Knee and ankle kinematics in schoolchildren with different body weights: a cross-sectional study

Abstract

Background

Obesity and overweight in children contribute to developing kinematic gait deficiencies; however, reports on this matter are not conclusive. This study aims to compare the kinematics of the knee and ankle during walking between schoolchildren with overweight or obesity and schoolchildren with normal weight.

Methods

We conducted a cross-sectional study on schoolchildren aged 6 to 12 who were classified into three groups according to body mass index Z score: obesity ≥ 2.0 (n = 30), overweight 1.0 to 1.99 (n = 30), or normal weight − 1.99 to 0.99 (n = 35). The evaluation of the kinematics of the knee and ankles was performed through stereophotogrammetry, and the results obtained were compared between the study groups.

Results

We included 95 schoolchildren, primarily female (60%). We found no significant difference in flexion-extension movements between the study groups; however, we observed that the group with obesity presented a mean difference of 8.6° degrees (95% CI 1.55–15.65); p = 0.041 in right knee adduction during the initial swing compared to the group with normal weight and a mean difference of 6.35° (95% CI 0.26–12.94); p = 0.048 in the left knee also compared to the normal weight group. Additionally, we found differences in plantarflexion between the group with obesity and the normal weight group, with a mean difference of 6.41° (CI95% 2.61–10.22); p = 0.001 on the right side and the left side of 6.15° (CI95% 2.60–9.70); p = 0.003.

Conclusions

The kinematic differences in the knees and ankles between the obese and normal groups suggest “adaptations” during walking as their body weight increases.

Introduction

Overweight and obesity are global health problems. According to the World Health Organization, around 390 million school children were living with overweight or obesity in 2022 [1]. Furthermore, these health problems have been associated with repercussions on the musculoskeletal system due to excessive joint loading [2–5], leading to greater wear and tear of articular cartilage and limitations in the performance of activities of daily living [6].

Some authors have described different implications for movement and quality of life in adults and older adults due to childhood obesity. A systematic review included 25 cross-sectional studies comparing spatiotemporal, kinematic, kinetic, center of mass, and muscle activation/strength parameters of gait in children and adolescents aged 8 to 18 years with overweight (OW)/obesity (OB) and normal weight (NW) [7]. Moderate evidence supported greater knee movement in the frontal plane and an abducted knee position during gait in the OW/OB group compared with the NW group. However, sagittal-plane kinematic data were inconsistent, and transverse-plane evidence was insufficient to determine differences between OW/OB and NW gaits [7].

Furthermore, throughout the full gait cycle, inconsistent evidence reported differences in ankle and foot kinematics in all three planes (sagittal, frontal, and transverse) [7].

Despite the above, nowadays, there are still contradictions in reports on the kinematic impact of the knee and ankle joints on children with obesity. Some authors, for example, have reported a slight displacement of their knee flexion movement angle with an increase in extensor moment in children with obesity when compared to normal weight children [8, 9]. Other researchers, such as McMillan et al. (2009), reported greater knee abduction and hindfoot inversion during gait in boys with obesity [10], while others reported greater angular displacements towards adduction and knee abduction in children with obesity [8].

Furthermore, it has been reported that children with obesity have a lower angle of plantarflexion than those with normal weight [9, 11], although Nantel et al. (2006) reported a greater angle of plantarflexion in children with obesity than in those with normal weight [12].

Despite existing studies on kinematic alterations of the knee and ankle in schoolchildren, these studies do not allow us to identify alterations associated with weight gain, as the methodologies used and their results are diverse. Therefore, this study aimed to compare the kinematics of the knee and ankle during walking between schoolchildren with overweight or obesity and schoolchildren with normal weight. Furthermore, we correlated weight gain with the magnitude of angular displacement in the groups of schoolchildren studied.

Materials and Methods

Design and setting

We performed an analytical cross-sectional study from May 2021 to April 2023 at the Hospital Infantil de México Federico Gómez in Mexico City, Mexico. The participants were invited from the Obesity Clinic, Pediatric outpatient consultation, Dermatology Department, and the Child Development Centre (CENDI) of the Hospital Infantil de México Federico Gómez. Informed consent was obtained from parents or guardians. This study was approved by the Ethics and Research Committees, with registration number HIM-2020-042.

Participants

Using non-probabilistic sampling of consecutive cases, we included boys and girls aged 6 to 12 years, regardless of body weight. Informed consent or informed assent was requested, depending on the children’s age. Children who presented endocrinological or genetic alterations and those who received any drug associated with weight gain were excluded. Children with epilepsy, neuromuscular diseases, injuries to lower limbs younger than six months, or absence of a lower limb were also excluded. Figure 1 shows the selection of the participants.

Fig. 1.

Patient recruitment flow chart

Measures and procedures

We recorded the characteristics of the schoolchildren, such as their sex, age at the time of assessment, date of birth, caregiver relationship, weight, height, and comorbidities such as flat feet, varus or valgus of the knee, and scoliosis.

In addition, a clinical nutritionist performed the anthropometric assessment by measuring weight in kilograms and height in centimeters with a Seca^® model 755 stadiometer and measured the length of the lower limbs (greater trochanter-heel) in centimeters, the pelvic width (distance between anterosuperior iliac spines), pelvic depth (distance between the greater trochanter of the femur and the anterosuperior iliac spines), knee width (distance between medial and lateral epicondyles), and the distance between medial and lateral malleolus of the ankle, using a Lufkin EXECUTIVE^® anthropometric tape and the anthropometer included in the BTS Bioengineering^® software, it is a accurate, reliable state-of-the-art gait analysis system used in clinical and rehabilitation settings to assess posture and gait, it is considered the gold standard for the analysis of human movement, demonstrating the ability to provide comprehensive, accurate, robust, and reproducible data [13, 14]. Additionally, schoolchildren were grouped according to the World Health Organization (WHO) BMI Z-score classification: normal weight group n = 35 (-1.99-0.99), group with overweight n = 30 (1.0-1.99), and group with obesity n = 30 (≥ 2.0) [15].

Knee and ankle kinematics assessment

The gait study was performed at the Movement Analysis Laboratory of the Rehabilitation Department of the Hospital Infantil de México Federico Gomez by an expert in biomechanical clinical assessment. To assess the kinematics of the knees and ankles of the schoolchildren, we extracted data from the stereophotogrammetric study, such as angles of flexion, extension, abduction, and adduction movements of both knees and plantarflexion and dorsiflexion of the ankles during the phases of initial contact (0–2%), loading response (3–10%), mid (11–30%) and terminal stance (31–50%), as well as pre-swing (51–60%), initial (61–73%), mid (74–87%) and terminal swing (88–100%). We used the average value of 3 walking tests to compare the study groups.

Before the study, we placed 15-mm-diameter passive spherical photo-reflective spherical markers on the schoolchildren’s bodies according to the Davis protocol (seventh cervical vertebra, bilateral acromion, second sacral vertebra, anterior superior iliac spines, greater trochanters, lateral femoral condyles, fibular heads, lateral malleolus, fifth metatarsal head, and calcaneus) [16]. Once the markers were placed, each patient walked on a 5-m platform at a self-selected speed.

Knee and ankle kinematics in the sagittal and frontal planes during gait were measured using the BTS Bioengineering^® 3D stereophotogrammetry analysis system, model Smart DX 5000. It has four infrared cameras and the BTS SmartDX motion analysis system, along with BTS bioengineering^® software, which was used to analyze the videos.

After the analysis had been carried out, the results were given and explained to the participants’ keepers. Patients who presented alterations in the kinematics of the ankle or knee that affected their functionality were referred to the Rehabilitation Department of the Hospital Infantil de México Federico Gómez for assessment and treatment.

Statistical analysis

We used a mean difference formula with a power of 80% to calculate the sample size. Considering a two-tailed hypothesis, the sample calculated was 30 participants per group.

We used the Kolmogorov-Smirnov test to assess data normality, and the mean and standard deviation, or the median and interquartile range, for quantitative variables, and absolute and relative frequencies for qualitative variables. We used the Chi-square test to analyze differences between categorical variables; when the cell counts were less than 5, we used Fisher’s exact test. We used the one-way ANOVA with Bonferroni’s post hoc test to determine differences in kinematic parameters between the comparison groups.

We calculated Pearson’s correlation coefficient to assess the relationship between increases in body weight and the angular displacement of the knee and ankles across the comparison groups. A p-value < 0.05 was considered significant. Data were captured in an Excel spreadsheet and analyzed using Stata version 15.1 for Windows (StataCorp, Texas, USA).

Results

A total of 95 schoolchildren were included in this study; 55.8% were male, although the group of normal-weight schoolchildren was predominantly female (60%). The distribution of the participants and their demographic characteristics are shown in Table 1.

Table 1.

Clinical and demographic characteristics of 95 participants, according to the BMI Z-score classification

Characteristics	Normal weight n = 35	Overweight n = 30	Obesity n = 30	All n = 95	p-value*
Sex, n (%)
Male	14 (40)	16 (53.3)	23 (76.7)	53 (55.8)	0.012†
Female	21 (60)	14 (46.7)	7 (23.3)	42 (44.2)
Age in months, median (IQR)	107 (43)	116 (28)	108 (29)	113 (35)	0.81∞
Height in centimeters, median (IQR)	130 (25)	141 (15)	135 (22)	136 (21)	0.31∞
Body Mass Index (BMI), mean ± SD	16.56 ± 2.41	22.42 ± 2.10	25.92 ± 2.64	21.37 ± 4.6	< 0.001
Pre-existing changes, n (%)
Bilateral flatfoot	6 (17.1)	2 (6.7)	7 (23.3)	15 (15.8)
Bilateral knee valgus	1 (2.9)	0	0	1 (1.05)	0.070††
Bilateral ankle valgus	0	0	1 (3.3)	1 (1.05)
Other comorbidities	2 (5.7)	8 (26.7)	15 (50)	25 (26.3)

Other comorbidities (unilateral flatfoot, unilateral knee valgus, unilateral ankle valgus, lumbar scoliosis, lumbar hyperlordosis, bilateral cavus foot)

n number of participants, % percentage, IQR Interquartile range

^†Chi² test

^††Fisher’s exact test

^∞ Median test

*p value < 0.05

Knee kinematics

Comparison of flexion-extension movements between the study groups

Even though we observed that the right knee kinematics during flexion-extension was lower in the group with obesity (See Fig. 2a and b), it was not statistically significant (p > 0.05) nor in the left side (p > 0.05), Table 2.

Fig. 2.

a. Right knee flexion-extension displacement during gait, according to weight classification; b. Left knee flexion-extension displacement during gait, according to weight classification

Table 2.

Knee kinematics data in study groups during maximum displacement by movement and subphase

Movement and subphase	Normal weight n = 35	Overweight n = 30	Obesity n = 30	p value†
Knee
Right flexion-extension
Initial contact (2%)	16.4 ± 9.1	15.1 ± 8.4	14.8 ± 12.5	0.80
Load response (7%)	19.3.±9.1	17.6 ± 8.9	17.2 ± 12.3	0.66
Mid stance (11%)	18.6 ± 9.1	17.7 ± 9.2	17.5 ± 12.6	0.90
Final stance (37%)	10.1 ± 8.9	11.1 ± 8.2	10.9 ± 10.9	0.89
Final stance (50%)	18.6 ± 9.0	19.1 ± 9.8	20.3 ± 13.0	0.81
Pre-swing (60%)	44.4 ± 11.3	45.9 ± 10.1	43.7 ± 14.6	0.77
Initial swing (69%)	57.8 ± 11.1	56.2 ± 10.1	52.6 ± 15.6	0.24
Mid-swing (74%)	52.7 ± 12.8	48.9 ± 13.6	45.8 ± 18.4	0.19
Final swing (88%)	15.3 ± 13.1	11.8 ± 11.5	11.6 ± 12.6	0.41
Final swing (94%)	8.6 ± 9.9	7.1 ± 8.4	7.5 ± 10.6	0.81
Left flexion-extension
Initial contact (2%)	16.9 ± 9.2	18.4 ± 10.4	19.8 ± 11.0	0.53
Load response (7%)	19.5 ± 10.0	21.6 ± 10.2	21.9 ± 10.9	0.61
Mid stance (11%)	18.9 ± 10.1	21.7 ± 10.3	21.8 ± 10.9	0.44
Final stance (37%)	9.5 ± 10.3	12.6 ± 9.1	13.6 ± 9.2	0.21
Final stance (50%)	19.1 ± 10.6	19.3 ± 11.3	24.7 ± 9.5	0.08
Pre-swing (60%)	45.5 ± 12.3	45.3 ± 12.5	49.5 ± 11.4	0.31
Initial swing (68%)	57.7 ± 11.6	59.2 ± 10.3	58.9 ± 11.4	0.84
Mid-swing (74%)	52.1 ± 13.9	54.8 ± 12.9	52.3 ± 15.4	0.70
Final swing (88%)	15.5 ± 13.1	16.7 ± 11.5	15.9 ± 13.1	0.93
Final swing (94%)	8.6 ± 9.9	7.1 ± 8.4	7.5 ± 10.6	0.83
Right abduction-adduction
Initial contact (2%)	9.4 ± 6.0	11.2 ± 6.9	12.3 ± 9.6	0.29
Load response (10%)	10.6 ± 6.5	11.6 ± 6.9	12.6 ± 9.4	0.58
Mid stance (13%)	10.7 ± 6.2	11.5 ± 6.9	12.2 ± 9.2	0.71
Mid stance (29%)	9.7 ± 5.4	10.4 ± 6.2	9.8 ± 8.2	0.91
Final stance (50%)	11.3 ± 7.3	12.8 ± 7.8	12.9 ± 11.02	0.71
Pre-swing (60%)	13.6 ± 11.5	17.7 ± 11.9	19.8 ± 16.6	0.17
Initial swing (73%)	16.5 ± 12.7	23.2 ± 14.6	25.1 ± 15.8	0.041 a
Mid-swing (87%)	9.8 ± 5.4	11.2 ± 6.3	12.7 ± 9.9	0.28
Final swing (94%)	8.1 ± 4.9	8.8 ± 5.4	9.4 ± 7.4	0.64
Final swing (100%)	9.3 ± 5.9	10.8 ± 6.5	10.8 ± 8.5	0.60
Left abduction-adduction
Initial contact (2%)	7.7 ± 5.3	9.0 ± 3.8	9.1 ± 8.5	0.60
Load response (10%)	8.2 ± 5.4	10.0 ± 4.6	8.9 ± 8.6	0.54
Mid stance (12%)	8.2 ± 5.3	9.9 ± 4.7	8.6 ± 8.4	0.54
Mid stance (22%)	7.7 ± 4.9	8.8 ± 4.5	7.5 ± 8.0	0.68
Final stance (44%)	8.6 ± 5.3	9.5 ± 4.8	8.2 ± 8.5	0.74
Pre-swing (51%)	8.6 ± 7.1	9.9 ± 5.8	8.8 ± 10.7	0.78
Initial swing (73%)	11.2 ± 13.5	17.6 ± 10.0	18.0 ± 14.7	0.03 a
Mid-swing (87%)	6.9 ± 5.4	9.9 ± 5.5	9.4 ± 7.2	0.11
Final swing (92%)	6.4 ± 4.0	7.2 ± 3.4	7.2 ± 6.3	0.72
Final swing(100%)	7.5 ± 5.2	8.4 ± 3.6	8.5 ± 8.1	0.75

 (%): Percentage of the gait cycle where the maximum displacement angle occurred. Positive values for knee flexion and adduction; negative values for knee extension and abduction. The expected maximum point for each movement is shown

^†Differences through one-way ANOVA test with Bonferroni post hoc test

^aDifference between normal weight and obesity groups

Comparison of abduction-adduction movements between the study groups

We observed a tendency in both knees to remain in adduction, especially in the groups with overweight and obesity, but without statistically significant difference (See Fig. 3a and b). The knee adduction was more evident during the initial swing with a mean difference on the right side of 8.6° degrees (CI 95% 1.55–15.65); p = 0.041 between the normal weight and obesity group, and a mean difference of 6.8° (CI 95% 0.15–13.82); p = 0.03 in the left side, between the same study groups, Table 2.

Fig. 3.

a. Right knee adduction-abduction displacement during gait, according to weight classification; b. Left knee adduction-abduction displacement during gait, according to weight classification

Ankle kinematics

Comparison of plantarflexion-dorsiflexion movements between the study groups

During the initial swing phase, the group with obesity showed less plantarflexion than the normal weight group in both ankles (See Fig. 4a and b), with a mean difference on the right side of 6.41° (CI95% 2.61–10.22); p = 0.001 and on the left side of 6.15° (CI95% 2.60–9.70); p = 0.003. We also observed less plantarflexion in the group with obesity compared to the group with overweight, with a mean difference of 5.5° (CI95% 2.08–8.93); p = 0.002 in the initial swing phase. We found no statistical differences in the rest of the gait cycle subphases (Table 3).

Fig. 4.

a. Right ankle dorsiflexion-plantarflexion displacement during gait, according to weight classification; b. Left ankle dorsiflexion-plantarflexion displacement during gait, according to weight classification

Table 3.

Ankle kinematics data in study groups during maximum displacement by movement and subphase

Movement and subphase	Normal weight n = 35	Overweight n = 30	Obesity n = 30	p value†
Ankle
Right plantarflexion
Initial contact (0%)	-2.0 ± 5.5	-1.1 ± 4.9	-1.0 ± 4.9	0.69
Load response (10%)	4.6 ± 4.2	4.1 ± 4.0	4.8 ± 5.3	0.81
Mid stance (30%)	11.6 ± 3.8	12.2 ± 3.6	12.1 ± 5.2	0.85
Final stance (36%)	12.4 ± 4.5	13.4 ± 4.0	13.5 ± 5.2	0.53
Final stance (50%)	3.5 ± 7.7	4.0 ± 5.8	5.8 ± 8.6	0.44
Pre-swing (60%)	-12.8 ± 7.8	-9.8 ± 5.9	-6.4 ± 7.5	0.002 a
Initial swing (61%)	-12.9 ± 7.9	-9.7 ± 6.0	-6.5 ± 7.3	0.002 a
Initial swing (73%)	-0.9 ± 6.4	0.2 ± 5.7	1.5 ± 6.3	0.29
Mid-swing (81%)	2.8 ± 5.6	3.2 ± 4.6	3.2 ± 5.8	0.95
Mid swing(87%)	1.2 ± 6.0	2.3 ± 5.5	1.5 ± 5.6	0.72
Final swing (98%)	-2.9 ± 5.8	-1.4 ± 4.7	-2.1 ± 4.9	0.50
Left plantarflexion
Initial contact (0%)	-1.9 ± 4.6	-1.7 ± 5.1	-1.8 ± 4.5	0.99
Load response (10%)	4.4 ± 4.6	2.8 ± 3.9	4.6 ± 4.3	0.21
Mid stance (30%)	11.3 ± 4.2	10.8 ± 3.7	11.7 ± 4.5	0.65
Final stance (36%)	12.1 ± 4.6	12.3 ± 4.0	13.0 ± 4.7	0.69
Final stance (50%)	3.1 ± 6.6	5.3 ± 8.2	5.8 ± 8.0	0.31
Pre-swing (60%)	-12.1 ± 7.3	-11.2 ± 7.0	-5.9 ± 7.0	0.002 a, b
Initial swing (61%)	-12.1 ± 7.6	-11.6 ± 6.5	-6.1 ± 6.8	0.003 a, b
Initial swing (73%)	-0.2 ± 7.0	0.9 ± 5.6	1.6 ± 5.3	0.26
Mid-swing (80%)	2.6 ± 4.7	2.7 ± 4.9	3.3 ± 4.3	0.82
Mid swing (87%)	0.3 ± 5.8	1.4 ± 5.8	0.6 ± 4.8	0.74
Final swing (98%)	-2.8 ± 4.9	-3.0 ± 6.3	-2.9 ± 3.9	0.99

(%): Percentage of the gait cycle where the maximum displacement angle occurred. Positive values for ankle dorsiflexion; negative values for ankle plantarflexion. The expected maximum point for each movement is shown

^†Differences through one-way ANOVA test with Bonferroni post hoc test

^a Difference between normal weight and obesity groups

^b Difference between overweight and obesity groups

Correlation between weight increase and joint displacement angle at the knees and ankles

During the load response, we found a weak negative correlation between weight increase and right knee flexion-extension (r=-0.24; p = 0.019).

In addition, we observed a moderate positive correlation between weight increase and ankle plantarflexion on the same side in the pre-swing (r = 0.36; p < 0.01) and initial swing (r = 0.36; p < 0.01) but without correlation in the other cycle gait subphases.

In the left ankle, we observed a moderate positive correlation between weight increase and ankle plantarflexion movement in the pre-swing (r = 0.35; p < 0.01) and initial swing (r = 0.33; p < 0.01). We found no correlation in the other subphases.

Discussion

This study aimed to compare knee and ankle kinematics during walking between schoolchildren with overweight or obesity and schoolchildren with normal weight.

The main findings of this study were differences in knee and ankle kinematics during gait between patients with obesity and children with overweight and normal weight, mainly in knee adduction and ankle plantar flexion during the swing phase.

In this study, children with obesity showed greater knee adduction movement than the normal weight group, which could be associated with the increase in the patient’s thigh circumference and the displacement of their body’s center of mass. This adaptation mechanism promotes the transfer of the weight vector to the lateral side of the knee, with an increase in the adductor moment of this joint [17–24].

The increase in knee adduction moment is relevant because, as well as the relative medial movement of the knee, it could cause stress in the menisci, ligaments, and cartilage, and increase the risk of damage to these structures [25].

Furthermore, we also observed in the group with obesity differences in the kinematics of their knees in the sagittal and frontal planes compared to the normal weight group, previously described by Blakemore et al., 2013 [26], which could probably be explained as an “adaptation” generated by the reduction in the muscle contraction force of the vastus lateralis and tibialis anterior during the swing phase [12]. The power required without appropriate muscle strength could lead to musculoskeletal fatigue, and, subsequently, to a poor performance of children with overweight and obesity during gait [27].

Regarding our findings in the sagittal plane, we observed a weak correlation between the increase in the knee flexion angle and the increase in weight of the children studied, unlike what was reported by other authors, who reported greater knee flexion in children with obesity [9].

Moreover, regarding the movement of the ankles, we observed lesser plantar flexion in the groups of children with overweight and obesity groups compared to the normal group. As reported by other authors, this lesser plantar flexor movement predominated during the initial swing and not in the complete gait cycle [28].

Lower plantarflexion displacement could increase the participation of other joints, like hip flexion, as compensation for pulling the limb into the swing [6].

On the other hand, a decrease in the dorsiflexion angle has also been described as associated with schoolchildren’s gait speed and hip kinematics in the frontal plane [29]. In the schoolchildren with obesity in our study, this angle was smaller, even without reaching the neutral position; however, we did not analyze the gait speed.

Furthermore, among our most relevant findings was the positive correlation between weight and movement of the ankle joint, which suggests that the greater body weight increases the difficulty of controlling this movement during the swing with the need for deceleration and forward propulsion to achieve the next step [30].

This study’s strengths include a large sample with three comparison groups according to body weight, which allows for the establishment of differences between overweight and obesity. In addition, an analysis was carried out with a 3D motion-capture system, considered the gold standard. Unlike previous reports, we included both sexes and both sides of the body. We evaluated the individuals’ anthropometric measurements to normalize the measurements using movement analysis software, which favors the generalizability of our results and even though our sample of normal weight children was small, it provides valuable information for other groups of researchers who study kinematic alterations of the knee and ankle in populations with characteristics similar to ours.

However, we recognize that the limitations of this study are mainly related to the soft tissue artifact in the overweight and obesity groups, which affects the results for the frontal and axial axes of the evaluated joints [31]. We also did not analyze some spatiotemporal parameters that could have explained some of our findings. Besides, we did not explore the hip and pelvis movements necessary to analyze the knee and ankle movements. Another limitation of this study is that we do not have data on the measurement of knee and ankle valgus since this information was extracted from each participant’s file and in most cases, only the clinical diagnosis was reported. Finally, we did not analyze differences between males and females, which could be interesting for future studies.

Conclusions

The differences between the overweight and obese groups and the group with normal weight suggest that schoolchildren’s increased body weight causes kinematic modifications in their knees and ankles while walking.

Authors’ contributions

CGC, JLAA, and XTC; wrote the main manuscript and methodology, XTC, JLAA formal analysis and investigation, CGC, resources and data curation, XTC and IJVM; wrote the original draft, CGC, CMSM, XTC, HMG; writing—review and editing, CGC, XTC, IJVM and CMSM; supervision, CGC; project administration, CGC, HMG; funding acquisition.

Funding

The authors did not receive support from any organization for the submitted work.

Data availability

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study is part of the protocol approved by the Ethics and Research Committees of the Hospital Infantil de México Federico Gómez in Mexico City, with registration numberHIM-2020-042, and all our investigations with human subjects were conducted inaccordance with the Declaration of Helsinki. Informed consent from parents or informedassent from children was requested before participation, depending on the children age.

Consent of publication

Not applicable.

Competing interests

The authors declare no competing interests.