Health Literacy in Schools? A Systematic Review of Health-Related Interventions Aimed at Disadvantaged Adolescents

Socioeconomically disadvantaged populations are at greater risk of adopting unhealthy behaviours and developing chronic diseases. Adolescence has been identified as a crucial life stage to develop lifelong healthy behaviours, with schools often suggested as the ideal environment to foster healthy habits. Health literacy (HL) provides a possible solution to promote such healthy behaviours. The aim of this study was to review school-based HL-related interventions targeting socioeconomically disadvantaged adolescents and to identify effective intervention strategies for this population. Searches were performed in six databases. Inclusion criteria included age: 12–16; the implementation of a school-based intervention related to HL aimed at socioeconomically disadvantaged populations; an intervention focused on: physical activity (PA), diet, mental health, substance abuse or sleep. Forty-one articles were included, with the majority focusing on PA and diet (n = 13), PA (n = 9) or mental health (n = 7). Few interventions focused solely on substance abuse (n = 2) or sleep (n = 1), and none targeted or assessed HL as an outcome measure. There was huge heterogeneity in study design, outcomes measures and effectiveness reported. Effective intervention strategies were identified that can be used to guide future interventions, including practical learning activities, peer support and approaches targeting the school environment, the parents or that link the intervention to the community.


Introduction
Morbidity and mortality rates from obesity and non-communicable diseases (NCDs), such as type 2 diabetes and cardiovascular disease, are rapidly rising [1,2]. Poor lifestyle behaviours, including low levels of physical activity (PA), high levels of sedentary behaviour, poor dietary habits, substance abuse, poor sleeping habits and mental illness, are said to be large contributors to the NCD burden [3][4][5]. This not only places a heavy burden on society but also on limited health resources [6].
Furthermore, there is extensive literature describing socioeconomic health inequalities in terms of morbidity and mortality. Populations from socioeconomically disadvantaged backgrounds are less likely to avoid unhealthy behaviours [7], and are consequently more likely to develop chronic diseases [8]. There is, therefore, a growing need for the promotion of a healthy lifestyle as a form of primary prevention, particularly in low socioeconomic populations [9].
The World Health Organisation (WHO) defines health literacy (HL) as the cognitive and social skills which determine the motivation and ability of an individual to gain access to, understand and use information in ways which promote and maintain good health [10]. The concept of HL has evolved from what was initially a focus on an individual's ability to read and understand health information, to focusing on factors that affect an individual's knowledge, motivation and competencies in relation to health [11,12]. HL has been 1.
Included typical adolescents with a reported mean age between 12 and 16 years; 2.
Self-reported that the participants were from a socioeconomically disadvantaged (or equivalent) background; 3.
Included the implementation of an intervention related to health literacy (increases health knowledge, understanding, awareness, motivation, confidence) in at least one of the following areas: physical activity, sedentary behaviour, dietary habits, sleeping habits, mental health or substance abuse; 4.
Included school-based interventions, interventions that could be feasibly implemented in a school setting or interventions that could be linked to a school curriculum; 5.
Aimed to increase health knowledge/comprehension, understanding, behaviour, value, well-being, motivation, self-efficacy or self-monitoring in relation to any of the following domains: physical activity, sedentary behaviour, dietary habits, sleeping habits, mental health or substance abuse.
Exclusion: Studies identified in the literature search were excluded if: 6. They included special populations (e.g., children with learning difficulties, pregnant adolescents, exclusively obese individuals, or those with a specific health condition); 7.
The intervention did not include an educational element or a component targeting health literacy (increases health knowledge, understanding, awareness, motivation, confidence); 8.
They were book chapters, case studies, student dissertations, conference abstracts, review articles, meta-analyses, editorials, protocol papers or systematic reviews; 9.
They were not published in English or in a peer-reviewed journal; 10. The full-text article was not available.
For full the PICO statement, see Appendix A.

Information Sources, Search Strategy and Study Selection
Six electronic databases were searched-MEDLINE/PubMed, ERIC, CINAHL, Psych-INFO, and EMBASE-to identify relevant evidence. 'English' and 'peer reviewed' filters were marked on all searches. The search strategy was developed using Boolean operators (AND/OR), incorporating the relevant terms (Appendix B). The search was not limited to any publication time frame. The search was conducted between June and August 2020. All records were exported to Mendeley reference managing software for screening and all duplicates were removed.
Two reviewers (C.S. and H.R.G.) independently assessed the eligibility of the studies. Following title and abstract screening, full-text copies of potentially relevant studies were obtained and screened for full-text inclusion. In the case of disagreement, a third author (S.B. or J.I.) was contacted for discussion until consensus was reached. Figure 1 below details the search and screening process. The literature search yielded 8074 publications; after removing 1908 duplicates, 6796 publications were subsequently screened. Of these publications, 6417 were excluded based on title and abstract because they did not fulfil one or more of the inclusion criteria. The remaining 379 publications were retrieved for full-text review. A total of 338 failed to meet the inclusion criteria. The main reasons for excluding full texts were that the intervention targeted a population which was outside of the age range, the intervention was not implemented with populations which were socioeconomically disadvantaged. Finally, 41 publications were included for review.

Quality Appraisal
The quality appraisal of the included studies is summarised in Table 2. Many studies were reported as 'unclear risk' as they did not provide adequate detail to assess the level of bias. Furthermore, studies which were not RCTs could not be assessed for selection bias as there was no random allocation. Of the studies that were assessed for selection bias, nine reported as 'low risk' across both assessments. Due to the lack of information, 15 of studies were reported as having an 'unclear risk' of detection bias. Of the remaining studies, seven were reported as 'high risk' and 19 as 'low risk'. The majority (n = 35) were scored as 'low risk' for attrition bias, with two reported as 'high risk'. Reporting bias was reported as 'low risk' in all included studies.

Intervention Results
In the interventions targeting both PA and diet, the behavioural outcomes reported were mixed. Five of the interventions improved both PA levels and dietary habits significantly [55,65,66,69], whereas two observed no significant improvement in either behaviour [64,84]. In the interventions aiming to improve PA and nutrition knowledge, both were effective (no significance level reported) [49,69]. Interventions targeting changes in anthropometric measures, such as BMI and body composition, also showed mixed levels of effectiveness. One intervention did not improve any measure of anthropometry [77], while another significantly improved body fat percentage but not BMI [70]. Furthermore, one significantly improved the percentage of individuals who were overweight or obese but did not improve total fat percentage, fat mass or fat free mass of participants [67], and finally Lubans et al. [71] improved BMI and body fat percentage slightly, but not significantly. The two interventions targeting psychosocial correlates associated with PA and dietary habits resulted in significant improvements [55,84]. Interventions which assessed the student's perception of the intervention displayed positive responses [55,81].
In the interventions targeting PA only, five targeted behavioural outcomes (levels of PA). Four of which were unsuccessful in improving PA levels in their targeted cohorts [58,73,74,80], and one observed a significant improvement in girls but not in boys [61].
Health-related quality of life, however, was significantly improved in one these interventions [74] and muscular fitness in another [73]. Furthermore, interventions focused on reducing screen time did not observe any positive changes in behaviour [76]. Interventions assessing anthropometric measures had higher levels of success, with one study reporting a significant reduction in weight and BMI among the intervention group [75], while another observed a significant reduction in body fat percentage [73] and one observed modest positive changes in body fat percentage [47]. An intervention assessing the dose, reach and fidelity reported positive findings on all three variables [48].
The majority of the mental health interventions targeted psychological functioning, with the authors reporting mixed levels of success. Depressive symptoms were measured in two interventions-one reporting a non-significant improvement [51] and the other finding no intervention effects [85]. In addition, Schleider et al. (2019) [51] reported no improvement in social anxiety symptoms following the intervention. Similarly, Dray et al. [72] found no significant differences between the control and intervention group's internalising problems, externalising problems and prosocial behaviour scores following the intervention. In a park-based intervention, results showed no significant improvements in parent-reported social skills or problem behaviours. However, staff-reported findings highlighted significant improvements in both problem behaviours and social skills at follow up [56]. In other interventions targeting mental health, social, emotional and academic performance [53], attention and self-efficacy [83], and school climate [82] were significantly increased.
Interventions specifically targeting diet had varied levels of success. Three programs targeted dietary behaviours specifically. One intervention reported a significant increase in the intake of fruit and vegetables and a statistically insignificant decrease in highly processed foods, compared with baseline [54]. Significant increases, however, were seen in psychosocial mediators, and qualitative assessments suggest that the intervention promoted skill building, but environmental barriers made these difficult to use. The two other interventions targeting dietary behaviours reported significant improvements as a result of the interventions [62,78]. School nutrition practice and policy was also significantly improved in the intervention by Alaimo et al. [62]. Finally, one intervention aimed to measure the students', teachers' and parents' perceptions of the intervention, and to identify attributes that were highly valued within [60]. Four key themes ((1) development of life skills, (2) food and health, (3) family and community, and (4) experiential and participatory learning environment) were identified, and all stakeholders positively appraised the intervention.
Both interventions solely targeting substance abuse aimed to improve behavioural outcomes. An intervention by Robinson et al. [50] targeting a reduction in cigarette smoking, alcohol consumption and marijuana use reported significant reductions in cigarette and marijuana use, but not in alcohol consumption. Vicary et al. [68] reported mixed findings on cigarette use and alcohol consumption over the assessment time points in the two intervention groups. Similarly, measurements of skills related to substance abuse (decision making skills, refusal skills, media awareness and resistance skills) resulted in mixed findings.
A multi-health domain intervention which targeted PA, dietary habits, substance abuse and mental health [59] resulted in significantly improved cardiovascular health knowledge scores in males and females. In females, dietary habits, total cholesterol and estimated VO2max also significantly improved. All other risk factors were non-significant in males and females. A sleep intervention qualitatively assessed the acceptability of two mobile phone apps to improve sleep hygiene [57]. The overall feedback on the application was positive, although several barriers were identified, and the students were sceptical about successfully adopting sleep hygiene practices. Furthermore, an intervention targeting both mental health and sleep significantly improved some psychological functioning measures (anxiety, negative coping approaches and self-reported anger) but not cortisol levels or sleep scores [63]. An intervention targeting mental health and PA, which used qualitive analysis to understand the perceived impact of the intervention on youth participants, found that the program was a promising method to improve psychological, social and physical well-being [87]. The adolescents, parents and program implementers described benefits across seven main areas, including dancing and related skills, behaviours (e.g., reduced television viewing), physical well-being, psychological well-being, relationships, respect for others and for diversity, and school performance. A PA, diet and mental health intervention assessed predictors to dropout in the 'Mexa Se' intervention [79]. It was reported that in the intervention group, age, body mass, height and BMI were all significant predictors of drop out. Finally, in the intervention implemented by Kerr et al. [52], aimed at improving PA, diet and substance abuse, it was reported that general health knowledge scores increased significantly more in the intervention group compared to the control. Health behaviours, however, did not significantly differ between groups post-intervention.

'Hands-On' or Practical Learning
Of the interventions identified within this review, many adopted 'hands-on' or practical learning components, a strategy that proved effective across various health domains. For example, an intervention targeting social, emotional and academic function by Mendelson et al. [53] incorporated mindfulness sessions into the program; resulting in significant improvements across all three health scores (social, emotional and academic function) post-intervention. Similarly, participants in an intervention by Sibinga et al. [63], who practiced mindfulness and yoga within the intervention, reported significantly improved psychological functioning. Furthermore, in an intervention targeting dietary habits [54], participants prepared and ate minimally processed meals with the objective of making these foods look 'cool and fun'. The intervention resulted in a large significant increase in the participant's consumption of minimally processed foods. Another example of practical learning was articulated in Jackson et al. [69], where the students learned pertinent nutrition and PA information that was later incorporated into writing and performing their own "healthy" skits in a theatre-based program. This intervention resulted in improved knowledge around PA and dietary habits, improved health-related choices, and the students reported that the intervention was an enjoyable experience.

Peer Support
The use of peer support was another effective intervention strategy identified within this review. For example, a social marketing intervention by Aceves-Martins et al. [63], which aimed to encourage adolescents to increase their fruit and vegetable intake, PA levels and reduce screen time, was led and designed by peer students (who were trained by university specialists). The social marketing intervention resulted in a significant increase in fruit intake, PA levels and a reduction in screen time. Similarly, two interventions [64,66], also targeting dietary habits and PA levels, which used the Transtheoretical Model to tailor intervention strategies and feedback suggestions to adolescents, also adopted a peer-led approach. Participants in the preparation, action and maintenance stages of change were used as peer models for students in the pre-contemplation and contemplation stages of change and led interventions sessions targeting healthy eating and exercise promotion with the assistance of nursing students. Both interventions produced findings supporting the peer-led approach, with one intervention reporting significant changes in dietary behaviour and PA in the intervention group compared to the control [66], and the other reporting nonsignificant changes in dietary habits and significant changes in PA levels in the intervention group compared to the control [64].

Holistic Approaches
Finally, this review identified the use of holistic approaches to school-based health interventions as an effective strategy. It has been suggested that for health interventions to be effective, they should focus on more than just the educational component [88], with research stating that interventions should also target the school environment and engagement with families or communities (or both) [34]. The benefit of adopting an intervention framework which targets all three areas is demonstrated by Hollis et al. [75]. The intervention, which incorporated teaching strategies to maximise PA and PE lessons (formal health curriculum), the development of school policies and PA lunch programs (school environment), and the development of a community PA expo and community newsletters (community links), reported significant improvements in weight, BMI and PA [89] (these effects were reported in a paper outside of this review). Although no other interventions included in this review met all three of the criteria of the WHO HPS framework as comprehensively as Hollis et al. [75], other interventions which targeted either the community/parent links, or the school environment also reported successful findings. For example, the intervention presented in Casey at al. [74], which was shown to significantly improve health-related quality of life, implemented a PE component that was linked to a community component addressing previously reported barriers to PA participation. Furthermore, the pedagogical approach of the PE program in this study [74] aligned with recent developments in the community sports club. Additionally, a park-based program targeting social emotional learning by engaging the parents through family sessions (supplementing the youth's sessions) developed specific strategies that the families could model and reinforce at home [56], ultimately leading to significant improvements in the staff reported measures of problem behaviours and social skills in the adolescent's cohort. Similarly, an intervention by Jackson et al. [69], which was successful in improving PA and dietary knowledge and behaviours (although no significance was reported), involved the parents by hosting healthy eating recipe sessions, completing home-based intervention sessions and inviting the parents to a performance by the adolescents where pertinent dietary or PA information was translated into a theatrical performance. Finally, an intervention by Alaimo et al. [62] involved the assembly of a 'Coordinated School Health Team' (CSHT) to improve school nutrition and practices and policies. The CSHT, which was made up of representatives from various sectors of the school (including students), met to discuss school nutrition policies, nutrition environment, health education programs and school food service programs. Intervention schools adopted significantly more nutrition policies and practices than schools in the control group. In addition, students from the intervention schools consumed significantly more fruit and fibre, and less cholesterol than students from the control schools.  Physical Health Module: 17% of the participants passed the pre-test (60% or higher). 67% percent of all participants passed their post-test (60% or higher, achieving that objective). 75% of these participants improved their test scores by at least 10%, also achieving that objective. Following the end of the intervention program, the survey showed a positive change in behaviour in 75% of the participants. Healthy Eating Habits Module: 29% of the participants passed the pre-test (60% or higher) .50% of the participants passed the post-test (60% or higher), failing to achieve that objective. 86% of all the participants improved their scores by at least 10%, successfully achieving that objective. The survey following the intervention showed that 50% of the participants positively changed their personal eating habits.

No
Black et al. [67] Health promotion and obesity prevention program A manualised 12-session intervention ("Challenge") included a rap music video promoting healthy eating and PA principles of mentorship (role modelling and support), participatory learning, goal setting. In addition to setting dietary and PA goals, tracking and evaluating progress, and revising goals as necessary, intervention adolescents prepared and tasted healthy snacks and engaged in PA.
Number of participants = 121. The percentage of overweight/obese adolescents declined from 54% to 36% in the intervention group, while the percentage declined from 36% to 32% among the control group. The percentage overweight/obese adolescents further declined to 35% in the intervention group, but increased to 38% among the control group at delayed follow up, resulting in a significant difference between groups (χ 2 = 5.8, p = 0.02, GEE). There were no statistically significant interactions between intervention and time in either post-intervention or delayed follow up for total percent body fat, fat mass or fat free mass in the multi-level modelling with the whole sample. No intervention effects were found for consumption of breakfast, fruit, soft drinks or sweet and savoury snacks. Positive intervention effects were found for self-efficacy for having a daily breakfast (p < 0.01), positive attitude towards vegetables consumption (p < 0.01) and towards lower soft drink consumption (p < 0.001). A trend towards significance (p < 0.10) was found for self-efficacy for reaching the PA guidelines. For sports participation, no significant intervention effect was found. 92 students completed the process evaluation questionnaire, the feedback was largely positive.

Yes
Frenn et al. [66] To improve the adoption of a diet lower in fat and duration of PA The primary classroom strategy for the sessions was consciousness raising and self-re-evaluation.Separate smaller group sessions were held for students in the preparation, action, and maintenance stages of change Examples of content used include food pyramids, food diaries, peer leadership, planning exercise sessions. (Specific strategies detailed in the paper.)Number of participants = 60.
Paediatric nursing students 4 classroom intervention sessions (duration of intervention not reported)

Transtheoretical
Model and Health Promotion model When the Health Promotion/Transtheoretical Model interventions were used in 4 classroom sessions, students had a significantly (p <.05) reduced trend toward choosing a diet higher in fat and increased duration of PA, as compared with a control group. Internet and video sessions for students in the precontemplation and contemplation stages of change focused on raising awareness of current eating and exercise, identifying pros (or benefits) of both low-fat diet and exercise, and overcoming cons (or barriers) to consuming low-fat diets and participating in exercise. Those in the precontemplation, action, and maintenance stages of change were prepared as "peer models" and led the healthy snack and exercise labs with the assistance of senior nursing students and faculty. Online feedback was given to all students in the intervention for each internet session. (Specific strategies detailed in the paper.)Number of participants = 67.

No
Peer models (students), nursing students and other faculty members Academic year

Transtheoretical
Model and Health Promotion model The difference in percentage of dietary fat intake between the intervention and control groups as a whole was not significant. Both control and intervention groups decreased their amount of MVPA, but the level of decrease in MVPA was less among the intervention group (-8.58 min) as compared to the control (-37.61 min; p = 0.024)

No
Frenn et al. [65] To increase PA and reduce dietary fat among low-income, culturally diverse, 7th-grade students  The adolescents appreciated influencing the components of the intervention and collaborating with peers in active learning activities such as practicing sports and preparing meals. They also reported acquiring new health information, that trying new activities was inspiring, and the use of pedometers and photo-food diaries helped them reflect on their health behaviours. This was echoed by teachers. Issner et al. [55] To motivate urban, minority youth to make healthy changes in diet and PA All participants engaged in goal discussion with a health coach that lasted 3-5 min. The enhanced intervention group continued the intervention after the goalsetting portion wherein facilitators used probes to discuss previous experience, elicit positive change talk, discussion of benefits, and ideas for potential solutions.Number of participants = 51.

No
"Health coaches" (first and second author and research assistants) 20 min-1 session  Jackson et al. [69] To engage low-income, urban, African American adolescents and their families in learning ways to adopt a healthy lifestyle Interventions components included nutrition and PA information education, preparation and performance of their own "healthy" skits, team building exercises designed to introduce theatre dynamics, which progressed into script-writing activities, making healthy recipes or snacks. Each session ended with some form of PA (circuit training or a dance routine). At the end of the 6 week program, students performed 'Getting on Track' for family and friends. Parents were engaged in three ways: (1) participating in a health information and recipe session, (2) completing home-based activities, and (3)   Yes

PA (only) Interventions (n = 9)
Bandeira et al. [76] To promote PA and reduce the screen time Teacher training, which was focused on lifestyle behaviours, including excessive screen time, and its implications for health support material, was delivered to teachers to assist them in organising classes on the topic. Component opportunities were created to encourage PA and decrease screen time in the school environment so that adolescents could play games/sports during free time at school. Supervised sessions of 10-15 min ("Gymnastics at School") were performed twice a week. Health education messages were spread across the schools, and pamphlets were handed over to students/parents. The messages were also addressed to the psychosocial (self-efficacy, attitude, and social support) and environmental aspects of the practice of PA and reducing sedentary behaviour, especially screen time.
Differences in PedsQL were also present in the 3-group analysis (intervention completers, intervention non-completers and control), whereby the intervention non-completers had significantly higher scores (84.0 ± 0.8, p = 0.021; 80.4 ± 0.9, p = 0.003; and 81.7 ± 0.8, p = 0.002, respectively) than the control group (80.9 ± 0.8, 76.1 ± 0.9 and 77.8 ± 0.8, respectively  Hollis et al. [75] To reduce the decline in PA typically observed during adolescence The intervention components targeted the school curriculum, school environment, and broader community and parental support. School curriculum included teaching strategies to maximise student PA in health and PE lessons, development and monitoring of student PA plans within PE lessons and the implementation of a 10 week enhanced school sports programme.School environment included the development and modification of school policies, PA programmes during school breaks and promotion of community PA providers. Additional interventions strategies included an in-school PA consultant 1 day per week, establishing leadership and support, teacher training resources, teacher prompts and intervention implementation performance feedback to schools. Parent engagement: information was regularly sent to the parents via school newsletters, the school website and newsletters on PA recommendations, school-based PA strategies, promotion of community PA providers and strategies to support their child's PA. Number of participants = 645. months, with group-by-time effects also found for BMI z-score (−0.08 (−0.14; −0.02), p = 0.02) favouring the intervention group. Intervention effects were significant for all adiposity outcomes at 12 and 24 months in both the complete cases and multiple imputation analyses. There was weak evidence of a differential treatment on effect on weight in males compared with females (three-way interaction p = 0.22). Among males there, was a statistically significant treatment effect at 24 months in favour of the intervention group (−1.26 kg (−2.11; −0.41), p = 0.01). There were no significant effects on weight, BMI and BMI z-score at either 12 or 24 months for females. There was weak evidence of a differential treatment on effect on weight in males compared with females (three-way interaction p = 0.22). Among males, there was a statistically significant treatment effect at 24 months in favour of the intervention group (−1.26 kg (−2.11; −0.41), p = 0.01). There were no significant effects on weight, BMI and BMI z-score at either 12 or 24 months for females. Weight status at baseline: minimal evidence of differential treatment effects depending on baseline weight for weight (p = 0.50), BMI (p = 0.57) or BMI z-score (p = 0.64).
PA level at baseline: no evidence of differential treatment effects depending on activity status at baseline for weight (p = 0.94), BMI (p = 0.95) or BMI z-score (p = 0.31). There was no significant effect on weight, BMI or BMI z-score for either active or inactive students at 12 or 24 months.   There was no evidence of any clinically important differences between the intervention and control arms in depressive symptoms scores at 3 months (adjusted difference in means, −0.19; 95% CI, −1.22 to 0.84; p = 0.72) or at 12 months. The adjusted difference in the primary outcome at 3 months between trial arms was −0.15 (95% CI, −1.12 to 0.81; p = 0.75) with 20 imputed full data sets.

Yes
Dray et al. [72] To increase the provision of universal strategies targeting multiple internal and external resilience protective factors A framework of sixteen intervention strategies. Each strategy was designed to address one or more internal or external resilience protective factor. Intervention schools were asked to meet the prescribed set of strategies; however, schools were given the flexibility to select which specific programs or resources to implement to address each of the strategies. ( There was no significant difference between intervention and control groups for the outcomes of total SDQ, internalising problems and prosocial behaviour There was a significant difference for the outcome of externalising problems in favour of the control group, though the magnitude of effect was small (b 1     Relative to girls in the control group, girls receiving the GM-SSI reported modest but significantly greater reductions in depressive symptoms (d = 0.23) and likelihood of reporting elevated depressive symptoms (d = 0.29) from baseline to follow up. GM-SSI effects were non-significant for social anxiety symptoms, although a small effect size emerged in the hypothesised direction (d = 0.21), and non-significant for change in conduct problems (d = 0.01). Sethi et al. [83] To improve attention and self-efficacy

No
All the students participated in the yoga course for 5 days. The module was selected from Integrated Approach of Yoga Therapy for positive health. Number of participants = 60.
Not reported 5 days None reported The intervention resulted in a significant increase in self-efficacy (p = 0.001) and attention scores (p < 0.001).

No
Shinde et al. [82] To improve school climate and health-related outcomes The intervention identifies four priority areas for action: promoting social skills among adolescents; engaging the school community in school-level decision-making processes; providing access to factual knowledge about health and risk behaviours to the school community; and enhancing problem-solving skills among adolescents. The intervention strategies were organised at three levels-whole school, group, and individual levels-and include school health committee, awareness generation fun activities, speak out box (letterbox), wall magazine, competitions (debates, poster making, quizzes, etc.), health policies, peer groups, workshops, individual counselling.
Number  There was a significant increase in mean frequency of fruit and vegetable intake at post-test compared with pre-test (t = 3.359, p < 0.01)-an effect size that is considered to be large (d = 0.59). Small effect sizes (d = 0.34) were seen for mean intakes of highly processed foods but change in score was not statistically significant (p = 0.06). Statistically significant increases in outcome expectations and self-efficacy for fruit and vegetable intake occurred from pre-test to post-test; all other mediators showed no statistically significant changes. Youth discussed supports for eating fruit and vegetable intake, including social support and modelling, the application of self-regulation skills to increase intake, and their expressed preferences for fruit and vegetable. Youth were also preoccupied with the negative physical outcome expectations of eating highly processed foods, such as developing diabetes.

No
Brito Beck da Silva et al. [78] To promote adequate and healthy eating Eight meetings lasting 50 min each were provided to promote healthy eating and PA. The topics covered were: 1) healthy eating; 2) PA and sports; 3) fats, sugars and salt: effects of a poor diet; 4) nutritional evaluation: The basics; 5) use of supplements in PA: A critical approach; 6) food labelling: food and nutritional safety; 7) role of nutrients in health promotion: functional foods with a focus on fruits, vegetables and legumes; 8) good food handling practices. A webpage was created in a social network used by the adolescents, in which videos, trivia and general guidelines on healthy eating were posted.For parents and/or guardians, didactic-educational materials were sent through adolescents to encourage them to maintain a healthy lifestyle. Number of participants = 387.  Four primary themes emerged from the focus group data: (1) development of life skills, (2) food and health, (3) family and community, and (4) experiential and participatory learning environment. These core themes and subcategories of the themes were organised into levels of the socioecological model.

No
Alaimo et al. [62] To improve school nutrition practices (including nutrition education) and policies, and to improve student dietary intake Schools were asked to convene a Coordinated School Health Team (CSHT) with representatives from various sectors of the school (administration, faculty, food service, health care, and students).Schools were provided with a trained facilitator to meet with their CSHT on one time to complete the healthy School Action Tool (HSAT) healthy eating and nutrition topic area (questions on the following topics: school nutrition policies, school nutrition environment, school health education programs including nutrition education, and school food service programs). At the end of each module, schools were to identify several "bright ideas" they could implement. Schools were asked to prioritise their goals and received $1000 to implement nutrition education or nutrition marketing activities in their action plans.Number of students unknown (40 schools).

Not reported 2 years None reported
• Adoption of school nutrition practices (including nutrition education) and policies (questionnaire) Schools that completed the HSAT prior to but not during the School Nutrition Advances Kids (SNAK) project reported adopting more nutrition policies than schools that never completed the HSAT or a similar program (2.2 vs. 0.4 nutrition policies).Schools that completed the HSAT at any time (prior to but not during the SNAK, during but not prior to, and both prior and during the SNAK project) reported adopting significantly more nutrition practices than schools that never completed the HSAT or a similar program (6.8, 5.8, 7.0, vs. 1.6 nutrition practices, respectively). Schools that completed a similar assessment or grant program before or after the SNAK project also reported adopting significantly more nutrition practices than schools that never completed the HSAT (4.3 vs. 1.6 nutrition practices). Students in schools that were randomised to complete the HSAT reported consuming significantly more fruit (17.5%) and fibre (4.9%) and less cholesterol (4.2%) than students in the control schools. Students in schools that completed the HSAT during the intervention reported consuming significantly more fruit (20.1%) and fibre (5.1%) and less cholesterol (8.4%) than students in schools that had never completed the HSAT. Vicary et al. [68] To reduce the risk for initiation of substance abuse or reduce increased use in high-risk females

No
The life skills training (LST) condition, which is usually taught by a limited number of teachers in a series of classes dedicated to substance abuse prevention. The infused (I)-LST condition integrates life skills training and alcohol, tobacco or other drugs (ATOD) information into a variety of the existing grade level subject curricula by the teachers for these subject areas.
The goal of such an approach is to make prevention an integral part of the total curriculum. I-LST teachers were trained by university/project staff. Number of participants (LST) = 234. Number of participants (I-LST) = 297.

Teachers 2 years None reported
Students' self-report of: The LST low-risk females reported a significantly lower frequency of alcohol use, less binge-drinking, and less marijuana use, while the I-LST low-risk females reported significantly less cigarette smoking. However, the only significant substance use effect that remained for the low-risk females by the end of 8th grade was [less] cigarette smoking for I-LST.Both LST and I-LST positively affected knowledge of alcohol, tobacco, and other drugs. The I-LST program demonstrated a desirable effect on the normative beliefs of the low-risk females at the end of year two. Among skill variables, treatment effects were found for LST low-risk females for decision making, communication, and coping skills at the end of year one, although, at the end of year two, these effects had disappeared. LST females demonstrated significantly worse media resistance skills at the end of year two. Two skill treatment effects existed for the low-risk I-LST females. At the end of year one, I-LST positively affected decision-making skill; however, this effect was reduced to a non-significant level by the end of year two. The I-LST program resulted in greater coping skills by the end of year two for the low-risk females. More positive results were observed, however, for the females at higher risk, with significant treatment effects found in a number of substance use categories. After the first year of programming, high-risk females in the LST program were less likely to use alcohol (both for any drinking and for binge drinking), marijuana, and inhalants. A significant treatment effect was also shown among high-risk females in the I-LST program for drinking, binge drinking, and marijuana use at the end of the first year. The LST program significantly affected pro drug attitudes, normative beliefs, and knowledge of ATOD myths and realities although the effects did not remain at a significant level by the end of year two.The LST program resulted in two treatment effects for the high-risk females, assertiveness and refusal skills. A significant treatment effect was observed for the high-risk females in the I-LST for normative beliefs by the end of year one, although the effect did not remain through year two. A significant I-LST treatment effect was observed for attitudes toward ATOD and refusal skills at year two among the high-risk females. For the analysis of cigarette smoking, a significant grade 3 SBHC interaction effect was found, F(1, 585) 5 3.83, p 5 0.05. The SBHC students smoked slightly less than non-SBHC students in the 9th grade; but, by 11th grade, SBHC students were smoking significantly less than non-SBHC students.The SBHC 3 grade interaction effect for alcohol use was non-significant, F(1, 586) 5 0.39, p 0.50, although students from SBHC schools (M 5 1.32, SD 5 3.24) reported drinking slightly less frequently than students from non-SBHC schools (M 5 1.60, SD 5 3.75), this difference was not significant, F(1586) 5 2.45, p 5 0.12 For marijuana, a significant grade 3 SBHC interaction effect was found, F(1, 587) 5 12.72, p, 0.001. By 11th grade, marijuana use had significantly decreased among SBHC students while marijuana use among non-SBHC students dramatically increased.

PA, Diet, Substance Abuse and Mental Health Interventions
Fardy et al. [59] To promote health knowledge and behaviour, coronary risk factors, and cardiovascular (CV) fitness 20-25 min circuit training classes followed by 5 min of health behaviours lecture/discussion (topics included exercise, nutrition, smoking cessation, stress management, heart disease, cancer, and motivation Cardiovascular health knowledge scores significancy increased in the intervention group, whereas they decreased in the control group. Significant changes in self-reported dietary behaviour were observed in female subjects. Significant changes in risk factors were restricted to lowered total cholesterol in girls. Mean cholesterol values in female subjects decreased from 165 to 149 mg/dl in the treatment group, whereas female controls decreased from 154 to 150 mg/dl. There were no significant differences in blood pressure, obesity, and self-reported PA. Estimated mean CV improved in females from 33 to 38 mL/kg per min (p < 0.0001), at heart rates of 176 and 152, respectively, whereas control subjects increased only from 33 to 34 mL/kg per min at heart rates of 178 and 172, respectively. In male subjects, treatment and control groups improved from 43 to 52 mL/kg per min and 41 to 49 mL/kg per min, respectively, although the differences between groups were not significant. To improve mental health and reduce stress

No
Mindfulness-based stress reduction (MBSR) programs consist of three components: didactic material related to mindfulness, meditation, yoga, and the mind-body connection; experiential practice of various mindfulness meditations, mindful yoga, and the "body scan" during group meetings and encouragement of home practice; group discussion focused on the application of mindfulness to everyday situations and problem solving related to barriers to effective practice. MBSR participants had less anxiety and a reduction in negative coping approaches (p = 0.06). MBSR participants showed an increase in self-reported anger (p =0.06). Otherwise, there were no significant differences between groups. Among all participants, there was an association between the mindfulness subscale "act with awareness" and lower anxiety (p < 0.01). Among MBSR participants, mindfulness subscales were associated with less self-reported angry temperament (pb0.02) and less anger reactivity (p = 0.05). Total cortisol output was not statistically significantly different between groups at baseline or follow up. Overall, cortisol output was higher post-program (p = 0.05). There was a trend towards increasing cortisol over time among 'healthy-topic' participants (113.6 to 167.5, p = 0.07); but not among MBSR participants (128.3 to 138.5, p = 0.33).Regression analyses of actigraphy data showed no differences between groups in sleep latency (p = 0.29), WASO (p = 0.42), or sleep efficiency (p = 0.97); also, sleep diaries showed no differences in sleep quality (0.67). PHAT participants had significantly higher knowledge scores than FOY (control) participants (p ≤ 0.0001), and the rate of increase in condition was significantly higher among PHAT participants than FOY participants (p ≤ 0.0001). Participants with greater general health knowledge for the centred health knowledge variable had significantly higher intercepts for past week fruit consumption (p ≤ 0.01), past week vegetable consumption (p ≤ 0.0001), past month vegetable consumption (p ≤ 0.0001) past week moderate PA (p ≤ 0.0001), past week PA to strengthen or tone muscles (p ≤ 0.01) lifetime alcohol use (p ≤ 0.01), and lifetime marijuana use (p ≤ 0.05). Participants with lower general health knowledge had higher intercepts for past month alcohol use (p ≤ 0.01) and past month marijuana use (p ≤ 0.0001).Participants with lower general health knowledge scores had significantly greater slopes for lifetime alcohol use (p ≤ 0.01). The growth curves for moderate PA and PA to strengthen and tone muscles were not significantly different between PHAT and FOY participants. The level of engagement of vigorous PA was not significantly different between experimental conditions; however, the rates of increase for participants in PHAT were higher than those in FOY. There were no significant differences in growth curve results between experimental conditions for all past month substance abuse behaviour variables, lifetime alcohol use, and lifetime tobacco use. There were significantly higher rates of increase for PHAT participants in lifetime marijuana use (p < 0.0001). Growth curve modelling indicated that participants in PHAT had significantly more gains in health knowledge than participants in FOY, but the effects on behaviour were modest. There were several barriers identified in relation to the adoption of sleep hygiene interventions, namely reluctance to follow scheduled sleep routines on weekends and concern about "parting" with electronics at bedtime.

Yes
Participants were intrigued by the idea of adopting an app-based sleep intervention but were sceptical that they could successfully adopt sleep hygiene practices and were more interested in making changes on school days than on weekends. The overall feedback on two commercial sleep apps was positive, with a good adherence and engagement rate, and perceived health benefits.

Mental Health and PA
Beaulac et al. [87] To promote psychological, social, and physical well-being The findings suggested that the community-based intervention was a promising program for the promotion of youth psychological, social, and physical well-being. The adolescents, parents, and/or personnel described benefits across seven main areas, including dancing and related skills, behaviours (e.g., reduced television viewing), physical well-being, psychological well-being, relationships, respect for others and for diversity, and school performance.

PA, Diet and Mental Health
Berria et al. [79] To In the crude analysis for the entire sample and among students with adequate BMI at baseline, there was a greater probability of dropping out with increasing age and BMI. Students classified as overweight were more likely to drop out with increasing age. In the adjusted analysis, the association with age remained for the entire sample, including students with adequate BMI and with overweight. In addition, for the overweight students, participation in the intervention during the afternoon period and the higher socioeconomic status were associated with dropping out of the intervention.

Discussion
The objective of this study was to review the evidence for school-based interventions aimed at HL-related areas in socioeconomically disadvantaged adolescents and to identify effective intervention strategies for this population. To the best of our knowledge, this is the first study to review this topic which is not restricted to a specific region of the world. The evidence collected gives insight into the interventions carried out, and in particular, the success of various strategies implemented with this population, which can be used to guide future intervention development.
Forty-one intervention studies were identified, with the majority carried out in the US. This review aimed to identify studies targeting five health domains (PA, diet, mental health, substance abuse and sleep). The majority of interventions targeted PA, diet and mental health, while very few focused on substance abuse and even less on sleep. This is despite the well-known harmful impacts of substance abuse and poor sleeping habits on adolescent's health and well-being [90][91][92], and the evidence supporting the benefits of behavioural interventions in improving sleeping habits [93] and substance abuse [92] in youth. Thus, future studies are needed to assess whether interventions targeting sleep and substance abuse are effective in socioeconomically disadvantaged adolescents.
The identified interventions varied greatly in research designs, aims, components, outcome measures and effects, increasing the difficulty to compare and analyse effectiveness per health domain, or across health domains. In addition, although highly recommended [94], only seventeen of the included interventions measured fidelity. This adds to the difficulty of comparing the quality or effectiveness of an intervention, as implementation fidelity of certain intervention strategies within and between studies may have varied greatly. Future studies should include fidelity or process evaluation evidence to allow for a better understanding of the implementation process and to determine whether disappointing results may have been related to poor program delivery [95]. Furthermore, there was a high variance in the risk of bias scores obtained, with numerous interventions scoring as 'high risk' or 'unclear' on multiple aspects of their study designs, adding to the difficulty of interpreting intervention outcomes. Nevertheless, numerous successful intervention strategies were identified. These include 'hands-on' or practical learning, peer support and adopting a holistic intervention approach.

Effective Intervention Strategies
The successful implementation of 'hands-on' or practical learning intervention strategies in the studies identified within this review [53,54,63] aligns with previous research that states educational activities which are carried out through interactive tasks and are focused on context specific learning, may improve health-related decision making and motivate adolescents' to improve behaviour [24,96]. An example of this is 'LifeLab' in Southampton; an innovative 'hands-on' science-based approach targeting adolescents' HL through scientific knowledge and lifestyle behaviours [97].
Interventions identified in this review have demonstrated the effectiveness of peer support in adolescent health interventions [63,64,66]. During adolescence, peer relationships begin to develop, and these relationships are reported to have positive or negative influences on health [98]. Connections with supportive and prosocial peers can lead to healthier behaviours and reduced likelihood of risky behaviours [99]. In addition, peer modelling and awareness of peer norms can be protective against behaviours such as sexual risk [100] and substance abuse [101]. The findings from this review have further underlined the benefits of peer support by using adolescent peers to design and deliver intervention components [63,64,66].
This review has provided further evidence for the benefits of adopting a holistic approach to school-based health interventions, rather than just targeting on the curriculum element. The WHO HPS framework suggests that school-based interventions should adopt an eco-holistic approach by targeting three key areas-the formal health curriculum, the school environment, and engagement with families or communities (or both) [34]. While most of the interventions within this review targeted the formal school curriculum dimension of the framework, less were concerned with targeting all three critical elements. The benefit of adopting an intervention framework which targets all three areas is demonstrated by Hollis et al. [75]. Although only one intervention was successful in incorporating all components of the HPS framework [34], others which have targeted critical elements, such as parent or community engagement or the school environment [56,62,69,74], appear to foster positive improvements in the health and well-being of low socioeconomic adolescents.

Implications of the Identified Interventions Strategies for Socioeconomically Disadvantaged Adolescents
As previously mentioned, socioeconomically disadvantaged populations tend to benefit less from health interventions than those from more affluent backgrounds, often resulting in greater health disparities [40]. Yet, to date, there has been limited evidence to inform the design of health interventions for this population [102]. This review, therefore, has attempted to add to the evidence base by identifying effective and attractive intervention strategies which have been shown to work specifically in this socioeconomically disadvantaged adolescent cohort.
Intervention strategies that involve students carrying out practical activities appear to be more effective than those involving didactic learning. This has been reported elsewhere in lifestyle behavioural interventions targeting adolescents from socioeconomically disadvantaged populations, where students have reported enjoying 'getting away' rather than learning from the classroom [103]. In addition, providing such 'hands-on' activities which are financially inexpensive is crucial when designing intervention strategies for this population, as cost has been reported as being a considerable barrier for participation [70]. The strategies implemented in successful interventions included in this review, such as mindfulness activities [53], yoga [83] and performing "healthy" skits [53], are very cost effective and are therefore suitable options.
Peer learning, in which peers design or deliver (or both) the intervention, appears to be an effective strategy in this cohort. This approach is said to give the students a sense of leadership and empower them to change their lifestyle [104]. Furthermore, including the adolescents in the design and implementation of the intervention is crucial, particularly in a socioeconomically disadvantaged populations, as it ensures the voice of the student is central to the design, which in turn maximises the potential for the intervention to meet the preferences and needs of the population [105]. The WHO strongly advocates the inclusion of peers in the design and implementation of health interventions in socioeconomically disadvantaged populations, as they are trusted by the participants, it leads to greater acceptability and it provides the adolescents with a sense of ownership [106].
Adopting interventions that consist of more than just a school-based educational element appears to be critical in socioeconomically disadvantaged populations. In particular, involving a parents in the intervention as a strategy is important, as parents play a vital role in the lifestyle behaviours of their children [107]. In addition, reaching a range of settings in which youth spend most of their time increases the likelihood of long-term intervention effects [108]. In a review study, the importance of engaging the parents of socioeconomically disadvantaged adolescents in lifestyle behaviour interventions has been highlighted [109], with another study, which failed to change dietary behaviours, stating that the adolescents felt that this was due to a lack of support from their parents [103]. Furthermore, adolescents who perceive their parents to be leading a healthy lifestyle are more likely to also partake in healthy behaviours [110]. Fostering this sort of modelling and parental support through engaging parents in school-based health interventions appears to be a viable solution to improve the impact of interventions in this population.

The Intervention Strategies and Health Literacy
If HL is understood as an observable set of skills, intervention efforts should focus on improving an individual's skills and capacities [111]. Despite this, previous reviews have indicated that existing HL research has been driven by health concerns, which potentially underplays the development of educational outcomes, such as critical thinking, the development of capabilities and motivation for behaviour change [34,112]. Nevertheless, the current review identifies a number of intervention strategies potentially useful for influencing HL. The classifications of HL provided by Nutbeam (2000) may be a useful way to interpret the strategies identified in the current review [113]. Considering interactive HL, the ability to extract health information, apply new information in changing circumstances and engage with others to make decisions, strategies that include practical learning activities and interaction with peers may be of particular use. Many studies included providing the opportunity for shared decision making as part of the intervention [80][81][82]. Knapp et al. [60] held interactive classes outside of the classroom (in the kitchen and garden). Vicary et al. [68] used a life skills training programme to improve the adolescent's confidence and capabilities, specifically targeting decision making skills, refusal skills, media awareness and resistance skills in relation to substance abuse. Including the social support network in an intervention may also be of benefit to increase motivation to change behaviour [56,62,74]. Strategies relating to critical HL, the ability to critically analyse information from a wide range of sources, were less prevalent. Only one study [78] attempted to target critical HL by incorporating discussions around a critical approach to the use of supplements in PA. However, many studies used pedagogical techniques and educational resources that could be structured around developing critical HL. For example, discussions [56,59], problem solving [82], role play [52,56], provision of resources [76,78,80], and the development of resilience [72] were all used when targeting health behaviour change. Such components could be easily delivered through practical activities or involve peer interaction to enhance learning. The use of smaller sessions, with groups of similar individuals, to target consciousness raising and self-re-evaluation, with regular feedback could be offered to improve basic functional HL behaviours, which enables individuals to function effectively in everyday situations, as this can improve the motivation and maintenance of behaviour change [66]. Other strategies to consider to improve HL in school aged children may be teacher training [75,76], as existing HL research has suggested pedagogical guidance will be needed to deliver HL informed curricula [14,18,112,114,115], and the holistic approach (previously outlined in this review), which may be a way to incorporate HL as part of the wider HPS framework [14,34,56,62,69,74,75].
It should also be acknowledged that strategies that lead to health behaviour change may or may not lead to improvements in HL and vice versa. One such example of this may be that even in a person considered to have high level of observable HL skills, they may experience real challenges in applying those skills in an unfamiliar environment [111]. This should be considered when interpreting the findings of the current study, which found that none of the included interventions explicitly targeted HL, nor did any use HL as an outcome measure. Despite this, as indicated earlier, strategies may be useful and transfer across HL, health promotion and health behaviours, and future research exploring the transferability of these skills, and the relationship between these connected fields is warranted. A factor to consider in the dearth of school-based HL interventions aimed at socioeconomically disadvantaged populations may be the lack of an appropriate tool to measure HL in this context, with interventions to date using adult-adapted HL tools which some have argued to be inaccurate [116]. As a result, it may have been problematic to place HL as a primary outcome of an intervention. Recently, a tool has been developed specifically for use in adolescent populations, but its applicability is yet to be tested [117]. Furthermore, another HL measurement tool is currently in development, which is following a rigorous co-design process with young people to understand the context and needs of adolescent [118]. The development of valid and sensitive measurement tools can be utilised to guide intervention designs to target the specific needs of a particular population and allow for the potential effects to be easily and accurately tracked.

Applying Effective Intervention Strategies
As adolescents encounter unique health issues based on their level of puberty, development, social environment and social context [22], caution should be exercised when using previous interventions to inform future approaches. 'Ready-made' or 'one-size-fitsall' health promotion approaches have become increasingly popular [119], yet evidence suggests that interventions developed outside of the targeted schools are rarely meaningful or effective [120]. For health interventions, and in particular HL interventions, to ensure that the health needs and priorities are met, it is recommended that they should be co-produced with all relevant stakeholders [115]. This approach values the knowledge and input of key personnel to facilitate deeper engagement in intervention content, and to ensure that interventions are contextual, sustainable, and equity driven; all of which have been demonstrated in previous HL intervention development approaches [121,122]. Based on this, effective intervention strategies identified within this review can be used to guide the development of future interventions but should always be contextualised and tailored to suit the targeted population.

Strengths and Limitations
To our knowledge, this is the first study to systematically review school-based HLrelated interventions in socioeconomically disadvantaged adolescents. The study provides evidence on interventions carried out globally and was not aimed at a specific region of the world. It must be noted that due to the contentiousness of defining socioeconomically disadvantaged populations, studies which self-identified as socioeconomically disadvantaged (or the equivalent) were included. It is acknowledged that included studies would have used different methods to define socioeconomically disadvantaged populations and, therefore, this may be a limitation when comparing results of studies within this review. The current review used a standardised and well-established appraisal tool for the assessment of study quality to enable comparison between studies. The Cochrane Risk of Bias Tool, however, was originally intended to appraise RCTs [46] and many of the studies included in this review scored poorly or did not report information related to certain aspects of study bias. As a result, the variation between study quality indicates that caution should be exercised when interpreting the results of intervention effects. Due to the large volume of peer-reviewed publications identified and screened in this review, and the practicalities of running and managing this review, the authors did not include a grey literature search. It is acknowledged that by excluding non-published articles, there is an increased risk of publication bias and the possibility of missing out on additional information that a grey literature search may have provided. This review included interventions reporting on outcome and process evaluation, yet many interventions did not report on the process evaluation or intervention fidelity. Future research should consider comprehensively reporting intervention evaluation to provide a deeper insight into the intervention delivery.

Conclusions
This systematic review provides evidence on the interventions implemented which aimed to improve HL-related areas in socioeconomically disadvantaged adolescents. This review highlights the lack of interventions targeting sleeping habits and substance abuse in this demographic. In addition, no interventions have explicitly aimed to improve HL. Nevertheless, successful interventions strategies were identified that could be used to inform future intervention development. These include the integration of practical-based learning activities and the use of peer educators. Furthermore, evidence supported linking the intervention to the parents and local community.  Acknowledgments: The author team would like to thank colleagues in the Irish Heart Foundation (Janis Morrissey and Laura Hickey) and University College Dublin (Dr Celine Murrin and Dr Ailbhe Spillane) who contributed to the refinement of the final search criteria used in this review, as part of a working group on a larger Irish Heart Foundation Schools Health Literacy Demonstration project.

Conflicts of Interest:
The authors declare no conflict of interest. Table   Inclusion Criteria Exclusion Criteria

Population
Adolescents with a reported mean age between 12 and 16 years from a socioeconomically disadvantaged (or equivalent) background.
Mean age not between 12 and 16 or special populations (e.g., children with learning difficulties, pregnant adolescents, exclusively obese individuals, or those with a specific health condition).

Intervention
The implementation of an intervention related to health literacy (increases health knowledge, understanding, awareness, motivation, confidence) in at least one of the following areas physical activity, sedentary behaviour, dietary habits, sleeping habits, mental health or substance abuse.
The intervention did not include an educational element or a component targeting health literacy (increases health knowledge, understanding, awareness, motivation, confidence).

Context
School-based interventions, interventions that could be feasibly implemented in a school setting, or interventions that could be linked to a school curriculum.
Book chapters, case studies, student dissertations, conference abstracts, review articles, meta-analyses, editorials, protocol papers or systematic reviews.

Outcome
Aimed to increase health knowledge/comprehension, understanding, behaviour, value, well-being, motivation, self-efficacy or self-monitoring in relation any of the following domains: physical activity, sedentary behaviour, dietary habits, sleeping habits, mental health or substance abuse.
Outside of the targeted health domains.

Study Design
Published in English and in a peer-reviewed journal.
Systematic review, meta-analysis or full text not available.