Effectiveness of intervention programs in schools to reduce health risk factors in adolescents: a systematic review

The aim of this study was to systematically review studies on intervention programs involving physical activity promotion and/or nutritional education to reduce health-related risk factors (overweight/obesity and metabolic profilealterations) in Brazilian adolescent students. A search was performed in the following electronic databases: Medline (PubMed), Lilacs, Embase, Scielo and Capes Thesis Database. A total of 1,568 studies were identified, of which 21 full papers were analyzed and 5 were included in the systematic review. The majority of the studies involved physical activity practice combined with nutritional education in Brazilian adolescent students and all interventions showed positive effect on reducing health-related risk factors among Brazilian adolescents.It was concluded that all studies included in this review showed positive alterations in healthrelated risk factors after interventions, reinforcing the importance of intervention programs to promote a healthier lifestyle and reduce health-related risk factors in adolescents.


INTRODUCTION
Population studies have demonstrated an increased prevalence of obesity among adolescents [1][2][3][4][5] , as well as the early appearance of a series of risk factors for the development of cardiovascular and metabolic diseases such as type-2 diabetes, hypertension and dyslipidemia 3,6,7 .Studies conducted in Brazil also showed the same results in the last decade, a high percentage of individuals in adolescence presenting several health risk factors such as abdominal obesity (32%), hypertension (15%), metabolic syndrome (7, 7%) [8][9][10] .
Childhood and adolescence are important stages for the development of a healthy lifestyle, since most habits acquired during this period of life are perpetuated into adulthood 11 .However, decreased levels of physical activity and increased sedentary behavior have been observed at this stage, both national and internationally [12][13][14][15] .Acquiring healthy habits in this age group should be a priority for the educational and social sectors, particularly schools, because they are privileged spaces for the development of interventions towards health, since they gather most adolescents in the country, which are over 30 million students enrolled in the Elementary and High-School public school system 16 .
There is evidence that physical activity interventions performed in schools are capable of promoting increased levels of physical activity and combat sedentary behavior 11 , but the beneficial effects of these actions on cardiometabolic risk factors remain inconsistent.Recent review studies have shown that interventions in schools in different countries, with different models of physical activity practices were effective in reducing cardiometabolic risk factors 17,18 , increase the level of physical activity and cardiorespiratory fitness, increase the active time during Physical Education classes and active transport to school 19 , and indicate that these interventions should focus both on promoting physical activity and nutrition education 11 .
In Brazil, few school interventions on physical and / or nutritional activities have been performed in order to reduce health risk factors [20][21][22][23][24] , moreover, little is known about the methodology of the studies, types of intervention, evidence of effects and results on the health of adolescents.Thus, this review study will assist in explaining the differences among studies, being useful for the systematization, planning and implementation of school intervention programs in order to define a favorable health policy in this environment.
In this context, the aim of the study was to conduct a systematic review of intervention programs with planned physical activity and / or nutrition education actions to reduce health-related risk factors (overweight / obesity and metabolic profile alterations) in Brazilian adolescent students aged 10-19 years.

METHODOlOgICAl PROCEDURES
A systematic review was conducted in August and September 2014 that aimed to identify articles published in the last 10 years in order to highlight the scope of the most recent studies in Medline (PubMed), Lilacs, Embase, Scieloelectronic databases and also the portal of dissertations and theses of the Coordination for the Improvement of Higher Education Personnel -CAPES (http://capesdw.capes.gov.br/capesdw/), using keywords and combinations in Portuguese and English with Boolean operators "and" and"or": Intervention, Students, Adolescents, Physical Activity, Sedentary Lifestyle, Sedentary Behavior, Eating Behavior, Nutrition, Risk Factors, Overweight and Obesity.A search through the reference lists of relevant studies and systematic reviews that addressed the topic of interest was also conducted.Consultation to journals to ensure that there was peer review and conference of methods described by the authors through the search and reading of articles to verify their methodological quality were used as strategies to reduce bias.
The following inclusion criteria for the selection of studies were used: (i) studies published in journals, dissertations and theses available in selected databases, as well as in the references of the selected studies; (ii) samples that include adolescents aged 10-19 years; and (iii) programmed physical activity and / or nutrition interventions conducted in public and / or private schools in order to reduce health risk factors (overweight / obesity and metabolic profile alterations) regardless of duration; (iv) interventions conducted in Brazil; (v) articles containing pre-and post-test data.
The review study excluded articles with sample of adolescents in specific health conditions (hypertension, diabetes, hyperlipidemia or other chronic diseases); sample with children (<10 years) or adults only (> 19 years); interventions performed totally outside the school environment (clubs, communities, clinics, hospitals, laboratories) or actions involving families or community; interventions carried out only with teachers / school staff; articles with exclusive description of the methodological design of intervention programs and / or transversal studies; Brazilian articles with international data; systematic review articles or meta-analysis.
After the search for studies in electronic databases, the selection started from the analysis of titles and abstracts by two independent reviewers in accordance with the inclusion and exclusion criteria.In the case of lack of information in the abstract, the study was evaluated by the full text.The results were crossed to verify agreements, and disagreements were resolved by consensus and with the presence of a third appraiser, and at the end of the search, and data extracted were checked.The assessment by reviewers was unmasked as for the authors and the study results.
The methodological quality of studies selected was evaluated by the scale proposed by Downs & Black 25 .This scale aims to evaluate both studies that have randomized clinical trial design and those who are not random, including five sub-items related to: the way results are reported (if the information presented in the study allows the reader to interpret data and results without bias), external validity, biases, confounding factors and study power.The maximum score to be achieved through the 27 items is 31 points 25 .From this, the evaluators classified the articles with scores greater than 20 points as high methodological quality, 10 to 20 points, moderate, and below 10 points low methodological quality.
The articles selected are shown in the tables in increasing order of year of publication and, when in the same year, in alphabetical order considering the first author.This study followed the PRISMA recommendations for systematic reviews 26 .

RESUlTS
Figure 1 shows the flowchart of the study selection process.Overall, 1,568 studies were found and of these, 1,515 were excluded after thorough analysis of titles and then 32 by reading the abstracts for not meeting the inclusion criteria.After the first stage, 21 original articles were selected for reading the full text, of these, 16 were excluded for not meeting the inclusion criteria.A search in the reference list of articles read in full was also performed and 1 article was selected.Finally, five studies were chosen for the quality verification step according to the scale proposed by Downs & Black 25 .Two studies were conducted in the southeastern 21,22 , one in the midwestern 23 , one in the northern 20 in one in the southern regions of Brazil 24 .All interventions were conducted with boys and girls at the same time, and the age group ranged from 7 to 17 years, the sample size from 34 to 416 students and the duration of interventions from 10 weeks to 1 academic year.It was observed that all interventions, with the exception of only one study 20 occurred in public schools.
As for the sample selection, all studies [20][21][22][23][24] showed selection of participants for convenience.It was found that all studies used body mass index (BMI) as the dependent variable, as well as some skin folds to estimate body fat percentage (% BF) 20,21,24 , anthropometric girth 20,21,23 , body composition through bioelectrical impedance (BIA) 22 , blood variables 21,23 , blood pressure measurement 23 and physical fitness control 23,24 (Table 1).Some results should be highlighted such as the significant body fat reduction observed in all studies, positive changes in blood parameters 21,23 , improvement in physical abilities 23,24 and blood pressure 23 .The effect size was calculated from the results obtained after the intervention programs of the main variables that showed significant pre-and post-test differences according to the equation of Cohen 27 .According to Cohen 27 , values greater than or equal to 0.8 represent large effect size; values between 0.8 to 0.2 are considered intermediate effect size and less than 0.2 small effect size, and positive values correspond to a higher average in the intervention group and negative values to a higher average in the control group (Table 1).When observing the aims, all studies focused on the effects of programmed physical activity and / or nutrition education in school about health risk factors such as changes in body composition [20][21][22][23][24] and / or metabolic profile 21,23 in adolescents.
After assessing the methodological quality of studies, it was found that only one article 23 had excellent methodological quality and the others had moderate.The items that most studies have not achieved refer to the sample randomization or the intervention groups, as well as the control of confounding variables 25 .

DISCUSSION
The first fact that drew attention in this systematic review was the limited number of available studies who developed intervention programs in Brazil promoting physical activity and / or nutrition education aiming to reduce health-related risk factors among adolescent students [20][21][22][23][24] .Another highlight is the occurrence of positive effects in all studies found, regardless of type of intervention (physical activity, nutrition or both), duration (10 weeks to 1 academic year) or nutritional status of participants (overweight / obesity or normal weight; overweight and obese included in the same sample).
The inclusion criteria of this study were selected due to the high prevalence of overweight and obesity, as well as to the poor health-related habits in this phase, as we believe that this phase is of utmost importance for the development of a healthy lifestyle.Thus, there are gaps in this area of study, making it possible for future investigations using a reliable methodology using methods of better accuracy and validity.
Improvement in the body composition of participants was observed after intervention [20][21][22][23][24] ; however, all studies used BMI or methods for assessing body composition considered doubly indirect (developed from an indirect assessment method) as skinfold thickness 20,[22][23][24] or bioelectrical impedance 21 .Despite being correlated with cardiometabolic risk factors in adolescents 28,29 and easy to be used in the field with a high number of individuals, as in the school environment, these methods may have some limitations related to the lack of accuracy, especially in individuals with higher body fat 30 .
Among the interventions protocols applied, three studies involved controlled exercise, using heart rate as a parameter for the monitoring physical exercise intensity 20,22,23 , and this variable is considered valid as an indicator of physical activity intensity in adolescents 31 .Most of them added education programs 20,23,24 to physical activity intervention nutrition and only one study used nutritional intervention and also demonstrated positive changes in body proportions and effectiveness in reducing BMI 21 , as previously observed in a systematic review 32 , where most of the studies showed a decrease in the prevalence of overweight and obesity among participants.
Analyzing the intervention period of the selected studies, it was observed that even programs with shorter duration (12 and 10 weeks) 20,23 demonstrated positive changes in variables.Sun et al. 18 claim that higher intervention periods provide greater health benefits and a high adhesion to physical activity practices in the daily lives of adolescents; in the studies analyzed, it was not possible to verify the influence of intervention duration on the health risk factors of students, because based on the findings, it was found that the effectiveness of interventions was independent of the duration of interventions.
The lack of control of the level of physical activity and nutrition intake of participants is also a point to be considered after the evaluation of the quality of studies 25 .Only two studies evaluated physical activity and eating habits, one of them using previously validated questionnaires and classifying adolescents by TEM / week (considering the resting metabolic equivalent of 3.5 mL • kg -1 • min -1 ) and the frequency of the consumption of different types of food 23 , and the other through a recall of time spent in physical and sedentary activities, classifying them according to the intensity of reported activity 24 .
Physical activity questionnaires and recalls are subjective instruments well accepted and widely used in population studies, and generally present good correlation with objective physical activity measures such as accelerometers, being valid to estimate the level of physical activity in adolescents 33 .Furthermore, the lack of criterion-methods to measure physical activity in adolescents can be considered a limitation of these studies, since self-reports can present significant bias associated with the outcomes [34][35][36] .
Another important aspect to be considered is the fact that the changes do not always occur in a sensitive way due to metabolic adaptations during the intervention process and mainly due to growth changes and body composition of the growth spurt and sexual maturation 22 , and most interventions included in this review did not control this variable.Therefore, some of the positive results presented by intervention programs can be related to body changes.Another fact observed in most studies was that adolescent males and females are equally compared within groups, but there are differences between them at the end of puberty such as the fact that girls present proportionately twice as much fat than boys, which may cause bias the interpretation of the reported findings 37 .Similarly to the inclusion of subjects with different nutritional status in the same group, differences in body composition after interventions tend to be higher in overweight and obese adolescents compared to those with normal weight 18 .
Furthermore, the limitations of these studies were not clearly described by the authors, except by Silva et al. 24 , who reported that the design adopted by the authors (quasi-experimental and sample not randomized for convenience) was chosen due to operability reasons and to the presence of complex programs with extracurricular activities twice a week, making it impractical to move participants from the intervention group to places far from their homes to participate in activities.The low score obtained by studies in item related to the intervention design in the assessment of the methodological quality should also be emphasized.
The inclusion of a control group can minimize this limitation, ensuring similarities among gender, age and nutritional status; however, there was a study 20 that did not include it, dividing participants only into different intervention groups and presenting a very small number of individuals in the sample.Another limitation is the lack of generalization of results to populations with similar characteristics, because none of them is characterized as randomized clinical trial, considered the gold standard for interventions, and so little representative to the population in question.
The results presented here can assist in the planning of future intervention studies.Thus, even considering that some studies have shown medium and small effects, it could be inferred that the regular practice of physical activity along with nutrition education contributes to the improvement of health-related risk factors in adolescents.In addition, it is essential to consider that school is critical in the formation and actuation of people in all fields of social life, playing a decisive role in the development of students, the perception and construction of citizenship and access to public policies.Thus, school is the best place to carry out health promotion activities for children, adolescents and young adults 38,39 .

FINAl COMMENTS
The increased incidence of obesity and the early onset of a number of risk factors for the development of cardiometabolic diseases in adolescents are stimulating factors for carrying out actions to improve health in this population, but there are few intervention studies in Brazil that are often not properly documented and evaluated.
Thus, the few studies involving students and methodological differences between them, as well as the observed biases, made it difficult to assess the effectiveness of these actions.However all studies in this review showed positive changes after intervention, even if some are not statistically significant, reinforcing the importance of intervention programs to promote a healthier lifestyle and reduce health-related risk factors in adolescents.
It is believed that the results and criticisms presented in this systematic review can assist in improving the methodological quality of intervention programs aimed at students, enriching the scientific evidence for improving the quality of life and health in adolescents.

Table 1 .
Characteristics of the intervention studies included in the systematic review.

Table 2 .
Characteristics of intervention programs IG: Intervention Group; CG: control group; N / A: Not applicable; P.A. .: Physical Activity; Nutri: Nutrition.PE: Physical Education; HR: heart rate; ': Symbol for minutes.˟ Own intervention model -developed by the author (s).