BJMB
Brazilian Journal of Motor Behavior
Research Notes
!
Leopoldo et al.
2023
VOL.17
N.5
222 of 227
Effects of anxiety, visual target predictability and pain on gaze behavior during a
visuomotor task
MAYSA P. G. LEOPOLDO
1
| CASSIO M. MEIRA JUNIOR
1
| RENATA M. SILVA
1
| CARLOS E. B. RODEGUER
1
|
MARCIO F. GOETHEL
2
| FERNANDO H. MAGALHÃES
1
| ULYSSES F. ERVILHA
1
1
University of Sao Paulo, School of Arts, Sciences, and Humanities, São Paulo, SP, Brazil
2
University of Porto, Faculty of Sport, Porto, Portugal
Correspondence to:!Cassio M. Meira Jr.
Av. Arlindo Bettio 1000, São Paulo, SP 03828-000 Brazil.
email: cmj@usp.br
https://doi.org/10.20338/bjmb.v17i5.371
HIGHLIGHTS
• High-anxious individuals exhibit greater cognitive effort
compared to low-anxious individuals during a postural
control task with visual pursuit.
• Minimal visual attention is directed toward a
deterministic, predictable target during a postural control
task with visual pursuit.
• Acute muscle pain and placebo muscle pain produce
similar effects on gaze behavior during a postural control
task with visual pursuit.
ABBREVIATIONS
CTGT Continuous time of gaze on target
M Means
PVD Pupil diameter variability
SD Standard deviations
STAI/IDATE State and Trait Anxiety Inventory
PUBLICATION DATA
Received 15 06 2023
Accepted 21 08 2023
Published 30 09 2023
BACKGROUND: Attention and cognitive effort during postural control can be influenced by
under threatening situations, such as pain, particularly in anxious individuals when visually
tracking a target.
AIM: This study aimed to investigate the effects of anxiety, visual target type, and acute
lumbar muscle pain on gaze behavior during a visual pursuit/postural control task.
METHOD: Nine young adult participants underwent testing over three time periods: (1) pre-
infusion, (2) intramuscular (multifidus muscles) infusion of hypertonic (acute pain)/isotonic
(placebo) solution, and (3) post-infusion (after 40 minutes after pain is vanished). The two
sessions were separated by one week, in a counterbalanced order. During each session,
participants performed a postural control task with visual pursuit, focusing on three targets
(fixed, stochastic, and deterministic - 3 trials per target) while wearing an eye-tracker.
STAI/IDATE was used to assess the participants' level of state anxiety: n=4 high-anxious and
n=5 low-anxious. Continuous time of gaze on target (CTGT) and pupil diameter variability
(PDV) were grouped into blocks of 3 trials.
RESULTS: High-anxious participants exhibited greater variability in PDV during the infusion
period. The deterministic target required less visual attention (shorter CTGT) compared to the
fixed and stochastic targets. Both injected solutions (hypertonic and isotonic) had similar
effects on CTGT and PDV.
CONCLUSION: During the postural control task with visual pursuit: (1) high-anxious
participants exerted greater cognitive effort, (2) participants deflected visual attention on the
deterministic target, (3) acute and placebo muscle pain did not affect visual attention and
cognitive effort.
KEYWORDS: Posture | Lumbar Pain | Visual Search | Motor Control | Anxious
INTRODUCTION
Postural control is a complex motor function influenced by the integration of sensory information and cognitive processes. The
perception of environmental cues plays a pivotal role in shaping postural responses. For instance, research has shown that body sway
amplitude is influenced by the coordination of visual information, as observed during saccadic eye movements (rapid, simultaneous shifts
of gaze between fixation points)
1-3
. Additionally, postural oscillations can be modulated by the visual tracking of moving targets within the
surroundings
4
. Beyond visual cues, factors such as acute pain, psychological states like anxiety, and threat perception are considered
significant contributors to the intricate interplay between perception and postural control
5-8
. Notably, anxiety, characterized by heightened
vigilance and responsiveness to potential threats, has been associated with physiological changes such as increased heart rate,
perspiration, and enhanced attention to threat-related stimuli
9-11
. These effects extend to postural adjustments and gaze behaviors,
reflecting the comprehensive influence of emotional states on motor coordination
12,13
.
Given the intricate relationship between emotional states, sensory processing, and postural adjustments, it becomes essential
to investigate how individuals with varying levels of anxiety interpret and respond to environmental cues during postural tasks.
BJMB! ! ! ! ! ! ! ! !
Brazilian(Journal(of(Motor(Behavior(
(
Leopoldo et al.
2023
VOL.17
N.5
223 of 227
Research Notes
Furthermore, in the context of pain perception, the interplay between anxiety and pain can modulate attentional focus and cognitive effort,
thereby impacting postural responses. To gain insights into these multifaceted interactions, this study focuses on examining the effects of
anxiety, visual target characteristics, and acute lumbar muscle pain on gaze behavior during a visuomotor task. For this purpose, we
evaluated gaze time on target and variability of pupil diameter, as these variables have been used to infer visual attention and cognitive
effort, respectively
14-16
. More specifically, two critical variables were employed: (A) continuous time of gaze on target (CTGT) and (B)
pupil diameter variability (PDV). The former, CTGT, represents the time duration an individual fixates on the visual target, reflecting the
allocation of visual attention to task-relevant information, while the latter, PDV, captures the variability in pupillary diameter, serving as a
proxy for cognitive effort and engagement during the task
17-19
.
By manipulating three key independent variables—(A) pain induction (hypertonic vs. isotonic solution) in lumbar erector
muscles, (B) visual target types (fixed vs. deterministic vs. stochastic), and (C) levels of state anxiety (high vs. low anxious participants)
— we aimed to address three specific hypotheses during the visuomotor task: (a) Injection of hypertonic solution into the lumbar region
was expected to lead to shorter CTGT and greater PDV as compared to isotonic solution injection, (b) CTGT was expected to be
prolonged and PDV to be reduced when fixating on deterministic and stochastic targets as opposed to a fixed target, (c) High-anxious
participants were expected to exhibit shorter CTGT and higher PDV as compared to their low-anxious counterparts. Through this
investigation, we strove to shed light on the complex interactions between emotional states, sensory processing, and motor responses in
postural control tasks. By elucidating the role of anxiety and pain perception in shaping attentional allocation and cognitive engagement,
we aimed to provide a comprehensive understanding of the interwoven nature of motor behavior and cognitive processing.
METHODS
Participants
We intentionally selected thirteen healthy volunteers, but three did not finish the experiment and the data of one participant
were not registered. The nine participants included in the analysis, one female and eight males, aged between 18 and 35 years (mean
age 24.22 ± 5.63 years), with normal vision, who reported no fear of needles, no lower limb surgery within the previous six months, no
use of antidepressant or anxiolytic medication within the previous three months, no pain on the day of the experiment or pain lasting more
than three consecutive days within a six-week period before the experiment, and no high-intensity exercise within two days before the
experiment. All participants received prior information about the procedures and risks and signed an informed consent form. The project
was approved by the University Ethics Committee (CAAE: 91074218.2.0000.5390).
Procedure
The State and Trait Anxiety Inventory (STAI; IDATE - Brazilian version)
17
was used to evaluate the participants' state anxiety
levels, which were determined based on the sample median (n=9). The cutoff point was set at 45 (the scale ranges from 20 to 80 points),
classifying four participants as high-anxious and five participants as low-anxious. Data collection with each individual consisted of two
sessions (days), with a one-week interval between them. On day 1, after signing the consent form and completing the STAI/IDATE
questionnaire, each participant performed the visuomotor task, which included three blocks of nine trials, each lasting 60 seconds. The
first block of trials (series 1 - pre-infusion) started without any solution injection. Series 2 (infusion 1) involved intramuscular injection of 2
ml of solution (6% hypertonic saline or 0.9% isotonic) into the multifidus muscles at L4, 2 cm laterally from the spinous process.
Immediately after the infusion, block 2 (series 2a - immediately after the first infusion) began and was divided into two parts: the first part
consisted of five trials, and the second part consisted of four trials. After the fifth trial, the second infusion (infusion 2 and series 2b) of an
intramuscular injection of 1 ml of solution (hypertonic or isotonic) into the same muscles was administered. Immediately after, the
remaining four trials were performed. The order of the conditions was counterbalanced among the participants, with five in the hypertonic-
isotonic condition and four in the isotonic-hypertonic condition.
Forty minutes after pain was vanished, the third block of nine trials (series 3) took place without any injection. The infusions
were administered by an experienced researcher. In the pain condition, as one infusion did not provide enough pain for the entire
session, saline solution (hypertonic) was administered twice: the first infusion containing 2 ml and the second 1 ml. Pain was quantified
before and immediately after infusion of solution and was stopped as soon as the participant no longer reported pain. The average pain
value declared on a visual analogue (posteriori digitally converted on a scale of 0-10) was 5.5, which is considered adequate
6
.
Participants reported feeling moderate pain for approximately 12 minutes.
The visual pursuit task was developed in Matlab
©
software. Participants were instructed to keep their eyes fixed on the targets
for 60 seconds. For analysis, the first and last five minutes of each trial were excluded, resulting in a trial duration of 50 seconds. The
nine trials of each block were presented randomly, with three trials in each pattern: stochastic - red target (2 cm diameter) moving
randomly from left to right and up and down at a frequency of 0.5 Hz; deterministic - red target (2 cm diameter) moving constantly from
left to right and up and down at a frequency of 0.5 Hz; fixed (static) - black target (2 cm diameter) in the center of the screen that
remained motionless for 60 seconds.
BJMB! ! ! ! ! ! ! ! !
Brazilian(Journal(of(Motor(Behavior(
(
Leopoldo et al.
2023
VOL.17
N.5
224 of 227
Research Notes
During the visual pursuit task, participants also performed the postural control task by standing upright in a bipedal orthostatic
posture, facing the projection screen, and looking at the center of the screen at eye height. Participants stood barefoot with their feet
separated by hip width. The gaze behavior of participants was tracked using the Mobile Eye XG (Applied Science Laboratories) eye
tracking system. This system operates within a safe range of infrared illumination. Before starting the acquisition of gaze behavior data,
the equipment was calibrated to adjust to the participant's eye.
Data analysis
The eye data were analyzed using ASL Results software. Continuous tracking gaze time (CTGT) on target was obtained by
tracking the targets frame by frame on each trial. For analysis, the sum of the three trials was calculated, resulting in a total of 150
seconds (50 seconds per trial). Pupil diameter variability (PDV) was the standard deviation in pixels for each trial. Data were organized in
blocks of three trials. Anxiety levels were determined based on the median of the sample, with a cutoff point of 45 (20-80). Participants
scoring above this cutoff were classified as high anxious (four participants), while those scoring below were classified as low anxious (five
participants).
Data were tabulated and organized in Microsoft Excel spreadsheets and analyzed using IBM SPSS software, version 24.
Descriptive analysis and the Shapiro-Wilk statistical test were conducted to examine the data distribution. Due to the non-normality of the
distributions, inferential analyses were performed using nonparametric techniques (Mann-Whitney tests for pain and anxiety
comparisons; Kruskal-Wallis tests for target comparisons). The significance level was set at 5% for all analyzes. We reported effect sizes
for the significant effects.
RESULTS
The variables continuous time of gaze on target (CTGT) and pupil diameter variability (PDV) were collected in two different
conditions (painful and placebo) and in three time windows, namely, pre-infusion (before the intramuscular infusion of hypertonic saline
solution/painful condition or isotonic saline solution/placebo condition), during-infusion (immediately after the infusion), and post-infusion
(thirty minutes after the pain has ceased completely). No significant differences were found for CTGT (Pre: U=179.5, p=0.1; During:
U=211, p=0.11; Post: U=284.5, p=0.942) and PDV (Pre: U=246, p=0.386; During: U=259, p=0.912; Post: U=253, p=0.49) (Table 1).
The deterministic pattern resulted in shorter CTGT compared to the fixed and stochastic targets during all three time periods
(Pre: X
2
=20.84, p=0.0001, ƞ
2
H=0.42; During: X
2
=15.53, p=0.0004, ƞ
2
H=0.3; Post: X
2
=21.36, p=0.0001, ƞ
2
H=0.44). No differences
among targets were found for PDV (Pre: X
2
=2.26, p=0.323; During: X
2
=3.23, p=0.199; Post: X
2
=0.7, p=0.704) (Table 2).
Data from high and low anxious participants are presented in Table 3. Although no differences were observed for CTGT (Pre:
U=189, p=0.146; During: U=277, p=0.82; Post: U=248, p=0.409), high-anxious participants exhibited greater variability in pupil diameter
during the infusion period compared to low-anxious participants (U=155, p=0.02, ƞ
2
=0.13). No significant effects were found for PDV in
the pre-infusion (U=270, p=0.71) or post-infusion (U=218, p=0.217) periods.
Table 1. Continuous time of gaze on target (CTGT) and pupil diameter variability (PDV) means (M) and standard deviations (SD) obtained during the
painful (hypertonic saline infusion) and placebo (isotonic saline infusion) conditions performed in three time windows: pre-infusion, during-infusion, and
post-infusion.
Isotonic
Hypertonic
M
SD
M
SD
CTGT (seconds)
Pre-infusion
83.24
14.24
88.77
13.56
During-infusion
80.79
15.05
81.47
17.70
Post-infusion
77.33
16.73
81.14
15.15
PDV (pixels)
Pre-infusion
22.88
06.93
20.13
8.62
During-infusion
22.88
17.88
25.94
10.89
Post-infusion
24.91
11.24
22.84
9.45
* pre-infusion (before the intramuscular infusion of hypertonic saline solution/painful condition or isotonic saline solution/placebo condition), during-
infusion (immediately after the infusion), and post-infusion (thirty minutes after the pain has ceased completely).
BJMB! ! ! ! ! ! ! ! !
Brazilian(Journal(of(Motor(Behavior(
(
Leopoldo et al.
2023
VOL.17
N.5
225 of 227
Research Notes
Table 2. Continuous time of gaze on target (CTGT) and pupil diameter variability (PDV) means (M) and standard deviations (SD) according to the type
of visual target used in the experiment, whether fixed, deterministic or stochastic in three time windows: pre-infusion, during-infusion, and post-infusion.
Fixed
Deterministic
Stocastic
M
SD
M
SD
M
SD
CTGT (seconds)
Pre-infusion
93.85
10.43
73.25 *
12.85
90.92
8.86
During-infusion
89.36
12.81
69.08 *
13.33
84.95
15.61
Post-infusion
90.78
11.90
64.64 *
9.73
82.29
13.31
PDV (pixels)
Pre-infusion
19.75
8.66
22.35
7.73
22.15
7.59
During-infusion
22.92
9.15
34.27
20.54
24.49
7.65
Post-infusion
22.75
10.76
22.05
5.36
26.65
13.05
* significant differences. p <0.001.
* pre-infusion (before the intramuscular infusion of hypertonic saline solution/painful condition or isotonic saline solution/placebo condition), during-
infusion (immediately after the infusion), and post-infusion (thirty minutes after the pain has ceased completely).
Table 3. Continuous time of gaze on target (CTGT) and pupil diameter variability (PDV) means (M) and standard deviations (SD) of high and low anxiety
individuals in three time windows: pre-infusion, during-infusion, and post-infusion.
Low Anxiety
High Anxiety
M
SD
M
SD
CTGT (seconds)
Pre-infusion
85.44
14.71
86.58
13.63
During-infusion
82.82
16.49
79.44
16.19
Post-infusion
79.75
17.23
78.72
14.82
PDV (pixels)
Pre-infusion
20.11
7.78
22.78
7.99
During-infusion
24.90
16.35
29.80
12.41
Post-infusion
21.40
7.30
26.21
12.24
* significant differences. p <0.03.
* pre-infusion (before the intramuscular infusion of hypertonic saline solution/painful condition or isotonic saline solution/placebo condition), during-
infusion (immediately after the infusion), and post-infusion (thirty minutes after the pain has ceased completely).
DISCUSSION
The analyses of CTGT and PDV revealed no significant differences between painful and placebo conditions, which contradicts
the hypothesis that the injection of a hypertonic solution affects visual attention
5
and cognitive effort compared to an isotonic solution
injection. It is worth noting that both solutions appeared to have led to increased PDV, indicating greater cognitive effort required to
perform the visuomotor task. One possible explanation for this finding is that the injection itself may have created a sense of threat for the
participants.
The hypothesis that the fixed target would increase visual attention during pain was refuted, as the findings indicated that CTGT
was shorter when participants gazed at the deterministic target compared to when they focused on the fixed and stochastic targets in all
three time periods. It can be inferred that in all three time periods, the deterministic condition attracted less visual attention, likely because
participants perceived an up-and-down pattern of the target and employed an energy-saving strategy by anchoring their gaze near the
center of the screen based on visual prediction
4
. The shorter CTGT for the deterministic target may also be associated with easier
pattern recognition, which could reduce focal vision
16
. Our findings also do not support the notion that the unpredictability of the
stochastic target would result in shorter CTGT
18
. The PDV analysis revealed no significant differences among targets, suggesting that
the cognitive effort employed to perform the task was similar in all three time periods, regardless of target type. This finding contradicts
studies on neutral and exciting scene recognition situations and complex cognitive task execution
19,20
.
The hypothesis that high-anxious participants during the visuomotor task would exhibit shorter CTGT compared to low-anxious
participants was refuted, as both low-anxious and high-anxious participants showed similar values across the three time periods.
Individuals with high levels of state anxiety who perform tasks under threat tend to compensate for this stress process by increasing heart
rate, perspiration, and altering gaze behavior
10,11
. These psychophysiological variables should be investigated in future studies. Our
findings support the hypothesis that high-anxious participants would show higher PDV compared to low-anxious participants after
receiving the injection. This provides evidence that high-anxious individuals are more likely to exhibit weaker gaze behavior parameters
when confronted with threatening stimuli
14,15
.
BJMB! ! ! ! ! ! ! ! !
Brazilian(Journal(of(Motor(Behavior(
(
Leopoldo et al.
2023
VOL.17
N.5
226 of 227
Research Notes
The main limitation of this study was the small sample size due to (1) refusal to initiate the experiment (owing to needle fear)
and (2) withdrawal from data collection (as a result of the prolonged data collection sessions and the invasive technique of substance
injections). Also, we highlight as a limitation the loss of gaze behavior data during some trials owing to blinks or abrupt head movements.
CONCLUSION
In conclusion, the findings of the present study indicate that during a visuomotor task: (1) high-anxious individuals exerted
greater cognitive effort compared to low-anxious individuals, (2) the deterministic, predictable target attracted less visual attention, and
(3) acute and placebo muscle pain did not affect visual attention and cognitive effort.
REFERENCES
1. Aguiar SA, Polastri PF, Godoi D, Moraes R, Barela JA, Rodrigues ST., Effects of saccadic eye movements on postural control in older adults.
Psychol & Neurosci. 2015;8(1):19-27. Doi: 10.1037/h0100352
2. Rodrigues ST, Polastri, PF, Carvalho JC, Barela, JA, Moraes R, Barbieri FA, Saccadic and smooth pursuit eye movements attenuate postural sway
similarly. Neurosci Lett. 2015;584:292-295. Doi: 10.1016/j.neulet.2014.10.045
3. Rodrigues ST, Polastri P, Godoi D, Moraes R, Barela JA. Effects of saccadic eye movements on postural control stabilization. Motriz.
2013;19(3):614-619. Doi: 10.1590/S1980-65742013000300012
4. Carvalho LR, Almeida LG, Aspectos sensoriais e cognitivos do controle postural. [Sensorial and Cognitive Aspects of Postural Control]. Rev.
Neurociências. 2009;17(2):156–160. Doi: 10.34024/rnc.2009.v17.8576
5. Fashler SR, Katz J. Keeping an eye on pain: Investigating visual attention biases in individuals with chronic pain using eye-tracking methodology. J.
Pain Res. 2016;9:551–561. Doi: 10.2147/JPR.S104268
6. Ervilha UF, Farina D, Arendt-Nielsen L, Graven-Nielsen T. Experimental muscle pain changes motor control strategies in dynamic contractions. Exp.
Brain Res. 2005;164(2):215–224. Doi: 10.1007/s00221-005-2244-7
7. Hirata RP, Ervilha UF, Arendt-Nielsen L, Graven-Nielsen T. Experimental muscle pain challenges the postural stability during quiet stance and
unexpected posture perturbation. J Pain. 2011;12(8): 911–919. Doi: 10.1016/j.jpain.2011.02.356
8. Soares JC, Weber P, Trevisan ME, Trevisan CM, Mota CB, Rossi AG. Influência da dor no controle postural de mulheres com dor cervical
[Influence of pain on postural control in women with neck pain]. Rev Bras Cineantrop Desempenho Hum. 2013;15(3):371–381 Doi: 10.5007/1980-
0037.2013v15n3p371
9. Causer J, Holmes PS, Smith NC, Williams AM. Anxiety, movement kinematics, and visual attention in elite-level performers. Emotion.
2011;11(3):595-602. Doi: 10.1037/a0023225.
10. Wilson M. From processing efficiency to attentional control: a mechanistic account of the anxiety–performance relationship. Int. Rev. Sport Exercise
Psychol. 2008;1(2):184-201. Doi: 10.1080/17509840802400787
11. Eysenck MW, Derakshan N, Santos R, Calvo MG. Anxiety and cognitive performance: attentional control theory. Emotion. 2007;7(2):336-352. Doi:
10.1037/1528-3542.7.2.336
12. Pergamin-Hight L, Naim R, Bakermans-Kranenburg MJ, van Ijzendoorn MH, Bar-Haim Y. Content specificity of attention bias to threat in anxiety
disorders: a meta-analysis. Clin Psychol Rev. 2015;35:10-8 Doi: 10.1016/j.cpr.2014.10.005.
13. Rosa GM. A influência da ansiedade sobre o equilíbrio postural em estudantes universitários [The influence of anxiety on postural balance of
university students]. Fisioterapia Brasil. 2011;12(4):244-249. Doi: 10.33233/fb.v12i4.919
14. Sege CT, Bradley MM, Lang PJ. Motivated action: Pupil diameter during active coping. Biol Psychol. 2020;153:107885. Doi:
10.1016/j.biopsycho.2020.107885
15. Shechner T, Jarcho JM, Wong S, Leibenluft E, Pine DS, Nelson EE. Threats, rewards, and attention deployment in anxious youth and adults: An
eye tracking study. Biol Psychol. 2017;122:121-129. Doi: 10.1016/j.biopsycho.2015.10.004
16. Williams AM, Davids K, Williams JG. Visual perception and action in sport. e-book, Taylor & Francis, London, 2005.
17. Fioravanti ACM, Santos LF, Maissonette S, Cruz APM, Landeira-Fernandez J. Avaliação da estrutura fatorial da Escala de Ansiedade-Traço do
IDATE. [Evaluation of the factorial structure of the trait anxiety scale of the STAI]. Aval Psicol. 2006;5(2):217-224.
http://pepsic.bvsalud.org/scielo.php?script=sci_arttext&pid=S1677-04712006000200011&lng=pt
18. Caminha LQ, Azevedo Neto RM, Teixeira LA. Use of peripheral vision in the interception of moving targets. Motriz, 2010;16(4):885-868.
Doi.org/10.5016/1980-6574.2010v16n4p858
19. Henderson RR, Bradley MM, Lang PJ. Emotional imagery and pupil diameter. Psychophysiol., 2018;55(6):e13050. Doi: 10.1111/psyp.13050
20. Thigpen NN, Bradley MM, Keil A. Assessing the relationship between pupil diameter and visuocortical activity. J. Vision. 2018;18(6):7. Doi:
10.1167/18.6.7
BJMB! ! ! ! ! ! ! ! !
Brazilian(Journal(of(Motor(Behavior(
(
Leopoldo et al.
2023
VOL.17
N.5
227 of 227
Research Notes
ACKNOWLEDGMENTS
We are thankful to the participants and to Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brasil) for
the grant to the first author.
Citation: Leopoldo MPG, Meira Junior CM, Silva RM, Rodeguer CEB, Goethel MF, Magalhães FH, Ervilha UF. (2023).!Effects of anxiety, visual target predictability and
pain on gaze behavior during a visuomotor task. Brazilian Journal of Motor Behavior, 17(5):222-227.
Editor-in-chief: Dr Fabio Augusto Barbieri - São Paulo State University (UNESP), Bauru, SP, Brazil. !
Associate editors: Dr José Angelo Barela - São Paulo State University (UNESP), Rio Claro, SP, Brazil; Dr Natalia Madalena Rinaldi - Federal University of Espírito Santo
(UFES), Vitória, ES, Brazil; Dr Renato de Moraes – University of São Paulo (USP), Ribeirão Preto, SP, Brazil.!
Copyright:© 2023 Leopoldo, Junior, Silva, Rodeguer, Goethel, Magalhães and Ervilha and BJMB. This is an open-access article distributed under the terms of the
Creative Commons Attribution-Non Commercial-No Derivatives 4.0 International License which permits unrestricted use, distribution, and reproduction in any medium,
provided the original author and source are credited.
Funding: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brasil) for the grant to the first author.
Competing interests: The authors have declared that no competing interests exist.
DOI:!https://doi.org/10.20338/bjmb.v17i5.371