Lung function in obese children and adolescents without respiratory disease: a systematic review

Background Obesity in children and adolescents is associated with increased morbidity and mortality due to multisystemic impairment, including deleterious changes in lung function, which are poorly understood. Objectives To perform a systematic review to assess lung function in children and adolescents affected by obesity and to verify the presence of pulmonary changes due to obesity in individuals without previous or current respiratory diseases. Methods A systematic search was performed in the MEDLINE-PubMed (Medical Literature Analysis and Retrieval System Online), Embase (Excerpta Medica Database) and VHL (Virtual Health Library/Brazil) databases using the terms “Lung Function” and “Pediatric Obesity” and their corresponding synonyms in each database. A period of 10 years was considered, starting in February/2008. After the application of the filters, 33 articles were selected. Using the PICOS strategy, the following information was achieved: (Patient) children and adolescents; (Intervention/exposure) obesity; (Control) healthy children and adolescents; (Outcome) pulmonary function alterations; (Studies) randomized controlled trial, longitudinal studies (prospective and retrospective studies), cross-over studies and cross-sectional studies. Results Articles from 18 countries were included. Spirometry was the most widely used tool to assess lung function. There was high variability in lung function values, with a trend towards reduced lung function markers (FEV1/FVC, FRC, ERV and RV) in obese children and adolescents. Conclusion Lung function, measured by several tools, shows numerous markers with contradictory alterations. Differences concerning the reported results of lung function do not allow us to reach a consensus on lung function changes in children and adolescents with obesity, highlighting the need for more publications on this topic with a standardized methodology.


Authors summary
What is known?
(i) Obesity in children and adolescents is a risk factor to higher morbidity and mortality due to multisystemic impairment; (ii) Obesity as a deleterious factor in lung function is poorly understood.

What is new?
(i) Spirometry was the most widely used tool to assess lung function in obesity and showed a high variability in its values, with a trend towards reduced lung function markers in children and adolescents with obesity; (ii) Differences regarding the reported results of lung function do not allow us to reach a consensus on lung function changes in children and adolescents with obesity, highlighting the need for more publications on this topic with a standardized methodology.

Background
Obesity is a dysfunction that interferes with systems of the body and whose prevalence increases in epidemic proportion [1]. The comprehensive impact of obesity prompts researchers to reflect on its deleterious effects, which progressively worsen the quality of life of increasingly younger individuals, leading children and adolescents to suffer from impairments that had been previously observed in adults only [2,3].
The respiratory system is one of the systems affected by obesity. Among adults, the most frequent findings in the comparison of lung function of healthy individuals versus individuals affected by obesity are the reduction in functional residual capacity (FRC) and expiratory reserve volume (ERV). One of the main reasons for these changes is the impairment of the respiratory mechanics. Excessive adipose tissue, mainly in the thorax and abdomen, causes an increase in the intra-abdominal pressure on the diaphragm and in the pressure of the fat tissue on the rib cage, hindering thoracic expansion and, consequently, lung compliance. This change leads to a reduction in lung volumes and capacities and is characteristic of a restrictive lung disease [4,5].
Respiratory mechanics are not the only way of compromising lung function in obesity. There is also an inflammatory component that causes obstructive pulmonary disorders. Adipose tissue macrophages produce proinflammatory substances and adipocytes secrete hormones (adipokines), which reach the systemic circulation, and are able to act directly on the respiratory system or alter the immune response. The entire pathophysiological process favors the induction of bronchial hyperreactivity and may compromise pulmonary air flow [6][7][8].
The prevalence of obesity has increased worldwide, and although it is a relevant public health problem that affects all age groups, the role and methods to evaluate its impact on lung function in children and adolescents is still unclear, and full understanding of the topic is still far from being attained.
As the obesity and lung function are complex phenotypes and their interaction has not been well understood, we included Fig. 1 that summarizes the mechanisms of lung function impairment due to obesity. Despite the relevance of the topic, it is still not well-established when lung function damages related to obesity start. Studies with children and adolescents diverge in their conclusions. Body changes during childhood and adolescence, variations in age range as well as in ethnic/environmental/genetic specificities make understanding of the cause-effect relationship between obesity and lung function more difficult.
Obesity and its association with lung function have been more often studied to ascertain the diagnosis of patients with asthma. Obesity and asthma have shown increasing prevalence in the last decades, and at the same time, they share common aspects, including the inflammatory process [8][9][10].
Obesity and asthma have been described as concomitant risk factors, with characteristics of a cause-effect relationship. The assessment of lung function in individuals with obesity and without asthma has yielded mixed results. Comprehensive studies are needed to understand whether mechanical and inflammatory changes are present in childhood obesity and during the growth process. Further studies should verify whether the disorders manifest differently during childhood and adolescence due to body changes throughout these periods, especially in individuals without a known lung disease, which is the focus of this systematic review [8][9][10].
This systematic review aimed to assess lung function in children and adolescents with obesity and to verify the presence of pulmonary restrictive or obstructive damages due to obesity in individuals without previous or current respiratory diseases, including asthma.

Methods
A systematic search was conducted in MEDLINE-PubMed (Medical Literature Analysis and Retrieval System Online, Public Medline), Embase (Excerpta Medica Database) and VHL (Virtual Health Library -Brazil) databases. Platform-specific tools were used, considering the descriptors (PubMed -MeSH Terms, VHL -DeCS and Embase -Emtree Terms) and equivalent terms, as well as excluding the descriptor "asthma" in order to achieve the objective of the study. The terms used in each database are described in Table 1.
The articles were selected in three stages, as detailed in Fig. 2. The titles were first read and the articles that were not relevant to the review were excluded (358, 1, 010 and 68 articles found in PubMed, Embase and VHL, respectively, were excluded). After the first database search filter was used, the articles were selected based on the abstracts (40,35 and six articles found in PubMed, Embase and VHL, respectively, were excluded). In the third stage, the articles were read in full and then, carefully screened according to their relevance to the topic (10, 23 and seven articles found in PubMed, Embase and VHL, respectively, were excluded). Also, using the PICOS strategy, the following information was achieved: Patientchildren and adolescents; Intervention (Exposure)obesity; Controlhealthy children and adolescents; Outcomepulmonary function variability; Studiesrandomized controlled trial, longitudinal studies (prospective and retrospective studies), cross-over studies and cross-sectional studies.

Eligibility criteria
In our study we included the following types of studies: randomized controlled trial, longitudinal studies (prospective and retrospective studies), cross-over studies and cross-sectional studies. Also, there was no restriction regarding the length of follow-up, and we considered only studies published between 2008 and 2018. We considered studies published in English, Spanish or Portuguese, and that were available in full text.

Information sources
All the studies were collected from the following databases: MEDLINE-PubMed, Embase and VHL databases. All the data was extracted directly from the studies and there was no contact with the study authors.

Study selection
In brief, the study selection was carried out as represented in Fig. 2. Also, two authors (MFS and FALM or MFS and VLWW) decided about the eligibility before including the study in the review. In the presence of ambiguous conclusion, a third author (RTM or JDR) was contacted to perform a full consideration. Afterwards, a fourth author revised all the studies and the dataset to reach a final decision (RTM or JDR).

Data collection process
The data collection was carried out by two authors (MFS and FALM or MFS and VLWW), in this way, the data collection was performed twice for each study. Also, after the data extraction, the study was described as Table 2 and both authors included a summary using both datasets generated in the individual data collection.
In the presence of ambiguous information, a third author (RTM or JDR) was contacted to perform a full consideration.

Results
As described in the methods, in brief, the articles were selected in three stages. A total of 48 articles were selected. After the exclusion of duplicates, 33 articles were included in the systematic review. Table 2 shows a detailed informative and descriptive summary of the articles in this review: authorship, year of publication, place of study (country), study objective, presence or absence of respiratory disease, type of study, type of evaluation of lung function and type of posture and markers used in the analysis, main results and conclusions.
It is important to highlight the exclusion of studies that included participants with respiratory diseases, since the focus was to evaluate the pulmonary changes resulting exclusively (or with the least possible influence of other factors) from obesity. Therefore, including children with previous respiratory diseases could evolve into sampling (selection of individuals), confounding (proven impact on outcomes) or information bias (previous knowledge of an existing problem). In this context, 66.7% (22/33) [12-15, 17-21, 25, 26, 29, 32-37, 39-42] of the studies excluded individuals with respiratory diseases, 27.3% (9/33) [16, 22-24, 27, 28, 30, 31, 38] did not mention respiratory diseases as a factor of exclusion or non-inclusion, and 6% (2/33) [11,43] excluded only individuals with a history of smoking. Among the studies that excluded previous respiratory diseases, several exclusion criteria could be observed: some authors excluded only individuals with exacerbation of asthma or cough; others excluded any respiratory conditions that might impair the evaluation; and others used standardized instruments such as the ISAAC questionnaire (The International Study of Asthma and Allergies in Childhood). This demonstrates the variability of methods adopted by the authors of the different studies, low standardization of exclusion/inclusion criteria, and difficulties/limitations to evaluate, diagnose and exclude patients with possible respiratory diseases in some specific cases among the evaluated children and adolescents [44].
In the evaluation of lifestyle habits, 15.2% (5/33) [20,24,30,41,42] of the studies assessed the participation of individuals in physical activities. Also, 9.1% (3/33) [16,20,42] of the studies assessed screen time of the participants and one [24] study mentioned the use of a lifestyle habits questionnaire but did not detail the assessed variables.
The comparison of the studies with a control group that showed comparative values or significant associations is described in Table 3. In this context, a great number of studies for each variable could be observed due to the selection of the various markers evaluated. No clear pattern emerged, as regards FEV 1 and FVC, with about half of the studies reporting no association between obesity and lung function parameters [11, 14, 16-20, 26, 27, 29-38, 41-43]. A clearer pattern emerged as regards FEV 1 /FVC and FEF 25-75% , as most studies found either a negative association or no association between obesity and lung function [11, 14, 16-20, 26, 27, 29-38, 41-43], with only one study reporting a positive association (Table 3) [13].

Effects of growth and development on lung function
Childhood and adolescence are characterized by major changes in the structure and functions of the human body systems. The physiological processes that influence lung function in 6-year-old children are different from those influencing 15-year-old adolescents, even if we disregard other multiple factors, such as gender, ethnicity, environment and genetics. Changes in lung function of adults with obesity are related to increased intraabdominal pressure due to the deposition of fat in this region, which compromises the efficiency of diaphragmatic mobility, as well as the deposition of fat on the rib cage, which reduces its compliance. However, such aspects are insufficient to understand the influence of obesity on lung function of children and adolescents [48,49]. Figure 3 lists some factors that influence lung function in children and adolescents and that should be considered in the discussion about the variability of the findings in this systematic review. The first factor is the increase in lung volume and surface for gas exchange until approximately eight years of age. In this context, 45.5% (14/33) [11, 13, 16, 17, 22, 24-28, 33, 34, 38, 41] of the studies in this review included individuals under 8 years olda period characterized by airway growth and development. Among these studies, 13 [11, 13, 16, 17, 22, 24-27, 33, 34, 38, 41] included, at the same time, individuals over eight years of agea period, in which the respiratory tract is anatomically formed.
According to the literature, until the end of the preschool age, the respiratory tract growth follows a dysanaptic pattern, i.e., the growth of the airways is slower than the growth of the lung parenchyma. After this period, the growth is isometric, showing greater homogeneity. This pattern of growth up to early childhood may lead to increased airway resistance and higher risk of obstructive processes in these individuals, especially in males, who have proportionally smaller airways than females during this maturation period. Thus, the findings regarding lung function in children and adolescents should be analyzed considering the lung growth phase and the child's development [49,50].
The second and third factors described in Fig. 3 are interrelated and associated with changes arising from growth. The onset of puberty marks the beginning of a process of maturation, characterized by body and psychological changes. The changes vary according to gender, and in women, the pubertal development begins approximately two years earlier than in men. Moreover, hormonal changes may directly influence lung function: at first, by growth spurts, followed by increase in trunk height and ribcage diameter, which influence the increase in lung capacity and volumes. Another example is the increase in the production of male testosterone during puberty, which triggers a muscle growth peak, which includes the respiratory muscles and favors the increase of FVC and respiratory flows [51,52].
Among individuals affected by obesity, the changes mentioned in the previous paragraph tend to occur at an earlier stage. Although the mechanisms are not yet well- Peak expiratory flow measured by spirometry 2 4 3 9 Expiratory reserve volume -3 2 5 Functional residual capacity -3 -3 Total lung capacity --3 3 Residual volume Of the studies included in the systematic review, 21.2% (7/33) [14,16,21,23,32,38,43] analyzed pubertal development [53,54]. The last item described in Fig. 3 is fundamental to understand lung function in children and adolescents and is associated with physical activities and sedentary lifestyle habits. In this context, we should emphasize that the comprehensive knowledge of the studied sample is of utmost importance. Today, the majority of the population is sedentary, and sedentary behaviors are not exclusively associated to the group of individuals affected by obesity. So, if the sample is composed of individuals with obesity under treatment/follow-up and HC with sedentary lifestyle habits, there may be a bias in relation to cardiorespiratory conditioning, which may affect lung function. Therefore, in the analysis of lung function in children and adolescents with obesity, information about physical activities and screen time is fundamental [55,56].

Measurement tools to define obesity
Body mass index (BMI) was the most commonly used tool to determine obesity in children and adolescents. However, the criteria to define obesity varied among the studies. The option of using country-specific standards of normality or those published by the CDC or WHO, determined the lack of homogenization of the samples. Also, the cut-off points for the definition of obesity within the selected references were different. Therefore, the use of comparable criteria between the studies would allow a more precise definition regarding the presence of obesity in children and adolescents. Comprehensive references, including data collection at a global level, are best indicated as they analyze normality patterns, taking ethnic differences into account.
The amounts of body fat and lean mass make the assessment of lung function in individuals with obesity more reliable. This can be explained because BMI, which is the most commonly used indicative of obesity, shows some limitations. BMI measures excess weight rather than excess fat and the variability due to gender, age, ethnicity and lifestyle habits may act as modifiers. Thus, well-trained individuals with high lean mass indexes are classified with obesity [57,58].
In this systematic review, only 12.1% (4/33) [15,18,23,38] of the studies used instruments that allowed the quantification of body fat. The first study used bioimpedance (BIA) [23], which estimates fat mass, fat-free mass and total body water. The second study [18] used BIA and the skinfold measurement test, which estimates the amounts of fat in each segment. However, the literature reports that this method has its limitations for the assessment of individuals with obesity [59]. The third study [38] only analyzed the skinfolds, and the fourth [15] used BIA and dual energy X-ray absorptiometry (DXA), which is a more accurate tool, as DXA assesses the amount and distribution of fat and lean body mass. The analysis of the distribution of body fat is an important tool to detect alterations caused by impaired respiratory mechanics, which is greater with the increase in fat in the thorax and abdomen.

Obesity epidemic reflected on the diversity of the studied populations
The inclusion of studies with individuals from almost all continents, except for Africa, is relevant for the analysis of the findings. They reflect a global epidemic of obesity among children and adolescents. According to the WHO, the prevalence of overweight and obesity among children aged 5 to 19 years increased from 4% in 1975 to 18% in 2016. Currently, more than 124 million children and adolescents are affected by excess of weight. Weight gain trends include both developed and underdeveloped countries, and currently, overweight and obesity are more prevalent and more often associated with causes of death than underweight, except in some parts of Africa (especially sub-Saharan Africa) and Asia [60,61].

Trends related to lung function in children and adolescents with obesity
The variability of the results showed an inability to establish lung function changes in children and adolescents with obesity. However, some trends have been detected and are discussed as follows.
The analysis of inflammatory markers to identify systemic impacts caused by obesity was reported in only 6% (2/33) [11,21] of the studies. One study [21] did not make any references to comparisons with the HC group. But, in another study [11], a positive association of FeNO with BMI was determined, suggesting that inflammation was more often detected in individuals affected by obesity and that these inflammatory changes should be considered in the clinical evaluation.
In short, FEV 1 /FVC was the spirometry marker with the greatest sensitivity to identify a possible obstructive process due to obesity in children and adolescents. However, the variability of this markerand, even more of other indications of obstructionwas high, regardless of the study. Thus, the development of cohort studies on indicatives of growth stages and body development is of utmost importance, since these factors are different between individuals affected by obesity and HC and also influence lung function. Pubertal development or age cohorts considering the time of dysanaptic growth and isometric growth would considerably reduce confounding factors and allow better understanding of lung function changes due to obesity in children and adolescents.
Besides the spirometry variables, some other measures contributed to the analysis of lung function in children and adolescents with obesity. TLC, FRC and residual volume (RV) were markers used in only 9.1% (3/33) [19,20,43] of the studies comparing individuals with obesity and HC. All these studies reported that individuals affected by obesity showed lower value or a negative association of FRC and RV with obesity. However, there were no differences in relation to TLC. If TLCwhich is the sum of the inspiratory capacity (IC), and FRC (FRC = RV + ERV)does not present a difference between individuals with obesity and HC, and if FRC and RV are reduced, it can be assumed that IC should be higher in individuals affected by obesity. Only 2 [18,43] studies analyzed this variable and found a higher value or a positive association of IC with obesity.
These results are in agreement with a systematic review published in 2016, which evaluated the effects of obesity on lung volume and capacity in children and adolescents, and found a reduction in some markers in obesity, especially the reduction in FRC, ERV and RV [66].

Issues on respiratory physiology and biomechanics requiring further investigation
The values for respiratory mechanics of individuals with obesity are incoherent. Some hypothesis and questions can be raised, namely: (i). Are individuals with obesity actually "stronger" and are, therefore, able to inspire more air? (ii). However, if there is an increase in RMS, considering the 9 [11,14,17,25,29,32,33,36,42] studies that analyzed peak expiratory flow measured by spirometry (individuals with obesity x HC), why did only 2 [11,29] studies find a positive association with obesity? Why did 4 [17,25,32,36] studies find lower value or a negative association between obesity and peak expiratory flow measured by spirometry? And, why did 3 [14,33,42] studies find no differences between groups? (iii). Are lung function changes in individuals with obesity due to the differences between males and females, since in females, there is a pattern of gynoid obesity, with higher fat concentration on the hip and legs; whereas in males, an android pattern occurs, with more volume of fat in the chest and abdomen? (iv).It is well-known that individuals affected by obesity tend to initiate pubertal development earlier than healthy individuals. So, is there greater muscle development among individuals with obesity, which tends to be balanced at the end of puberty? Can impairment of lung function in individuals with obesity be clearly observed after puberty?
Confounding biases to clarify the mechanisms that interfere with lung functions in children and adolescents affected by obesity Obesity is a multisystemic dysfunction, and therefore it is difficult to control the variables in order to understand the damage caused to lung function. For this reason, we found high variability in the results. Given the studies included in this systematic review, we are not able to establish which ventilatory changes are due to obesity in children and adolescents, even excluding data whose focus was the influence of asthma on obesity, which is a bias in this analysis. There are mechanisms that correlate both dysfunctions and the causal relationship between them may hinder perception of what is actually a consequence of obesity and/or asthma [8][9][10].
The findings of this review, although inconclusive, may give us a direction for future research. The strategies include: greater sampling control; reduction of confounding variables; conducting interdisciplinary and longitudinal studies with individuals with obesity versus HC; detailed analysis of environmental and social aspects; validation of findings among different populations; larger sample size; inclusion of measurements of lean mass and fat mass in order to unify and establish better criteria to define obesity; and future studies aiming to associate different genetic aspects, with predisposition to variability for weight gain, as well as for the individual nuance of lung function.
Thus, the inclusion and analysis of lung function in children and adolescents have become fundamental. Pubertal staging should be considered in order to avoid the influence of early maturation of individuals affected by obesity on the overestimation of lung capacity. It is important to analyze fat distribution, considering the concentration of abdominal and thoracic fat as factors that directly influence lung function. For this analysis, the use of instruments, such as DXA, may help determine the influence of the distribution and amount of body fat on lung function.
Assessing RMS with manuvacuometry or the distribution of lean mass with DXA, or even amount of lean mass using BIA, also favours the understanding of the physical conditions that influence lung function in children and adolescents. In order to analyze physical conditions, it is also important to include the evaluation of programmed and non-programmed physical activities as well as the screen time.
It is essential to exclude previous respiratory conditions that may influence lung function and control variables that are indicatives of inflammation, including CRP, erythrocyte sedimentation rate, FeNO or serum levels of leptin, adiponectin, IL-6 and TNF-α, which will allow precise determination of the influence of overweight in lung function of children and adolescents with obesity.

Meta-analysis
Meta-analysis is the gold standard in order to interpret a specific topic such as the importance of lung function in cases of obesity in the pediatric population. However, as described in our data there is no standardization in the studies about lung function in children and adolescents with obesity. To perform a meta-analysis, a minimum of standardization should be applied in the data acquisition. However, looking for the data included in Table 2, the studies were performed using different methodologies and/or lung function tools and/or lung functions measures (markers). Moreover, the age range was not equal, and the objectives were different among the studies. In brief, articles [11, 13-17, 19, 20, 22, 25, 26, 28-30, 32-38, 40-42] were described as cross-sectional studies on children-adolescents without respiratory diseases, where spirometry markers (such as FVC, FEV 1 and FEV 1 /FVC) were assessed, and the relation between lung function and body mass, expressed as either BMI or norm weight/overweight/obesity, was estimated. Those studies did not allow us to perform a meta-analysis because there is a disparity of study objectives, population type (age range, sex distribution), origin of the population, presence of other lung function measurement, obesity as an independent variable and the exercise analysis. The information about the disparities among the studies is shown in Table 2 and Table 4.

Conclusion
The different results observed for lung function in children and adolescents with obesity show that there is no consensus on the impairment in such individuals in the literature. Considering the influence of growth and development on the function of all systems, it is fundamental to control the variables to reduce sampling, information and confounding biases, as well as to enable the analysis of the deleterious effects of obesity. In this context, new studies should require greater control of variables that influence growth and development to better understand the influence of obesity on lung function of children and adolescents.    Table 4 Disparities between the markers of cross-sectional studies on children-adolescents without respiratory diseases, where spirometry markers were assessed, and the relation between lung function and body mass, expressed as either body mass index or norm weight/overweight/obesity preventing the performance of meta-analysis (Continued) Authors ( However, studies on individuals with obesity describe a trend towards lower FEV 1 /FVC, FRC, ERV and RV, suggesting that both mechanical and inflammatory impairments influence lung function throughout childhood and adolescence.
Studies on pubertal development would be significant for a standard comparison including hormonal and structural changes in this period and the onset and duration of maturation. The quantification and distribution of body fat and the analysis of lifestyle habits would promote coherence and standardization on this subject, favoring the clinical approach to individuals.
The prevalence of obesity has increased worldwide, and although it is a relevant public health problem that affects all age groups, the role and methods to evaluate its impact on lung function in children and adolescents have not been established yet, and full understanding of the topic is still far from being attained.