Exercise-Induced Muscle Damage During the Menstrual Cycle: A Systematic Review and Meta-Analysis : The Journal of Strength & Conditioning Research

Secondary Logo

Journal Logo

Brief Review

Exercise-Induced Muscle Damage During the Menstrual Cycle: A Systematic Review and Meta-Analysis

Romero-Parra, Nuria; Cupeiro, Rocío; Alfaro-Magallanes, Victor M.; Rael, Beatriz; Rubio-Arias, Jacobo Á.; Peinado, Ana B.; Benito, Pedro J.;  on behalf of the IronFEMME Study Group

Author Information
Journal of Strength and Conditioning Research 35(2):p 549-561, February 2021. | DOI: 10.1519/JSC.0000000000003878
  • Free


Romero-Parra, N, Cupeiro, R, Alfaro-Magallanes, VM, Rael, B, Rubio-Arias, JA, Peinado, AB, and Benito, PJ, IronFEMME Study Group. Exercise-induced muscle damage during the menstrual cycle: A systematic review and meta-analysis. J Strength Cond Res 35(2): 549–561, 2021—A strenuous bout of exercise could trigger damage of muscle tissue, and it is not clear how sex hormone fluctuations occurring during the menstrual cycle (MC) affect this response. The aims of this study were to systematically search and assess studies that have evaluated exercise-induced muscle damage (EIMD) in eumenorrheic women over the MC and to perform a meta-analysis to quantify which MC phases display the muscle damage response. The guidelines of the Preferred Reported Items for Systematic Reviews and Meta-Analysis were followed. A total of 19 articles were analyzed in the quantitative synthesis. Included studies examined EIMD in at least one phase of the following MC phases: early follicular phase (EFP), late follicular phase (LFP), or midluteal phase (MLP). The meta-analysis demonstrated differences between MC phases for delayed onset muscle soreness (DOMS) and strength loss (p < 0.05), whereas no differences were observed between MC phases for creatine kinase. The maximum mean differences between pre-excercise and post-exercise for DOMS were EFP: 6.57 (4.42, 8.71), LFP: 5.37 (2.10, 8.63), and MLP: 3.08 (2.22, 3.95), whereas for strength loss were EFP: −3.46 (−4.95, −1.98), LFP: −1.63 (−2.36, −0.89), and MLP: −0.72 (−1.07, −0.36) (p < 0.001). In conclusion, this meta-analysis suggests that hormone fluctuations throughout the MC affect EIMD in terms of DOMS and strength loss. Lower training loads or longer recovery periods could be considered in the EFP, when sex hormone concentrations are lower and women may be more vulnerable to muscle damage, whereas strength conditioning loads could be enhanced in the MLP.


Scientific knowledge regarding female training responses and adaptations lags behind the body of research on men (15) despite findings reporting differences between the sexes (1,40,64). For instance, women present less anti-inflammatory gene regulation (1) and less fatigability after isometric contractions in comparison with their male counterparts (40). In addition, based on body mass, women must work harder to perform less total work, and they are less capable to complete an upper-body exercise protocol than men (64). Thus, it cannot be assumed that sex differences either do not exist or are irrelevant beyond the reproductive system (4). In fact, the menstrual cycle (MC) effects on training responses, adaptations, and performance remain unclear because most studies including women have tested them in the “low-hormone” phase of the MC, when the hormonal profile is most similar to that of men (58). The effect of such hormonal changes on exercise responses should be investigated because of the increasing number of women engaging in sports and training programs (58).

The MC is a sequence of circamensal rhythms domineered by the feedback loops within the hypothalamus-pituitary-ovarian axis (19,34,67). The entire cycle takes approximately 28 days, with an interindividual range of regular MCs of 21–35 days in healthy adult women (49), and it involves repetitive cycles of follicle development, ovulation, and preparation of the endometrium for possible implantation of an embryo (32). The signaling process begins in the hypothalamus with a pulsatile secretion of gonadotropin-releasing hormone (GnRH) that travels through the pituitary portal venous system to the anterior pituitary gland (33) and stimulates the synthesis and pulsatile release of luteinizing hormone (LH) and follicle-stimulating hormone (FSH). These 2 gonadotropins stimulate the secretion of sex steroid hormones estradiol β-17 and progesterone in the ovaries (33), which are the main contributors to the reproductive function in women (19). The endocrine communication between the aforementioned hormones and glands determine the 2 major phases of the MC, separated by mid-cycle ovulation: the follicular phase, which is focused on maturing a reproductive cell (19), and the luteal phase, which is characterized by the formation of the corpus luteum and its regression (19,32,33). In the absence of fecundation, the endometrium loses its supply and degenerates, sloughing off with menstrual bleeding (32), which is considered the beginning of the MC. In the early follicular phase (EFP), FSH triggers rapid proliferation of cells and follicle growth (34). The follicle with the greatest number of granulosa cells and hence the highest sensitivity to FSH will be the dominant follicle, which is the first to increase its ability to aromatize androgens (19,28,32,34). With the development of aromatase activity, estrogen levels gradually rise halfway through the midfollicular phase (MFP), producing a high-estrogen microenvironment and, as a result, downregulation of FSH from the anterior pituitary (19,32,34). Estradiol concentration keeps increasing during the late follicular phase (LFP), reaching its maximum at around day 14 of the cycle (43). The high level of estrogen triggers positive feedback in the anterior pituitary that stimulates the midcycle surge of LH (and a little spike in FSH as a side effect) (32–34). This occurs approximately 14–26 hours before ovulation (43), and it is necessary for final follicular growth. Furthermore, LH surge causes weakening of theca externa, contributing to follicle rupture, ovulation, and subsequent conversion of the remaining granulosa and theca cells into mainly progesterone-secreting cells. Hence, the follicle becomes the corpus luteum (34). This organ secrets increasing amounts of progesterone (33) while the overall concentration of estradiol β-17 decreases after LH surge (19) to moderately rise again in the midluteal phase (MLP), when progesterone reaches its peak. Therefore, secretion of both hormones provide negative feedback effects on the anterior pituitary gland and the hypothalamus (33), but the drop of both hormones during the late luteal phase (43) provides the necessary feedback to stimulate the hypothalamus and start a new cycle.

The influence of the MC and its underlying changes in the hormonal environment should therefore be considered when exercise and performance are evaluated in women. Accordingly, some findings have demonstrated that the aforementioned hormones affect the cardiovascular system, ventilation, thermoregulation, injury mechanisms, and substrate metabolism in humans (14). Heart rate, ventilation, and temperature increased during the luteal phase, indicating higher cardiorespiratory strain in this phase (3). Further anabolic effects of estrogens have been observed in comparison with the catabolic effect of progesterone or the major insulin sensitivity triggered by estrogens in comparison with the insulin resistance elicited by progesterone (61). In addition, studies with female animal models have provided evidence that estrogen may influence muscle membrane stability and limit exercise-induced muscle damage (EIMD) (26,77). Exercise-induced muscle damage is a process that takes place when muscle is exposed to strenuous and unaccustomed exercise (47). The high mechanical strain, especially produced by eccentric actions, elicits myofibrillar disruption resulting in membrane damage, which is exacerbated by free radicals present in metabolically active tissues (47,63,65). As a consequence, uncontrolled calcium release into myofibrillar cytosol enhances not only a coupling dysfunction in the cross-bridge cycle but also the action of some proteolytic enzymes that could promote muscle protein degradation (47,65). Some of the consequences derived from this process are the loss of muscle strength, delayed onset muscle soreness (DOMS), decrease in the range of motion (ROM), increase in limb girth, increase in systemic leakage of muscle enzymes and proteins (e.g., creatine kinase [CK] and myoglobin), or a combination of these (22,47,63).

Estrogens have been proposed to act as antioxidants against lipid peroxidation of the cell membrane (66) due to the similarities in their structure with tocopherol, which is known as an antioxidant. Thus, they may facilitate the donation of hydrogen atoms, leading to limitation of peroxidation chain reactions and therefore contributing to membrane stability (66). Accordingly, estrogen replacement therapy has demonstrated a beneficial effect on limiting muscle disruption and inflammation after damaging exercise in postmenopausal women (21). However, the role of endogenous estrogen in this process in female human models has not been clearly elucidated. This inconclusiveness in the literature may be explained by the lack of intrasubject study designs, reporting results for women and men together (9,25) or measurements performed in only one phase of the MC, not covering the complete hormonal environment of subjects. Even data from eumenorrheic and oral contraceptive users have been reported together (79), although the hormonal environment is quite different between these groups. Finally, inconsistencies could also be due to conflicting definitions of “reproductive status,” inaccurate consideration and verification of MC phases (36,44,54,56), different modalities of exercise performed, or the recruitment of nonhomogenous groups of subjects with different training status (14,24). Therefore, to clarify the influence of sex hormone environment on EIMD, the objectives of this study were (1) to systematically review the existing literature evaluating EIMD in women considering the MC, (2) to perform a meta-analysis to quantify the effect of the MC phases on muscle damage response, and (3) to identify the MC phases that are more sensitive to muscle damage (43).


Experimental approach to the problem

The methodological process was based on the recommendations indicated by the guidelines of the Preferred Reported Items for Systematic Reviews and Meta-Analysis (50).The eligibility criteria were established by the authors. The systematic review registration number PROSPERO 2019 is CRD42018110290.

A comprehensive database search was systematically conducted (PubMed and Web of Science) up to February 12, 2020, and performed independently by 2 authors (N.R.P. and P.J.B.) obtaining the same results. The search strategy included variations on terms related to muscle damage, MC, and exercise. The complete search term combinations are included as a Supplemental Digital Content 1 (see Supplementary file, https://links.lww.com/JSCR/A240) to this article.

The specific inclusion criteria were (1) studies reporting information on muscle damage markers in at least one specific MC phase (EFP, LFP, or MLP because these show the most pronounced changes in hormonal environment) on EIMD and (2) articles with full text of original research published in peer-reviewed scientific journals in English. Any exercise modality aimed to elicit muscle damage was included, but it was classified into low-damaging or high-damaging exercise, according to the amount of EIMD produced (63), fully explained below. Research studies were excluded if they were (1) consisting of women as subjects without reporting MC phase; (2) studies with only men as subjects or reporting male and female outcomes as the same group; (3) based on populations with severe or chronic diseases that may require medical treatment; (4) studies with premenarchial, menopausal women, or oral contraceptive users; (5) studies with animals; (6) published reviews, conference communications, opinion articles, commentaries, book chapters, case studies, or presentations; (7) interventional or longitudinal studies except those reporting an acute response; or (8) based on muscle fatigue measurement without muscle damage markers.


The initial search, which analyzed the influence of MC on muscle damage, identified 3,736 articles from the databases and 13 articles from other sources. After removing duplicates, 3,198 articles were screened for eligibility on the basis of their title and abstract, with 2,278 subsequently excluded. The list of retrieved articles was screened independently by 2 authors (N.R.P. and P.J.B.) to choose potentially relevant articles. A total of 920 were assessed as full texts, and 26 studies were included in qualitative synthesis, excluding 894 articles on the basis of the aforementioned exclusion criteria. Disparities found were discussed by the authors to reach a consensus. From these, 7 studies were excluded because variables did not correspond to the final outcome variables analyzed (DOMS, CK, and strength loss), which are fully explained below, or the phases of measurement do not correspond to the EFP (days 2–7), LFP (days 9–13), or MLP (days 18–24) of the MC (43). In this regard, 2 criteria were used to classify the MC phases of measurement: (1) the indication in the article of the phases/days of measurement or (2) the inclusion of blood hormone concentration, or both criteria. When blood hormone concentrations provided in a study disagreed with those expected for the days of measurement indicated (71), this measurement was relocated in the adequate phase to ensure that the studies meet the criteria of low estrogen and progesterone (EFP), high estrogen (LFP), and high estrogen and progesterone (MLP). A total of 19 studies were included in the quantitative synthesis (meta-analysis). The flow diagram of the search process is shown in Figure 1.

Figure 1.:
Flowchart demonstrating the step-by-step process of articles elimination to find the final studies to be included in the meta-analysis.

Risk of Bias Assessment

The methodological quality of the selected studies was assessed with the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies (35) that analyses the following factors: (1) study question clearly stated; (2) study population well defined; (3) all eligible subjects enrolled; (4) eligibility criteria clearly specified; (5) sample size estimation provided; (6) exposure(s) of interest measured before the outcome measurement; (7) sufficient timeframe to see an association between exposure and outcome; (8) different level of exposure examined; (9) exposure measures clearly described, valid, and reliable; (10) exposure assessed more than once over time; (11) outcome measures clearly described, valid, and reliable; (12) blinding of outcome assessors; (13) follow-up rate; and (14) confounding variables measured and adjusted.

Data Extraction and Outcomes

From the included studies, three authors (R.C., B.R., and V.M.A.M) extracted to a previously designed data sheet the following information was collected: authors, study design, number of subjects, age, height, body mass, training status and training experience, MC phases and the corresponding days of measurements when protocols were performed, hormone concentrations in each phase when provided, method to determine and verify phases, modality of exercise (e.g., endurance, resistance…), and load (e.g., sets and repetitions or duration, rest between sets, and intensity). The outcome variables for muscle damage were DOMS; blood biomarkers such as CK, myoglobin, interleukin-6, and lactate dehydrogenase; strength loss; ROM; circumferences; and countermovement jump (CMJ). Delayed onset muscle soreness is defined as the sensation of muscle discomfort during muscle contractions 24–48 hours after intense exercise (7). It was obtained from questionnaires providing soreness perception in terms of a numerical scale, usually from 0 (no pain at all) to 10 or 100 (unbearable pain) (7). Some of the included studies used 10-point scales, whereas others used a 100-mm visual analogue scale (5); hence, all the studies evaluating DOMS were unified to the same scale (10-point) to facilitate calculations and statistics. Regarding these variables, the information extracted was the timing of measurement (baseline, 0, 24, 48, and 96 hours post-exercise; minimum one post-exercise time point), the units of measurement, and the technique performed. Some authors were contacted to complete missing data regarding the main outcomes. N.R.-P. and P.J.B. reviewed this data sheet.

Statistical Analyses

The meta-analysis and the statistical analysis were conducted using the Review Manager software (RevMan 5.2; Cochrane Collaboration, Oxford, United Kingdom). The effects of exercise on muscle damage outcomes in each phase of the MC were expressed as the standardized mean differences (SMDs) between baseline and post-exercise measurements and their confidence intervals (CIs), which were analyzed using a random-effects model (39). Each mean difference was weighted according to the inverse variance method (20). Heterogeneity between studies was evaluated through I2 statistics and the interstudy variance using the tau-square (Tau2). A Tau2 >1 indicated substantial statistical heterogeneity. The chi-square was defined as the indicator of heterogeneity presence. A scale of magnitude was implemented for the interpretation of heterogeneity of the results, where <25% was assessed as low magnitude, between ≥25% and ≤75% was set as medium magnitude, and >75% was considered high magnitude (39).The publication bias was statistically evaluated through an asymmetry test as estimated from a funnel plot. In addition, Egger's test was also used to assess publication bias. A p value of less than 0.05 was considered to be statistically significant.

Supplementary analyses were conducted to determine the effects of exercise modality on the heterogeneity of muscle damage outcomes (confounding variables) attending to the severity of muscle damage. In each study, the SMD (CI 95%) of the EIMD severity was calculated as the difference between baseline and post-exercise measurements for DOMS, CK, and strength loss. Likewise, the estimation of effects was calculated according to the inverse variance method (20). Three authors independently scored the exercise protocols according to the amount of EIMD produced following criteria previously described (63). Each protocol was assessed considering (1) the predominant component during contractions, rating as “1” eccentric and isometric contractions at long muscle lengths and “0” concentric or isometric contractions at short muscle lengths; (2) the exercise performance at long vs. short muscle lengths, rated as “1” or “0,” respectively; (3) the exercise performance of single joint vs. multiple joints exercises rated as “1” or “0,” respectively; (4) the exercise using arms vs. legs, rated as “1” or “0,” respectively; and (5) the use of knee flexors vs. knee extensors rated as “1” vs. “0,” respectively. Peake et al. (2017) also considered other 2 criteria to describe the amount of EIMD produced by an exercise: a high or low number of eccentric contractions and high or low eccentric torque. However, these 2 criteria were not applicable to all the studies included in the meta-analysis, and thus, they were not included and measured to obtain our classification. The exercise protocols were classified into the following 3 categories: light, moderate, or severe muscle damage (Table 1). The results from this classification are provided in the Supplemental Digital Content 1 (see Supplementary file, https://links.lww.com/JSCR/A240).

Table 1 - Description of groups and studies included.*
Study Cycle phase/days Hormone concentrations
E2 (pg/ml)
P4 (ng/ml)
N Age (y) Mass (kg) Height (cm) Training status Training experience EIMD protocol description Classification attending to damage severity
Anderson et al. (2) EFP
2 ± 1
E2 115.0 ± 82.9 9 25 57.2 162 Able to achieve an electrically induced force output of ≥15% of a MVIC Not involved in consistent, structured resistance or aerobic training 6 mo before the study 100 left knee extensions with neuromuscular electrical stimulation (NMES) (80 pps frequency, 300 μs pulse duration, 5:10 s on:off time, and 1.5 s ramp-up) at ≥25% MVIC. Moderate
22 ± 3
E2 140.2 ± 85.7 8 24 61.8 163
Brown et al. (8) MLP NR 12 20 57.4 162 Recreational dancers 14 ± 3 y; 324 ± 188 min·wk−1 10 × 1 min movement phrase of contemporary dance. Each phrase movements includes jumps, rolls to the floor, and weight transferences. Each phrase was separated by 2 min rest. Light

12 19 60.0 166 15 × 30-m sprints with a rapid 10 m deceleration phase, every 65 s.
Chaffin et al. (10) EFP
E2 68.4 ± 28.7
P4 8.7 ± 3.4
9 27 57.2 166 Competitive, physically active <40.2 km·wk−1 running Two rounds of: 3× (2 min running at 100% vV̇o 2peak with 2 min jogging or walking at 50% V̇o 2peak) with 3 min walk between rounds; 5 min recovery at a self-selected pace; 5× (1 min running at 110% vV̇o 2peak + 3 min walk). Light
E2 85.8 ± 14.8
P4 13.7 ± 2.5
Hackney et al. (30) MFP
8 ± 2
E2 23.8 ± 6.6
P4 1.1 ± 0.4
8 25 58.9 168 Competitive, physically active 4–5 d·wk−1 for 45–120 min 90 min running on a treadmill at 70% V̇o 2max. Running speed was 13.0 ± 0.5 km·h−1 Light
23 ± 3
E2 134.3 ± 38.8
P4 6.7 ± 3.9
Hicks et al. (37) LFP
day 14
NR 11 21 63.0 165 Recreationally active <1 h·wk−1 physical activity 6 sets of 12 reps of eccentric knee extensions at MVEC and 2 min recovery between sets. Knee ROM was set at 20°–90° (0° = full extension). Eccentric phase at an isokinetic angular velocity of 30°·s−1. The concentric phase was performed at an angular velocity of 60°·s−1. Moderate
Hicks et al. (38) LFP
day 14
E2 433 ± 147 9 21 63.2 165 Recreationally active <1 h·wk−1 physical activity 6 sets of 12 reps of eccentric knee extensions at MVEC and 2 min recovery between sets. Knee ROM was set at 20°–90° (0° = full extension). Eccentric phase at an isokinetic angular velocity of 30°·s−1. The concentric phase was performed at an angular velocity of 60°·s−1. Moderate
Jang et al. (42) LFP NR 6 21 55.9 162 Recreationally active <1 h·wk−1 physical activity A single bout of 30 min of bench stepping at 60 steps per min at a step height of 110% of the lower leg length. Light
Joyce et al. (45) EFP
NR 9 22 63.4 170 Physically active 1–3 d/wk for 30 min 240 MVEC of knee flexors: 4 sets of 6×10 with each leg at an isokinetic angular velocity of 30°·s−1 (24 total sets per leg). 1 min rest between sets. Severe
Keane et al. (46) MLP NR 11 22 62.7 167 Premier or national league players Rugby, soccer, and netball 15 × 30-m maximal sprints departing every 65 s. Moderate
Mackay et al. (53) EFP
E2 3.0 ± 1.1
P4 0.01 ± 0.04 (salivary)
10 29 67.4 163 Not involved in consistent, structured resistance or aerobic training 6 mo before the study 30 min of eccentric cycling on a recumbent ergometer with a motor moving the cranks backward at a cadence of 60 rpm, maintaining 90% POmax Moderate
14 ± 2
E2 3.8 ± 1.9
P4 0.04 ± 0.05 (salivary)
Minahan et al. (55) EFP
36.0 ± 11.4 8 22 62.3 169 Habitually active Moderate-intensity endurance activities (walking, jogging, and swimming) for 30 min; 2–3 d·wk−1 240 MVEC of knee flexors: 4 sets of 6 × 10 with each leg at an isokinetic angular velocity of 30°·s−1 (24 total sets per leg). 1 min rest between sets. Severe
Nikolaidis et al. (59) MLP
2–6 after ovulation
NR 12 23 54.0 165 Healthy <3 h·wk−1 of low-intensity leisure activities (jogging, swimming, and dancing) and no experience in eccentric exercise 6 mo before 5 sets of 15 MVEC of knee flexors at an angular velocity of 60°·s−1 in the prone position. 2 min rest between sets. Severe
Oosthuyse et al. (60) EFP
4 ± 1
E2 45.2 ± 24.2
P4 0.7 ± 0.3
5 21 55.6 158 Sedentary <2 h·wk−1 movements associated with everyday life 20 min running on treadmill at 9 km·h−1 at a −10% gradient. Speed was maintained constant for all subjects. Moderate
14 ± 3
E2 269.9 ± 110.8
P4 1.3 ± 0.5
5 23 57.6 162
20 ± 2
E2 136.2 ± 67.2
P4 8.1 ± 2.6
5 21 60.8 162
Pal et al. (62) LFP
NR 22 16 44.8 154 Sedentary Not take part in any physical conditioning program Running at 80% of HRmax until onset of fatigue as indicated by volitional exhaustion Moderate
Savage et al. (68) EFP
day 3
E2 31.5 ± 11.2 14 21 63.4 165 Sedentary Not take part in any physical conditioning program 2 × 25 eccentric contractions (3 s each) on the nondominant arm with 12 s rest between contractions and 5 min rest between sets. Subjects were seated at a preacher curl modified by the addition of an attachment bar as a lever arm to force the extension of the subject's arm while they attempted to resist it from being lowered. Severe
Sipavičienė et al. (70) EFP
E2 32.4 ± 16.8 18 20 56.2 167 Physically active 100 maximum drop jumps from a height of 0.75 m platform with an immediate maximum rebound on a Kistler force plate. Jumps were performed with a countermovement to a 90∘ angle in the knee at an interval of 20 s. Moderate
1–2 of ovulatory phase
E2 62.5 ± 22.3
Thompson et al. (74) MLP
6–8 after ovulation
E2 116.0 ± 18.42 6 25 63.8 176 Recreationally active 2.7 ± 0.8 h·wk−1 of cardiovascular activity 50 min bench stepping exercise at a cadence of 70 b·min−1 and a height equal to 110% of each individual's leg length, within 10% of their age-predicted max HR. Light
Williams et al. (80) MFP
7 ± 2
E2 39.8 ± 18.3 10 21 61.3 164 Highly trained 3–5 d·wk−1, 45–120 min·session−1 of aerobic activity 60 min of treadmill running at 65% of V̇o 2peak Light
23 ± 3
E2 148.1 ± 35.2
Wolf et al. (81) EFP
NR 7 22 65.6 169 Resistance-trained 4.7 ± 4.4 y or resistance training, at least 6 mo 6 sets of 5 squats at 90% of their 1 RM with 3 min rest between sets. Moderate
*EIMD = exercise-induced muscle damage; EFP = early follicular phase; MLP = midluteal phase; E2 = estradiol; NR = not reported; MVIC = maximal voluntary isometric contraction; P4 = progesterone; LFP = late follicular phase; MFP = midfollicular phase; MVEC = maximal voluntary eccentric contraction; ROM = range of movement; RM = repetition maximum. V̇o2max = maximal oxygen uptake; V̇o2peak = peak oxygen uptake; POmax = maximal concentric power output.


Subjects and study characteristics are summarized in Table 1. Of the 19 included studies, 12 tested during one phase of the MC alone. Four of these conducted testing in the EFP (45,55,68,81), 4 during the LFP (37,38,42,62), and 4 during the MLP (8,46,59,74). The remaining 7 studies compared between MC phases. Two of these compared between EFP and LFP (53,70), 2 between EFP and MLP (2,10), 2 between MFP and MLP (30,80), and one compared between EFP, LFP, and MLP (60). However, the results of the 2 studies testing during the MFP were allocated to the EFP according to sex hormone concentrations. In addition, in these 2 studies, the MFP was clearly stated as the one with low hormone concentrations (30,80).

Studies were classified according to the variable analyzed—(a) DOMS, (b) CK, and (c) strength—which were separated into 3 subgroups corresponding to the aforementioned phases of the MC: EFP, LFP, and MLP.

Data regarding other muscle damage and inflammation markers such as lactate dehydrogenase, myoglobin, limb circumference, ROM, CMJ, and interleukin-6 were not included in the quantitative analysis because the number of studies evaluating these variables' response throughout the MC was small, especially when different cycle phases were analyzed separately; thus, the subgroup analysis was not possible.

Risk of Bias Assessment and Heterogeneity

On the basis of the aforementioned quality assessment tool for the risk of bias, where each item was rated up to 10 points, the average sum of quality criteria was 9.46, so quality assessment was classified as “good” as the studies met the criteria with a high percentage. The lowest percentages were observed in items regarding the existence of confounding variables, blinded assessors to the exposures, or different exposure levels. In addition, the funnel plots suggested the presence of significant publication bias (Figure 2), whereas the results of Egger's test showed a significant heterogeneity among studies when comparing differences between pre-exercise and peak values for DOMS (Z = 9.98, p < 0.001, Figure 2A), CK (Z = 4.56, p < 0.001, Figure 2B), and strength (Z = 2.02, p = 0.044, Figure 2C).

Figure 2.:
Funnel plots for the meta-analysis of published studies for DOMS (A), CK (B), and strength loss (C). Each dot represents the SE and standardized mean difference (SMD) between pre-exercise and post-exercise maximum value of each study. Vertical dashed line represents mean SMD. Dots distribution shows a deviation from a pyramid shape suggesting higher publication bias and hence, the need to investigate possible causes. CK = creatine kinase; DOMS = delayed onset muscle soreness.

Effects of the Menstrual Cycle Phase on Delayed Onset Muscle Soreness

Our results indicated differences between phases from 24 to 72 hours post-exercise (Table 2). The lowest mean value in DOMS was observed at the MLP for all post-exercise time points, whereas the highest was observed at the EFP. A significant increase in DOMS during all post-exercise time points in comparison with baseline was observed in all phases.

Table 2 - Effect of menstrual cycle phases on muscle damage (DOMS, CK, and strength) at different post-exercise time points.*
Phases Studies (n) Subjects (n) SMD (95% CI) Test for overall effect Z (p) I2% (p) Differences between phases Differences between phases
 0 h
  EFP 4 42 2.46 (1.21 to 3.71) Z = 3.85 (p < 0.001) 75% (0.007) χ2 = 4.11; p = 0.13 χ2 = 2.39; p = 0.30
  LFP 1 5 1.89 (0.26 to 3.53) Z = 2.24 (p = 0.02) N/A
  MLP 11 114 1.18 (0.79 to 1.58) Z = 5.90 (p < 0.001) 42% (0.07)
 24 h
  EFP 5 61 7.25 (3.21 to 11.28) Z = 3.52 (p < 0.001) 94% (<0.001) χ2 = 7.62; p = 0.02 χ2 = 8.34; p = 0.02
  LFP 3 33 4.45 (0.44 to 8.45) Z = 2.18 (p = 0.03) 94% (<0.001)
  MLP 8 93 2.01 (1.44 to 2.58) Z = 6.87 (p < 0.001) 57% (0.02)
 48 h
  EFP 8 86 7.79 (4.80 to 10.78) Z = 5.11 (p < 0.001) 92% (<0.001) χ2 = 12.60; p = 0.002 χ2 = 11.36; p = 0.003
  LFP 5 53 6.94 (3.07 to 10.80) Z = 3.51 (p < 0.001) 94% (<0.001)
  MLP 12 123 2.87 (1.92 to 3.83) Z = 5.89 (p < 0.001) 84% (0<0.001)
 72 h
  EFP 5 61 5.79 (2.43 to 9.15) Z = 3.38 (p < 0.001) 94% (<0.001) χ2 = 7.38; p = 0.02 χ2 = 7.20; p = 0.03
  LFP 3 33 4.60 (0.10 to 9.09) Z = 2.01 (p = 0.04) 95% (<0.001)
  MLP 10 105 1.61 (1.09 to 2.13) Z = 6.09 (p < 0.001) 58% (0.01)
 96 h
  EFP 5 72 2.87 (1.23 to 4.51) Z = 3.43 (p = 0.001) 88% (<0.001) χ2 = 5.29; p = 0.07 χ2 = 4.36; p = 0.11
  LFP 3 48 1.68 (0.38 to 2.98) Z = 2.54 (p = 0.01) 77% (0.01)
  MLP 8 89 1.03 (0.69 to 1.37) Z = 5.95 (p < 0.001) 0% (0.91)
  EFP 9 95 6.57 (4.42 to 8.71) Z = 5.99 (p < 0.001) 89% (<0.001) χ2 = 9.79; p = 0.007 χ2 = 9.27; p = 0.011
  LFP 5 53 5.37 (2.10 to 8.63) Z = 3.22 (p = 0.001) 93% (<0.001)
  MLP 13 132 3.08 (2.22 to 3.95) Z = 6.98 (p < 0.001) 80% (<0.001)
 0 h
  EFP 3 21 1.89 (−0.24 to 4.02) Z = 1.74 (p = 0.08) 85% (0.001) χ2 = 2.67; p = 0.26 N/A
  LFP 3 38 0.18 (−0.28 to 0.63) Z = 0.76 (p = 0.45) 0% (0.81)
  MLP 6 71 0.41 (0.03 to 0.78) Z = 2.14 (p = 0.03) 16% (0.31)
 24 h
  EFP 6 52 1.34 (0.58 to 2.09) Z = 3.48 (p < 0.001) 62% (0.02) χ2 = 0.21; p = 0.90 N/A
  LFP 3 37 1.36 (0.57 to 2.14) Z = 3.39 (p < 0.001) 50% (0.14)
  MLP 8 82 1.55 (0.87 to 2.24) Z = 4.44 (p < 0.001) 70% (0.002)
 48 h
  EFP 5 46 1.69 (0.91 to 2.47) Z = 4.25 (p < 0.001) 56% (0.06) χ2 = 2.67; p = 0.26 N/A
  LFP 5 57 1.02 (0.19 to 1.86) Z = 2.40 (p = 0.02) 74% (0.004)
  MLP 7 70 0.97 (0.62 to 1.33) Z = 5.33 (p < 0.001) 0% (0.65)
 72 h
  EFP 4 37 1.47 (−0.18 to 3.12) Z = 1.75 (p = 0.08) 88% (<0.001) χ2 = 1.48; p = 0.48 N/A
  LFP 2 15 0.44 (−0.78 to 1.65) Z = 0.70 (p = 0.48) 59% (0.12)
  MLP 8 78 1.23 (0.55 to 1.91) Z = 3.53 (p < 0.001) 69% (<0.001)
 96 h
  EFP 2 24 2.52 (−1.48 to 6.51) Z = 1.23 (p = 0.22) 95% (<0.001) χ2 = 1.69; p = 43 N/A
  LFP 3 30 0.85 (0.31 to 1.38) Z = 3.10 (p = 0.002) 0% (0.57)
  MLP 3 30 1.28 (0.71 to 1.85) Z = 4.42 (p < 0.001) 0% (0.96)
  EFP 8 66 1.66 (0.75 to 2.58) Z = 3.56 (p < 0.001) 78% (<0.001) χ2 = 2.10; p = 0.35 N/A
  LFP 4 52 2.98 (1.43 to 4.52) Z = 3.78 (p < 0.001) 84% (0.0003)
  MLP 9 88 2.12 (1.46 to 2.78) Z = 6.25 (p < 0.001) 64% (<0.001)
Strenght loss
 0 h
  EFP 1 10 −3.46 (−4.95 to −1.98) Z = 4.57 (p < 0.001) N/A χ2 = 12.70; p = 0.002 χ2 = 10.25; p = 0.006
  LFP 3 27 −1.87 (−2.55 to −1.20) Z = 5.46 (p < 0.001) 0% (0.38)
  MLP 5 64 −0.89 (−0.34 to −1.43) Z = 3.18 (p = 0.001) 53% (0.08)
 24 h
  EFP 1 10 −2.68 (−3.96 to −1.41) Z = 4.12 (p < 0.001) N/A χ2 = 8.52; p = 0.01 χ2 = 11.78; p = 0.003
  LFP 2 16 −1.64 (−3.70 to 0.42) Z = 1.56 (p = 0.12) 82% (0.02)
  MLP 5 64 −0.71 (−0.25 to −1.18) Z = 3.00 (p = 0.003) 38% (0.17)
 48 h
  EFP 1 10 −2.76 (−4.05 to −1.46) Z = 4.17 (p < 0.001) N/A χ2 = 13.92; p < 0.001 χ2 = 11.9; p = 0.003
  LFP 4 36 −1.36 (−2.27 to −0.46) Z = 2.96 (p = 0.003) 64% (0.04)
  MLP 9 88 −0.47 (−0.17 to −0.77) Z = 3.04 (p = 0.002) 0% (0.74)
 72 h
  EFP 1 10 −1.54 (−2.57 to −0.52) Z = 2.95 (p = 0.003) N/A χ2 = 6.19; p = 0.05 χ2 = 7.45; p = 0.02
  LFP 2 16 −0.86 (−2.52 to 0.80) Z = 1.02 (p = 0.31) 78% (0.03)
  MLP 9 97 −0.21 (−0.55 to 0.13) Z = 1.20 (p = 0.23) 19% (0.15)
 96 h
  EFP 1 10 −0.93 (−1.87 to 0.00) Z = 1.96 (p = 0.05) N/A χ2 = 0.59; p = 0.75 χ2 = 0.41; p = 0.81
  LFP 4 36 −0.87 (−1.80 to 0.06) Z = 1.84 (p = 0.07) 69% (0.02)
  MLP 6 48 −0.10 (−0.76 to 0.57) Z = 0.28 (p = 0.78) 59% (0.03)
  EFP 1 10 −3.46 (−4.95 to −1.98) Z = 4.57 (p < 0.001) N/A χ2 = 15.84; p < 0.001 χ2 = 8.90; p = 0.01
  LFP 4 36 −1.63 (−2.36 to −0.89) Z = 4.32 (p < 0.001) 41% (0.16)
  MLP 9 88 −0.72 (−1.07 to −0.36) Z = 3.99 (p < 0.001) 19% (0.27)
*SMD = standard mean difference; CI = confidence interval; DOMS = delayed onset muscle soreness; CK = creatine kinase; EFP = early follicular phase; LFP = late follicular phase; MLP = midluteal phase; N/A = not applicable; EMID = exercise-induced muscle damage.
Different between phases adjusted according to the severity of the exercise protocol.

The analysis considering the severity of exercise protocols revealed differences from 24 hours until 96 hours post-exercise: 0 hours (χ2 = 0.79; df = 2; p = 0.68; I2 = 0%), 24 hours (χ2 = 9.71; df = 2; p = 0.008; I2 = 79.4%), 48 hours (χ2 = 9.81; df = 2; p = 0.007; I2 = 79.6%), 72 hours (χ2 = 7.56; df = 2; p = 0.02; I2 = 73.6%), 96 hours (χ2 = 6.98; df = 2; p = 0.03; I2 = 71.4%), and maximum value (χ2 = 11.6; df = 2; p = 0.04; I2 = 82.1%). On the basis of this, the analysis of EIMD severity was repeated removing the protocols eliciting light muscle damage because they may confound the results. However, the differences between phases remained. These differences after adjusting for exercise severity are shown in Table 2.

Effects of the Menstrual Cycle Phase on Creatine Kinase

The meta-analysis demonstrated no significant differences between MC phases for CK in response to exercise (Table 2). Significant increases of CK in comparison with baseline were observed in all phases at 24 and 48 hours post-exercise, in the MLP 72 hours post-exercise, and in the LFP and MLP at 96 hours post-exercise.

Regarding the supplementary analysis, when the severity of EIMD was considered, no differences were observed in CK between baseline and 0 hours (χ2 = 2.67; df = 2; p = 0.26; I2 = 25.2%), 24 hours (χ2 = 0.21; df = 2; p = 0.09; I2 = 0%), 48 hours (χ2 = 2.67; df = 2; p = 0.26; I2 = 25%),72 hours (χ2 = 1.48; df = 2; p = 0.48; I2 = 0%), and 96 hours (χ2 = 1.69; df = 2; p = 0.43; I2 = 0%) post-exercise and in comparison with the maximum value (χ2 = 2.10; df = 2; p = 0.35; I2 = 4.6%).

Effects of the Menstrual Cycle Phase on Strength Loss

Regarding strength loss, significant differences between phases were observed until 72 hours post-exercise (Table 2). The lowest mean value in strength loss was observed at the MLP for all post-exercise time points, whereas the highest was observed at the EFP. Significant strength loss in comparison with baseline was observed for almost all post-exercise measurements, except in the LFP at 24, 72, and 96 hours post-exercise (Table 2).

The supplementary analysis for strength loss considering the muscle damage elicited by the different exercise protocols revealed differences at 48, 72, 96 hours, and at the maximum value: 0 hours (χ2 = 2.79; df = 2; p = 0.25; I2 = 28.4%), 24 hours (χ2 = 3.28; df = 2; p = 0.19; I2 = 39%), 48 hours (χ2 = 5.88; df = 2; p = 0.05; I2 = 66%), 72 hours (χ2 = 5.50; df = 2; p = 0.06; I2 = 63.6%), 96 hours (χ2 = 12.58; df = 2; p = 0.002; I2 = 84.1%), and maximum value (χ2 = 5.86; df = 2; p = 0.05; I2 = 65.9%). As well as with DOMS, the EIMD severity analysis was repeated without the protocols eliciting light muscle damage, and the differences between phases remained, except at 96 hours post-exercise (Table 2).


The major finding of this study was that muscle damage, specifically DOMS and strength loss, is affected by MC phases. The meta-analysis showed that the higher the concentration of sex hormones estrogen and progesterone (as reported or assumed in the MLP according to the included studies) the lower the DOMS and strength loss differences observed between pre-exercise and post-exercise. By contrast, MC phases do not seem to affect CK response. The knowledge of muscle damage response relating to the exercise modality is important for prescription of adequate training loads and recovery periods to meet athletes' requirements. To the best of our knowledge, this is the first meta-analysis to investigate the effects of sex hormone variations throughout the MC on post-exercise muscle damage response.

Regarding DOMS, the major finding from our results is the difference between phases observed from 24 to 72 hours post-exercise. What could be clearly observed from our results is that the slightest DOMS increase was found in the MLP along all-time measurements in comparison with EFP and LFP. This finding agrees well with previous research (60), which has shown the underlying protective effect of estrogen (26) when the concentrations of this hormone are high as in the MLP. However, on the basis of this finding, it could also be suggested that the increase of progesterone in the MLP may not necessarily modulate estrogen's protective effects as has been suggested in the literature (75). In addition, results also suggest higher soreness values in the EFP, which, indeed, could seem reasonable according to the menstrual-associated dysmenorrhea reported by women (17) that could contribute to increase muscle pain perception and decrease their predisposition to perform a physical task (16). Nevertheless, this should be interpreted with caution because the mean differences in the LFP in some time points are also high. This hypothesis would need to be investigated to confirm whether a relationship exists between dysmenorrhea and higher muscle pain perception because women can also experience periovulatory pain widely known as intermenstrual pelvic pain (29).

Concerning heterogeneity analysis, the MLP showed lower values in all time points, which may be related to the larger number of studies analyzing EIMD in this phase (Table 2) in comparison with the EFP and LFP. In general, the smaller amount of studies in these phases could impair achievement of statistical power required. Although our supplementary analysis could not confirm muscle damage severity as a source of heterogeneity, within the same severity level, high heterogeneity is still observed when analyzing DOMS values within the different phases, which may be related to the different combination of the variables that determine exercise load (e.g.,: reps, duration, recovery etc). In this regard, from 5 studies evaluating DOMS over the MC (2,10,53,60,70), only 3 consisted of endurance exercise protocols (10,53,60), and only 2 of these 3 were classified with moderate muscle damage (53,60). However, one of the studies reported a lack of differences between MC phases (53), whereas the other showed a trend to lower muscle soreness in the MLP (60). This variety of features regarding exercise variables additionally impairs comparisons between studies. Further reasons involved in DOMS response could be the lack of entirely homogeneous rates of pain perception (60) or other highly variable factors between subjects, as for instance, the architecture of the muscle fiber type. Fusiform muscles and type II fibers seem to be more susceptible to EIMD than pinnate muscles and type I fibers, respectively (41). Nevertheless, it may also be related to other reasons such as the complexity of nociceptive mechanisms, variations in single nucleotide polymorphisms associated to the level of muscle soreness (41), or individual release of several substances (e.g., serotonin) in the degeneration–regeneration process of tissues that stimulate pain afferent nerve endings (11). Altogether, this meta-analysis suggests that the MC phase affects DOMS, although the different exercise modalities and limited comparisons over the MC indicate further research in the area is needed.

As previously mentioned, sarcomeres are overstretched during strenuous exercise, prompting an increase in membrane permeability and subsequent sarcolemma-associated disruption, which could lead to CK leakage from the damaged muscle to the blood stream (48). Peak values in serum CK have been observed 8 hours after training, although the most pronounced increase in CK seems to occur between 2 and 7 days post-exercise (6,63). The time course of CK release into plasma depends on the training status, exercise modality, and load. Our analysis revealed significant differences from baseline at 24 and 48 hours for all phases. In some phases, these differences were observed at 72 hours (MLP) and even 96 hours (LFP and MLP) post-exercise. This could be expected after a damaging protocol due to the delayed CK response and slow diffusion out of the damaged muscle as a secondary stage of muscle damage (57,63). The lack of differences in some time points could be explained by the small amount of studies and subjects evaluated in some phases as EFP and LFP in comparison with MLP (Table 2), which may not have been sufficient to reach the statistical power required. Intriguingly, the current meta-analysis did not demonstrate an effect of MC phase on CK response, and thus the results are not in agreement with the whole body of literature that suggests the positive influence of estrogen in muscle membrane stability and EIMD restraint (13,26,77). Nonetheless, it is worth mentioning that the use of plasma enzyme activities for estimating the amount of muscle damage should be cautioned (75,76). On one hand, high interindividual CK response variability has been observed, which could contribute to the high heterogeneity observed. This may be explained by its release by the muscle and its clearance by the reticuloendothelial system, which depends on the lymph flow because CK molecular weight does not permit direct passage to the blood (51). On the other hand, the increase of muscle enzymes in plasma does not necessarily reflect structural damage (76). The changes in baseline serum CK activity may indicate normal adaptations to physiological demands and be a part of the shift in muscle homeostasis that accompanies habitual and chronic exercise training (75). In addition, CK response is related to the susceptibility of being a high or low responder (6,48,52) and thus, the presence of variation in some genetic coding for specific myofibrillar proteins such as α-actinin 3, myosin light chain kinase, or the myosin heavy chain isoform expressed by the cell and hence, the fiber type (12).

The lacking differences between MC phases in CK results are possibly derived from the fact that most studies analyzed evaluating exercise-induced CK in more than one phase of the MC did not find differences between them (53,60,70), although some of them did find differences (30,80). Intriguingly, the studies reporting differences between MC phases were classified as light in terms of EIMD severity, whereas those not reporting differences were classified as moderate. However, EIMD severity could not explain the results because no differences in CK were revealed by our supplementary analysis when exercise protocols where compared regarding their severity. In fact, a mild relationship has been observed between volume load (mass × reps × sets) and plasma CK (48). However, duration may be relevant in this relationship because the 2 studies reporting differences between MC phases were running protocols longer in duration than the exercise protocols not reporting differences. This finding agrees well with previous research showing higher increase in both histological muscle damage and enzyme release after longer duration exercises and even disproportionally higher in much longer exercises, such as marathons (76). Interestingly, in these 2 studies, higher post-exercise CK levels were observed in the MLP, when subjects presented higher concentrations of sex hormones, than in the MFP when low concentrations were obtained. This result therefore would coincide with the pattern observed for muscle soreness for the EFP and MLP, but this meta-analysis considering all the studies evaluating this response revealed no effects of MC phases on CK.

The main finding related to exercise-induced strength loss provided by our analysis is the differences between phases observed in all post-exercise measurements except 96 hours. As with DOMS, the lower mean difference in strength loss from pre-exercise to post-exercise was shown in the MLP, whereas the EFP seems to elicit the higher mean difference in this variable. This finding suggests that higher concentrations of sex hormones may attenuate strength loss after exercise and may be related to the fluctuation of estrogen receptors throughout the MC. Strength loss during maximal voluntary contraction has been consolidated by previous research as the main EIMD indirect marker because it seems to coordinate the responses of other indirect markers (18). However, the body of evidence evaluating exercise-induced strength loss throughout the MC is limited. In this meta-analysis, 3 studies specifically evaluated strength loss (2,53,70) over the MC. All 3 studies measured subjects in the EFP, whereas 2 of them measured in the LFP (53,70) and one in the MLP (2). However, in contrast to our results, 2 of these studies showed higher strength loss in the LFP or MLP when estrogen or both estrogen and progesterone present high concentrations. This might also be related to previous findings showing worse strength development in ovulation and MLPs (72). By contrast, high estrogen concentrations have been related to greater expression of estrogen receptors (23,31), which seem to elicit the myogenic activation of satellite cells. Therefore, estrogens seem to assist muscle repair and regeneration during recovery, although this process needs further investigation.

Regarding the analysis of EIMD severity, a difference between studies was observed. Although the exercise protocols considered were all lower limb based except one, some of them were classified as severe while others were classified as moderate in terms of muscle damage. In addition, these differences provided by our supplementary analysis may also be related to different factors, such as subjects' training status. Nonetheless, heterogeneity values were lower in this variable, specifically in the MLP, possibly related to the greater number of studies evaluating strength loss in this phase (Table 2) in comparison with the EFP and LFP. Hence, despite this meta-analysis suggests lower strength loss in MLP, this finding should be taken with caution because the number of studies evaluating exercise-induced strength loss is small in some phases of the MC, for instance, the EFP with only one study (53).

A major limitation of the research to date and therefore in this systematic review is that so far only 7 studies have actually compared EIMD over the MC phases. The other 12 studies included in this review only reported on responses to exercise at one point of the MC. From a total of 19 studies evaluated, only one performed a damaging protocol in 3 phases of the MC (60). This study used a randomized controlled trial design with only 5 subjects tested in each phase and thus not considering the whole range of hormone fluctuations from the same person. By contrast, 6 studies evaluated the same subject in 2 MC phases (2,10,30,53,70,80), being one of these phases with sex hormone concentrations at their nadirs (EFP) or with low levels (MFP). However, in 2 of these studies (53,70), the phase with low sex hormone concentrations was compared with the phase with high estrogen and low progesterone concentrations (LFP), whereas in the rest of them, the comparison was between low hormone concentrations and high estrogen and progesterone levels (MLP) (2,10,30,80). Therefore, phases are not entirely comparable among them.

The low number of studies evaluating more than one phase of the MC leads to another limitation, which is the high heterogeneity observed in some phases and time points, especially the LFP. This highlights the existence of confounding variables influencing the results, such as the training status of subjects, whose influence was not possible to analyze due to the scarce information provided by some studies. The protocol used to verify MC phases is another limitation of several of the studies. It should include retrospective monitoring of the cycle length, blood sampling to confirm hormone concentrations, and ovulation kits to detect the urine LH surge occurring just before ovulation and, consequently, the peak of estrogen (44,69). Also, temperature monitoring could help to depict a map of the MC phases (73). These instruments provide information not only to properly identify LFP and MLP measurements but also to confirm the cycle is not anovulatory. From a total of 19 studies, none included the three-step methodology to verify the MC (44,69): calendar-based counting, serum hormone analysis, and urinary LH measurement. Specifically, 10 of the included studies performed blood testing (2,10,30,38,55,60,68,70,74,80) and one salivary testing (53) to analyze sex hormones, whereas only 3 studies predicted ovulation by using LH urine-based kits (53,60,74). The rest of the studies only took into consideration calendar-based counting. Hence, further research should include more accurate methodologies to verify MC phases.

Finally, it should be considered that this meta-analysis is focused on the study of indirect markers of EIMD that are more attainable and useful for exercise coaches. Further research on histological analysis of muscle tissue damage through biopsies would be interesting to directly measure damage from muscle tissue. The hormone fluctuations may influence the presence and distribution of estrogen receptors between muscle fibers, which are important to mediate the transcriptional activity in physiological responses (78). As previously mentioned, these receptors seem to favor muscle tissue repair and adaptation to physical training, and they may determine further actions of estrogens related to skeletal muscle. Therefore, on the basis of the outlined limitations, future research should focus on intrasubject designs, evaluating as many phases as possible of the MC, especially the EFP, LFP, and MLP, and including both an adequate monitoring and verification of hormone fluctuations and when possible, histological analysis of muscle tissue.

In conclusion, the main finding from our analysis is that sex hormones influence muscle damage response over the MC. Specifically, DOMS and strength loss mean differences in response to exercise are lower in the MLP, when sex hormone concentrations are high, whereas the EFP seems to elicit higher DOMS and strength loss differences. These differences between MC phases extended to 72 hours post-exercise for both variables, with higher mean differences in comparison with baseline around 24 and 48 hours post-exercise. No differences between MC phases were observed for CK. Hence, it is evident that further research should consider the response of muscle damage throughout the MC.

Practical Applications

Muscle damage has been demonstrated to acutely affect performance by increasing not only pain discomfort but also the time needed for optimal recovery to achieve the appropriate training stimuli and adaptations (27). The findings of this meta-analysis could provide helpful guidance on exercise prescription in women. During the EFP, muscle tissues may not support such strenuous loads as those endured when sex hormone concentration is higher, and the muscle damage response seems to be attenuated. Hence, these results suggest key moments to increase loads (LFP and MLP) or recovery periods (EFP) according to athletes’ pain perception and muscular performance and strength after EIMD. Furthermore, the 2 variables showing differences between phases are DOMS and strength loss, which provide attainable and useful information to coaches to evaluate EIMD compared with blood markers, such as CK. However, load administering should still be carefully controlled and individualized.

Finally, this meta-analysis demonstrated peak muscle damage responses in women around 24–72 hours post-exercise. Future research could be therefore recommended to standardize time points of testing to 24, 48, and 72 hours post-exercise. This would reduce testing days, making volunteers' enrolment more attainable.


N.R.P. and V.M.A.-M. are both supported by a grant provided by Universidad Politécnica de Madrid. The IronFEMME Study takes place with the financial support of the Ministerio de Economía y Competitividad, Convocatoria de ayudas I+D 2016, Plan Estatal de Investigación Científica y Técnica y de Innovación 2013–2016 (Contract DEP2016-75387-P). Authors have no other funding or conflicts of interest to declare.

Systematic review registration number PROSPERO 2019: CRD42018110290.


1. Abbasi A, Vieira RdP, Bischof F, et al. Sex-specific variation in signaling pathways and gene expression patterns in human leukocytes in response to endotoxin and exercise. J Neuroinflammation 13: 289, 2016.
2. Anderson LJ, Baker LL, Schroeder ET. Blunted myoglobin and quadriceps soreness after electrical stimulation during the luteal phase or oral contraception. Res Q Exerc Sport 88: 193–202, 2017.
3. Barba-Moreno L, Cupeiro R, Romero-Parra N, Janse de Jonge XAK, Peinado AB. Cardiorespiratory responses to endurance exercise over the menstrual cycle and with oral contraceptive use. J Strength Cond Res 2019. doi: 10.1519/JSC.0000000000003447. Epub ahead of print.
4. Beery AK, Zucker I. Sex bias in neuroscience and biomedical research. Neurosci Biobehav Rev 35: 565–572, 2011.
5. Bijur P, Silver W, Gallagher E. Reliability of the visual analog scale for measurement of acute pain. Acad Emerg Med 8: 1153–1157, 2001.
6. Brancaccio P, Maffulli N, Limongelli FM. Creatine kinase monitoring in sport medicine. Br Med Bull 81–82: 209–230, 2007.
7. Brentano MA, Martins Kruel LF. A review on strength exercise-induced muscle damage: Applications, adaptation mechanisms and limitations. J Sports Med Phys Fitness 51: 1–10, 2011.
8. Brown MA, Howatson G, Keane K, Stevenson EJ, Exercise-induced muscle damage following dance and sprint specific exercise in females. J Sports Med Physical Fitness 56: 1376–1383, 2016.
9. Cadore EL, González-Izal M, Grazioli R, et al. Effects of concentric and eccentric strength training on fatigue Induced by concentric and eccentric exercises. Int J Sports Physiol Perform 14: 91–98, 2019.
10. Chaffin ME, Davis JE, Berg KE, et al. Interleukin-6 and delayed onset muscle soreness do not vary during the menstrual cycle. Res Q Exerc Sport 82: 693–701, 2011.
11. Chen TC, Lin KY, Chen HL, Lin MJ, Nosaka K. Comparison in eccentric exercise-induced muscle damage among four limb muscles. Eur J Appl Physiol 111: 211–223, 2011.
12. Choi SJ. Differential susceptibility on myosin heavy chain isoform following eccentric-induced muscle damage. J Exerc Rehabil 10: 344–348, 2014.
13. Clarkson PM, Hubal MJ. Are women less susceptible to exercise-induced muscle damage? Curr Opin Clin Nutr Metab Care 4: 527–531, 2001.
14. Constantini NW, Dubnov G, Lebrun CM. The menstrual cycle and sport performance. Clin Sports Med 24: 51–82, 2005.
15. Costello JT, Bieuzen F, Bleakley CM. Where are all the female participants in Sports and Exercise Medicine research? Eur J Sport Sci 14: 847–851, 2014.
16. Cristina-Souza G, Santos-Mariano AC, Souza-Rodrigues CC, et al. Menstrual cycle alters training strain, monotony, and technical training length in young. J Sports Sci 37: 1824–1830, 2019.
17. Daley AJ. Exercise and primary dysmenorrhoea. Sports Med 38: 659–670, 2008.
18. Damas F, Nosaka K, Libardi CA, Chen TC, Ugrinowitsch C. Susceptibility to exercise-induced muscle damage: A cluster analysis with a large sample. Int J Sports Med 37: 633–640, 2016.
19. Davis HC, Hackney AC. The hypothalamic–pituitary–ovarian axis and oral contraceptives: Regulation and function. In: Sex Hormones, Exercise and Women. Hackney AC, ed. Cham, Switzerland: Springer International Publishing, 2017. pp. 1–17.
20. Deeks JJ, Higgins JPT, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Cochrane Handbook for Systematic Reviews of Interventions version 6.1. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, eds. Cochrane, 2020. Available at www.training.cochrane.org/handbook. (updated September 2020).
21. Dieli-Conwright CM, Spektor TM, Rice JC, Schroeder ET. Hormone therapy attenuates exercise-induced skeletal muscle damage in postmenopausal women. J Appl Physiol 107: 853–858, 2009.
22. Douglas J, Pearson S, Ross A, McGuigan M. Eccentric exercise: Physiological characteristics and acute responses. Sports Med 47: 663–675, 2017.
23. Ekenros L, Papoutsi Z, Fridén C, Dahlman Wright K, Lindén Hirschberg A. Expression of sex steroid hormone receptors in human skeletal muscle during the menstrual cycle. Acta Physiol 219: 486–493, 2017.
24. Elliott KJ, Cable NT, Reilly T, Diver MJ. Effect of menstrual cycle phase on the concentration of bioavailable 17-beta oestradiol and testosterone and muscle strength. Clin Sci (Lond) 105: 663–669, 2003.
25. Ely MR, Romero SA, Sieck DC, et al. A single dose of histamine-receptor antagonists before downhill running alters markers of muscle damage and delayed-onset muscle soreness. J Appl Physiol 122: 631–641, 2017.
26. Enns DL, Tiidus PM. The influence of estrogen on skeletal muscle. Sports Med 40: 41–58, 2010.
27. Fatouros IG, Jamurtas AZ. Insights into the molecular etiology of exercise induced inflammation: Opportunities for optimizing performance. J Inflamm Res 9: 175–186, 2016.
28. Gougeon A. Dynamics of follicular growth in the human: A model from preliminary results. Hum Reprod 1: 81–87, 1986.
29. Grandi G, Xholli A, Ferrari S, et al. Intermenstrual pelvic pain, quality of life and mood. Gynecol Obstet Invest 75: 97–100, 2013.
30. Hackney AC, Kallman AL, Ağgön E. Female sex hormones and the recovery from exercise: Menstrual cycle phase affects responses. Biomed Hum Kinet 11: 87–89, 2019.
31. Haines M, McKinley-Barnard SK, Andre TL, et al. Skeletal muscle estrogen receptor activation in response to eccentric exercise up-regulates myogenic-related gene expression independent of differing serum estradiol levels occurring during the human menstrual cycle. J Sports Sci Med 17: 31–39, 2018.
32. Hall JE. Neuroendocrine control of the menstrual cycle. In: Yen and Jaffe's Reproductive Endocrinology Physiology, Pathophysiology, and Clinical Management. Jerome I, Strauss F, Barbieri RL, eds. Philadelphia, PA: Elsevier, 2019. pp. 149–166.
33. Hall JE, Guyton AC. Chapter 81: Female physiology before pregnancy and female hormones. In: Guyton and Hall Textbook of Medical Physiology. Philadelphia, PA: Saunders Elsevier, 2011. p. 987–1000.
34. Hawkins SM, Matzuk MM. The menstrual cycle: Basic biology. Ann N Y Acad Sci 1135: 10–18, 2008.
35. Heart N. Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. The Cochrane Collaboration, 2019. Study Quality Assessment Tools. Availabel at: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools.
36. Heavens KR, Szivak TK, Hooper DR, et al. The effects of high intensity short rest resistance exercise on muscle damage markers in men and women. J Strength Cond Res 28: 1041–1049, 2014.
37. Hicks KM, Onambélé GL, Winwood K, Morse CI. Muscle damage following maximal eccentric knee extensions in males and females. PLoS One 11: e0150848, 2016.
38. Hicks KM, Onambélé-Pearson G, Winwood K, Morse CI. Oral contraceptive pill use and the susceptibility to markers of exercise-induced muscle damage. Eur J Appl Physiol 117: 1393–1402, 2017.
39. Higgins JPT. Measuring inconsistency in meta-analyses. BMJ 327: 557–560, 2003.
40. Hunter SK. Sex differences in fatigability of dynamic contractions. Exp Physiol 101: 250–255, 2016.
41. Jamurtas AZ, Theocharis V, Tofas T, et al. Comparison between leg and arm eccentric exercises of the same relative intensity on indices of muscle damage. Eur J Appl Physiol 95: 179–185, 2005.
42. Jang HJ, Lee JD, Jeon HS, et al. Metabolic profiling of eccentric exercise-induced muscle damage in human urine. Toxicol Res 34: 199–210, 2018.
43. Janse de Jonge XAK. Effects of the menstrual cycle on exercise performance. Sports Med 33: 833–851, 2003.
44. Janse de Jonge XAK, Thompson B, Han A. Methodological recommendations for menstrual cycle research in sports and exercise. Med Sci Sports Exerc 51: 2610–2617, 2019.
45. Joyce S, Sabapathy S, Bulmer AC, Minahan C. The effect of prior eccentric exercise on heavy-intensity cycling: The role of gender and oral contraceptives. Eur J Appl Physiol 114: 995–1003, 2014.
46. Keane KM, Salicki R, Goodall S, Thomas K, Howatson G. Muscle damage response in female collegiate athletes after repeated sprint activity. J Strength Cond Res 29: 2802–2807, 2015.
47. Kendall B, Eston R. Exercise-induced muscle damage and the potential protective role of estrogen. Sports Med 32: 103–123, 2002.
48. Koch AJ, Pereira R, Machado M. The creatine kinase response to resistance exercise. J Musculoskelet Neuronal Interact 14: 68–77, 2014.
49. Lebrun CM, McKenzie DC, Prior JC, Taunton JE. Effects of menstrual cycle phase on athletic performance. Med Sci Sports Exerc 27: 437–444, 1995.
50. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. Plos Med 6: e1000100, 2009.
51. Machado M, Brown LE, Augusto-Silva P, Pereira R. Is exercise-induced muscle damage susceptibility body segment dependent? Evidence for whole body susceptibility. J Musculoskelet Neuronal Interact 13: 105–110, 2013.
52. Machado M, Pereira R, Willardson JM. Short intervals between sets and individuality of muscle damage response. J Strength Cond Res 26: 2946–2952, 2012.
53. Mackay K, Gonzalez C, Zbinden-Foncea H, Penailillo L. Effects of oral contraceptive use on female sexual salivary hormones and indirect markers of muscle damage following eccentric cycling in women. Eur J Appl Physiol 119: 2733–2744, 2019.
54. Markofski MM, Braun WA. Influence of menstrual cycle on indices of contraction-induced muscle damage. J Strength Cond Res 28: 2649–2656, 2014.
55. Minahan C, Joyce S, Bulmer AC, Cronin N, Sabapathy S. The influence of estradiol on muscle damage and leg strength after intense eccentric exercise. Eur J Appl Physiol 115: 1493–1500, 2015.
56. Moreira A, Nosaka K, Viveiros L, et al. Changes in muscle damage markers in female basketball players. Biol Sport 31: 3–7, 2013.
57. Muanjai P, Mickevičius M, Sniečkus A, et al. Slow torque recovery after eccentric exercise and the repeated bout effect; the role of primary and secondary muscle damage. J Musculoskelet Neuronal Interact 19: 207–214, 2019.
58. Mujika I, Taipale RS. Sport science on women, women in sport science. Int J Sports Physiol Perform 14: 1013–1014, 2019.
59. Nikolaidis MG, Paschalis V, Giakas G, et al. Favorable and prolonged changes in blood lipid profile after muscle-damaging exercise. Med Sci Sports Exerc 40: 1483–1489, 2008.
60. Oosthuyse T, Bosch A. The effect of gender and menstrual phase on serum creatine kinase activity and muscle soreness following downhill Running. Antioxidants 6: 16, 2017.
61. Oosthuyse T, Bosch AN. The effect of the menstrual cycle on exercise metabolism. Sports Med 40: 207–227, 2010.
62. Pal S, Chaki B, Chattopadhyay S, Bandyopadhyay A. High-Intensity exercise induced oxidative stress and skeletal muscle damage in postpubertal boys and girls. J Strength Cond Res 32: 1045–1052, 2018.
63. Peake JM, Neubauer O, Della Gatta PA, Nosaka K. Muscle damage and inflammation during recovery from exercise. J Appl Phys 122: 559–570, 2017.
64. Price TB, Sanders K. Muscle and liver glycogen utilization during prolonged lift and carry exercise: Male and female responses. Physiol Rep 5: 1–11, 2017.
65. Proske U, Morgan DL. Muscle damage from eccentric exercise: Mechanism, mechanical signs, adaptation and clinical applications. J Physiol 537: 333–345, 2001.
66. Rankin P, Lawlor MJ, Hills FA, et al. The effect of milk on recovery from repeat-sprint cycling in female team-sport athletes. Appl Physiol Nutr Metab 43: 113–122, 2018.
67. Reilly T. The menstrual cycle and human performance: An overview. Biol Rhythm Res 31: 29–40, 2000.
68. Savage KJ, Clarkson PM. Oral contraceptive use and exercise-induced muscle damage and recovery. Contraception 66: 67–71, 2002.
69. Schaumberg MA, Jenkins DG, Janse de Jonge XAK, Emmerton LM, Skinner TL. Three-step method for menstrual and oral contraceptive cycle verification. J Sci Med Sport 20: 965–969, 2017.
70. Sipavičienė S, Daniusevičiutė L, Klizienė I, Kamandulis S, Skurvydas A. Effects of estrogen fluctuation during the menstrual cycle on the response to stretch-shortening exercise in females. Biomed Res Int 2013: 1–6, 2013.
71. Stricker R, Eberhart R, Chevailler MC, et al. Establishment of detailed reference values for luteinizing hormone, follicle stimulating hormone, estradiol, and progesterone during different phases of the menstrual cycle on the Abbott ARCHITECT analyzer. Clin Chem Lab Med 44: 883–887, 2006.
72. Tenan MS, Hackney AC, Griffin L. Maximal force and tremor changes across the menstrual cycle. Eur J Appl Physiol 116: 153–160, 2016.
73. Tenan MS, Peng YL, Hackney AC, Griffin L. Menstrual cycle mediates vastus medialis and vastus medialis oblique muscle activity. Med Sci Sports Exerc 45: 2151–2157, 2013.
74. Thompson HS, Hyatt JP, De Souza MJ, Clarkson PM. The effects of oral contraceptives on delayed onset muscle soreness following exercise. Contraception 56: 59–65, 1997.
75. Thompson HS, Scordilis SP, De Souza MJ. Serum creatine kinase activity varies with ovulatory status in regularly exercising, premenopausal women. Horm Res 65: 151–158, 2006.
76. Van der Meulen JH, Kuipers H, Drukker J. Relationship between exercise-induced muscle damage and enzyme release in rats. J Appl Physiol (1985) 71: 999–1004, 1991.
77. Velders M, Diel P. How sex hormones promote skeletal muscle regeneration. Sports Med 43: 1089–1100, 2013.
78. Wiik A, Ekman M, Johansson O, Jansson E, Esbjornsson M. Expression of both oestrogen receptor alpha and beta in human skeletal muscle tissue. Histochem Cell Biol 131: 181–189, 2009.
79. Wikström-Frisén L, Boraxbekk CJ, Henriksson-Larsén K. Effects on power, strength and lean body mass of menstrual/oral contraceptive cycle based resistance training. J Sports Med Phys Fitness 57: 43–52, 2017.
80. Williams T, Walz E, Lane AR, Pebole M, Hackney AC. The effect of estrogen on muscle damage biomarkers following prolonged aerobic exercise in eumenorrheic women. Biol Sport 32: 193–198, 2015.
81. Wolf MR, Fragala MS, Volek JS, et al. Sex differences in creatine kinase after acute heavy resistance exercise on circulating granulocyte estradiol receptors. Eur J Appl Physiol 112: 3335–3340, 2012.

sex hormones; muscle soreness; women; creatine kinase; eumenorrheic; eccentric exercise

Supplemental Digital Content

© 2020 National Strength and Conditioning Association