Key words
behavioral modification - physical fitness - exercise - maximum oxygen uptake - heart
rate variability - sleep intervention
Introduction
Behavioral modification (BM) is a strategy designed to sustain or restore well-being
through effects such as enhanced relaxation, reduced stress, and improved sleep. It
is often implemented as a primary or adjunct therapy in managing a multitude of disorders
ranging from cardiovascular disease [18] to psychiatric illness [41]. Specifically, one of the most common applications of BM is to enhance participation
in physical activity and exercise interventions. Over the past decade alone, accumulating
evidence suggests that BM improves exercise adherence in patients with obesity [7], cancer [34], and heart failure [11]. Although these results show promise for individuals impacted by disease, few studies
exist that examine the potential role of BM in the healthy, active individual. While
promoting exercise adherence in athletes is advantageous, its potential impact is
limited due to the high baseline adherence rate in this population. Instead, BM may
offer improved physiological benefits of training secondary to stress reduction and
enhanced sleep.
Sleep deprivation and sleep disorders affect 50–70 million Americans and result in
billions of dollars in medical costs annually [25], yet sleep health is frequently overlooked as part of maintaining wellness. Poor
sleep adversely affects physiological functioning, causing decreased autonomic tone,
increased blood pressure, and deleterious effects on inflammation and hormone balance,
even in healthy adults without medical comorbidities [37]
[45]. Despite clear evidence that poor or insufficient sleep adversely impacts health,
access to high quality behavioral interventions outside of medical settings remains
rare and has not been formally studied.
Physical activity itself is known to improve sleep quality, total sleep time, and
decreased sleep onset latency [27]. In contrast, psychological stress has been shown to impede individuals’ capability
to exercise at optimal levels [44]. Thus, implementation of BM into an exercise training program could be an important
component of both health and performance optimization. Studies of athletes suggest
that working to improve sleep habits can increase total sleep time [16]
[31]
[46] and may improve performance, however, studies have been small and to our knowledge,
controlled trials of these behavioral approaches have not been conducted.
This study investigated, using a randomized controlled study design, the potential
benefits to fitness club members of incorporating BM in the form of a novel, multicomponent
intervention, delivered by fitness professionals with specialized training in behavior
change theory as well as in aerobic and resistance training. We hypothesized that
participants who received BM in addition to an exercise training program would demonstrate
greater improvements in fitness and favorable changes in heart rate variability (HRV)
than those who engaged in the exercise training program alone.
Methods
Participants
We recruited club members meeting the following inclusion criteria: (i) apparently
healthy men and women, (ii) 22–44 years of age; (iii) history of exercising 4–8 days
per month at the fitness club over the previous three months. In total, 43 members
were enrolled into the study. All volunteers completed a pre-participation physical
activity readiness questionnaire (PAR-Q) [1] and an exercise history questionnaire. Exclusion criteria were unstable cardiovascular,
pulmonary, musculoskeletal or metabolic disorders that would preclude high intensity
exercise training as well as anabolic drug use in the past 6 months. All participants
were determined to be low risk for sleep apnea as measured by the Berlin Sleep Apnea
Questionnaire [39]. Five participants failed to complete baseline assessments, resulting in a total
of 38 randomized participants (22 men; age: 33.0±5.4 years; height: 1.74±0.8 m). Randomization
was 1:1 (19 per group) following CONSORT guidelines ([Fig. 1]). The UCLA Institutional Review Board reviewed and approved the study and all participants
gave their written informed consent. This study was conducted according to international
standards for sport and exercise science research [22].
Fig. 1 CONSORT diagram showing participant flow through the study.
Study design
All participants received a 12-week evidence-based, remotely guided aerobic exercise
program [15] and a recently validated, periodized functional resistance training regimen [43] in combination with either the behavioral modification intervention (BM group) or
a non-directive, equal-attention wellness education program (EA group). Training sessions
were 3 times weekly (for a total of 36 sessions) and participants were asked to refrain
from additional vigorous activity. Both groups received basic nutritional information
via weekly emails (Precision Nutrition, Inc., Toronto, Canada); however, dietary intake
and macronutrient portions were not assessed or controlled in either group.
Interventions
Certified personal trainers conducted the BM and EA interventions and coached all
participants through the resistance training portions of the exercise program.
Intervention group: Behavioral Modification (BM)
The weekly BM intervention comprised twelve, 10 min interactive slide presentations
(PowerPoint; Microsoft Corporation, Redmond, WA), each covering a different topic
related to enhancing relaxation, reducing stress, or improving sleep, coupled with
individualized recommendations for implementation ([Table 1]). Between sessions, participants tracked progress in a paper-and-pencil sleep diary,
which was reviewed weekly by the trainer. The BM intervention content was centered
on education commonly provided to individuals with behaviorally induced insufficient
sleep syndrome, including sleep education [6]
[, 35 ]plus components of motivational interviewing [33], to encourage participants to improve healthy sleep habits (e. g., avoiding sleep
deprivation during the work week), problem solving around challenges to adherence,
and stress management techniques such as relaxation. Coaches had a weekly consultation
call with an experienced clinical sleep psychologist (JLM), who reviewed recommendations
and provided guidance in facilitating client motivation to implement strategies for
enhanced relaxation, reduced stress, and improved sleep.
Table 1 Topics and information provided at each session of the behavioral modification (BM)
intervention.
|
Week
|
Material Covered
|
Activities
|
Homework
|
|
1
|
Topic: Welcome to the SCORE study
|
-
Introduction to behavioral modification
-
Importance of sleep in relaxation and stress reduction
-
Introduction to the sleep diary
|
|
|
|
2
|
Topic: Introduction to relaxation and healthy sleep habits
|
|
|
-
Review of sleep questionnaires
-
Review/discuss sleep diary
-
Action plan: sleep hygiene changes
|
|
|
3
|
Topic: Biological Sleep Need: Part 1
|
-
Review other sleep questionnaire scores
-
Learn about biological sleep need and spending the right amount of time in bed.
-
Learn rationale for consistent sleep schedules
|
|
|
|
4
|
Topic: Biological Sleep Need: Part II
|
|
|
-
Review/discussion of sleep diary
-
Adjust sleep timing if needed
-
Action plan: revise sleep schedule
|
|
|
5
|
Topic: The Circadian Clock
|
-
Review circadian rhythm questionnaire scores
-
Learn about circadian rhythms
-
Adjust your sleep schedule if needed
|
|
|
|
6
|
Topic: Resetting the circadian clock
|
|
|
|
|
|
7
|
Topic: Travel and Jet Lag
|
|
|
|
|
|
8
|
Topic: Summary and review of progress
|
|
|
-
Review/discussion of sleep diary
-
Action plan: continue with what is working, and identify other improvements
|
|
|
9
|
Topic: Stress impairs sleep and well-being
|
-
Identify daily stress patterns
-
Learn to “put the day to rest”
-
Guidelines for technology
|
|
|
|
10
|
Topic: Stress Reduction - Putting the mind to rest
|
-
Techniques for stress reduction
-
Learn to quiet the mind at bedtime
-
Awareness, acceptance and letting go
|
|
Action plan: sleep hygiene+sleep schedule+bedtime routine+mindfulness Daily sleep
diary
|
|
11
|
Topic: Relaxation - Separating from thoughts
|
|
Techniques for relaxationObstacles to long-term prioritization of sleep
|
Thoughts as leaves on a stream
|
|
|
12
|
Topic: Surfing the sleep wave
|
|
|
|
|
Control group: Equal Attention (EA)
Participants randomized to the EA group received equal time periods of attention with
their assigned trainer viewing 10 min slide presentations covering generalized healthy
lifestyle advice absent of any specific guidance related to relaxation, stress reduction
or sleep. The EA intervention included the following topics: healthy relationships,
brain power, movement therapy (ergonomics), general health screening, environmental
health, cancer screening, tobacco/nicotine, time management, basic hygiene practices,
and preventable diseases/immunizations.
Exercise training
Supervised, periodized, functional resistance training
Resistance training was provided to all participants using a previously validated,
evidence-based functional resistance training program [43]. The training regimen consisted of a 3-cycle, nonlinear program in which acute program
variables including exercise selection, volume, and intensity were varied over both
the 4-week mesocycles and within the weekly microcycles. The volume and/or intensities
of each training session were categorized as high (H), moderate (M), or low (L) and
applied to both upper- and lower-body exercises on a given day during the course of
each week of training. Individual sessions lasted approximately 20–25 min.
Remotely-guided aerobic exercise training
Aerobic training was performed on treadmill ergometers and each session lasted 30 min.
Heart rate was recorded using a physiological status monitor affixed to a chest strap
(BioHarness-3™; Zephyr Technologies, Annapolis, MD, USA) paired via Bluetooth to a
smartphone. Individualized target heart rate zones (THRZ) were displayed and at the
end of each training session, the data was uploaded remotely to a storage server.
Prior to the next session, study personnel reviewed the data via the UCLA Exercise
Physiology Research Laboratory Digital Health Network (DHN) – a secured, encrypted
web portal used to gather and analyze participant data [5] – and adjusted exercise parameters for future sessions accordingly. The remote-guidance
process, outlined in more detail in a previous publication [15], was used to define individualized program progression without requiring a dedicated
strength coach to supervise each session. The automated system enabled aerobic training
zones to be defined for each participant based on the physiological events underlying
the incremental exercise response. These zones were presented in terms of lower- and
upper-limit heart rates at the ventilatory and metabolic thresholds, Respectively.
Individualized exercise prescription for all subsequent treadmill training sessions.
The target heart rate zones were adjusted during the study to account for progression
of training effects.
Outcome measures
All outcomes were measured at baseline and at completion of the 12-week intervention.
Fitness measures
Aerobic performance
Aerobic capacity (VO2max), the metabolic threshold (lactate threshold determined by gas exchange, VO2θ) and the ventilatory threshold (the onset of hyperventilation in response to metabolic
acidosis, VCO2θ) were measured from an incremental, symptom-limited maximal treadmill exercise test
using a validated portable metabolic measurement system (Oxycon™ Mobile; CareFusion,
Yorba Linda, CA, USA) and standardized approach [14].
Muscle strength and endurance
Upper- and lower-body isotonic muscle strength was measured by determining 1-repetition
maximum (1-RM) on a seated chest press (Technogym, Cesena, Italy) and a seated leg
press (Eagle NX; Cybex International, Medway, MA, USA) using the procedure described
by the National Strength and Conditioning Association [3]. After 10 min of rest, muscle endurance was measured as the number of repetitions
to failure using 85% of 1-RM values.
Lower-body power
Participants stood on a previously validated electronic jump mat (Just Jump; Probotics,
Inc., Huntsville, AL, USA) [28] with feet hip-width apart and then performed a countermovement jump for maximal
height. Three trials were given with 20–30 s rest between each attempt. The best trial
was used to calculate peak and average lower-body power using published equations
that require vertical jump height and body mass [21].
Body composition
Percentage body fat (BF%) was measured using a recently validated [13] octipolar, multi-frequency, multi-segmental bioelectrical impedance device (R20;
InBody Co., Seoul, South Korea). Because hydration state has a marked influence on
bioelectrical impedance analysis (BIA) results, participants were instructed to remain
hydrated and not exercise during the 2 h period before testing. Data were collected
after at least 3 h of fasting and voiding.
Sleep measures
All participants completed a battery of sleep questionnaires using the DHN. Questionnaires
included the Pittsburgh Sleep Quality Index (PSQI) [9] to measure general sleep quality, the Sleep Hygiene Index (SHI) [32] to assess sleep hygiene behaviors, and the Morningness-Eveningness Questionnaire
(MEQ) [23] to determine chronotype tendency and circadian phase preference. The information
was used both to measure outcomes and to inform individualization of the sleep coaching
intervention recommendations. Participants were also asked to complete a daily sleep
diary, which was used to compute average time in bed (TIB), total sleep time (TST),
and sleep efficiency (SE) calculated as TST/TIB. The sleep diary, which was completed
electronically on a smartphone-type device and uploaded to the DHN, was based on that
described by Carney et al. [10] and included items to assess bedtime at night, time to fall asleep, time awake at
night, final morning wake time and morning rise time.
Heart rate variability
Autonomic balance, which is thought to reflect stress, was evaluated using heart rate
variability (HRV) derived from a 5 min, resting single-channel electrocardiogram chest
strap biosensor (Zephyr Bioharness) obtained with participants seated in a darkened,
quiet room. Time domain indices (ms) included standard deviation of normal-to-normal
intervals (SDNN), and root mean square differences of the standard deviation (RMSSD).
Frequency domain analysis included low-frequency (LF) component (frequency range 0.04–0.15 Hz)
and high-frequency (HF) component (frequency range 0.15–0.4 Hz), in absolute units
(ms2), and the low-frequency to high-frequency ratio (LF/HF). Elaborated methodology is
presented in Supplemental Digital Content.
Statistical analysis
We determined from pilot testing, and allowing for 15% missing data, that a sample
size of 44 participants would be sufficient to assess changes in fitness outcomes
based on α=0.05 and β=0.20. All data were exported to IBM SPSS Statistics for Windows,
version 24 (IBM Corp., Armonk, N.Y., USA) for analysis. Descriptive statistics are
presented as mean (SD). Grubbs’ test was employed to detect potential outliers and
none were found. Prior to comparisons, all variables were assessed for normality via
Shapiro-Wilk tests. Within group comparisons at baseline and after 12 weeks were made
by paired t-tests and Wilcoxon signed-rank tests for normally and non-normally distributed variables,
respectively. Changes between groups were analyzed by Welch’s independent t-tests (normal) or Mann-Whitney U tests (non-normal). Statistical significance was determined by α=0.05 and all tests
were two-tailed. Given that this is one of the first randomized trials testing a behavioral
modification model, we did not employ strict type 1 error control; however, we limited
the number of main outcome measures and based our interpretation on the pattern of
results seen for each domain rather than on individual statistical tests.
Results
Thirty-eight participants completed the 12-week training program (36 aerobic exercise
and resistance sessions) without injuries or serious adverse events, although three
participants required an additional week to complete the program due to minor illness
or vacation. Due to scheduling conflicts, six participants were not able to return
for 12-week follow-up testing, therefore 32 participants (15 receiving BM and 17 receiving
EA) were included in the final analysis of fitness measures ([Table 2]). HRV data for one additional participant was corrupted during recording and was,
therefore, not included in the analysis ([Table 3]). Eleven participants did not perform 12-week follow-up sleep measurements ([Table 3]). Sleep diary data were available for 19 participants. The losses to follow-up did
not result in uneven group size for final analyses.
Table 2 Fitness measures at baseline and after 12 weeks for the equal-attention control and
behavioral modification groups.
|
Fitness Measures
|
Equal-Attention (n=17; 10 males)
|
Behavioral Modification (n=15; 9 males)
|
P-between
|
|
Baseline
|
12 Weeks†
|
Change
|
P-within
|
Baseline
|
12 Weeks‡
|
Change
|
P-within
|
|
Age
|
33.3 (6.3)
|
-
|
-
|
-
|
32.7 (4.7)
|
-
|
-
|
-
|
0.748
|
|
Body weight (kg)
|
72.7 (12.6)
|
72.5 (12.0)
|
−0.3 (1.8)
|
0.530
|
74.5 (11.6)
|
74.1 (10.5)
|
−0.4 (2.7)
|
0.538
|
0.859
|
|
BMI (kg/m
2
)
|
23.8 (2.6)
|
23.7 (2.3)
|
−0.1 (0.6)
|
0.577
|
25.0 (3.1)
|
24.9 (2.9)
|
−0.1 (0.9)
|
0.630
|
0.922
|
|
Body fat (%)
|
20.1 (7.1)
|
18.6 (6.8)
|
−1.4 (1.9)
|
0.006
|
22.1 (6.5)
|
18.5 (6.5)
|
−3.6 (2.6)
|
<0.001
|
0.011
|
|
Fat mass (kg)
|
14.6 (5.7)
|
13.3 (5.1)
|
−1.3 (1.7)
|
0.006
|
16.5 (5.7)
|
13.8 (5.6)
|
−2.7 (1.7)
|
<0.001
|
0.021
|
|
Fat-free mass (kg)
|
58.3 (12.1)
|
59.2 (12.1)
|
0.9 (1.9)
|
0.083
|
58.1 (10.5)
|
60.4 (9.4)
|
2.3 (3.4)
|
0.021
|
0.165
|
|
CP 1-RM (kg)
|
62.1 (26.5)
|
73.8 (28.9)
|
11.6 (9.7)
|
<0.001
|
58.9 (25.4)
|
75.0 (31.0)
|
16.1 (16.3)
|
0.002
|
0.551
|
|
LP 1-RM (kg)
|
92.4 (33.4)
|
120.7 (36.2)
|
28.3 (14.5)
|
<0.001
|
104.8 (39.3)
|
134.2 (54.4)
|
29.5 (20.4)
|
0.001
|
0.766
|
|
CP 85% 1-RM (kg)
|
52.8 (22.5)
|
62.7 (24.6)
|
9.9 (8.3)
|
<0.001
|
50.0 (21.6)
|
63.7 (26.4)
|
13.7 (13.8)
|
0.002
|
0.551
|
|
CP 85% 1-RM (reps)
|
5.2 (1.9)
|
7.1 (2.0)
|
1.9 (2.6)
|
0.008
|
5.0 (1.4)
|
7.9 (1.9)
|
2.9 (1.7)
|
0.001
|
0.207
|
|
LP 85% 1-RM (kg)
|
78.5 (28.4)
|
102.6 (30.8)
|
24.0 (12.3)
|
<0.001
|
89.1 (33.4)
|
114.1 (46.2)
|
25.0 (17.3)
|
0.001
|
0.766
|
|
LP 85% 1-RM (reps)
|
7.2 (2.6)
|
9.2 (4.1)
|
1.9 (3.7)
|
0.067
|
7.5 (3.9)
|
8.6 (2.3)
|
1.1 (4.0)
|
0.076
|
0.823
|
|
Leg power
peak
(W)
|
7861 (349)
|
8161 (471)
|
300 (208)
|
<0.001
|
7760 (339)
|
8283 (374)
|
523 (190)
|
<0.001
|
0.006
|
|
Leg power
avg
(W)
|
1597 (145)
|
1733 (126)
|
137 (58)
|
<0.001
|
1534 (138)
|
1746 (167)
|
212 (79)
|
<0.001
|
0.005
|
|
VO
2
max (L/min)
|
3.06 (0.64)
|
3.29 (0.65)
|
0.23 (0.08)
|
<0.001
|
3.13 (0.53)
|
3.57 (0.51)
|
0.44 (0.12)
|
<0.001
|
<0.001
|
|
rVO
2
max (mL/min/kg)
|
42.1 (5.1)
|
45.4 (4.9)
|
3.3 (1.0)
|
<0.001
|
42.6 (7.9)
|
48.7 (7.1)
|
6.1 (2.1)
|
<0.001
|
<0.001
|
|
VO
2
θ (L/min)
|
2.03 (0.39)
|
2.38 (0.42)
|
0.34 (0.13)
|
<0.001
|
2.12 (0.32)
|
2.76 (0.37)
|
0.64 (0.13)
|
<0.001
|
<0.001
|
|
VO
2
θ/VO
2
max (%)
|
67 (5)
|
73 (6)
|
6 (3)
|
<0.001
|
68 (6)
|
77 (5)
|
9 (3)
|
<0.001
|
0.009
|
|
VCO
2
θ (L/min)
|
2.79 (0.54)
|
3.01 (0.59)
|
0.22 (0.10)
|
<0.001
|
2.90 (0.53)
|
3.26 (0.50)
|
0.36 (0.13)
|
<0.001
|
0.098
|
Values are mean (SD). No significant differences were observed between groups at baseline.
VO2max=maximum oxygen uptake; rVO2max=maximum oxygen uptake normalized by body mass; VO2θ=oxygen uptake at metabolic threshold; VO2θ/VO2max=metabolic threshold as a percent of maximum oxygen uptake; VCO2θ=oxygen uptake at ventilatory threshold; BMI=body mass index; CP=chest press; 1-RM=1-repetition
maximum; LP=leg press
Table 3 Sleep and heart rate variability measures at baseline and after 12 weeks for the
equal-attention control and behavioral modification groups.
|
Sleep Measures
|
Equal-Attention (n=14)
|
Behavioral Modification (n=13)
|
P-between
|
|
Baseline
|
12 Weeks†
|
Change
|
P-within
|
Baseline
|
12 Weeks‡
|
Change
|
P-within
|
|
PSQI
|
4.9 (3.0)
|
3.1 (1.5)
|
−1.7 (3.1)
|
0.049
|
4.5 (2.7)
|
3.0 (1.3)
|
−1.5 (2.9)
|
0.063
|
0.830
|
|
MEQ
|
39.0 (8.1)
|
40.4 (8.1)
|
1.4 (2.7)
|
0.070
|
37.2 (6.8)
|
38.9 (5.3)†
|
1.8 (2.7)
|
0.044
|
0.322
|
|
SHI
|
17.1 (5.5)
|
16.5 (4.2)
|
−0.6 (4.2)
|
0.577
|
17.2 (4.5)
|
15.2 (2.6)
|
−2.0 (4.3)
|
0.121
|
0.416
|
|
TST (hrs)
|
6.5 (0.7)‡
|
6.9 (0.6)‡
|
0.4 (0.7)
|
0.119
|
6.7 (0.6)#
|
6.9 (1.0)#
|
0.2 (1.1)
|
0.237
|
0.898
|
|
TIB (hrs)
|
7.3 (0.8)‡
|
7.4 (0.8)‡
|
0.1 (0.7)
|
0.619
|
7.6 (0.4)#
|
7.7 (1.0)#
|
−0.1 (1.1)
|
0.604
|
0.816
|
|
SE (%)
|
88.7 (5.9)‡
|
92.6 (4.7)‡
|
3.9 (4.3)
|
0.015
|
88.6 (7.1)#
|
89.8 (7.2)#
|
1.2 (8.0)
|
0.866
|
0.368
|
|
HRV Measures
|
Equal-Attention (n=16)
|
Behavioral Modification (n=15)
|
P-between
|
|
Baseline
|
12 Weeks†
|
Change
|
P-within
|
Baseline
|
12 Weeks‡
|
Change
|
P-within
|
|
fC (beats/min)§
|
68.1 (1.9)
|
67.2 (1.8)
|
−0.9 (1.6)
|
0.031
|
69.3 (1.3)
|
67.2 (2.1)
|
−2.1 (2.2)
|
0.002
|
0.104
|
|
SDNN (ms)
|
60.7 (4.8)
|
63.5 (3.9)
|
2.8 (3.0)
|
0.002
|
59.5 (4.1)
|
63.0 (4.3)
|
3.5 (3.5)
|
0.002
|
0.446
|
|
RMSSD (ms)
|
53.8 (3.9)
|
56.5 (3.4)
|
2.7 (3.5)
|
0.008
|
52.7 (4.8)
|
57.5 (3.3)
|
4.8 (4.3)
|
0.001
|
0.086
|
|
HF (ms
2
)§
|
43.0 (2.9)
|
43.2 (2.0)
|
0.3 (2.3)
|
0.616
|
40.1 (3.0)
|
42.7 (1.5)
|
2.6 (3.3)
|
0.010
|
0.036
|
|
LF (ms
2
)
|
53.6 (2.3)
|
52.3 (2.5)
|
−1.3 (2.7)
|
0.075
|
53.6 (2.1)
|
51.0 (3.0)
|
−2.6 (2.8)
|
0.003
|
0.197
|
|
LF/HF
|
2.6 (0.4)
|
2.4 (0.4)
|
−0.2 (0.2)
|
0.002
|
2.6 (0.5)
|
2.1 (0.4)
|
−0.5 (0.8)
|
0.007
|
0.740
|
Values are mean (SD). PSQI=Pittsburgh Sleep Quality Index; MEQ=Morningness-Eveningness
Questionnaire; SHI=Sleep Hygiene Index; TST=total sleep time derived from sleep diary;
TIB=time in bed; SE=sleep efficiency (TST/TIB); HRV=heart rate variability; fC=heart
rate; SDNN=standard deviation of normal-to-normal intervals; RMSSD=root mean square
differences of the standard deviation; HF=high-frequency component; LF=low-frequency
component; LF/HF=ratio of low- to high-frequency components; †n=12; ‡n=10; #n=9; §significantly
different at baseline (P<0.05).
Fitness measures ([Table 2])
Aerobic performance: improved significantly in both groups, with an almost 2-fold
greater improvement in nearly all measures in participants who received BM. VO2max increased in the BM group compared to the EA group. Additionally, metabolic threshold
(VO2θ), and the percentage of maximum oxygen uptake at which metabolic threshold occurred
(VO2θ/VO2max) increased in the BM compared to the EA group.
Both upper- and lower-body muscle strength and endurance improved significantly in
all participants, although there was no difference in the magnitude of improvement
between groups. The number of repetitions performed during lower-body endurance testing
did not decrease despite performing the test with roughly 30% more weight. Peak and
average lower-body power also increased significantly in both groups; however, there
was significantly greater improvement in the BM group versus the EA group for both
variables.
Furthermore, body composition improved in both groups, with two-fold greater improvement
in the BM group. Body fat percentage decreased in the BM group compared to the EA
group. Similarly, absolute fat mass decreased nearly 2-fold more in the BM group.
Sleep measures ([Table 3])
For all symptom questionnaires and sleep diary variables, mean baseline values fell
within normal ranges and significant differences between groups for changes in sleep
measures were not observed. However, within the EA group sleep quality was improved
(reduced PSQI score, increased diary SE), whereas the BM group demonstrated a shift
toward morningness in circadian tendency (MEQ).
Heart rate variability ([Table 3])
High-frequency signal (HF) improved in the BM group compared to no change in the EA
group. Although this improvement between groups was statistically significant, it
is notable that this measure was also significantly different between groups at baseline.
Discussion
Research in behavioral modification evaluates the efficacy of various interventions
designed to enhance relaxation, reduce stress, and improve sleep. Sleep research mainly
addresses issues specific to patients with clinically diagnosed sleep disorders in
an effort to restore sleep quality to healthy levels. Considering that a high proportion
of the general population receives inadequate sleep on a nightly basis [30]
[47], even in the absence of diagnosable sleep disorders, the potential exists for improvements
in health and fitness in broader subclinical populations as well. We believe this
to be the first randomized, double-blind controlled trial integrating a novel, multicomponent
BM intervention into an exercise training program in healthy volunteers, delivered
by fitness professionals under the supervision of a clinical sleep psychologist. This
delivery method is novel and contributes to existing literature by expanding the models
through which BM interventions related to sleep can be delivered.
Physical performance
Prior research has shown decrements in aerobic performance resulting from sleep loss.
Mougin et al. [36] observed reduced VO2max and earlier onset of VO2θ following partial sleep loss. Several other researchers have concluded chronic sleep
deprivation leads to decreased athletic performance in various athletes [19]
[20]. The current study demonstrates that this decrement in aerobic performance can be
reversed when behavioral strategies designed to improve sleep are introduced. Interestingly,
these aerobic improvements were realized without compromising development of muscular
strength or endurance.
Similarly, a few other studies have examined the effect sleep has on lower-body power.
Using a jump mat similar to the one in the present investigation, Prentice et al.
[40] reported a significant decrease in lower-body power immediately following a night
of curtailed sleep. As with aerobic performance, we believe our study is the first
to demonstrate an increase in this measure after an intervention targeting sleep-related
habits and behaviors. Therefore, our data strengthens the evidence for a relationship
between sleep and physical performance; however, the mechanism by which such a relationship
exists remains to be established.
Body composition
Recent evidence suggests that sleep deprivation is linked to increased fat mass and
obesity [26]
[49]; however, the literature is equivocal on the effect of exercise as an additional
factor. Farnsworth et al. [17] found that body composition, independent of physical activity levels, predicted
the risk for developing a sleep disorder. Conversely, a meta-analysis of patients
with obstructive sleep apnea concluded that exercise—and not solely a decrease in
body mass—was responsible for the reduction in symptom severity [2]. Although the results in the present study suggest improved sleep habits augment
the effect of exercise on body composition, the lack of significant changes in traditional
sleep metrics makes it difficult to draw conclusions about the pathway through which
these improvements may occur. It is possible that improved regularity of sleep habits,
which is not easily captured in questionnaires or short-term sleep diaries, leads
to improved health habits in general.
Sleep quality
Interestingly, although our BM program had a substantial component focused on improving
sleep, the only significant change in sleep outcomes was a slight shift toward morningness
on the MEQ in the BM group. It is possible that regular exercise in combination with
sleep coaching served to strengthen the sense of alertness during the daytime hours
via changing circadian rhythmicity and perhaps improve overall daytime energy levels,
subjectively. Despite these improvements, we did not see major differences between
groups in other sleep outcome measures. At baseline, participants did not demonstrate
clinically significant sleep disorders, and commonly used sleep questionnaires may
not have captured the degree of improvements in nighttime sleep quality one would
see in clinical populations. Since our population consisted of healthy volunteers,
any sleep deficiencies would be considered sub-clinical and, therefore, improvement
in sleep resulting from our intervention could constitute a shift from adequate to
optimal, rather than pathologic to healthy. In addition, because the intervention
was individualized for each participant, some participants improved the regularity
of their sleep schedule across weekdays and weekends, whereas others increased the
total amount of time they spent in bed on a nightly basis, and others reduced sleep-disruptive
environmental factors to decrease nighttime awakenings. This variation cannot be captured
with aggregate measures. One limitation of our ability to monitor changes in sleep
habits and routines was low adherence to completion of the daily sleep diaries during
the assessment phases. The reason for this is somewhat unclear but may have been related
to a preference for the paper-and-pencil sleep diaries used during the intervention
(rather than the electronic diaries used during the assessment phases).
Heart rate variability
Another important goal of our BM program was to reduce stress, and we attempted to
capture such an effect by measuring HRV – a reliable, noninvasive marker that reflects
the balance of the autonomic nervous system (i. e., sympathetic and vagal neural influences)
on heart rate. Previous studies have shown that increased sympathetic activity and
decreased parasympathetic activity results in reduced HRV, which has been strongly
associated with developing cardiovascular pathology [12]
[24]. In the present investigation, SDNN and RMSSD increased in both groups while resting
fC and LF/HF ratio decreased. Because a lower LF/HF ratio suggests dominance of parasympathetic
activity over sympathetic activity, these results suggest that exercise may exhibit
a cardioprotective effect by simultaneously reducing sympathetic outflow and augmenting
vagal tone. Furthermore, a significantly greater change in HF in the BM group compared
to the EA group suggests that the BM intervention may have had an additive effect
of increasing parasympathetic modulation. Many investigators have previously shown
a strong positive association between HF and sleep quality in a variety of populations
[8]
[48]. This observed difference in HF HRV improvement parallels the benefit of our BM
intervention.
Strengths and limitations
One intriguing physiologic process that should be explored in future studies is circadian
rhythms. A recent investigation found that, independent of sleep deprivation, chronobiological
misalignment led to decreased insulin sensitivity and increased systemic inflammation
[29]. In addition to being linked to caloric intake and glucose metabolism [4], misaligned circadian rhythms have been shown to interfere with sleep and hinder
athletic performance [38]. As described above, we did see a change in MEQ scores coupled with reduced fatigue
during the intervention. Prescribed sleep scheduling—a component of the BM intervention—is
regarded as a potential treatment for circadian rhythm disruption [42]; therefore, it is plausible that sleep coaching stimulated circadian realignment
which, in turn, increased the productivity of exercise training sessions even if it
did not improve sleep quality as measured by the questionnaires employed in this trial.
The use of targeted physiological measures (e. g., melatonin levels, dim light melatonin
onset time) is needed to definitively test this hypothesis.
Although statistical power was calculated in advance, one limitation is the relatively
small sample included in our study, which may have increased the risk for both type
1 and type 2 errors. This is partially ameliorated by careful selection of outcome
measures and interpretation of results based on the pattern of findings in each domain;
however, future larger studies are recommended to confirm our findings. Another limitation
was the inability to directly detect meaningful changes in sleep quality using traditional
clinical questionnaires in addition to the limited adherence to daily sleep diaries.
Future studies should incorporate 24 h objective monitoring with actigraphy to evaluate
whether objective changes in sleep can be achieved with this sleep coaching model.
Additionally, although all participants received generalized dietary guidance, we
did not assess or control dietary intake. It is possible that nutritional differences
may have emerged between the groups and impacted our results. Similarly, we did not
control for other behavioral and lifestyle factors, such as stress, emotional and/or
spiritual health, or family and community support.
A unique aspect of our study was the use of trained fitness professionals to deliver
the BM intervention program. The intervention program was developed by an experienced
sleep psychologist (JLM) and was structured so it could be delivered within the health
club setting (e. g., use of visual materials that can be shown on a computer screen
or printed) in a relatively short period of time (10 min). This has the likelihood
of increasing access to accurate information about sleep for fitness clients, and
for delivering messages about the importance of sleep health by linking sleep-related
behavioral changes (e. g., getting out of bed at a consistent time, avoiding stimulating
activities near bed time) to fitness goals. Notably, the trainers in this study did
consult weekly with the sleep psychologist (JLM), which facilitated their ability
to master the sleep coaching program and work effectively with clients in changing
their sleep-related behaviors. We cannot conclude that use of the intervention materials
in the absence of this ongoing learning would have lead to the same positive outcomes.
The improved aerobic fitness profiles in both the BM and EA groups over baseline measures
serve to further validate the use of a remotely-guided digital health system in directing
an effective aerobic training program. In this study, as in previous studies that
have incorporated the DHN, participants were able to capture and transmit biometric
data from a fitness club or their home, facilitating rapid translation of this information
into personalized exercise progression and performance feedback, which was then available
to inform the very next exercise session. Previously validated in a population of
emergency responders [15], semi-automated systems, such as our DHN-enabled platform, could serve to enhance
the reach and productivity of medical practitioners, personal trainers, and health
coaches working with a wide variety of clientele without sacrificing fitness results.
Conclusions
This study demonstrates that a novel, multicomponent behavioral modification intervention
delivered by fitness professionals with the goal of optimizing fitness outcomes was
more efficacious at improving various measures of aerobic performance, body composition,
lower-body power, and HRV compared with identical exercise training with an equal-attention
control. Our findings indicate not only that sleep interventions have the potential
to directly enhance physical performance and autonomic cardiac regulation, but that
these interventions can be successfully administered by trained fitness professionals
in a non-medical setting. Future investigations should explore the replicability and
utility of these results in special populations such as elite athletes and those afflicted
with lifestyle diseases.
