, the suicide rate has fluctuated around 15 deaths per 100 000 inhabitants per year, for more than ten years. In the period 2006 – 2016, 20 deaths were registered as suicide in the age group of 10 to 19-year-olds. These suicides represent approximately 19% of all deaths registered in this age group. In the Luxembourgish HBSC study conducted in 2014, 875 adolescents indicated to have contemplated suicide in the last 12 months, which amounts to 15.1% of the adolescents in the study. In the same year, 811 adolescents (14.0%) indicated to have made a suicide plan in the last 12 months, and 448 adolescents (7.7%) to have attempted suicide (at least once) in the last year. In first instance, bivariate logistic regressions analyses were conducted for 24 independent variables with three suicidal behaviours (contemplation of suicide, planning of suicide and suicide attempt) and sadness as dependent variables in order to identify potential risk factors. These risk factors were further tested in multivariate logistic regressions, in order to make a statement about the relevance of these factors for suicidal behaviour of adolescents in Luxembourg, while taking into account the dependence between the risk factors. Results from multivariate logistic regressions indicate that subjective health complaints are the most important risk factor for suicidal behaviour. Adolescents who have recurrent multiple health complaints are at higher risk for suicidal behaviour than adolescents who do not have health complaints. Life satisfaction is the second most important risk factor for suicidal behaviour. Adolescents with lower levels of life satisfaction are at higher risk for suicidal behaviour than adolescents who have higher levels of life satisfaction. Gender-specific analyses show that the risk factors differ between girls and boys for suicidal behaviour. [less ▲]

School-Class Co-Ethnic and Immigrant Density and Current Smoking among Immigrant Adolescents

in International Journal of Environmental Research and Public Health (2020), 17(2),

Although the school-class is known to be an important setting for adolescent risk behavior, little is known about how the ethnic composition of a school-class impacts substance use among pupils with a ... [more ▼]

Although the school-class is known to be an important setting for adolescent risk behavior, little is known about how the ethnic composition of a school-class impacts substance use among pupils with a migration background. Moreover, the few existing studies do not distinguish between co-ethnic density (i.e., the share of immigrants belonging to one’s own ethnic group) and immigrant density (the share of all immigrants). This is all the more surprising since a high co-ethnic density can be expected to protect against substance use by increasing levels of social support and decreasing acculturative stress, whereas a high immigrant density can be expected to do the opposite by facilitating inter-ethnic conflict and identity threat. This study analyses how co-ethnic density and immigrant density are correlated with smoking among pupils of Portuguese origin in Luxembourg. A multi-level analysis is used to analyze data from the Luxembourg Health Behavior in School-Aged Children study (N = 4268 pupils from 283 classes). High levels of co-ethnic density reduced current smoking. In contrast, high levels of immigrant density increased it. Thus, in research on the health of migrants, the distinction between co-ethnic density and immigrant density should be taken into account, as both may have opposite effects. [less ▲]

Suicide Prevention: Using the Number of Health Complaints as an Indirect Alternative for Screening Suicidal Adolescents

in Journal of Affective Disorders (2020), 260

Background: Suicide is the second leading cause of death in adolescents. Screening for persons at risk usually includes asking about suicidal ideation, which is considered inappropriate in some societies ... [more ▼]

Background: Suicide is the second leading cause of death in adolescents. Screening for persons at risk usually includes asking about suicidal ideation, which is considered inappropriate in some societies and situations. To avoid directly addressing suicide, this paper investigates whether the Health Behaviour in School-aged Children Symptom Checklist (HBSC-SCL), a validated non-clinical measure of eight subjective health complaints (e.g. headache, feeling low), could be used as a tool for screening suicidal ideation and behavior in adolescents. Methods: 5262 secondary school students aged 12-18 answered the Luxembourgish HBSC 2014 survey, including the HBSC-SCL items and suicidal ideation and behavior questions. Results: Each HBSC-SCL item correlates with suicidal ideation and behavior. A sum score was calculated ranging from zero to eight health complaints to predict respondents who considered suicide (area under the ROC curve = .770). The ideal cut-off for screening students who consider suicide is three or more health complaints: sensitivity is 66.3%, specificity is 75.9% and positive predictive value is 32.9%. Limitations: One limitation is HBSC-SCL's low positive predictive value. This is a general problem of screening rare events: the lower the prevalence, the lower the positive predictive value. Sensitivity and specificity could be improved by taking age-, gender- and country-specific cut-off values, but such refinements would make the score calculation more complicated. Conclusions: The HBSC-SCL is short, easy to use, with satisfactory screening properties. The checklist can be used when suicide cannot be addressed directly, and also in a more general context, e.g. by school nurses when screening adolescents. [less ▲]

The influence of well-being, social support, media use and sociodemographic factors on problematic social media sue among Luxembourgish adolescents

in Cogent Medicine (2020), 7(1),

Background: Adolescents spend an increasing amount of time communicating online. Previous research has indicated that electronic media communication has been associated with positive outcomes on ... [more ▼]

Background: Adolescents spend an increasing amount of time communicating online. Previous research has indicated that electronic media communication has been associated with positive outcomes on adolescent well-being and development, however, problematic social media use is on the rise. This study investigates factors that influence problematic social media use (PSMU), based on previous empirical research and the Differential Susceptibility to Media Effects Model by Valkenburg and Peter (2013). Methods: The data used in this study stems from the 2018 Health Behaviour for School-aged Children (HBSC) study in Luxembourg. Data from elementary and secondary school students aged 11 to 18 was used (N = 6164), which was collected through a written survey. A four-stage hierarchical multiple regression analysis was conducted using SPSS, with problematic social media use as the dependent variable. 14 independent variables were included in the model, added in four blocks: sociodemographic factors, social support factors, well-being factors and media use factors.Results: The results indicate that in stage one of the hierarchical regression, the sociodemographic predictors accounted for 3% of the variation in problematic social media use. The addition of the social support factors to the model in stage two explained an additional 7% of the variation in problematic social media use, and the addition of the well-being factors in stage three an additional 5.3%. In stage four of the hierarchical regression media use factors were added to the model, and the four blocks of predictors accounted for 22.2% of the variation in problematic social media use (Adjusted R2 = 0.222). The most important predictors for problematic social media use were preference for online social interaction (β = 0.205, p < .001), the intensity of electronic media communication (β = 0.155, p < .001), psychosomatic complaints (β = 0.136, p < .001), perceived stress (β = 0.122, p < .001) and cyberbullying perpetration (β = 0.117, p < .001). Conclusions: The block of sociodemographic factors contributed minimally to the explanation of the variance in problematic social media use in the model. The most important predictors for problematic social media use were preference for online social interaction, the intensity of electronic media communication, psychosomatic complaints, perceived stress and cyberbullying perpetration. This suggests that there are several starting points for the prevention of problematic social media use among adolescents. [less ▲]

Patterns of health related gender inequalities – a cluster analysis of 45 countries

in Journal of Adolescent Health (2020), 66(6S), 29-39

Purpose: The paper explores gender inequalities between 45 countries across 10 health indicators among adolescents and whether those differences in health correlate with gender inequality in general ... [more ▼]

Purpose: The paper explores gender inequalities between 45 countries across 10 health indicators among adolescents and whether those differences in health correlate with gender inequality in general. Methods: Data from 71,942 students aged 15 years from 45 countries who participated in the 2018 Health Behaviour in School-aged Children survey were analyzed. For this purpose, 10 indicators were selected, representing a broad spectrum of health outcomes. The gender differences in the countries were first presented using odds ratios. Countries with similar risk profiles were grouped together using cluster analyses. For each of the 10 indicators, the correlation with the Gender Inequality Index was examined. Results: The cluster analysis reveals systematic gender inequalities, as the countries can be divided into seven distinct groups with similar gender inequality patterns. For eight of the 10 health indicators, there is a negative correlation with the Gender Inequality Index: the greater the gender equality in a country, the higher the odds that girls feel fat, have low support from families, have low life satisfaction, have multiple health complaints, smoke, drink alcohol, feel school pressure, and are overweight compared with boys. Four indicators show a divergence: the higher the gender equality in a country in general, the larger the differences between boys and girls regarding life satisfaction, school pressure, multiple health complaints, and feeling fat. Conclusions: Countries that are geographically and historically linked are similar in terms of the health risks for boys and girls. The results challenge the assumption that greater gender equality is always associated with greater health equality. [less ▲]

Trends in cannabis consumption among youth in Luxembourg

Scientific Conference (2019, September)

Background: Cannabis is the most widely consumed illegal drug worldwide. Among adolescents, cannabis use is a risk factor for cognitive decline, mental illness, social problems, and the use of other ... [more ▼]

Background: Cannabis is the most widely consumed illegal drug worldwide. Among adolescents, cannabis use is a risk factor for cognitive decline, mental illness, social problems, and the use of other psychoactive drugs. The current study presents trends in cannabis consumption among adolescents in Luxembourg. Methods: The Health Behaviour in School Aged Children (HBSC) Study in Luxembourg collected data in 2006, 2010, 2014 and 2018 using a standardized paper-pencil questionnaire. In total, 23,346 secondary schools students aged 11 to 18 years old (M=15.51, SD=1.53) responded to questions on cannabis, tobacco and alcohol consumption (lifetime and the past 30 days). Findings: In general, students who never used cannabis significantly increased over the four HBSC study waves (78%, 81.2%, 81%, 84%), whereas trends are similar for boys (74.5%, 77%, 78.2%, 81.4%), but not for girls (81.5%, 85%, 83.2%, 86.3%). Cannabis use (past 30 days) significantly differ for girls (94.1%, 94.1%, 92.8%, 93.7%), but not in general (91.7%, 92%, 90.9%, 91.7%), neither for boys (89.3%, 90.1%, 88.6%, 89.6%). Discussion: Cannabis lifetime use remains high for both genders. While consumption in the last 30 days remained stable for boys, it increased for girls over the past years. Tailored preventive interventions, based on health psychological models, are essential to educate adolescents about the social-cognitive risks of cannabis use and strengthen their capacities and resilience to resist experimental drug use and social pressure. In a context where legalization policies are discussed in various European countries, e-health approaches, for example, could be widely implemented in a cost-effective manner. [less ▲]

"Are you a boy or girl?" Who are the non-responders

Scientific Conference (2019, June 20)

Background: In many studies, participants who do not state their gender are excluded from the analysis. This may be appropriate if they do not answer the questionnaire seriously. However, some ... [more ▼]

Background: In many studies, participants who do not state their gender are excluded from the analysis. This may be appropriate if they do not answer the questionnaire seriously. However, some participants may have understandable reasons for not reporting their gender, e.g. questioning their gender identity. Objective: How many students and which students do not answer the question on gender? Methods: HBSC 2018 raw data from Ireland, Luxembourg, Belgium and France are compared. To explore the reasons for non-response, we divided the participants into 3 groups: 1. Responders answered both sociodemographic questions (age and gender) 2. age non-responders did not answer the question on age. 3. Gender non-responders answered the question on age, but not the one on gender. Results: Between 0.8% (Ireland) and 1.2% (Luxembourg) of participants did not report their gender. About half of them did not answer the age question either. However, the other half belong to the group of gender non-responders and this group is disadvantaged compared to responders: they report lower life satisfaction, lower self-rated health, more health complaints, less peer support and their WHO-5 Well-being score is lower. Not answering the question on gender is rare. If the participants answered the question on age, but not the question on gender, then the variable gender is missing not at random. Implication: The question arises whether the group of gender non-responders should be included in the analysis and whether the question on gender should be asked differently in the future. [less ▲]

Is Life Satisfaction Contagious?

Scientific Conference (2019, June)

Background: Life satisfaction (LS) is a major component of adolescents’ subjective well-being, facilitating adaptive development and influencing health. Literature shows that social support influences ... [more ▼]

Background: Life satisfaction (LS) is a major component of adolescents’ subjective well-being, facilitating adaptive development and influencing health. Literature shows that social support influences adolescents LS. In addition, the social network can affect health-related behaviors of adults - individuals that smoke or exercise tend to group together. However, the effects of others` LS on adolescents’ individual evaluation of LS (the contagion hypothesis) is still to be addressed. Objective(s): To test the contagion hypothesis of adolescents’ life satisfaction (how LS of proxies influences the individual LS appraisal). Method: Data is from 9738 students (aged 9-20) from the 2018 HBSC Luxembourg survey. A multilevel analysis was used to evaluate LS, with the school classes as subjects (model 1) to estimate the influence of being in a certain school class. Later, FAS, age and gender were entered as control variables (model 2). Results: The grand mean (intercept) for LS in model 1 was 7.57 (SE=.03, p<.001). For model 2, FAS (b=.47, SE=.03, p<.001), age (b=-.14, SE=.01, p<.001) and gender (b=-.23, SE=.04, p<.001) were significantly predictive of LS. The grand mean for LS, conditioned on the presence of FAS, age and gender, was 9.02 (SE=.05, p<.001). Interclass Correlation Coefficient decreased from model 1 (ICC=.08) to model 2 (ICC=.04). Conclusions: Results suggest that part of the variance of LS can be explained by the school class level. In other words, school class clusters have an influence on their LS, indicating that the LS of adolescents from a class partially accounts for individual LS. [less ▲]

Adolescent suicide behaviour profiles: comparing unplanned attempts, planned attempts and ideation

in European Journal of Public Health (2019)

Suicide Prevention in Luxembourg: Using the HBSC Symptom Checklist as an Alternative Tool for Screening

Poster (2018, December 07)

Introduction: Suicide is one of the leading causes of death of young people and as such, screening for suicidal ideation is a major public health concern. However, there is fear that exposure to suicide ... [more ▼]

Introduction: Suicide is one of the leading causes of death of young people and as such, screening for suicidal ideation is a major public health concern. However, there is fear that exposure to suicide-related content would encourage suicide attempt. This false idea is a great barrier to effectively screen. Hence, the need for tools without suicide content. Purpose: The goal of this study is to present a short tool that could be helpful for suicide screening and prevention. Materials and Methods: This study is based on the 2014 HBSC Luxembourg survey. A total of 5595 students aged from 12 to 18 years old in secondary school responded to a questionnaire translated to both French and German. Among others, it included the HBSC Symptom Checklist, a scale developed to measure eight health complaints (headache, abdominal pain, backache, feeling low, irritability, feeling nervous, sleeping difficulties and dizziness), as well as 4 questions asked in a logical sequence concerning sadness, suicide ideation, suicide planning and suicide attempt. Results: Multivariate Logistic Regression analyses with 24 potential indicators showed the number of health complaints as the most predictive indicator for suicide attempt (OR=1.248; C.I.: 1.175-1.325). Following, a significant ROC curve (area under the curve of 0.76; sensitivity=0.68 and specificity=0.73) and Youden Index (0.41) indicates the optimum cut-off at three complaints, with a predictive value of 17%. An alternative cut off point at four (sensitivity=0.57 and specificity=0.82; Youden Index of 0.39) has a predictive value of 21%. Conclusion: The HBSC Symptom Checklist seems to be a good predictor as each additional health complaint increases the risk of suicide attempt by 25%. In addition, it could be an alternative to traditional suicide screening measures, as its sensitivity, specificity and predictive values are similar to measures more commonly used, such as the Columbia Suicide Screen or the Beck Depression Inventory. Further work should be invested to validate the HBSC Symptom Checklist as a screening tool for suicide prevention. [less ▲]