Prevalence of Brazilian children and youth who meet health criteria for cardiorespiratory fitness : systematic review

The study aimed to systematically review the literature to identify the prevalence of Brazilian children and adolescents that meet health criteria for cardiorespiratory fitness (CRF). The search was performed in the electronic databases (PubMED, SciELO, LILACS, Scopus, SPORTDiscus and Web of Science) and list of references of identified articles. Inclusion criteria were: population composed of children and/or adolescents (0-19 years or average age up to 19 years); studies with Brazilian children and/ or adolescents and observational studies with cross-sectional or longitudinal design. The process of analysis of studies involved reading titles, abstracts and full texts. After these phases, 60 articles were eligible. Regarding the design, two studies were longitudinal. Of the total studies included, 49 presented moderate methodological quality values and 11 presented low methodological quality values, according to National Heart, Lung, and Blood Institute instrument. The prevalence of children and adolescents who met CRF health criteria varied among studies (7.5% to 70.4%), and this variation was higher in girls than in boys. Total of 49.093 individuals aged 6-19 years were surveyed to compute all of these studies, resulting in 32.2% of subjects (27.7% of boys and 28.4% of girls) with CRF levels adequate for health around Brazil. These results may help to screen the pediatric population at risk if CRF is considered as a health measure.


INTRODUCTION
Physical fitness is a characteristic of a multifactorial nature, usually understood as the ability to perform daily movements and tasks that involve physical-motor efforts without showing excessive fatigue 1 .The distinctive aspect of physical fitness is the multidimensional expression that includes cardiorespiratory, morphological, motor, metabolic and muscular dimensions 1 .Cardiorespiratory fitness (CRF) has been one of the most studied components in the universe of physical fitness [2][3][4] .Defined as the ability to provide oxygen to muscles and to use it to generate energy during physical exercise, the importance of CRF is highlighted in the strong relationship with performance in sports modalities and in association with a more positive health profile 1,2,5 .
Epidemiologically, CRF has gained increasing importance in the expression of population health, since it provides a good indication of the ability of various organs and body systems involved in the movement, being therefore considered an important health marker in the pediatric context 2,6 .It has been reported that moderate to high CRF levels in children and young adults are associated with lower incidence of cardiometabolic risk factors, improved mental and bone health, and academic performance 6 .Considering the period of childhood and adolescence as critical phases for the development of CRF 6,7 and potentially sensitive to the influence of environmental factors, the evaluation and monitoring of CRF levels can significantly contribute to the identification of children and adolescents with unsatisfactory levels, predict future behaviors and states, and for the promotion of health throughout life 2,6,7 .
CRF can be evaluated through different laboratory and field tests and interpreted in normative and criterial terms 2,6 .The normative reference assigns meaning to individual performance in differential terms, since it is possible to determine percentage performance ranks by comparing pairs of the same age and sex and identifying high, medium and low performance; however, without a direct link with health outcomes 2 .The criterion assessment allows the description in individual terms relative to a criterion measure that identifies the minimum amount of CRF required for good health (healthy cardiometabolic profile) 2,3,8 .The interpretation referenced by criterion represents a way to identify the percentage of children and young people in the healthy zone or not regarding CRF.This specification was constructed differently in each of the batteries and physical-motor tests, with different cut-points emerging in literature, which may lead to some confusion regarding the cardiorespiratory health status of children and adolescents 2,3,8 .
Regardless of evaluation (normative or criterial), literature has revealed worrying data regarding the CRF levels of children and adolescents.Tomkinson et al. 4 , in a secular trend study with children and adolescents from 19 countries, showed a moderate decline in CRF equivalent to 7.3% over the study period (1981 to 2014).In a meta-analysis 5 , the CRF of children and adolescents from 50 countries was investigated, concluding that Brazil was among the countries with the worst cardiorespiratory performance (42 nd position).In this scenario, studies carried out in Brazil have shown that approximately 80% of children at school ages had CRF below that established for health 9,10 and that this condition has not changed in the last decade (1999 to 2010) 10 .
Through the above-mentioned evidence, as well as the undeniable importance of CRF as a health marker, the international literature shows an effort to systematize the cardiorespiratory performance of children and adolescents 2,5,6 .In Brazil, although there is information associated with CRF behavior in the pediatric population from different cities and contexts, these results are not systematized, and there is no national level knowledge.In this sense, the present study aims to systematically review the literature to identify the prevalence of Brazilian children and adolescents who meet the health criteria regarding CRF.The relevance of this analytical and interpretative effort has epidemiological and public health impact, since, in addition to quantifying children and adolescents with adequate or inadequate CRF levels, it also contributes to: (1) identify Brazilian sites and regions of higher and lower prevalence in the healthy CRF zone and thus to recognize places with greater needs of public policies for health promotion; (2) understand how the relationship between CRF and health has been studied in pediatric populations in Brazil; and (3) identify which tests are used and their cut-points.

METHODOLOGICAL PROCEDURES
The method of this systematic review was consistent with the PRISMA statement 11 .The systematic search was performed on PubMED, SciELO, LILACS, Scopus, SPORTDiscus and Web of Science databases.For the search, three blocks of descriptors were developed: Block 1: "physical fitness", "aerobic capacity", "aerobic fitness", "cardiorespiratory capacity", "cardiovascular fitness", "aerobic power", "aerobic endurance", "cardiorespiratory endurance", "oxygen consumption", "maximum oxygen consumption", "maximal oxygen uptake", "VO2 maximal"; Block 2: child *, schoolchildren, adolescent *, student; Block 3: Brazil *.Within each block, the "OR" Boolean operator was used and between blocks, the "AND" operator was used.The "prevalence" indicators were not included in the search to reduce possible losses of relevant publications.Parentheses were used to combine search terms by outcome, exposure and population categories.Quotation marks were used to search for exact terms or expressions.Asterisk was used to search for all words derived from the same prefix.Automatic database filters were not used.The search process was finalized in February 2018, which is considered the final observation period for all databases.
The results of the search in each database were exported to the End-Note® reference manager software version X7 (Thomson Reuters, New York, USA).

Inclusion criteria
The inclusion criteria were: a) population consisted of children and/or adolescents (age 0-19 years or with mean age of up to 19 years) according to the World Health Organization 12 ; (b) studies carried out with sample of children and/or adolescents of Brazilian nationality; (c) observational studies (with cross-sectional or longitudinal design).

Exclusion Criteria
The exclusion criteria were: (a) athletes (members of competitive sports teams); (b) children and/or adolescents with special needs (with diagnosis of acute or chronic diseases, physical or intellectual disability); (c) theses, dissertations, monographs, abstracts, book chapters, points of view and review articles, validation and/or reproducibility articles, articles of cut-point determination; (d) articles that did not present numerical data, classifying individuals according to CRF (quantity or prevalence of fit and/or unfit, regardless of parameter considered).

Procedures
The search, extraction and reading of articles were carried out by two independent evaluators (CASAJ and HEGN).If there was no consensus among peers, a third evaluator (DASS) was required to resolve the disagreement.Data extracted were the authors' name/year, methodological quality score, study site, age group, population and sample, study design, stratification, test used to measure CRF, cut-point used to classify CRF and prevalence/ quantity of eligible individuals in relation to CRF.
To evaluate the methodological quality of studies, the instrument proposed by the National Heart, Lung, and Blood Institute 13 was used.This instrument evaluates 14 criteria of the internal validity of studies, and the higher the risk of bias, the lower the score referring to the methodological quality considered for the study.The following are among the criteria analyzed: clarity of the study's objective, definition, selection and participation of the study population, definition, selection and participation of the study population, definition and evaluation of the study variables, study period, and statistical analysis.Each question is scored with "0" or "1", with "0" applied to questions answered with "no" and "1" for those answered with "yes" or "not applicable".The "not applicable" option was used when it was not possible to evaluate one of the instrument criteria due to the type of study (as in the case of the cross-sectional design).The total score was obtained by summing the score of each question, and studies with values of 13 and 14 are considered with good methodological quality, from 9 to 12, of reasonable quality and values below nine are considered of low quality 14 .

RESULTS
A total of 1,106 articles were found; however, 440 were duplicates, resulting in 666 articles.After reading the titles and abstracts, 502 studies were excluded, and then 164 articles were read in their entirety.Of these, 44 studies were included because they presented the eligibility criteria and, after reading the references, 16 studies were included, totaling 60 studies in the present review (Figure 1).The characteristics of the 60 studies included in the review were presented in Table 1.The population evaluated covered the age group from 6 to 19 years.Studies were published between 2005 and 2017.Twelve studies were carried out in the southeastern region, 6 in the northeastern, 39 in the southern, 1 in the northern region and two in all regions of the country (northern, northeastern, southern, southeastern and midwestern).Of these studies, two had longitudinal design 15,16 , the others presented cross-sectional design.Regarding the aims of studies, 23 had as main aim to estimate the prevalence of CRF in children or adolescents who met (or did not meet) the health criteria 9,10, .
The methodological quality of studies was presented in Table 2. Of included studies, 49 presented moderate methodological quality values and 11 presented low values 16,17,22,23,28,31,35,36,[38][39][40] .Among studies that presented low methodological quality, questions 6, 7 and 14 were not answered in any of them, and nine studies did not answer questions 3 and 5 and eight did not answer questions 2 and 4.

Cross-sectional
To examine the independent and combined influence of cardiorespiratory fitness (CRF), body mass index (BMI) and percentage of fat (% fat) on total cholesterol (TC) and blood pressure (BP) in male and female youth.

Longitudinal
To verify the four-year incidence of overweight (Ow) and obesity (Ob) and cardiorespiratory fitness (CRF) on levels in a sample of Brazilian children and adolescents and to identify the associated socio-demographic and nutritional characteristics of these subjects' parents.

Longitudinal
To check the influence of the holidays in physical fitness of schoolchildren.

Cross-sectional
To examine the association between classifications obtained with the CRF and physical activity level in adolescents, as well as the agreement between tertiles and z-score distribution of the variables generated with these methods (distance covered and total physical activity score).NA: Not applicable.Q1: Was the research question or objective in this study clearly stated?; Q2: Was the study population clearly specified and defined?; Q3: Was the participation rate of eligible persons at least 50%?; Q4: Were all the subjects selected or recruited from the same or similar populations (including the same time period)?Were inclusion and exclusion criteria for being in the study prespecified and applied uniformly to all participants?;Q5: Was a sample size justification, power description, or variance and effect estimates provided?;Q6: For the analyses in this study, were the exposures of interest measured prior to the outcome(s) being measured?;Q7: Was the time frame sufficient so that one could reasonably expect to see an association between exposure and outcome if it existed?; Q8: For exposures that can vary in amount or level, did the study examine different levels of the exposure as related to the outcome (e.g., categories of exposure, or exposure measured as a continuous variable)?;Q9: Were the exposure measures (independent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?;Q10: Were the exposures assessed more than once over time?; Q11: Were the outcome measures (dependent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?;Q12: Were the outcome assessors blinded to the exposure status of participants?; Q13: Was loss to follow-up after baseline 20% or less?; Q14: Were key potential confounding variables measured and adjusted statistically for their impact on the relationship between exposures and outcomes?
Studies verified the CRF of children and/or adolescents by different tests: 6-minute walk/run test, 9-minute walk/run test, treadmill ergospirometry, mCAFT test and 20-meter shuttle-run test.Different … continue cut-points were found in the reviewed literature, with 17 studies using the FITNESSGRAM classification 41,42,43,44,45 , 14 studies used PROESP-BR 46,47,48,49,50,51 , 11 studies used AAHPERD 52,53,54 , eight the Physical Best 54 , seven used the CSEP 55 , one study used cut-point proposed by Bergmann et al. 56 , one study the parameter of Rodrigues et al. 57 and one used classification from the sample standards 58 .For one study 39 , it was not possible to identify the cut-point used, since the reference presented by the authors 59 did not report classification information (Table 3).
The prevalence of children and adolescents with adequate CRF varied among studies, and the lowest prevalence found was 7.5% 60,61 and the highest prevalence was 70.4% 23 , both studies carried out in the southern region of the country.The northeastern region showed variation from 13.1% 62 to 54.4% 29 of healthy CRF levels.For the southeastern region, the variation was from 24.1% 63,64 to 64.6% 33,[65][66][67] .

DISCUSSION
This study indicated that, between 2005 and 2017 (period of publications of articles included in this systematic review), the prevalence of children and adolescents in Brazil meeting the health criteria for CRF was 32.2% (the highest prevalence of 70.4% in the city of Lapa, Paraná, and the lowest prevalence of 7.5% in the city of Criciúma, Santa Catarina).These differences in the prevalence of CRF adequate for health can be explained by the variation in the CRF assessment protocols and by the variation in cut-points for CRF classification 91 .
The use of submaximal protocols to estimate VO 2 peak has less precision than maximum protocols 92 ; however, they are more practical to apply in samples with greater number of individuals 93 .Direct methods are more accurate when compared to indirect methods, but in addition to the high cost, there is need for specialized personnel for the application of tests and time spent with each individual, so they are less used 92 , as observed in this review.

… continue
The variability of the different protocols used in studies of this review makes it difficult to compare results, considering that each method used to estimate CRF can produce different results due to different construct validity levels 2,5,8 .Treadmill protocols, for example, result in VO 2 peak approximately 9% higher than cycle ergometer protocols 94 .Furthermore, the same test may have different parameters.The 20-meter shuttle-run test (used by 19 studies of the 60 included) allows the analysis of performance by number of laps, predicted VO 2 peak and CRF stages.The question as to which estimate (VO 2 peak or number of laps) should be used seems to be frequent among researchers 95 .In addition, there are at least five prediction equations available to estimate VO 2 peak using the 20-meter shuttle-run test 2,5,8 .Even when applied to the same collective test, such as the 20-meter shuttle-run test, test performance can be altered by the level of motivation and competitive spirit among students, which may also influence the results 37 .Therefore, cautious interpretation of results is necessary.
It is possible to make evaluative decisions through two types of references: references by norms and references by criterion.This discrepancy in cut-points to classify subjects with adequate or inadequate CRF levels was also found in this review.The norm-based evaluative decision (usually presented in percentage tables) makes it possible to compare the performance of individuals with pairs (same sex and age) 8 .For example, a 12-year-old adolescent who had VO 2 peak estimate of 30 mL/kg/min can be analyzed by means of a percentage table, if that value was higher or lower when compared to other adolescents in the same age group.However, it is not possible to determine from these normative standards if the VO 2 peak value of 30mL/kg/min is considered good or not for a healthy CRF zone.Therefore, evaluation referenced by criterion becomes more appropriate in the approach of indicators directed to health, considering that this type of reference does not aim to indicate individuals regarding the position they occupy in the group, but to the position in relation to the cut-point to ensure better health levels 96 .On the other hand, the norm-referenced evaluation is useful when the purpose is to establish comparisons among individuals by information presented by the group to which they belong, thus allowing a precise identification of the magnitude of inter-subjective differences that may arise 96 .
Among the 60 studies included in this review, seven different cut-points were identified; the most used were references by PROESP-BR 49 (three studies) and FITNESSGRAM 44 criteria (seven studies); and references by PROESP-BR 48 (three studies) and AAHPERD 54 standards (six studies).Systematic review of standards referenced by criteria for CRF identified 10 different patterns for children and adolescents, which produced different percent estimates of healthy CRF 8 .Thus, when analyzing a 17-year-old adolescent who has walked 1,535 meters in the 9-minute walk/run test, for PROESP-BR 49 criteria, this adolescent is in the healthy CRF zone.However, analyzing this same adolescent by the AAHPERD 54 cut-points, there would still be an additional 448.33 meters (approximately 30% more) of the test course for this adolescent to have aerobic performance considered good for health (Box 4).Therefore, when interpreting the results of motor tests using analysis referenced by criterion, it is necessary to take into account that the same value, produced by the same subject at the same time, can receive different judgment if analyzed before another cut-point proposal, which emphasizes the need for standardization of classification criteria.
Despite the fact that the northern region of Brazil is considered the largest region of the country (covering seven states), only one study was developed in this region, and the concentration of research development occurred in the southern region (39 studies).In fact, there is concentration of greater scientific development in the southeastern and southern regions of Brazil and scarcity in the northern region 97 .The southeastern and southern regions of Brazil accounted for more than three-quarters of the total Brazilian scientific production in the period between 2007 and 2009, while the northern and midwestern regions together did not reach 10% of the national total 97 .Regional inequality in scientific production is closely associated with the marked disparities in the distribution of scientific and technological resources 98 .In this context, the southeastern and southern regions of Brazil are favored by the concentration of historically consolidated universities and research institutes 99 and the greater availability of human and financial resources 100 .
When analyzing the prevalence of healthy CRF levels separated by Brazilian regions, it was observed that studies carried out in the northeastern region presented variation from 13.1% to 54.4% 29 .For the southeastern region, the variation was from 24.1% to 64.6%.However, the region that presented the largest variation was the southern region, with the lowest prevalence of 7.5% and the highest of 70.4% of healthy CRF levels.Brazil is a country with continental dimension that has different ethnicities and different socio-cultural and behavioral aspects from one region to another 101 .For example, when analyzing self-reported skin color, 22.6% and 70.3% of adolescents living in the northern region are white and brown/black skin color, respectively, while in the southern region, only 30.8% of adolescents described themselves as brown/black and 64.1% white 102 .When healthy eating habits are analyzed, 70.3% of adolescents in the southeastern region reported eating healthy foods five days or more a week.The prevalence of healthy eating habits for young people in the northern region is almost half (39.3%).In addition, physical activity is practiced on a regular basis (greater than or equal to 150 minutes per week) by approximately 67% of adolescents in the southern region, and for adolescents in the northeastern region, prevalence decreases by almost 10% (54.6%).All these factors (skin color, healthy eating habits and regular practice of physical activity) influence CRF 20,81 .For these reasons, comparisons between the prevalence of CRF levels should be interpreted with caution 2,5,8 .This systematic review identified some gaps in Brazilian studies.Among the 60 included studies, 11 studies had low methodological quality regarding the instrument used 13 .The methodological quality of studies is of great relevance for the systematic review, as it influences the magnitude of results.When quality is not adequate, the results may become untrue, hampering the completion of the review 103 .It was observed that most studies did not use sexual maturation (51 studies) and level of physical activity (40 studies) as control variables in CRF data analyses.These variables directly influence CRF levels in children and adolescents 60 .
This systematic review has some limitations that should be considered.First, the included studies are heterogeneous in terms of age, sample size and cut-points to evaluate CRF.This heterogeneity, together with potential confounders, such as maturational stage and level of physical activity of adolescents, may have partially affected the results.Second, some studies did not present the prevalence of CRF in the total sample or stratified by sex.Therefore, these values were calculated by the authors of these studies, which may have affected the results.
The present review was carried out in a systematic way, with the use of six different databases, including studies that were rigorously reviewed in pairs through a tool to analyze their methodological quality, which allows verifying aspects of internal and external validity.In addition, this research did not stipulate specific period of publication of results, which could suggest publication bias, depending on the period of publication.In addition, this review gathered data on the prevalence of healthy CRF of 49,093 Brazilian participants during the entire phase of childhood and adolescence (zero to 19 years), making it possible to compare Brazilian results with other countries of different cultures and customs, helping researchers of the health area in the screening of the pediatric population at risk if CRF is considered as a health measure.
The results of this study help complement standards referenced by criterion on CRF, especially in children and adolescents, who are likely to have healthy aerobic performance.Together, normative reference standards and standards referenced by criterion could be used in physical education and other health research areas to screen pediatric populations at risk and to compare the aerobic performance of children and adolescents in different settings, different populations, and different times.In addition, repeated measurements in the same population could identify temporal trends of CRF.In fact, time trends can help monitor the impact of policies implemented to improve CRF and the overall health of a given population and can help predict the health of future generations.
On average, three out of ten Brazilian adolescents presented healthy CRF.Six different tests to evaluate CRF and eight different reference criteria were used by studies, and years (and data) of the most commonly used reference criteria were also different.The region with the highest and lowest concentration of studies on the prevalence of CRF was the southern and northern regions, respectively.
The monitoring and standardization of CRF guidelines should be performed to provide a uniform interpretation of CRF among children and adolescents.Consistent and periodic CRF monitoring could help guide and evaluate health promotion policies and interventions aimed at reducing physical inactivity and encouraging regular physical activity and healthy behaviors to improve the general health of the population.The school environment is the most conducive for the promotion of physical activity, considering that the physical education class is the main way to create habits associated with health improvement and maintenance 104 .

Figure 1 -
Figure 1 -Flow chart of search, selection and exclusion of articles.
450; ♀: 513)Cross-sectionalTo identify the behavior of CRF during adolescence and to describe the prevalence of adolescents who attended and did not met the recommended criteria for health by sex, age and socioeconomic level.Note.♂ male; ♀ women; mean.
.2%* (n=15.789)Sex ♂: 27.7%* (n = 7.291) ♀: 28.4%* (n = 7.195) Note.♂ male; ♀ women; : median; PROESP: Project Sports Brazil; CRF: cardiorespiratory fitness; CSEP: Canadian Society for Exercise Physiology; AAHPERD: American Alliance for Health Physical Education Recreation and Dance; a without reference because the article does not include the year of the cut-point of PROESP-BR; * value calculated based on the following equation: [(number of subjects with adequate aerobic fitness/number of subjects in the total sample) * 100]; † sum of the total sample of the studies collected, without stratifying by sex -for the studies that only investigated one of the two sexes, the value of the sample was computed in the sex-specific information.

Table 1 .
Description of studies on the prevalence of aerobic fitness in Brazilian children and/or adolescents.

Table 2 .
Evaluation of the methodological quality of included studies.Yes Yes Yes Yes No No No Yes No NA Yes NA NA No Castro et al. 62 Yes Yes Yes No Yes No No Yes Yes NA Yes NA NA Yes Coledam et al. 77 Yes Yes Yes Yes Yes No No Yes No NA No NA NA Yes Coledam et al. 78 Yes Yes Yes Yes Yes No No Yes Yes NA Yes NA NA Yes Coledam and Ferraiol 79 Yes Yes Yes Yes Yes No No No Yes NA Yes NA NA No Costanzi et al. 80 Yes Yes Yes Yes Yes No No Yes Yes NA Yes NA NA Yes Dorea et al. 18 Yes Yes Yes Yes Yes No No Yes Yes NA Yes NA NA No Fronza et al. 69 Yes Yes Yes Yes Yes No No No Yes NA Yes NA NA Yes Author(s), year Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Total score Andreasi et al. 73 Yes Yes Yes Yes Yes No No No Yes NA Yes NA NA No Araujo et al. 17 Yes Yes No Yes No No No No Yes NA Yes NA NA No Barbosa Filho et al. 58 Yes Yes Yes No Yes No No Yes Yes NA Yes NA NA No Bergmann et al. 74 Yes Yes Yes Yes Yes No No Yes Yes NA Yes NA NA No Brito et al. 75 Yes Yes Yes Yes Yes No No No Yes NA Yes NA NA No Casonatto et al. 76 23 Yes No No No No No No Yes Yes NA Yes NA NA No Mello et al. 24 Yes Yes Yes Yes Yes No No No Yes NA Yes NA NA No Minatto et al. 25 Yes Yes Yes Yes Yes No No Yes Yes NA Yes NA NA Yes Minatto et al. 65 Yes Yes Yes Yes Yes No No Yes Yes NA Yes NA NA No

Table 3 .
Studies on the prevalence of aerobic fitness in Brazilian children and/or adolescents.

Table 4 .
54, points of the PROESP-BR48,49, AAHPERD54and Fitnessgram 44 for children and adolescents stratified according to sex and age Project Sports Brazil; AAHPERD: American Alliance for Health, Physical Education, Recreation and Dance; m: meters.