Secondary Logo

Journal Logo

Brief Review

Factors Contributing to Diurnal Variation in Athletic Performance and Methods to Reduce Within-Day Performance Variation: A Systematic Review

Kusumoto, Hirofumi1; Ta, Canhnghi1; Brown, Symone M.2; Mulcahey, Mary K.2

Author Information
Journal of Strength and Conditioning Research: December 2021 - Volume 35 - Issue - p S119-S135
doi: 10.1519/JSC.0000000000003758
  • Free

Abstract

Introduction

Recent work in exercise physiology has attempted to elucidate the effects that homeostatic changes throughout the day may have on athletic performance. It is well established that many athletes display a marked difference in performance depending on the specific time of day (TOD) at which exercise is performed (8,12–15,17,21,23,24,26,30,47,56,59,65). Most studies relating to the influence of circadian processes on athletic performance report greater athletic ability in the evening than in the morning (1–3,5,6,10,11,18,19,22,25,34,35,39,44–46,49,53,54,60–63,71). For example, it has been shown that long-term combined strength and endurance training in the evening can lead to a greater increase in muscle hypertrophy and mass than training in the morning (33). Such findings are consistently observed in tests of cardiovascular fitness, such as swimming and cycling, as well as tests of strength, such as standing vertical jump (countermovement jump [CMJ]) and isometric muscle contraction (40,69). This variation in athletic ability has a strong association with mechanisms of the circadian rhythm and carries important implications for scheduling training regimens in preparation for competitions (6,8,17,22,24,26,30,35,44,49,56,61).

The circadian rhythm governs rhythmic changes in human physiology and coordinates timing for processes such as the sleep-wake cycle, fluctuating activity levels, and synchronizing skeletal muscle functionality with photic light cues (66,70). In addition, the circadian rhythm is critical for managing major bodily processes including metabolism and energy balance through the regulation of hormones such as cortisol, norepinephrine, and melatonin (41,43). Dysregulation of the circadian rhythm has been associated with adverse metabolic and cardiovascular illnesses, such as diabetes, metabolic syndrome, and sudden cardiac death (58).

Numerous factors relating to the circadian rhythm have been shown to contribute to diurnal variation in athletic performance, including body temperature (35), chronotype (morning type vs. evening type) (31), training schedules (29), and diurnal fluctuations in biochemical markers (1). The questions guiding this review were “Why do some athletes show variable athletic performance output at different times of day?” and “What can athletes or coaches do to achieve more consistent performance?” Therefore, the first aim of this review was to identify statistically meaningful differences (p ≤ 0.05) in physical performance based on TOD and summarize the proposed mechanisms driving these diurnal variations in athletic performance. The second aim of this review was to summarize interventions that have been shown to eliminate previously statistically significant (p ≤ 0.05) TOD differences in the aforementioned measures of athletic performance. By recognizing contributors to diurnal variation and methods to reduce these differences, athletes and coaches may be able to optimize training regimens to achieve consistent competitive or recreational performance regardless of the TOD of the physical activity.

Methods

Experimental Approach to the Problem

This systematic review was conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (42). PubMed, Web of Science, Cochrane, EMBASE, and Ovid were searched using database-specific strategies. The PubMed search was created first and used as a model for the other search engines. A “diurnal variation” search concept was created by combining related terms such as “circadian rhythm” or “within-day,” with the “or” Boolean operator. An “athletic performance” concept was created in the same way by combining “sports” and related terms such as “athletics” and “exercise.” These concepts were queried together using the “and” operator to find articles related to diurnal variations in athletic performance (Figure 1). The search terms were deliberately broad to capture as many relevant articles as possible. The combined concept was then limited by the following filters, “Languages: English,” and “Species: Humans.” Articles relevant to either of the 2 aims of this review were collected from this search. The athletic performance measures accepted during the screening process were kept broad such that the findings of this review could be applied to a wider variety of physical activity.

F1
Figure 1.:
PubMed search strategy.

Subjects

Only studies involving subjects at least 18 years of age were included. The authors of the included studies obtained informed consent from subjects before the start of data collection. This review was conducted at Tulane University School of Medicine in New Orleans, Louisiana, USA. Because this review does not constitute human subjects research, it was exempt from review by the institutional review board. Subjects who underwent intentional sleep manipulation through the use of stimulants, depressants, light therapy, or jet lag (simulated or otherwise) were excluded. This exclusion criterion was implemented to limit potentially confounding effects of inadequate sleep quality and duration on athletic performance.

Procedures

An independent electronic search of the literature was performed by 2 reviewers using the customized search strategies for PubMed, Web of Science, Cochrane, EMBASE, and Ovid databases described previously. Articles with relevant titles were collected in a spreadsheet from primary search results and reference lists of review articles. Duplicates were removed. An initial screen of the articles was completed based on the study type and population. The inclusion and exclusion criteria were then applied to the abstracts of the remaining articles. A third round of screening was conducted on the full text of the remaining articles. A quality assessment of each study was conducted by the 2 reviewers at this stage.

Studies were included if they were peer-reviewed articles reporting a significant TOD difference (p ≤ 0.05) in any quantitative measure of athletic performance (e.g., muscle contractile force) and offered explanations for discrepancies between performances at different times of day. Studies that described a method to narrow or eliminate diurnal variation in athletic performance were also included. Publications were excluded if they were literature reviews, involved nonhuman subjects or subjects younger than 18 years, intentionally manipulated sleep duration or quality through the use of stimulants, depressants, light therapy, or jet lag, or were deemed to be of poor quality according to the NIH Quality Assessment tools (4).

Quality Assessment

Risk of bias of each included study was assessed independently by 2 authors using NIH Study Quality Assessment Tools (4). These standardized critical appraisal tools aided in evaluating potential flaws in study methods or implementation based on the criteria, including sample size, confounding factors, study design, strength of causality in the association between interventions and outcomes, withdrawal, and other factors. The appropriate assessment tool was selected based on study design (e.g., controlled intervention studies or observational cohort and cross-sectional studies). Criteria and conditions to assess the methodological quality of each study were marked in the tool as “yes,” “no,” or “not applicable.” Each reviewer assigned a rating to each study (“good,” “fair,” or “poor”), and then the reviewers deliberated on the ratings until a consensus was reached for each study. Studies rated as “poor” were removed from the analysis. Of the included studies, 30 were rated “good” and 19 were rated “fair.” Two studies were excluded from this review for having a rating of “poor.”

Data Extraction

The following data were extracted: authors, study design, year of publication, sample size, subject demographics, main contributing factor to diurnal variation, measurement test or tool, methods to reduce diurnal variation, and key findings. Data were tabulated with means, SDs, SEMs, ranges, p values, or intraclass correlation coefficients where applicable.

Results

The customized search strategies identified 3,618 citations across 5 different search engines (PubMed, Ovid, EMBASE, Web of Science, and Cochrane Libraries). Eight additional articles were identified by searching the references of the review articles found in the initial search. After removal of duplicates and screening by title, 207 studies remained. After the inclusion and exclusion criteria were applied to abstracts, 55 studies met the criteria for full-text extraction. After final analysis and study quality assessment, 49 publications were included for qualitative synthesis (Figure 2). Of the included publications, 35 (71.4%) primarily discussed factors that contribute to diurnal variation in athletic performance, whereas 14 studies (28.6%) discussed specific methods to reduce diurnal variation.

F2
Figure 2.:
Preferred Reporting Items for Systemic Reviews and Meta-Analysis (PRISMA) flowchart of study selection process.

Effect of Body Temperature on Exercise Capacity

Thirteen studies investigated body temperature as the primary contributor to diurnal variation in athletic performance (Table 1). Studies examined Wingate cycle ergometer performance (6,30,34,35,61), other cycle ergometer tests (3,54,55,57), maximal voluntary contractions (MVCs) of leg extensors (47,53), MVCs of elbow flexors (22), and treadmill sprints (49). Two studies included highly trained subjects (3,34), and 11 studies included recreational athletes (6,22,30,35,47,49,53–55,57,61). The most common outcomes measured were power output metrics including peak power, mean power, and force generated during muscle contraction. All studies demonstrated a diurnal variation in body temperature with higher values measured in the evening relative to the morning. Kin-Isler (30) reported diurnal variation in athletic performance that did not correlate with diurnal fluctuations in temperature. In addition, Reilly and Garrett (57) reported no diurnal difference in 60-minute cycle ergometer mean power tested at different times of day. All others showed a correlation between body temperature and athletic performance (3,6,22,34,47,49,53–55,61).

Table 1 - Summary of studies examining temperature as the key contributor to diurnal variation in athletic performance.*
Authors Journal (year) Population studied (age) Measurement test/tool Key findings
Atkinson et al. (3) J Sports Sci. (2005) 8 highly trained male cyclists
(24.9 ± 3.5 y)
Cycle ergometer time trial and power output
Chronotype
Intra-aural temperature
RPE
Serum metabolites
Cycling performance was worse in the morning even for athletes who tend toward higher “morningness”
Vigorous 25-min warm-up eliminated a statistically significant (p < 0.05) am-pm pretrial temperature difference. Warm-up resulted in modest improvement in am 16.1-km completion time.
am time trial performance was 35 s worse than pm performance (95% CI = 33–66 s, p < 0.01)
Bernard et al. (6) Eur J Appl Physiol Occup Physiol. (1998) 23 male physical education students
(23 ± 3 y)
Rectal temperature
Wingate test
Jump power
Rectal temperature was significantly higher as the day progressed (p < 0.05)
Mean cycling power and jump power were significantly higher at 18:00 compared with 9:00
 (14.8 ± 2.6 vs. 14.3 ± 2.4 W·kg−1, p < 0.01) and
 (56.6 ± 9.7 vs. 53.9 ± 8.8 W·kg−1, p < 0.05), respectively
Gauthier et al. (22) Eur J Appl Physiol. (2001) 8 physical education students
(21.2 ± 0.4 y)
Oral temperature
Dynamometer parameters
MVC of elbow flexors
There was a circadian rhythm in oral temperature (F = 18.30, p < 0.01)
Elbow flexion torque varied by time of day concomitantly with temperature
Temperature variability amplitude was low (0.9° C, SEM 0.06) and likely insufficient to explain performance variation
Kin-Isler (30) Isokinet Exerc Sci. (2006) 14 untrained undergraduate males
(22.6 ± 2.6 y)
(19–29 y)
Oral temperature
Wingate test
Oral temperature was lowest at 09:00 and highest at 17:00 (mean diff = −0.593° C, p < 0.01)
Ppeak and Pmean were lowest at 09:00 and highest at 13:00
 (mean diff = −0.459 W·kg−1, p < 0.05)
 (mean diff = −0.227 W·kg−1, p < 0.02), respectively
No significant difference in power output at 17:00 vs. 09:00
Diurnal variation in Wingate performance did not match the circadian rhythm of oral temperature
Lericollais et al. (34) Chronobiol Int. (2009) 16 male competitive cyclists
(23.5 ± 1.1 y)
Wingate test
Chronotype
Oral temperature
Mean power output is higher in the evening (568.1 ± 30.5 W vs. 530.0 ± 28.7 W, p < 0.01)
Oral temperature correlated with performance (36.3 ± 0.1 vs. 35.4 ± 0.2° C; p < 0.01)
Lericollais et al. (35) Scand J Med Sci Sports. (2011) 20 active males
(21.4 ± 1.4 y)
Gastrointestinal temperature
Wingate test
Core temperature was significantly higher at 18:00 vs. 06:00 (fitted cosine curve r = 0.97, p < 0.01; peak to trough amplitude 1.0 ± 0.16° C)
Ppeak 18:00 vs. 06:00
 (850.1 ± 120.0 vs. 799.7 ± 116.9, p < 0.01)
Pmean over 30 and 60 s at 18:00 vs. 06:00
 (612.8 ± 69.7 vs. 593.9 ± 70.9, p < 0.01) and
 (466.6 ± 49.3 vs. 454.2 ± 47.8, p < 0.01), respectively
Pereira et al. (47) Biol Sport. (2011) 30 healthy, untrained male subjects
(22 ± 1 y)
Leg extension MVC
Axillary temperature
EMG readings
Body temperature, contractile impulse (CI), rate of force development (RFD), and peak force of the rectus femoris and vastus lateralis muscles were significantly greater at night compared with morning (p < 0.05)
Pearson product-moment correlation coefficients of temperature vs. RFD and temperature vs. CI were correlated (all r ≥ 0.73 at p = 0.01)
No significant difference in mean frequency morning vs. afternoon vs. night
Pullinger et al. (49) Chronobiol Int. (2014) 20 active males with ≥2 years of experience playing soccer, field hockey, or rugby
(21 ± 2.2 y)
(19–27 y)
Treadmill repeated sprints
Rectal temperature
Heart rate
Blood draws
Breath-by-breath gas values
Muscle and rectal temperatures were higher in the pm vs. am at rest. Mean difference, respectively:
 0.57° C, CI = 0.34–0.80° C; p < 0.01
 0.46° C, CI = 0.39–0.54° C; p < 0.01
And after warm-ups
 0.38° C, CI = 0.14–0.62° C; p < 0.01
 0.42° C, CI = 0.32–0.52° C; p < 0.01
Peak velocity, average velocity, average power, and distance covered were higher in pm vs. am
Significant positive correlations between:
 Trec and Ppeak (r = 0.25); p < 0.05
 Trec and Pavg (r = 0.49); p < 0.05
 Trec and Vpeak (r = 0.26); p < 0.05
 Trec and Vavg (r = 0.57); p < 0.05
 Trec and distance (r = 0.56); p < 0.05
 Trec and RPE (r = 0.88); p < 0.05
Racinais et al. (53) Med Sci Sports Exerc. (2005) 11 male physical education students
(26 ± 4 y)
MVC of knee extensors
Chronotype
Ambient temperature and humidity
Rectal temperature
Skin temperature
Exposure to a moderately warm and humid environment significantly increased Tskin in am alone (F = 186.494, df = 10, p < 0.01)
Force/root-mean-square ratio during MVC testing showed a significant interaction effect between time of day and environmental condition (F = 7.865, df = 10, p < 0.02)
pm vs. am MVC values showed a difference of +12% in the control condition and −3% in the warm condition (F = 5.815, df = 10, p < 0.05)
Racinais et al. (55) J Sci Med Sport. (2009) 7 male physical education students
(27 ± 2 y)
Cycle ergometer power output
Chronotype
Rectal temperature
Leg skin temperature
Direct local heating of the leg in a water bath only increased muscle power in the morning.
Significant diurnal differences (p < 0.05) under control conditions in maximal power, maximal force, and maximal velocity were eliminated after local heating.
Direct local cooling of the leg decreased Pmax, Fmax, and Vmax in both am and pm, despite a diurnal variation in core temperature
Racinais et al. (54) Chronobiol Int (2005) 9 active, male physical education students (22 ± 4 y) RSA cycle parameters
Body temperature
pm rectal temperature was significantly higher than am temperature (F 1,8 = 10.20; p < 0.05)
Ppeak of the first sprint was higher in the evening than in the morning (958 ± 112 vs. 915 ± 133 W; p < 0.05)
Power decrement during the RSA test was higher in the evening than in the morning (F 1,8 = 10.57; p < 0.05)
Reilly and Garrett (57) J Sports Med Phys Fitness (1995) 7 male college-level soccer players (20.9 ± 1.6 y) (19–24 y) Rectal temperature
60-min cycle ergometer test
Pretrial mean rectal temperature was lower in the morning (37.2 ± 0.2 vs. 37.8 ± 0.6° C; p < 0.01)
Mean power output difference at 08:30 and 17:30 not significant (159.7 ± 10.3 vs. 156.9 ± 6.9 W; p > 0.05)
Significant difference in power output favoring evening testing during the first 20 min (p < 0.01)
Significant difference in power output favoring morning testing during the middle 20 and final 20 min (p < 0.05)
Souissi et al. (61) Int J Sports Med. (2004) 19 healthy male physical education students (21.8 ± 0.6 y) Cycle ergometer power output
Wingate test
Oral temperature
Oral temperature, Pmax, Ppeak, and Pmean fluctuate concomitantly
Toral acrophase at 18:22 ± 00:34 h
Pmax acrophase at 17:10 ± 00:52 h
Ppeak acrophase at 17:24 ± 00:36 h
Pmean acrophase at 18:00 ± 01:01 hours (mean ± SE)
*RPE = rating of perceived exertion; MVC = maximal voluntary contraction; CI = confidence interval; EMG = electromyography; RSA = repeated sprint ability.

Central vs. Peripheral Factors Influencing Diurnal Variation in Athletic Performance

Ten studies evaluated electromyographic (EMG) data in recreational athletes to determine TOD effects on central nervous system (CNS) inputs to muscle (i.e., central factors) or on intrinsic contractile properties of muscle (i.e., peripheral factors) to explain diurnal variations in athletic performance (Table 2). The outcomes assessed included root-mean-square (RMS) value, MVC, torque developed, neuromuscular efficiency, and mean power frequency. Most experimental designs involved MVC of leg extensors (24,44,46,60,71) or upper extremity muscles (10,21,39,45). One study used the Wingate test (12). In 1996, Gauthier et al. (21) cited both central and peripheral factors as influential in determining diurnal performance variation of the muscle output. All other publications, however, cited peripheral factors as more influential (10,12,24,39,44–46,60,71).

Table 2 - Studies examining EMG and muscle properties as the key contributors to diurnal variation in athletic performance.*
Authors Journal (year) Population studied (age) Measurement test/tool Key findings
Castaingts et al. (10) Chronobiol Int. (2004) 8 males
(18–30 y)
Intra-aural temperature
EMG parameters of the triceps surae
MVC
Drop test
Intertest variability <5% in 8/11 subjects
Temperature was significantly higher in the evening (p < 0.05)
Significant diurnal variation in drop jump height and MVC torque (10.9 ± 4.5%, p < 0.05 and 8.6 ± 2.6%, p < 0.05) but not EMG M wave in the triceps surae
Changes in TS function appear to be due to mainly peripheral changes
Temperature seems inadequate to explain the diurnal variation (amplitude 2.5 ± 0.5° C)
Chtourou et al. (12) Chronobiol Int. (2011) 22 active, male physical education students
(23.2 ± 1.9 y)
Oral temperature
EMG parameters
Wingate test
Significantly higher temperature at 17:00 vs. 05:00 (p < 0.01)
Significant time-of-day effect for:
 Power output: (F 1, 21 = 8.57, p < 0.01)
 MPF: (F 1, 21 = 33.98, p < 0.01)
 NME: (F 1, 21 = 4.34, p < 0.05)
Nonsignificant time-of-day effect for: RMS: (F 1, 21 = 1.05; p > 0.05)
Temperature amplitude was 0.6° C and seems inadequate to explain the diurnal variation
Peripheral factors are a more likely source of diurnal variation
Gauthier et al. (21) Chronobiol Int. (1996) 13 physical education students

7 males (22.0 ± 0.6 y)

6 females (23.1 ± 0.7 y)
Oral temperature
EMG parameters
MVC of elbow flexors
Slope of RMS/torque ratio was significantly higher at 09:00 vs. 18:00, meaning it took more effort to produce less torque in the morning (p < 0.05)
Temperature showed a diurnal rhythm with acrophase at 16:42 and batyphase at 4:42 (cosine curve r = 0.87)
Temperature seemingly sufficient to explain diurnal variation in performance (amplitude 0.4° C)
Central and peripheral inputs modulate MVC parameters
Guette et al. (24) Chronobiol Int. (2005) 10 male physical education students
(26.4 ± 1.4 y)
MVC torque of knee extensors
Intra-aural temperature
Significant 24-hour variation in temperature of athletes (p < 0.05), max voluntary contraction torque of knee extensors (p < 0.01)
EMG did not differ by TOD (maximal at all times)
Temperature may have an indirect effect on torque by increasing excitation-contraction coupling processes and intracellular inorganic phosphate levels.
Martin et al. (39) Muscle Nerve. (1999) 13 healthy subjects
12 males and 1 female (22–40 y)
Oral and skin temperature
MVC and tetanic force of the adductor pollicis
Central vs. peripheral mechanisms on muscle contractility
Oral and skin temperature significantly higher at 18:00 vs. 07:00 (p ≤ 0.03)
MVC force, but not RMS, significantly higher at 18:00 vs. 07:00 in the experimental group (96.7 ± 11.5 vs. 90.2 ± 12.4; p = 0.02)
M wave amplitude not significantly different at the 2 times of day
Direct muscle heating caused a significant decrease in RMS (610.7 ± 106.9 vs. 514.8 ± 79.8; p < 0.05) with no change in force.
Force variation is due to peripheral factors such as excitation-contraction coupling of muscle, with perhaps some contribution from temperature
Force variation is not due to central factors such as modification at the neural level
Nicolas et al. (44) Chronobiol Int. (2005) 12 male subjects
(24 ± 0.7 y)
Oral temperature
MVC of knee extensor muscles
EMG parameters
Maximal torque significantly higher in the evening than in the morning (169.5 ± 6.3 N·m−1 vs. 155.8 ± 5.3 N·m−1; p < 0.01) (F 1,11 = 14.39, p < 0.05)
NME for knee extensors significantly higher at 18:00 h vs. 06:00 h
 Vastus lateralis: (F 24,264 = 2.21, p < 0.01)
 Vastus medialis: (F 24,264 = 1.60; p < 0.05)
Diurnal variation in MVC likely due to peripheral changes such as muscle-twitch activation
Temperature may partly contribute to DVAP
Nicolas et al. (45) Isokinet Exerc Sci. (2007) 10 active male subjects
(28.6 ± 4.4 y)
MVC of elbow flexors
Torque
EMG
Strength recovery
Torque produced was greater at 18:00 vs. 06:00 for 4 different angular velocities of elbow flexion/extension
 240°: F 1,9 = 6.41; p < 0.04
 60°: F 1,9 = 7.01; p < 0.03
 0°: F 1,9 = 14.42; p < 0.01
 −60°: F 1,9 = 5.27; p < 0.05
Elbow flexor muscles more fatigable but faster to recover in the late afternoon compared with morning likely due to peripheral mechanisms
Extensor exercises not TOD dependent
Onambele-Pearson et al. (46) J Biomech. (2007) 12 healthy young men
(27 ± 2.0 y)
Maximal isometric knee-extension torque
EMG parameters
Patellar tendon mechanical properties
Knee extensor muscle properties
am to pm increase in
 Knee angle-torque: (16 ± 3.0%; p < 0.01; ICC = 0.97)
 Knee angle force: (16 ± 3.0%; p < 0.01)
No diurnal variation in knee extensor activation capacity (AC) as calculated by MVC/(MVC + interpolated twitch torque) (90.6% am vs. 87.9 pm; p > 0.05; ICC = 0.97)
Tendon stiffness decreased significantly from am to pm (21%; p < 0.01)
Sedliak et al. (60) Int J Sports Med. (2008) 32 untrained male volunteers
(32 ± 7 y)
MVC of knee extensors
EMG parameters
Barbell jump squat
pm squat jump power output significantly higher than am power output (p < 0.05)
pm MVC peak torque higher than am MVC peak torque (p < 0.01)
No diurnal variation in knee extensor RMS (p > 0.05)
Knee extensor RMS/torque ratio significantly lower at 20:30 vs. 07:00 (p < 0.05)
Peripheral changes are likely responsible for diurnal variation in peak torque
Zarrouk et al. (71) Int J Sports Med. (2012) 12 male subjects
(21.2 ± 1.8 y) (all >18 y)
Cycling repeated sprint test
Oral temperature
EMG activity
% peak power decrement higher in pm vs. am (t = 2.32, p < 0.05)
No significant correlation between changes in peak power and changes in EMG RMS activity
Peripheral changes are likely responsible for diurnal variation in sprint performance
*EMG = electromyography; MPF = mean power frequency; NME = neuromuscular efficiency; MVC = maximal voluntary contraction; TOD = time of day; ICC = intraclass correlation coefficient; RMS = root mean square; TS=triceps surae; DVAP=diurnal variation in athletic performance

Serum Biomarker Findings

Of the 5 studies evaluating the influence of serum biomarkers on diurnal variation in athletic performance, 4 investigated the effects of TOD-dependent hormonal changes on resistance exercise performance (1,2,32,59), and 1 investigated hormonal influences on cycle ergometer time trial results (19) (Table 3). Three studies included elite athletes alone (1,2,32), and 2 studies included recreational athletes alone (19,59). Overall, afternoon performance was superior to morning performance in all tested outcomes except for the CMJ and bench press task in a study by Kraemer et al. (32).

Table 3 - Summary of studies examining serum biomarkers as the key contributor to diurnal variation in athletic performance.*
Authors Journal (year) Population studied (age) Contributor to diurnal variation studied Measurement test/tool Key findings
Ammar et al. (1) Chronobiol Int. (2015) 9 male elite weightlifters
(21 ± 0.5 y)
Antioxidant enzymes Total volume of weight lifted for snatch, clean and jerk, and squat
Blood draws
RPE
Oral temperature
Greater performance, lower RPE values, and higher antioxidant levels in pm vs. am (p < 0.05)
Pretest catalase at 08:00, 14:00, and 18:00, respectively, were significantly different from each other: (F 2,16 = 88.4, p < 0.01, = 0.9)
Pretest glutathione peroxidase at 08:00, 14:00, and 18:00, respectively, were significantly different from each other: (F 2,16 = 92.4, p < 0.01, = 0.9)
Ammar et al. (2) J Sports Sci. (2015) 9 male elite weightlifters
(21 ± 0.5 y)
Muscle damage markers Total volume of weight lifted for snatch, clean and jerk, and squat
Blood draws
RPE
Oral temperature
Greater performance and lower RPE values in pm vs. am (p < 0.05)
pm training had a lowest muscle damage rate increase as quantified by *creatine kinase, lactate dehydrogenase, aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl, and alkaline phosphate (respective rates below)
Morning: 59.91 ± 7.71, 25.94 ± 9.24, 38.05 ± 5.53%, 22.91 ± 3.26%, 32.28 ± 3.28%, and 42.01 ± 2.38%
Afternoon: 178.43 ± 125.60%, 19.46 ± 8.52%, 29.95 ± 5.10%, 28.41 ± 8.23%, 28.07 ± 7.57%, and 32.24 ± 2.96%
Evening: 29.89 ± 3.08%, 26.25 ± 3.7%, 23.22 ± 3.77%, 26.98 ± 2.61%, and 36.99 ± 3.31%
*Rate increase not reported for evening.
Fernandes et al. (19) PLoS One. (2014) 9 male recreational cyclists
(31 ± 7.3 y)
(18–43 y)
Cortisol, insulin, growth hormone, norepinephrine, and glucose Cycle ergometer time trial
Blood draws
Physiologic response
TT was faster in the evening than in the morning t(8) = 2.33, p < 0.05; = 0.63, 95% CI = 0.7–12.8 s
Insulin, cortisol, and testosterone were lower in the evening (p < 0.05)
Growth hormone and plasma glucose were higher in the evening (p < 0.05)
Kraemer et al. (32) J Strength Cond Res. (2014) 10 male elite collegiate track and field athletes
(20.4 ± 1.6 y)
Melatonin Countermovement jump
Bench press
Quick feet reaction test
Blood draws
am vs. pm power performance not different (5,407.1 ± 1,272.9 vs. 5,384.6 ± 888.3W, p > 0.05)
pm quickness and accuracy were better
 Test time: 5.14 ± 1.06 vs. 4.39 ± 0.76 (p < 0.05)
 Mistakes: 2.68 ± 1.61 vs. 2.10 ± 1.08 (p < 0.05)
Melatonin levels higher in the morning (p < 0.01)
Sedliak et al. (59) Chronobiol Int. (2007) 38 healthy males matched by the training level
(20–24 y)
Testosterone and cortisol 1 repetition maximum MVC of half-squats, loaded squat jumps, and knee extensions
Chronotype
RPE
Blood draws
Testosterone (T) and cortisol (CORT) showed a significant TOD variation
 T: F = 132.85; p < 0.01
 CORT: F = 619.84; p < 0.01
10 weeks of TOD specific training in the am resulted in lower baseline am CORT (319 vs. 295 nmol·L−1)
It also blunted the diurnal variation in MVC: significant interaction between pre-post training, TOD, and peak torque difference from 07:00 and 12:00 h F = 5.14; p < 0.05
*RPE = rating of perceived exertion; TT = time trial; MVC = maximal voluntary contraction; TOD = time of day.

Chronotype and Habitual Training Time

Three studies examined the influence of chronotype on performance at various TOD in highly trained athletes (8,17,56) (Table 4). One study assessed the effects of a 5-week TOD specific training program on performance in the morning and afternoon in 12 recreational athletes (26). Outcomes assessed included rowing time trials, 200-m swimming time trials, cycle ergometer time to exhaustion, and shuttle run times. All studies reported improved performance when testing TOD matched athlete habitual training time or chronotype.

Table 4 - Summary of studies examining chronotype or training adaptation as the key contributor to diurnal variation in athletic performance.*
Authors Journal (year) Population studied (age) Measurement test/tool Key findings
Brown et al. (8) J Strength Cond Res. (2008) 8 male and 8 female university club crew rowers
(19.6 ± 1.5 y) (18–23 y)
Rowing ergometer times
Broad jump
Basic Language Morningness (BALM) Scale
Morningness-Eveningness Questionnaire (MEQ)
Morning types rowed faster in the morning (1.1%, p < 0.01)
Evening/neither types showed no significant difference (p = 0.104, p = 0.907, respectively)
Each individual's am-pm performance change correlated well with BALM (r = 0.88, p < 0.01) and MEQ scores (r = 0.86, p < 0.01)
Facer-Childs and Brandstaetter (17) Curr Biol. (2015) 20 competition-level field hockey players
(22.5 y)
(18–25 y)
BLEEP endurance test
Maximum oxygen uptake
ECT = early chronotype
ICT = intermediate chronotype
LCT = late chronotype
Significant difference in peak performance time of day by phenotype (peak performance for ECTs at 12.19 ± 1.43 h; ICTs at 15.81 ± 0.51 h; and LCTs at 19.66 ± 0.67 h)
ECT and ICT performance not different when analyzed by time since awakening (ECTs at 5.60 ± 1.44 h and ICTs at 6.54 ± 0.74 h; Kruskal-Wallis; p > 0.05)
LCT peak performance remained significantly different from ECT and ICT peak times (11.18 ± 0.93 h; Kruskal-Wallis, p < 0.01)
Hill et al. (26) Med Sci Sports Exerc. (1998) 12 recreationally active college-aged females
(20 ± 1 y)
Cycle ergometer time to exhaustion
Oral temperature
Breath-by-breath metabolic response
Heart rate
Subjects in the pm-trained group had 13% greater time to exhaustion in the pm than am (p = 0.03)
The pm group had greater O2 deficit in pm vs. am (36% greater; p = 0.01)
The am-trained group did not show greater time to exhaustion in pm than am (−13%; NS)
No time-of-day effect on oral temperature
Rae et al. (56) Eur J Appl Physiol. (2015) 18 male and 8 female highly trained swimmers
(25–50 y)
200-m swimming time trials (TT)
50-m split times
Horne-Ostberg Chronotype Inventory
Oral temperature
RPE
Morning types swam faster in the morning, whereas neither types were faster in the evening (p < 0.04)
TT and RPE results were improved when testing time matched training time (p = 0.011 TT results by chronotype; Train am group: 06h30–18h30 RPE difference = −1.0 ± 1.4; Train pm group: 06h30–18h30 RPE difference = 0.8 ± 1.9, p < 0.02)
Sleep and body temperature were not correlated with performance
*NS=not significant; RPE = rating of perceived exertion.

Oxygenation Kinetics

The role of oxygenation kinetics in observed diurnal variations in athletic performance was reported in 3 studies (Table 5). Faull et al. (18) included competition-level rowers, whereas the other authors assessed performance in recreational athletes (25,62). Outcomes assessed were power outputs of rowing ergometers (18) and cycle ergometers (25,62). In addition, physiologic parameters such as heart rate, blood pressure, ventilation, and oxygenation were measured. All studies reported improved ventilation or oxygenation in the afternoon.

Table 5 - Summary of studies examining oxygenation kinetics as the key factor of diurnal variation in athletic performance.
Authors Journal (year) Population studied (age) Contributor to diurnal variation studied Measurement test/tool Key findings
Faull et al. (18) Scand J Med Sci Sports. (2015) 13 healthy male highly trained rowers
(21 ± 3 y)
(18–30 y)
Cerebral perfusion and oxygenation Rowing ergometer power output
Hemodynamic/gas measurements
Muscle oxygenation
Hemoglobin (Hb) volume
Afternoon rowing time trials were faster than in the morning (n = 10; p = 0.03; 95% CI ± 2.5 s)
Mean power output was greater in the afternoon (337 vs 347 W; p = 0.04)
Oxygenated Hb volume decreased less in am trials (137 ± 66 vs. 238 ± 96 μM·cm, p < 0.01)
Hill (25) Appl Physiol Nutr Metab. (2014) 20 untrained healthy males
(22 ± 3 y)
(20–26 y)
Exercise efficiency and muscle oxygen Cycle ergometer
Chronotype
Oral temperature
Time to exhaustion (Texhaustion)
Maximal oxygen consumption (V̇o 2max)
Temperature, SBP, and HR were higher in the pm vs. am (p < 0.05)
Texhaustion (+20%, p < 0.01 ES = 1.72), V̇o 2max (4% p < 0.05; ES = 0.31), V̇o 2 kinetics (p ≤ 0.05), and anaerobic capacity were higher in the afternoon despite lower exercise efficiency (oxygen consumption/work rate)
Souissi et al. (62) Chronobiol Int. (2007) 11 untrained healthy male physical education students
(21.8 ± 2.4 y)
Aerobic contribution Oxygen uptake
Blood lactate
Oral temperature
Wingate test parameters
Peak power, mean power, power, total work output, and V̇o 2 were greater in the afternoon than in the morning (p < 0.01)
Power decrease greater in the morning than in the afternoon (p < 0.01)
Absolute V̇o 2 and aerobic contribution were higher in the afternoon than in the morning
SBP = systolic blood pressure; HR = heart rate; ES = effect size.

Methods to Reduce Diurnal Variation

Fourteen studies described methods to reduce performance variation at different points in the day (Table 6). The study populations included highly trained athletes (5,64), recreational athletes (11,13–15,23,50–52,63,65,67), and a mix of both (20). Nine studies used various forms of warm-ups to decrease within-day differences in athletic performance (5,11,14,20,50–52,63,67). Outcomes for the warm-up intervention studies included shuttle run times, cycle ergometer power outputs, and CMJ performances. Belkhir et al. (5) and Chtourou et al. (14) found that listening to music during the warm-up augmented athletic performance in the morning and afternoon, but preferentially in the morning.

Table 6 - Methods to reduce diurnal variation in athletic performance.*
Authors Journal (year) Population studied (age) Measurement test/tool Key findings
Belkhir et al. (5) Physiol & Behav. (2019) 12 male semiprofessional soccer players
(21.82 ± 2.47 y)
(all >18 y)
Warm-up with music
5-m shuttle run
(Music warm-up)
pm total distance significantly higher than am distance TOD effect: (F = 66.17, p < 0.001, = 0.85)
Warm-up with neutral or motivational music resulted in significantly increased total distance run (p < 0.01)
am total distance and higher distance gains with music were greater than pm gains
am: (Δ-change: 3.51 and 6.97%, respectively)
pm: (Δ-change 3.56 and 5.26%, respectively)
Chaari et al. (11) Biol Rhythm Res. (2015) 11 untrained male physical education students
(22.6 ± 2.5 y)
(20–22 y)
Wingate test
Chronotype
Oral temperature
(Warm-up)
Significant diurnal variation in Ppeak (p < 0.01) and Pmean (p < 0.05)
The 5-min warm-up group performed better than the 5-min warm-up group on the whole for both recovery and no recovery conditions
15-min WU in the am approached the 5-min warm-up Pmean and Ppeak for the recovery group
15-min warm-up is better when combined with 5-min rest pre-exercise in the afternoon
Chtourou et al. (13) J Strength Cond Res (2012) 31 recreationally active male physical education students
(23.1 ± 1.98 y)
(all >18 y)
Oral temperature
Wingate
Jump tests
(Training adaptation)
Statistically significant (p < 0.01) diurnal variations in Wingate Pmean and Ppeak, squat jump, and countermovement jump were eliminated in the MTG after a 12-wk training program (ICC >0.81, 0.86, 0.86, 0.96, respectively)
Chtourou et al. (14) Int J Sports Med. (2012) 12 male physical education students
(22.4 ± 1.7 y)
Wingate test
RPE
(Music warm-up)
Ppeak and Pmean increased following the music warm-up in comparison with the no music group (p < 0.01 in the morning and p < 0.05 in the afternoon)
Diurnal variation in Pmean was significant in the no music group (p < 0.01) but nonsignificant in the music group
Chtourou et al. (15) J Strength Cond Res. (2012) 30 recreationally active male physical education students
(22.9 ± 1.3 y)
Squat jump
Countermovement jump
1 repetition maximum half-squat
Wingate test (training adaptation)
Previously significant diurnal variation in leg extension (amplitude 3.62%; ICC >0.86), leg curl (amplitude 4.17%; ICC >0.86), and squat (amplitude 3.45 ± 3.7%; ICC ≥0.86) (all p < 0.05) were eliminated after training adaptation in the morning-trained group
Significant diurnal variation remained for these tasks in evening-trained and control groups (all p ≤ 0.05)
Frikha et al. (20) Biol Rhythm Res. (2015) 12 trained and 10 untrained male physical education students
(21.51 ± 1.25 y)
Wingate test
RPE
Oral temperature
(Warm-up)
Longer warm-up lead to higher Ppeak in both groups in the am sessions 15-min warm-up vs. 5-min warm-up (p < 0.01) and 20-min warm-up vs. 5-min warm-up (p < 0.01)
Diurnal variation in Ppeak and Pmean was significantly higher in the nontrained group compared with the trained group
Gueldich et al. (23) Int J Sports Med. (2017) 20 recreationally active male physical education students
(22.06 ± 1.98 y)
Oral temperature
EMG readings
MVC of knee extensors
5-wk temporal adaptation ES program (training adaptation)
MTG = morning-trained group
ETG = evening-trained group
MPF = mean power frequency
Oral temperature increased significantly from 07:00 to 17:00 (F = 434.56; p < 0.01)
Regarding pre-training vs. post-training MVC, RMS, and MPF:
MTG and ETG showed pre-post improvement (p < 0.01)
Diurnal variation in all parameters eliminated in MTG
Diurnal variation in all parameters significant in ETG (p < 0.05)
Changes to neural drive, MU recruitment, or muscle membrane potentials may explain diurnal variation in performance
Racinais et al. (53) Med Sci Sports Exerc (2005) 8 moderately active male physical education students
(27 ± 8 y)
Maximal force during cycling sprint
Muscular power
Rectal temperature
(Warm-up)
Rectal temperature, maximal force, and muscular power were higher in the afternoon than in the morning (p < 0.05)
Active warm-up (AWU) significantly increases Pmax regardless of the time of testing and independent of initial core temperature (p < 0.05)
AWU in the morning resulted in similar muscular power as afternoon control condition
Racinais et al. (50) Chronobiol Int (2004) 12 male physical education students
(27 ± 4 y)
Skin and rectal temperature
Squat jump, countermovement jump, and cycle ergometer power output
(Warm environment)
Moderately humid and warm condition blunted diurnal variation in muscular performance
Power in neutral, not warm, environment during CMJ was significantly higher (p < 0.05) in the afternoon than in the morning
Warm environment enhanced both CMJ and SJ power in the morning (p < 0.05) but not in the afternoon
Racinais et al. (51) Int J Sports Med. (2004) 23 recreationally active physical education students (15 males and 8 females)
(22.8 ± 3 y)
Rectal temperature
Vertical jump test
Maximal cycling power
(Warm and humid environment)
Rectal temperature significantly higher in the afternoon than in the morning (p < 0.01)
No variation in anaerobic power between 08:00, 13:00, and 17:00 in a hot and humid environment
Souissi et al. (64) Biol Rhythm Res. (2013) 12 elite judokas
(21.1 ± 1.2 y)
Wingate test power output
Reaction time test
(Caffeine intake)
Main effect for caffeine (F = 19.4, p < 0.05)
Caffeine × TOD interaction: (F = 18.12, p < 0.05)
Ppeak and Pmean increased in response to caffeine ingestion in the morning alone (p < 0.01)
Souissi et al. (63) Chronobiol Int (2010) 12 untrained healthy male physical education students
(23.5 ± 3.1 y)
Wingate test parameters
Rectal temperature
(Warm-up)
Ppeak and Pmean were significantly higher in the afternoon than in the morning for 5- and 15-min AWU (p < 0.01)
5-min AWU resulted in significantly higher Ppeak and Pmean at 08:00 compared with 5-min AWU (p < 0.05)
No significant difference in Ppeak and Pmean between 5- and 15-min AWU at 18:00
Souissi et al. (65) Pharmacol Biochem Behav (2019) 15 healthy male physical education students
(20 ± 1 y)
(all >18 y)
5-m shuttle run
(Caffeine intake)
Cognitive and physical performances varied by time of day
Reaction time was best in the late am and afternoon (F = 48.51; p < 0.05; = 0.78)
Shuttle run total distance was best in the late afternoon (F = 87.82; p < 0.05; = 0.86)
Caffeine improved RT at all times of day but most at worst performance times—early am and early pm (p < 0.05)
Caffeine improved total distance mostly in early am (p < 0.05)
Taylor et al. (67) Int J Sports Med. (2011) 8 recreationally trained males
(29.8 ± 5.2 y)
Countermovement jump
Chronotype
Body temperature
(Warm-up)
4–6% difference in jump performance measures between am and pm sessions (ES range = 0.2–0.4; coefficients of variation all <6%)
Warming-up am temperature to that of pm temperature reduced TOD differences in explosive muscular performance (mean difference <1%, ES range = 0.0–0.1).
*RPE = rating of perceived exertion; EMG = electromyography; ES = electrostimulation; TOD = time of day; CMJ = countermovement jump; SJ = squat jump; ICC = intraclass correlation coefficient; MVC = maximal voluntary contraction; RMS = root mean square; MU=motor unit; RT=reaction time

To reduce diurnal variation in athletic performance, 3 studies examined temporal adaptation training (13,15,23), and Gueldich et al. (23) used temporal electrostimulation (ES) training. Outcomes included Wingate, squat, jump, and lower extremity MVC tests. All studies successfully eliminated diurnal variation in athletic performance in morning-trained groups.

Finally, 2 studies found that caffeine intake improved Wingate test metrics, reaction time tests, and shuttle run performance in the morning (64,65).

Discussion

This systematic review summarizes factors that contribute to or reduce diurnal variation in athletic performance. Although there is debate about the most important contributory factors, the current body of literature provides information that may be leveraged to optimize physical performance at various levels of athletic competition. In addition, to the best of our knowledge, this is the first review to examine various methods to reduce diurnal variation in athletic performance in order to achieve more consistent physical output throughout the day.

Body temperature was the most commonly analyzed and well-established contributory factor to diurnal variation in athletic performance (Table 1). Since the 1990s, many research groups have observed strong diurnal patterns in the athlete's core body temperature that demonstrate troughs in the early morning and peaks in the late afternoon to early evening (6,10,12,21–25,30,34,35,39,47,49,51,54,57,61). Temperature troughs often correlated with relatively poor performance, whereas peaks coincided with relatively high performance in numerous tasks and metrics (e.g., peak power, mean power, peak velocity, average velocity, and ratings of perceived exertion (RPEs) for various physical tasks) (2,6,10,24,34,35,39,44,45,47,49,53–55,61,71).

Although body temperature and athletic performance are often positively correlated, there is discord regarding the degree to which temperature affects athletic performance. For example, Atkinson et al. (3) suggest that increasing body temperature before exercise can modestly improve performance, but temperature alone is not enough to completely account for variable exercise capacity throughout the day. This conclusion, which has been reported by several others (10,12,35), was supported by 2 separate studies led by Gauthier (21,22), showing that temperature and elbow flexor function were higher in the evening. However, because the temperature range was small (less than 1° C), it was concluded that temperature alone was likely insufficient to explain the difference in muscle output (21,22). Furthermore, other studies have observed significant differences in body temperature that did not correlate with diurnal variation in athletic performance measures (30,57).

The mechanism underlying the relationship between temperature and athletic performance remains unclear. Because temperature has been shown to alter nerve conduction properties and muscle contractility (36,37), many research groups have studied muscle EMG activity during exercise. There are 2 primary mechanisms that appear to influence muscular output, central mechanisms and peripheral mechanisms. Central mechanisms such as neural input from the CNS are commonly assessed with the RMS value, which quantifies EMG signal power as a function of time (i.e., level of muscular activation). Peripheral mechanisms such as muscle contractile properties are commonly measured by force output of muscle contractions. There is currently no consensus regarding which mechanism has the greater contribution to athletic performance. Several groups have reported significant increases (p < 0.05) in both RMS and MVC in the evening relative to the morning (23,24), suggesting that although muscle output was higher in the evening, it took greater neural activation to achieve increased contractile force. These findings implicate both central and peripheral mechanisms in diurnal performance variation.

Gueldich et al. (23) demonstrated that a 5-week ES program led to increased RMS and MVC power output at the same TOD as routine stimulation (p < 0.01). Although routine morning training eliminated previously significant diurnal variation in MVC of leg extensors, the subjects demonstrated increased RMS and MVC after ES training. These results suggest that TOD variation in muscle contraction may be caused by a compensatory increase in central input or by changes in motor unit properties and muscle membrane potentials.

On the contrary, other researchers who observed a diurnal variation in MVC did not observe a concomitant pattern in central mechanisms. For example, Pereira et al. (47) reported that differences in explosive force parameters in short time intervals correlated significantly with diurnal body temperature variations (r ≥ 0.73; p = 0.01) and not mean frequency variations. This suggests that changes in the muscle recruitment level—a central mechanism—were not different at various times of day (47). This finding, which was recapitulated by several others (12,24,39,44–46,60), suggests that peripheral mechanisms were more likely the driver of variable athletic performance.

In support of the peripheral mechanism hypothesis, previous studies have directly explored the modulatory effect of temperature on muscle function. Racinais et al. (53) found that exposure to a moderately warm, humid environment increased both morning body temperature and morning knee extensor contractile force in 11 male physical education (PE) students (mean age = 26 ± 4 years). In a later study, Racinais et al. (55) tested the effects of direct local muscle heating on athletic performance. The authors found that directly heating PE students' legs (n = 7 men; mean age = 27 ± 2 years) for 30 minutes before exercise eliminated differences (morning vs. afternoon) in skin temperature, relative power, maximal force, and maximal velocity observed under control conditions (55). Collectively, these data suggest that better evening performance is a function of improved intrinsic muscle properties (rather than central neural activity) that may be mediated by increases in temperature. This explanation is supported by the aforementioned studies reporting EMG patterns that show diurnal variation in athletic performance measures without a concomitant increase in muscle neural input activation levels.

Another contributor to diurnal variation in athletic performance is the rhythmicity or exercise-related fluctuation of various serum biomarkers (Table 2). These biomarkers include antioxidant enzymes, stress-related markers, and mood and sleep-mediating hormones. During resistance training exercises, high levels of antioxidant enzymes and low rates of increase of muscle damage markers were reported to coincide with superior weightlifting ability (1,2). During endurance exercises, hormonal control of glucose metabolism appeared to be the main driver of performance. Fernandes et al. (19) found that recreational cyclists (n = 9 men; mean age = 31 ± 7.3 years) demonstrated a significant improvement in completion time (p < 0.05) in a 1,000-m cycling time trial during evening tests. Insulin, cortisol, and total free testosterone were higher, and plasma glucose levels were lower in the morning than the evening (p < 0.05). The authors concluded that the observed morning hormonal balance creates a low glycemic metabolic “milieu” not conducive to peak athletic performance (19).

Sedliak et al. (59), however, reported that healthy male subjects (n = 38; age range: 20–24 years) displayed diurnal variation in isometric strength of knee extensors that was attenuated after a 10-week morning training regimen. Morning cortisol levels were reduced in this experimental group, whereas a concomitant shift was not observed in testosterone levels. These findings suggest that cortisol, not testosterone, is the key modulatory hormone of athletic performance, and that decreased morning cortisol levels are associated with improved performance. Although these findings somewhat contradict the metabolic milieu hypothesis posed by Fernandes et al. (19), Sedliak et al. (59) did not measure insulin and growth hormone, so the results are difficult to compare.

Accurate interpretation of the impact of serum biomarkers is difficult given the number of biomarkers that can be analyzed and the counteracting effects and individual circadian rhythms of the selected hormones. The studies in the current literature evaluate a variety of factors including hormones and nutrients (19,32,59), cell counts, enzymes, and inflammatory markers (1,2,59), but there is considerable disagreement about the type of markers that contribute most to diurnal variation in athletic performance. Moreover, some hematologic markers exhibit complex diurnal rhythms that vary with the athlete's training level and chronotype. Finally, body temperature fluctuations may mask serum biomarker effects, and unfortunately, studies that examine serum biomarkers as an explanatory factor for diurnal variation in athletic performance do not consistently evaluate the potential influence of body temperature on these biomarkers throughout the testing protocol (19,32,59).

Several groups have postulated that athletes perform better when testing sessions align with their circadian phenotype (i.e., early bird, night owl, or intermediate) (Table 3). Some data suggest that early birds, but not intermediate-type or night owls, are affected by testing outside of their preferred TOD (8), whereas others report that habitual time of training is the key determining factor in TOD performance differences (2,68). Hill et al. (26) explored this hypothesis by training a cohort of 12 women (mean age = 20 ± 1 years) in high-intensity cycling. Six subjects were assigned to train in the morning and 6 in the afternoon. After 5 weeks, they tested each group's morning and afternoon cycling performance. They found that time to exhaustion was longer at the habitual training TOD (F1,10 = 8.29, p = 0.02), but only the afternoon-trained group had significantly different morning vs. afternoon time to exhaustion (373 ± 222 vs. 422 ± 252 seconds; p = 0.03) (26).

More recent work implicates both athlete chronotype and training schedule as key determinants of diurnal variation. Rae et al. (56) found that in a morning-evening 200-m time trial of highly trained swimmers (n = 26; 18M/8F; age range: 25–50 years), grouping performance by athlete chronotype resulted in significant diurnal variation (p < 0.04). Furthermore, faster times and lower RPE scores were reported when training and testing sessions were in circadian alignment. Taken together, these data suggest that athletes perceive submaximal effort exercise to be easier when testing aligns with their biological rhythm or training time (56).

Alternatively, Facer-Childs and Brandstaetter (17) reported significant diurnal variation in a cardiovascular endurance test of competitive field hockey players (n = 20; mean age = 20.4 years) who differed by early chronotype (ECT), intermediate chronotype (ICT), and late chronotype (LCT). When performance was analyzed as a function of time since awakening, there was no variation in peak performance time between ECTs and ICTs. Their findings suggest that at least for ECT and ICT subjects, peak athletic performance was most influenced by the time gap between rising and testing. Specifically, the gap of time between awakening and peak performance was significantly longer for LCTs than ECTs and ICTs (p < 0.01). The authors proposed that decreased morning cortisol levels in LCTs limited their athletic capabilities in the morning and resulted in delayed peak performance time (17). However, although these findings suggest that a certain amount of time must pass before the optimal temperature/hormonal conditions are achieved for peak athletic performance, considering cortisol as a driver of improved athletic performance contradicts the conclusions of previously mentioned studies (19,59).

Several studies have implicated physiologic responses to exercise based on TOD as a factor that contributes to performance variation (Table 4). In several studies, performance in ergometer tests (e.g., time to exhaustion and power output) was better in the afternoon (18,25,62). Athletes had higher maximal oxygen consumption (V̇o2max), faster peripheral oxygenation kinetics, greater anaerobic capacity (as measured by maximal accumulated oxygen deficit), and higher body temperature at the time of higher performance (p < 0.05 for all measures) (25). Although it was not addressed in this study how temperature might change oxygen kinetics or athletic performance (25), it is possible that increased muscle temperature facilitates oxygen extraction from blood in times of high metabolic demand (18).

Because body temperature was found to be among the most important factors influencing athletic performance, a prolonged warm-up regimen was reported to be the most effective method to reduce diurnal variation likely as a result of increased morning pre-exercise body temperature (3,53,55). Taylor et al. (67) assessed explosive lower-body power output in 8 male athletes (mean age = 29.8 ± 5.2 years) using the CMJ test. They found that raising morning body temperatures to approximate afternoon temperatures eliminated diurnal variations in jump height (mean difference <1%; effect size = 0.0–0.1) (67). This finding is consistent with other studies showing that extending the length of warm-up regimens in the morning can reduce diurnal variation in Wingate performance (11,20). In addition, a warm environment is sufficient to increase morning power in jump tests to approximate outputs seen in the afternoon (50) and has been shown to result in no diurnal variation in vertical jump and cycling tests (51).

Interestingly, music can also reduce diurnal variation. Belkhir et al. (5) demonstrated that warming-up with music results in increases in short sprint exercise metrics at all times of day (p < 0.01) with preferential improvement in relatively poor morning performance. These results agree with those of Chtourou et al. (14), showing that within their sample of PE students (n = 12 men; mean age = 22.4 ± 1.7 years), Wingate mean power output was no longer dependent on TOD after implementing a 10-minute warm-up with music. The authors speculated that music may have resulted in a higher warm-up intensity. Although it was not addressed in this study, improved performance may have been in part due to differences in body temperature after warming-up with and without music.

The ergogenic effects of caffeine at various times of day have been assessed in several studies. For example, in a Wingate test of elite judokas (n = 12; mean age = 21 ± 1.2 years), peak and mean power were found to be significantly greater (p < 0.05) after caffeine ingestion in morning but not evening trials (64). Recently, it has been found that caffeine produces significant ergogenic effects on short sprint tests and reaction time tests, especially at times of day where performance in these domains is weakest (p < 0.05) (65).

Training adaptation was also evaluated as a method to reduce diurnal variation in athletic performance. Athletes have been shown to perform various resistance exercises best at the same TOD as habitual training over an 8-week period (15). Time of day–specific improvements apply to ES training as well (23). Moreover, morning-specific training can eliminate diurnal differences in various leg resistance exercises (15) as well as Wingate and CMJ measures (13). Further support of the equalizing effects of temporal adaptation is observed by Chtourou et al. (16), which found that in a cohort of elite male judokas, mean repeat sprint and CMJ results did not differ between the morning and afternoon (16). The authors provided several potential explanations for the negative result; the athletes habitually trained during morning hours and completed a relatively long 15-minute warm-up before testing (16).

There is currently no consensus on how these factors interact to influence diurnal variation in athletic performance. Future work examining performance at a standardized core body temperature and time after awakening may be useful in stratifying the relative importance of the various contributing factors to diurnal performance variation. In addition, it would be interesting to see how body temperature rhythms relate to RPE at different TOD because most studies examining chronotype training effects make no mention of body temperature.

Despite our efforts, there are some limitations to this study that should be addressed. First, this review is susceptible to publication bias because the search criteria only sought out studies reporting diurnal variation in athletic performance. Although there is insight to be gained from studies that do not show differences in performance measures, the “significant variation” limitation was incorporated to maintain a manageable scope of investigation. To mitigate negative result bias, studies that described interventions that reduced or eliminated previously significant performance variation within a given study population were included. Second, the study populations cited in this review were small, had varying baseline training levels, and included primarily male subjects in their 20s. In addition, direct comparison of all the included studies was often not possible because there were several different types of tasks performed and different levels of control stringency—not all investigators controlled for subject sleep schedules, diet, and chronotype matching within experimental groups (e.g., recruiting the same number of morning types and evening types). Furthermore, the included studies recruited subjects of various training levels. Although this fact may make the analysis of this review more applicable to the general population, it makes direct comparisons across studies difficult.

Finally, temperature fluctuations associated with the menstrual cycle may have confounded the results of the referenced studies that included female subjects. The transition from the follicular to luteal phase of the menstrual cycle is associated with a 0.1–0.5° C elevation in basal temperature in many women (27,38). However, there is a persistent diurnal pattern of relatively low core temperatures in the early morning and elevated temperatures in the afternoon regardless of menstrual phase status (9). Unfortunately, the literature regarding athletic performance across different phases of the menstrual cycle is conflicting, and many physiologic changes occur throughout the cycle that can confound a temperature effect on exercise in women (7,27,28,48). Of the 7 studies that enrolled female subjects in this review, only one controlled for menstrual phase status (21). Of note, almost no women were enrolled in the studies examining temperature as the primary contributor to diurnal variation in athletic performance or the studies examining methods to reduce diurnal sports performance discrepancies. This fact highlights the need for further investigation into am-pm performance variation across multiple phases of the menstrual cycle within the same subject cohort.

Practical Applications

Understanding the mechanisms behind fluctuations in athletic performance could have great implications for optimizing the preparation for competitive sporting events. Coaches may consider using the information provided in this review for the development of more effective training regimens allowing athletes to attain their peak performance regardless of competition time. However, because of population heterogeneity and the broad definition of “athletic performance” used, careful application of these findings is necessary. Training programs should consider the nature of the event, the athlete's circadian phenotype, and the habitual time of training vs. time of competition. In addition, preparation should emphasize an adequate warm-up that raises the athlete's body temperature. Caffeine ingestion may also be considered for tasks requiring increased concentration and physical capacity. These considerations are particularly important if the athlete is a night owl attempting to maximize morning performance.

Acknowledgments

The authors would like to thank Elaine Hicks from the Matas Library at Tulane University School of Medicine for her expertise on systematic reviews and Miranda J. McDaniel for her superb editing. The authors declare no conflicts of interest. No funding was received for this work.

References

1. Ammar A, Chtourou H, Hammouda O, et al. Acute and delayed responses of C-reactive protein, malondialdehyde and antioxidant markers after resistance training session in elite weightlifters: Effect of time of day. Chronobiol Int 32: 1211–1222, 2015.
2. Ammar A, Chtourou H, Trabelsi K, et al. Temporal specificity of training: Intra-day effects on biochemical responses and Olympic-weightlifting performances. J Sports Sci 33: 358–368, 2015.
3. Atkinson G, Todd C, Reilly T, Waterhouse J. Diurnal variation in cycling performance: Influence of warm-up. J Sports Sci 23: 321–329, 2005.
4. Available at: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools. Accessed June 20, 2018.
5. Belkhir Y, Rekik G, Chtourou H, Souissi N. Listening to neutral or self-selected motivational music during warm-up to improve short-term maximal performance in soccer players: Effect of time of day. Physiol Behav 204: 168–173, 2019.
6. Bernard T, Giacomoni M, Gavarry O, Seymat M, Falgairette G. Time-of-day effects in maximal anaerobic leg exercise. Eur J Appl Physiol Occup Physiol 77: 133–138, 1998.
7. Birch K, Reilly T. The diurnal rhythm in isometric muscular performance differs with eumenorrheic menstrual cycle phase. Chronobiol Int 19: 731–742, 2002.
8. Brown FM, Neft EE, LaJambe CM. Collegiate rowing crew performance varies by morningness-eveningness. J Strength Cond Res 22: 1894–1900, 2008.
9. Cagnacci A, Volpe A, Paoletti AM, Melis GB. Regulation of the 24-hour rhythm of body temperature in menstrual cycles with spontaneous and gonadotropin-induced ovulation. Fertil Sterility 68: 421–425, 1997.
10. Castaingts V, Martin A, Van Hoecke J, Perot C. Neuromuscular efficiency of the triceps surae in induced and voluntary contractions: Morning and evening evaluations. Chronobiol Int 21: 631–643, 2004.
11. Chaâri N, Frikha M, Mezghanni N, et al. Does post-warm-up rest interval affect the diurnal variation of 30-s Wingate cycle ergometry? Biol Rhythm Res 46: 949–963, 2015.
12. Chtourou H, Zarrouk N, Chaouachi A, et al. Diurnal variation in Wingate-test performance and associated electromyographic parameters. Chronobiol Int 28: 706–713, 2011.
13. Chtourou H, Chaouachi A, Driss T, et al. The effect of training at the same time of day and tapering period on the diurnal variation of short exercise performances. J Strength Cond Res 26: 697–708, 2012.
14. Chtourou H, Chaouachi A, Hammouda O, Chamari K, Souissi N. Listening to music affects diurnal variation in muscle power output. Int J Sports Med 33: 43–47, 2012.
15. Chtourou H, Driss T, Souissi S, et al. The effect of strength training at the same time of the day on the diurnal fluctuations of muscular anaerobic performances. J Strength Cond Res 26: 217–225, 2012.
16. Chtourou H, Engel FA, Fakhfakh H, et al. Diurnal variation of short-term repetitive maximal performance and psychological variables in elite Judo athletes. Front Physiol 9: 1499, 2018.
17. Facer-Childs E, Brandstaetter R. The impact of circadian phenotype and time since awakening on diurnal performance in athletes. Curr Biol 25: 518–522, 2015.
18. Faull OK, Cotter JD, Lucas SJ. Cerebrovascular responses during rowing: Do circadian rhythms explain morning and afternoon performance differences? Scand J Med Sci Sports 25: 467–475, 2015.
19. Fernandes AL, Lopes-Silva JP, Bertuzzi R, et al. Effect of time of day on performance, hormonal and metabolic response during a 1000-M cycling time trial. PLoS One 9: e109954, 2014.
20. Frikha M, Chaâri N, Souissi N. Effect of sport practice and warm-up duration on the morning–evening difference in anaerobic exercise performance and perceptual responses to it. Biol Rhythm Res 46: 497–509, 2015.
21. Gauthier A, Davenne D, Martin A, Cometti G, Van Hoecke J. Diurnal rhythm of the muscular performance of elbow flexors during isometric contractions. Chronobiol Int 13: 135–146, 1996.
22. Gauthier A, Davenne D, Martin A, Van Hoecke J. Time of day effects on isometric and isokinetic torque developed during elbow flexion in humans. Eur J Appl Physiol 84: 249–252, 2001.
23. Gueldich H, Zarrouk N, Chtourou H, et al. Electrostimulation training effects on diurnal fluctuations of neuromuscular performance. Int J Sports Med 38: 41–47, 2017.
24. Guette M, Gondin J, Martin A. Time-of-day effect on the torque and neuromuscular properties of dominant and non-dominant quadriceps femoris. Chronobiol Int 22: 541–558, 2005.
25. Hill DW. Morning-evening differences in response to exhaustive severe-intensity exercise. Appl Physiol Nutr Metab 39: 248–254, 2014.
26. Hill DW, Leiferman JA, Lynch NA, Dangelmaier BS, Burt SE. Temporal specificity in adaptations to high-intensity exercise training. Med Sci Sports Exerc 30: 450–455, 1998.
27. Janse De Jonge XAK, Thompson MW, Chuter VH, Silk LN, Thom JM. Exercise performance over the menstrual cycle in temperate and hot, humid conditions. Med Sci Sports Exerc 44: 2190–2198, 2012.
28. Julian R, Hecksteden A, Fullagar HHK, Meyer T. The effects of menstrual cycle phase on physical performance in female soccer players. PLoS One 12: e0173951, 2017.
29. Kim HK, Konishi M, Takahashi M, et al. Effects of acute endurance exercise performed in the morning and evening on inflammatory cytokine and metabolic hormone responses. PLoS One 10: e0137567, 2015.
30. Kin-Isler A. Time-of-day effects in maximal anaerobic performance and blood lactate concentration during and after a supramaximal exercise. Isokinet Exerc Sci 14: 335–340, 2006.
31. Knaier R, Meister S, Aeschbacher T, et al. Dose-response relationship between light exposure and cycling performance. Scand J Med Sci Sports 26: 794–801, 2016.
32. Kraemer WJ, Boyd BM, Hooper DR, et al. Epinephrine preworkout elevation may offset early morning melatonin concentrations to maintain maximal muscular force and power in track athletes. J Strength Cond Res 28: 2604–2610, 2014.
33. Kuusmaa M, Schumann M, Sedliak M, et al. Effects of morning versus evening combined strength and endurance training on physical performance, muscle hypertrophy, and serum hormone concentrations. Appl Physiol Nutr Metab 41: 1285–1294, 2016.
34. Lericollais R, Gauthier A, Bessot N, Sesboue B, Davenne D. Time-of-day effects on fatigue during a sustained anaerobic test in well-trained cyclists. Chronobiol Int 26: 1622–1635, 2009.
35. Lericollais R, Gauthier A, Bessot N, Davenne D. Diurnal evolution of cycling biomechanical parameters during a 60‐s Wingate test. Scand J Med Sci Sports 21: e106–e114, 2011.
36. Mallette MM, Green LA, Gabriel DA, Cheung SS. The effects of local forearm muscle cooling on motor unit properties. Eur J Appl Physiol 118: 401–410, 2018.
37. Mallette MM, Green LA, Hodges GJ, et al. The effects of local muscle temperature on force variability. Eur J Appl Physiol 119: 1225–1233, 2019.
38. Marshall J. Thermal changes in the normal menstrual cycle. Br Med J 1: 102–104, 1963.
39. Martin A, Carpentier A, Guissard N, Van Hoecke J, Duchateau J. Effect of time of day on force variation in a human muscle. Muscle Nerve 22: 1380–1387, 1999.
40. Martin L, Nevill AM, Thompson KG. Diurnal variation in swim performance remains, irrespective of training once or twice daily. Int J Sports Physiol Perform 2: 192–200, 2007.
41. Mayeuf-Louchart A, Staels B, Duez H. Skeletal muscle functions around the clock. Diabetes Obes Metab 17: 39–46, 2015.
42. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 6: e1000097, 2009.
43. Morgan JA, Corrigan F, Baune BT. Effects of physical exercise on central nervous system functions: A review of brain region specific adaptations. J Mol Psychiatry 3: 3, 2015.
44. Nicolas A, Gauthier A, Bessot N, Moussay S, Davenne D. Time‐of‐day effects on myoelectric and mechanical properties of muscle during maximal and prolonged isokinetic exercise. Chronobiol Int 22: 997–1011, 2005.
45. Nicolas A, Gauthier A, Michaut A, Davenne D. Effect of circadian rhythm of neuromuscular properties on muscle fatigue during concentric and eccentric isokinetic actions. Isokinet Exerc Sci 15: 117–129, 2007.
46. Onambele-Pearson NG, Pearson SJ. Time-of-day effect on patella tendon stiffness alters vastus lateralis fascicle length but not the quadriceps force–angle relationship. J Biomech 40: 1031–1037, 2007.
47. Pereira R, Machado M, Ribeiro W, et al. Variation of explosive force at different times of day. Biol Sport 28: 3–9, 2011.
48. Petrofsky J, Al Malty A, Suh HJ. Isometric endurance, body and skin temperature and limb and skin blood flow during the menstrual cycle. Med Sci Monit 13: Cr111–117, 2007.
49. Pullinger SA, Brocklehurst EL, Iveson RP, et al. Is there a diurnal variation in repeated sprint ability on a non-motorised treadmill? Chronobiol Int 31: 421–432, 2014.
50. Racinais S, Hue O, Blonc S. Time-of-day effects on anaerobic muscular power in a moderately warm environment. Chronobiol Int 21: 485–495, 2004.
51. Racinais S, Hue O, Hertogh C, Damiani M, Blonc S. Time-of-day effects in maximal anaerobic leg exercise in tropical environment: A first approach. Int J Sports Med 25: 186–190, 2004.
52. Racinais S, Blonc S, Hue O. Effects of active warm-up and diurnal increase in temperature on muscular power. Med Sci Sports Exerc 37: 2134, 2005.
53. Racinais S, Blonc S, Jonville S, Hue O. Time of day influences the environmental effects on muscle force and contractility. Med Sci Sports Exerc 37: 256–261, 2005.
54. Racinais S, Connes P, Bishop D, Blonc S, Hue O. Morning versus evening power output and repeated‐sprint ability. Chronobiol Int 22: 1029–1039, 2005.
55. Racinais S, Blonc S, Oksa J, Hue O. Does the diurnal increase in central temperature interact with pre-cooling or passive warm-up of the leg? J Sci Med Sport 12: 97–100, 2009.
56. Rae DE, Stephenson KJ, Roden LC. Factors to consider when assessing diurnal variation in sports performance: The influence of chronotype and habitual training time-of-day. Eur J Appl Physiol 115: 1339–1349, 2015.
57. Reilly T, Garrett R. Effects of time of day on self-paced performances of prolonged exercise. J Sports Med Phys Fitness 35: 99–102, 1995.
58. Scheer FAJL, Hilton MF, Mantzoros CS, Shea SA. Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci 106: 4453–4458, 2009.
59. Sedliak M, Finni T, Cheng S, Kraemer WJ, Häkkinen K. Effect of time-of-day-specific strength training on serum hormone concentrations and isometric strength in men. Chronobiol Int 24: 1159–1177, 2007.
60. Sedliak M, Finni T, Cheng S, Haikarainen T, Häkkinen K. Diurnal variation in maximal and submaximal strength, power and neural activation of leg extensors in men: Multiple sampling across two consecutive days. Int J Sports Med 29: 217–224, 2008.
61. Souissi N, Gauthier A, Sesboue B, Larue J, Davenne D. Circadian rhythms in two types of anaerobic cycle leg exercise: Force-velocity and 30-s Wingate tests. Int J Sports Med 25: 14–19, 2004.
62. Souissi N, Bessot N, Chamari K, et al. Effect of time of day on aerobic contribution to the 30‐s Wingate test performance. Chronobiol Int 24: 739–748, 2007.
63. Souissi N, Driss T, Chamari K, et al. Diurnal variation in Wingate test performances: Influence of active warm-up. Chronobiol Int 27: 640–652, 2010.
64. Souissi M, Abedelmalek S, Chtourou H, et al. Effects of time-of-day and caffeine ingestion on mood states, simple reaction time, and short-term maximal performance in elite judoists. Biol Rhythm Res 44: 897–907, 2013.
65. Souissi Y, Souissi M, Chtourou H. Effects of caffeine ingestion on the diurnal variation of cognitive and repeated high-intensity performances. Pharmacol Biochem Behav 177: 69–74, 2019.
66. Tahara Y, Aoyama S, Shibata S. The mammalian circadian clock and its entrainment by stress and exercise. J Physiol Sci 67: 1–10, 2017.
67. Taylor K, Cronin JB, Gill N, Chapman DW, Sheppard JM. Warm-up affects diurnal variation in power output. Int J Sports Med 32: 185–189, 2011.
68. Torii J, Shinkai S, Hino S, et al. Effect of time of day on adaptive response to a 4-week aerobic exercise program. J Sports Med Phys Fitness 32: 348–352, 1992.
69. West DJ, Cook CJ, Beaven MC, Kilduff LP. The influence of the time of day on core temperature and lower body power output in elite rugby union sevens players. J Strength Cond Res 28: 1524–1528, 2014.
70. Yamanaka Y, Hashimoto S, Masubuchi S, et al. Differential regulation of circadian melatonin rhythm and sleep-wake cycle by bright lights and nonphotic time cues in humans. Am J Physiol Regul Integr Comp Physiol 307: R546–R557, 2014.
71. Zarrouk N, Chtourou H, Rebai H, et al. Time of day effects on repeated sprint ability. Int J Sports Med 33: 975–980, 2012.
Keywords:

circadian; chronotype; sports; temperature; warm-up

© 2020 National Strength and Conditioning Association