Heart Rate Variability and Training Load Among National Collegiate Athletic Association Division 1 College Football Players Throughout Spring Camp

Abstract

The purpose of this study was to determine if recovery of cardiac-autonomic activity to baseline occurs between consecutive-day training sessions among positional groups of a collegiate football team during Spring camp. A secondary aim was to evaluate relationships between chronic (i.e., 4-week) HRV and training load parameters. Baseline HRV (lnRMSSD_BL) was compared with HRV following ~20 h of recovery prior to next-day training (lnRMSSDpost20) among positional groups comprised of SKILL (n = 11), MID-SKILL (n = 9) and LINEMEN (n = 5) with a linear mixed model and effect sizes (ES). Pearson and partial correlations were used to quantify relationships between chronic mean and coefficient of variation (CV) of lnRMSSD (lnRMSSD_chronic and lnRMSSDcv, respectively) with the mean and CV of PlayerLoad (PL_chronic and PL_cv, respectively). A position × time interaction was observed for lnRMSSD (p = 0.01). lnRMSSD_BL was higher than lnRMSSDpost20 for LINEMEN (p <0.01; ES = Large) while differences for SKILL and MID-SKILL were not statistically different (p >0.05). Players with greater body mass experienced larger reductions in lnRMSSD (r = -0.62, p <0.01). Longitudinally, lnRMSSDcv was significantly related to body mass (r = 0.48) and PL_chronic (r = -0.60). After adjusting for body mass, lnRMSSDcv and PL_chronic remained significantly related (r = -0.43). The ~20 h recovery time between training sessions on consecutive days may not be adequate for restoration of cardiac-parasympathetic activity to baseline among LINEMEN. Players with a lower chronic training load throughout camp experienced greater fluctuation in lnRMSSD (i.e., lnRMSSDcv) and vice-versa. Thus, a capacity for greater chronic workloads may be protective against perturbations in cardiac-autonomic homeostasis among American college football players.

Full text

Downloaded from https://journals.lww.com/nsca-jscr by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3bhnalqTQiPvqBhGoLqqjrnzC9EarABoyogS5oDBbVEGPxeWHXR9pnA== on 11/02/2018
Downloadedfromhttps://journals.lww.com/nsca-jscr by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3bhnalqTQiPvqBhGoLqqjrnzC9EarABoyogS5oDBbVEGPxeWHXR9pnA== on 11/02/2018
HEART RATE VARIABILITY AND TRAINING LOAD
AMONG NATIONAL COLLEGIATE ATHLETIC
ASSOCIATION DIVISION 1COLLEGE FOOTBALL
PLAYERS THROUGHOUT SPRING CAMP
ANDREW A. FLATT,
1,2
MICHAEL R. ESCO,
1
JEFF R. ALLEN,
3
JAMES B. ROBINSON,
3
RYAN L. EARLEY,
4
MICHAEL V. FEDEWA,
1
AMY BRAGG,
5
CLAY M. KEITH,
3
AND JONATHAN E. WINGO
1
1
Exercise Physiology Laboratory, Department of Kinesiology, University of Alabama, Tuscaloosa, Alabama;
2
Department of
Health Sciences, Armstrong State University, Savannah, Georgia;
3
Department of Athletics, Sports Medicine, University of
Alabama, Tuscaloosa, Alabama;
4
Department of Biological Sciences, University of Alabama, Tuscaloosa, Alabama; and
5
Department of Athletics, Sports Nutrition, University of Alabama, Tuscaloosa, Alabama
ABSTRACT
Flatt, AA, Esco, MR, Allen, JR, Robinson, JB, Earley, RL,
Fedewa,MV,Bragg,A,Keith,CM,andWingo,JE.Heartrate
variability and training load among National Collegiate
Athletic Association Division 1 college football players
throughout spring camp. J Strength Cond Res 32(11):
3127–3134, 2018—The purpose of this study was to deter-
mine whether recovery of cardiac-autonomic activity to baseline
occurs between consecutive-day training sessions among posi-
tional groups of a collegiate football team during Spring camp. A
secondary aim was to evaluate relationships between chronic (i.e.,
4-week) heart rate variability (HRV) and training load parameters.
Baseline HRV (lnRMSSD_BL) was compared with HRV after
;20 hours of recovery before next-day training (lnRMSSDpost20)
among positional groups composed of SKILL (n= 11), MID-
SKILL (n= 9), and LINEMEN (n= 5) with a linear mixed model
and effect sizes (ES). Pearson and partial correlations were used
to quantify relationships between chronic mean and coefficient of
variation (CV) of lnRMSSD (lnRMSSD_chronic and lnRMSSDcv,
respectively) with the mean and CV of PlayerLoad (PL_chronic
and PL_cv, respectively). A position 3time interaction was
observed for lnRMSSD (p= 0.01). lnRMSSD_BL was higher than
lnRMSSDpost20 for LINEMEN (p,0.01; ES = large), whereas
differences for SKILL and MID-SKILL were not statistically differ-
ent (p.0.05). Players with greater body mass experienced larger
reductions in lnRMSSD (r=20.62, p,0.01). Longitudinally,
lnRMSSDcv was significantly related to body mass (r= 0.48) and
PL_chronic (r=20.60). After adjusting for body mass,
lnRMSSDcv and PL_chronic remained significantly related
(r=20.43). The ;20-hour recovery time between training ses-
sions on consecutive days may not be adequate for restoration of
cardiac-parasympathetic activity to baseline among LINEMEN.
Players with a lower chronic training load throughout camp expe-
rienced greater fluctuation in lnRMSSD (i.e., lnRMSSDcv) and
vice versa. Thus, a capacity for greater chronic workloads may
be protective against perturbations in cardiac-autonomic homeo-
stasis among American college football players.
KEY WORDS parasympathetic, autonomic, monitoring, sport
physiology, sport science, recovery
INTRODUCTION
Despite a demanding training schedule and high
injury rate in American football (26), research
pertaining to training status monitoring among
American football players pales in comparison
with sports such as soccer and rugby. American football
players vary in physical and performance characteristics
because of unique positional requirements. For example, re-
ceivers and defensive backs (SKILL) experience the greatest
running demands and thus tend to have the fastest sprinting
speeds, lowest body and fat mass, and greatest aerobic fitness
level among positional groups (32,36). By contrast, offensive
and defensive linemen (LINEMEN) have the lowest run-
ning demands but regularly encounter physical bouts in
which they must displace their opponent to gain or defend
field position (26,36). LINEMEN, therefore, have the high-
est maximal strength, greatest body and fat mass, and lowest
aerobic fitness level of the various positional groups (32).
Linebackers, running backs, and tight-ends (MID-SKILL)
experience playing demands characteristic of both SKILL
and LINEMEN and thus tend to display physical and fitness
characteristics intermediate to these positions (26,32,36).
Given that body mass, playing demands and physiological
responses to training vary by position (26), recovery duration
Address correspondence to Andrew A. Flatt, aflatt@crimson.ua.edu.
32(11)/3127–3134
Journal of Strength and Conditioning Research
Ó2017 National Strength and Conditioning Association
VOLUME 32 | NUMBER 11 | NOVEMBER 2018 | 3127
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

requirements may also differ among position. However, this
topic has received little investigation.
Wearable devices using triaxial accelerometers (micro-
sensors) are capable of measuring movement profiles of
athletes during sports play. Wearable microsensors are
attractive to coaches because the physical demands of
competition or training can be quantified with minimal
burden to the athlete. Though training load monitoring is
becoming more popular in American football (35–37), its use
in conjunction with recovery status indicators among foot-
ball players has not been well studied. This research is
needed because external training load does not provide
information regarding internal physiological responses and
there exists substantial interindividual variation in responses
and adaptation to training (24).
A recovery status metric gaining popularity among
sports teams is resting heart rate variability (HRV). Heart
rate variability reflects autonomic modulation of the heart
and can be measured noninvasively with inexpensive field
tools such as smartphone applications (13,15,28). Vagally
mediated HRV is considered a global marker of homeo-
stasis and reflects cardiovascular recovery after a training
session (33). For example, vagal-HRV suppression is
observed for 24–48 hours after intense training (33) with
the return to baseline possibly reflecting the optimal
state for subsequent intensive training (22). Cardiac-
parasympathetic recovery from exercise is affected by fac-
tors such as fitness level, exercise intensity, and alterations
in fluid balance (33). Because fitness level, body mass, as
well as fluid balance and thermoregulatory responses to
training vary among positional groups in American foot-
ball (9,32), daily cardiac-autonomic responses inferred
from vagal HRV may also differ among positions and thus
provide useful recovery status information.
Greater daily fluctuation in cardiac-parasympathetic activity
(assessed by the coefficient of variation [CV]) has been
associated with lower maximal oxygen uptake (V
_
O
2
max) (14),
lower intermittent running performance (2,7,14,16), and greater
fatigue during intensified training (17,18). An athlete’s chronic
training load (i.e., 4-week average) is sometimes used as an
indicator of training capacity or fitness level (21). Hypotheti-
cally, players who perform greater chronic training loads (i.e.,
SKILL) should have a lower CV of vagal HRV, whereas those
who perform lower chronic workloads (i.e., LINEMEN) would
be expected to display a greater CV. This would build on pre-
vious findings that less fit individuals experience greater pertur-
bations in cardiac-autonomic homeostasis than more fit
individuals (7,14,16). However, no previous studies have eval-
uated relationships between chronic training load and chronic
vagal-HRV trends in American football players.
The purpose of this study was to determine whether
recovery of cardiac-autonomic activity to baseline occurs
between consecutive-day training sessions among positional
groups of a collegiate football team. A secondary aim was to
assess relationships between football players’ chronic work-
loads and chronic vagal-HRV trends (i.e., mean and CV)
from an annual Spring training camp.
METHODS
Experimental Approach to the Problem
This was a prospective observational cohort study that
evaluated cardiac-autonomic responses to training among
positional groups of an elite college football team during
their 2016 Spring training camp. The research design and
methodology were devised according to the predetermined
program and structure of the training camp, which the
researchers did not influence. Vagal-HRV and external
training load were acquired each football training day
throughout the 4-week Spring camp. For the first objective,
we assessed whether vagal HRV returned to baseline levels
between consecutive-day training sessions (i.e., after ;20
hours of recovery) among positional groups. For the second
objective, relationships between chronic external training
load and chronic vagal HRV (i.e., mean and CV) from the
4-week camp were evaluated.
Subjects
Twenty-five Division 1 football players from a National
Collegiate Athletics Association team volunteered for this
study. This was the national championship winning team
from the preceding competitive season. Only players aged 18
years and older and on athletic scholarship were included.
Volunteers were grouped based on position as previously
described (26) (SKILL: n= 11; age = 20 61 years; height =
187.7 64.0 cm; body mass = 90.2 64.3 kg; MID-SKILL: n
= 9; age = 20 61 years; height = 188.8 65.5 cm; body mass
= 103.8 64.4 kg; and LINEMEN: n= 5; age = 22 61 years;
height = 192.5 62.8 cm; body mass = 131.1 610.6 kg). All
athletes obtained medical clearance from the sports medi-
cine staff and provided written informed consent for their
involvement as research participants. Study approval was
granted by the University of Alabama institutional review
board.
Procedures
Spring Camp. Spring camp was held between mid-March and
mid-April and involved 4 weeks of football training. Two-
hour football practices were held on Monday, Wednesday,
and Friday of week 1; Monday, Wednesday, Friday, and
Saturday of weeks 2 and 3; and Tuesday, Thursday, and
Saturday of week 4. Forty-five minutes full-body resistance
training sessions of fewer sets and lower intensity (#85% of 1
repetition maximum) relative to workouts preceding Spring
camp were held on Tuesdays and Thursdays throughout
weeks 1–3. Sundays were reserved for passive rest. Before
Spring camp, all players participated in an 8-week off-season
strength and conditioning program. Outdoor temperatures
during training were 21.7 62.88C.
Training Load. Training load parameters were obtained from
football sessions using triaxial accelerometers (Catapult
HRV in College Football Players
3128
Journal of Strength and Conditioning Research
the
TM
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Innovations, Melbourne, Australia) at a sampling rate of 100
Hz. These devices measure full-body acceleration in 3
planes: anteroposterior, mediolateral, and vertical. Subjects
wore the same device each session, positioned between the
scapulae, and fixed in place on their shoulder pad in
a custom-built cartridge. After each practice session, training
load data were downloaded to a laptop for analysis. The
training load parameter used for this study was total Play-
erLoad. This parameter has demonstrated acceptable reli-
ability and reflects total external workload including running,
jumping, changes of direction, and body contacts (3). In
addition, variation in PlayerLoad has been related to injury
occurrence in college football players (37). PlayerLoad is
expressed as the square root of the sum of the squared
instantaneous rate of change in acceleration in each vector
and divided by 100 (3).
Physiological Assessments. Fasted body mass was measured on
a calibrated digital scale (Tanita Corporation, Arlington
Heights, IL, USA) at the training facility before Spring
camp. Heart rate variability data were obtained in the
athletic training facility at least 90 minute after team
breakfast and before any physical activity (25). Subjects ver-
bally confirmed that caffeinated beverages were not con-
sumed before data acquisition. Heart rate variability
recordings were obtained while subjects were seated com-
fortably on an athletic training table 60–90 minute before
training. Once seated, participants were handed a tablet
device (iPad2; Apple, Inc., Cupertino, CA, USA) with a val-
idated optical pulse-wave finger sensor (HRV Fit Ltd.,
Southampton, United Kingdom) inserted into the head-
phone slot (13). Subjects were instructed to insert their left
index finger into the finger sensor cuff and to select their
name from the team roster previously uploaded onto the
application (ithlete Team; HRV Fit Ltd). Subjects would
then initiate a 1-minute HRV recording while remaining
quiet and breathing naturally (30). All HRV recordings were
preceded by at least 1 minute for stabilization (24). The
application provides the time domain vagal-HRV index of
the natural logarithm of the root mean square of successive
RR intervals (lnRMSSD). The lnRMSSD is multiplied by 20
to fit an approximate 100-point scale for simplified interpre-
tation (15). The application is equipped with an irregular
pulse-rate detection algorithm which excludes interpulse in-
tervals ,500 ms and .2000 ms. In addition, adjacent normal
pulse–pulse (PPn) interval differences are automatically
examined using the following formula: (PPn 2[PPn 21])
2
,(40 3Exp [120/PRave])
2
, where PRave is the average
pulse rate calculated since commencement of the recording.
Statistical Analyses
Data normality was confirmed with the Shapiro-Wilk test (p.
0.05). To determine whether lnRMSSD returned to baseline
between consecutive-day football training sessions, we com-
pared baseline lnRMSSD (lnRMSSD_BL) to lnRMSSD after
the ;20-hour recovery period between training sessions
(lnRMSSDpost20). The intraindividual mean lnRMSSD from
Monday, Wednesday, and Friday of week 2 represented
lnRMSSD_BL as each of these training sessions were sepa-
rated by $44 hours. Saturday of the same week represented
lnRMSSDpost20 as it was preceded by only ;20-hour rest
before the next session. Compliance for HRV measures for this
component was 100%. A linear mixed model was used to
examine variation in lnRMSSD among positions and between
lnRMSSD_BL and lnRMSSDpost20. Position was included as
a fixed effect, time (lnRMSSD_BL vs. lnRMSSDpost20) as
a fixed within-subjects repeated measure, the position 3time
interaction as a fixed effect, and athlete identification as a ran-
dom effect. Tukey honest significant difference (HSD) was used
for post hoc analyses. We evaluated the magnitude of the
change in lnRMSSD between conditions with Cohen’s deffect
sizes (ES) 690% confidence limits (CLs) (8). Effect sizes were
interpreted qualitatively as follows: ,0.2, trivial; 0.2–0.59, small;
0.6–1.19, moderate; 1.2–1.9, large; and .2.0, very large (20).
The effect was deemed unclear if the CL crossed thresholds
for both substantially positive (0.20) and negative (20.20)
values (1). Individual change variables for lnRMSSD were
calculated (lnRMSSDpost20 2lnRMSSD_BL, DlnRMSSD).
TABLE 1. Model effects and the natural logarithm of the root mean square of successive RR interval difference
(lnRMSSD) values among and between positional groups and across time.*
Model
effect Fdf p Mean 6SD lnRMSSD
Position 4.24 2, 22 0.028 SKILL: 82.4 67.6 MID-SKILL: 76.4 68.0 LINEMAN: 71.0 610.3†
Time 18.47 1, 22 ,0.001 BL: 80.0 67.3 Post: 76.0 610.7z
Position
3time
5.46 2, 22 0.012 SKILL BL: 82.8 67.2 MID-SKILL BL: 78.4 66.2 LINEMAN BL: 76.6 68.6
SKILL post: 82.0 68.3 MID-SKILL post: 74.4 69.5 LINEMAN post: 65.4 69.5§
*BL = baseline; post = 20-hour post-training.
†p#0.05 vs. SKILL across all time-points.
zp#0.05 vs. BL across all groups.
§p#0.05 vs. BL for LINEMEN.
Journal of Strength and Conditioning Research
the
TM
|
www.nsca.com
VOLUME 32 | NUMBER 11 | NOVEMBER 2018 | 3129
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

The relationship between DlnRMSSD and body mass was
quantified via Pearson correlation. PlayerLoad values derived
from the same days as lnRMSSD_BL were compared among
positions with a 1-way analysis of variance (ANOVA), Tukey
HSD post hoc analysis, and ES.
Pearson correlations were used to quantify relationships
between the mean and CV of PlayerLoad, (PL_chronic and
PLcv, respectively) with the mean and CV of lnRMSSD
(lnRMSSD_chronic and lnRMSSDcv, respectively) derived
from the entire 4-week training camp (i.e., all training days).
The thresholds used for qualitative assessment of the
correlations were as follows: ,0.1, trivial; 0.1–0.29, small;
0.3–0.49, moderate; 0.5–0.7, large; 0.7–0.89, very large;
and .0.9 nearly perfect (20). Positional differences in
lnRMSSDcv were evaluated
with a 1-way ANOVA, Tukey
HSD post hoc analysis, and ES.
Compliance for HRV measures
for this component was 93.4 6
8.7%. Statistical procedures were
performed using JMP Pro 13
(SAS Institute, Inc., Cary, NC,
USA) and Excel 2016 (Microsoft
Corp., Redmond, WA, USA).
pvalues #0.05 were considered
statistically significant. Data are
reported as mean 6SD unless
noted otherwise.
RESULTS
A significant effect was found for
PlayerLoad values derived from
the same days as lnRMSSD_BL
(p,0.0001). Baseline Player-
Load values for SKILL, MID-
SKILL, and LINEMEN were
623.7 660.6 au, 556.8 653.6
au, and 450.0 629.2 au, respec-
tively. Post hoc comparisons re-
vealed that PlayerLoad values
for SKILL were significantly
higher than LINEMAN (p,
0.0001, ES = 3.27 61.27, very
large) and MID-SKILL (p=
0.029, ES = 1.16 60.79, moder-
ate). In addition, MID-SKILL
baseline PlayerLoad was signifi-
cantly greater than LINEMEN
(p= 0.005, ES = 2.25 61.10,
very large).
A significant position 3
time interaction was found
for lnRMSSD. Model effects
and lnRMSSD values are pre-
sented in Table 1. Post hoc
comparisons revealed that lnRMSSD_BL was signifi-
cantly higher than lnRMSSDpost20 for LINEMEN (p,
0.01; ES = large), whereas differences for SKILL (p=
0.998; ES = unclear) and MID-SKILL (p= 0.343; ES =
unclear) were not statistically significant. Interaction ES
and individual lnRMSSD responses are graphically
displayed in Figure 1A, B, respectively.
SKILL had greater lnRMSSD than LINEMEN (Table 1;
ES = 1.35 60.96; large). In addition, evaluated as a group (n=
25), lnRMSSD_BL was higher than lnRMSS Dpost20 (Table 1;
ES = 0.44 60.47; small). We observed a large and significant
negative relationship between DlnRMSSD and body mass
(Figure 2A), indicating that greater reductions in lnRMSSD
occurred among players with greater body mass.
Figure 1. A) Effect size 690% confidence limits (ES 690% CLs) comparison between baseline natural
logarithm of the root mean square of successive RR interval differences (lnRMSSD_BL) with lnRMSSD after 20
hours of rest (lnRMSSDpost20) among positional groups. The horizontal dashed lines represent thresholds for
a small effect (0.20 to 20.20). B) Individual lnRMSSD_BL and lnRMSSDpost20 values among positional groups.
HRV in College Football Players
3130
Journal of Strength and Conditioning Research
the
TM
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

A significant effect for lnRMSDcv was observed
(p,0.05). SKILL, MID-SKILL, and LINEMEN lnRMSSDcv
values were 6.5 62.9%, 8.5 61.7%, and 10.5 62.7%, respec-
tively. Post hoc analysis showed that lnRMSSDcv for SKILL
was significantly lower than LINEMEN (p#0.05, ES = 21.41
60.97, large) but not MID-SKILL (p.0.05, ES = 20.82 6
0.76, moderate). In addition, MID-SKILL lnRMSSDcv was
not significantly different from LINEMEN (p.0.05, ES =
20.96 61.28, unclear).
The lnRMSSD_chronic did not show significant relation-
ships with PL_chronic (r=0.33,p= 0.109), PLcv (r= 0.16, p=
0.436), or body mass (r=20.36, p= 0.073). There was a large,
significant negative relationship between lnRMSSDcv and
PL_chronic (Figure 2B) and a moderate, significant positive
relationship between lnRMSSDcv and body mass (Figure
2C). The lnRMSSDcv did not relate with PLcv (r=0.025
p= 0.907). Body mass showed a large, significant negative
relationship with PL_chronic (Figure 2D). After adjusting for
body mass via partial correlation analysis, the negative relation-
ship between lnRMSSDcv and PL_chronic remained signifi-
cant (r=20.42, p= 0.034, moderate).
DISCUSSION
This study evaluated daily and chronic lnRMSSD responses
to training among an elite college football team throughout
Spring camp. The novel finding was that daily lnRMSSD
responses to a football training session differ by position.
;Twenty hours after a practice session, LINEMEN show
substantial reductions from baseline, whereas SKILL and
MID-SKILL were recovered to within or near baseline val-
ues. Longitudinally, lnRMSSDcv was inversely related to
PL_chronic, independent of body mass.
Previous studies among youth and elite adult soccer and
rugby players have found that daily cardiac-parasympathetic
activity is not significantly affected by training sessions at the
group level (4,12,34), but is substantially reduced 1-day post-
competition among youth players (4,12). During a ,1 week
training camp, elite adult rugby players showed small reduc-
tions in lnRMSSD after the first day of training and
remained suppressed for the duration of camp (25). Exclud-
ing LINEMEN, our findings tend to agree with the previous
literature considering SKILL players are reasonably compa-
rable to soccer players and MID-SKILL players are reason-
ably comparable to rugby players with regards to physical
characteristics (23).
Despite substantially smaller PlayerLoad values than
SKILL and MID-SKILL, LINEMEN demonstrated large
reductions in lnRMSSDpost20, indicating inadequate car-
diovascular recovery between sessions. A recent systematic
review determined that cardiac-parasympathetic recovery
Figure 2. A) Scatterplot representing the relationship of the difference in natural logarithm of the root mean square of successive RR interval differences from
baseline (lnRMSSD_BL) and 20-hour post-training (lnRMSSDpost20) (i.e., lnRMSSDpost20 2lnRMSSD_BL = DlnRMSSD) with body mass. B) Scatterplot
representing the relationship between the coefficient of variation of lnRMSSD (lnRMSSDcv) and average PlayerLoad (PL_chronic) from the entire 4-week camp.
C) Scatterplot representing the relationship between lnRMSSDcv and body mass. D) Scatterplot representing the relationship between PL_chronic and body
mass.
Journal of Strength and Conditioning Research
the
TM
|
www.nsca.com
VOLUME 32 | NUMBER 11 | NOVEMBER 2018 | 3131
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

from exercise is slower in individuals with lower aerobic
fitness and is suppressed for longer durations (;48 hours)
after high intensity, anaerobic exercise (33). We speculate
that numerous factors related to body mass, fitness, and
exercise intensity may have contributed to the suppressed
lnRMSSDpost20 in LINEMEN. For example, during 60 mi-
nutes of simulated football training, LINEMEN displayed an
average exercise intensity corresponding to 79% of maxi-
mum HR, a respiratory exchange ratio .0.90, and blood
lactate levels .5.0 mmol$L
21
(19). As such, LINEMEN
may experience a substantial anaerobic workload during
training due to both lower aerobic fitness and movement
demands that depend more heavily on the expression of
strength and power (e.g., repeated blocking, tackling, and short
sprints). Moreover, LINEMEN are more sedentary than other
positions between efforts because of smaller distances covered,
necessitating less jogging (29). This provides less active recov-
ery for LINEMEN in addition to reduced convective heat loss
compared with SKILL (11). Indeed, LINEMEN experience the
greatest internal core temperatures and fluid loss during training
which may increase cardiac strain (10). Deren et al. (10) dem-
onstrated that LINEMEN experience a progressive increase in
exercise HR beyond 30 minutes of cycling in the heat at a work-
load that elicits a heat production of 350 W$m
22
, whereas
SKILL demonstrated no such change in HR. Although con-
tinuous cycling is dissimilar from football training, the cardio-
vascular response may be relevant. Dehydration-induced
hypovolemia is thought to contribute to suppressed lnRMSSD
in the 24- to 48-hour postexercise period (33). Accordingly,
unloading of cardiopulmonary baroreceptors facilitates sympa-
thoexcitation to maintain peripheral resistance and combat the
reduction in cardiac output consequent to reduced stroke vol-
ume (31). It has been hypothesized that lnRMSSD may not
return to baseline after intense exercise until plasma volume has
been restored (6). Collectively, lower aerobic fitness, larger body
mass, greater reliance on anaerobic-glycolytic metabolism
during training, and disturbed fluid balance may all have con-
tributed to the delayed cardiac-parasympathetic recovery in
LINEMEN.
Whether unrecovered lnRMSSD affects performance or
injury risk in football players remains unclear, particularly
during periods when sessions areheldmorefrequently(e.g.,
preseason). Hypothetically, consistent inadequate recovery may
lead to suppressed lnRMSSD over prolonged periods which
has been associated with high perceived fatigue, illness, and
decrements in running performance (16,18,27). We speculate
that LINEMEN may be at greater risk of inadequate recovery
when training over several consecutive days based on their
daily lnRMSSD responses to training observed in this study.
Thus, future research evaluating cardiac-autonomic responses
to more frequent training among LINEMEN is warranted.
Relationships between chronic training load and
lnRMSSD parameters have not previously been investi-
gated in American football. Although indirect, chronic
training load is sometimes used as a surrogate for fitness
level (21). We did not find a significant relationship between
lnRMSSD_chronic and PL_chronic, possibly because weekly
changes in averaged lnRMSSD relate more with fitness (5,16)
than lnRMSSD averaged over chronic periods. Previous studies
found that lnRMSSDcv is inversely related to V
_
O
2
max and
intermittent-running performance in soccer players (2,7,14,16).
We observed a significant inverse relationship between
lnRMSSDcv and PL_chronic after adjusting for body mass.
This suggests that training capacity (inferred from PL_chronic)
may be a determinant of lnRMSSDcv in football players
during Spring camp, a period where overreaching is
unlikely to occur due to mostly nonconsecutive-day train-
ing. This distinction is important because more intensified
(e.g., preseason) or stressful (e.g., competitive season)
periods may increase lnRMSSDcv and alter its relation-
ship with PL_chronic. Another explanation may be that
lnRMSSDcv is reflecting position-specific training effects,
where LINEMEN experience more physiologically dis-
ruptive demands from line-play, exacerbated by their size
and fitness level relative to other positions. Additional
research is needed to determine what information
lnRMSSDcv provides in the context of fitness and training
adaptation in football players.
Although HRV data collection was standardized ac-
cording to recent recommendations (25), there is greater
potential for “noise” when acquired at the facility versus
postwaking and thus is a limitation of this study. In addi-
tion, HRV data were only collected on football training
days, and therefore the lnRMSSD values do not reflect
all days of the week. The inclusion of only 5 LINEMEN
is also a limitation, and thus further research is needed to
support the current findings. It must also be considered
that the lnRMSSD responses from this observation period
may not reflect what occurs during training in hotter con-
ditions with more frequent training sessions. Finally, rela-
tionships between chronic workloads and fitness have not
been investigated in American football players; thus, we
caution readers that our interpretation of lnRMSSDcv re-
flecting training capacity is, therefore, speculative.
Approximately 20 hours after a football training session,
lnRMSSD values were suppressed for LINEMEN, whereas
SKILL and MID-SKILL values returned to near or within
baseline. Players with greater body mass showed the greatest
reductions in lnRMSSD in response to training. Over the 4-
week camp, players who performed the lowest chronic
workloads experienced greater fluctuation in lnRMSSD (i.e.,
lnRMSSDcv) and vice versa. Therefore, training capacity
may be a determinant of lnRMSSD trends in football players
during Spring camp.
PRACTICAL APPLICATIONS
During Spring camp, LINEMEN did not attain cardiac-
parasympathetic recovery to baseline between consecutive-
day training sessions, potentially making them susceptible
to autonomic nervous system imbalance during more
HRV in College Football Players
3132
Journal of Strength and Conditioning Research
the
TM
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

intensive training periods. Longitudinally, a capacity for
greater chronic workloads may be protective against daily
perturbations in cardiac-autonomic homeostasis based
on the inverse relationship between PL_chronic and
lnRMSSDcv.
Obtaining compliance from athletes to perform daily HRV
measures at home after waking is challenging and in some
cases prohibitive of the implementation of HRV monitoring.
This study showed that standardized, 60-second lnRMSSD
recordings acquired via mobile devices at the training facility
may provide meaningful information regarding training
responses among football players. This may make HRV
monitoring a more convenient process for coaching and
support staff.
ACKNOWLEDGMENTS
The authors extend their thanks to the Alabama football
players and staff for their participation in this study. The
authors also thank Aaron Brosz for his assistance throughout
the study. The authors have no conflicts of interest to
disclose.
REFERENCES
1. Batterham, AM and Hopkins, WG. Making meaningful inferences
about magnitudes. Int J Sports Physiol Perf 1: 50–57, 2006.
2. Boullosa, DA, Abreu, L, Nakamura, FY, Mun
˜oz, VE, Domı
´nguez, E,
and Leicht, AS. Cardiac autonomic adaptations in elite Spanish
soccer players during preseason. Int J Sports Physiol Perf 8: 400–409,
2013.
3. Boyd, LJ, Ball, K, and Aughey, RJ. The reliability of MinimaxX
accelerometers for measuring physical activity in Australian football.
Int J Sports Physiol Perf 6: 311–321, 2011.
4. Bricout, VA, DeChenaud, S, and Favre-Juvin, A. Analyses of heart
rate variability in young soccer players: The effects of sport activity.
Autonom Neurosci 154: 112–116, 2010.
5. Buchheit, M, Chivot, A, Parouty, J, Mercier, D, Al Haddad, H,
Laursen, P, and Ahmaidi, S. Monitoring endurance running
performance using cardiac parasympathetic function. Eur Appl
Physiol 108: 1153–1167, 2010.
6. Buchheit, M, Laursen, PB, Al Haddad, H, and Ahmaidi, S. Exercise-
induced plasma volume expansion and post-exercise
parasympathetic reactivation. Eur J Appl Physiol 105: 471–481, 2009.
7. Buchheit, M, Mendez-Villanueva, A, Quod, MJ, Poulos, N, and
Bourdon, P. Determinants of the variability of heart rate measures
during a competitive period in young soccer players. Eur J Appl
Physiol 109: 869–878, 2010.
8. Cohen, J. Statistical Power Analysis for the Behavioral Sciences (2nd
ed.). Hillsdale, NJ: L. Erlbaum, 1988.
9. Davis, JK, Baker, LB, Barnes, K, Ungaro, C, and Stofan, J.
Thermoregulation, fluid balance, and sweat losses in American
football players. Sports Med 46: 1–15, 2016.
10. Deren, TM, Coris, EE, Bain, AR, Walz, SM, and Jay, O. Sweating is
greater in NCAA football linemen independently of heat
production. Med Sci Sports Exerc 44: 244–252, 2012.
11. Deren, TM, Coris, EE, Casa, DJ, DeMartini, JK, Bain, AR, Walz, SM,
and Jay, O. Maximum heat loss potential is lower in football linemen
during an NCAA summer training camp because of lower self-
generated air flow. J Strength Cond Res 28: 1656–1663, 2014.
12. Edmonds, RC, Sinclair, WH, and Leicht, AS. Effect of a training
week on heart rate variability in elite youth rugby league players. Int
J Sports Med 34: 1087–1092, 2013.
13. Esco, MR, Flatt, AA, and Nakamura, FY. Agreement between
a smart-phone pulse sensor application and ECG for determining
lnRMSSD. J Str Cond Res 31: 380–385, 2017.
14. Flatt, AA, Esco, M, Nakamura, FY, and Plews, DJ. Interpreting daily
heart rate variability changes in collegiate female soccer players. J
Sports Med Phys Fitness 57: 907–915, 2017.
15. Flatt, AA and Esco, MR. Validity of the ithlete smart phone
application for determining ultra-short-term heart rate variability. J
Hum Kinet 39: 85–92, 2013.
16. Flatt, AA and Esco, MR. Evaluating individual training adaptation
with Smartphone-derived heart rate variability in a collegiate female
soccer team. J Strength Cond Res 30: 378–385, 2016.
17. Flatt, AA and Esco, MR. Smartphone-derived heart-rate variability
and training load in a women’s soccer team. Int J Sports Physiol Perf
10: 994–1000, 2015.
18. Flatt, AA, Hornikel, B, and Esco, MR. Heart rate variability and
psychometric responses to overload and tapering in collegiate
sprint-swimmers. J Sci Med Sport 20: 606–610, 2017.
19. Hitchcock, KM, Millard-Stafford, ML, Phillips, JM, and Snow, TK.
Metabolic and thermoregulatory responses to a simulated American
football practice in the heat. J Strength Cond Res 21: 710–717, 2007.
20. Hopkins, W, Marshall, S, Batterham, A, and Hanin, J. Progressive
statistics for studies in sports medicine and exercise science. Med Sci
Sports Exerc 41: 3–13, 2009.
21. Hulin, BT, Gabbett, TJ, Lawson, DW, Caputi, P, and Sampson, JA.
The acute: Chronic workload ratio predicts injury: High chronic
workload may decrease injury risk in elite rugby league players. Br J
Sport Med 50: 231–236, 2016.
22. Kiviniemi, AM, Hautala, AJ, Kinnunen, H, and Tulppo, MP.
Endurance training guided individually by daily heart rate variability
measurements. Eur J Appl Physiol 101: 743–751, 2007.
23. Kuhn, W, Reilly, T, Clarys, J, and Stibbe, A. Comparative Analysis of
Selected Motor Performance Variables in American Football, Rugby
Union and Soccer Players. In: Science and Football II. T Reilly, J
Clarys, and A Stibbe, eds. London, England: Chapman & Hall, 1993.
pp. 62–69.
24. Mann, TN, Lamberts, RP, and Lambert, MI. High responders and
low responders: Factors associated with individual variation in
response to standardized training. Sports Med 44: 1113–1124, 2014.
25. Nakamura, FY, Pereira, LA, Esco, MR, Flatt, AA, Moraes, JE, Cal,
AC, and Loturco, I. Intra-and inter-day reliability of ultra-short-term
heart rate variability in rugby union players. J Strength Cond Res 31:
548–551, 2017.
26. Pincivero, DM and Bompa, TO. A physiological review of American
football. Sports Med 23: 247–260, 1997.
27. Plews, DJ, Laursen, PB, Kilding, AE, and Buchheit, M. Heart rate
variability in elite triathletes, is variation in variability the key to
effective training? A case comparison. Eur J Appl Physiol 112: 3729–
3741, 2012.
28. Plews, DJ, Scott, B, Altini, M, Wood, M, Kilding, AE, and Laursen,
PB. Comparison of heart rate variability recording with smart
phone photoplethysmographic, Polar H7 chest strap and
electrocardiogram methods. Int J Sport Physiol Perf, 2017. Epub
ahead of print.
29. Rhea, MR, Hunter, RL, and Hunter, TJ. Competition modeling of
American football: Observational data and implications for high
school, collegiate, and professional player conditioning. J Strength
Cond Res 20: 58–61, 2006.
30. Saboul, D, Pialoux, V, and Hautier, C. The impact of breathing on
HRV measurements: Implications for the longitudinal follow-up of
athletes. Eur J Sport Sci 13: 534–542, 2013.
31. Saitoh, T, Ogawa, Y, Aoki, K, Shibata, S, Otsubo, A, Kato, J, and
Iwasaki, K. Bell-shaped relationship between central blood
volume and spontaneous baroreflex function. Autonom Neurosci 143:
46–52, 2008.
Journal of Strength and Conditioning Research
the
TM
|
www.nsca.com
VOLUME 32 | NUMBER 11 | NOVEMBER 2018 | 3133
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

32. Smith, D and Byrd, R. Body composition, pulmonary function and
maximal oxygen consumption of college football players. J Sports
Med Phys Fitness 16: 301–308, 1976.
33. Stanley, J, Peake, JM, and Buchheit, M. Cardiac parasympathetic
reactivation following exercise: Implications for training
prescription. Sports Med 43: 1259–1277, 2013.
34. Thorpe, RT, Strudwick, AJ, Buchheit, M, Atkinson, G, Drust, B, and
Gregson, W. Tracking morning fatigue status across in-season
training weeks in elite soccer players. Int J Sport Physiol Perf 11: 947–
952, 2016.
35. Wellman, AD, Coad, SC, Goulet, GC, Coffey, VG, and McLellan,
CP. Quantification of accelerometer derived impacts associated with
competitive games in NCAA Division I college football players. J
Strength Cond Res 31: 330–338, 2017.
36. Wellman, AD, Coad, SC, Goulet, GC, and McLellan, CP. Quantification
of competitive game demands of NCAA Division I college football
players using global positioning systems. JStrCondRes30: 11–19, 2016.
37. Wilkerson, GB, Gupta, A, Allen, JR, Keith, CM, and Colston, MA.
Utilization of practice session average inertial load to quantify
college football injury risk. J Strength Cond Res 30: 2369–2374, 2016.
HRV in College Football Players
3134
Journal of Strength and Conditioning Research
the
TM
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.