The influence of rest break frequency and duration on physical performance and psychophysiological responses: a mining simulation study

A repeated measures ANOVA power calculation (α = 0.05, 1–β = 0.8) was conducted with G*Power (Version 3.1.9.3) to determine a sample size based on our primary variable: physical performance (total distance covered). Based on the existing literature, the average effect size (ES) for a difference in physical performance, derived from prolonged self-pacing running laboratory studies is 0.47 (Skein et al. 2012; Wright & Davison, 2013). To express our results with 95% confidence, a minimum sample size of 12 participants was calculated for this study. Therefore, to account for dropout, thirteen male participants were recruited from the general population (mean ± SD; age: 39.3 ± 15.3 y; height: 181.5 ± 4.4 cm; body mass: 78.7 ± 11.1 kg; body mass index: 23.9 ± 3.2). All participants provided informed consent and ethics approval was granted by the host institution’s human research ethics committee (RA/4/20/6536). Testing was performed during the winter months so that participants were not heat acclimatised.

This study used a counterbalanced, randomised, cross-over design. Participants visited the laboratory on three separate occasions: a familiarisation session, followed by two experimental trials scheduled 10 ± 8 days apart (no less than 5 days in-between). Of the two trials performed (Fig. 1), one involved usual work-to-rest break ratios undertaken by employees on a mine site in the North-west of Australia (current practice: CP trial), while the second trial had the same overall exercise and rest break duration, but rest breaks were shorter and more frequent (experimental protocol: EXP trial). Total duration of each trial was 225 min, and consisted of 180 min of simulated work, separated by 45 min of rest breaks. Simulated work was performed in a climate chamber, where the ambient environment replicated summer air temperature and relative humidity experienced on mine sites located in the North-west region of Australia (37 °C, 40% RH, 0 ms⁻¹ wind speed, 0 Wm⁻² solar radiation). Rest breaks were taken in an air-conditioned room (22℃, 35% RH, 0 Wm⁻² solar radiation) to simulate the indoor environments provided on a mine site. All laboratory visits were performed at the exact same time of day. Physical activity (Craig et al. 2003; Hagströmer 2006; O'Brien-Cousins 1994), food and fluid intake were recorded in diaries during the 24 h period prior to the first trial and replicated prior to the second trial. Participants brought in their own lunch during their rest break, which was weighed and replicated for the subsequent trial. Finally, particular attention was given to familiarise participants with perceptual scales.

During a preliminary visit, participants completed the Exercise & Sports Science Australia pre-exercise screening tool to identify any contraindications for physical activity. This was then followed by anthropometric measurements (body mass, height, waist and hip girth). Participants were fitted for the clothing to be worn during the trials and then practiced the counting span task (see below), which they performed five times to minimise any learning effects (Saldaris et al. 2020). Participants then completed 5 min of treadmill walking at various ratings of perceived exertion (RPE: ranging between 9 and 13) in the climate chamber (~ 27 °C, ~ 24% RH) for habituation.

Simulated work involved walking on a treadmill at a perceptually regulated intensity that equated to a RPE of 11 (or ‘light’) on the 6–20 Borg scale (Borg 1982). This corresponded to the average RPE determined from thirteen outdoor workers (with comparable anthropometrics to our participants), whilst working on a mine site in the North-west of Australia in the afternoon during the summer months (37 °C, 40% RH, 0 ms⁻¹ wind speed, 0 Wm⁻² solar radiation: unpublished observations). Walking was chosen as it represents a common activity performed by all outdoor workers across all trades (e.g. gardeners, grounds staff and tradesmen). Participants wore standard mining personal protective equipment (long-sleeve industrial shirt and pants with steel-cap boots and hat) which equated to a clothing insulation value of > 0.78 clo (American Society of Heating Refrigerating and Air-Conditioning Engineers [ASHRAE], 2013). Participants were assessed before, during and after the simulation in both trials in respect to physical and cognitive performance, perceived fatigue, physiological and perceptual parameters and hydration status.

Upon arrival, baseline measures (nude body mass, perceived fatigue, thermal sensation [TS] and thermal comfort [TC], HR, T_c and T_sk, and urine specific gravity [USG]) were recorded. After entering the chamber, cognitive function was assessed before participants commenced walking. They were reminded then and every 10 min thereafter, to adjust treadmill speed at any time so to match their current work rate to an RPE of 11. Simulated work was terminated when: i) the protocol was completed (n = 23), ii) if T_c reached 40 °C (n = 0), or iii) if volitional exhaustion occurred (n = 3). If participants decided to stop walking due to exhaustion (whilst maintaining an RPE of 11), they were given the opportunity to recommence walking when they felt ready or to terminate the session. Whilst walking in the heat chamber, participants could drink room temperature water (~ 37 °C) ad libitum and had unlimited access to refrigerated water (5 °C) during rest breaks, with all water intake recorded. Urine output was also measured during toilet breaks using a labelled container. Various measures were taken during the protocol (details provided below), while all measurements were recorded again on completion of the protocol.

Treadmill distance (m) was used to assess physical performance and was recorded via a computerised program (Treadmill Measuring System Version 2.0, UWA, Australia) with data averaged every 30 min. The treadmill display board (speed and distance) was hidden from participant viewing to ensure condition blinding.

Six hours prior to arrival, participants ingested a radiotelemetry pill to measure gastrointestinal T_c (CorTemp, HQ Inc, USA). On arrival and at the end of the protocol, a mid-stream urine sample was collected to determine USG (Kavouras, 2002), using a desktop refractometer (Model T3-NE, Atago Co. Ltd., Japan) to establish hydration status. Nude body mass was then measured to the nearest 0.05 kg using a digital platform (Model ED3300; Sauter Multi-Range Ebingen, Germany); this was also repeated immediately after the protocol (towel dried) for the calculation of sweat loss (pre – post nude body mass + food + water intake + urinary output). A HR monitor (Polar RS400, Finland) was fitted around the participant’s chest, and thermistors (Skin Sensor SST-1, Physitemp Instruments Inc, NJ, USA) were taped to three sites: sternum, right medial forearm, and right mid-posterior calf to determine mean T_sk via a computerised program (DASYLab Light, National Instruments, Ireland Resources Ltd.). Mean T_sk was calculated using the formula of Burton (1935): T_sk = (0.5 × sternum temperature) + (0.14 × forearm temperature) + (0.36 × calf temperature). Measures of TS and TC were obtained every 5 min during the protocol on visual analog scales ranging from green to red scale (0 = ‘very cold’ to 20 = ‘very hot’) and from white to black graduation (0 = ‘very comfortable’ to 20 = ‘very uncomfortable’), respectively (Gaoua et al. 2012). These physiological (T_c, T_sk and HR) and perceptual (TS and TC) measures were obtained at baseline and every 5 min thereafter, with the exception of T_sk which was not recorded during rest breaks.

Short-term working memory was assessed using the counting span task (Case et al. 1982; Millisecond Software, Seattle, USA), with assessments made upon entering the climate chamber and after every 60 min of simulated work. Briefly, participants were presented with a number of cards (‘spans’) which included yellow and green dots. They were required to remember the number of green dots presented on each card in the order of presentation within the span and then to recall this information later. Counting span performance was recorded as the highest level in which participants successfully recalled the spans, plus one-third of a point for each span recalled at a higher level (Case et al. 1982).

Perceived fatigue was assessed before and after the treadmill exercise using the Multidimensional Fatigue Inventory (Smets et al. 1995). This questionnaire measures the sub-dimensions of general, mental and physical fatigue, as well as reduced motivation and reduced activity. The Multidimensional Fatigue Inventory has displayed good internal consistency between general, physical and mental fatigue sub-dimensions (Cronbach’s alpha > 0.80) and adequate reliability for reduced activity and reduced motivation sub-dimensions (Cronbach’s alpha > 0.65) (Lin et al. 2009). Test–retest reliability had also been reported to be between 0.59 and 0.68 for each of the sub-dimensions of perceived fatigue (Chung et al. 2014).

Data are expressed as mean ± SD. Statistical analysis was conducted using IBM SPSS Statistics (Version 24.0 for Windows; SPSS Inc, Chicago, IL). Two-way repeated measures ANOVAs (time and condition) were used to compare distance covered, USG, fluid intake, T_c, T_sk, HR, TS, TC, cognitive function and perceived fatigue. One-way repeated measures ANOVAs were used to compare sweat loss between trials. All main effects were assessed using Cohen’s d (Cohen, 1988), derived from the partial eta and eta² to minimise publication of error in ES (Levine & Hewett, 2002). Where appropriate, post hoc comparisons using Bonferroni adjustments were conducted. Due to our sample size (n < 20), Hedge’s g ES were assessed to determine meaningful differences, with only moderate (g = 0.38–0.75) and large ES (g ≥ 0.76) reported (Brydges, 2019; Cumming, 2012). Mean difference ± 95% confidence intervals were also determined. Statistical significance was accepted at p ≤ 0.05.

While there were no significant main effects of time (p = 0.273) and condition (p = 0.086) or an interaction effect between these factors (p = 0.143) for treadmill distance, moderate ES were found between trials at 30 min (CP: 2221 ± 295 m; EXP: 1938 ± 350 m; g = 0.68 [0.08,1.26]) and 60 min (CP: 2292 ± 352 m; EXP: 2055 ± 90 m; g = 0.59 [0.01, 1.16]; Fig. 2). A moderate ES was also found in total accumulative distance after 3 h of simulated work (CP: 12,858 ± 2207 m; 12,094 ± 2174 m; g = 0.49 [–0.07, 1.02]).

Physical performance (treadmill distance) did not differ significantly between trials. This contradicts our hypothesis that shorter, more frequent rest breaks would subsequently improve exercise capacity in the heat. However, these observations are in line with a field-based study by Sawka et al. (2015) reporting no differences between continuous and intermittent (work–rest ratio: 30–10 min) military pacing at a moderate intensity (~ 400 W) in the heat (~ 29 °C wet bulb globe temperature). It is possible that circulatory adjustments incurred via the transition from rest to work may have disrupted physical performance in EXP. During exercise in the heat, the demand for blood flow to the periphery (for cooling) is contested by the working muscles (for movement) (Périard et al. 2021). During rest, however, participants opted to be seated, thereby reducing the demand for skeletal muscle pump activity (Crandall and Wilson, 2015). Consequently, blood flow during rest breaks would have been displaced from the working muscles (legs) and redirected from the working muscles (legs) to the periphery (skin) for heat dissipation (Johnson and Rowell, 1975). Blood flow redistribution may have accentuated venous pooling (in the legs), which would have consequently reduced venous return and cardiac filling when commencing subsequent work-intervals (Périard et al. 2021). With RPE strongly determined by cardiovascular redistribution (Périard et al. 2011), this may have increased participants’ perception of work at the commencement of each work-interval (following rest breaks), influencing their adjustment of intensity (RPE of 11). Therefore, the intervention of additional rest breaks in EXP compared to CP may have induced circulatory adjustments at the resumption of each subsequent work interval (following rest breaks), and thus disrupted physical performance.

It was expected that shorter, more frequent rest breaks would encourage water ingestion (Pryor et al. 2018), in turn leading to better hydration status in EXP compared to CP. Surprisingly, urine specific gravity was stable across conditions and from before to after the exercise protocol. Conversely, Greenleaf and Sargent (1965) reported that individuals failed to drink enough water ad libitum to compensate for sweat losses, when walking (at 6.4 kmh⁻¹) for 4 h in extreme heat (49 °C, 25% RH). Hydrated individuals are able to tolerate higher increases in T_c under heat stress, compared to dehydrated (as they experience increased cardiovascular strain; Périard et al. 2021). As many participants were either physically active and/or had a background in exercise sciences, much of the sample was likely well informed of the importance of fluid ingestion and the potential detrimental impacts of dehydration. These individuals may have consumed water in excess to overcompensate for fluid loss and prevent dehydration. Altogether, results showed that participants were able to self-manage in both trials, which may have aided in their physiological tolerance to heat stress and maintenance of physical performance throughout the exercise protocol.

Importantly, water intake inside the climate chamber (during work) in both trials may have been greater than anticipated due to the monotony of simulated work. While participants were permitted to have entertainment (i.e. music, television shows, podcasts; controlled between trials) inside the climate chamber, anecdotal evidence suggests boredom may have prompted them to drink. Water was available ad libitum throughout work and rest breaks, resulting in a total water consumption of ~ 3.6 L in both trials. Total water intake contributed to an increase in nude body mass of ~ 1.5% in CP and ~ 1.6% in EXP. Water consumption during CP and EXP may have also prevented the deterioration of complex cognitive performance (discussed below). It is unclear whether these findings would be consistent if replicated in a field-based study, whereby participants will be required to perform occupational tasks, often in the company of fellow workers. This may provide a distraction from hydration, as work tasks involve increased stimuli and decision-making, compared to treadmill walking. Therefore, actual labour tasks may lower water ingestion during a prolonged shift, in line with field-based evidence (Hunt et al. 2013).

The T_c and HR responses observed in CP and EXP concur with previous observations (Zhai et al. 2019) in that rest breaks in cooler environments can lower cardiovascular strain incurred from prolonged physical activity in the heat. This may have increased heat dissipation capacity via the attainment of a larger gradient of ambient temperature and water vapour pressure between the skin and the surrounding environment (Wendt et al. 2007). Notably, the higher T_c and HR values observed at 120 and 180 min time points in CP compared to EXP (with moderate-to-large ES) may represent the accumulative elevation of hourly work intervals in CP. Additional rest breaks in EXP, separated into 30 min work intervals may have aided in the alleviation of thermal strain via the restoration of thermoregulation that occurs with (1) exposure to a cooler ambient environment (Wendt et al., 2007) and (2) the cessation of exercise (Kenny & McGinn, 2017).

Of importance, the EXP trial was successful in maintaining T_c values below 38 °C. Therefore, shorter, more frequent rest breaks may be appropriate for the protection of worker health and safety. Nonetheless, our findings also indicated that most participants were able to tolerate hyperthermic conditions (> 38 °C) without negatively impacting physical performance. Heat exhaustion resulted in the early termination of the protocol on three occasions. While two of the three cases failed to report T_c values (due to inability to detect the telemetric capsule), the final T_c of one participant was reported to be 39.09 °C at the time of termination in CP (completed 195 of the 225 min protocol). Notably, the participant completed the EXP trial without the termination of the protocol.

Better maintenance of thermal homeostasis in the EXP trial may also be attributed to the greater consumption of cold water compared to CP. Ingestion of cold water can delay the autonomic heat dissipation mechanisms activated during activity in the heat via the reduction of thermal strain on the cardiovascular system, possibly resulting in lower TS and TC values (Wilson et al. 2002). Of relevance, the EXP trial reported lower TS at 120 min, and lower TC at 120 and 180 min (similar to reductions in T_c and HR). This concurs with a report by Lee et al. (2018), whereby the ingestion of cold water (4 °C) lowered TS at the end of the rest break period (45 min) and throughout subsequent time-to-exhaustion exercise in the heat (cycling), compared to warm water ingestion (37 °C). Additionally, these authors reported that 900 ml of cold water lowered T_c by 0.5 °C, compared to warm water, during 30 min of rest. Exposure to cooler environments and the provision of refrigerated water in rest breaks, likely explained the reduction in physiological and perceptual responses reported for EXP compared to CP.

Our results showed no impairment of complex cognitive performance in either protocol. It was anticipated that 180 min of walking in thermally stressful conditions while wearing protective personal equipment, with fewer rest periods exposed to a cool environment, would induce moderate hyperthermia (T_c > 38.5 °C) in CP, and hence contribute to deterioration of complex cognitive performance. In both CP (T_c = 38.3 °C) and EXP (T_c = 37.9 °C), however, T_c did not exceed the proposed threshold (38.5 °C) above which complex cognitive performance is impaired (Gaoua et al. 2011; Schlader et al. 2015). Additionally, EXP prevented mean T_c from reaching hyperthermic levels (< 38.0 °C). Our findings, in which cognitive impairment did not occur in either protocol, were supported by previous studies which assessed complex cognitive performance during hyperthermia using short-term working memory tasks. For example, Schlader et al. (2015) reported that mild hyperthermia (T_c = 38.0 °C) did not impair complex cognitive performance. It was speculated that rest breaks in an air-conditioned room and access to cold water (5 °C) ad libitum likely reduced T_c. Results from Lee et al. (2008) suggest that the ingestion of 900 mL of cold fluid (4 °C) reduces T_c by ~ 0.5 °C relative to ingestion of 900 mL of warm fluid (37 °C). In our study, the combination of cold water consumption and exposure to thermo-neutral conditions during rest breaks likely prevented T_c from exceeding 38.5 °C in CP and EXP, hence averting complex cognitive impairment.

All sub-dimensions of perceived fatigue increased from pre- to post-exercise protocol (albeit not significantly for mental fatigue and reduced motivation), independent of condition. This contradicts our hypothesis, as we expected lower perceived fatigue scores following completion of EXP compared to CP. Although fatigue levels were not different between trials, lower physiological strain was observed in EXP relative to CP. As fatigue likely results from extended periods of physically demanding activity (Caldwell et al. 2012), it was assumed incorporation of more frequent, shorter duration rest breaks may have resulted in lower levels of fatigue. By the end of the protocol, TC, T_c and HR were higher in CP than in EXP, indicating greater thermal discomfort and physiological strain. However, this did not correspond to increased perceived fatigue post-trial in CP. One potential explanation is because exercise intensity was perceptually regulated (as opposed to imposed), and therefore similar between protocols, given participants were required to maintain the same RPE in both trials. Although T_c and HR were higher at 120 min and 180 min in CP, the alleviation of thermal strain during the lunch break (CP: 30 min; EXP: 15 min) was comparable between conditions (all p > 0.05). Fatigue, measured perceptually, may have been influenced by participants’ preference of trials and/or their preconceived ideas, whereby they may have believed that longer rests alleviated fatigue. Overall, our results suggest that altering work-to-rest schedules to incorporate more frequent, shorter duration rest breaks did not result in lower fatigue accumulation relative to schedules with fewer, longer duration rest breaks.

This study is not without limitations. First, our results showed that physical performance did not differ between trials over a single simulated shift. However, with rostered swings often lasting for ~ 14 days, the impact of rest break interventions on accumulative heat strain (and its consequences on physical performance) on outdoor mine workers was not assessed. While a laboratory-based study allowed for strict control of ambient environmental conditions and the removal of confounding factors, it prevents the generalisation of our findings to real-world settings. With literature suggesting that time to exhaustion occurs sooner in laboratory versus field-based studies (Sawka et al., 2015), this indirectly highlights the distinction between simulative- and naturally incurred heat stress.

Second, while participants were selected to best replicate the demographic of outdoor mine workers, volunteer recruitment may have attracted physically fit individuals who were confident in their ability to withstand thermally stressful conditions. Heat-tolerant individuals often possess inter-individual characteristics (younger, male, with higher aerobic capacity, and lower adipose body composition) that can improve their ability to maintain thermal homeostasis under heat stress (Kenny and Jay, 2013).

One additional limitation is the absence of solar radiation in our study, which is considered one of the biggest contributors to cognitive impairment (Piil et al. 2020; Ioannou et al. 2021). While we did not observe any impairment of cognitive performance over a shift, the presence of solar radiation may have influenced this response. Alternatively, the complexity of the cognitive task may have not been suitable for the tested population. Eight trials resulted in the attainment of the maximum counting span performance for all time points within a trial, while three participants achieved the maximum counting span performance across all time points in both CP and EXP.

Finally, it is likely (yet unknown) that behavioural thermoregulation strategies may differ between laboratory and field-based studies conducted in thermally stressful conditions. Field-based research has suggested that outdoor (agricultural) workers spend ~ 15% of total work shift time on irregular unprescribed rest breaks (Ioannou et al. 2017). This contrasts with our laboratory-based protocol where rest breaks were carefully timed and controlled. While the present study reported no significant differences in total water intake between protocols, real-world replication may limit workers from sufficient ingestion due to limited water accessibility (Piil et al. 2018), which may subsequently influence worker’s urine specific gravity post trial. Overall, field-based research is needed to capture the actual behavioural worker adjustments to prolonged heat stress on work productivity and water intake.