Acute physiological, endocrine, biochemical and performance responses associated with amateur boxing: A systematic review with meta-analysis

ABSTRACT

Previous research has explored the demands of amateur boxing-specific activity; however, no holistic review of the acute responses to such activity currently exists. This paper aimed to provide a systematic review of the available literature on the acute physiological, endocrine, biochemical, and performance responses to amateur boxing-specific activity. Following a search of EBSCOhost, SPORTDiscus, PubMed and Google Scholar databases, 25 studies were identified as meeting the inclusion criteria for the review. The methodological quality of the included studies was assessed via a modified Downs and Black checklist. Random-effects meta-analysis of standardised mean differences (SMD) revealed large (SMD = 4.62) increases in pre-post blood lactate (BLa), cortisol (SMD = 1.33), myoglobin (Mb) (SMD = 1.43), aspartate transaminase (AST) (SMD = 1.37), and alanine aminotransferase (ALT) (SMD = 0.97), in addition to moderate increases in creatine kinase (CK) (SMD = 0.65). Small pre–post increases in counter-movement jump (CMJ) height (SMD = 0.33) were observed. Consistently greater pre–post alterations were observed in competitive bouts, followed by sparring, and less so in boxing-specific simulations. Considerable physiological, endocrine, and biochemical responses are elicited following amateur boxing. Interestingly, neuromuscular and task-specific performance may not deteriorate following boxing-specific activity. The findings of the review may assist in the designing and periodising of boxing-specific training, dependent on the desired physical adaptations, training phase, and recovery status of the amateur boxer.

Highlights

• Amateur boxing elicits a considerable acute physiological, hormonal, and biochemical response.

• Such responses are typically greater in competitive bouts, followed by sparring, and less so in simulated activity.

• The considerable demands of amateur boxing-specific activity do not appear to negatively affect neuromuscular or task-specific performance.

• Amateur boxers may be conditioned to preserve performance despite the acute demands of the sport, or the lack of performance decrement may reflect the short duration of amateur boxing.

KEYWORDS:

• Competition
• training
• physiology
• endocrinology
• performance

Introduction

Senior elite male amateur boxing consists of intermittent periods of high-intensity activity across 3 × 3-minute rounds, interspersed with 1-minute periods of passive and active recovery (Davis, Benson, Pitty, Connorton, & Waldock, 2015; Davis, Connorton, Driver, Anderson, & Waldock, 2018). Competitors throw between 61 and 78 punches per round, in addition to defensive actions (Davis et al., 2015, 2018), corresponding to a work-to-rest ratio of 18:1 (Slimani, Chaabene, Davis Franchini, Cheour, & Chamari, 2017). This activity and the high-pressures involved in competitive combat provoke a substantial physiological, biochemical, and psychological demand (Slimani et al., 2018; Slimani, Chaabene, et al., 2017).

Researchers have gained an appreciation of the acute physiological demands of boxing-specific activity, utilising heart rate (HR), blood lactate (BLa), oxygen uptake (V̇O₂) and perceptual measures of intensity. The findings of these studies suggest that these settings generally elicit moderate-to-high demands on aerobic and anaerobic metabolism, and perceptual responses (Slimani, Chaabene, et al., 2017; Slimani, Davis, Franchini, & Moalla, 2017). In contrast to our physiological understanding, little is known about the endocrine response to boxing-specific activity. Indeed, the most recent review on the hormonal responses to combat sports by Slimani et al. (2018), unfortunately, did not include boxing. Nevertheless, the same authors found moderate-to-extremely large pre–post increases in the following hormonal parameters in other combat sports; cortisol (p < 0.001, ES = 0.79); adrenaline (p < 0.001, ES 4.22); noradrenaline (p = 0.005, ES = 3.40) and human growth hormone (HGH) levels (p < 0.001, ES = 3.69). Furthermore, an elevated hormonal response was associated with competitive bouts, when compared to simulated activity.

The acute physiological and hormonal demands elicited by boxing-specific activity may evoke muscle damage and inflammation, though again, this has been afforded only limited attention (Bianco et al., 2005; Graham et al., 2011; Kaynar, 2019; Kilic et al., 2019; Obminski & Turowski, 2014; Zuliani et al., 1985). Increased skeletal muscle damage, inflammation, and physiological stress biomarkers have been found up to 24 hrs following mixed martial arts (MMA) and wrestling (Ghoul et al., 2019; Kraemer et al., 2001; Lindsay et al., 2017). Wiechmann, Saygili, Zilkens, Krauspe, and Behringer (2016) found the volume of strikes received in MMA, significantly contributed to the variance of creatine kinase (CK) and myoglobin (Mb) concentrations in the blood, and so, given the similarities in the striking elements between the sports, scrutiny of the above parameters appears warranted in boxing.

Researchers have also been concerned with how the above acute responses may influence neuromuscular and task-specific performance in boxers (Hukkanen & Häkkinen, 2017; Loturco et al., 2021; Nikolaidis, Clemente, Busko, & Knechtle, 2017) and other combat disciplines (Andreato et al., 2015; Detanico, Dal Pupo, Franchini, & Dos Santos, 2015; Detanico, Dellagrana, Saldanha, Kons, & Goes, 2017; Ghoul et al., 2019; Kraemer et al., 2001) in single and repeat-bout scenarios. For example, significant decreases in strength and lower-body impulse have been observed in repeat-bout collegiate wrestling, Brazilian jiu-jitsu and Judo competition (Andreato et al., 2015; Detanico et al., 2015, 2017; Kraemer et al., 2001). Both domestic amateur boxing championships and Olympic boxing competitions are characterised by repeat-bout scenarios, whereby successful boxers are required to compete on consecutive days, or on several days across a week (England Boxing, 2021). The detriments to the performance observed in the previous repeat-bout combat literature may not be generalisable to amateur boxing, therefore further examination of the acute effects of repeat-bout amateur boxing on task-specific and neuromuscular performance may also be warranted.

In view of our understanding of the endocrine, biochemical and performance responses to other combat sports, the data may not be generalisable to boxing due to its unique characteristics (duration, rules, activity profile), thus, there is a need to review the body of research on common biomarkers in amateur boxing. Likewise, a greater understanding of the acute responses associated with competition, sparring and simulated activity may assist in the optimisation of training and recovery design for coaches and practitioners. Moreover, findings may assist in reducing the likelihood of injury, overtraining syndrome or non-functional overreaching (Kellmann et al., 2018), perhaps of increased importance during repeat-bout tournaments (England Boxing, 2020). Therefore, this study aims to provide the first holistic review of the literature pertaining to the acute physiological, endocrine, biochemical and performance responses to competitive boxing bouts, sparring and simulated boxing activity.

Methods

Inclusion criteria

To be included in the review, studies were required to meet the following criteria: (i) studies included either competitive bouts, sparring or simulated amateur boxing activity, (ii) studies monitored either the physiological, endocrine, biochemical or performance response, (iii) participants were carded senior amateur boxers within their respective governing bodies, (iv) full-text versions of studies could be accessed in English language peer-reviewed journals.

Search strategy

A review of the literature was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (Moher, Liberati, Tetzlaff, & Altman, 2009). EBSCOhost, SPORTDiscus, Pubmed and Google Scholar databases were searched until May 2021 for all studies pertaining to acute responses to boxing. The following combination of search terms was used: “Physiological” OR “Biomechanical” OR “Physical” OR “Biochemical” OR “Heart rate” OR “Blood lactate” OR “Recovery” OR “Fatigue” OR “Endocrine” OR “Hormonal” OR “Neuromuscular” OR “Performance” OR “Jump” OR “Punch force” OR “Muscle damage” OR “Inflammation” OR “Creatine kinase” OR “Repeat-bout” OR “Tournament” AND “Boxing” OR “Striking” OR “Combat sport”. The literature search was conducted independently by two authors (MF and CB).

Study coding and data extraction

Two authors (MF and CB) performed the study coding and data extraction. The following information from studies meeting the inclusion criteria was extracted where possible: (i) participant characteristics (age, sex, sample size, nationality and competitor level); (ii) boxing activity performed (competitive bouts, sparring, simulated activity and the rounds and duration performed); (iii) acute responses quantified (physiological, endocrine, biochemical and performance responses) and (iv) time of measurements (i.e. pre–post and/or across rounds).

Methodological quality

To assess the methodological quality of each study, a modified Downs and Black (1998) checklist was utilised by the same two authors who conducted the literature search and study coding and data extraction (MF and CB). There were no disagreements between raters during this process and hence a third reviewer (arbitrator) was not needed. Eleven of the original 27 criteria were deemed appropriate for inclusion in the methodological assessment as follows: 1 = hypothesis, aims and objectives stated; 2 = main outcomes described; 3 = subject characteristics described; 4 = intervention described; 5 = main findings described; 6 = established random variability; 7 = actual probability reported; 8 = potential recruits representative of entire population; 9 = participants representative of the entire population; 10 = appropriate statistical tests; 11 = outcome measures valid and reliable. These criteria to be addressed were classified as “yes” or “no”, followed by a total summation of the criteria scores, out of 11.

Statistical analysis

Mean ± SD values of pre–post and/or across round responses to boxing-specific activity were tabulated. Select studies did not disclose the mean ± SD data, thus, data were requested by contacting the authors in the first instance, and extracted from a valid and reliable web-based plot digitising tool (WebPlotDigitiser) in the second instance (Drevon, Fursa, & Malcolm, 2017). The authors reported such data with a ∼ preceding the results, emphasising that the data are approximations due to varied standards of graph presentation. A random-effects meta-analysis of pre and post standardised mean differences (SMD) was performed using the meta-analysis software (RevMan 5.4, Cochrane). Specifically, this comprised the physiological response, as quantified by BLa levels, hormonal response as quantified by cortisol, and the muscle damage and inflammation responses as quantified by CK, Mb, alanine transaminase (ALT) and aspartate transaminase (AST). Lastly, the pre–post SMD of the neuromuscular responses to boxing-specific activity was performed, as quantified by the counter-movement jump (CMJ) height. The SMD was calculated via the adjusted Hedges g comprising the following thresholds: small = 0.20–0.49, moderate = 0.50–0.80 and large = ≥0.8 (Hedges & Olkin, 1985). Heterogeneity between studies was assessed by observing the I² statistic, at the following thresholds: low = <25%, moderate <25%, and high >75%. The combined effect sizes (Hedges g) for each boxing-specific mode were also compared to explore potential differences in the acute responses (Figure 2).

Results

Search results

A total of 78 studies pertaining to the search criteria were identified for further analysis, of which 20 duplicates were removed (Figure 1). Titles and abstracts of 58 manuscripts were initially screened for their relevance, followed by a full-text review of the remaining 39 relevant manuscripts to assess the eligibility in accordance with the eligibility criteria. Studies were excluded if they included any of the following: (1) an inappropriate simulation activity profile (e.g. maximum amount of punches per round), (2) participants of mixed-combat disciplines, (3) any type of physical or ergogenic intervention and (4) case study. A total of 25 studies were selected for inclusion in the review (Figure 1), published from 1985 to 2021. Of the 25 studies included, 18 were further selected for meta-analysis. A total of 380 participants from the 25 studies were included, comprising 321 males and 30 females, the sex of 29 participants was not provided. The average age of the boxers was 21.0 ± 2.2. Elite boxers accounted for 65% of all boxers, whilst 35% were described as non-elite (novice or intermediate). The physiological responses to boxing-specific activity in the literature are discussed in the text and displayed in Table 1, whilst endocrine, biochemical and performance responses are presented in Table 2.

Figure 1. Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) flow diagram of literature screening process.

Table 1. Physiological and perceptual alterations in competitive bouts, sparring, and boxing-specific simulations.

Reference	Participants	Activity	Time	HRave (b.min^−1)	% HRmax	HRpeak (b.min^−1)	% HRmax	V̇O2ave (ml.kg^−1 min^−1)	% V̇O2max	V̇O2peak (ml.kg^−1min^−1)	% V̇O2max	BLa (mmol.l^−1)	RPE (Borg 6–20)
Simulated Activity
Finlay et al. (2020)	14 M English elite amateur boxers age 22 ± 2	3 × 3 Punch bag simulation	Baseline Round 1 Round 2 Round 3 Post (5 mins)	– 144 ± 15 150 ± 17* 157 ± 16*# –	– 75 79 82 –	– 158 ± 20 165 ± 20* 170 ± 17*# 118 ± 14	– 83 86 89 –	– 33.2 ± 6.8 35.0 ± 7.9 35.1 ± 7.8 –	– 59 62 63 –	– 42.4 ± 6.0 43.8 ± 7.3 44.4 ± 5.6 –	– 76 78 79 –	1.2 ± 0.5 2.1 ± 0.8 2.7 ± 1.0* 3.3 ± 1.2*# 2.5 ± 0.9	– 11 ± 1 13 ± 1* 14 ± 2*# –
		3 × 3 Pads simulation	Baseline Round 1 Round 2 Round 3 Post (5 mins)	– 154 ± 10 163 ± 9 * 167 ± 9*# –	– 81 85 87 –	– 167 ± 12 175 ± 12* 178 ± 9*# 114 ± 10	– 87 92 93 –	– 37.7 ± 6.9 38.1 ± 6.7 38.2 ± 6.9 –	– 67 68 68 –	– 46.9 ± 8.1 47.3 ± 8.5 47.0 ± 8.4 –	– 84 84 84 –	1.1 ± 0.5 2.3 ± 0.7 3.1 ± 0.8* 4.0 ± 1.0*# 3.0 ± 0.9	– 11 ± 2 13 ± 2* 14 ± 2*# –
El-Ashker et al. (2018)	11 M Elite amateur boxers age 21.4 ± 2.1	3 × 3 Semi contact pads simulation	Round 1 Round 2 Round 3	– – –	– – –	176 ± 4 179 ± 1* 182 ± 5*#	90 91 92	– – –	– – –	55.3 ± 4.4 53.8 ± 6.5 50.4 ± 7.9	100 97 91	– – –	– – –
Finlay et al. (2018)	9 M English elite amateur boxers age 21 ± 4	3 × 3 Punch bag simulation	Baseline Round 1 Round 2 Round 3	– 150 ± 15 156 ± 16* 157 ± 18*	– 79 83 83	– 162 ± 12 166 ± 13* 169 ± 14*	– 86 88 89	– 32.3 ± 4.9 33.0 ± 5.8 32.7 ± 6.0	– 59 60 59	– 43.8 ± 7.3 43.3 ± 8.2 43.5 ± 8.5	– 80 79 79	1.1 ± 0.9 2.4 ± 1.3 3.3 ± 1.7* 4.3 ± 2.6*#	– 11 ± 1 13 ± 1* 14 ± 2*#
Davis et al. (2014)	10 M Novice boxers 23.7 ± 4.1	3 × 2 Pads	Round 1 Round 2 Round 3	– – –	– – –	166 ± 19 173 ± 12 174 ± 13*	87 90 91	– – –	– – –	42.6 47.5* 47.2*	– – –	6.7 ± 1.4 8.6 ± 1.5* 9.5 ± 1.8*#	– – –
Halperin et al. (2019)	15 M Australian elite amateur boxers age 21 ± 4	3 × 2 punch simulation (Differing performance feedback)	Round 1 Round 2 Round 3	– – –	– – –	– – –	– – –	– – –	– – –	– – –	– – –	– – –	14 ± 1 16 ± 1* 17 ± 1*#
Arseneau et al. (2011)	9 M Canadian regional level amateur boxers age 22 ± 3.5	3 × 2 Pads simulation	Average Round 1 Round 2 Round 3	– 152 ± 14 168 ± 12* 172 ± 12*	– 78 86 88	– – – –	– – – –	41.1 ± 5.1 – – –	∼66 – – –	– – – –	– – – –	– – – 4.0 ± 1.7	– – – 13.6 ± 2.1
Thomson and Lamb (2017)	28 M Non-elite amateur boxers age 22.4 ± 3.5	3 × 3 Pads simulation	Average Round 1 Round 2 Round 3	– 165 ± 11 172 ± 10* 175 ± 10*#	– – – –	– 178 ± 13 187 ± 8* 189 ± 11*	– – – –	– 42.1 43.7 40.7	∼ 69 – – –	– – – –	∼92 – – –	– – – 4.6 ± 1.3	(CR-10 scale) 5.8 ± 1.4 6.8 ± 1.1* 8.1 ± 1.1*#
Ghosh (2010)	6 M Indian elite amateur boxers age 21.4 ± 3	4 × 2 Punch bag simulation	Round 4	–	–	192 ± 6	–	–	–	–	–	13.6 ± 3.2	–
Lutoslawska et al. (1995)	9 M Polish elite amateur boxers age 24.6 ± 3.7	3 × 3 Punch simulation	Average Baseline Round 3	– – –	– – –	183 ± 8.9	– – –	– – –	– – –	– – –	– – –	– 0.9 ± 0.2 14.0 ± 1.4#	– – –
Sparring and simulated bouts
Khanna and Manna (2006)	21 M Indian elite amateur boxers age 17.7	3 × 2 Competitive sparring (all weight classes)	Round 1 Round 2 Round 3	171 ± 16 179 ± 13* 183 ± 15*	93 97 99	– – –	– – –	– – –	– – –	– – –	– – –	6.2 ± 3.4 6.3 ± 3.4 8.3 ± 4.5	– – –
Hukkanen and Hakkinen. (2017)	7 M Finish elite amateur boxers age 20.3 ± 2.7	3 × 3 Sparring (pre-comp phase) 3 × 3 sparring (Comp phase)	Baseline Round 2 Round 3 Baseline Round 1 Round 2 Round 3	– – – – 183 ± 6 ∼185 ± 11 191 ± 7	– – – – – – –	– – – – – – –	– – – – – – –	– – – – – – –	– – – – – – –	– – – – – – –	– – – – – – –	5.0 ± 1.0 14.0 ± 2.0 17.0 ± 2.0 3.0 ± 1.0 – 12.0 ± 3.0 13.0 ± 3.0	– – – – – – –
Obminski et al. (1993)	11 M Elite amateur boxers Age 22.1	3 × 3 Sparring	Baseline Round 1 Round 2 Round 3	– – – –	– – – –	– 181 ± 9 186 ± 10 187 ± 9	– – – –	– – – –	– – – –	– – – –	– – – –	1.6 ± 0.1 – – 14.4 ± 2.2*	– – – –
Nassib et al. (2017)	15 M Tunisian elite amateur boxers age 19.56 ± 3.6	3 × 3 Competitive sparring	Average Round 1 Round 3	173 ± 7 – –	93 – –	– – –	– – –	– – –	– – –	– – –	– – –	– 1.8 ± 0.9 8.9 ± 2.0#	– – –
Ghosh (2010)	6 M Indian elite amateur boxers age 21.4 ± 3	6 × 2 Competitive sparring	Average Round 6	– –	–	190 ± 7	–	–	–	–	–	– 14.5 ± 0.6	– –
Arseneau et al. (2011)	9 M Canadian regional level amateur boxers age 22 ± 3.5	3 × 2 Sparring LAB	Average Round 1 Round 2 Round 3	– 162 ± 10 168 ± 12* 172 ± 12*	– – – –	– – – –	– – – –	43.4 ± 5.9 – – –	70 – – –	– – – –	– – – –	– – – 6.1 ± 2.3	12.2 ± 1.3 – – –
		3 × 2 Sparring GYM	Round 1 Round 2 Round 3	174 ± 8 181 ± 8* 183 ± 8*	89 93 94	– – –	– – –	– – –	– – –	– – –	– – –	– – 9.4 ± 2.2	– – –
Nikolaidis et al. (2017)	15 M Greek novice Amateur boxers age 21.4 ± 6.3	3 × 3 Simulated boxing (sitting and standing rest)	Round 1 Round 2 Round 3	177 ± 12 184 ± 7* 187 ± 7*	– – –	– – –	– – –	– – –	– – –	– – –	– – –	– – –	12.9 ± 3.0 14.2 ± 2.5* 14.6 ± 2.7*
de Lira et al. (2013)	6 M 4 F Brazilian experienced boxers age 17.9 ± 1.8	3 × 2 Simulated bouts	Round 1 Round 2 Round 3	175 ± 11 183 ± 6* 186 ± 7*	91 95 96	190 ± 8 194 ± 6 199 ± 5*	98 101 103	44.2 ± 7.2 48 .0 ± 6.1* 49.6 ± 5.6*	85 92 95	50.8 ± 5.6 52.6 ± 5.2 54.6 ± 6.2*	99 102 105	– – –	– – –
Chatterjee et al. (2006)	20 F Elite Indian amateur boxers Age 19.7 ± 2.2	3 × 2 Sparring	Round 1 Round 2 Round 3	171 ± 11 181 ± 8 184 ± 7	– – –	183 ± 6 189 ± 6 193 ± 7	– – –	– – –	– – –	– – –	– – –	– – –	– – –
Competitive Bouts
Davis et al. (2013)	32 M British novice amateur boxers age 19.3 ± 1.4	3 × 2 Competitive bouts	Baseline Round 3 Post (5 mins) Post (7 mins)	– – – –	– – – –	– – – –	– – – –	– – – –	– – – –	– – – –	– – – –	1.3 ± 0.5 11.8 ± 1.6# 10.8 ± 2.2 9.9 ± 2.5	– – – –
Hanon et al. (2015)	33 M Elite amateur boxers age 21.5 ± 2.1	3 × 3 Competitive bouts	Round 3	–	–	–	–	–	–	–	–	13.6 ± 2.4	–
Obminski et al. (2009)	15 U Elite amateur boxers age 20–25	3 × 3 Competitive bouts	Post (3 mins)	–	–	–	–	–	–	–	–	12.6 ± 4.7	–

Data are mean ± SD, % where applicable, or ∼ for estimates. M = male; F = female; HRave = average heart rate; HRmax = heart rate maximum; HRpeak = heart rate peak; V̇O₂ave = average oxygen consumption; V̇O₂max = maximal oxygen uptake; V̇O₂peak; maximum oxygen consumption; BLa = blood lactate; RPE = rating of perceived exertion; LAB = laboratory.

*Denotes significant (p < 0.005) difference compared to round 1.

#Denotes significant (p < 0.005) difference compared to previous round.

Table 2. Endocrine, biochemical and performance alterations in competitive bouts, sparring, and boxing-specific simulations.

Reference	Participants	Activity	Measures	Pre	Post-activity			ES
Endocrine response
Graham et al. (2011)	16 M Novice amateur boxers Full sparring n = 8 Age 19.1 ± 3.2 Body sparring n = 8 age 17.6 ± 5.3	Boxing (Full sparring) Boxing (Body sparring)	Serum cortisol (nmol.1^−1)	373 ± 202 416 ± 140	5 mins post 756 ± 93* 417 ± 135	– –	– –	2.30 0.01
Kilic et al. (2019)	20 M Elite amateur boxers age 25.9 ± 3.3	3 × 3 Competitive bout	Plasma cortisol (ug/dl)	∼9 ± 7	10 mins post ∼28 ± 13*	–	–	1.86
Obminski et al. (1993)	11 M Elite amateur boxers age 22.1	3 × 3 Competitive bout 3 × 3 min Sparring	Salivary cortisol (nmol.1^−1)	84.4 ± 17.3 28.6 ± 14	3 mins post 73.5 ± 11.6 40.7 ± 29.2	15 mins post 93.5 ± 19 51.2 ± 17.1	30 mins post 103.9 ± 17.9* 52.5 ± 9.6*	1.02 1.80
Obminski et al. (2009)	15 U Elite amateur boxers age (range 20–25)	3 × 3 Competitive bout	Serum cortisol (nmol^−1) Testosterone (nmol^−1)	– –	3 mins post 600 ± 104 11.6 ± 4.7	– –	– –	– –
Biochemical response
Graham et al. (2011)	16 M Novice amateur boxers Full sparring n = 8 age 19.1 ± 3.2 Body sparring n = 8 age 17.6 ± 5.3	Boxing (full sparring) Boxing (body sparring)	CK (u/L) CK (u/L)	207 ± 107 150 ± 43	5 mins post 244 ± 118* 195 ± 63*	– –	– –	0.33 0.79
Kilic et al. (2019)	20 M Turkish elite amateur boxers age 25.9 ± 3.3	3 × 3 Competitive bout	AST (u/L) ALT (u/L)	∼14 ± 4 ∼5 ± 4	10 mins post ∼35 ± 13* ∼14 ± 4*	– –	– –	2.14 2.21
Zuliani et al. (1985)	8 M Non-elite amateur boxers age (range 15–32)	3 × 3 Competitive bout 3 × 3 Shadow boxing	CK (u/L) Mb (ng/ml) LDH (u/L) CK(u/L) Mb (ng/ml) LDH (u/L)	119 ± 38 15 ± 3 136.75 114 ± 38 15 ± 3 127.06	15 mins post 139 ± 43 52 ± 27* 152.38 149 ± 61 28 ± 11* 135.03	5 hrs post 194 ± 82* 75 ± 36* 143.52 156 ± 61* 31 ± 13* 147.34	16 hrs post 183 ± 78* 27 ± 21 141.58 129 ± 53* 18 ± 9 134.94	1.11 2.22 – 0.78 1.60 –
Kaynar (2019)	14 U Novice amateur boxers age 17.9 ± 5.6	3 × 3 Body sparring	AST (u/L) ALT (u/LH) LDH (u/L) CK (u/L)	23.21 ± 4.12 14.93 ± 3.83 222.21 ± 36.00 242.14 ± 183.55	5 mins post 25.96 ± 4.69* 13.79 ± 4.82 287 ± 61.30* 332.14 ± 198.36*	– – – –	– – – –	0.62 0.26 1.29 0.46
Bianco et al. (2005)	15/18 M Italian elite amateur boxers age 24.9 ± 3.7	4 × 2 Competitive bout	CK (u/L) Mb (ng/ml)	208.7 ± 149.8 40.0 ± 11.9	– –	12 hrs post 300.1 ± 136.2* 53.9 ± 16.1*	–	0.65 0.96
Neuromuscular performance response
Nikolaidis et al. (2017)	15 M Greek novice amateur boxers age 21.4 ± 6.3	3 × 3 Simulated boxing (sitting and standing rest)	CMJ height (cm)	32.2 ± 4.6	5–10 mins post 34.2 ± 4.6*	–	–	0.42
Loturco et al. (2021)	8 M 6 F Brazilian elite amateur boxers Age 24.3 ± 4.3	3 × 3 Competitive bout	CMJ height (cm) HS power output (W.kg^−1) BP power output (W.kg^−1)	37.5 ± 7 7.70 ± 1.49 6.38 ± 1.06	1 hr post 39.4 ± 8.4 7.44 ± 1.61 6.43 ± 1.20	– – –	– – –	0.24 0.16 0.04
Task-specific performance response
				Pre	Round 1	Round 2	Round 3
Halperin et al. (2019)	15 M Australian elite amateur boxers age 21 ± 4	3 × 2 Punch simulation (Differing performance feedback)	Punch force (N) Various punch types	–	2081 ± 273	2119 ± 210	2141 ± 326	–
Hukkanen and Hakkinen (2017)	7 M Finnish elite amateur boxers age 20.3 ± 2.7	3 × 3 Competitive sparring (PC and C phase)	CP – Punch force (kg) Rear straight Jab Reaction time (ms) Rear straight Jab PCP – Punch force (kg) Rear straight Jab Reaction time (ms) Rear straight Jab	∼174 ± 36 ∼134 ± 37 354 ± 24 ∼303 ± 53 ∼192 ± 42 ∼148 ± 40 ∼347 ± 46 ∼293 ± 33	∼174 ± 33 ∼142 ± 31 ∼342 ± 42 ∼311 ± 45 ∼191 ± 41 ∼150 ± 39 332 ± 29 272 ± 62	∼169 ± 32 ∼138 ± 35 ∼344 ± 55 ∼290 ± 53 ∼194 ± 54 ∼148 ± 51 ∼346 ± 49 ∼312 ± 53	176 ± 52 149 ± 36 312 ± 36 323 ± 96 209 ± 61 156 ± 51 390 ± 57 324 ± 65	– – – – – – – –
Thomson and Lamb. (2017)	28 M Non-elite amateur boxers Age 22.4 ± 3.5	3 × 3 Pads simulation	Punch accelerations (g)	–	2697.3 ± 134.3	2768.1 ± 107.7#	2782 ± 100.1	–
Finlay et al. (2020)	14 M English elite amateur boxers Age 22 ± 2	3 × 3 Pads simulation 3 × 3 Punch bag simulation	PLTotal (a.u.) Cervical Lumbar Cervical Lumbar Overall	– – – – – –	39.48 ± 8.90 44.62 ± 7.59 44.04 ± 6.40 48.34 ± 9.91 44.12 ± 8.64	41.37 ± 8.32 46.48 ± 6.43 45.46 ± 6.03 51.82 ± 11.25 46.28 ± 8.84#	42.09 ± 8.13 47.33 ± 7.86 44.88 ± 6.17 51.23 ± 12.80 46.38 ± 9.42#	– – – – –
Finlay et al. (2018)	9 M English elite amateur boxers Age 21 ± 4	3 × 3 Punch bag simulation	PLTotal (a.u.) Cervical	– –	46.85 ± 6.88	48.69 ± 7.02	47.56 ± 6.13	–

Notes: Data are mean ± SD, or ∼ for estimates. M: male; F: female; ES: effect size (g) of pre-post (peak) alterations. CK: creatine kinase; Mb: myoglobin; AST: aspartate transaminase; ALT: alanine transaminase; LDH: lactate dehydrogenase; CMJ: counter-movement jump; HS: half squat; BP: bench press; PL_Total: PlayerLoad_TM total; PCP: pre-competition phase; CP: competition phase; ms: milliseconds; W.kg⁻¹: watts per kilogram per minute; kg: kilogram; g: g-force.

*Denotes significant (p < 0.005) pre-post difference in measure.

#Denotes significant (p < 0.005) difference compared to previous time-point.

Methodological quality

The methodological quality of 9 studies was scored 9 out of 11 (82%), with 14 studies scoring 8 (73%). The remaining 2 studies scored 7 (64%). Criteria 8 and 9 were unable to be determined in all studies, therefore no studies obtained the maximum score of 11.

Meta-analysis results

Six studies included BLa measurements pre–post boxing-specific activity (Davis, Wittekind, & Beneke, 2013; Finlay, Greig, McCarthy, & Page, 2020; Finlay, Greig, & Page, 2018; Hukkanen & Häkkinen, 2017; Nassib et al., 2017; Obminski, Stupnicki, Eliasz, Sitowski, & Klukowski, 1993). There were significant (P < 0.001; SMD = 4.62; CI 2.89, 6.36) increases in peak BLa post-boxing-specific activity when compared to baseline. Considerable heterogeneity was observed (I² = 90%, P < 0.001). Three studies analysed cortisol levels pre–post boxing-specific activity (Kaynar, 2019; Kilic et al., 2019; Obminski et al., 1993). There were significant (p = 0.002; SMD = 1.33; CI 0.50, 2.17) increases in cortisol post boxing-specific activity, compared to baseline. Substantial heterogeneity was observed (I² = 64%, P = 0.03). Five studies investigated muscle damage and inflammation levels pre–post boxing-specific activity, as quantified by CK, Mb, ALT and AST levels (Graham et al., 2011; Kaynar, 2019; Kilic et al., 2019; Zuliani et al., 1985). There were significant increases in CK (P = 0.0006; SMD = 0.65; CI 0.28, 1.01) and Mb (P < 0.0001; SMD = 1.43; CI 0.72, 2.15); and non-significant increases in both ALT (P = 0.43; SMD 0.97; CI −1.44–3.38) and AST (P = 0.07; SMD = 1.37; CI −0.14–2.87) post boxing-specific activity, compared to baseline. No heterogeneity (I² = 0%, P = 0.91), low heterogeneity (I² = 30%, P = 0.24) and considerable heterogeneity (I² = 95%; P < 0.0001; I² = 87%; P = 0.006) were observed for CK, Mb, ALT and AST, respectively. Two studies analysed counter-movement jump (CMJ) performance pre–post boxing-specific activity (Loturco et al., 2021; Nikolaidis et al., 2017). There were non-significant increases in CMJ height (p = 0.21; SMD = 0.33; CI 0.19, 0.85) post boxing-specific activity compared to baseline. No heterogeneity was observed (I² = 0%, P = 0.73).

Pre–post changes for each boxing-specific activity

Comparisons of the responses to each individual boxing-specific activity are available in Figure 2, as quantified by the combined effect sizes (hedges g).

Figure 2. Combined effect sizes of pre-post responses for each boxing-specific activity. Bla: blood lactate; CK: creatine kinase; AST: aspartate transaminase; ALT: alanine transaminase; SC: semi-contact; NM: neuromusuclar; CMJ: counter-movement jump. Time course range of post measures: BLa = immediately – 7 mins post; Myoglobin = 15 mins–16 hrs post; Cortisol = 3 mins–30 mins post; CK = 5 mins–16 hrs post; ALT/AST = 5 mins–10 mins; CMJ = 5 mins–1 hr.

Discussion

This review provides the most thorough analysis of the acute physiological, biochemical, endocrine and performance responses experienced by amateur boxers following competitive bouts, sparring, and simulated activity to date. Our findings show that amateur boxing-specific activity induces a considerable physiological, biochemical and hormonal response; however, task-specific or neuromuscular performance may remain intact, or indeed in some instances, improve. Competitive bouts consistently proved the more demanding activity, followed by sparring, and lastly simulated activity, with high intensity sparring more closely replicating the demands of competitive bouts.

Physiological and perceptual responses

Due to the combative nature and restrictions on data collection within boxing, physiological data obtained from competitive bouts are limited to pre–post BLa in only a few studies; (Hanon, Savarino, & Claire, 2015; Davis et al., 2013; Obminski et al., 2009). The large pre–post increases (ES = 8.75) compared to baseline highlight the critical role played by anaerobic glycolysis in actual competitive bouts (Davis et al., 2013), whereby boxers are required to be conditioned to tolerate an incremental increase in BLa across successive rounds (Chaabène et al., 2015; Smith, 2006). The current review found post-competition BLa values ranged from 11.8 ± 1.6 to 13.6 ± 2.4 mmol.l⁻¹. However, it is worth noting that muscle lactate values may exceed that of BLa (Knuttgen & Saltin, 1972), therefore the true metabolic cost of boxing-specific activity in the literature may still be underestimated. In some instances, higher BLa values were observed in sparring, when compared to studies on competitive bouts, highlighting its ability to somewhat replicate the demands of competition. However, the combined pre–post change was lower (ES = 5.03) than competitive bouts. Likewise, large yet inferior pre–post BLa alterations (ES = 2.41) were observed in simulated activity when compared to sparring, and more so in competitive bouts. Where boxing-specific simulation had induced a comparable or greater BLa response to sparring and competitive bouts, this had included invalid protocols such as maximum frequency, speed or power protocols (Ghosh, 2010; Lutoslawska, Eliasz, & Sitowski, 1995). Other factors that have influenced BLa values in boxers in the previous literature, include boxer weight category, training phase, and the environment (i.e. boxing gym or laboratory) where the activity is performed in (Arseneau, Mekary, & Léger, 2011; Hukkanen & Häkkinen, 2017; Khanna & Manna, 2006).

Researchers have also sought to quantify the HR and V̇O₂ responses to sparring and simulated activity, thus allowing for a deeper understanding of the cardiovascular and cardiorespiratory demands that are not quantifiable during competition. Alike the BLa response, this review shows a typical incremental increase in HR across successive rounds, mirroring the trend in other combat sports (Bridge, McNaughton, Close, & Drust, 2013; Campos, Bertuzzi, Dourado, Santos, & Franchini, 2012; Ouergui et al., 2016). Khanna and Manna (2006) reported that HRave in sparring equated to ∼93% of HRmax in round 1, increasing to 99% in round 3. Similarly, high HRave values have been found in other sparring scenarios and in simulated activity (de Lira et al., 2013; Hukkanen & Häkkinen, 2017; Nikolaidis et al., 2017). Similar to sparring, this work found both HR and RPE typically increased across subsequent rounds of simulated activity; however, consistently lower responses were observed in simulated activity compared to sparring. Oxygen uptake was seemingly consistent across rounds of boxing-specific simulations, corresponding to ∼59–69% of V̇O_2max. It is worth noting that there may often be a disconnect between the HR and V̇O₂ kinetics in combat sports, due to the stress response of actual combat (de Lira et al., 2013) and the work:rest ratios associated with the boxing-specific activity. Whilst no direct data exists in competitive bouts, de Lira et al. (2013) reported higher average (∼85–95%V̇O_2max) and peak (∼99–105%V̇O_2max) V̇O₂ in sparring, though this was estimated via the HR–V̇O₂ relationship. Nevertheless, one can assume that the V̇O₂ cost of actual bouts may be higher than that of simulated activity. Indeed, as with the BLa and HR response, average and peak V̇O₂ values in the literature are typically higher in semi-contact pad protocols with an opponent, when compared to punch bag protocols (Arseneau et al., 2011; Finlay et al., 2020; 2018; Thomson & Lamb, 2017). This is not surprising, as sport-specific simulations typically elicit a lower stress response when compared to the competition (Bridge et al., 2018). Unfortunately, there is no evidence of the use of RPE in competitive bouts and extremely limited use in sparring (Arseneau et al., 2011; Smith, 2006), despite it being a valid and cost-effective method of quantifying internal load (Slimani, Davis, et al., 2017). Alike the physiological response, the perceptual response to sparring may increase across rounds (Nikolaidis et al., 2017); however, more research is needed on the use of this method in sparring. The incremental increase across rounds is more evident in boxing-specific simulations (Finlay et al., 2020; Finlay, Greig, & Page, 2018; Halperin, Chapman, Thompson, & Abbiss, 2019; Thomson & Lamb, 2017).

The literature would suggest that sparring elicits a valid BLa response in comparison with competitive bouts, in addition to a high and incremental HR response similar to that which we may expect in actual bouts. Thus, sparring evidently represents an appropriate training modality in replicating the physiological demands of a competitive bout; however, this may still be somewhat reduced. The seemingly appropriate training stimuli provided by sparring must also be balanced against the inherent fatigue and injury risk associated with actual combat. In comparison, boxing-specific simulations produce a high, yet weakened physiological response compared to all other modes, likely due to the decreased stress response involved. More specifically, the greater combined physiological and psychological stressors involved in competition, perhaps due to the increased threat or risk of injury, can lead to the release of stress hormones via the hypothalamic–pituitary–adrenocortical axis (HPA) and sympathetic–adrenal–medulla (SAM) (Bridge et al., 2013). This is discussed further in the endocrine responses section. Boxing-specific simulations do offer a unique opportunity to monitor several acute physiological responses in a controlled manner, thus being a useful tool for the practitioner. Specifically, they could be used to obtain benchmark fitness data or used to assess the efficacy of interventions due to their standardised activity.

Endocrine responses

The physiological demands of amateur boxing-specific activity are accompanied by an endocrine/hormonal response. A previous review on the hormonal response to combat, but not boxing activity, found moderate to very large increases in cortisol (ES = 0.79); adrenaline (ES = 4.22); noradrenaline (ES = 3.40) and human growth hormone (ES = 3.69) when compared to baseline (Slimani et al., 2018).

Cortisol, a steroid hormone produced and secreted by the adrenal cortex via the HPA (Slimani et al., 2018), was the most frequently measured hormone found in this review. Large pre–post increases (ES = 1.33) in salivary or blood cortisol were found when all boxing-specific modes were considered, evidently greatest post-sparring (ES = 2.09), followed by competition (ES = 1.57), with minimal differences in one study that considered semi-contact sparring (ES = 0.01). However, results should be interpreted with caution due to the variation in the type of cortisol concentration and the time course associated with the measurements between studies. Whilst peak serum and saliva cortisol have been shown to be significantly correlated (Rc = 0.728; P = .001) (Van Bruggen, Hackney, McMurray, & Ondrak, 2011), peak salivary concentrations may appear 30-minutes post-exercise (Obminski et al., 1993), as opposed to serum cortisol which may appear immediately post-exercise. Previous research has shown that activity associated with greater stress, for example sporting competition or simulated bouts such as high-intensity sparring, results in elevated cortisol concentrations (Slimani et al., 2018), a finding reflected in the current review. This would inevitably influence the physiological response to competitive situations, as seen in Table 1. Chronic increases in cortisol may indicate high physical or psychological stress (Hug, Mullis, Vogt, Ventura, & Hoppeler, 2003; Kilic et al., 2019), impaired recovery, overreaching (Lee et al., 2017), or it may also reflect the natural time course of regular boxing training, as reported by Nassib et al. (2016). Future research should aim to integrate the monitoring of other hormonal markers alongside cortisol (Kilic et al., 2019) to assess the health and potential overreaching or overtraining status of the amateur boxer (Lee et al., 2017). Compared to what is known from a physiological standpoint, research on the endocrine response to boxing-specific activity is in its infancy. Nevertheless, the small body of research suggests boxers experience a substantial hormonal demand.

Biochemical responses

This review was also concerned with the muscle damage and inflammation responses associated with amateur boxing-specific activity. Two common biomarkers used by practitioners for the determination of muscular or metabolic stress are CK and Mb (Lee et al., 2017; Urhausen, Gabriel, & Kindermann, 1995). This review found moderate (ES = 0.65) and large (ES = 1.43) pre–post increases in CK and Mb following boxing-specific activity. Likewise, large pre-post alterations in other commonly used markers of skeletal muscle damage, such as ALT (ES = 0.97) and AST (ES = 1.37), have been found up to 10-minutes following boxing-specific activity. This demonstrates that amateur boxers may experience considerable muscle tissue damage, trauma or injury, in addition to inflammation. When comparing the individual boxing-specific activities, CK and Mb pre–post changes were the greatest in sparring (ES = 0.78; 1.60) and competitive bouts (ES = 0.65; 1.47) whilst CK levels following the more restrictive semi-contact sparring were lower (ES = 0.57). A similar trend was found in the inflammation data, whereby large pre–post increases in ALT (ES = 2.21) and AST (ES = 2.14) were found post competitive bouts. This suggests sparring activity induces a considerable amount of damage, inflammation and potential injury to the amateur boxer, which in the case of muscle damage, may even surpass that obtained in a competitive bout. Unfortunately, no studies quantified CK or Mb concentrations post boxing-specific simulations. Graham et al. (2011) found lower CK levels in boxers following semi-contact sparring when compared to full sparring inclusive of punches to the head. This could suggest that exposure to greater levels of head punches, or blunt force trauma, accounted for greater levels of muscle damage, as has been proposed in MMA research (Wiechmann et al., 2016). Future research could utilise reliable and sensitive head trauma biomarkers such as serum neurofilament light chain (NFL) to assess this in a boxing cohort and to differentiate from more general markers such as CK (Shahim, Zetterberg, Tegner, & Blennow, 2017). Research in MMA, which shares some similar physical and technical requirements to striking in boxing, suggests CK concentration may peak at 24 hrs post activity (Lindsay et al., 2017; Tabben et al., 2018; Wiechmann et al., 2016). Whilst CK values were generally higher after several hours, no studies monitored this at a post 24 hrs time point, which may potentially underestimate the peak CK response. There was evidence of elevated CK at 12 and 16 hrs post competitive bouts (Bianco et al., 2005; Zuliani et al., 1985). Amateur boxers are exposed to considerable pre–post alterations in muscle damage and inflammation, highlighting the importance of recovery strategies.

A greater emphasis on the quantification of muscle damage biomarkers is required to further understand the demands placed on the amateur boxer and to bring this in line with the substantial research on the physiological demands.

Performance responses

The above responses may induce fatigue, and thus may be expected to influence the neuromuscular and task-specific performance of athletes. Interestingly, the pre–post changes in neuromuscular and task-specific performance in the current review, albeit based on a small body of research, would suggest performance is not negatively affected. Specifically, Loturco et al. (2021) found that CMJ height and upper-and-lower-body power output were maintained following a competitive bout. Indeed, Nikolaidis et al. (2017) observed CMJ improvements following a 3 × 3 simulated boxing bout in Greek amateur boxers, when compared to baseline values. The latter was attributed to the sport comprising predominantly upper-body activity; however, we know that the lower extremities contribute a considerable amount to punch technique (Stanley, Thomson, Smith, & Lamb, 2018) whilst lower-body strength and power are highly associated with punch force (Dunn, Humberstone, Franchini, Iredale, & Blazevich, 2022; Loturco et al., 2016). It should be noted that differences in the timings of the CMJ tests (1 hr prior and immediately post) and (5–10 mins prior and 5–10 mins post) in the studies of Loturco et al. (2021) and Nikolaidis et al. (2017) respectively, means the results should be interpreted with caution. Non-boxing-specific lower-body exercise has been found to negatively influence subsequent punch force performance (Dunn, Humberstone, Franchini, Iredale, & Blazevich, 2021) whilst a negative effect may not materialise following boxing-specific activity. Hukkanen and Häkkinen (2017) found punch force increased by 7.7% from baseline to the 3rd round of sparring in amateur boxers during their pre-competition phase. Punch force also remained consistent from baseline to the 3rd round during the boxers’ competition phase trials. In the same study, reaction time lengthened across subsequent rounds of sparring in the pre-competition phase. An opposite trend was apparent in the competition phase however, whereby the shortest reaction times were observed in the 3rd round, when compared to the 1st round. This suggests that whilst the sparring load may have had a slightly negative effect on reaction times in some instances, it did not impair punch performance. Indeed, in the case of the competition phase, the punch force improved round-on-round. The authors noted that the differences between the two phases could be due to the boxers’ exposure to maximal strength and power training during the pre-competition phase. Another possible reason could lie in a warm-up or a post-activation performance enhancement (PAPE) effect of the earlier activity, but this is speculative. The difference in body mass of the boxers between the two phases and the potential impact this may have on punch force and reaction time were not clear. Elite amateur boxers exhibited slight improvements in punch force across rounds of a 3 × 2-minute punching simulation with differing attention focus strategies (Halperin et al., 2019), further evidencing a lack of performance decrement.

Thomson and Lamb (2017) showed that amateur boxers’ punch accelerations increased slightly across rounds when performed to an externally valid pad-based boxing simulation (BOXFIT). Likewise, external load as quantified by a tri-axial accelerometer (PlayerLoad_TM) generally increases across rounds of boxing-specific simulation (Finlay et al., 2018, 2020), whereby a PL_total·min⁻¹ ranging from 13.66 to 16.82 au represents a comparable external demand to that of MMA simulation (∼14.91 au) (Kirk, Hurst, & Atkins, 2015). The aforementioned examples on neuromuscular and task-specific performance could suggest that the familiarity of the motor task and the short-duration of boxing-specific activity included in the review (6–9 minutes) result in little to no performance decrements across rounds, and pre–post activity. The longer duration of professional boxing may provide interesting comparisons. The lack of performance detriment found in the current review may also indicate that the amateur boxers observed in the studies were sufficiently conditioned to cope with the demands of the activity.

Conclusion

This review found that amateur boxing-specific activity places a considerable physiological, endocrine, and biochemical demand on amateur boxers. Such demands are seemingly not detrimental to neuromuscular or task-specific performance, as evidenced by the maintenance of, and in some cases, improved performance post boxing-specific activity. Unsurprisingly, competitive bouts and sparring consistently induced greater acute responses compared to externally valid boxing simulations. The use of boxing-specific simulations is a paradox, whereby despite the opportunity of increased mechanistic rigour in scientific investigations, researchers are aware that acute responses are consistently lower than that elicited in actual combat. Sparring may provide the most valid representation of competitive bouts, though the inherent risks of actual combat, such as cuts, concussion and other injuries, exist. The findings of this review may assist in the periodisation of boxing-specific exercise modes dependent on the desired physiological adaptations, and for recovery management in relation to the endocrine, biochemical,and performance responses. For example, high-intensity sparring might be reserved for later phases of preparation, in which the priority is to replicate the demands of competition as closely as possible. Furthermore, though beyond the scope of this review, the accumulation of head impacts and the potential for concussion in activities such as sparring close to competition, must also be considered. Implementing 'less demanding' boxing-specific activity (such as bag and/or pad work protocols) in the early stages of preparation, or indeed during a tapering phase prior to competition once a cessation of sparring has occurred, may be useful as conditioning tools. Manipulation of the intensity of this activity may provide an appropriate stimulus at select phases of preparation, whilst avoiding an accumulation of fatigue, overreaching, or overtraining. Future research, regardless of boxing-specific mode, should aim to analyse the acute responses to boxing-specific activity more holistically. This may also be extended to the monitoring of longer durations or applying this to repeat-bout scenarios, replicating domestic or Olympic level amateur boxing tournaments.

Disclosure statement

No potential conflict of interest was reported by the authors.