Journal Logo


Massage Timing Affects Postexercise Muscle Recovery and Inflammation in a Rabbit Model


Author Information
Medicine & Science in Sports & Exercise: June 2013 - Volume 45 - Issue 6 - p 1105-1112
doi: 10.1249/MSS.0b013e31827fdf18
  • Free


Delayed onset muscle soreness (DOMS) is characterized by skeletal muscle strength loss, soreness, swelling, and weakness, which are known to accompany intense or prolonged physical activity, particularly eccentric exercise (EEX) (3,4). DOMS is thought to be the result of a delayed secondary inflammatory response to a primary mechanical disruption that arguably leads to further injury (44,17). A review of both human and animal studies has shown that soreness typically appears between 8 and 24 h postexercise, peaks at 24 to 48 h, and can last for up to 7 d (11,13,14). There are multiple theories of the mechanism of injury accompanying intense EEX, including the following: local tissue inflammation; tears in the muscle fibers; disruption of passive, noncontractile components in the sarcomere, sarcoplasmic reticulum, and connective tissue that strain on the sarcomeres causing disruption of Z lines and A bands; and impairment of plasmalemmal action potential conduction (25,10,26,28).

EEX-induced muscle injury initiates an inflammatory cascade, leading to the production of pro-inflammatory cytokines within damaged muscle tissue, the systemic release of leukocytes and cytokines, and the alterations in leukocyte receptor expression and functional activity (30). Research has shown inflammation as one of the most prevalent causes for the general decline of physical function and loss of muscle strength in both the young and the elderly (9,29,33). Moreover, it has been postulated that the inflammatory process is responsible not only for successful repair but also in extending the damage process and may be coupled with significant loss of muscle mass (23,40,47). This progression of injury could occur through a variety of mechanisms, including neutrophil infiltration and activation of the respiratory burst (40). In fact, recent studies suggest that infiltrating neutrophils may play a key role in both injury and repair (42).

A variety of techniques, including therapeutic ultrasound (39), electrical stimulation (43), ice immersion (cryotherapy) (34), static and ballistic stretching (37), therapeutic massage (15), and nonsteroidal anti-inflammatory medications (24), are used clinically to minimize the pain and to enhance functional recovery after exercise-induced muscle damage (8,32). Although research on the efficacy of these treatments has increased, there remains a void in the mechanistic understanding and therefore optimal clinical strategies, especially concerning therapeutic massage, to mitigate muscle pain and weakness associated with EEX.

The science of massage is of interest to athletes, patients presenting with muscle pain, and all personnel associated with medical care. EEX-associated muscle pain and loss of function are often the impetus for athletes to receive massage therapies as a means for the rapid resolution of pain and recovery of muscle function (20,45). Although evidence to support or refute its effects on muscle function is insufficient at this time, new reports help formulate an understanding of massage and its possible role in mitigating exercise-related muscle pain (27,2). It does appear that under certain conditions, massage can reduce muscle soreness associated with eccentric muscle contractions, reduce inflammation, and promote mitochondrial biogenesis, although whether muscle force output recovers more quickly is still unclear (21,27,12). Despite the lack of conclusive scientific evidence, providers spend a large portion of their time prescribing and using massage-based therapies for a variety of soft tissue pathologies (16). Massage treatments can last up to 2 h, with therapists often advising a series of appointments, without evidence for the efficacy of repeated sessions of this therapy (35). Another important question is the optimal timing for use of massage after EEX for maximal results. The contrast between current scientific understanding of sports massage and its practice is notable, and scientific evidence to corroborate or refute an effect of massage on muscle recovery remains an important area of investigation (27).

Our laboratory has shown in several recent studies that massage-like compressive loading (MLL) incorporated into a laboratory model of EEX can improve the recovery of muscle function and may decrease leukocyte infiltration and myofiber damage (8,18,19). In the current study, we wanted to determine whether the timing of MLL after EEX could be an important variable in recommending this therapy to individuals to enhance recovery after intense EEX. On the basis of our previous work demonstrating an optimal MLL protocol (0.5 Hz, 10 N, 15 min) for the recovery of muscle function (isometric joint torque) after a bout of intense EEX (18), we hypothesized that (i) the immediate application of MLL post-EEX results in greater recovery of muscle function than delayed MLL and (ii) the immediate application of MLL decreases leukocyte infiltration as compared with 48 h of delayed MLL. The primary goals of this study were 1) to use an experimental approach that simulated MLL to evaluate the time effects of MLL application on the recovery of muscle function from EEX-induced muscle damage and 2) to evaluate the effects of the different protocols on tissue inflammation (leukocyte infiltration) via quantitative immunohistochemistry to shed further light on the optimal strategy for massage post-EEX.


Animal surgery for nerve cuff implantation and experimental design

Under a protocol approved by the Institutional Laboratory Animal Care and Use Committee (Ohio State University), 18 skeletally mature (18–20 wk) female New Zealand White rabbits (3.66 ± 0.27 kg) were anesthetized and instrumented with peroneal nerve cuffs for the stimulation of hindlimb tibialis anterior (TA) muscles as previously described (8,18).

Previous work in our laboratory determined an optimal MLL combination of magnitude, frequency, and duration for the recovery of peak isometric torque post-EEX (0.5 Hz, 10 N, 15 min) (18). The current study uses this protocol to study the effects of MLL timing postexercise-induced muscle injury on the recovery of active muscle properties and the magnitude of inflammatory cell infiltration. To this end, three conditions were compared: MLL immediately post-EEX, 48 h post-EEX, or no-MLL exercised control. Seven days after nerve cuff implantation, each rabbit was randomly assigned to a protocol, with all 18 animals initially undergoing an EEX protocol and then receiving MLL or no-MLL (control) either immediately postexercise or 48 h postexercise. Six rabbits were assigned to each of the three conditions.

Torque–joint angle properties and EEX protocol

Rabbits were anesthetized and secured supine with one foot attached to a foot pedal that allowed isometric TΘ analysis and implemented a bout of EEX (8,18). Twenty-one measurements obtained in 5° increments from 55° to 155° were taken to obtain a TΘ relationship. An isometric contraction was elicited by supramaximal stimulation, three times the α-motoneuron threshold voltage of the peroneal nerve, at each 5° increment with a 2-min rest between each measurement to minimize fatigue.

Seven sets of 10 cyclic lengthening contractions comprised the bout of EEX. Muscle activation was preceded by stretch of the TA muscle–tendon unit by 100 ms for all 70 lengthening contractions. The ankle moved within a tibiotarsal angle of 95° to 145° of plantarflexion at 150°·s−1 during each contraction. Two minutes post-EEX, repeat TΘ measurements were performed to assess the effect of EEX.

MLL protocol

All protocols were carried out daily (∼24 h apart) for 15 min on four consecutive days after initial application (day of injury [day 1] for immediate protocol and 2 d after injury [day 3] for the delayed protocol). The MLL was applied using a customized device for the application of lengthwise strokes (19,48). The self-designed MLL device allowed the capability to produce repeatable deficits in muscle function after EEX and to apply quantifiable, repeatable compressive loads via a computerized, mechanical device to methodically follow recovery and healing. This device has been shown to be far more accurate and precise in quantifying tissue forces (19) than the one used in our initial studies (8). After daily MLL, animals were recovered from anesthesia, returned to cages, and given food and water ad libitum. One day after the final MLL or no-MLL bout, a final TΘ relationship was obtained to assess the effects of the 4-d protocol. After final TΘ measurements, animals were euthanized with intravenous potassium chloride administered during deep anesthesia (5% Isoflurane) (day 5 for immediate MLL and day 7 for delayed MLL).

Analysis of TΘ recovery

A recovery index (RI) was calculated to compare the three protocols and their effects on the recovery of peak isometric torque. RI was the difference of peak torque at three fixed time points (pre-EEX, immediate post-EEX, and after 4 d of MLL protocol) (18).

The denominator measured loss of torque postexercise relative to pre-EEX and the numerator measured torque after 4 d of MLL protocol relative to pre-EEX.

To investigate the effect of MLL over the entire tested angle range, the area under the curve (AUC) for the entire TΘ relationship (55 angles) was calculated using the trapezoidal method. The RI equation was also used to calculate an RI for AUC. In addition, the angle corresponding to the peak torque output for pre-EEX, post-EEX, and post-MLL or no-MLL protocol was compared to investigate the extent MLL shifted peak torque angle back to that of preexercise (18).

Immunohistochemical analysis

Immediately after sacrifice, the tissue was harvested, weighed, flash frozen in liquid nitrogen, and carefully mounted on cork and oriented perpendicular to the long axis of the muscle. TA muscles were sectioned at 8-μm thickness for immunohistochemistry staining. Sectioned tissues were fixed for 10 min in ice-cold acetone, washed in 0.1 M phosphate-buffered saline (PBS), then quenched for 10 min in 3% H2O2 and washed again in 0.1 M PBS. Sections were then blocked for 1 h at room temperature (RT) with 2% bovine serum albumin in PBS. Muscle sections were incubated with the primary antibody (mouse antirabbit macrophage clone RAM11 [1:50], Dako, Carpinteria, CA; mouse antirabbit CD11b [1:50] or mouse antirabbit T-cells and neutrophils RPN3/57 [1:50], AbD Serotec, Raleigh, NC) for 90 min at RT. Sections were then washed with 0.1 M PBS with 0.1% Tween-20 before incubation in Immpress antimouse Ig Detection Kit (Vector Labs, Burlingame, CA) for 30 min at RT. Sections were then washed in 0.1 M PBS with 0.1% Tween-20 before incubation in Tyramide Signal Amplification (TSA) kit for 20 min at RT. After a final wash in 0.1 M PBS with 0.1% Tween-20, sections were stained with Dapi and coverslipped with Vectashield (Vector Labs).

Muscles were randomly sectioned within the MLL area and viewed with a Zeiss Axio Imager M.1 microscope (Carl Zeiss, Thornwood, NY) at 20× magnification. Five random fields for each muscle with each of the three antibodies were photographed for cell quantification. Positively labeled cells were then counted for each of the five photos, and the numbers were averaged for each animal limb by two blinded individuals. Only stained areas that colocalized with Dapi-stained nuclei were counted as positively stained cells. CD11b antibody was used primarily to identify neutrophils; however, it also recognizes other granulocytes, macrophages, blood monocytes, lymphocytes, and bone marrow cell (36), and the RPN3/57 antibody detects an uncharacterized antigen on neutrophils, T lymphocytes, thymocytes, and platelets. However, RAM11 is specific to the detection of rabbit macrophages (46).

Statistical analysis

RI for both peak torque relationship and AUC as well as joint angle shift and muscle wet weight were described with a sample mean and SE and were compared between protocols using ANOVA to assess the effects of MLL protocol on muscle function, with the level of significance set at P < 0.05. For the immunohistochemical data, as the observations from the same rabbit are correlated, the linear mixed models were used with the level of significance set at P < 0.05. PS: Power and Sample Size Calculation version 3.0, 2009, was used to perform a post hoc power analysis to differentiate significant effects with a statistical power of 0.87 for torque output analyses. All other analyses were performed using SAS/STAT software (version 9.2; SAS Institute Inc., Cary, NC).


EEX produced an average 51% ± 13% decrease in peak isometric torque (n = 18: 6 immediate, 6 delayed, and 6 no-MLL). Both the immediate and the delayed application of four consecutive days of MLL showed significant improvement in the recovery of peak torque output as compared with the EEX no-MLL control (P < 0.0001 and P = 0.02, respectively). The immediate application of MLL produced an average 129% difference in peak torque recovery as compared with the exercised no-MLL control (P < 0.0001). The 48-h delay of the same 4-d MLL protocol produced an average 82% difference of peak torque recovery compared with control (P = 0.02). There was a 64% difference in peak torque recovery between the immediate and the delayed application of MLL in favor of immediate (P = 0.0012) (Fig. 1).

Immediate application of MLL produced greatest recovery as compared with delayed MLL (RI = 0.93 ± 0.08 vs 0.48 ± 0.09, respectively). However, both applications of MLL produced a significantly greater recovery than control (RI = 0.20 ± 0.06).

Area under the curve

The area under the TΘ curve (AUC) for preexercise, postexercise, and post-MLL protocol was examined as a secondary indicator of recovery. Similar to peak torque analysis, the application of MLL beginning immediately post-EEX showed the highest AUC RI at 0.86 ± 0.08 with a 119% difference from control (Fig. 2). There was a significant difference for both applications of MLL as compared with control (RI = 0.86 ± 0.08 vs 0.22 ± 0.06, P < 0.0001 immediate and RI = 0.22 ± 0.06 vs 0.61 ± 0.10, P = 0.0051 delayed). There was greater recovery for the immediate MLL as compared with the delayed MLL with a 34% difference of RI_AUC between the two groups, but it was not significant (P = 0.0526).

RI_AUC served as a secondary indicator of muscle functional recovery with the immediate application of MLL producing greatest the recovery of peak torque and AUC as compared with both delayed MLL and control.

Joint-angle shift

The effect of the EEX bout on the torque–angle curve was calculated by subtracting the mean peak isometric torque angle preexercise bout from the corresponding postexercise angle and from each MLL protocol. Exercise produced an average 10° ± 0.2° rightward shift from preexercise peak isometric torque angle (n = 18, P = 0.006). Four consecutive days of MLL applied immediately produced an average −5.6° ± 2.8° leftward angular shift from the postexercise peak isometric torque angle (day 5), whereas delayed MLL produced a −9.7° ± 0.8° leftward angular shift from postexercise peak isometric torque angle (day 7) (P = 0.28 and P = 0.03, respectively). Exercised no-MLL control produced a −3.8° ± 3.9° leftward angular shift from postexercise peak isometric torque angle (day 5, P = 0.47) (Fig. 3).

The angle corresponding to peak torque output shifted rightward because of EEX. All three protocols, control, immediate MLL, and delayed MLL, produced a leftward shift of peak torque angle back toward preexercise angle indicating recovery from muscle injury.

Muscle wet weight and immunohistochemistry

The average muscle wet weight for the exercised control group (3.77 ± 0.09 g) was higher than for both the immediate (2.47 ± 0.12 g) and the delayed (3.19 ± 0.19 g) MLL groups (P < 0.0001 and 0.012, respectively). There was also a difference in wet weight between the immediate and the delayed MLL groups (P = 0.002) (Fig. 4).

Muscle wet weight for the control was higher than both the immediate (P < 0.0001) and delayed MLL (P = 0.012). There was also a difference in wet weight between immediate and delayed (P = 0.002).

There was a 0.94 correlation between the two blinded counters. Immediate MLL showed a 53.4% and 69.8% decrease in the respective number of RPN3/57 and CD11b-positive cells infiltrating the muscle compared with EEX no-MLL animals (30.9 ± 7.7 vs 14.4 ± 2.4 for RPN3/57 and 29.3 ± 9.2 vs 8.8 ± 2.9 for CD11b, P = 0.28 and 0.33, respectively) (Fig. 5). Delayed MLL showed a 35.09% decrease and a 40.8% increase in the number of RPN3/57 and CD11b-positive cells, respectively, infiltrating the muscle compared with EEX no-MLL animals (30.9 ± 7.7 vs 14.4 ± 2.4 for RPN3/57 and 29.3 ± 9.2 vs 8.8 ± 2.9 for CD11b, P = 0.28 and 0.33, respectively). There was a 39.3% and 366.0% increase in the number of RPN3/57 and CD11b-positive cells from immediate to delayed MLL protocols (P = 0.71 and 0.12, respectively) (Fig. 6A). There was a 38.9% and 7.9% decrease in the number of RAM11-positive macrophages between the EEX no-MLL (11.1 ± 2.6) and the immediate (6.8 ± 1.9, P = 0.45) and delayed MLL (10.3 ± 2.5, P = 0.88) protocols, respectively. Delayed MLL produced a 50.9% increase in the number of RAM11-positive macrophages as compared with immediate MLL, P = 0.54 (Fig. 6B).

Representative immunohistochemistry images for the control, delayed, and immediate protocols, showing florescence for both macrophage and neutrophil presence. Both control and delayed muscle showed a greater presence of RPN3/57 and CD11b-positive neutrophils and RAM11-positive macrophages as compared with immediate MLL.
A, B. Neutrophil (RPN3/57 and CD11b) infiltration per protocol and macrophage (RAM11) infiltration per protocol. Delayed MLL showed an increase in the number of RPN3/57 and CD11b-positive cells as compared with immediate MLL. There was a decrease in the number of RAM11-positive macrophages between the EEX no-MLL and the immediate and delayed MLL protocols. Delayed MLL showed an increase in the number of RAM11-positive macrophages as compared with immediate MLL.


This study shows using an in vivo animal model the ability of both immediate and 48 h delayed MLL to accelerate the recovery of muscle and joint function after a controlled bout of EEX. To our knowledge, we showed for the first time that there is a time-dependent effect of MLL on the recovery of mechanical properties as well as evidence for MLL to decrease muscle edema and modulate inflammatory cell infiltration. The immediate MLL protocol produced the greatest recovery of peak isometric torque as compared with both delayed MLL and EEX no-MLL protocols. Our animal model and self-designed device allowed the capability to produce repeatable deficits in muscle function after EEX and to apply quantifiable, repeatable compressive loads via a computerized, mechanical device to methodically follow recovery and healing. Similar experiments in humans will be important in determining the optimal use and indications for MLL and recovery of muscle function after intense EEX.

Our findings build on our previous work (8,18) by demonstrating that massage carried out immediately after a bout of intense EEX favors a quicker recovery of muscle and joint function than massage delayed by 48 h. Interestingly, in the current study, the immediate application of 0.5 Hz, 10 N, 15-min protocol produced an average RI of 0.94, similar to our previous study in which this set of loading conditions produced an RI = 1.08. The RI takes into account the loss of torque postexercise relative to preexercise and therefore is a more complete measure of recovery than peak isometric torque alone, building on our initial work where we only evaluated peak torque as a measure of recovery (8). The area under the TΘ curve was shown in the present study to be a secondary indicator of recovery with MLL from the bout of exercise. The same trend as was seen for peak torque recovery was seen using this measure with immediate MLL producing the greatest RI followed by delayed and control in order of average RI value. The current study, in combination with previous investigations in our laboratory, shows a highly reproducible exercise and injury protocol that allows us to systematically investigate several dependent variables using the smallest sample size possible (8,18).

Research has suggested that the inflammatory process is not only responsible for successful repair but also may act to extend the damage process (23,40). This progression of injury could occur through a variety of mechanisms, including neutrophil infiltration and activation of the respiratory burst (40,42). Invading leukocytes, neutrophils and macrophages, can exacerbate the initial mechanical damage and elicit unwanted tissue destruction to healing muscle through the release of oxygen-derived free radicals and proteases, which potentially cause injury (31,41). In the current study, muscle wet weight was measured as an indicator of tissue edema. Results showed a significant difference between the three protocols, with weight of the exercised control group higher than both the immediate and the delayed MLL groups. There was also a significant difference in wet weight between the immediate and the delayed MLL protocols, suggesting a difference in fluid dissipation from the exercised muscle dependent on timing of MLL application. Surprisingly, we did not observe statistical differences in leukocyte infiltration for either the immediate or delayed MLL conditions when compared with the nonmassaged controls. Despite a lack of statistical differences, the large percent differences (up to 366%) between the immediate MLL and the control conditions may still be of clinical significance; however, this is challenging to test using our animal model. Immediate MLL also produced a modulation of the immune response as compared with control (53.4% and 69.7% decrease for neutrophils and 38.9% for macrophages), again suggesting an effect on the host inflammatory response. It may be noteworthy that we used quantitative immunohistochemistry to definitely characterize the presence of both neutrophils and macrophages rather than simple hematoxylin and eosin staining used in our previous work (8).

Smith et al. (38) studied the effect of massage on inflammation and DOMS in humans, with results indicating that a 30-min massage protocol applied 2 h after the exercise interfered with neutrophil emigration. Crane et al. (12) found that a 10-min massage after a controlled bout of upright cycling in humans reduced pro-inflammatory cytokine expression while promoting mitochondrial biogenesis. Previous work in our laboratory has shown that four consecutive days of MLL led to an accelerated recovery of isometric torque production in addition to a decrease in the number of torn fibers (18). Additional studies have shown that MLL can modulate the passive stiffness of skeletal muscle with both an acute (daily) and a cumulative (4-d treatment) change in the muscle’s viscoelastic properties (19). These observations provide, to our knowledge, the first objective and quantifiable evidence to support the various assertions about massage and its physical effect on tissues (22). Evidence suggests that massage may work by affecting multiple mechanisms that cause DOMS, and it may be a combination of these factors that allow for the beneficial effect of this modality. In the current work, we chose to compare the 0-h (immediate) and the 48-h (delayed) strategies as a starting point to determine an optimal timing for use of massage. Given the findings noted herein, future studies may examine additional time points such as massage-like loading commenced 24 h after exercise to determine its effect on the recovery of muscle function and tissue inflammation and whether this strategy provides for an even more enhanced recovery. Moreover, possible biologic mechanisms to account for the accelerated recovery of muscle function such as reduced pro-inflammatory cytokine expression and myogenic up-regulation should be explored in our model.

The rightward shift in peak torque angle produced by the bout of EEX is consistent with previous studies from our laboratory and others (1,6,18). Herein, both MLL protocols resulted in a leftward shift from the postexercise peak torque angle, indicating some degree of muscle recovery from EEX (7). However, the shift in peak torque has been shown not to simply be an indicator of muscle damage but possibly a combination of both muscle damage and fatigue (1). Although delayed MLL produced the greatest leftward angular shift in the current study, there was no statistically significant difference from the immediate MLL angular shift. The clinical significance of this finding is not certain but does suggest a benefit to MLL.

Although we acknowledge that the injury created in our animal model may not be completely analogous to the injury produced in humans with EEX, the high torque deficits produced by the exercise in our model allow for differentiation of the effects of MLL and natural healing. In addition, in our model, we use maximum motor unit recruitment by stimulating the muscle to tetanus to determine the full extent of muscle damage by activating any motor unit that is capable of firing. Although this is not how human muscle functions, our results demonstrate the effects of timing of MLL on the recovery of muscle function, irrespective of motor unit recruitment and potential confounding variables such as pain and motivation. Moreover, there may be a difference between statistical and clinical implications of our findings. It is well documented that regional strains within exercised muscle result in focal damage and, therefore, regional and localized tissue inflammation (15). Although muscle tissue is a three-dimensional structure, the immunohistochemical analysis in the present study is based on a set of samples from a single slice of the muscle, which may not be representative of the entire muscle and therefore may not accurately characterize the nonnormal distribution of injury occurring throughout that three-dimensional structure. In addition, the nonnormal distribution of injury may cause the appearance of substantial variation within animal, thereby making statistical significance difficult to serve as a true representation of the biological response.

In conclusion, we have shown that skeletal muscle responds to MLL in a time-dependent fashion with the application of MLL immediately after damaging EEX attenuating the infiltration of neutrophils and macrophages in skeletal muscle and facilitating the recovery of function to near preexercise levels. The delayed application of MLL after EEX was less effective in restoring function and was associated with more edema and greater infiltration of immune cells. These data provide support for the immunomodulating effects of MLL and its effect on the recovery of function after muscle injury. This study allows greater understanding of how EEX-induced injury alters muscle function and the ability of quantifiable massage-like loading to both enhance muscle recovery and influence the host inflammatory response, suggesting further the need for human studies to confirm the efficacy of this modality in recovery from EEX.

The authors thank Scott Crawford and Ryan DeCoster. Research reported in this publication was supported by the National Center for Complementary and Alternative Medicine of the National Institutes of Health under award number R01AT004922. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

None of the authors have any professional relationships with companies or manufacturers that would benefit from the results of the present study.

Haas contributed in study design, experiments, analysis, and writing.

Butterfield contributed in experiments, analysis, and writing.

Abshire contributed in immunohistochemistry experimentation and analysis.

Zhao contributed in experiments, analysis, and writing.

Zhang contributed in study design, analysis, and writing.

Jarjoura contributed in study design, analysis, and writing.

Best contributed in study design, experiments, analysis, and writing.

Results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Agarwal SJ, Deschner P, Long A, et al. Role of NF-kappaB transcription factors in anti-inflammatory and proinflammatory actions of mechanical signals. Arthritis Rheum. 2004; 50: 3541–48.
2. Best TM, Hart L. A growing concern: the older athlete. Clin J Sport Med. 2008; 18 (6): 477–78.
3. Best TM, McCabe RP, Corr DT, Vanderby R. Evaluation of a new method to create a standardized muscle stretch injury. Med Sci Sports Exer. 1998; 30 (2): 200–5.
4. Bowers EJ, Morgan DL, Proske U. Damage to the human quadriceps muscle from eccentric exercise and the training effect. J Sports Sci. 2004; 22: 1005–14.
5. Brickson S, Ji LL, Schell K, Olabisi R, St Pierre B, Best TM. M1/70 attenuates blood-borne neutrophil oxidants, activation, and myofiber damage following stretch injury. J Appl Physio. 2003; 95: 969–76.
    6. Brockett CL, Morgan DL, Proske U. Human hamstring muscles adapt to eccentric exercise by changing optimum length. Med Sci Sports Exerc. 2001; 33 (5): 783–90.
    7. Butterfield TA, Herzog W. The magnitude of muscle strain does not influence serial sarcomere number adaptations following eccentric exercise. J Appl Physiol. 2005; 100: 1489–98.
    8. Butterfield TA, Zhao Y, Agarwal S, Haq F, Best TM. Cyclic compressive loading facilitates recovery after eccentric exercise. Med Sci Sports Exerc. 2008; 40 (7): 1289–96.
    9. Cesari M, Penninx BW, Pahor M, et al. Inflammatory markers and physical performance in older persons: the InCHIANTI study. J Gerontol A Biol Sci Med Sci. 2004; 59 (3): 242–48.
    10. Clarkson PM, Tremblay I. Exercise-induced muscle damage, repair and adaptation in humans. J Appl Physiol. 1988; 65 (1): 1–6.
    11. Cleak MJ, Eston RG. Muscle soreness, swelling, stiffness and strength loss after intense eccentric exercise. Br J Sports Med. 1992b; 26: 267–72.
    12. Crane J, Ogborn D, Cupido C. Massage therapy attenuates inflammatory signaling after exercise-induced muscle damage. Sci Transl Med. 2012; 4: 119ra13.
    13. Ebbeling CB, Clarkson PM. Exercise-induced muscle damage and adaptation. Sports Med. 1989; 7: 207–34.
    14. Fridén J. Muscle soreness after exercise: implications of morphological changes. Int J Sports Med. 1984; 5: 57–66.
    15. Fridén J, Sjöström M, Ekblom B. A morphological study of delayed muscle soreness. Experientia. 1981; 37 (5): 506–7.
    16. Galloway S, Watt J, Sharp C. Massage provision by physiotherapists at major athletic events between 1987 and 1998. Br J Sport Med. 2004; 38 (2): 235–37.
    17. Gleeson M, Almey J, Brooks S, Cave R, Lewis A, Griffiths H. Haematological and acute-phase responses associated with delayed-onset muscle soreness in humans. Eur J Appl Physiol. 1995; 71: 137–42.
    18. Haas C, Butterfield TA, Zhao Y, Zhang X, Jarjoura D, Best TM. Dose-dependency of massage-like compressive loading on recovery of active muscle properties following eccentric exercise: rabbit study with clinical relevance. Br J Sports Med. 2013; 47: 83–88.
    19. Haas C, Best TM, Wang Q, Butterfield TA, Zhao Y. In vivo passive mechanical properties of skeletal muscle improve with massage-like loading following eccentric exercise. J Biomech. 2012; 45 (15): 2630–36.
    20. Herman PM, Craig BM, Caspi O. Is complementary and alternative medicine (CAM) cost-effective? A systematic review. BMC Complement Altern Med. 2005; 5: 11.
    21. Jonhagen S, Ackermann P, Eriksson T, Saartok T, Renstrom P. Sports massage after eccentric exercise. Am J Sports Med. 2004; 32 (6): 1499–1503.
    22. Kennedy AB, Blair SN. Massage therapy for those who exercise [Internet]. 2011 [cited 2012 May 1]. Available from:
    23. Ley K. Healing without inflammation? Am J Physiol Regul Integr Comp Physiol. 2003; 285: R718–19.
    24. McAnulty S, McAnulty L, Nieman D, Morrow J, Dumke C, Henson D. Effect of NSAID on muscle injury and oxidative stress. Int J Sports Med. 2007; 28 (11): 909–15.
    25. McHugh MP, Connolly DEJ, Eston RG, Gleim GW. Electromyographic evidence of exercise resulting in symptoms of muscle damage. J Sports Sci. 2000; 18: 163–72.
    26. Miles MP, Ives JC, Vincent KR. Neuromuscular control following maximal eccentric exercise. Eur J Appl Physiol. 1997; 76: 368–74.
    27. Moraska A. Therapist education impacts the massage effect on postrace muscle recovery. Med Sci Sports Exer. 2007; 39 (1): 34–7.
    28. Ogilvie RW, Armstrong RB, Baird KE, Bottoms CL. Lesions in the rat soleus muscle following eccentrically biased exercise. Am J Anat. 1988; 182: 335–46.
    29. Omoigui S. The biochemical origin of pain: the origin of all pain is inflammation and the inflammatory response. Part 2 of 3—inflammatory profile of pain syndromes. Med Hypotheses. 2007; 69: 1169–78.
    30. Peake J, Nosaka K, Suzuki K. Characterization of inflammatory responses to eccentric exercise in humans. Exerc Immunol Rev. 2005; 11: 64–85.
    31. Pizza FX, Koh TJ, McGregor SJ, Brooks SV. Muscle inflammatory cells after passive stretches, isometric contractions, and lengthening contractions. J Appl Physiol. 2002; 92: 1873–8.
    32. Pratt M, Macera CA, Wang G. Higher direct medical costs associated with physical inactivity. Phys Sportsmed. 2000; 28: 63–70.
    33. Schaap LA, Pluijm SM, Deeg DJ, Visser M. Inflammatory markers and loss of muscle mass (sarcopenia) and strength. Am J Med. 2006; 119 (6): 526e9–17.
    34. Sellwood KL, Brukner P, Williams D, Nicol A, Hinman R. Icewater immersion and delayed-onset muscle soreness: a randomized controlled trial. Br J Sports Med. 2007; 41 (6): 392.
    35. Sherman K, Ezzo J, Khalsa. Massage therapy: an introduction [Internet]. 2010 [cited 2012 May 1]. Available from:
    36. Smet EG, Smet WD, Brys L, Baetselier PCD. A monoclonal antibody recognizing the complement type 3 receptor (CD3) in the rabbit. Immunology. 1986; 59: 419–25.
    37. Smith LL, Brunetz MH, Chenier TC, et al. The effects of static and ballistic stretching on delayed onset muscle soreness and creatine kinase. Res Q Exerc Sport. 1993; 64 (1): 103–7.
    38. Smith LL, Keating MN, Holbert D, et al. The effects of athletic massage on delayed onset muscle soreness, creatine kinease, and neutrophil count: a preliminary report. J Orthop Sports Phys Ther. 1994; 19: 93–9.
    39. Stay JC, Richard MD, Draper DO, Schulthies SS, Durrant E. Pulsed ultrasound fails to diminish delayed-onset muscle soreness symptoms. J Ath Train. 1998; 33 (4): 341–6.
    40. Tidball JG. Inflammatory processes in muscle injury and repair. Am J Physiol Regul Integr Comp Physiol. 2005; 288: R345–53.
    41. Toumi H, Best TM. The inflammatory response: friend or enemy for muscle injury? Br J Sports Med. 2003; 37 (4): 284–6.
    42. Toumi H, F’guyer S, Best TM. The role of neutrophils in injury and repair following muscle stretch. J Anat. 2006; 208: 459–70.
    43. Tourville TW, Connolly DA, Reed BV. Effects of sensory-level high-volt pulsed electrical current on delayed-onset muscle soreness. J Sports Sci. 2006; 24 (9): 941–9.
    44. Warren GL, Hayes DA, Lowe DA, Prior BM, Armstrong RB. Materials fatigue initiates eccentric contraction induced injury in rat soleus muscle. J Physiol. 1993; 464: 477–89.
    45. Weerapong P, Hume P, Kolt G. The mechanisms of massage and effects on performance, muscle recovery and injury prevention. Sports Med. 2005; 35 (3): 235–56.
    46. Wilkinson JM, Galea-Lauri J, Sellars RA, Boniface C. Identification and tissue distribution of rabbit leucocyte antigens recognized by monoclonal antibodies. Immunology. 1992; 76: 625–30.
    47. Yumet G, Shumate ML, Bryant P, Lin CM, Lang CH, Cooney RN. Tumor necrosis factor mediates hepatic growth hormone resistance during sepsis. Am J Physiol Endocrinol Metab. 2002; 283: E472–81.
    48. Zeng H, Butterfield TA, Agarwal S, Haq F, Best TM, Zhao Y. An engineering approach for quantitative analysis of the lengthwise strokes in massage therapies. J Med Dev. 2008; 2: 041003–1.


    ©2013The American College of Sports Medicine