Keywords
sleep quality - athletic injuries - testosterone - cortisol - sports training
Introduction
Many athletes initiate their participation in competitions and engage in sport-specific
training in adolescence.[1] Monitoring training-induced psychobiological adaptations in young athletes is essential
to increase longevity in sports participation. Additionally, monitoring training-related
variables such as sleep, injuries, internal and external loads, and recovery is essential
for adequate training prescription during a competitive season.[2]
[3]
The risk of injuries in sports is multifactorial and includes nonlinear relationships
between different factors such as athlete biomechanics, training characteristics,
recovery, sport specificity, and physiological aspects.[4] Sleep debt is related to injury risk, changes in mood, and athletes' physiology
and performance.[5] Sleep plays a fundamental role in recovery, cognition, energy restoration, and metabolic
functions. Moreover, it has an important role in the development and maturation processes
of adolescents.[6] The American Academy of Sleep Medicine recommends that adolescents should sleep
between 8 and 10 hours per night.[7] However, factors such as a delayed circadian rhythm typical of puberty, school activities,
use of electronic devices at night, caffeine intake, and training schedules can compromise
sleep in adolescent athletes.[8]
Thus, adolescent track and field athletes, for example, have a total sleep time worse
than recommended,[9] and the quantity and quality of sleep have been associated with musculoskeletal
injuries.[10] In contrast, sleep complaints have been associated with a history of occurrence,
recurrence, and severity of injuries in this population.[9]
In addition to sleep, Foster et al.[11] observed an association between excessive training load and the occurrence of injuries.
When excessive training demands are associated with insufficient recovery, athletes
may suffer preventable injuries or even experience overtraining syndrome.[12] Self-reported scales and hormonal measures can be used to quantify an athlete's
response to training. There is evidence that hormonal concentration oscillates concomitantly
with changes in training volume and intensity. The measurement of testosterone and
cortisol concentrations and the calculation of the testosterone/cortisol (T/C) ratio
have been used as endocrine stress biomarkers in the sports medicine context and a
marker of the anabolic/catabolic balance.[13] Tyndall et al. observed increased cortisol and reduced testosterone levels in swimmers
subjected to intensity and volume variations at different training phases.[14]
Training loads are constantly adjusted during a season to promote optimized performance
in target competitions.[15] The preparatory phase is associated with the highest loads (i.e., training intensity
and volume), whereas the competitive phase is associated with sport-specific stimuli.
Thus, the individual monitoring of psychophysiological and performance adaptations
at different stages of training can help practitioners implement strategies to prevent
injuries, identify flaws in training planning, and take care of the athletes' physical
and mental health.
Our objective was to compare the quantity and quality of sleep, musculoskeletal injuries,
and the concentrations of testosterone and cortisol (and their ratio) in different
training phases in adolescent athletes, and we also investigated whether there was
a relationship between injury occurrence, cortisol and testosterone levels, with sleep
quantity and quality.
Materials and Methods
Study Design
This was a prospective cohort study. The protocol was approved by the Research Ethics
Committee of the authors' institution (64492016.8.0000.5149). After being informed
about the study (objectives, procedures, and risks), the athletes were invited to
participate, and those who accepted signed a written informed consent form.
Participants
The sample was obtained by convenience and 30 track and field athletes of both sexes,
aged between 12 and 21 years - considered adolescents according to the American Academy
of Pediatrics[16]–were recruited at the Sports Training Center of the authors' institution. They were
runners (i.e., 100, 200, and 400 m) and participated in competitions at local, regional,
and/or national levels at least in the six months before data collection.
Procedures
This was a study with a 6-month follow-up. Initially, anthropometric data were collected.
In female athletes, information about the date of last menstruation and whether they
were using birth control pills were also collected. The athletes were followed through
6 months, and we monitored the occurrence of musculoskeletal injuries resulting from
their sports practice. Injury data were collected from the Physical Therapy Department
of the Sports Training Center. Athletes were evaluated by physical therapists and
a questionnaire to describe the characteristics of these injuries was provided to
them. Sleep and hormone concentrations were monitored at three specific time points
(using the same procedures) that corresponded to different training phases, as described
below:
Preparatory phase: from August to early October and corresponded to the preparation period for the competitions.
According to the coaches, the training sessions had a high volume and intensity.
Competitive phase: from October to December. According to the coaches, the principle of volume versus
intensity interdependence was emphasized, and the training volume was reduced while
the intensity was kept high.
Post-Competitive phase: from December to January. According to the coaches, both the training volume and
intensity were reduced, with sessions being characterized by regenerative and more
recreational activities.
Actigraphy
Actigraphy has been considered the gold standard method to evaluate the sleep-wake
cycle in the general population.[17] It is the less invasive way to objectively measure sleep quality and quantity, with
minimal impact on athletes' sleep and training habits.[17] The participants were instructed to wear an actigraph (Philips Respironics®, Andover,
MA, USA) on the non-dominant wrist and to use it for 10 days, maintaining their usual
sleep habits during this period. They received the actigraph on a Thursday to englobe
two weekends. Simultaneously, the athletes filled out a sleep diary to facilitate
the data analysis. The data was analyzed using the Action-W version 02 software (Ambulatory
Monitoring Inc®, Andover, MA, USA). We extracted the following variables: time awake
(TA), total sleep time (TST), sleep onset latency (SOL), sleep efficiency (SE), and
wakefulness after sleep onset (WASO).
Hormonal Collection and Analysis
Hormonal collection for analysis of free testosterone, cortisol, and testosterone/cortisol
(T/C) ratio was done through saliva samples collected during five consecutive days
of training at 3:00 pm, 30 minutes before each training session. Samples were always collected at the same
time of day in the three training phases to avoid circadian rhythm interference. The
athletes were instructed not to feed or ingest fluids except water for 30 minutes
and to remain at rest for one hour before collection. They were also instructed to
avoid brushing their teeth at least two hours before collection and to perform gentle
mouthwash immediately before collection. Salivette ® tubes with a synthetic swab for
cortisol determination and a cotton swab for testosterone analysis were used. The
athletes were instructed to remain for at least 60 seconds with the swab in their
mouth stimulating salivation before returning it to the tube.[18] The collection of saliva samples in women was performed between the first to the
14th day of the menstrual cycle (follicular phase) following the same pattern of collection
days and times for males.
After collection, the Salivette® tubes were centrifuged (3600 rpm) for 20 minutes
at 4°C.[19] The salivary content was then removed, stored in sterile falcon tubes, and frozen
at -20°C until posterior analysis. After the end of the three training phases, the
samples were thawed and centrifuged again to remove any impurities that could be present.
Free testosterone concentration was analyzed using the enzyme-linked immunosorbent
assay (ELISA), with a detection limit ranging from 7.00 to 1500.00 ng/dL (0.24 to
52.05 nmol/L). Cortisol concentration analysis was performed using the electrochemiluminescence
method, with a detection limit ranging from 1.5 nmol/L to 1750 nmol/L (0.054 to 63.4 μg/dL).[20]
Musculoskeletal Injury Monitoring
To record musculoskeletal injuries, we used an adapted version of the questionnaire
developed by Fuller et al., which was used in a consensus manuscript by the Fédération
Internationale de Football Association (FIFA).[21] The following information was recorded during the 6 months of monitoring: date,
classification, location (i.e., body part), diagnosis, recurrences, mechanism, withdrawal
duration from training and competition, physiotherapy treatment duration, and when
the injury occurred (i.e., training or competition). We followed the injury definition
by Fuller et al.: “any physical complaint reported by an athlete that resulted from
a training or competition, regardless of the need for medical attention or withdrawal
from activities”[21].
We asked the athletes to complete a form daily during the six months to record all
the variables mentioned above. However, when the athletes reported an injury, they
always underwent a physiotherapy evaluation in the Physical Therapy Department to
exclude complaints that were not classified as injuries according to our definition
(e.g., delayed onset muscle soreness).
Subjective Rating of Perceived Exertion (RPE)
RPE was obtained daily, 30 minutes after each training session of each trained day.[11] Athletes indicated the level of difficulty experienced during the session, ranging
from 0 (rest condition) to 10 (greatest physical effort). We calculated the internal
training load by multiplying the RPE score (intensity) by the session duration expressed
in minutes (volume) of every day, and the results were expressed in arbitrary units.
Total Recovery Quality (TQR) Scale
Before each training session, the athletes answered the question “How do you feel
about your recovery?” to verify their perceived recovery related to the previous training
session. The answer is based on a scale ranging from 6 (nothing recovered) to 20 (fully
recovered).[22]
Statistical Analysis
The Shapiro-Wilk test was performed to verify the data normality. Means and SDs were
used to describe sleep characteristics. One-way analyses of variance (ANOVAs) with
repeated measures were used to compare the variables between training phases. When
statistically significant differences were found, Fisher's Least Significant Difference
(LSD) post hoc test was performed to compare pairs of means. For data that did not
present normal distribution, Friedman's ANOVAs on Ranks with repeated measures were
performed, using Tukey's post hoc tests whenever applicable. The Spearman test was
used to analyze the correlation between sleep, injury, and hormone levels. The significance
level was set at α < 0.05. The SPSS software (version 20.0, Chicago, Illinois, US)
was used.
Results
Of the 30 athletes recruited, 19 (13 male and 6 female) were included in our sample.
Eleven athletes who did not attend all three evaluations were excluded from our sample.
The characteristics of our sample are detailed in [Table 1]. There were statistically significant differences in the body mass and height between
boys and girls (p = 0.001 and p = 0.038, respectively), though no differences were observed for age, BMI, and training
frequency. None of the female athletes used birth control pills during the study.
Table 1
Sample characteristics
|
Age (years)
|
Weight (kg)
|
Height (m)
|
BMI (kg/m2)
|
Weekly training frequency (days)
|
|
Total (n = 19)
|
16.8 ± 2.7
|
62.3 ± 8.3
|
1.72 ± 0.11
|
21.1 ± 1.4
|
4.6 ± 0.7
|
|
Males (n = 13)
|
17.0 ± 2.9
|
65.4 ± 8.3
|
1.75 ± 0.11
|
21.2 ± 1.6
|
4.4 ± 0.8
|
|
Females (n = 6)
|
16.6 ± 2.2
|
55.6 ± 2.1
|
1.64 ± 0.05
|
20.6 ± 0.8
|
5.0 ± 0.0
|
Abbreviation: BMI, body mass index.
Values are presented as means ± SD.
All the athletes combined presented 25 injuries and attended 55 physiotherapy sessions
during the 6-month follow-up, where fourteen athletes suffered injuries and five of
them had no injuries. [Supplementary Table S1] (online only) describes individual data for injury variables. During the preparatory
phase, the athletes presented the highest incidence of injuries,[12] of which resulted in 41 physiotherapy sessions. In the competitive phase, 8 injuries
and 8 sessions were recorded, whereas, during the post-competitive phase, the athletes
suffered 5 injuries and attended 6 physiotherapy sessions. [Table 2] shows the characteristics of injuries, such as the need to withdraw from sports
practice, the body part and side affected, and the mechanism of injuries (overuse
or trauma). It is noteworthy that most of the athletes (92%) did not need to withdraw
from training due to injury. The most injured body regions were the thigh (28%) and
ankle (28%), and the most common injury mechanism was muscle overuse (92%).
Table 2
Injury characteristics during the whole period evaluated and according to the training
phase
|
Injury characteristics
|
Whole period
|
Preparatory Phase
|
Competitive Phase
|
Post-Competitive
|
|
Withdrawal from sports practice
|
|
|
|
|
|
Yes
|
1
|
1
|
0
|
0
|
|
Partial
|
1
|
1
|
0
|
0
|
|
No (only physical therapy treatment)
|
23
|
10
|
8
|
5
|
|
Body part
|
|
|
|
|
|
Lumbar/sacrum/pelvis
|
4
|
2
|
2
|
0
|
|
Shoulder/collarbone
|
2
|
0
|
2
|
0
|
|
Handle
|
1
|
0
|
0
|
1
|
|
Hip/groin
|
1
|
1
|
0
|
0
|
|
Thigh
|
7
|
4
|
2
|
1
|
|
Knee
|
1
|
1
|
0
|
0
|
|
Achilles tendon/ Calf
|
7
|
3
|
2
|
2
|
|
Ankle
|
2
|
1
|
0
|
1
|
|
Body side
|
|
|
|
|
|
Dominant
|
10
|
6
|
1
|
3
|
|
Non-dominant
|
3
|
1
|
1
|
1
|
|
Bilateral
|
7
|
3
|
3
|
1
|
|
Does not apply
|
5
|
2
|
3
|
0
|
|
Injury mechanism
|
|
|
|
|
|
Stress fracture
|
1
|
1
|
0
|
0
|
|
Ligament sprain/injury
|
1
|
0
|
0
|
1
|
|
Muscle strain/tension/injury/cramp
|
22
|
10
|
8
|
4
|
|
Tendon injury/rupture or tendinosis
|
1
|
1
|
0
|
0
|
|
Over-use
|
23
|
11
|
8
|
4
|
|
Trauma
|
2
|
1
|
0
|
1
|
Values are presented as absolute frequency.
[Table 3] shows the comparison of sleep variables, perceived recovery and exertion, and injuries
between the three training phases. In the post-competitive phase, the TST (7.5 ± 0.9 hour)
was statistically higher than in the competitive phase (6.9 ± 0.7 hour) (p = 0.01); and TA was statistically lower in the competitive (15.9 ± 1.1) and post-competitive
(15.4 ± 1.0) phases (p = 0.004). In addition, the athletes presented lower WASO values in the competitive
(29.2 ± 9.9 minute) and post-competitive (36.1 ± 15.2 minute) phases compared with
the preparatory (46.0 ± 12.4 minute) phase (p = 0.00). There were no significant differences in the other variables analyzed.
Table 3
Comparison of sleep variables, perceived recovery and exertion, and injury incidence
during ten days of actigraph use in three different training phases (n = 19)
|
Preparatory Phase
|
Competitive Phase
|
Post-Competitive
|
P-value
|
|
TA (h)
|
15.9 ± 1.1
|
16.3 ± 1.1
|
15.4 ± 1.0*
|
0.004
|
|
TST (h)
|
7.2 ± 0.7
|
6.9 ± 0.7
|
7.5 ± 0.9*
|
0.012
|
|
SOL (min)
|
20.8 ± 10.4
|
24.1 ± 13.5
|
18.4 ± 13.0
|
0.201
|
|
WASO (min)
|
46.0 ± 12.4
|
29.2 ± 9.9 #
|
36.1 ± 15.2 #
|
0.001
|
|
SE (%)
|
82.5 ± 3.6
|
83.2 ± 6.5
|
84.2 ± 4.9
|
0.447
|
|
TRQ
|
5.55 ± 0.77
|
5.36 ± 0.63
|
5.66 ± 0.64
|
0.360
|
|
Internal Load
|
492.4 ± 88.8
|
447.6 ± 61.5
|
495.7 ± 148.3
|
0.343
|
Abbreviations: h, hours; min, minutes; SE, sleep efficiency; SOL, sleep onset latency;
TA, time awake; TRQ, total recovery quality; TST, total sleep time; WASO, wakefulness
after sleep onset.
Values are presented as means ± SD.
* Significantly differs from the Competitive Phase.
# Significantly differs from the Preparatory Phase.
[Table 4] shows the comparison of hormone levels between the three training phases. We only
found statistically significant differences in cortisol levels. There was a difference
without stratification by sex (p = 0.01) and among male athletes (p = 0.04). Post hoc analyses showed that cortisol levels were higher in the preparatory
phase than in the post-competitive phase (p = 0.02, all samples; p = 0.03, males). No differences were found for testosterone levels and T/C ratio.
Table 4
Comparison of hormonal variables between the three training phases in boys and girls
|
Preparatory Phase
|
Competitive Phase
|
Post-Competitive Phase
|
P
|
|
Cortisol (nmol/L)
|
|
All (
n
= 19)
|
1.99 ± 0,61
|
2.42 ± 1,12
|
2.81 ± 1.60
|
0.01*
|
|
Males (
n
= 13)
|
1.98 ± 0.59
|
2.36 ± 1.18
|
2.81 ± 1.39
|
0.04*
|
|
Females (
n
= 6)
|
2.01 ± 0.61
|
2.53 ± 1.08
|
2.79 ± 2.15
|
0.38
|
|
Testosterone (pg/mL)
|
|
All (
n
= 19)
|
67.25 ± 39.85
|
57.94 ± 31.68
|
56.86 ± 30,53
|
0.30
|
|
Males (
n
= 13)
|
87.19 ± 31.34
|
69.86 ± 31.64
|
72.08 ± 23.27
|
0.19
|
|
Females (
n
= 6)
|
24.05 ± 9.21
|
32.12 ± 6.53
|
23.89 ± 12.25
|
0.32
|
|
Testosterone/Cortisol Ratio
|
|
All (
n
= 19)
|
0.10 ± 0,06
|
0.08 ± 0,05
|
0.08 ± 0,06
|
0.09
|
|
Males (
n
= 13)
|
0.13 ± 0.05
|
0.10 ± 0.05
|
0.10 ± 0.05
|
0.09
|
|
Females (
n
= 6)
|
0.04 ± 0.02
|
0.04 ± 0.01
|
0.03 ± 0.02
|
0.76
|
Values are presented as means ± SD.
* = statistically significant.
No significant correlation was found between sleep, hormone levels, and incidence
of injuries in the different phases of training (p > 0.05).
Discussion
This study aimed to compare the quantity and quality of sleep, the incidence of musculoskeletal
injuries, and the concentrations of testosterone and cortisol in different training
stages in adolescent track and field athletes. The results showed that the mean TST
and SE were below the recommended (i.e., < 8h and 85% respectively) in all phases,
and SOL and WASO were within the recommended (≥ 30 minutes and 20 minutes respectively).[23] Moreover, the athletes presented higher TST and lower WASO in the post-competitive
phase, compared with the competitive and preparatory phases, respectively. These findings
corroborate previous studies showing that athletes, on average, sleep less than the
recommended.[24] We found a mean TST of 7h. Previous studies have shown that athletes who experience
sleep restriction may exhibit motor and cognitive deficits, altered mood, increased
reaction time, and fatigue.[25]
The athletes' sleep quality can be influenced by sports schedules. Previous studies
have found a poorer sleep quality in periods close to competition,[26] corroborating our findings. In the present study, a higher TST was observed in the
post-competition compared with the competitive phase, in addition to higher WASO in
the preparatory phase. However, the increase in TST in the post-competition phase
still does not correspond to what is recommended for adolescents.[23] The reduction in sleep quantity and quality in periods of preparation and competition
can be explained by factors such as anxiety related to competitions, long travel routines,
and training schedules.[27] It is common for athletes to present sleep restriction conditions, mainly due to
the higher intensity and volume of training.[28] Silva et al.[29] evaluated 146 Olympic athletes from the Brazilian national team, during the preparatory
phase for the RIO 2016 Games and recorded a total of 250 sleep complaints. Furthermore,
the athlete and adolescent population is subject to specific demands that combine
school and sports calendars, which can have an impact on sleep variables.[30] It was also observed by Oliveira et al. that the practice of physical exercise in
adolescents and a physically active life may have positive effects on sleep patterns,
corroborating the findings of this study.[31]
Regarding the internal training load, no statistically significant differences were
observed between the training phases. This may be related to why there were no significant
variations in testosterone and T/C ratio. Furthermore, the lack of hormonal variation
may indicate that athletes did not have poor adaptation to training loads during the
season, or that the training load and volume imposed on athletes were not high enough
to generate variations in these variables.[32]
[33]
Training loads, depending on the dose, can disrupt muscle homeostasis and cause a
series of physiological responses generating strength and function. During strenuous
exercises, there is a great caloric expenditure, with high production of ATP through
metabolic pathways, where glycogen depletion has been related to muscle fatigue during
exercise.[34] In addition, muscle microlesions may occur from the high demand of training and
repeated muscle contractions and mechanical loads on the myofibrils during exercise.[35] In the present study, injuries occurred in all three evaluation phases; however,
we did not find statistically significant differences between the three phases. The
most injured sites were the thighs and ankles, and most injuries were due to overuse.
These findings corroborate Edouard et al.[36] who found a higher injury incidence in thighs and feet during an international athletics
championship.[36]
Salivary cortisol is recommended and widely used as a training stress index.[13] The cortisol levels in our sample were within the range expected for athletes (1.8–19.9 nmol.
L −1), but it increased as the season progressed and the stress and chronic load of
training and competitions. The increase in cortisol levels induced by training stress
was shown to be lower in trained individuals than in untrained individuals,[37] which may explain the little hormonal variation in the present sample between training
phases.
Testosterone is an anabolic hormone that has functions such as tissue repair, and
muscle tissue growth and is related to athletes' motivation.[13] We did not find differences in testosterone concentration between training phases.
It was expected that testosterone levels would be higher in the competitive period,
given that adequate hormonal signaling is essential for physical adaptations, and
low testosterone levels are associated with decreased performance, energy, and strength.[13] Thus, as there were no significant changes, there was also no variation in the T/C
ratio and this is a positive point, since its prolonged decrease in the T/C ratio
may be associated with losses in sports performance due to proteolysis and a decrease
in protein synthesis.[13]
Some limitations of the present study should be mentioned. We had a relatively small
sample size and a large sample loss, reflecting the difficulties in objectively monitoring
sleep in adolescent elite athletes. Also, we did not control the training schedule
and parameters in the three different phases; however, we tried to mitigate it by
evaluating the internal load in each phase. Thus, more studies with larger sample
sizes and control of training parameters should be performed.
We conclude that adolescent athletes presented a higher TST and lower WASO in the
post-competitive phase compared with the competitive and preparatory phases, respectively.
However, cortisol was higher in the preparatory phase. Despite this, no differences
were observed in the other sleep variables, testosterone, and incidence of musculoskeletal
injuries, nor were the variables correlated.
The present study suggests that there is a need to encourage better sleep habits,
especially in the pre-competitive and competitive phases in adolescent athletes. In
addition, it seems that in adolescent track and field athletes, monitoring with biomarkers
was not correlated with the incidence of injuries, and, therefore, we suggest that
new methods be used to seek to investigate the relationship of injuries with other
variables related to training in this population.
