This article was downloaded by: [Memorial University of Newfoundland]On: 18 July 2014, At: 14:58Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Sports SciencesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/rjsp20

Effect of time-of-day-specific strength trainingon maximum strength and EMG activity of the legextensors in menMilan Sedliak a , Taija Finni a , Jussi Peltonen a & Keijo Häkkinen aa Department of Biology of Physical Activity and Neuromuscular Research Centre , Universityof Jyväskylä , Jyväskylä, FinlandPublished online: 17 Jul 2008.

To cite this article: Milan Sedliak , Taija Finni , Jussi Peltonen & Keijo Häkkinen (2008) Effect of time-of-day-specificstrength training on maximum strength and EMG activity of the leg extensors in men, Journal of Sports Sciences, 26:10,1005-1014, DOI: 10.1080/02640410801930150

To link to this article: http://dx.doi.org/10.1080/02640410801930150

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Effect of time-of-day-specific strength training on maximum strengthand EMG activity of the leg extensors in men

MILAN SEDLIAK, TAIJA FINNI, JUSSI PELTONEN, & KEIJO HAKKINEN

Department of Biology of Physical Activity and Neuromuscular Research Centre, University of Jyvaskyla,

Jyvaskyla, Finland

(Accepted 18 January 2008)

AbstractIn this study, we examined the effects of time-of-day-specific strength training on maximum strength and electromyography(EMG) of the knee extensors in men. After a 10-week preparatory training period (training times 17:00–19:00 h), 27participants were randomized into a morning (07:00–09:00 h, n¼ 14) and an evening group (17:00–19.00 h, n¼ 13). Bothgroups then underwent 10 weeks of time-of-day-specific training. A matched control group (n¼ 7) completed all testing butdid not train. Unilateral isometric knee extension peak torque (MVC) and one-repetition maximum half-squat were assessedbefore and after the preparatory training and after the time-of-day-specific training at times that were not training-specific(between 09:00 and 16:00 h). During training-specific hours, peak torque and EMG during MVC and submaximumisometric contraction at 40% MVC were assessed before and after the time-of-day-specific training. The main finding wasthat a significant diurnal difference (P5 0.01) in peak torque between the 07:00 and 17:00 h tests decreased after time-of-day-specific training in the morning group but not in the evening or control groups. However, the extent of this time-of-day-specific adaptation varied between individuals. Electromyography during MVC did not show any time-of-day-specificadaptation, suggesting that peripheral rather than neural adaptations are the main source of temporal specificity in strengthtraining.

Keywords: Diurnal, time-of-day-specific training, muscle strength, electromyography

Introduction

Time of day has been shown to affect various indices

of maximal neuromuscular performance in humans.

It has been shown repeatedly that voluntary muscle

strength varies across a day in a predictable manner.

According to Reilly and colleagues (Reilly, Atkinson,

& Waterhouse, 2000), peak values of muscle strength

are observed in the early evening. The diurnal

minimum is almost always found in the early

morning. Such a diurnal pattern is typically seen in

maximal strength of the knee extensor muscles, both

in maximal voluntary isometric (Callard, Davenne,

Gauthier, Lagarde, & Van Hoecke, 2000; Coldwells,

Atkinson, & Reilly, 1994; Guette, Gondin, & Martin,

2005) and isokinetic (Deschenes et al., 1998;

Nicolas, Gauthier, Bessot, Moussay, & Davenne,

2005) conditions. Peak-to-trough variation has been

reported to range from 6% (Guette et al., 2005) to

18% (Coldwells et al., 1994).

The exact mechanism responsible for diurnal

variation in strength remains unclear. Recent find-

ings by Guette et al. (2005) suggest that neural drive

to the muscle does not vary with time of day. In line

with Martin and colleagues (Martin, Carpentier,

Guissard, Van Hoecke, & Duchateau, 1999), they

proposed that circadian changes in muscle contrac-

tile properties could be partially responsible for

diurnal variation in strength. In contrast, it has been

reported that both central (neural input to the

muscles) and peripheral (contractile state of the

muscle) mechanisms may be altered across a day

(Castaingts, Martin, Van Hoecke, & Perot, 2004;

Gauthier, Davenne, Martin, Cometti, & Van

Hoecke, 1996).

As well as diurnal changes in maximum strength

performance, time of day was proposed to affect the

adaptation to strength training – so-called circadian

specificity in training (Hill, Cureton, & Collins,

1989). In other words, after training at a particular

Correspondence: M. Sedliak, Department of Biology of Physical Activity, University of Jyvaskyla, PO Box 35 (VIV226), FIN-40014 Jyvaskyla, Finland.

E-mail: milan.sedliak@sport.jyu.fi

Journal of Sports Sciences, August 2008; 26(10): 1005–1014

ISSN 0264-0414 print/ISSN 1466-447X online � 2008 Taylor & Francis

DOI: 10.1080/02640410801930150

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

time of day, an individual would be stronger at that

time of day than at other times.

To our knowledge, only one scientific article has

dealt with circadian (temporal) specificity to strength

training. Souissi and colleagues (Souissi, Gauthier,

Sesboue, Larue, & Davenne, 2002) conducted a 6-

week training study during which two groups of

males performed an identical training programme

either only in the morning (07:00–08:00 h) or only in

the evening (17:00–18:00 h). They found that

‘‘adaptation to strength training is greater at the

time of day at which training was conducted than at

other times’’ (Souissi et al., 2002). No difference was

observed in the absolute increase in one-repetition

maximum between the two groups.

These results suggest that a typical diurnal pattern

might be altered by time-of-day-specific training.

However, the roles of central and/or peripheral

mechanisms involved in the temporal specificity of

exercise training have yet to be addressed. The

recording of voluntary muscle strength together with

electromyographic (EMG) activity of trained agonist

muscles would enable the determination of whether

diurnal variation in muscle strength and the adapta-

tion to strength training are solely due to events

inside muscle fibres or there is a contribution of

alternating neural drive and/or properties of fibre

membrane and motor units. Therefore, the main

purpose of this study was to examine the effects of

time-of-day-specific training on maximum isometric

strength and maximum myoelectrical activity of the

knee extensors in moderately trained men.

Methods

Participants

Initially, 75 participants with similar health status

and physical condition volunteered for the study (52

participants formed two training groups, 23 partici-

pants formed a control group). However, the data for

only 34 men were included in the present report.

Sixteen participants dropped out and the data for a

further 25 (mostly originally in the control group)

were affected by a recording error discovered during

the final analysis phase.

The participants had no experience of lower

extremity strength training and no participant had

performed regular physical activity more than once a

week during the 3 years prior to the experiment.

Before the experiment, background information

on physical activity, sleep habits, motivational

characteristics, and medical history was collected

by questionnaires. All participants were considered

healthy and had no medical contraindications that

would affect the results of the study. Shift workers

were excluded. The Circadian Type Questionnaire

(Folkard, Monk, & Lobban, 1979) was used to

evaulate ‘‘morningness’’ versus ‘‘eveningness’’, abil-

ity to overcome drowsiness, and flexibility of sleeping

habits. In addition, length of sleep was recorded

throughout the test days and the day before. This

study was approved by the Ethics Committee of the

Central Hospital of Central Finland and an informed

consent form was signed by participants before the

investigation.

Experimental protocol

The chronology of the experiment is presented in

Figure 1. The familiarization session was used to

accustom the participants with the measurement

apparatus in the week preceding the experiment.

Subsequently, all participants were tested using two

different designs:

. Time-of-day-non-specific. The tests in this design,

which were administered on three occasions

(Baseline, Middle, and End), beginning at a

random time of day (between 09:00 and

16:00 h), were identical for each individual,

although the start times differed between

participants. For the purpose of this report,

unilateral isometric knee extension peak torque

of the right leg at a knee angle of 1208 (MVC)

and one-repetition maximum (1-RM) half-

squat were selected among the variables mea-

sured.

. Time-of-day-specific. In this design, the tests

were performed on two separate occasions –

Pre and Post (Figure 1). Both the Pre and Post

test consisted of the same set of strength

Figure 1. Chronology of the experiment. The distance between two consecutive vertical bars is equal to 1 week. Weeks connected with

horizontal line are the training weeks; weeks with no fill represent the test weeks.

1006 M. Sedliak et al.

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

measurements performed repeatedly at four

different time points across two consecutive

days (Day 1 and Day 2). Morning on Day 1 was

the first session for all participants. The time

points were as follows: morning, 07:00–08:00 h;

noon, 12:00–13:00 h; afternoon, 17:00–

18:00 h; evening, 20:30–21:30 h. Each test

session lasted 30 min per participant and con-

sisted of 15 min of resting quietly in a supine

position followed by an approximate 15-min

warm-up and then the test procedure. The

warm-up consisted of ten technically performed

half-squats with a 25-kg load. Maximal (MVC)

and submaximal (40% MVC) unilateral iso-

metric knee extensions were performed while

recording myoelectrical activity from the vastus

lateralis, rectus femoris, and vastus medialis

muscles.

The participants were allowed to leave the labora-

tory but requested to maintain their habitual daily

routine between test sessions. To ensure control of

confounding factors, the participants were requested

to refrain from alcohol, sexual and strenuous physical

activity in the day before and throughout the two test

days. The diet was identical for all participants during

the two test days. The overall daily energy intake goal

was set at 10.5 MJ (2500 kcal) per capita per day.

Meals were served within 10 min of the end of each

measurement session. Water and non-caffeine bev-

erages could be consumed ad libitum during and

between meals. Since the participants came to the

laboratory after an overnight fast, 250 ml of orange

juice was given 5 min before the morning test to

decrease the possible risk of hypoglycaemia during

the measurements.

For the purpose of this report, only the 07:00 h

and 17:00 h time points were selected for further

analyses, so as to present data for the training-

specific times during the day with typical and

significant diurnal rhythm in peak torque. Therefore,

Day 2 was selected, since it showed more pro-

nounced diurnal variation in peak torque than Day 1

(Sedliak, Finni, Cheng, Haikarainen, & Hakkinen,

2008).

Training protocol

The entire 20-week training period was completed by

27 participants randomized as a training group.

Seven participants assigned to the control group

(mean age 34 years, s¼ 8; body mass 81.6 kg, s¼ 10;

height 1.79 m, s¼ 0.08) did not train but were

instructed to maintain their pre-experimental physi-

cal activity. The groups were matched according to

their age, body mass, height, and maximum strength

performance before the experiment. During the first

10 weeks of a preparatory training period, all training

sessions took place between 17:00 and 19:00 h.

Thereafter, the participants were randomized either

to a Morning (n¼ 14; mean age 32 years, s¼ 7; body

mass 82.6 kg, s¼ 8.5; height 1.79 m, s¼ 0.09) or

Evening (n¼ 13; mean age 33 years, s¼ 7; body mass

79.9 kg, s¼ 11.3; height 1.80 m, s¼ 0.07) training

group for a time-of-day-specific training period. The

Morning training group performed all training

sessions between 07:00 and 09:00 h, while the

Evening training group trained between 17:00 and

19:00 h. Both the preparatory training and time-of-

day-specific training periods were planned as whole-

body periodized programmes with the main focus on

the knee extensor muscles. Half squats (*908 knee

angle), loaded squat jumps, leg presses, and knee

extensions were the primary exercises. For the details

of the preparatory training period and time-of-day-

specific training period, see Table I. One-repetition

maximum values for all exercises were obtained

during the first sessions at training weeks 1, 11, and

16 according to McDonagh and Davies (1984).

Data collection

1-RM half-squat. Maximal dynamic strength of the

lower limbs was measured using a 1-RM squat test.

The protocol consisted of several submaximal trials

of half-squats separated by 3-min rest periods. The

starting load was estimated to be approximately 75%

of 1-RM and it was progressively increased by 5–

10% at each trial. A barbell with weights, inbuilt in a

Smith machine (Kraftwerk, Finland), was held on

the shoulders. A trial started from the standing

position, squatting slowly down to a 908 knee angle

(announced by an audio cue) followed by extension

of the knees and hips back to the standing position.

The last successful trial performed with the correct

technique was used for further analyses.

MVC. Maximum voluntary isometric force of the

knee extensors was tested using unilateral knee

extensions of the right leg at a knee angle of 1208(1808¼ knee fully extended) in both the time-of-day-

specific and time-of-day-non-specific designs. The

participants were secured in a sitting position on a

knee extension device (Leg Ext/Curl Research, Hur

Oy, Kokkola, Finland) and the same individual

position was ensured in the repeated measures.

Two horizontal safety belts were strapped around

the torso at the chest and waist, and both thighs were

fastened with the cushioned straps placed about the

knee joint. The upper extremities were placed next to

the body holding handgrips. The participants were

asked to produce maximal force rapidly and maintain

it for 3 s. Loud verbal encouragement was given by

the test personnel. The force produced on a

Temporal specificity in strength training 1007

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

transducer attached to the lever arm was amplified

and the analog signal was collected by a biomonitor

ME6000T8 (MEGA Electronics Ltd., Kuopio, Fin-

land) and telemetrically transmitted and stored in the

MegaWin software. The sampling frequency was

1000 Hz. Three trials were performed with a rest

period of 1 min between trials. The trial with the

highest peak force was taken for further analyses.

40% MVC. The participants were asked to produce

force of 40% MVC and maintain it as accurately as

possible with real-time visual feedback and verbal

encouragement for 5 s. Three trials were performed

separated by 5s rest intervals. The data were

recorded and processed as for the MVC. Average

torque and EMG activity were calculated from the

second and third trial; the first trial was considered as

preparatory trial.

EMG. In the time-of-day-specific design, the surface

bipolar electromyogram (EMG) was recorded during

MVC and 40% MVC actions from the vastus

lateralis, rectus femoris, and vastus medialis of the

right leg. The SENIAM recommendations (Her-

mens et al., 1999) were followed when placing silver/

silver chloride pre-gelled electrode pairs with a

sensor diameter of 13 mm and reference electrodes

(Blue Sensor M, AMBU, Ballerup, Denmark). The

location of the electrodes was marked before the first

test session with intradermal ink dots to ensure

reliable positioning over the sessions. On time-of-

day-specific test days, the same pairs of electrodes

were kept in place for the entire test day. Regarding

the EMG signal detection characteristics, the com-

mon mode rejection ratio was 4110 dB, gain was

305, and input impedance was 4100 MO. The

EMG signal amplified by a factor 1000 was recorded

and processed in an identical manner to the MVC

force described above.

Data analysis

Peak torque (MVC) and submaximal torque (40%

MVC) were calculated from low-pass filtered torque

signals. The raw EMG signals were rectified, band-

pass filtered (10–500 Hz), and expressed as the root

mean square from the 100-ms time window around

the MVC peak torque. In addition, the root mean

square was calculated from a 1-s period of the most

constant torque signal during 40% MVC. Subse-

quently, the root mean squares of the vastus lateralis,

rectus femoris, and vastus medialis were averaged to

provide overall activity of the surface knee extensor

muscles (knee extensor EMG; Guette et al., 2005).

Peak torque, power output, and knee extensor EMG

of MVC were normalized with respect to the highest

value recorded for each individual and expressed as a

percentage of this maximum value. The root mean

square to torque ratio (EMG/torque) was calculated

from the normalized values in MVC and absolute

values in 40% MVC.

Statistical analysis

Statistical significance was set at P5 0.05. Standard

descriptive statistics (means and standard deviations)

were calculated. The effects of the preparatory and

time-of-day-specific training on time-of-day-non-

specific MVC and 1-RM were examined by a one-

factor (Pre-to-Post) general linear model with

repeated measures. The effects of time-of-day-

specific training on diurnal variation in performance,

Table I. Basic variables of the preparatory and time-of-day-specific training.

Preparatory training

sessions per week

1th–4th week

2 Percentage of total volume

5th–10th week

2, 5 Percentage of total volume

High-load protocol* 40–70% 100% 60–90% 91%

sets/reps 2–4/6–15 2–5/3–12

Hypertrophy* – –

sets/reps – –

Explosive* – 25–60% 9%

sets/reps – 1/6–10

Time-specific training

sessions per week

11th–15th week

2, 6 Percentage of total volume

16th–20th week

3 Percentage of total volume

High-load protocol* 70–85% 36% 80–100% 38%

sets/reps 2–4/3–8 2–4/1–5

Hypertrophy* 60–70% 49% 60–80% 40%

sets/reps 2–3/8–15# 3–4/8–12

Explosive* 40–55% 15% 50–60% 22%

sets/reps 2/7–8 1–3/5–8

*Load expressed as a percentage of individual one-repetition maximum measured at week 1, week 11, and week 16. #Minimum last set,

maximum last two consecutive sets performed until concentric failure.

1008 M. Sedliak et al.

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

EMG, and EMG/torque were tested by a two-factor

(Time-of-day and Pre-to-Post) general linear model

with repeated measures. Group was included as a

between-participant factor in both models. When the

repeated-measures general linear model revealed

significant F-ratios, pairwise comparisons with

Ryan-Holm-Bonferroni adjustment were employed

to localize significant differences (Atkinson, 2002).

In addition, a paired-samples t-test was applied to

examine Pre to Post changes within the groups in the

time-of-day-specific MVC. Hierarchical clustering

was applied to identify possible sleep inertia or sleep

disturbance subgroups in diurnal change of MVC,

40% MVC, and EMG.

Results

Time-of-day-non-specific design

Strength performance. During the first 10 weeks of the

preparatory training, 1-RM maximum half-squat

increased significantly in all three groups (P5 0.05

in Control and P5 0.001, within-group compar-

ison). However, the increase was more apparent in

the Morning and Afternoon training groups than in

the Control group. When expressed as percentage

difference between the Baseline and Middle test, 1-

RM increased by 28.3% in the Morning group

and 26.6% in the Afternoon group, which were

both significantly higher strength gains (P5 0.05,

between-groups comparison) than that of 10.2% in

the Control group. A similar trend was observed for

peak torque: 10.1%, 9.3%, and 1% in the Morning,

Afternoon, and Control groups, respectively.

The second 10-week period of time-of-day-speci-

fic training did not result in significant improvements

of time-of-day-non-specific strength. The increases

in 1-RM performance were 6.7%, 5.7%, and 2.8% in

the Morning, Afternoon, and Control group, respec-

tively. Peak torque changed by 0.9%, 6.9%, and

3.4%, respectively. The absolute values are shown in

Table II.

Time-of-day-specific design

Peak torque. Significant Time-of-day and Pre-to-Post

main effects were found for both absolute and

relative peak torque values when using the re-

peated-measures general linear model (P5 0.001)

(Figure 2). The interaction between Time-of-day

and Group in the relative values reached P¼ 0.061.

The other interactions (Pre-to-Post6Group,

Time-of-day6Pre-to-Post, Time-of-day6Pre-to-

Post6Group) were not significant. To assess the

biological importance of the Time-of-day6Group

interaction found (P¼ 0.061) in the two-factor

general linear model, further statistical analyses were

conducted to study this phenomenon within each

group separately.

When examining the groups separately with the

repeated-measures general linear model, Time-of-

day and Pre-to-Post main effects were significant for

all three groups (P5 0.001 in Control and After-

noon group, P5 0.01 in Morning group, and

P5 0.05 for all groups, respectively) (Table II).

The Time-of-day6Pre-to-Post interaction reached

significance in the Morning group only (P5 0.05).

When comparing the percentage changes from

Pre07:00 to Post07:00 against percentage changes

from Pre17:00 to Post17:00 by paired-samples t-test,

only the Morning group showed a significant

difference (7.8%, P5 0.05). In this group, the

increase in peak torque was higher at 07:00 h

(13%) than at 17:00 h (5.2%). The above percentage

differences were 0.4% (P¼ 0.882) and 4.3%

(P¼ 0.445) in the Afternoon and Control group,

respectively. The Afternoon group could exert

10.3% and 9.9% higher peak torque at 07:00 and

17.00 h, respectively. The respective values were

11.8% and 7.5% in the Control group. Similar

statistical results (P5 0.05 in the Morning and

non-significant in the Afternoon and Control group)

were obtained from a paired-samples t-test compar-

ison of Pre07:00 to Pre17:00 percentage changes

versus Post07:00 to Post17:00 percentage changes

(Figure 3).

Table II. Peak torque and one-repetition maximum (1-RM) in Time-of-day-non-specific design (Baseline, Middle, End) and Time-of-day-

specific design (Pre at 07:00 and 17:00 h, Post at 07:00 and 17:00 h) (mean+ s).

Test Group Baseline Middle Pre 07:00 Pre 17:00 End Post 07:00 Post 17:00

1-RM semi-squat (kg) Morning (n¼14) 141+ 20 181+ 25* – – 193+29*x – –

Afternoon (n¼12) 127+ 16 158+ 13* – – 168+16* – –

Control (n¼7) 140+ 24 153+ 24* – – 157+21 – –

Peak torque MVC (N � m) Morning (n¼14) 256+ 57 274+ 38 256+31 282+ 40¤ 276+51 287+ 35# 295+ 55

Afternoon (n¼12) 248+ 21 270+ 29 259+48 281+ 47¤ 288+40* 283+ 43# 308+ 53#

Control (n¼7) 205+ 50 205+ 46 194+46 221+ 42¤ 210+41 215+ 52 236+ 39

Note: The tests are presented in chronological order. Statistical significance¼P50.05: *significantly higher than Baseline; xsignificantly

higher than Middle; ¤significantly higher than 07:00 h of the same day; #significantly higher than the same time point in Pre.

Temporal specificity in strength training 1009

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

EMG. Both absolute and normalized EMG were

higher after training in both the Morning and

Afternoon training groups than the Controls

(Pre-to-Post main effect, P5 0.001) (Figure 4).

Percentage changes from Pre-test to Post-test were

as follows: at 07:00 h by 19.6%, 13.8%, and 71.2%,

and at 17:00 h by 16.8%, 8.1%, and 3.2%, in

the Morning, Afternoon, and Control group,

Figure 2. Normalized MVC peak torque at 07:00 h and 17:00 h before and after 10 weeks of time-of-day-specific training.

100%¼ individual highest peak torque from all four time points. M¼Morning training group, A¼Afternoon training group, C¼Control

group. *07:00 h significantly lower than 17:00 h within the same day (P5 0.05).

Figure 3. Peak torque in the Morning and Afternoon training groups expressed as the percentage difference between 07:00 h and 17:00 h

before (PRE) and after (POST) 10 weeks of time-of-day-specific training. Individual values are connected with thin lines, mean values with

bold dash line. Statistical analysis was performed on the mean values.

Figure 4. Absolute EMG (knee extensor root mean square) at 07:00 h and 17:00 h before and after 10 weeks of time-of-day-specific training.

M¼Morning training group, A¼Afternoon training group, C¼Control group *Post-test mean values significantly higher than Pre-test

mean (P5 0.05).

1010 M. Sedliak et al.

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

respectively. However, no significant main effects of

Time-of-day or interactions were found. The relative

difference between 07:00 h and 17:00 h was less

than 2% for all groups in the Pre- test. In the Post-

test, the respective values were 72.3%, 75%, and

7.9% (non-significant) in the Morning, Afternoon,

and Control group, respectively.

EMG/torque ratio. A significant Time-of-day main

effect was observed in normalized EMG/peak torque

ratios in all three groups in both the Pre- and Post-

tests (P5 0.001), the ratios being higher at 07:00 h

than at 17:00 h. The Pre-to-Post main effect was not

significant (P¼ 0.603). During submaximal 40%

MVC, the Pre-to-Post decrease in the ratios was

not significant but there was a Time-of-day main

effect for all three groups (P5 0.001).

Sleep. For all participants, the mean habitual length

of sleep was 7.3 h (s¼ 0.8) and 7.4 h (s¼ 0.8) in the

Pre- and Post-test, respectively. The average sleep

length during the night prior to Day 1 and the night

between Day 1 and Day 2 was 6.8 h (s¼ 0.8) and

7.0 h (s¼ 0.8) in the Pre-test respectively, and 6.9 h

(s¼ 0.8) and 7.0 h (s¼ 0.9) in the Post-test respec-

tively. The waking hours ranged between 05:45 and

06:45 h in the present study; the habitual waking

hours were not recorded. Five participants at the

Pre-test and three participants at the Post-test

reported their sleep length before or during the

experiment to be two or more hours shorter or longer

as compared with their individual average sleep

length. However, these participants were not identi-

fied as a distinct subgroup in terms of MVC, 40%

MVC or EMG diurnal changes.

Discussion

The main finding of the present study was that a 10-

week time-specific strength training period per-

formed in the morning may alter the typical diurnal

pattern in peak torque. However, the extent of this

time-of-day-specific adaptation might vary between

individuals. Based on the stable diurnal EMG signal,

adaptations beyond the muscle fibre membrane

could be the main source of temporal specificity in

strength training. The magnitudes of the increase in

MVC and 1-RM half-squat were similar regardless of

the training time.

Preparatory training

The preparatory training period was applied to

accustom the training groups to resistance training

before the actual time-of-day-specific training and to

‘‘synchronize’’ the activity patterns of both training

groups by training exclusively in the afternoon. By

matching the preparatory training period training

times with the diurnal peak in maximum strength

(for a review, see Drust, Waterhouse, Atkinson,

Edwards, & Reilly, 2005), we attempted to prevent

disturbance of the typical diurnal pattern (Souissi

et al., 2002). The preparatory training period

induced expected adaptation in both the Morning

and Afternoon groups. Their maximum strength

increased from the Baseline to Middle test as

determined both by MVC and 1-RM half-squat.

However, even the Control group increased their 1-

RM half-squat significantly but not their MVC. This

indicates that in all three groups, learning effects

could be a substantial part of the improvements in 1-

RM half-squat but not in the MVC test. The most

likely explanation is the difference in skill demands.

MVC knee extension is a simple one-joint test,

whereas the half-squat requires a more complex

dynamic multi-joint movement. This suggests that

more than one or two familiarization session may be

needed for dynamic tests such as the 1-RM half-

squat in untrained or unaquainted participants.

When measured at Pre-test, the Control and

Training groups exhibited similar diurnal patterns

with significantly lower peak torques at 07:00 h than

at 17:00 h (Fig. 2, left). The minimum-to-maximum

difference was 9%, 10%, and 14% in the Morning,

Afternoon, and Control group, respectively. This is

in line with the data of Guette et al. (2005) on peak

torque recorded during isometric knee extension.

They reported a 12% minimum-to-maximum differ-

ence in physical education students. Similarly,

Giacomoni and colleagues (Giacomoni, Edwards,

& Bambaeichi, 2005) reported a diurnal amplitude

(mesor to peak) in peak torque of 5.9% (non-

significant) in physically active male participants,

suggesting the minimum-to-maximum difference

was about 12%. As shown above, both training

groups had a somewhat lower difference between

07:00 h and 17:00 h than the Controls. In contrast,

normalized knee extensor EMG did not differ

significantly between 07:00 and 17:00 h at the Pre-

test, as reported by Nicolas et al. (2005) and Guette

et al. (2005). The latter study, using the twitch

interpolation technique and EMG normalized to M-

wave, concluded prevailing variation at the muscular

level rather than changes in neural input to knee

extensors muscles. In contrast to our findings,

Gauthier et al. (1996) reported significantly in-

creased EMG activity in the morning compared with

the afternoon. These conflicting results could be in

part accounted for by differences in the test mode

and muscle groups tested (upper vs. lower extremi-

ties). EMG/torque ratios in MVC and 40% MVC

reflected the EMG and peak torque behaviour

across a day. This ratio is considered by some

authors to be a measure of neuromuscular efficiency

Temporal specificity in strength training 1011

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

(Milner-Brown, Mellenthin, & Miller, 1986), but its

actual physiological significance is questionable. In

the present study, neuromuscular efficiency im-

proved significantly from 07:00 h to 17:00 h in both

maximal and submaximal voluntary contractions.

Gauthier et al. (1996) also reported significantly

higher neuromuscular efficiency of the biceps brachii

muscle in the afternoon than in the morning.

However, their findings were the result of a

combined effect of a significant increase in peak

torque and a significant decrease in EMG from the

morning to the afternoon hours.

Time-of-day-specific training

The second 10-week training period resulted in less

pronounced strength increases when tested in the

time-of-day-non-specific design (Table II). The

preceding 10-week preparatory training led most

probably to attenuation in training adaptation during

the second 10-week period as usually observed

(Hakkinen, 1994). Another reason may stem from

the effect of training and/or testing on the subsequent

tests. In the Control group, higher time-of-day-

specific MVC (mean value in both the Pre- and Post-

tests) than time-of-day-non-specific MVC (Middle

and End) in both the Pre- and Post-tests suggests a

training/learning effect from the subsequent testing

separated by 5–7 days. Prevost and colleagues

(Prevost, Nelson, & Maraj, 1999) showed that after

only 2 days of strength training, males can increase

their peak torque by as much as 22%. Therefore,

both the Middle and End tests could in part act as

training sessions in untrained participants of the

Control group. However, unchanged EMG in the

Controls from Pre to Post suggests that, if present,

the training effects were of short duration and/or that

skill learning was a substantial reason for the higher

time-of-day-specific MVC.

When tested in the time-of-day-specific design

(Post), both the Afternoon and Control group

showed significantly lower peak torque values at

07:00 h than at 17:00 h, as examined by t-test. In

contrast, the peak torque of the Morning group

measured at 07:00 h and at 17:00 h did not differ.

The relative increase at the training-specific time

(07:00 h) was 12.1% versus 4.6% at 17:00 h in the

Morning group. The present results are in line with

those of Souissi et al. (2002). However, the time-of-

day-specific adaptation of the Morning group in the

present study was somewhat less pronounced (sig-

nificant by the paired-samples t-test comparison,

non-significant when tested by the more complex

two-factor repeated-measures general linear model)

than in the study of Souissi and co-workers. They

found a strong temporal specificity of the Morning

group with equal or even higher absolute values at

07:00 h than at 17:00 h after training (isokinetic

peak torque during knee extension). It is possible

that in untrained participants, the first weeks of

strength training that involve primary neural adap-

tations (Hakkinen, 1994; Moritani & DeVries,

1979) and learning are more sensitive to time-of-

day-specific training. In the present study, the

participants underwent a 10-week preparatory train-

ing period prior to the time-of-day-specific training

to discount those factors. Therefore, as the training

proceeds with a larger contribution of peripheral

factors, the actual time-specific training effects

might be less pronounced and/or might require

longer than a 10-week training period to reach

statistical significance. Another possible explanation

may stem from inter-individual differences in

responsiveness to time-of-day-specific training.

Some participants in the present Morning group

showed a clear temporal adaptation with higher

peak torque at 07:00 h than at 17:00 h in the Post-

test (Figure 3). Furthermore, another subgroup of

participants (n¼ 9) in the Morning group showed a

decrease in 07:00–17:00 h difference from Pre to

Post. These 9 of the 14 morning group participants

could be considered ‘‘high responders’’ to the

morning time-of-day-specific training. However, it

must be noted that three participants in the ‘‘high

responder’’ subgroup initially (at Pre) had a very

high 07:00 h to 17:00 h difference (420%). At the

Post-test, their 07:00 h to 17:00 h difference was

dramatically decreased but still within the range of

typical diurnal variation (45%). The remaining five

participants more or less maintained their diurnal

variation or even increased the 07:00–17:00 h

difference (‘‘low responders’’). A chronotype could

be one of the mechanisms involved in the variation

in responsiveness. Of all 14 members of the

Morning group, 6 participants (5 low responders,

1 high responder) scored ‘‘neutral’’ on all three

factors of the Circadian Type Questionnaire (Folk-

ard et al., 1979): ability to overcome drowsiness,

flexibility of sleeping habits, and morningness vs.

eveningness. None of the factors could satisfactorily

explain differences in the response to the training at

07:00 h. However, the morningness vs. eveningness

factor seemed to show the most consistent trend in

this regard. Eight high responders were neutral

types; one was a medium morning type. The low

responder group consisted of three neutral type

participants, one medium evening type, and one

extreme evening type. This might mean that the low

responders tend to towards eveningness while the

high responders tend to towards morningness. In

line with this speculation, Souissi et al. (2002)

reported that all seven members of their morning

training group were categorized either as ‘‘moderate

morning type’’ or ‘‘neutral type’’. Therefore, such a

1012 M. Sedliak et al.

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

homogeneous group of potential ‘‘high responders’’

could achieve a clear temporal adaptation to

strength training at 07:00 h. In contrast, in the

present study, six low responders in the Morning

group might have caused blunted temporal adapta-

tion to strength training at 07:00 h.

The present study showed that the EMG activity

of the trained muscles could not explain the trend of

the morning group to increase their performance

more at the time of day of training (07:00 h) than at

another time (17:00 h). The second 10-week time-

of-day-specific training period resulted in a further

significant increase in EMG root mean square in

both training groups compared with the Controls,

but no significant diurnal variation was observed in

either group. Based on this finding, adaptations at

the muscle fibre level rather than changes of the

neural drive, motor unit properties, and/or muscle

membrane properties seem to be the main source of

temporal specificity to strength training. However,

surface EMG has some limitations (e.g. Farina,

Merletti, & Enoka, 2004) and the quantity and

quality of EMG may be affected by the recording

procedure (e.g. surface electrodes vs. intramuscular

electrodes). Our study does not allow us to identify

the actual locus/loci within a muscle tissue or the

mechanisms. Previous studies have suggested several

candidate pathways via which the chronic time-of-

day-specific training stimulus could act upon the

muscles involved, including the hormonal system

(Bird & Tarpenning, 2004; Nindl, Hymer, Deaver,

& Kraemer, 2001; Tuckow et al., 2006) and/or

modulation of muscle DNA transcription activity

(Zambon et al., 2003). All these findings imply that

chronic resistance exercise may have the potential to

alter normal diurnal variation in muscle strength.

However, it is important to note that all the above

studies used an acute exercise protocol. Therefore,

chronic training experiments are needed to study

the mechanism of temporal specificity of strength

training.

In summary, the present results confirm previous

findings that temporal adaptation to strength train-

ing may exist, at least during repeated training in

the morning hours. However, the level of adaptation

to time-of-day-specific training varied widely among

individuals. Some individuals, by training repeat-

edly in the morning, might be able to improve

typically poor morning performance to the same or

even higher level as their normal daily peak typically

observed in the late afternoon. In contrast, other

individuals would still perform poorly in the

morning in spite of training in the morning hours.

It is suggested that a chronotype of the participants

may be partly responsible for the inter-individual

variation in responsiveness to time-of-day-specific

training.

References

Atkinson, G. (2002). Analysis of repeated measurements in

physical therapy research: Multiple comparisons amongst level

means and multi-factorial designs. Physical Therapy in Sport, 3,

191–203.

Bird, S. P., & Tarpenning, K. M. (2004). Influence of circadian

time structure on acute hormonal responses to a single bout of

heavy-resistance exercise in weight-trained men. Chronobiology

International, 21, 131–146.

Callard, D., Davenne, D., Gauthier, A., Lagarde, D., & Van

Hoecke J. (2000). Circadian rhythms in human muscular

efficiency: Continuous physical exercise versus continuous

rest. A crossover study. Chronobiology International, 17, 693–

704.

Castaingts, V., Martin, A., Van Hoecke, J., & Perot C. (2004).

Neuromuscular efficiency of the triceps surae in induced and

voluntary contractions: Morning and evening evaluation.

Chronobiology International, 21, 631–643.

Coldwells, A., Atkinson, G., & Reilly, T. (1994). Sources of

variation in back and leg dynamometry. Ergonomics, 37, 79–86.

Deschenes, M. R., Kraemer, W. J., Bush, J. A., Doughty, T. A.,

Kim, D., Mullen, K. M. et al. (1998). Biorhythmic influences

on functional capacity of human muscle and physiological

responses. Medicine and Science in Sports and Exercise, 30, 1399–

1407.

Drust, B., Waterhouse, J., Atkinson, G., Edwards, B., & Reilly, T.

(2005). Review: Circadian rhythms in sports performance – an

update. Chronobiology International, 22, 21–44.

Farina, D., Merletti, R., & Enoka, R. (2004). The extraction of

neural strategies from the surface EMG. Journal of Applied

Physiology, 96, 1486–1495.

Folkard, S., Monk, T. H., & Lobban, M. C. (1979). Towards

a predictive test of adjustment to shift work. Ergonomics, 22, 79–

91.

Gauthier, A., Davenne, D., Martin, A., Cometti, G., & Van

Hoecke, J. (1996). Diurnal rhythm of the muscular

performance of elbow flexor during isometric contractions.

Chronobiology International, 13, 135–146.

Giacomoni, M., Edwards, B., & Bambaeichi, E. (2005). Gender

differences in the circadian variations in muscle strength

assessed with and without superimposed electrical twitches.

Ergonomics, 48, 1473–1487.

Guette, M., Gondin, J., & Martin, A. (2005). Time-of-day effect

on the torque and neuromuscular properties of dominant and

non-dominant quadriceps femoris. Chronobiology International,

22, 541–558.

Hakkinen, K. (1994). Neuromuscular adaptation during strength

training, aging, detraining, and immobilization. Critical Reviews

in Physical and Rehabilitation Medicine, 6, 161–198.

Hermens, H. J., Freriks, B., Merletti, R., Hagg, G., Stegeman,

D. F., Blok, J. et al. (1999). SENIAM 8: European Recommenda-

tions for Surface ElectroMyoGraphy. Enschede, Netherlands:

Roessingh Research & Development.1399–1407.

Hill, D. W., Cureton, K. J., & Collins, M. A. (1989). Circadian

specificity in exercise training. Ergonomics, 32, 70–92.

Martin, A., Carpentier, A., Guissard, N., van Hoecke, J., &

Duchateau, J. (1999). Effect of time of day on force variation in

a human muscle. Muscle and Nerve, 22, 1380–1387.

McDonagh, M., & Davies, C. (1984). Adaptive response of

mammalian skeletal muscle to exercise with high loads: A

review. European Journal of Applied Physiology, 52, 139–155.

Milner-Brown, H. S., Mellenthin, M., & Miller, R. G. (1986).

Quantifying human muscle strength, endurance and fatigue.

Archives of Physical Medicine and Rehabilitation, 67, 530–535.

Moritani, T., & DeVries, H. A. (1979). Neural factors

versus hypertrophy in the time course of muscle strength gain.

American Journal of Physical Medicine, 58, 115–130.

Temporal specificity in strength training 1013

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4

Nicolas, A., Gauthier, A., Bessot, N., Moussay, S., & Davenne, D.

(2005). Time-of-day effects on myoelectric and mechanical

properties of muscle during maximal and prolonged isokinetic

exercise. Chronobiology International, 22, 997–1011.

Nindl, B. C., Hymer, W. C., Deaver, D. R., & Kraemer W. J.

(2001). Growth hormone pulsatility profile characteristics

following acute heavy resistance exercise. Journal of Applied

Physiology, 91, 163–172.

Prevost, M. C., Nelson, A. G. E., & Maraj, B. K. V. (1999). The

effect of two days of velocity specific isokinetic training on

torque production. Journal of Strength and Conditioning Research,

13, 35–39.

Reilly, T., Atkinson, G., & Waterhouse, J. (2000). Chronobiology

and physical performance. In W. E. Garrett, Jr. & D. Kirkendall

(Eds.), Exercise and sport science (pp. 351–372). Philadelphia,

PA: Lippincott Williams and Wilkins.

Sedliak, M., Finni, T., Cheng, S., Haikarainen, T., & Hakkinen,

K. (2008). Diurnal variation in maximal and submaximal

strength, power and neural activation of leg extensors in men:

Multiple sampling across two consecutive days. International

Journal of Sports Medicine, 29, 217–224.

Souissi, N., Gauthier, A., Sesboue, B., Larue, J., & Davenne, D.

(2002). Effects of regular training at the same time of day on

diurnal fluctuations in muscular performance. Journal of Sports

Sciences, 11, 929–937.

Tuckow, A. P., Rarick, K. R., Kraemer, W. J., Marx, J. O.,

Hymer, W. C., & Nindl, B. C. (2006). Nocturnal growth

hormone secretory dynamics are altered after resistance

exercise: Deconvolution analysis of 12-hour immunofunctional

and immunoreactive isoforms. American Journal of Physiology:

Regulatory, Integrative and Comparative Physiology, 291, R1749–

R1755.

Zambon, A. C., McDearmon, E. L., Salomonis, N., Vranizan,

K. M., Johansen, K. L., Adey, D. et al. (2003). Time- and

exercise-dependent gene regulation in human skeletal muscle.

Genome Biology, 4, 10:R61.

1014 M. Sedliak et al.

Dow

nloa

ded

by [

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and]

at 1

4:58

18

July

201

4