Effect of rapid weight loss incorporating hot salt water immersion on changes in body mass, blood markers, and indices of performance in male mixed martial arts athletes

In sports that have weight class restrictions, athletes attempting to ‘make weight’ frequently practice short-term weight loss, termed acute or rapid weight loss (RWL), in the ~ 72 h before weigh-in (Reale et al. 2017; Burke et al. 2021). Methods of RWL focus on the acute reduction of the contents of the gastrointestinal tract, and of total body water, through methods such as a low carbohydrate and low-residue diet, fluid restriction, and increasing sweat losses through exercise and/or heat exposure (Barley et al. 2018a; Hillier et al. 2019; Connor and Egan 2019). In mixed martial arts (MMA) athletes, combinations of these methods typically induce losses of ~ 5% to ~ 10% of body mass in the week before weigh-in (Coswig et al. 2015, 2019; Matthews and Nicholas 2017; Barley et al. 2018a; Brechney et al. 2019; Connor and Egan 2019; Hillier et al. 2019).

One method of heat exposure used to induce passive fluid loss is hot water immersion (HWI), which is often employed as part of weight-making strategies in combat sports (Pettersson et al. 2013; Matthews and Nicholas 2017; Brandt et al. 2018; Connor and Egan 2019; Kasper et al. 2019; Park et al. 2019; Gordon et al. 2021). Colloquially known as hot baths, this method typically involves short-duration HWI followed by ‘wrapping’ in warm clothing for a period of time before further exposures to HWI and wrapping (Connor et al. 2020; Kasper et al. 2019). The HWI is often performed in salt water, usually by the addition of Epsom salt, with the aim of augmenting the loss of body mass compared to that achieved by immersion in fresh water (Connor et al. 2020). Indeed, there is some empirical evidence for salt water immersion to augment fluid loss when compared to fresh water immersion (Whitehouse et al. 1932; Hertig et al. 1961; Hope et al. 2001), albeit these studies involve prolonged (≥ 3 h) HWI. The mechanistic basis for this effect is proposed as the addition of salt increasing the osmotic pressure difference between the immersion medium and body fluids, and/or attenuating the inhibitory effect on sweating that occurs during prolonged HWI, and thereby resulting in the greater fluid loss compared to fresh water (Whitehouse et al. 1932; Buettner 1953, 1959; Peiss et al. 1956; Hertig et al. 1961; Brebner and Kerslake 1964; Hope et al. 2001).

Our recent work has investigated a hot bath protocol incorporating short-duration (2 × 20 min) HWI with an Epsom salt concentration of ~ 1.6%wt/vol, but found no difference in body mass losses comparing fresh water and salt water immersion either at a fixed water temperature of 37.8 ºC (Connor et al. 2020), or when athletes self-adjusted the water temperature to the maximum temperature each could tolerate (~ 39.0 ºC) (Connor and Egan 2021). Our choice of a ~ 1.6% salt solution using Epsom salt was chosen for its ecological validity based on our knowledge of applied practice, and responses to a questionnaire in which fighters described typically the addition of 1 to 2 kg of Epsom salt to a standard sized bath (Connor et al. 2020). However, higher concentrations of salt, which would induce a larger osmotic gradient between the bath water and body fluids, may be required to augment fluid loss when compared to fresh water. For example, even in thermoneutral water, i.e., in the absence of sweating, immersion in a strong salt solution (either 11.5 or 20.0% salt as sodium chloride) induced passive fluid loss (Whitehouse et al. 1932). In water heated to 36/37 ºC, addition of 5% sodium chloride (~ 1709 mOsmol/kg) allowed for higher sweat rates during 3 h of immersion when compared to fresh water, with the effect more pronounced at salt concentrations of 10 and 15% (Hertig et al. 1961). Finally, immersion in seawater (~ 3.5% salt; ~ 1000 mOsmol/kg to ~ 1200 mOsmol/kg) resulted in ~ 32% greater loss of body mass compared to fresh water during 4 h of immersion at ~ 38 ºC (Hope et al. 2001). Given these observations, it may be that the concentration of salt in a hot bath should be at least 3.5% (Hope et al. 2001), or possibly greater (Buettner 1953, 1959; Hertig et al. 1961), if the aim is to augment the rate of passive fluid loss that would otherwise occur during HWI in fresh water.

Therefore, the present study investigated the magnitude of body mass losses in MMA athletes using a hot bath protocol with immersion in fresh water or salt water at a concentration of ~ 5%wt/vol of Epsom salt. Extending our previous work (Connor et al. 2020; Connor and Egan 2021), we also investigated the effects of ~ 28 to ~ 30 h of RWL on blood markers (plasma volume, kidney function, and electrolytes) and indices of performance. Hypohydration is well established as negatively impacting indices of performance (Savoie et al. 2015; Deshayes et al. 2020), and importantly, recent evidence suggests this to be the case even when major confounders such as expectancy effects because of a lack of blinding, and inadequate familiarization with methods of dehydration, are addressed (James et al. 2019). While RWL inherently involves dehydration processes, in professional combat sports such as MMA and boxing the competitive bout typically takes place ~ 30 to ~ 36 h after weigh-in (Burke et al. 2021). This time-period may allow for rehydration and recovery of muscle glycogen (Burke et al. 2021), but several studies have observed a residual negative impact on indices of performance even after ~ 16 to ~ 24 h of recovery (Oöpik et al. 1996; Moghaddami et al. 2016; Alves et al. 2018; Yang et al. 2018; Barley et al. 2018b; Kurylas et al. 2019). Therefore, we also investigated the effect of RWL followed by ~ 24 to ~ 26 h of recovery in the form of a high carbohydrate diet and rehydration on body mass, hydration status, blood markers, and indices of performance.

Given that our previous work using ~ 1.6% salt solutions did not reveal an effect of salt to augment body mass loss during a hot bath protocol (Connor et al. 2020; Connor and Egan 2021), the present study investigated body mass losses when the salt concentration is increased to ~ 5.0%wt/vol. This higher concentration is more similar to immersion studies where an effect of salt to augment the loss of fluid and/or body mass has been observed (Whitehouse et al. 1932; Hertig et al. 1961; Hope et al. 2001). However, the present study demonstrates that the body mass lost during a hot bath protocol using fresh water (FWB) is similar to a protocol using ~ 5.0%wt/vol of Epsom salt (SWB).

Body mass losses induced by ~ 26 to ~ 28 h of restriction of fluid intake combined with a low-residue, low-carbohydrate diet were ~ 2.6% of body mass. This is similar in magnitude to the suggestion of ~ 3% reduction to be expected by short-duration restriction of carbohydrate and fluid, and emptying of the gastrointestinal contents using a low-residue diet (Reale et al. 2017; Burke et al. 2021), and is also similar to our previous work (Connor et al. 2020; Connor and Egan 2021). However, the percentage of body mass lost during the entire RWL process was greater in the present study at ~ 5.3% compared to those previous studies where ~ 4.3% (Connor et al. 2020) and ~ 4.5% (Connor and Egan 2021) were observed. The larger magnitude is explained by differences in percentage of body mass lost in the bathing protocol, which was ~ 2.7% loss of body mass in this study compared to ~ 2.1% in the other studies. This difference may simply reflect inter-individual differences in RWL between studies. Alternatively, commencing bath in water of higher temperature (e.g. ~ 40.3 ºC versus 37.8 ºC), and using a sauna blanket for the wrap periods rather than cotton clothing in a warm room, may result in more efficient loss of body mass per unit of time invested in such a protocol.

Together, these findings across three studies suggest that the addition of salt to HWI does not augment the loss of body mass compared to fresh water, at least in the hot bath protocol employed. The caveat that this conclusion only applies to the hot bath protocol employed is important, because several prior studies do indeed demonstrate an effect of salt to augment immersion-induced loss of fluid and/or body mass in various experimental models including whole-body immersion, and localized immersion of an arm/hand or leg/foot (Whitehouse et al. 1932; Buettner 1953, 1959; Peiss et al. 1956; Hertig et al. 1961; Brebner and Kerslake 1964; Hope et al. 2001). There are two suggested mechanisms for this phenomenon. First, that during immersion in salt water, the osmotic pressure difference between the immersion medium and body fluids results in greater fluid loss compared to fresh water, and/or second, that salt water serves to attenuate an inhibitory influence on the decline in sweat rate that usually occurs with prolonged immersion in hot fresh water (Whitehouse et al. 1932; Buettner 1953, 1959; Peiss et al. 1956; Hertig et al. 1961; Brebner and Kerslake 1964; Hope et al. 2001).

The absence of an effect of salt in our work may be explained by duration of immersion being much shorter than those previous studies observing an effect. For example, those studies have used immersion times of 3 h (Hertig et al. 1961), 4 h (Hope et al. 2001) and 5 h (Whitehouse et al. 1932; Brebner and Kerslake 1964). Despite our protocol comprising of 2 h of passive heating, HWI only accounts for 2 × 20 min of this time-period. Hope et al. (2001) observed a difference of ~ 600 g of body mass lost in 4 h when comparing immersion in fresh water to salt water (sea water) at 38ºC (Hope et al. 2001). This difference between conditions would be the equivalent of ~ 2.5 g per minute assuming linearity in the response. Translating this rate into our 40 min of total time spent immersed in water would result in an expected difference of just 100 g between FWB and SWB. Therefore, for the addition of salt to have the desired impact of augmenting loss of body mass through passive fluid loss, much longer immersion times than the 2 × 20 min employed in this study may be required.

Another consideration, however, is the osmolality of the salt water given the proposed mechanism around the osmotic pressure difference between the immersion medium and body fluids. While the %wt/vol of salt is most commonly used as the descriptor of the salt water condition, the osmolality will be a function of both the concentration and type of salt. Our previous work using 1.6%wt/vol of Epsom salt (Connor et al. 2020; Connor and Egan 2021), and the present study using 5.0%wt/vol of Epsom salt, would result in an osmolality of ~ 130 and ~ 406 mOsmol/kg, respectively, which would decline somewhat with the addition of boiling water to maintain or increase the water temperature while bathing. Thus, these salt water baths were, respectively, hypotonic and only mildly hypertonic relative to the osmolality of body fluids (i.e., ~ 280 to ~ 295 mOsmol/kg). In contrast, when augmented body mass losses have been previously observed, these salt water baths were markedly hypertonic, i.e., 5%wt/vol of sodium chloride (Hertig et al. 1961) being ~ 1709 mOsmol/kg, and seawater (Hope et al. 2001) being ~ 3.5% salt and ~ 1000 to ~ 1200 mOsmol/kg. Therefore, while Epsom salt was used for its ecological validity, a salt such as sodium chloride may be more effective on a %wt/vol basis. Alternatively, Epsom salt would need to be used at > 12.3%wt/vol to produce an osmolality of > 1000 mOsmol/kg. These points assume that the osmotic gradient is an important mechanism by which salt water augments loss of body mass during HWI, and tentatively suggest that > 1000 mOsmol/kg is a valid threshold above which these effects would be observed.

The present study extends our previous work by measuring heart rate during the hot bath protocol, and measuring changes in blood markers during the RWL process, in addition to investigating effects of the RWL followed by ~ 24 to ~ 26 h of recovery on indices of performance. The heart rate data during the hot bath protocol demonstrate that a moderate degree of cardiovascular stress was induced as indicated by heart rate averaging ~ 110 bpm throughout the 2 h period and a measured peak at 128 ± 19 bpm and 127 ± 21 bpm during FWB and SWB, respectively. These values are equivalent to ~ 68% of the participants’ age-predicted maximum heart rate.

The concentrations of several analytes in blood were increased during the hot bath protocol. Specifically, in blood samples taken immediately after the 2nd wrap period, concentrations of BUN, chloride, creatinine, hemoglobin, and sodium were each increased, as was the hematocrit value, but each returned to values similar to baseline by weigh-in on Day + 1. Calculation of plasma volume from hemoglobin and hematocrit revealed an average decrease in plasma volume induced by RWL of ~ 14% when measured upon completion of the 2nd wrap. This value is somewhat greater than that observed by (Hope et al. 2001) of ~ 7 to ~ 12%, but perhaps unsurprising given that the overall loss of body mass during the RWL protocol in the present study was approximately double of that previous work. In contrast, when elite amateur boxers undertook RWL in which a similar quantity of body mass was lost (5.6 ± 1.7%), the reduction in plasma volume was smaller at 8.6 ± 3.9% (Reljic et al. 2013). In that study, the RWL process was over a 5 day period, which potentially suggests that a shorter time frame of RWL and/or exposure to HWI may lead to greater loss of plasma volume or differential effects on different compartments of body water. There are two caveats that apply to the interpretation of these data for plasma volume. First, the i-STAT blood analyzer derives the value for hemoglobin using a proportionality constant after the measurement of hematocrit by a conductometric method, so the plasma volume data are based on an estimation of hemoglobin rather than direct measurement. Alternatively, using hematocrit only and thereby calculating loss of blood volume (Dill and Costill 1974), RWL resulted in a decrease in blood volume by ~ 8% in both conditions upon completion of the 2nd wrap. Second, postural changes are known to acutely influence measures of plasma volume (Pivarnik et al. 1986; Lippi et al. 2015), and a reduction in plasma volume of ~ 4.8% was previously observed within the initial 5 min after moving from a supine to seated (Pivarnik et al. 1986). Although the seated posture and rest period was consistent before blood sampling on Day − 2, before the 1st bath on Day 0, and on Day + 1, the sample taken upon completion of the 2nd wrap was preceded by 40 min in a supine position and only ~ 3 min of equilibration in a seated position. Therefore, the change in posture from a supine to seated position may have also contributed to decrease in plasma volume observed in response to the hot bath protocol.

Also of note is the observation of increased BUN and creatinine concentrations as these are often used as biomarkers of acute kidney injury (AKI) (Edelstein 2008; Kellum and Lameire 2013; Ostermann et al. 2020). RWL of > 4% of body mass consistently results in an increase in BUN and creatinine, which has been suggested as an indication of AKI being caused by RWL (Lakicevic et al. 2021). AKI has been previously defined as an increase in serum creatinine concentration by ≥ 0.3 mg/dL within 48 h (Kellum and Lameire 2013), a threshold which just two of our participants exceeded and which occurred within the 2 h bathing period. Additionally, the utility of circulating BUN and creatinine concentrations as sensitive and specific markers of AKI has been questioned (Edelstein 2008; Ostermann et al. 2020), whereas traditional measures of AKI are limited in their ability to classify AKI during heat stress, especially when combined with dehydration and/or exercise (Chapman et al. 2021). BUN and creatinine concentrations are indirect measures of AKI rather than direct measures of tissue injury such as with creatine kinase and cardiac troponin from skeletal muscle and heart, respectively. Direct measures of AKI in the circulation remain to be firmly established, especially those that can differentiate between ‘pre-renal’ and ‘intrinsic’ causes of change in circulating markers (Edelstein 2008; Ostermann et al. 2020; Chapman et al. 2021). Moreover, changes in BUN and creatinine concentrations are generally delayed in their response to AKI rather than acutely responsive (Edelstein 2008; Ostermann et al. 2020). Our data indicate an acute response in creatinine concentration to increase over the 2 h bathing period, whereas BUN concentration was already increased after the diet and fluid restriction, and increased further during bathing. Hence, it remains unclear whether these increases can indeed be considered to be evidence of AKI, or whether these simply reflect the well-established hemoconcentration effect of an acute decrease in plasma volume (Harrison 1985). In favor of the former is that it is well established that heat stress, especially when combined with physical exertion, can result in AKI (Chapman et al. 2021). Especially relevant to the present study is that heat stress-associated AKI is also influenced by hydrostatic pressure of water when HWI is used to apply the heat stress in experimental contexts (Chapman et al. 2021). Therefore, changes in markers of AKI during RWL and comprehensive assessment of kidney function should continue to be investigated by future research to better understand this phenomenon given its implications for the welfare of athletes who repeatedly undertake RWL.

Immediately before the performance testing on Day + 1 represented a ~ 24 to ~ 26 h recovery period at which point only 4 (FWB trial) and 2 (SWB trial) participants remained in a body mass deficit compared to Morning Day − 1. On average, there was a body mass surplus of 0.47 ± 1.48 kg and 0.69 ± 0.83 kg compared to Morning Day − 1 after recovery from FWB and SWB, respectively. This surplus is in contrast to the deficit observed on average in our previous work (Connor et al. 2020; Connor and Egan 2021), but is explained by the ~ 2 to ~ 4 h longer recovery time in the present study due to the inclusion of the performance tests. Therefore, despite the loss of ~ 5.3% of body mass in ~ 28 to ~ 30 h, blood markers had returned to values similar to baseline after ~ 24 to ~ 26 h of recovery. In practice, the time from weigh-in until official competition in professional MMA is usually longer, i.e., ~ 30 to ~ 36 h, but even with a longer time-period for rehydration, the majority of MMA athletes have been observed to be hypohydrated up to 2 h before competition (Jetton et al. 2013; Matthews and Nicholas 2017). Based on these observations, regain of body mass alone was suggested as potentially not being a good indicator of returning to a euhydrated state, but there is some debate about the validity of the classification of hypohydration through assessment of hydration status by spot analysis with urine measures (Cheuvront et al. 2015; Barley et al. 2020). For example, an alternative to the criterion of urine osmolality of > 700 mOsmol/kg being classified as hypohydration (Sawka et al. 2007) has been proposed as ≥ 925 mOsmol/kg (Armstrong et al. 2010). Using the Armstrong et al.’s threshold, only 3 (FWB trial) and 3 (SWB trial) participants were classified at hypohydrated at Morning Day + 1 compared to 7 (FWB trial) and 9 (SWB trial) using the Sawka et al.’s threshold.

No indices of performance were impacted by the RWL and recovery process when compared to pre-RWL values in either the FWB or SWB conditions. These results are in contrast to studies that have demonstrated a residual negative impact on indices of performance after ~ 24 h of recovery (Alves et al. 2018; Barley et al. 2018b; Kurylas et al. 2019). In one study, athletes were dehydrated by ~ 5% of body mass through exercise in a heated room, and performance tests were completed 3 and 24 h after the intervention (Barley et al. 2018b). Vertical jump was unaffected by dehydration and recovery; hand-grip strength was weaker at 3 but not 24 h; medicine ball chest throw distance was shorter at 24 h, but not 3 h; and repeated sled push performance was worse at both 3 and 24 h after dehydration (Barley et al. 2018b). Therefore, there are likely to be time course-specific effects on performance in response to RWL and recovery, and which may also be impacted by whether passive or active methods of dehydration are employed, and the choice of performance test. Active methods of dehydration, i.e., involving exercise, may lead to residual fatigue and depletion of energy stores (Savoie et al. 2015), and can produce divergent responses in relation to changes in plasma volume, serum and urine osmolality, and performance, compared to passive dehydration (Nielson et al. 1981; Caldwell et al. 1984; Muñoz et al. 2013). Moreover, if a chosen performance test is not sensitive enough to detect physiological and performance changes, if any, that may be happening in response to RWL, the conclusion that there are no negative performance consequences of RWL when followed by adequate recovery and rehydration may be a type II error, i.e., false-negative finding.

Relatedly, this study was powered using the primary outcome of change in body mass as a consequence of the 2 h bath and wrap protocol. Given the absence of effect in our previous research using a salt concentration of ~ 1.6% (Connor et al. 2020; Connor and Egan 2021), like our previous approach (Connor and Egan 2021) a priori we planned an interim data analysis for the assessment of futility, and therefore discontinuation. However, the sample size was based on data derived from pre-to-post differences in a crossover design, and therefore, it is likely that the sample size is underpowered for the analysis of serial time point data such as those analyzed by ANOVA. In this scenario, again a type II error for observing the lack of differences between FWB and SWB cannot be fully discounted. Additionally, there are several methodological limitations that could be addressed in future work including the measurement of body temperature with a valid measure of core temperature (either esophageal or rectal), and the inclusion of a body mass measurement immediately after each period of HWI to isolate the effects of salt versus fresh water during HWI specifically rather than the entire hot bath protocol including wrapping periods.

In summary, short-duration HWI combined with periods under an infrared sauna blanket is an effective method of RWL to induce a loss of ~ 2.7% of body mass during 2 h of bathing (2 × 20 min) and wrapping (2 × 40 min). Using this protocol, the total amount of body mass lost when the water was supplemented with ~ 5.0%wt/vol of Epsom salt was similar to fresh water. When an appropriate refueling and rehydration strategy was followed, the ~ 5.3% loss of body mass during the overall ~ 28 to ~ 30 h RWL period was not detrimental in terms of blood markers or indices of performance measured after the ~ 24 to ~ 26 h recovery period.