Back pain in children and adolescents: a cross-sectional study

Back pain is a growing problem worldwide, not only in adults but also in children. Therefore, it is becoming increasingly important to investigate and understand the factors that influence the early onset of back pain. The aim of this study was to determine the prevalence of back pain in children and adolescents and to identify predisposing risk factors and protective factors. A cross-sectional study was conducted between October and December 2019 in schools from northern Portugal, evaluating 1463 students aged 9 to 19 years, of both genders. The instruments used were the Spinal Mouse® to assess posture, the Inbody 230® to assess body composition, an online questionnaire to characterize the sample and back pain, and the FITescola® battery test to access physical fitness. Half of the subjects experienced back pain at least once in their lifetime. The most frequently mentioned were lumbar spine and thoracic spine, mostly with mild or moderate pain intensities. Age, female gender, percent body fat, prolonged smartphone and computer use, hyperkyphosis, and the lateral global spine tilt to the left side are all factors with higher relative risk of back pain. Practicing physical activity or sports regularly and video games have a protective effect. The prevalence of back pain in children and adolescents is very high: The study enhances the case for protective factors such as physical activity habits or video games while reinforcing risk factors such as percent body fat, prolonged smartphone or computer use, and posture.


Introduction
Back pain is a clinical condition that affects a wide spectrum of the world's population and has implications for public health, as well as economic and social concerns [1].There are several factors associated with the development of back pain, including repetitive activities or a sedentary lifestyle [2].Symptoms do not always reflect structural changes in the spine [3], and it is important to find the cause of symptoms as early as possible [4].Recent studies focused on how early in human life back pain sets in and how common it is [5,6].
Back pain in children and adolescents is a clinical condition that in recent years has seen increased attention from parents and the appropriate medical support services because of its early onset [6] but also due to a better understanding of the risk factors [7].To better characterize this condition, several studies have been conducted recently in several countries, including Portugal [8,9].
The prevalence of back pain is high among young students [5], with a higher representation of female children [10], but the precipitating factors have not been fully elucidated.A recent meta-analysis studied the relationship between physical fitness and back pain, but the results were inconclusive [11].
This study aims to characterize the prevalence of back pain in children and adolescents from Portugal and the possible protective and risk factors associated with this clinical condition.

Study design and participants
A cross-sectional study was carried out with children from a school cluster in northern Portugal (Fig. 1), between October and December 2019.
The study was explained to the physical education teachers at the school cluster.Subsequently, a description of the study was given to all students in these schools, and written informed consent was obtained from their parents or guardians after attending a presentation of the project, during which all doubts were resolved.All participants were given the opportunity to decline participation.
Exclusion criteria included students with acute musculoskeletal injuries or serious medical problems that prevented data collection.
This study was conducted according to the guidelines of the Declaration of Helsinki, and all procedures involving human participants were approved by the Ethics Committee of FADEUP-University of Porto (CEFADE 50, 2019).The directors of the participating schools gave their ethical approval and written consent.
The elaboration of the manuscript was based on the STROBE statement guidelines.

Instruments
An online questionnaire (Google Forms) was used to characterize the sample in terms of anthropometric data, physical activity and sedentary habits, as well as presence of back pain, its prevalence, and location (more than a region could be indicated).Back pain intensity was quantified by an 11-point numeric rating scale (NRS-11), associated with Faces Pain Scale-Revised [12].Pain classification was based on the study by Tsze et al. [13].
Physical fitness was analyzed using the FITescola® test protocol [14], where the tests are divided into three sections: aerobic condition, evaluated by the 20-m shuttle run; body composition, evaluated by body mass index (BMI); and neuromuscular fitness assessed in three components: trunk neuromuscular condition through abdominal strength (number of curl-ups), upper body neuromuscular condition through the maximal number of push-ups in one series, and lower body neuromuscular condition through the long jump.Flexibility was assessed with the sit and reach test.The test results were divided into three categories according to the reference values, namely low profile, normal profile, and high profile [14].
Body mass index (BMI) and percentage body fat (PBF) composition were assessed using the InBody 230 (InBody, Cerritos, California, USA) scoring system [15], a body analysis system based on the bioimpedance of the body.BMI and PBF were divided into three categories, namely low, normal, and high, adjusted for age and gender [16,17].
Posture was assessed through spinal evaluation performed using the Spinal Mouse® (Idiag, Voletswil, Switzerland), with proprietary software IDIAG M360pro® version 7.6.Spinal Mouse (SM) is a computerized wireless telemetry device, consisting of two wheels, sensors, and controllers in a protective casing, acquiring at 150 Hz, that is manually guided on the skin along the spine, from the seventh cervical vertebrae to the sacrum to quantify posture and spine mobility [18].
Spinal measurements were taken with the students in the orthostatic reference position and with minimal clothing on the trunk (the girls used tape to hold their bras, always accompanied by the researcher and a female teacher; the boys had their torsos unclothed).Postural analysis in the orthostatic position was performed in the sagittal and frontal planes.In the sagittal plane, posture was considered in four regions, each divided into three categories: thoracic spine: hypokyphosis/neutral/hyperkyphosis; lumbar spine: hypolordosis/neutral/hyperlordosis; pelvic tilt: anterior/ neutral/posterior; global spinal sagittal tilt: anterior/neutral/posterior.In the frontal plane, posture was also divided into three categories for the different regions studied: right/ neutral/left tilt.The reference angles for spinal curvatures in the sagittal plane in children are: thoracic kyphosis (33.3° ± 5.4°) and lumbar lordosis L1-L5 (39.6° ± 5.6°) and in adolescents: thoracic kyphosis (35.4° ± 4.9°) and lumbar lordosis L1-L5 (42.7° ± 4.5°) [19].The reference values for pelvic tilt in children and adolescents are 7.7° ± 11.3° [20].In the frontal plane and global spine in the sagittal plane, the reference values for neutral are 0º(± 3º).All the reference values were adjusted by ± 3º according to the value of the SM standard error of measurement (SEM) determined by Mannion et al. [18].
The privacy of the students was maintained by providing a private room for the examination.The average duration of each examination was approximately 5 min per participant.

Statistical analysis
Descriptive statistics were used to characterize the sample.Normality of the data was tested by the Kolmogorov-Smirnov test, and most of the variables did not have a normal distribution.Mann-Whitney U-test was used to estimate differences in the studied variables between the two gender groups.The Chi-squared test was used to estimate the differences between genders and back pain manifestation, as well as the NRS-11 intensity categories and their association with the back regions.
The Phi correlation coefficient test was used to measure the relationship between two binary variables (yes/no; female/male).
For the association between the manifestation of back pain and the variables studied, binary logistic regression was used to calculate the odds ratio (OR).We assumed as a missing value the answer option (I don't know the answer; it depends on the day/week) for those variables including it, because it did not allow to determine a specific answer.Statistical significance was set at α = 0.05.SPSS version 26 (IBM Inc., Chicago, IL, USA) was used for the statistical computations.

Results
From all school' students, aged 9 to 19 years, whose caretakers freely signed the informed written consent, 1463 agreed to participate in the study (719 females-49.1%;744 males-50.9%).
Through descriptive analysis of the data (Table 1), we can observe that there are significant differences between genders in mass, height, BMI, PBF, and NRS-11, but not in age.The largest significant difference relates to PBF and the smallest to pain intensity.
Back pain was present in half the children and adolescents, at least once in their lifetime (Table 2), with higher prevalence for females (57%).Most back pain complains mentioned occurred in the previous month for both genders, with a higher proportion of those occurring only once (30.1%).Females had a slightly higher percentage of limitations originating from back pain complains than males.The regions with the highest pain prevalence were the lumbar region, followed by the thoracic and cervical spine, and the combination of thoracic + lumbar spine, in both genders.
Analysis of the data in Table 3 shows that slight pain is the most prevalent (48.2%), followed by moderate pain.
When analyzing the odds ratio associated with the development of back pain through binary logistic regression (Table 4), we found that females have a 71% higher risk compared to males (OR 1.71).The risk of developing back pain also increases strongly with age and with PBF.
Competitively performing physical exercise reduces the probability of having back pain by 36% (OR 0.637) compared to this practice noncompetitively.However, for those performing physical exercise competitively or not, performing 2-3 h per day of physical exercise increases the risk by 58% (OR 1.579) compared to performing 0-1 h per day.Analysis of days and hours spent in physical activity, separated between competitive and noncompetitive, is presented in Table 5.
The use of smartphone or computer shows an increased risk of presenting back pain compared to not using them.A surprising finding is the use of video games for 2 to 3 h per day that significantly reduces the risk of developing back pain by 50% (OR 0.50), although based in a small number of students (56 in 1463).
In the analysis of posture, in the sagittal plane, hyperkyphosis showed a 44% (OR 1.437) increased risk of the manifestation of back pain.In the frontal plane, the lateral spine tilt (left side) is associated with an increased risk of developing back pain (OR 2.257), although only 52 students showing this lateral tilt.
Table 5 shows that most participants who performed competitive physical activity did so 2-3 h/day and 3-4 days/week, while those performing it noncompetitively spend 0-1 h/day and 1-2 days/week.

Discussion
The aim of this study was to investigate back pain in children and adolescents and factors that influence it.
In Portugal, some studies have been conducted, especially by Minghelly et al. [8], presenting disturbing data on the high prevalence of low back pain in adolescents (62.1% have had low back pain at least once in their lives).Our study shows that half of the students experienced back pain at least once in their lifetime (50.6%), and half of these students reported having had at least one episode of back pain in the previous month.Of the students who reported having back pain, 20.8% had a functional limitation related to that pain.
We considered different regions of the spine to better characterize back pain.The lumbar spine was the most commonly cited, accounting for nearly half of the complaints, followed by the thoracic spine, the cervical spine, and finally the pelvis, which is consistent with other studies [5,7].In our data, the greatest difference between genders was found in the cervical spine, with females representing 77% (53/69) of the total students complaining about this region.The most common pain intensities were "slight pain" and "moderate pain," with the female gender having a slightly higher mean score in NRS-11 (5.07 vs. 4.61).
Analyzing the influence of gender, females showed a higher prevalence of back pain and a higher risk of developing it than males.Age is another significant factor in the manifestation of back pain, especially by comparing the older adolescents to younger children.These two results are consistent with a systematic review by Calvo-Muñoz et al. [7].Body mass index showed no significant association with the risk of manifestation of back pain.However, our data suggest that children and adolescents with higher PBF have a 40% higher risk of developing back pain than those with normal PBF.These findings are consistent with studies showing the influence of excess body fat on the manifestation of musculoskeletal pain in children and adolescents [21].Furthermore, a recent review found an association between increased body fat and lower levels of moderate to vigorous physical activity [22].The association between posture and back pain has been highlighted in some studies [8,10], particularly in children.Minghelly et al. [8] found a relationship between posture (assessed with a scoliometer) and the occurrence of low back pain, especially when sitting in poor posture.In our study, a greater association with the manifestation of back pain showed only in the thoracic hyperkyphosis and the lateral global spine tilt (left side).We also see a large number of students with abnormal postures in the sagittal plane, such as anterior tilt of the spine and pelvis and hypolordosis, and in the frontal plane, right tilt of the pelvis and thoracic spine and left tilt of the lumbar spine.Although they are not significant risk factors for back pain, they may have an impact on adult life, and the consequences are often underestimated.These changes were also observed in other studies [23].Are we seeing the onset of a health problem of future generations?Considering the results for the sagittal and frontal planes, more studies are needed to clarify this question.
We did not find any association between physical fitness (assessed by the FITescola tests) and pain, a result in disagreement with a recent systematic review that found moderate evidence for this association [24].
Sports and physical activity in general acted as protective factors for the development of back pain.However, physical activity for 2-3 h/day increased the risk of developing back pain compared with 0-1 h/day.This increase in risk is related to the growth in the number of hours of physical activity, also present at over 4 h/day, although not significant statistically.These data suggest that practice of sport or physical activity for more than 1 h consecutively may increase the likelihood of developing back pain in children and adolescents.Sedentary habits are also associated with back pain [8], and in our case, the link with new technologies is particularly emphasized.The use of computers, but especially of smartphones by students, shows a higher risk of the manifestation of back pain.Smartphone use of 4 to 7 h per day increases the risk of developing back pain by four times.These results contrast with a study in which smartphone or computer use did not increase the likelihood of developing back pain [25].

Limitations
This study has natural limitations that are characteristic of cross-sectional studies when trying to understand a complex and multifactorial phenomenon.Thus, although we can establish associations between factors, we cannot establish direct causality between them.Although the study has a large sample, the fact that it was conducted in a limited geographical area is also a limitation.
For younger children, the limited understanding of the questionnaires and questions was overcome with the help of parents and teachers who kindly helped in a fundamental way in this process.

Conclusion
The prevalence of back pain is very high in children and adolescents, with some factors such as higher age, female gender, and sedentary habits contributing negatively

Table 1
Sample characterization by gender and its association with continuous variables *Mann-Whitney U test: (level of significance 95%)

Table 4
Binary logistic regression for the variable back pain and its relationship with the studied variables

Table 5
Detail of the entry for "physical exercise habits" presented in Table4Hours of physical activity or sport practice per day The posture of the thoracic spine, namely hyperkyphosis, and the lateral global spine tilt (left side) are also important factors in back pain prevalence.There are also protective factors such as sports and physical exercise practice and video games.