Abdominal muscular endurance in Brazilian children and adolescents : systematic review of cross-sectional studies Resistência muscular abdominal em crianças e adolescentes do Brasil : revisão sistemática dos estudos transversais

The literature has demonstrated the importance of musculoskeletal fitness in the prevention of chronic noncommunicable diseases. Although current recommendations of physical activities include strengthening and muscular endurance aspects, little is known about the muscular endurance levels of children and adolescents in a national scope. The aim of this study was to systematically review the literature to identify the prevalence of Brazilian children and adolescents who meet health criteria for muscular endurance. A hierarchical search was conducted in four databases (MEDLINE; Scopus; SciELO; LILACS) using the following terms: “muscular endurance”, “muscle endurance”, “physical fitness”, “child”, “adolescent”, “adults” “school” and correspondents in the Portuguese language. Overall, 2,652 articles (2,269 had their title and abstract read) were found and 70 were eligible for reading in full. Seventeen studies were reviewed and evaluated for risk of bias. Among 32,661 children and adolescents, only 40.2% of boys and 31.9% of girls presented abdominal muscular endurance adequate for health, and most studies were conducted in the southern region of Brazil. The variability in procedures for evaluating abdominal muscular endurance and cut-points used for interpretation occurred due to the use of different standardizations (PROESP/BR®, FITNESSGRAM® and AAHPERD®). Less than half of Brazilian adolescents of both sexes have adequate abdominal muscular endurance for health. Studies investigating the causes and consequences of inadequate abdominal muscular endurance may contribute to strategies for disease prevention and health promotion of children and adolescents.


INTRODUCTION
Physical fitness has been characterized as the ability to perform daily activities with vigor and resistance to fatigue 1 and has been considered an important health marker, also demonstrated by the strong association with cardiovascular, metabolic and motor competence outcomes in children and adolescents [2][3][4] .The health-related physical fitness components have been organized into cardiorespiratory, musculoskeletal, motor and body composition 2 .Although cardiorespiratory fitness has a notable importance for general health, the independent contribution of musculoskeletal fitness has been evidenced, basically by the expression of muscular strength and endurance 4 .The latter is defined by resistance to repeated muscle contractions over time or to sustain contraction for an extended period of time 4,5 .
Physical activities of muscular strength and endurance have been recommended in addition to aerobic activities for children and adolescents 6 .Conversely, the absence of adequate levels of musculoskeletal fitness has been common in this population and is associated with low back pain 7 and high body adiposity 8 .In this regard, monitoring of muscular endurance is important to verify if levels are adequate or inadequate to health.
Musculoskeletal fitness can be evaluated by laboratory tests that offer advantages such as greater validity and reproducibility, but they are not possible to be used in population studies due to the high cost and operation of instruments, and the long time required in the measurement 5 .On the other hand, field tests appear to be viable alternatives for large-sample studies, especially when performed in school settings 5,9 .Artero et al. 5 found that the most commonly used field tests for assessing musculoskeletal fitness were handgrip and arm support / flexion tests, which showed high and low quality and reproducibility scores, respectively.However, the abdominal muscular endurance test is used in most physical fitness test batteries 9,10 , possibly because it represents the strength and endurance of core muscles in dynamic, easy-to-use protocols using low-cost equipment 10 .The variety of tests and protocols used to assess muscular endurance may imply divergences among studies, also assuming the limitation of the interpretation of different cut-points to meet health criteria.
Studies carried out in Brazil have shown heterogeneity in relation to the selection, size and origin of the sample of participants 11,12 , in addition to the use of different muscular strength tests and cut-points for the interpretation of health suitability [13][14][15][16] , which makes it difficult to summarize findings on muscular strength in Brazilian children and adolescents.In addition, there are no estimates of the number of children and adolescents in Brazil who meet the health criteria for this physical fitness component.The compilation of evidences about the Brazilian pediatric population may contribute to the standardization of field tests and to the elaboration of strategies to improve musculoskeletal fitness.Therefore, the aim of the present study was to identify the prevalence of children and adolescents in Brazil who meet the health criteria for muscular endurance.

Registration Protocol
The present review is registered in the International prospective register of systematic reviews (PROSPERO) under identification number CRD 42018079881.

Eligibility Criteria
The following inclusion criteria were applied: studies published as a scientific article, studies measuring muscular strength based on curl-up test (trunk flexion) or push-up test (elbow flexion test) and those presenting the classification of the localized muscle endurance according to some reference physical test battery or cut-point.
Studies published in languages other than Portuguese or English, without cross-sectional design, with sample of children and/or adolescents who were not Brazilian, studies not performed with schoolchildren, studies with athletes, those carried out on a sample with chronic diseases and medical diagnosis of diseases such as muscular dystrophies, cerebral palsy and Down syndrome were excluded.Studies that did not present prevalence and classification of localized muscle endurance according to some reference criteria, studies that focused on other physical fitness components such as aerobic endurance, flexibility and muscle power were also excluded.In addition, literature reviews, case reports, monographs, dissertations/theses were not eligible for this review.No restriction was imposed on the date of publication.

Sources of information
Searches were carried out in four databases (Medical Literature Analysis and Retrieval System Online -MEDLINE, through PubMed; Scopus; Scientific Electronic Library Online -SciELO; Latin American and Caribbean Literature in Health Sciences -LILACS, through Bireme).The search in all databases was finalized in January 2018.

Search strategy
The process of search, selection and exclusion of articles was performed by two independent reviewers, debated at consensus meetings, and in case of disagreement, a third reviewer would resolve the disagreement situation.
The descriptors used in the search were divided into three blocks and dealt with the outcome in the first block (muscular endurance, muscle endurance, physical fitness) of the population in the second block (children and adolescents: "child", "adolescent", "adult") and the place of study in the third block "school".The Boolean operators "OR" (inside the blocks), "AND" (between blocks) and the "AND NOT" operator precede the "adult" descriptor.To achieve the variations of descriptors, the truncation symbol ("") was used.
Searches were carried out with descriptors in English and Portuguese.

Selection of Studies
Initially, selected studies were filtered and duplicates were excluded by the EndNote X7® reference manager software (Philadelphia, USA).Then, titles and abstracts were read and those that did not fit the review objective were excluded.After this step, the remaining articles were read in full and checked according to the eligibility criteria of the review (Figure 1).All eligible articles had the list of references checked and potential studies were included for reading in full.Subsequently, if these studies met the established selection criteria, they were included in the present review.

Risk of bias
The risk of bias was assessed based on the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies of the National Heart, Lung, and Blood Institute 17 .It is an instrument for the evaluation of cross-sectional and cohort studies, with 14 questions regarding the internal validity of studies.Through the instrument, it is possible to check selection, information, measurement and/or confusion biases.For each criterion evaluated, scores of 0 "no" and 1 "yes" were assigned, and at the end of the study classification, a total score was assigned to each study based on the number of positive responses to the questionnaire in relation to the total number of questions 18 .Questionnaire questions that could not be answered by the available information and/or that were not applicable to the study and / or aspects that had not been reported were excluded from the calculation to determine the final risk of bias score 18 .Based on the final score, studies were classified as: low (score ≥0.70), moderate (≥0.50); and high (<0.50)risk of bias.

Summarization Measures
The main summarization measure of results was the percentage of children and adolescents who met the criteria recommended for health according to cut-points used.

RESULTS
The initial search resulted in 2,652 articles, of which 383 were duplicate articles and were consequently excluded in the EndNote X7® reference manager (Philadelphia, USA), remaining 2,269 for titles and abstracts.After reading titles and abstracts, 70 studies were identified to read the text in full and to apply the previously established inclusion/exclusion criteria.After the additional step of analyzing the lists of references, 12 other studies that met the inclusion criteria were found.After the identification, selection and eligibility stages, 17 articles were considered as the result of this review (Figure 1).Studies were found from 1995 to 2017, and the period of greatest publications was 2005 and 2016, with three publications each year (Figure 2).The mapping of the Brazilian regions that developed studies on local-ized muscle endurance shows the predominance of investigations in the southern region of the country (n = 10), followed by the southeastern (n = 3), northeastern (n = 2) and midwestern (n=10) regions, with one study covering all regions.The age of subjects investigated ranged from 6 to 18 years.Most studies investigated both sexes (n = 16), while one investigation addressed only males (Table 1).
With regard to the aims of studies, eleven studies aimed at descriptively evaluating the health-related physical fitness of children and adolescen ts 11,12,[14][15][16]20,21,25,28,31,33 and other five studies 13,22 -24,27 evaluated the association of physical fitness with other variables (socioeconomic and cultural level, place of residence (rural and urban), anthropometric characteristics (body fatness) and professionals involved in physical education classes) (Table 1).
A total of 32,661 subjects aged 6-18 years participated in studies that used the curl-up test (abdominal endurance).Of this amount, 40.2% of boys and 31.9% of girls were classified with abdominal muscle endurance levels adequate for health.Table 2 shows the results of each study, which classified localized muscle endurance as adequate or inadequate for health.In two studies, information was not stratified by sex 23,24 .The cut-points used in studies to classify localized muscle endurance as adequate or inadequate were those from PROESP/BR 19 , FITNESSGRAM 30 and AAHPERD 26 .

DISCUSSION
The main finding of this review was that less than half of children and adolescents in Brazil presented adequate abdominal muscular endurance according to the reference criteria adopted.This low prevalence of Brazilian Has the question of the research or objective in this article been clearly indicated?Q2.Has the study population been clearly specified and defined?Q3.The participation rate of eligible persons was at least 50%?; Q4.Have all subjects been selected or recruited in the same or similar populations (including the same time period)?Were the inclusion and exclusion criteria to be included in the study pre-specified and uniformly applied to all participants?;Q5.Were provided justification of sample size, power description or estimates of variation and effect?Q6.For the analyses in this article, were the exposure (s) of interest measured before the outcome (s) to be measured?;Q7.Was the deadline sufficient to reasonably expect an association between exposure and outcome if existed?Q8.For exposures that may vary in quantity or level, has the study examined different exposure levels as related to outcome (e.g., exposure categories or exposure measured as a continuous variable)?Q9.Have exposure measures (independent variables) been clearly defined, validated, trusted and consistently implemented to all study participants?;Q10.Has the exposure (s) been assessed more than once over time?; Q11.Have outcome measures (dependent variables) been clearly defined, validated, trusted and consistently implemented to all study participants?;Q12.Were the outcome evaluators blind to participants' exposure status?Q13.The loss of follow-up after the baseline was 20% or less?; Q14.Were the main confounding variables measured and statistically adjusted for their impact on the relationship between exposure (s) and outcome (s)?; ND.Cannot be determined; NR.Not reported; NA.Not applicable; 0. No; 1. Yes; ≥ 0.70: low risk of bias; ≥ 0.50: moderate risk of bias; <0.50 high risk of bias; * to determine the total score, we considered the following equation: (total of positive answers / total of questions considered for that study).

… continue
study conducted with 192,848 children and adolescents from 725 schools located in urban and rural areas of the United States found that 76.4% of girls and 78.3% of boys evaluated also met the health criteria for this variable 35 .
Brazil is a middle-income country with large social discrepancies and quite heterogeneous in terms of per capita income distribution 36 .The literature points out that socioeconomic factor can influence the physical fitness of children and adolescents 16 , since families with higher schooling and economic levels have better understanding of the benefits of physical activity, and can provide the participation of children in activities and sports that require specific environments such as clubs and/or training schools 37 .In this sense, one of the differences between results presented in this review on the percentage of young people with adequate levels of abdominal muscular endurance and the results presented in Portugal 34 and the United States 35 may be the socioeconomic differences between these countries.
Four protocols were found in this review that aimed to evaluate the abdominal muscular endurance component of physical fitness 19,26,30,32 .The curl-up test recommended by PROESP/BR 19 and AAHPERD 26 , and by the protocol proposed by Guedes et al. 32 used the measurement of maximum repetitions during one minute, having as difference between them the positioning of the upper limbs during the test: (1) crossed on the trunk 19,32 ; (2) positioned behind the head 26 .On the other hand, the test proposed by the FITNESSGRAM 30 differed from the other tests by: (1) measuring unlimited repetitions (performed at a rate of 20 repetitions per minute until exhaustion) and/or until reaching maximum of 75 repetitions; (2) positioning of the upper limbs (extended along the body) and lower limbs (140º of knee flexion, while in the other tests, knee flexion was 90º) and (3) the amplitude of the trunk flexion required to perform of the curl-up test (until fingers reached the other end of a 75 cm wide band, whereas in the other tests, trunk flexion should be performed until forearms 19,32 and/or elbows 26 touched the thighs).In this context, some authors 30 reported that the curl-up test with the upper limbs positioned next to the body generates less compression in the lumbar region.In addition, other authors 38 reported that sixty seconds would not be the best procedure to measure abdominal muscular endurance, but rather the muscle power, suggesting that longer time, such as that provided by curl-up tests with unlimited duration, would be more appropriate for the evaluation of this variable 38 .
Regarding the cut-points used to assess the health-related physical fitness of children and adolescents in relation to the abdominal muscular endurance component, two batteries (PROESP/BR 19 and AAHPERD 26 ) used cut-points based on normative values, that is, curves in percentiles of the population, and one (FITNESSGRAM 30 ) used reference values that indicate the future risk of children and adolescents to develop cardiovascular diseases.Gulias-Gonzales et al. 39 developed percentile values of Spanish children aged 6-12 years and reported that it is more interesting to establish categories associated with the risk of developing cardiovascular diseases as proposed by FITNESSGRAM 30 than percentile curves that often fail to report the real risk of the low physical fitness found.
Regarding the aim of studies, no study included in this review had as main aim to measure the abdominal muscular endurance component in children and adolescents.In most studies, this component was evaluated only for comparison with other variables that could influence healthrelated physical fitness and/or in conjunction with the other components that compose the physical fitness construct.The lack of specific studies focusing on the muscular endurance component is an issue that deserves to be highlighted.Although it is important to investigate the general physical fitness of children and adolescents, the information of the individual components is also relevant, as it will allow specific improvement actions.
In Brazil, stricto sensu Graduate programs in Physical Education that constitute Area 21 of the Coordination of Improvement of Higher Education Personnel (CAPES) present high concentration in the southern and southeastern regions of the country 40 .This condition, coupled with the fact that large part of the scientific studies carried out in Brazil are linked to higher education institutions that have graduate programs, may justify the finding of the present review that most articles included were carried out in the southern and southeastern regions of Brazil.
With respect to the risk of bias, 14 of the 17 articles included in this review were classified as having low risk of bias 11,[13][14][15][16][20][21][22]24,25,27,28,31,33 and other three studies 12,23,29 as moderate risk of bias. None of the studies incuded met questions 6 and 12 of the methodological quality instrument, which referred to the evaluation of the exposure before the measurement of the result and if the evaluator of results was blind.One reason that may have contributed to the non-meeting of question six of the instrument was the cross-sectional design of studies, which does not allow the evaluation of variables at different moments. Regardin question 12, which deals with the blindness of evaluators, none of the studies included in this review reported this condition in methodological aspects, which suggests that more attention should be paid to this aspect in the writing of manuscripts.
Regarding question 14 of the risk of bias instrument, which assesses whether variables with confusion potential were measured or adjusted, only three studies have met this condition 13,22,33 .This finding may indicate that the authors did not consider possible variables that could confuse the association with the outcome of interest.Failure to meet this methodological item may lead to a false association between independent variables and muscular endurance, thus compromising the veracity of the results obtained.
The present study presents some limitations such as the investigation of only cross-sectional studies, the number of databases investigated (only four), the inclusion of only articles in English and Portuguese, the selection of only studies that evaluated muscular endurance by curl-up test and/or push-up test, which resulted in studies predominantly with the curl-up test.The following are strengths of the present review: (1) the provision of data regarding the physical fitness level in relation to the abdominal muscular endurance component of Brazilian children and adolescents.This information may support future studies aimed at evaluating this physical fitness component in children and adolescents, as well as stimulating the development of strategies aimed at increasing this physical fitness component in the scientific and school environments and public health policies; (2) to highlight the state of the art in relation to cut-points used to evaluate abdominal muscular endurance in Brazil, suggesting that cut-points based on health criteria should be used and created to evaluate this health-related physical fitness component.

CONCLUSION
With data obtained from this systematic review, it could be concluded that less than half of children and adolescents in Brazil meet health criteria for abdominal muscular endurance.Some suggestions for this area and for future studies are: 1) the proposal of more studies that aim to evaluate abdominal muscular endurance as primary outcome; 2) the investigation of factors related to abdominal muscular endurance considering the adjustment of statistical analyses and potential confusion variables; 3) the use and proposal of cut-points for abdominal muscular endurance based on health criteria is urgent for a more accurate classification of the prevalence of Brazilian children and young who meet health criteria.

Figure 1 .
Figure 1.Flowchart of the process of search, selection and exclusion of studies.

Figure 2 .
Figure 2. Timeline of publications with a sample of children and adolescents in Brazil that dealt with localized muscle endurance.

Table 2 .
Prevalence of subjects classified as having adequate muscular endurance according to some reference criteria from the curlup test.for the calculation of the percentage of subjects with adequate muscular endurance, the following equation was considered: (number of subjects classified as having adequate muscular endurance/total number of subjects in the studies considered) * 100; †: the calculation only considered the information of the abdominal muscular endurance test of the studies found; §: the calculation for sex-stratified information only considered studies that presented information stratified by sex, thus, two studies were not included for this calculation (e.g.,Sehn et al., and Tornquist et al.).
34uth with adequate levels of abdominal muscular endurance differs from that found in other countries.Santos et al.34evaluated a sample of 22,048 Portuguese children and adolescents and found that 81.3% of girls and 83.8% of boys met the health criterion for abdominal muscular endurance.Another *

Table 3 .
Risk of bias assessment of studies included in the systematic literature review, 2018.