Recently there has been an increased interest in the use of hypoxic conditioning (HC) as a strategy for promoting indicators of health and weight loss in individuals with obesity (Hobbins et al. 2017; Ramos-Campo et al. 2019). In comparison to normoxic constant-load exercise training programmes (e.g., 60-90 min walking/running, cycling or cross-training at 60-75% maximal oxygen uptake [V̇O_2max] or heart rate [HR_max]), HC with an FiO₂ in the range 13–16.5% for 4-8 weeks has been shown to elicit greater reductions in body mass and fat mass (Netzer et al. 2008; Wiesner et al. 2009), along with improvements in blood glucose concentrations (Haufe et al. 2008; De Groote et al. 2018), blood pressure (Kong et al. 2014) and exercise capacity (Chacaroun et al. 2020). Comparatively, perceptual responses and exercise-related sensations associated with this type of training have so far been overlooked.

Weight loss within the early phase of an exercise training intervention (within 2–4 weeks) is a prominent predictor of exercise adherence (Burgess et al. 2017). HC studies involving individuals with obesity have typically implemented a training duration of 4-8 weeks (Hobbins et al. 2017; Ramos-Campo et al. 2019). However, Morishima et al. (2015) noted improvements in V̇O_2max and mean blood pressure in adults with obesity following completion of twelve hypoxic (FiO₂ = 15.0%) 60-min cycling sessions at 65% of relative V̇O_2max over 2 weeks. Further, these adaptations were similar to those observed following the same number of sessions spread over 4 weeks in normoxia. Therefore, achieving similar (or better) positive health and functional fitness outcomes in a shorter time frame (with the same exercise load) may improve confidence and likelihood of adherence to regular exercise in obese populations.

Perceptually regulated exercise intensity is defined as maintaining a target intensity, through the use of, for example, rating of perceived exertion (RPE) to self-adjust the absolute intensity (i.e., velocity or power output) (Tucker, 2009). Previously, it has been suggested that this mode of exercise may be considered more enjoyable than absolute fixed-intensity exercise (Hobbins et al. 2017; Ramos-Campo et al. 2019). HC studies have typically implemented continuous, fixed-intensity exercise (60-75% V̇O_2max), which may prove detrimental in terms of exercise adherence and other indicators of exercise satisfaction and pleasure (Burgess et al. 2017).

Therefore, our intention was to compare the effects of a short-term (eight, 60-min sessions over two weeks) perceptually regulated (RPE = 14) interval training intervention in both hypoxic and normoxic conditions. The training intervention involved 15 × 2 min walking blocks interspersed with 2 min of rest and measurement of exercise-related sensations, cardio-metabolic markers, and functional performance in overweight-to-obese individuals were made. It was hypothesised that health outcomes would be improved more after training in hypoxia compared to normoxia, despite the perceptually regulated selection of a lower training workload (i.e., slower treadmill walking velocity).

Male and female adults were recruited to participate in this study from a University staff population of approximately 1700 individuals. Inclusion criteria required that participants were sedentary (<1 h of moderate-intensity exercise/week), did not smoke, had no current or recent (within 3 months) musculoskeletal injury or recent URT infection. Recruited participants were also classified in the BMI range of 27-35 kg.m^-2 (overweight: 25-29.9; obesity: 30+) and had not been exposed to hypoxic conditions within the previous 6 months prior to the start of the study (see Table 1 for anthropometric data of the sample). Written informed consent was obtained from all participants. This study was carried out in accordance with the Declaration of Helsinki. Ethical approval was received from the School of Applied Sciences Ethics Committee (SAS1822).

This study compared two separate groups of adults who were classified as overweight or obese, whereby, half of the participants completed their sessions in hypoxic conditions (FiO₂ = 13.0%, equivalent to ~3500 m elevation above sea level, HYP; n = 8), and the other half completed their sessions in normoxic conditions (sea level, NOR; n = 8). Participants were randomly allocated to training conditions through simple randomization. Participants completed eleven separate visits across three consecutive weeks and were instructed to maintain their normal daily habits in terms of activity, diet, social interactions and sleep patterns. The first session (visit 1) consisted of eligibility determination, familiarisation with the measures and treadmill walking (Pulsar, h/p/cosmos, Germany), and identification of the walking velocity associated with a RPE of 14, as described previously (Hobbins et al. 2019). Within 72 h, participants returned to the lab for pre-testing (visit 2), which consisted of assessment of anthropometrics (body mass and stature), physiological responses (blood pressure), exercise-related sensations (perceived mood change and exercise self-efficacy) and functional fitness (6-min walk test). After 24-72 h, participants undertook the first of eight supervised 60-min self-paced interval-walking training interventions. Visits 4–10 involved the remainder of the 8 training sessions, which were completed within a 2-week period. Within 72 h of the final perceptually regulated interval walking session, participants returned to the lab (visit 11) for post-intervention testing. Testing and training environments were maintained at 23iC and 45% relative humidity.

The eight supervised 60-min perceptually regulated interval-walk sessions were completed across two consecutive weeks in a commercial gym (Academy of Sport, London South Bank University). Each session began with a 5-min warm up at 3.0 km·h^-1 on the treadmill (Fusion Run Series3, Pulse Fitness, UK). A facemask connected to a portable hypoxic generator (see below) was then attached and remained in place for the entire session. During all sessions, the first 30 s of each 2-min interval began at the participants’ perceptually regulated walking velocity (RPE = 14) identified at visit 1. Following this, participants were able, every 30 s, to decide if and how treadmill velocity needed to be altered (i.e., increased or decreased by 0.5, 1.0 or 1.5 km·h^-1, or maintained) to ensure maintenance of an RPE of 14 whilst walking.

The facemask (Altitude Training Mask, Hypoxico Altitude Training Systems, USA) was connected via corrugated plastic tubing to a hypoxic generator (Everest Training Summit II, Hypoxico Altitude Training Systems, USA) to create hypoxic conditions. The hypoxic level provided in this study was an FiO₂ 13.0% (simulated altitude of ~3500 m), while the total hypoxic exposure corresponded to exactly 480 min for those in the HYP group. Participants in the normoxic group were blinded to the condition by being connect to the same hypoxicator system set at sea level equivalent FiO₂ (21%).

Treadmill velocity, HR (M400, Polar, Finland) and arterial oxygen saturation (SpO₂) (iHealth Air, iHealthLabs, USA) were recorded every 30 s during interval walking. Before each session, perceived recovery was assessed in response to a numeric scale, ranging from 0 being ‘very poorly recovered’ to 10 being ‘very well recovered’ (Laurent et al. 2011). Perceived motivation was assessed via a 20 cm visual analog scale, with 0 being ‘not very motivated’ (white colored) and 20 being ‘very motivated’ (black colored) (Crewther et al. 2016). Immediately after each training session, perceived breathlessness and limb comfort were determined in response to a numeric scale ranging from 0 being ‘nothing at all’ to 10 being ‘very, very severe’ (Ward and Whipp, 1989), whilst perceived pleasure was assessed via a 20-cm visual analog scale ranging from 0 being ‘not very pleasant’ (white colored) and 20 being ‘very pleasant’ (black colored).

Participants arrived at the lab following an 8-h fasting period (water exempt). Stature and body mass, and subsequently BMI, were assessed using an electric stadiometer (220, Seca GmbH, USA). After 10 min of rest, participants were asked ‘how are you feeling right now?’ and instructed to verbally specify a number on an 11-point scale anchored ‘very bad’ (-5) up to ‘very good’ (+5) for perceived mood state (Hardy and Rejeski, 1989). Exercise self-efficacy was determined by participants completing a six item, 11-point Likert scales (Smith et al. 2012). Blood pressure (systolic and diastolic) was assessed via an automated pressure cuff (Omron M4, Omron, Japan) attached, secured and inflated around the upper arm, level with the heart.

Participants completed a standardised warm up (5 min at 3 km·h^-1) before a functional fitness test involving 6 min of perceptually regulated continuous walking in normoxic conditions. The treadmill velocity was set at 50% of their self-selected walking velocity, and participants stepped on whilst the velocity was increased to their velocity associated with an RPE of 14 within 10 s. Participants were instructed to ‘walk as far as possible in six minutes without running or jogging’. They were able to maintain the velocity or increase/decrease it by 0.5, 1.0 or 1.5 km·h^-1 every 30 s. Following completion of the 6-min period the total distance covered was recorded (Gibson et al. 2015).

Preliminary analysis (paired-sample, equal variance t-test) was carried out to determine whether pre-tests measurements were statistically significantly different between HYP and NOR. If statistical differences were found, data collected during and post-training were normalized to the pre-training measurement. Velocity, HR and SpO₂ were averaged across each 60-min session. Velocity, HR and exercise-related sensations recorded during sessions 2-8 were calculated as a percentage change from session 1 (100%) due to differences in the initial velocity deemed equal to RPE 14 between HYP and NOR.

Data are presented as mean ± standard deviation. A t-test was used to determine any statistically significant differences in the absolute velocity, HR, SpO₂, perceived recovery, motivation, breathlessness, limb discomfort and pleasure values (averaged across the session) during session 1. A two-way repeated-measures ANOVA was used to investigate the main effect of condition (hypoxia vs. normoxia), time (pre-training vs. post-training or session 1 vs. 2, 3, 4, 5, 6, 7 and 8) and the condition × time interaction. A Bonferroni post hoc multiple comparison was performed if a significant main effect was observed. Effect sizes were described in terms of partial eta-squared (η_p², with η_p² ≥ 0.06 representing a moderate effect and η_p² ≥ 0.14 a large effect). All statistical calculations were performed using SPSS statistical software (IBM Corp., Armonk, NY, USA). The significance level was set at p < 0.05.

Treadmill velocity did not differ between conditions or over time (i.e., session 1: 6.5 ± 0.3 vs. 6.3 ± 0.6 km.h^-1; p > 0.05, Figure 1A; Table 2). During the first session, HR was higher in HYP vs. NOR (144 ± 16 vs. 129 ± 20 bpm; +10 ± 3%; p = 0.04; Table 2). While this difference persisted between conditions there were no changes within condition across the training sessions (p > 0.05, Figure 1B). Similarly, SpO₂ was lower during HYP vs. NOR (83 ± 1% vs. 96 ± 1%, p < 0.05), but did not vary over time (p > 0.05, Figure 1C; Table 2).

Perceived recovery was lower prior to session 4 (76 ± 16%, p < 0.01) and 7 (75 ± 19%, p < 0.01) compared to session 1, irrespective of condition (Figure 2A). Perceived motivation was lower prior to session 2 (75 ± 12%), 3 (69 ± 20%), 4 (69 ± 14%), 5 (77 ± 16%), 7 (73 ± 17%) and 8 (71 ± 20%) compared to session 1, irrespective of condition (p < 0.04, Figure 2B). Perceived breathlessness, limb discomfort and pleasure did not change with condition or time (p > 0.05, Figure 2C–E; Table 2).

Body mass, body mass index, systolic and diastolic blood pressures and functional fitness did not change with condition or time (p > 0.05, Table 1). Exercise self-efficacy (+7 ± 5%, p = 0.03) improved from pre- to post-tests, irrespective of condition (p > 0.05). Despite failing to reach statistical significance, perceived mood state (+12 ± 2%, p = 0.06) followed a similar trend.

The aim of this study was to investigate whether the addition of hypoxia during a short-term, perceptually regulated (RPE = 14) interval walking training program in adults with obesity leads to similar (or superior) improvements in psycho-physiological responses. During the training, treadmill velocity and perceptual responses did not differ between conditions, despite hypoxia-induced elevations in physiological strain (i.e., higher heart rate and lower SpO₂). From pre- to post-training perceived mood state improved in both groups. However, body mass, BMI, blood pressure and functional fitness did not change.

Over the course of our intervention, treadmill velocity that was self-selected by our participants to maintain a RPE of 14 did not differ either between conditions or across sessions. Despite similar external workload in both groups, higher HR and lower SpO₂ were measured during session 1 in HYP vs. NOR, and this internal load due to hypoxia exposure persisted during all remaining sessions. When training three times per week for 3-4 weeks, lower workload (-7-28%) during cycling (Haufe et al. 2008; Pramsohler et al. 2017) or running (Wiesner et al. 2009) exercise modes in moderate hypoxia (FiO₂ = ~15.0%) produced similar HR values compared to normoxia. In the aforementioned studies, training included 30–60 min of continuous exercise at a fixed, moderate intensity. In our study, the lack of a difference for treadmill velocity between conditions may be due to our original approach of maintaining a constant RPE target during 2-min interval walking workouts.

We further observed that neither external (treadmill velocity) nor internal load (HR, SpO₂) metrics changed during the course of the eight training sessions in both conditions. Contrastingly, Fernández Menéndez et al. (2018) showed that preferred walking speed (corresponding to RPE ~10) in obese adults became progressively faster over the course of a 3-week walking intervention in both hypoxic and normoxic training groups, despite selection of a ~7% slower velocity in hypoxia than normoxia. In the present study when exercise intensity is perceptually regulated (RPE = 14) during interval walking workouts, internal and external loads metrics remained unchanged from the first to the eighth training session over the course of the 2-week intervention. The difference in our findings to those of Fernández Menéndez et al. (2018), may be explained by the interval nature of the exercise such that the rest periods allowed for appropriate recovery and initiation of exercise in the next interval at a constant perceptually regulated intensity.

Walking velocities did not differ between conditions, therefore one would expect exercise-related sensations to be negatively impacted in HYP compared to NOR in the presence of lower SpO₂ and higher HR readings. One interesting finding, however, was that none of the perceptual measures were negatively affected by the addition of moderate hypoxia. Conversely, higher difficulty breathing and limb discomfort readings were reported by Soo et al. (2020) when completing repeated cycle sprints (8 × 5-s sprints, 25 s of rest) and by Hobbins et al. (2019) during perceptually regulated (RPE = 16), high-intensity intermittent runs (4 × 4-min, 3 min of rest) in deprived-O₂ conditions (FiO₂ = 13-15%). Jeffries et al. (2019) reported progressive arterial hypoxemia (lower SpO₂) and increases in ventilation as primary cues as an explanation for a shorter time to exhaustion during a cycling task (clamped at RPE 16) in severe hypoxia (FiO₂ = 11.4%) in comparison to normoxia. This suggests that more severe hypoxia levels than those used in the current study (FiO₂ = 13.0%) may be required to observe a negative influence on exercise-related sensations. The nature of our perceptually regulated exercise, involving short (2 min) exercise intervals, followed by similar recovery duration, may also explain why we failed to observe apparent differences between conditions in perceptual variables. Importantly, decreases in perceived recovery and motivation (i.e., already visible after the initial session) occurred across sessions, irrespective of environmental condition. A possible explanation may likely be the large number of training sessions in a short-time frame. This is an important consideration for implementation since, as described by Ekkekakis and Lind (2006), exercise-related sensations are important for adherence to regular exercise training.

Total body mass, BMI and functional fitness did not change in response to either interventions. Previously, greater (Kong et al. 2014; Netzer et al. 2008) and similar improvements (Gatterer et al. 2015) in body composition have been found after training in hypoxic than normoxic conditions. Similarly, gains in functional fitness were greater after HC compared to the normoxic equivalent training (Haufe et al. 2008; Wiesner et al. 2009). In adults with obesity, a 2-week training block may not be long enough to elicit positive changes in body composition and functional fitness, unlike the aforementioned studies implementing longer training periods (4-6 weeks) yet with a similar number (8-12) of training sessions (Haufe et al. 2008; Kong et al. 2014; Netzer et al. 2008). Other anthropometric measures not assessed in the current study (i.e., waist: hip ratio, fat mass and muscle mass) that are pertinent for improved body composition should be assessed in future investigations. In our study, blood pressure remained unchanged throughout the protocol. Perhaps a greater training dose (i.e., higher intensity and/or longer training duration) may be required in normotensive individuals (as recruited here) to positively impact on exercise capacity and cardio-metabolic health when in hypoxia compared to normoxia (Navarrete-Opazo and Mitchell, 2014). That said, despite greater improvement in exercise tolerance, HC (hypobaric hypoxia with a target SpO₂ of 80%) thrice weekly for 8 weeks was not associated with larger improvement in either body composition or vascular and metabolic functions in overweight-to-obese individuals compared to normoxic equivalent (Chacaroun et al. 2020).

Irrespective of condition, perceived mood change and exercise self-efficacy improved from pre- to post-tests. To our knowledge, no investigation exists that has compared exercise-related sensations before and after HC in a similar population. In obese individuals, fixed-intensity interval training (60 i 8-s sprint at 90% V̇O_2max/12 s recovery) including 20 sessions over 5 weeks in normoxia was perceived as being more enjoyable and easier to complete compared to moderate-intensity, continuous training (40 min at 65% V̇O_2max) (Kong et al. 2016). In our study, implementation of interval exercise at a perceptually regulated intensity, causing higher internal but similar external load levels between conditions, may have mitigated the potential onset of hypoxic-induced negative mood (Lane et al. 2004). Overall, similar improvements in perceived mood change and exercise self-efficacy result from perceptually regulated interval walking in HYP and NOR.

Limitations of this study include both the relatively small sample size and the approach to simple randomization increasing the likelihood of unequal distribution of participants (Kim and Shin 2014). This appears evident in both the mean age of the participants in the hypoxic and normoxic groups (32.1 and 41.1 years, respectively) and the assessment of baseline functional fitness (670 vs. 613 m, respectively). That being said, changes in functional fitness from pre- to post- the intervention did not differ between groups (hypoxic: >1.5% and normoxic: 1.0%). This implies that the ability to determine the impact of age on change in functional fitness is not evident. Rather the duration of the intervention was the key limiting factor in determining the impact of training under hypoxic compared to normoxic conditions. This suggests that adherence to exercise for longer than a 2-week period would be required if the beneficial effects of exercise in hypoxia vs. normoxia are to be realized.

In overweight adults, or those with obesity, eight perceptually regulated interval-walk sessions over 2-week training period led to similar treadmill velocity and perceptual responses between HYP and NOR, despite hypoxia-induced elevations in physiological strain (i.e., higher heart rate and lower SpO₂). While both interventions improved exercise-related sensations, there were no significant differences in weight loss, blood pressure and functional performance from either training condition. Hypoxic conditioning does not appear to modify some cardiometabolic risk factors and improve exercise tolerance in overweight-to-obese adults, at least over a short training period.