This study demonstrates that, in the same socioeconomic area, European and non-European immigrants had higher school difficulties than their French counterparts and their risks were clearly different. European immigrants had a higher risk for grade repetition; 24% of the risk was explained by father’s occupation and family structure, and 47% by these family factors, WHOQoL-Bref domains, and unhealthy behaviours. Non-European immigrants had a higher risk for grade repetition, low school performance, and school dropout ideation; approximately 35% of the risks were explained by father’s occupation and family structure and 66%-78% by all confounders. This study is a part of a recent survey on health among French adolescents [41, 42].
We focused on middle school students because a number of deleterious life events and health related issues started in the early adolescence period [10, 43–45], and school difficulties need to be solved sooner. Many adolescents have poor living conditions [28, 44, 46] and their impact on mental health may be higher among the younger. Indeed, the first years at middle school correspond approximately to the mean age of onset of substance use, sleep disorders, suicidal ideation, involvement in violence, and violence victimization [28, 47]. This is very troubling, especially because those issues generally persist over time.
In France, there were, in 2007, 2.56 million children aged less than 18 years in immigrant families. The birth country of the head of family was in Africa (1.37 million children, mainly in Algeria, Morocco, and Tunisia), Europe (642,302 children, mainly in Portugal, Spain, Italia, and United Kingdom), Asia (393,775 children, mainly in Turkey and Indochina), and America and Oceania (147,326 children) . Therefore, most European immigrants are from developed countries while most non-European immigrants were from developing countries. Our study reveals that non-intact families were more represented among European immigrants and more markedly among non-European immigrants. Parents divorced/separated and reconstructed families were more represented among European immigrants while single parents were more represented among non-European immigrants. European immigrants were more likely to be from manual worker families while non-European immigrants were more likely to be from inactive ones. An important finding of our study is that family structure and father’s occupation explained 24% of the risk for grade repetition among European immigrants and about 35% of the risk for all grade repetition, low school performance, and school dropout ideation among non-European immigrants. It may be noted that adding insufficient income to the logistic models did not change the results. This should be explained by the strong relationships between family structure, father’s occupation, and insufficient income. Family structure and socioeconomic status appeared thus to be important barriers for school achievement among immigrants. European immigrants had thus a higher risk for school difficulties but they may not be persistent over time because they may disappear after a grade repetition. Non-European immigrants had much more school difficulties which may not be solved after a grade repetition and may lead to school dropout at 16 years with no qualification. A possible explanation is that their parents offer them less material conditions as well as lower engagement on their academic achievement which is known to be related to adolescents' academic engagement . Their parents have too low educational levels to help the student with homework and to monitor their school results. The role of socioeconomic status, family structure, and families’ resources in school achievement is well known .
Another important finding of our study is that all physical health, psychological health, social relationships, and living environment (measured by WHOQoL-Bref domains) and unhealthy behaviours were strong factors for grade repetition, low school performance, and school dropout ideation. The present survey shows that adding these factors to the logistic models reduced the odds ratio for grade repetition by 47% for European immigrants while among non-European immigrants, it reduced the odds ratio by 66% for grade repetition, 73% for low school performance, and 78% for school dropout ideation. The contributions of confounders appeared here to increase with the severity of school difficulties. Our results call for several hypotheses:
European immigrants had altered physical health, psychological health, and living environment compared to their French counterparts. These factors effect learning capacity leading to grade repetition, but the school difficulties may not be long lasting to result in persistent low school performance and school dropout ideation.
Non-European immigrants had much more altered physical health, psychological health, social relationships, and living environment compared to their French and European immigrant counterparts. These factors are generally interdependent and persistent. We found that they were massive and may affect continuously learning capacity, motivation, leading to impressive school difficulties and especially to school dropout ideation. Because of these issues the students may think that learning is not worth for them or they were not able to do it over time.
Current use of tobacco, cannabis, and other illicit drugs were more common among European immigrants and much more common among non-European immigrants compared to their French counterparts. Only non-European students initiated earlier use of all substances including alcohol. Our results confirm the relationships previously reported of tobacco, alcohol, and cannabis use with low school performance and leaving school without qualification [13, 14]. Two school-based population studies in France and the United States demonstrated that early initiation of tobacco, alcohol, cannabis and other illicit drugs was associated with a higher risk for suicide ideation and suicide attempts [34, 45]. As previously stated, tobacco and alcohol consumption may effect physical and mental performance [6–9] and cannabis use may exacerbate mental difficulties . Positive attitude toward substance use is negatively associated with introversion, stability, and social conformity . This would explain their strong associations with all outcomes variables with a trend according to their severity: the relationships between substance uses were high for grade repetition, clearly stronger for low school performance (with gender-age-adjusted odds ratios reaching 5.33), and much more for school dropout ideation (with gender-age adjusted odds ratios reaching 10). We conducted here a further analysis which showed that school dropout ideation was associated with a 6.98-fold higher risk for last-12-month suicide ideation and a 4.53-fold higher risk for lifetime suicide attempts (after controlling for gender and age).
Therefore, school may be seen by some immigrants and more especially by non-European immigrants as uninteresting, unchallenging, overwhelming or non-supportive. We found that absenteeism was more frequent among immigrants, especially for skipping school reaching 11.1% among European immigrants and 22.2% among non-European immigrants, vs. 5.6% among their native counterparts. This situation may lead the students concerned to develop negative attitudes (classes irrelevant, boring, just passing time, poor efforts, not able to connect with benefit in later life, attribution of responsibility for learning to teachers rather than to themselves, etc.) , and finally to avoidance of school and all it represents. It may be indicated that, in the same study, suicide behaviours were much more common among immigrants than among French students and the higher risk was strongly explained by family characteristics, school difficulties, unhealthy behaviours, and mental health .
This study demonstrates that grade repetition was strongly related to increasing age with a gender-adjusted odds ratio of 2.11. Age reflects in fact the duration of exposure to school difficulties. Low school performance was also related to age but with a lower gender-adjusted odds ratio of 1.42. This may reveal that school difficulties increased over time. It is very troubling that school dropout ideation was not related to age. This would mean that students who thought to dropout the school do it since a very young age. Whatever the initial causes, the longer the students have lived with them the more likely they have negative thoughts and re-engaging difficulties. Our results suggest that their living and learning conditions may be harmful, unsupportable or insurmountable giving to students little chance to see an improvement and expectations from year to year. It is not surprising that family structure, father’s occupation, physical health, psychological health, social relationships, living environment, and unhealthy behaviours contributed to 78% of the excess risk for school dropout ideation among non-European immigrants.
Regarding gender disparities, only school dropout ideation was more frequent among boys than girls with an age-adjusted odds ratio of 3.45. We conducted further analysis to explore the roles of confounders. Further adjustment for family structure and ethnic group did not change the odds ratio (3.56, 95% CI 1.93-6.58). Further adjustment for father’s occupation increased the odds ratio to 3.81 (95% CI 2.05-7.08, by 15%). Finally, taking into account all confounders increased the odds ratio to 4.78 (95% CI 2.45-9.33, by 54%). The gender difference for school dropout ideation was thus higher when controlling for all confounders. Boys had thus a 4.78-fold higher risk independently of all confounders. However, the risk for school dropout ideation was similar for non-European boys (age-adjusted odds ratio vs. French counterparts 4.82, p < 0.001, 95% CI 1.84-12.60) and girls (4.19, p = 0.067, 95% CI 0.89-19.60).
Some methodological aspects warrant comments. First, the study was based on self-reported data, but a self-administered anonymous questionnaire is widely used and arguably a good tool to study the living conditions, mental health, and unhealthy behaviours of adolescents [10, 28, 33, 43, 53]. A study on family factors and substance use among adolescents showed that self-report data were corroborated by independent teacher reports . The European and non-European immigrants were well distributed in the 63 classes. The public school for each student was determined by his/her residence place and the precise class was attributed by the school. Second, as previously stated, most non-European immigrants are from Africa and Turkey. In France, it is not allowed to investigate specific ethnic groups. So we considered only three wide groups: French, European, and non-European immigrants which are clearly different in terms of socioeconomic and cultural characteristics. We focused this survey on the population from an urban area so that the subjects are in the same socioeconomic context, free of variations across regions. The interpretation of the results may be problematic if the native and immigrant students live in very different regions. Third, given the large number of statistical tests carried out, type I error may be a concern, but it has to be pointed out that most tests were significant at the 0.001 level, with very large odds ratios estimates. Although the socioeconomic conditions of the population studied are common through France and other developed countries, our findings cannot be generalized to other populations. They need to be confirmed by further studies.
Strengths of the study also deserve to be mentioned. The participation rate was high (93.6%). The data collection was made under the supervision of the research team and with the help of a teacher with no influence on the survey. The sample size allowed us to study European and non-European immigrants separately, which may be crucial because of their socioeconomic and cultural differences. All were made to guarantee the respondents’ anonymity. For this purpose, the questionnaire excluded the birthday, the birth place, and the residential town. Data collected and the respondents’ identification number do not allow the determination of school and the precise class. Some students needed however a confirmation about the anonymity when filling in the questionnaire. The quality of responses to the questionnaire was good. The different instruments were reliable and used in previous studies on wide samples of adolescents in France, the United Stated, and a number of other countries [10, 28, 33, 34]. Grade repetition and low school performance are objective measures of school difficulties. Studying grade repetition, low school performance, and school dropout ideation at 16 years shed light on the levels of school difficulties and their risk patterns among European and non-European immigrants in early adolescence. It should be noted that the population studied was closed to the ESPAD (The European School Survey Project on Alcohol and Other Drugs) survey conducted on a large representative sample of school-based adolescents in France [28, 33] for substance use, sleep disorders, asthma, depressive symptoms, suicide behaviours, victim of violence, and implication in violence measured using the same measures (Appendix A).