The aim of this study was to investigate whether, and to what degree, exercise-induced muscle damage will increase decompression strain, as indicated by increased vascular bubble formation, during a hypobaric decompression. The results demonstrated that muscle damage induced by eccentric arm work 24 h prior to ascent from sea level to 24,000 ft, significantly increased the formation of decompression bubbles during rest and arm-flex provocation, but not during knee-bend provocation. Furthermore, there was a significantly earlier onset of first VGE in the ECC exposures.

Strength loss after eccentric exercise is considered to be a valid and reliable marker of muscle damage (Warren et al. 1999). Thus, our finding that the eccentric exercise bout reduced MVC immediately and 24 h after exercise (reduction in MVC to 63% and 79% from baseline, respectively) is in line with other studies on eccentric exercise of the elbow flexors (Jones et al. 1987; Newham et al. 1987; Stauber et al. 1990) and implies that the exercise bout indeed induced microscopic muscle injury. One of the cardinal symptoms associated with tissue injury and inflammation is pain, and hence, the subjects reported delayed-onset muscle soreness in several upper body muscles after the exercise, which corroborates the notion of EIMD and suggests that it was not limited to the biceps brachii muscles. It is well documented that the overstretching of sarcomeres during high-intensity eccentric muscle actions results not merely in a drop in MVC, but also in a local inflammatory response characterized by DOMS (Proske and Morgan 2001). The choice to perform exercise 24 h prior to a hypobaric exposure was based on the fact that most literature on eccentric exercise has shown mechanical (Proske and Morgan 2001; Fridén et al. 1981), inflammatory (MacIntyre et al. 1995) and microvascular (Hotta et al. 2018) changes within this time window. Whether or not the decompression stress is maximum during that point in time has to be explored further.

Seven days might not have been enough to completely restore the EIMD; however, it appears from indirect markers of muscle damage (i.e. drop in strength, soreness and biomarkers) that the affected muscle will almost, if not completely, recover by day 7 after a maximal eccentric bout (MacIntyre et al. 1995; Jones et al. 1987). In the present study, the interval between the exposures carried out in the order CNT–ECC was (mean (range)) 12 (5–28) days and for ECC–CNT order it was 16 (8–28) days. In addition, the orderly sequence CNT/ECC exposure was alternated in a balanced fashion, securing that any carryover effect between trials was balanced between conditions.

That both the peak and time-integrated VGE scores were higher in the ECC than the CNT altitude exposure suggests that EIMD increased the formation of VGE during decompression. There appears to be no similar controlled experimental study on humans reported in the literature, but a possible relationship between prior musculoskeletal injury and susceptibility to DCS has been discussed in the literature concerning hypobaric decompression (Fryer 1969; Adler 1964). Thompson et al. tested the relationship between old injury and DCS susceptibility by collecting data on serious injuries of the extremities (i.e. fractures, sprains, etc.) in flight crews, prior to their first hypobaric exposure. In a group of 1220 men, 327 gave a history of a fracture or other injuries to a limb (Thompson et al. 1944). Upon decompression to altitude, merely 18% of these men developed pain in proximity to the location of the old injury, and the authors concluded that there was little, if any, relationship between old injury and the localization of DCS. It is, however, difficult to interpret the results of this study (Thompson et al. 1944), since there was no account of the DCS incidence in the control group (893 men without a previous serious limb injury). In the same report, it was concluded that DCS pain tended to localize more frequently in a region which had been subjected to a recent minor injury. Subjects given a blow on the tibia prior to being exposed to 35,000 ft, experienced increased incidence of bends in the same region (17% vs 8% in the control group) (Thompson et al. 1944). Houston et al. studied 1538 man-ascents to altitude and determined the coincidence of bends at the site of previous injuries to be 7.2%, which was considered a significant relationship (Houston et al. 1944). Notably, even though the occurrence of DCS correlates to the prevalence of decompression-induced VGE, comparison of results must be done with caution, between studies using DCS (as in the aforementioned studies) vs VGE (present study) as the main effect variable.

Light aerobic exercise in combination with preoxygenation prior to hypobaric decompression has been shown to significantly reduce the incidence of DCS compared to resting preoxygenation (Webb et al. 1996; Hankins et al. 2000; Loftin et al. 1997). Since the elimination and uptake of nitrogen is a perfusion-limited process, this additive effect is attributed to augmented denitrogenation resulting from increased blood flow through the exercising muscles (Webb et al. 1996). Indeed, moderate concentric arm or leg exercise performed during decompression from hyperbaric pressure has also been found to reduce VGE (Jankowski et al. 1997, 2004). However, exercise without preoxygenation (150 knee flexes over 10 min) just prior to depressurization led to significant increases in bubble formation (Dervay et al. 2002). This effect is probably unrelated to muscle damage and more likely, as the authors discuss, is due to generation of new micronuclei or enlargement of existing micronuclei. At altitude, deep knee bends followed by weighted arm extensions have been found to decrease time to maximum VGE score and to increase the incidence of DCS compared to sedentary conditions (Krutz and Dixon 1987). Similarly, isometric and dynamic arm and leg exercises induce DCS at otherwise considered symptom-free altitudes (Pilmanis et al. 1999). The increase in VGE and DCS is probably due to production of new bubbles, similar to what is observed following knee flexes prior to decompression.

Regarding exercise prior to hyperbaric exposure, it has been shown to reduce (Jurd et al. 2011; Blatteau et al. 2005; Dujic et al. 2004) or have no effect on circulating bubbles (Gennser et al. 2012) upon decompression. The results seem to vary depending on the type, duration and intensity of the exercise, as well as on how long before decompression the exercise was performed. The protocols used in these studies consisted of aerobic exercise (running or cycling), dominated by concentric muclse contractions, known to induce considerably less muscle damage than eccentric actions (Lavender and Nosaka 2006; Peñailillo et al. 2013). Presumably, these exercise regimens did not inflict substantial muscle damage, but rather, may have served to, possibly by way of increased peripheral blood flow, remove precursor micronuclei adhering to the endothelial walls (Vann et al. 1980; Arieli and Marmur 2017).

Also, with regard to studies in experimental animals, information is scarce and equivocal on effects of muscle damage on the susceptibility to decompression stress. Thus, Jørgensen et al., showed that eccentric exercise performed by rats prior to simulated diving had no effect on bubble formation (Jørgensen et al. 2015, 2013). Harvey, on the other hand, found that, in cats, skeletal muscle injury inflicted shortly prior to hypobaric decompression increased the presence of circulating VGE (Harvey 1945). The reason for the discrepancy between the results of the two animal studies may well be that the intervention by Jörgensen et al. was considerably milder than the one by Harvey. Jørgensen et al. used downhill running to induce eccentric exercise and hence muscle damage in the rats, whereas, Harvey inflicted damage in the thigh muscles by applying local pressure (squeezing them). The difference between present results and those by Jørgensen, remains to be settled. At least in humans, downhill running is associated with considerably less muscle damage than high-force eccentric actions (Clarkson and Hubal 2002).

The question arises as to the mechanisms underlying the augmented VGE prevalence in the ECC trial. It is rather well established that, in vivo, decompression VGE, and in particular altitude-induced VGE, originate from already pre-existing micronuclei attached to the endothelial walls (Christman et al. 1986; Lee et al. 1993; Vann et al. 1980; Arieli and Marmur 2011), since spontaneous (de novo) formation of gas bubbles in a solution with dissolved gas requires a marked supersaturation corresponding to a pressure differential of about 10.0 MPa (1000 msw) (Jones et al. 1999).

Our observation that there was a significant difference of bubble load in the ECC condition during rest indicates a humoral effect. However, the significant difference after arm flex but not leg flexion suggests that the general increase in VGE results from increased bubble formation locally in muscle-damaged regions. This assumption, as well as the extent to which the importance of total affected muscle mass, will have to be confirmed in future studies. Peak EB and KISS scores after knee flexions were the same in both conditions in five participants, resulting in no significant difference between conditions (Table 1 and Fig. 3). During deep knee bends, big muscle groups of the legs and the trunk are activated, and consequently, knee bends are usually followed by high VGE scores during prolonged exposure to 24,000 ft (Ånell et al. 2020; Elia et al. 2021). VGE scores in the CNT condition were similar to those seen in previous studies with a similar protocol of 24,000 ft for 60–90 min (Ånell et al. 2019, 2020; Elia et al. 2021), giving rise to peak bubble scores of 2–3 on the EB scale. There was also a substantially earlier occurrence of first VGE, regardless of grade, in the ECC than in the CNT condition. Conversely, Elia et al. found a significant latency before the first VGE was detected in a preconditioning study consisting of 30-min whole-body vibration prior to hypobaric decompression. They also found lower maximum VGE scores in the pre-exposure vibration condition than in the control condition (Elia et al. 2021). It thus seems that, depending on whether an intervention generates more or fewer bubbles, it also affects the onset of VGE.

Assuming that, following eccentric exercise, VGE is increased locally in the damaged muscle, then one might, albeit speculatively, consider a few mechanisms that might contribute to local formation of micronuclei. Firstly, analyses of biopsy samples after eccentric exercise have shown possible damage and disturbance to capillaries (Stauber et al. 1990) and in vitro experiments suggest microvascular hyperpermeability during EIMD (Hotta et al. 2018). Conceivably, the ECC-induced increase of VGE could be due to increased leakage of micronuclei from affected muscle to vessels, perhaps in combination with increased intramuscular formation of microbubbles due to inertial cavitation induced by the traction of sarcomeres (Kim et al. 2021). In addition, early in the recovery period, neutrophiles propagate an inflammatory response by secretion of cytokines. In the following days, both proinflammatory and anti-inflammatory cytokines are secreted, acting to clear damaged tissue and initiate regeneration of muscle fibres (Peake et al. 2017; Moldoveanu et al. 2001). Possibly, these molecules could coat the microbubbles, reducing the surface tension of the bubble by reducing the attractive forces between the gas molecules in the bubble. This would lead to a greater influx of nitrogen in existing microbubbles. Lastly, in vitro experiments have shown that on a completely hydrophobic surface (e.g. glass), the zero-contact angle means that no bubble can stick or remain, whereas with any positive contact angle gas nuclei can stick and be stable. EIMD might change the local structure on endothelia tissue, by increased areas of depressions and cavity, creating more positive contact angles for micronuclei to adhere. Nuclei might then enlarge and break away from the surface as a bubble, which continue to grow.

In summary, eccentric exercise induces muscle damage and the literature is unanimous in that this leads to local disruption of tissue and capillaries, with subsequent systemic inflammation. We hypothesize that this will increase the presence of gaseous micronuclei. When and for how long it is possible to observe increased bubble production following eccentric exercise, and whether the increase in VGE occurrence is due to local or humoral effects, remain to be settled.

The study was not performed it in a double- or single-blinded fashion and hence was limited by the fact that not only the subjects, but also the experimenters were aware of at what time point they had performed exercise. This was deemed a necessary safety precaution, to avoid confusing DOMS with DCS. On the other hand, the same subjects were used in both conditions, allowing for intra-individual comparison. The same ultrasound operator was present in all chamber runs and the scoring was verified after each experiment by another experienced sonographer.

EIMD increases the presence of circulating VGE; it is however unclear if it also increases the risk of DCS, although a large number of bubbles and high grades predispose to DCS (Francis and Mitchell 2003). To date, there is no recommendation regarding flying or diving after injury or EIMD. However, the results in the present study makes this question worthy of being explored further. Specifically, it will be interesting to examine whether, and to what extent, addition of muscle groups would further increase/advance VGE production. Additionally, it would be of interest to explore whether EIMD following simulated dive in a hyperbaric chamber increases circulating VGE. If so, caution might be needed both when planning diving and flying decompression procedures.

In conclusion, the present study demonstrates that 15 min of eccentric upper body exercise 24 h prior to decompression significantly increases the prevalence of VGE and advances the generation of VGE during a continuous 90-min exposure at 24,000 ft compared to a control flight in the same individuals.