Different Exercise Endurance Modalities

ORIGINAL RESEARCH
published: 13 August 2021

doi: 10.3389/fpsyg.2021.686661
Different Endurance Exercise

Modalities, Different Affective
Response: A Within-Subject Study
Katja Dierkes 1,2*, Felipe Mattioni Maturana 2,3 , Inka Rösel 3,4 , Peter Martus 4 ,
Andreas M. Nieß 2,3 , Ansgar Thiel 1,2 and Gorden Sudeck 1,2
1
Faculty of Economic and Social Sciences, Institute of Sports Science, University of Tübingen, Tübingen, Germany,
2
Interfaculty Research Institute for Sport and Physical Activity, University of Tübingen, Tübingen, Germany, 3 Department of
Sports Medicine, University Hospital of Tübingen, Tübingen, Germany, 4 Faculty of Medicine, Institute for Clinical
Epidemiology and Applied Biometry, University of Tübingen, Tübingen, Germany
Affect experienced during an exercise session is supposed to predict future exercise

behavior. However, empirical evidence reveals high variability in affective response to
different exercise modalities. Thus, the purpose of the present study was to compare
acute affective response and its variation during three different endurance exercise
modalities: (a) moderate-intensity continuous exercise (MICE), (b) vigorous-intensity
continuous exercise (VICE), and (c) high-intensity interval exercise (HIIE). Using the
dual-mode theory as a theoretical framework, cognitive and interoceptive factors were
Edited by: considered as potential predictors of in-task affective response. In a within-subject
Elizabeth Thomas, design, 40 insufficiently active healthy participants (aged from 20 to 40 years) attended
Université de Bourgogne, France
three sessions per exercise modality on a cycle ergometer. Affective valence (measured
Reviewed by:
Nicolas Zink,
by the Feeling Scale), two cognitive factors (perceived competence and awareness
University of California, Los Angeles, of interoceptive cues), and one interoceptive factor (heart rate) were assessed before,
United States
during, and after each exercise session. Mixed models with three levels (subject, exercise
Alberto Grao-Cruces,
University of Cadiz, Spain session, and time point) revealed more positive affective valence during MICE compared
*Correspondence: with VICE (p < 0.001) and HIIE (p < 0.01), while there was no significant difference
Katja Dierkes between the latter two. Levene’s test results showed the highest variability of in-task
[email protected]
affective valence during VICE (ps < 0.01). Regarding the course across the session,
Specialty section: MICE was associated with a constant slight increase in affective valence from pre- to
This article was submitted to post-exercise (p < 0.05), whereas VICE and HIIE caused a decline in pleasure, followed
Movement Science and Sport
Psychology,
by an affective rebound immediately after exercise termination (ps < 0.01). The highest
a section of the journal importance of cognitive and interoceptive factors for in-task affective valence was
Frontiers in Psychology
observed in VICE (ps < 0.05). The current findings provide support for the tenets of the
Received: 27 March 2021 dual-mode theory, however, indicating that there may be differences in the affect-intensity
Accepted: 19 July 2021
Published: 13 August 2021 relationship between continuous and interval exercise. In conclusion, the study results
Citation: concerning previously insufficiently active individuals extend the knowledge of how
Dierkes K, Mattioni Maturana F, exercise can positively shape affective well-being depending on exercise modality and
Rösel I, Martus P, Nieß AM, Thiel A
and Sudeck G (2021) Different
psychophysiological influences. This knowledge enables public health practitioners to
Endurance Exercise Modalities, design more individualized activity recommendations, thereby improving the subjective
Different Affective Response: A experience of exercise.
Within-Subject Study.
Front. Psychol. 12:686661. Keywords: affect, physical activity, exercise, dual-mode theory, variability, within-subject, cognitive factors,
doi: 10.3389/fpsyg.2021.686661 interoceptive cues
Frontiers in Psychology | www.frontiersin.org 1 August 2021 | Volume 12 | Article 686661

Dierkes et al. Affective Response to Different Exercise Modalities
INTRODUCTION represent a challenge to the adaptive capacity of the body,

without implying any concrete utility or danger. Given the
In the context of theory-based interventions to enhance adoption ambiguous adaptational implications of exercise within this
and maintenance of exercise programs, affective response during intensity range, cognitive factors are assumed to determine
an exercise session has been shown to predict future exercise the affective response (Ekkekakis and Acevedo, 2006). Third,
behavior (Ekkekakis and Dafermos, 2012; Williams et al., 2012; exercise within the severe-intensity domain (extending from
Rhodes and Kates, 2015). However, substantial interindividual MLSS to the level of maximal exercise capacity; above the
response variability has been demonstrated (Ekkekakis et al., VT2/LT2) is supposed to elicit again more interindividually
2011) and exercise characteristics (e.g., intensity, modality) seem homogeneous affective responses, but this time in the form of
to have a decisive influence (Stork et al., 2017). Affective response a decrease in pleasure (Ekkekakis et al., 2005a). Activities in
can be viewed as an umbrella term for numerous interrelated this intensity range are based on limited energetic resources of
constructs, while current literature has highlighted core affective the anaerobic metabolism and preclude the maintenance of a
valence as the crucial component (Ekkekakis et al., 2020; Stevens physiological steady state. A strong interoceptive influence (e.g.,
et al., 2020), which is primitive and limited to basic appraisals of heart rate, blood lactate) is assumed here, which signals the
pleasure and displeasure (Russell and Barrett, 1999). approaching state of exhaustion to the body (i.e., adaptational
To date, research has not sufficiently clarified which specific risk) and provokes the timely termination of the exercise in
factors influence the affective response and its variability order to prevent damage to the systems. Interoceptive stimuli
during various exercise modalities. In order to prevent thus act as a kind of protective mechanism, causing unpleasant
affective experiences that could obviate exercise adherence, sensations during exercise in this critical physiological range
this knowledge is of particular relevance. Thus, the purpose (Ekkekakis and Acevedo, 2006).
of the present study was to investigate affective response Exercise intensity-dependent patterns of affective response
during three endurance exercise modalities considering different that are apparent during exercise tend to dissipate rather rapidly
exercise intensities among insufficiently active adults. Providing as soon as the activity is terminated (Ekkekakis et al., 2011).
explanations for variability in individuals’ affective response will Accordingly, Backhouse et al. (2007) suggested that the “feel-
give insight into how exercise can be structured to achieve more better” effect of exercise may be an artifact of just measuring
positive affective states (i.e., how to make exercise more pleasant). pre-post affective response, failing to account for affective
fluctuations across an exercise session. While exercising within
Affective Response Depending on Exercise the moderate domain (i.e., intensities below VT1/LT1) should
Intensity be accompanied by a constant or slightly increasing trend in
A broader conceptual framework that encompasses exercise affective valence, exercising at intensities at or above VT1/LT1
intensity-dependent patterns of interindividual variability in is supposed to cause a decline in pleasure, followed by a robust
affective response is the dual-mode theory (DMT; Ekkekakis, positive “affective rebound” once exercise is terminated. That is,
2003, 2005). On the grounds of evolutionary arguments and a uniform shift toward pleasure after exercising in the heavy-
adaptational implications, differences in affective response or severe-intensity domain is predicted by DMT (Ekkekakis
to exercise are explained by the continuous interplay et al., 2005a). As a result, exercisers are supposed to return
between two general factors: (a) cognitive parameters (e.g., to or even exceed pre-exercise feeling states (Ekkekakis et al.,
perceived competence, awareness of interoceptive cues), and (b) 2011). Within the literature on DMT, this assumption is based
interoceptive stimuli (i.e., those emerging from the body such as on the “affect contrast” phenomenon described by Solomon
increased heart rate and ventilation). The relative contribution (1980). Here, the opposing trend in the pattern of affective
of these two factors is hypothesized to shift systematically as a response is attributed to an adaptive benefit of terminating
function of exercise intensity and the corresponding metabolic aversive stimuli and a concomitant restoration of homeostasis.
requirements (Ekkekakis and Acevedo, 2006; Ekkekakis, 2009). As a result of an increased production of neuromodulators (e.g.,
First, for exercise within the moderate-intensity domain (below endorphins, dopamine, serotonin), an increase in pleasure after
the first ventilatory or lactate threshold [VT1/LT1]), it is intense exercise is assumed (Solomon, 1991; Basso and Suzuki,
assumed that affective responses are interindividually relatively 2017).
homogeneous and mainly positive, with a small to moderate
influence of cognitive factors (Ekkekakis et al., 2005a). Due to the
low metabolic requirements, such intensities can be maintained Comparing Different Endurance Exercise
over a long period of time and do not pose a threat to the Modalities: Empirical Evidence
homeostasis of the organism, consequently resulting in pleasant Research on DMT provides broad support within the context
sensations (e.g., feelings of warmth; Ekkekakis and Acevedo, of continuous exercise regarding the valence and variability of
2006). Second, in the heavy-intensity domain (extending from affective responses (e.g., Ekkekakis et al., 2011, 2018; Oliveira
VT1/LT1 to the maximal lactate steady state [MLSS]), the et al., 2015). Thus, empirical findings show less positive and more
highest variability of affective response is postulated, with some variable responses during vigorous-intensity continuous exercise
individuals reporting increases and others decreases in pleasure (VICE; i.e., heavy domain) in comparison with moderate-
(Ekkekakis et al., 2005a). Rising blood lactate concentration intensity continuous exercise (MICE; i.e., moderate domain).
and associated physiological processes to maintain the activity In contrast, affective response during interval exercise is less

understood. Recently, high-intensity interval exercise (HIIE) is the negative influence of interoceptive stimuli (e.g., heart
has been gaining attention in the public health area as a rate) on affective response in the severe domain postulated
time-efficient, less monotonous exercise strategy to improve by DMT, although recent evidence supports the link between
cardiorespiratory and metabolic health (e.g., Batacan et al., affective valence and homeostatic perturbations (Hartman et al.,
2017; Mattioni Maturana et al., 2021a). HIIE is characterized 2019).
by relatively brief, repeated bouts at maximal or near-maximal Despite emerging trends, the comparison of affective
effort, interspersed with recovery periods of low intensity response between continuous and interval-based exercise is still
or complete rest (Gillen and Gibala, 2014). The numerous insufficiently elucidated. Due to the large number of existing
benefits of HIIE on physiological outcomes are well-documented; endurance exercise modalities, barely any reliable predications
however, concerns have been raised about the likelihood of HIIE can be made. In addition, the question arises as to which
(i.e., exercise in the severe-intensity domain) evoking a high psychophysiological factors are relevant in which exercise
degree of displeasure (Decker and Ekkekakis, 2017). modality, and to what extent these are related to the variability of
In recent years, an increasing number of studies have affective response.
investigated differences in affective response between HIIE and
continuous exercise, so review-based evidence is now available.
Niven et al. (2020) performed a meta-analytic synthesis of the Study Rationale
current research (HIIE vs. MICE: 15 studies; HIIE vs. VICE: 7 This study aimed to investigate acute affective response
studies) and concluded that, while HIIE is associated with more to different endurance exercise modalities in previously
negatively valenced affective responses during exercise when insufficiently active individuals. Covering the exercise intensity
compared with MICE, there is no difference in in-task affective domains described within DMT, the following three exercise
valence between HIIE and VICE. Importantly, a large degree modalities were compared: (a) MICE (i.e., moderate domain),
of heterogeneity was evident in both comparisons. A detailed (b) VICE (i.e., heavy domain), and (c) HIIE (i.e., severe domain).
discussion on the apparent inconsistency regarding affective Within this approach, we first wanted to explore differences
response to interval in comparison with continuous exercise can in valence and variation of in-task affective response among
be found elsewhere (Decker and Ekkekakis, 2017). the three different exercise modalities. With respect to
Regarding the affective-rebound effect, the empirical data the outlined theoretical arguments regarding continuous
confirmed the assumptions of DMT for exercise at or above exercise (MICE and VICE) and review-based evidence
the VT1/LT1. Resembling the numerous findings on continuous on HIIE (in comparison with MICE and VICE), we
exercise in the heavy domain (i.e., VICE; e.g., Ekkekakis et al., hypothesized that MICE would result in more positive
2008, 2011), the affective valence decreased consistently during in-task affective valence compared with VICE and HIIE,
HIIE. However, as assumed, this decline in pleasure was followed while there would be similar affective responses between
by an affective rebound immediately after exercise termination the latter two (MICE > VICE, HIIE [Hypothesis H1a]). We
or in the post-exercise period, returning to or even exceeding moreover assumed that, in line with the DMT postulate,
baseline values (Oliveira et al., 2013; Jung et al., 2014; Niven et al., VICE would be associated with a higher variability of in-task
2018; Stork et al., 2018; Alicea et al., 2020; Box et al., 2020). affective valence in comparison with MICE (VICE > MICE
To date, not much is known about exercise modality- [Hypothesis H1b]).
dependent factors of affective valence and its variation. However, Second, we examined whether the course of affective response
among other cognitive factors, concepts relating to perceptions across the session differs depending on the exercise modality.
of ability (e.g., perceived competence) and attentional focus (e.g., Considering the assumptions of DMT and the high consistency
awareness of interoceptive cues) have been shown to account of research findings, we hypothesized that MICE would be on
for variation in the valenced (pleasure–displeasure) response to average accompanied by a constant or slightly increasing trend
continuous exercise within the moderate and especially within in affective valence from pre- to post-exercise, while VICE and
the heavy domain (Rose and Parfitt, 2007; Ekkekakis et al., 2011). HIIE (i.e., intensities above the VT1/LT1) would cause a decline
Recently, a narrative review found perceived competence to be in pleasure, followed by a positive affective rebound immediately
a consistent explanatory variable for affective valence (Bourke after exercise termination (Hypothesis H2).
et al., 2020). This is in line with the important role attributed Lastly, we aimed to identify psychophysiological predictors of
to competency-related characteristics in prevalent theories of in-task affective response. As cognitive factors, we first considered
behavior change (e.g., Self-Determination Theory; Deci and perceived competence (PC) as an already highlighted important
Ryan, 1985, Social Learning Theory; Bandura, 1977). In contrast, explanatory variable for affective valence. Second, within the
mixed results were found for the concept of attentional focus concept of attentional focus, we examined the awareness of
(Bourke et al., 2020). Rose and Parfitt (2010) illuminate such interoceptive cues (AOI) as a less consistent variable. Based
mixed results in exercising around VT1, showing that some on the DMT postulate and previous studies, we hypothesized
individuals interpret interoceptive cues positively and others that cognitive factors (PC and AOI) would be more strongly
negatively. Thus, the directional influence of the awareness associated with in-task affective valence in VICE than in MICE
of interoceptive cues on affective valence seems to be highly (VICE > MICE; Hypothesis H3). Since evidence for interval
variable, depending on how the cues are interpreted (i.e., exercise is particularly scarce in this regard, no directional
individual’s cognitive appraisals of the stimuli). Still less studied hypothesis for HIIE was formulated. Third, we examined the

association of heart rate (HR), the interoceptive factor, with pregnancy or breastfeeding period, and (h) German as a native
in-task affective valence on an exploratory basis. language. Reasons for exclusion from the iReAct study included
the following:
MATERIALS AND METHODS - Chronic diseases or findings that result in a decreased ability
to exercise
Study Design - Medication or supplement intake within the previous 4-weeks
This study was part of a larger research initiative entitled
- Counter indication(s) for local anesthetics
“Individual Response to Physical Activity—A Transdisciplinary
- Clinically relevant deviations in the lab results
Approach” (iReAct; Thiel et al., 2020). The iReAct project is
- Pathological indications in the resting electrocardiogram
an interdisciplinary research network, investigating individual
- Vein conditions that do not allow for multiple blood sampling
physiological, affective, and cognitive responses based on a
- Participation in a medication study within the last 3 months
randomized, two-period sequential-training-intervention design.
- History of drug use or alcohol abuse
Over a period of ∼15-weeks, participants underwent two 6-week
- Current psychotherapy.
training periods starting with either HIIE or MICE (HIIE–MICE
vs. MICE–HIIE). The training programs were of significantly Recruitment occurred in six consecutive waves over a 2-year
different intensity, but matched for energy expenditure. Each period (March 2018 to March 2020). Eligibility was assessed
training period consisted of three training sessions per week (on during a telephone screening, as well as a medical examination
average). Participants underwent a physical fitness assessment prior to final enrollment in the study. A total of 58 participants
before the start of training (week 1), between the two training were assessed for eligibility, 49 of whom were included in the
periods (week 8), and at the end of the study (week 15), including randomization process and nine were excluded during medical
an incremental step test for the standardization of exercise diagnosis. Out of these nine excluded participants, two were
intensity and a VICE session (see Figure 1). excluded due to time management issues and seven for not
The present manuscript addressed a secondary research meeting the inclusion criteria (gastrointestinal issues [n = 2],
question of the iReAct project. As listed in detail above iron deficiency anemia [n = 2], under psychological treatment
(Hypotheses H1–3), a comparison of acute affective response [n = 1], drug consumption [n = 1], and BMI above the
and potential influencing factors was made among three different predetermined upper limit [n = 1]). One female, who exhibited
endurance exercise modalities, which were linked to the intensity a BMI below the specified range, was also included due to her
domains described within DMT (MICE moderate domain, normal percent body fat of 23.5 (normal range: 18–28%) as
VICE heavy domain, HIIE severe domain; Ekkekakis et al., measured at baseline. The included participants (N = 49) were
2005a). Our focus of interest was the in situ assessments, which provided with detailed information about the study procedure
were conducted at four time points (t0 -t3 ) in three exercise and associated risks prior to giving written informed consent.
sessions per exercise modality (Ses1−9 ), resulting in a within- During the baseline assessment, five participants dropped out
subject design with a total number of 36 observations per for different reasons (migraine episode [n = 1], lung condition
participant (see Figure 1). Therefore, it is important to note being discovered [n = 1], time management issues [n = 1], lack
that the group comparison between training type effects across of willingness to continue participation [n = 1], and withdrawal
sequences (i.e., MICE–HIIE vs. HIIE–MICE) was not relevant for during the VICE session due to discomfort with the exercise
the current study, and that the following description refers to the [n = 1]). Two other participants did not complete the first
relevant aspects for the present analysis. Further details on the training period due to illness and thus not being able to complete
clinical trial can be found in the study protocol (Thiel et al., 2020). the minimum adherence. Two non-native speakers included in
deviation from study protocol were subsequently excluded from
Eligibility and Recruitment this data analysis because comprehension of the questionnaires
The targeted study group consisted of insufficiently active could not be guaranteed. Thus, the final study sample comprised
adults at the time of recruitment following the health- 40 insufficiently active healthy adults (men and women) from 20
enhancing physical activity recommendations of the World to 40 years of age. An overview of participants’ demographic and
Health Organization (WHO). That is, <150 min/week of anthropometric characteristics can be found in Table 1.
moderate physical activity, <60 min/week of leisure-time
exercise (including sports participation, endurance-oriented Sample Size
activities, and muscle strengthening), and no regular exercise We did not reach the sample size of N = 60 as originally
engagement for several weeks during the last 6 months. Non- projected in our power calculation and as documented in the
adherence to the WHO recommendations was assessed using study protocol (Thiel et al., 2020). This calculation, however,
the validated German version of the European Health Interview aimed at group comparisons between training type effects across
Survey—Physical Activity Questionnaire (EHIS-PAQ; Finger the two training sequences (i.e., MICE–HIIE vs. HIIE–MICE),
et al., 2015). Further inclusion criteria were (a) age from 20 to which was not the focus of the current study. No separate power
40 years, (b) body mass index (BMI) from 18.5 to 30.0 kg/m2 , (c) analysis was performed for this secondary research question. A
non-smokers, (d) maximal oxygen uptake (V̇O2 max) from 25 to post-hoc sensitivity analysis (with the given sample size of N =
50 ml/min/kg, (e) no current or former eating disorder or obesity, 40) suggests that using a simple t-test for matched pairs and
(f) no severe internistic or neurological previous illness, (g) no assuming a type I error of 0.05 (2-sided) and a power of 80% effect

FIGURE 1 | Overview of the within-subject design. In situ assessments were conducted at 4 time points (t0 -t3 ) within 3 exercise sessions per exercise modality (VICE,
vigorous-intensity continuous exercise; MICE, moderate-intensity continuous exercise; HIIE, high-intensity interval exercise; Ses1−9 ), resulting in a total number of 36
observations per participant (N = 40). In laboratory visits at weeks 1, 8, and 15, exercise intensities were standardized using an incremental step test with lactate
diagnostics and spiroergometry.
TABLE 1 | Demographic and anthropometric characteristics of participants at Cognitive Factors

baseline (N = 40). Two cognitive factors were examined in this study. First,
Characteristic M ± SD Range
perceived competence (PC) was operationalized via the level of
agreement regarding the statement “I feel like I am very competent
Age (years) 27 ± 6 20–40 for the physical activity.” This single item has been formulated
Gender (female/male) 40a 29 (72%)/11 (28%)b in context-specific variation by Sudeck and Conzelmann (2014).
Height (cm) 171.2 ± 9.1 155.0–190.0 The version used in the present study was based on a 7-point
Weight (kg) 69.4 ± 11.1 45.0–101.4 bipolar rating scale ranging from 1 (strongly disagree) to 7
BMI (kg m−2 ) 23.6 ± 2.6 17.6c –30.3d (strongly agree). Second, awareness of interoceptive cues (AOI)
V̇O2max (ml kg−1 min−1 ) 31.4 ± 4.2 24.2–41.4 was assessed using a single-item designed on the basis of Rose
HRmax (b min−1 ) 191.3 ± 10.8 168.0–207.0 and Parfitt’s (2010) procedures. Participants were asked to rate
POpeak (W) 162 ± 26 112–217 the influence of interoceptive cues on their general affective
LTP1 (W) 68 ± 18 35–116 state during exercise by completing the statement “My physical
LTP2 (W) 122 ± 22 75–171 reactions and sensations were...” on a visual analog scale ranging
from 0 to 100 with three verbal anchor points: very disturbing (0),
BMI, body mass index; V̇O2max , maximal oxygen uptake; HRmax , maximal heart rate; neutral (50), and very beneficial (100).
POpeak , peak power output; LTP1, first lactate turning point; LTP2, second lactate
turning point.
a Total number of participants. b number (percentage) of females/males. Minimum and Interoceptive Factor
maximum values of BMI are outside the specified range for one participant each but: The extent of interoceptive stimuli was operationalized by heart
c had a normal percentage of body fat (23.5%), and d had a BMI below 30 at inclusion.
rate (HR), which was constantly monitored through a HR
belt (further information follows in the next section). Minor
artifacts in the HR data were cleaned using an anomaly detection
sizes of d = 0.45 can be detected (no Bonferroni adjustment) algorithm to delete noisy data points (implausible spikes or gaps
(G∗ Power Version 3.1.9.6). We further discuss this issue in the due to technical problems), which uses the interquartile method
limitations section. to find outliers (Upton and Cook, 1996). Each noisy data point
was deleted, and the HR was then interpolated on a second-by-
Measures second basis.
Affective Response
Core affective valence, as the primary outcome variable, Procedures
was assessed using the validated German version Standardization of Exercise Intensity
of the Feeling Scale (FS; Hardy and Rejeski, 1989; In the laboratory visits (weeks 1, 8, and 15; see Figure 1),
Maibach et al., 2020). The FS is a single-item, 11- participants undertook an incremental step test to volitional
point bipolar rating scale, ranging from −5 (very bad) exhaustion on a cycle ergometer (Ergoselect 200; Ergoline
through 0 (neutral) to +5 (very good) developed for the GmbH, Bitz, Germany) for determination of the V̇O2max , peak
assessment of affective response during exercise along a power output (POpeak ), and lactate thresholds (first lactate
displeasure-pleasure continuum. turning point [LTP1] and second lactate turning point [LTP2]).

Before starting the test, baseline blood pressure and capillary Data Collection
blood lactate concentration ([La− ]) were measured. The test An overview of measures used at the different survey time
began with a 2-min resting period on the bike, followed by points can be seen in Figure 2. Within each survey session,
25-watt (W) step increments every 3 min, starting at 50 W affective valence (measured by FS) was recorded prior (t0 ), at two
for males and at 25 W for females, until task failure. [La− ] time points in-task (t1 , t2 ), and immediately after exercise (t3 )
was analyzed (Biosen S-Line; EKF, Cardiff, UK) by collecting (i.e., four time points in total). In-task assessments for FS were
capillary blood samples (20 µL) from the right earlobe during performed after 14 and 30 min in HIIE (i.e., in the last 15 s of the
the last 20 s of each stage and immediately after volitional first and third loading intervals). To achieve temporal alignment
exhaustion. HR and electrocardiogram (ECG) were constantly of the measurement time points, minutes 20 and 40 were set for
monitored throughout the test (12-channel PC ECG; custo med the continuous exercise modalities (MICE and VICE). Cognitive
GmbH, Ottobrunn, Germany). Breath-by-breath pulmonary factors were assessed during and after the exercise session.
gas exchange and ventilation (V̇E) were measured using Data collection points for PC were minutes 22 and 38 in HIIE
a metabolic cart (MetaLyzer; CORTEX Biophysics, Leipzig, (i.e., in the last 15 s of the second and forth loading interval),
Germany). Calibration was performed before each test following minutes 20 and 40 in VICE, and minutes 31 and 53 in MICE.
the manufacturer’s instructions. Lactate thresholds were analyzed The time-lagged assessment in MICE and HIIE resulted from
using a segmented regression model at which two breakpoints other study interests as well as the rationale of not overloading
were estimated from the [La− ]–power output relationship. LTP1 individual measurement time points. Consequently, PC values in
was determined as the first rise in [La− ] above baseline levels MICE/HIIE were estimated using the next observation carried
(first breakpoint), which is accompanied by the first increase in backward (NOCB) method (Engels and Diehr, 2003). AOI was
V̇E as a function of V̇O2 (i.e., VT1). LTP2 was determined as the collected in the form of a retrospective rating immediately after
second abrupt increase in [La− ] (second breakpoint), which is exercise cessation (see Figure 2).
accompanied by the second sharp increase in V̇E as a function In-task assessments were implemented by presenting the
of V̇O2 (i.e., VT2; Binder et al., 2008; Hofmann and Tschakert, individual items on A3 posters as visual references so that
2017). the participants could concentrate on the exercise itself. The
questions, including scale anchors, were read aloud by the
Exercise Modalities investigator and the participant’s response was recorded via
All exercise modalities were performed on calibrated smartphone (Google Nexus 5; LG Group, Seoul, South Korea)
bicycle ergometers (ec5000; custo med GmbH, Ottobrunn, with the movisensXS application (movisens GmbH, Karlsruhe,
Germany). Based on the results of spiroergometric and lactate Germany) after consultation. Pre- and post-surveys were
measurements, the following exercise protocols were defined (for conducted independently by the participants with smartphone
a graphical illustration, see Figure 2): in hand. HR was collected throughout all sessions using a HR
belt (3-channel ECG; custo med GmbH, Ottobrunn, Germany).
(a) MICE was prescribed as 60 min of continuous cycling at
Based on this training monitoring, HR was adjusted to fitness
the power output (PO) corresponding to 90% of LTP1 (
changes over the weeks (for details, see Mattioni Maturana et al.,
moderate-intensity domain).
2021b). This ensured that participants were always exercising
(b) VICE was performed for 50 min at a constant PO
within the originally prescribed relative intensity of exercise.
corresponding to the midpoint between the first and the
second lactate threshold (i.e., 50% of the difference between
LTP1 and LTP2; heavy-intensity domain). The session
Statistical Analyses
A manipulation check was carried out to verify whether the
was introduced by a 10-min warm-up at a PO corresponding
participants were exercising at different exercise intensities in
to 90% of LTP1 ( intensity of MICE), totaling 60 min
the three exercise modalities as intended. Separate one-way
of exercise.
analysis of variance (ANOVAs) were executed for the variables
(c) HIIE involved 4 x 4-min intervals at a PO corresponding to
of %HRmax and %HRR.
90% of HRmax . This exercise intensity was chosen as such
Descriptive statistics using means (M) and standard
intensity would be within the severe-intensity domain for this
deviations (SD) were generated for continuous variables
population (i.e., all the exercise intensities were above LTP2).
according to the distribution; frequencies (n) and percentages
Each high-intensity load interval was interspersed with a 4-
(%) were generated for categorical variables. Intraclass
min active recovery at 30 W. The session was enclosed by a
correlation coefficients (ICCs) were calculated for the main
10-min warm-up (at 70% of HRmax ) and a 5-min cool-down
outcome measures. Furthermore, estimates of the within-person
at 30 W, totaling 43 min of exercise duration.
variability across the FS measurements (9 sessions × 4 time
All exercise sessions were supervised by trained personnel. points) and the PC, AOI, and HR measurements (9 sessions ×
While VICE took place under standardized controlled laboratory 3 time points), as well as the between-person variability in these
conditions, MICE and HIIE were completed in a health and outcomes, were calculated.
fitness orientated training environment. However, we considered For the first research question regarding in-task (t1 , t2 )
a potential modality-dependent influence of environment-related affective valence in the three different exercise modalities
characteristics on the affective response in preliminary analyses (Hypothesis H1a), we fit a multilevel model for repeated
(see Supplementary Table 1). measures with the levels subject (ID), session (s1, s2, s3), and

FIGURE 2 | Overview of exercise modalities and measures. Data was collected at 4 points in time: pre-exercise (t0 ), in-task (t1 , t2 ), and post-exercise (t3 ). The x-axis
describes the percentage of exercise completed (0–100%). FS, Feeling Scale; HR, heart rate; PC-VICE, perceived competence assessed in VICE; PC-MICE/HIIE,
perceived competence assessed in MICE/HIIE; AOI, awareness of interoceptive cues; MICE, moderate-intensity continuous exercise; VICE, vigorous-intensity
continuous exercise; HIIE, high-intensity interval exercise; LTP1, first lactate turning point; LTP2, second lactate turning point.
the crossed factor modality (MICE, VICE, HIIE) to examine affective response (Hypothesis H3), we extended Model 1 by
the effect of intensity conditions during the intervention on separately introducing one of three factors (PC, AOI, or HR)
the FS (Model 1). A step-up model construction strategy was and its interaction with modality as fixed effects (Model 3a, b,
applied, retaining fixed effects in the model if they demonstrated c, respectively).
statistical significance (p < 0.05) and successively adding random Due to the low numbers of units on the subject level,
effects to account for the correlated structure of the data. simple covariance structures (scaled identity) had to be chosen
Due to non-convergence caused by over-parameterization (the to reach convergence in all the models. Significant effects
random effects structure exhibited a complexity not supported were followed by pairwise post hoc comparisons applying
by the underlying data), we did not account for the nesting of Bonferroni adjustments. For significant interaction terms, post
individuals within group sequence (i.e., MICE-HIIE vs. HIIE- hoc probing was performed to describe the direction of
MICE). Thus, our final model included the fixed effect modality the interaction effect. We contrasted effects with one SD
and a random intercept on the subject and session level, as below and above the mean value (± 1 SD) using two-way
well as allowing for a random slope for exercise modalities. To interaction plots.
examine the in-task (t1 , t2 ) variability of affective valence in the Data preparation and statistical analyses were carried
three different exercise modalities (Hypothesis H1b), equality out using the Statistical Package for Social Sciences (SPSS,
of variances between exercise modalities was tested using the version 26; IBM Corp., Armonk, NY, USA). All p-values
median-based Levene’s test. were two-sided, and the statistical significance level was set
To address the second research question with regard to the at p < 0.05.
course of the affective valence across a session within the three
different exercise modalities (Hypothesis H2), data from the pre- RESULTS
exercise time point (t0 ), the in-task time points (t1 , t2 ), and
the post-exercise time point (t3 ) were examined in the model Descriptive Analyses
(Model 2). As fixed effects, we considered modality, time point, The manipulation check confirmed that participants were
and the interaction term modality x time point. A random exercising at different intensities in the three exercise modalities
intercept on session and subject level, as well as a random slope, based on mean in-task %HRmax , F (2,645) = 701.53, p < 0.001,
were included. η2 = 0.69, and %HRR, F (2,645) = 643.92, p < 0.001, η2 =
For the third research question concerning the identification 0.67. In addition, comparison of the HR data with reference
of exercise modality-dependent predictors of in-task (t1 , t2 ) values proposed by Binder et al. (2008) indicated that the

TABLE 2 | Descriptive statistics for study variables.
Variable Between-person variability Within-person variability
ICC N M SD Range N mV SD Range
FS [−5 to +5] 0.35 40 2.70 0.96 −0.12 to 4.46 1360a 80 1.29 0.17 to 2.66
PC [1 to 7] 0.63 40 5.29 1.13 2.70 to 6.89 1020b 60 0.85 0.19 to 2.37
AOI [0 to 100] 0.47 40 56.10 12.83 33.11 to 85.11 1020b 60 13.14 3.03 to 27.41
HR 0.25 40 149.74 12.01 122.35 to 169.34 958b 122 19.48 13.03 to 26.49
ICC, intraclass correlation coefficient; mV, missing values; FS, Feeling Scale; PC, perceived competence; AOI, awareness of interoceptive cues; HR, heart rate.
a FS was measured at 4 points in time (pre-exercise [t ], in-task [t , t ], and post-exercise [t ]). b PC, AOI, and HR were measured at 3 points in time (t , t , and t ).
0 1 2 3 1 2 3
exercise modalities were within the targeted intensity domains

(for descriptive statistics see Supplementary Table 2).
Descriptive statistics of the study variables can be found in
Table 2. It is important to note that there are missing values
due to disturbances in HR measurement as well as an early
termination of the last survey wave due to the COVID-19
pandemic. The grand means of FS, PC, and AOI were in the
upper third or upper half of the respective scales. The empirical
range of person means varied from −0.12 to 4.46 for FS, from
2.70 to 6.89 for PC, and from 33.11 to 85.11 for AOI, indicating
substantial between-person variability. The ICCs indicated that
35% (FS), 63% (PC), or 47% (AOI) referred to between-person
differences. In contrast, the ICC of HR indicated that 75% could
be attributed to within-person differences.
Main Analyses
Valence and Variation of In-task Affective Response
Figure 3 illustrates the differences of in-task (t1 , t2 ) affective
FIGURE 3 | Affective valence (Feeling Scale) and its variation within the three
valence and its variation among the three different exercise
different exercise modalities: MICE, moderate-intensity continuous exercise;
modalities (the corresponding descriptive statistics are provided VICE, vigorous-intensity continuous exercise; HIIE, high-intensity interval
in Supplementary Table 3). Model 1 revealed a significant main exercise. The diamonds represent the mean values. Raw data is presented for
effect for modality (F = 14.56, p < 0.001) on affective valence the in-task time points (t1 , t2 ) of each of the three sessions (Ses1−3 ) without
(FS). Pairwise post hoc comparisons (see Table 3) showed FS adjustment for the dependencies within clusters.
to be significantly higher in MICE than in VICE and HIIE.

No significant difference was found between VICE and HIIE
(confirmation of Hypothesis H1a). Levene’s test results indicated
significant variance differences among exercise modalities (F of Hypothesis H2). While the FS post value for VICE remained
= 8.57, p < 0.001). Pairwise post hoc comparisons showed below the pre-exercise value, the FS post value for HIIE increased
significantly higher variability of FS in VICE than in MICE (F = above baseline.
19.91, p < 0.001; confirmation of Hypothesis H1b) and in HIIE
(F = 8.98, p = 0.003). No differences in variability were found Psychophysiological Predictors of In-task Affective
between MICE and HIIE (F = 2.52, p = 0.113). Response
Figure 5 depicts the association of cognitive (PC, AOI) and
Course of Affective Response Across the Session interoceptive (HR) factors with FS as a function of exercise
Figure 4 shows that, while a slight increase of FS occurred in modality. Models 3a-c revealed significant main effects for all
MICE, there was an affective rebound in VICE and HIIE. Model fixed factors and interaction terms.
2 revealed significant main effects for modality (F = 6.08, p = For PC (Model 3a), we found significant main effects for
0.003) and time point (F = 24.22, p < 0.001) on FS; however, modality (F = 10.58, p < 0.001) and PC (F = 41.54, p < 0.001)
they were qualified by the interaction modality x time point (F = on FS; however, they were qualified by the interaction modality
13.33, p < 0.001). In the pairwise post hoc analysis (see Table 4), x PC (F = 7.55, p = 0.001). Pairwise post hoc comparisons (see
we observed a significant increase of FS from pre- to post-exercise Table 3) showed a significantly stronger association of PC with FS
in MICE. In contrast, a significant decrease in FS from pre- in VICE than in MICE (confirmation of Hypothesis H3) and in
exercise to t1 and t2 was evident for VICE and HIIE, followed HIIE, while no significant difference was observed between MICE
by a significant increase from t2 to post-exercise (confirmation and HIIE. Contrasting the interaction effect with one SD higher

TABLE 3 | Associations of exercise modality and psychophysiological factors with TABLE 4 | Comparisons of affective valence at four time points within exercise
affective valence. modalities.
Fixed effect B SE p Tp 1M SE p
MODEL 1: in-task affective valencea t1 vs. t0 0.250 0.107 0.119

MICE vs. VICE 0.886 0.165 <0.001*** t2 vs. t0 0.233 0.126 0.390
Exercise MICE vs. HIIE 0.535 0.166 0.005** t3 vs. t0 0.422 0.144 0.021*
Modality MICE
VICE vs. HIIE −0.350 0.166 0.114 t2 vs. t1 −0.017 0.090 >0.999
MODEL 3: Predictors of in-task affective valenceb t3 vs. t1 0.172 0.114 0.791
PC × VICE vs. MICE 0.342 0.107 0.006** t3 vs. t2 0.190 0.132 0.913
(a) Exercise MICE vs. HIIE −0.051 0.108 >0.999 t1 vs. t0 −0.441 0.109 <0.001***
Modality VICE vs. HIIE 0.291 0.088 0.003** t2 vs. t0 −0.946 0.129 <0.001***
AOI × VICE vs. MICE 0.013 0.007 0.186
t3 vs. t0 −0.468 0.147 0.009**
(b) Exercise MICE vs. HIIE 0.006 0.007 >0.999 VICE
Modality t2 vs. t1 −0.505 0.092 <0.001***
VICE vs. HIIE 0.019 0.007 0.015*
t3 vs. t1 −0.027 0.117 >0.999
HR × VICE vs. MICE −0.023 0.009 0.033*
(c) Exercise MICE vs. HIIE −0.009 0.011 >0.999 t3 vs. t2 0.477 0.135 0.003**
Modality VICE vs. HIIE −0.032 0.010 0.003** t1 vs. t0 −0.327 0.108 0.016*
t2 vs. t0 −0.602 0.128 <0.001***
The results represent pairwise post hoc comparisons of Feeling Scale (FS) values.
PC, perceived competence; AOI, awareness of interoceptive cues; HR, heart rate; x,
t3 vs. t0 0.558 0.146 0.001**
HIIE
interaction term; MICE, moderate-intensity continuous exercise; VICE, vigorous-intensity t2 vs. t1 −0.274 0.092 0.018*
continuous exercise; HIIE, high-intensity interval exercise. t3 vs. t1 0.885 0.116 <0.001***
a In Model 1, we examined in-task (t , t ) affective valence by including the levels subject,
1 2
t3 vs. t2 1.159 0.134 <0.001***
exercise session, and the crossed factor exercise modality (MICE, VICE, HIIE). b For Model
3, we extended Model 1 by separately introducing the interaction term of one of three
The results represent pairwise post-hoc comparisons of Feeling Scale (FS) values for
potential predictors (3a: PC, 3b: AOI, 3c: HR) with exercise modality (x exercise modality)
the 4 points in time (Tp: pre-exercise [t0 ], in-task [t1 , t2 ], and post-exercise [t3 ]; Model
as a fixed factor.
2). MICE, moderate-intensity continuous exercise; VICE, vigorous-intensity continuous
*p < 0.05. **p < 0.01. ***p < 0.001 (Bonferroni adjusted).
exercise; HIIE, high-intensity interval exercise.
*p < 0.05. **p < 0.01. ***p <0.001 (Bonferroni adjusted).
For AOI (Model 3b), we observed significant main effects for

modality (F = 8.40, p = 0.001) and AOI (F = 58.29, p < 0.001)
on FS; however, they were qualified by the interaction modality
x AOI (F = 4.29, p = 0.014). Pairwise post hoc comparisons (see
Table 3) revealed a significantly stronger association of AOI with
FS in VICE than in HIIE. In contrast, no significant differences
were found between VICE and MICE (rejection of Hypothesis
H3) or between MICE and HIIE. Effect contrasting of AOI
resulted in an FS change of 0.66 in VICE vs. 0.32 in HIIE (and
0.43 in MICE; see Figure 5B).
For HR (Model 3c), the model revealed significant main
effects for modality (F = 5.33, p = 0.006) and HR (F =
17.61, p < 0.001) on FS; however, they were qualified by the
interaction modality x HR (F = 6.61, p = 0.002). Pairwise post
hoc comparisons (see Table 3) showed a significantly stronger
association of HR with FS in VICE than in the two other
modalities (MICE and HIIE), where no significant difference
was observed. Effect contrasting of HR resulted in a FS change
FIGURE 4 | Course of affective valence (Feeling Scale) across the session of 0.90 in VICE vs. 0.39 in MICE and 0.19 in HIIE (see
within the three different exercise modalities: MICE, moderate-intensity Figure 5C). Importantly, the direction of the interaction effect
continuous exercise; VICE, vigorous-intensity continuous exercise; HIIE,
high-intensity interval exercise. The graphs represent estimated means and
here was opposite to that for the cognitive factors, such that
standard errors (Model 2) over 4 time points (pre-exercise [t0 ], in-task [t1 , t2 ], an increase in HR rate was accompanied by a decrease in
and post-exercise [t3 ]). affective valence.
DISCUSSION
and lower in PC resulted in an increase and decrease, respectively,
on the FS by 0.80 in VICE vs. 0.31 in MICE and 0.38 in HIIE (see The present study was designed to investigate acute affective
Figure 5A). response associated with endurance exercise modalities

domain (VICE) and an interval exercise in the severe-intensity

domain (HIIE) were compared in a within-subject study among
insufficiently active adults. Basically, the current findings provide
support for the tenets of DMT regarding continuous exercise, but
suggest that these are not directly applicable to the intermittent
nature of HIIE that allows periods of recovery between bouts of
severe exercise.
Consistent with the hypotheses, in-task affective valence was
more positive in MICE compared with VICE and HIIE, while
there was no significant difference between the latter two.
However, the descriptive statistics suggested that VICE was more
negatively valenced in comparison with HIIE. Taken together,
our results are in line with previous research, indicating similar
(Martinez et al., 2015; Niven et al., 2018; Alicea et al., 2020)
or even more positive affective responses (Jung et al., 2014;
Kilpatrick et al., 2015; Martinez et al., 2015) in HIIE as opposed
to VICE. On the basis of these results, it can be assumed that
the short periods of severe intensity in HIIE are not of sufficient
duration to disrupt homeostasis to such an extent that a more
negative affective response is induced. Rather, the rest periods
seem to mitigate the detrimental effect associated with the HIIE
load intervals. Reduced monotony, the prospect of getting a
break, and a feeling of pride after the completion of each interval
could positively influence the affective response in contrast to
continuous exercise in the heavy domain. In line with this
assumption, studies showed that the specific characteristics of
HIIE promote participants’ self-efficacy beliefs (Jung et al., 2014)
and result in more positively valenced post-exercise narrative
responses (e.g., feelings of reward and reenergization associated
with the rest interval) in comparison with continuous exercise
(Alicea et al., 2020). Due to the fact that, in our study, we
exclusively considered affective response at the very end of the
load intervals, it seems plausible that the beneficial influence of
the interval exercise modality in terms of affective valenced states
was not only reflected in the overall picture but also within the
specific severe-intensity exercise periods.
Further, a rebound to more positive affect was observed in
this study following exercise in the heavy and severe domain,
which aligns with the assumptions of DMT (Ekkekakis et al.,
2005a) and empirical evidence (e.g., Stork et al., 2018; Alicea
et al., 2020; Box et al., 2020). As hypothesized for the course
across the session, MICE was associated with a slight increase
in affective valence from pre- to post-exercise, whereas VICE
and HIIE caused a decline in pleasure, followed by an affective
rebound immediately after exercise termination. Regarding the
FIGURE 5 | Two-way interaction plots of psychophysiological predictors of amount of affective rebound, post values of affective valence in
in-task (t1 , t2 ) affective valence (Models 3a-c): (A) Perceived competence, (B) our study remained below baseline values in VICE, but exceeded
Awareness of interoceptive cues, and (C) Heart rate. Shown is the respective baseline values in HIIE. Importantly, the post-value was collected
expression of the Feeling Scale (FS) for the mean value of the predictor as well
in VICE directly following the exercise, whereas in HIIE, a 5-min
as for one standard deviation below the mean value (−1 SD) and for one
standard deviation above the mean value (+1 SD) with error bars representing cool-down was performed before the end of exercise. Results of
standard errors. MICE, moderate-intensity continuous exercise; VICE, previous studies suggested that the postulated affective rebound
vigorous-intensity continuous exercise; HIIE, high-intensity interval exercise. effect only develops (completely) in the post-exercise period (e.g.,
Decker and Ekkekakis, 2017; Stork et al., 2018; Box et al., 2020).
Thus, it can be assumed that a cool-down phase or a later survey
considering different exercise intensities. Based on the lactate time point of post-exercise affective response (e.g., 5 min post-
threshold concept proposed by DMT, two continuous exercise exercise) would also have revealed a rebound to more positive
protocols within the moderate- (MICE) or heavy-intensity valenced states in VICE.

The current results revealed the heavy-intensity range influence of HIIE observed in the study of Decker and Ekkekakis
(i.e., VICE) as the zone of the highest response variability. (2017) could have been due to their demanding protocol with a
Furthermore, the study provided an explanation for this finding work-to-rest ratio of 1:0.66 and a comparatively high intensity
by showing a higher importance of psychophysiological factors within the recovery periods (85% of VT1). In contrast, the
within the heavy domain. Both cognitive factors (perceived 1:1 work-to-rest ratio used in this and in other studies (e.g.,
competence and awareness of interoceptive cues) almost Jung et al., 2014; Alicea et al., 2020), with comparatively lower
consistently had a greater association with affective response dependence on anaerobic metabolism, resulted in equal or even
during VICE in contrast to MICE and HIIE (see Figures 5A,B). more positive affective states in HIIE vs. VICE. Interestingly,
A one SD higher score on either cognitive factor resulted in the negative impact of longer interval duration of 120 s in HIIE
twice the rate of change in affective valence in VICE compared on affective response in comparison with VICE that was found
with HIIE. Moreover, perceived competence was shown to be in the study by Martinez et al. (2015) was not evident for this
more relevant within continuous exercise in the heavy domain study which had an interval duration of 240 s. However, due to
(i.e., VICE) compared with the moderate domain (i.e., MICE), an equation of energy expenditure in their study, HIIE was of
associated with an even 2.5 times higher rate of change in affective longer duration than VICE (24 vs. 20 min), whereas in our study,
valence. Such a tendency could also be seen for the awareness we used a significantly shorter HIIE protocol compared with
of interoceptive cues with a factor of 1.5 for the comparison of VICE (43 vs. 60 min), since the main argument for promoting
impact in VICE vs. MICE, although here the significance level HIIE in public health is its time efficiency. Nonetheless, there is
was not achieved. increasing evidence that low-volume HIIE may diminish feelings
There was no difference in the association of cognitive of displeasure during exercise (e.g., Haines et al., 2020). Similarly,
factors on affective response in the moderate (i.e., MICE) and reducing the intensity of load intervals (from 100 to 85% of peak
severe (i.e., HIIE) modalities considered here. While previous power) has been shown to be a successful strategy for obtaining
research based on a comparison of imposed and self-selected more positive affective experiences in HIIE while maintaining a
HIIE concluded that (reflecting the assumptions of DMT on health-promoting heart rate stimulus (Malik et al., 2019).
continuous exercise) affective valence within the severe domain
is mediated by exercise intensity rather than the feeling of Strengths and Limitations
autonomy (Kellogg et al., 2019), the present results suggest an Some strengths of the current study are noteworthy. First,
influence of cognitive factors in HIIE. Thus, it is possible that responding to the call for psychophysiological perspectives in
the intermittent nature of HIIE prevents a switch to a mode of research examining the affective response to exercise (Acevedo
affect induction that relies primarily on interoceptive stimuli. and Ekkekakis, 2006), we standardized exercise intensities
Supporting this assumption, we found an analogous pattern for relative to metabolic landmarks with reference to a physiological
the modality-dependent association of the interoceptive factor framework. This allowed us to ensure accurate comparisons of
with affective response to that of the cognitive factors studied affective response across the three different exercise intensity
(see Figure 5C). That is, heart rate had a greater importance domains proposed by DMT. Second, in line with current
in the heavy vs. the other two domains, with a two times recommendations, we assessed the valence component of basic
(VICE vs. HIIE) or even five times (VICE vs. MICE) higher affect (using FS) before, during, and after the exercise session
rate of negative change in affective valence. This finding does to provide in-task as well as course-specific patterns of affective
not support the proposition of DMT for continuous exercise, response. Third, we applied a mixed model approach to account
that interoceptive stimuli have the greatest (negative) influence for the nested data structure of the within-subject design (9
in the severe domain. Importantly, the heart rate examined in sessions with 4 time points per participant). Finally, by recruiting
this study represents only one physiological factor and may adults who did not achieve the recommendations for health-
have a different influence on affective response in contrast to a promoting physical activity, the current findings are of direct
neurophysiological (e.g., heart rate variability), cardiorespiratory relevance to a segment of the population that is particularly in
(e.g., oxygen uptake), or metabolic marker (e.g., blood lactate). need of interventions for promoting exercise.
For example, a study by Roloff et al. (2020) demonstrated that Although the current study produced novel and important
valenced affective states closely track changes in oxygen uptake findings, some potential limitations should be mentioned. As
in four different HIIE protocols. this study addressed a secondary research question of the iReAct
Taking a closer look at HIIE, the contradictory study results project, no group comparison between training type effects
can potentially be explained by examining the exercise protocol across sequences (i.e., MICE–HIIE vs. HIIE–MICE) was made,
variables, such as work-to-rest ratio, total session duration, thus neglecting potential carry over effects. In addition, it is
and energy expenditure. Since these characteristics decisively important to consider that, because the participants completed a
determine the extent of reliance on limited energetic resources 15-week training program, strictly speaking, they were no longer
of the anaerobic metabolism, they are supposed to have a physically inactive during the study. However, through training
decisive influence on affective response to exercise (Ekkekakis monitoring, we accounted for fitness changes over the weeks and
et al., 2005a). Therefore, it is not surprising that the strenuous ensured that participants were exercising within the originally
HIIE protocol in a study from Oliveira et al. (2013) with 2- prescribed relative intensities.
min work intervals and <1-min recovery periods resulted in Other limitations relate to the in situ assessments. First, due to
less pleasure compared with VICE. Likewise, the detrimental concerns about overloading the surveys, not all measures were

collected at each survey time point during exercise, so the PC endurance exercise in order to achieve more positive affective
scores had to be estimated using the NOCB method. Although states. In addition to this acute experimental manipulation
we consider this missing data approach to be reasonable due to study of affective response to prescribed endurance exercise
the assumption of a certain latency period of the competence modalities, future studies should expand the scope of the
experience, the shift in data may have been associated with present investigation to real-world settings (i.e., increase external
an over- or underestimation of effects. Second, for HIIE, validity) and long-term exercise adherence monitoring.
we considered only responses within the load intervals (as In the process of personalized exercise programming, trait
representants of the severe-intensity domain). However, looking differences (i.e., individuals’ preference for and tolerance of
at the whole, the intermittent nature of interval exercise seems exercise intensity or modality; Ekkekakis et al., 2005b) as well
to play an important role (Stork et al., 2018). Especially when it as state differences (i.e., individuals’ pre-exercise physical and
comes to predicting subsequent exercise behavior on the basis of mental “readiness-to-exercise”; Strohacker and Zakrajsek, 2016)
in-task response, fluctuations in affect during both the rest and need to be considered. Thus, HIIE may be a viable, time-efficient
load intervals should be considered in future studies. Third, the strategy for some individuals and certain occasions in obtaining
5-min cool-down period in HIIE compared with VICE made it positive psychological responses and long-term health benefits.
difficult to compare the extent of the affective rebound. It would A combination of both HIIE and continuous exercise should
certainly have been worthwhile to have considered affective be considered, in order to better utilize the advantages of each
valence not only immediately after the end of exercise, but also modality and to bring more variety and flexibility into the daily
at several time points in the post-exercise period (e.g., 10, 20, and exercise routine.
30 min post-exercise; Decker and Ekkekakis, 2017), but this was Our results suggest that, in addition to bolstering one’s self-
not possible due to other study interests. Last, in situ assessments perception of competence, changing individual interpretations
were carried out in different environmental conditions with a of interoceptive cues (i.e., cognitive reframing) may be one
rather sterile laboratory setting (VICE) on the one hand and avenue to increase pleasure during exercise, especially in the
a less standardized training area (MICE and HIIE) on the heavy intensity domain. Consequently, in addition to dissociative
other. Although we took this potential confounder into account strategies of directing attention away from bodily symptoms
in preliminary analyses, a minor influence of environment- (i.e., producing an external attentional focus by exercising with
related characteristics on affective response to exercise cannot be music/video; Karageorghis et al., 2021 or exercising in natural
completely ruled out. environments; Bourke et al., 2020), there is also potential for
Moreover, the three exercise modalities assessed in the present the use of associative strategies that consciously address the
study should be considered as only a selection of a large pool (unpleasant) bodily sensations themselves (e.g., mindfulness
of possible options. In particular, we solely examined one HIIE practices during exercise; Cox et al., 2020).
protocol, so no conclusions can be drawn about which specific Collectively, this study provides insight into how exercise
variables of HIIE influenced acute affective response in our study. can be structured to elicit more positive affective states, and
Because there is high variability in HIIE protocol configurations contributes to a theory-based foundation for the development
among studies that limits the generalizability of our and previous and implementation of more individualized exercise promotion
findings, future research should emphasize a more detailed interventions, thereby improving the subjective experience
comparison of different HIIE sessions to determine optimally of exercise.
configured protocols. A balanced work-to-rest ratio as well as
shorter total duration compared with continuous exercise seems
to be promising.
DATA AVAILABILITY STATEMENT
Lastly, only 40 participants could be recruited for the elaborate The raw data supporting the conclusions of this article will be
within-subject design with 15-weeks of training and three made available by the authors, without undue reservation.
extensive diagnostic blocks that did not allow for extended
absences. Due to this low sample size, observed effects have
to be interpreted with caution. In future studies, accounting ETHICS STATEMENT
for interindividual differences, it should be tested how stable
these effects are, in order to improve the generalizability of the The studies involving human participants were reviewed and
results. In this context, due to the highly standardized ergometer approved by Ethics Committee of the Medical Faculty, University
training, the limited external validity should be mentioned as of T(# 882/2017BO1) and was registered at the German Clinical
another limiting factor. Trials Register (# DRKS00017446). The participants provided
their written informed consent to participate in this study.
Practical Implications and Conclusion
The current findings provide us with a more comprehensive AUTHOR CONTRIBUTIONS
understanding of insufficiently active people’s acute affective
response to MICE, VICE, and HIIE. When it comes to KD, FM, AN, AT, and GS conceptualized the paper. KD and FM
aerobic exercise prescriptions, one size does not fit all due to performed the measurements and prepared the data. AN, AT, and
interindividual differences. Thus, health promotion practitioners GS supervised the study. PM contributed to the planning of the
should offer beginners the opportunity to try different forms of statistical procedures. KD and IR analyzed the data. KD wrote the

first draft of the manuscript. FM provided the information for ACKNOWLEDGMENTS

the procedures on the standardization of exercise intensity and
IR for the model descriptions in the statistics section. All authors We want to thank the iReAct Team for their support in planning
discussed and revised the manuscript before submission and read the multidisciplinary project and carrying out the extensive
and approved the final manuscript. data collection.
FUNDING SUPPLEMENTARY MATERIAL

The iReAct study was funded as a doctoral training network The Supplementary Material for this article can be found
by the Ministry of Research, Science, and the Arts of Baden- online at: https://www.frontiersin.org/articles/10.3389/fpsyg.
Württemberg; State Postgraduate Fellowship Programme, 2021.686661/full#supplementary-material
Germany (GZ.: II.2–7631 02). We also acknowledge
support by the Open Access Publishing Fund of University
of Tübingen.
REFERENCES other improbable dualistic creatures, exercising. Int. Rev. Sport Exerc. Psychol.
2, 139–160. doi: 10.1080/17509840902829323
Acevedo, E. O., and Ekkekakis, P. (2006). “Psychobiology of physical activity: Ekkekakis, P., and Acevedo, E. O. (2006). “Affective responses to acute exercise:
integration at last!,” in Psychobiology of Physical Activity, eds E. O. Acevedo and toward a psychobiological dose-response model,” in Psychobiology of Physical
P. Ekkekakis (Champaign, IL: Human Kinetics), 1–14. Activity, eds E. O. Acevedo and P. Ekkekakis (Champaign, IL: Human
Alicea, S. K., Parrott, A. D., Manos, T. M., and Kwon, Y. S. (2020). Comparison Kinetics), 91–109.
of the affective responses to continuous training and high-intensity interval Ekkekakis, P., and Dafermos, M. (2012). “Exercise is a many-splendored thing but
training protocols: application of the dual-mode model. J. Strength Cond. Res. for some it does not feel so splendid: staging a resurgence of hedonistic ideas in
33, 1295–1304. doi: 10.1519/JSC.0000000000003282 the quest to understand exercise behavior,” in The Oxford Handbook of Exercise
Backhouse, S. H., Ekkekakis, P., Biddle, S. J., Foskett, A., and Williams, C. (2007). Psychology, ed E. O. Acevedo (New York, NY: Oxford Press), 295–333.
Exercise makes people feel better but people are inactive: paradox or artifact? J. Ekkekakis, P., Hall, E. E., and Petruzello, S. J. (2005b). Some like it vigorous:
Sport Exerc. Psychol. 29, 498–517. doi: 10.1123/jsep.29.4.498 measuring individual differences in the preference for and tolerance of exercise
Bandura, A. (1977). Social Learning Theory. Englewood Cliffs, NJ: Prentice Hall. intensity. J. Sport Exerc. Psychol. 27, 350–374. doi: 10.1123/jsep.27.3.350
Basso, J. C., and Suzuki, W. A. (2017). The effects of acute exercise on mood, Ekkekakis, P., Hall, E. E., and Petruzzello, S. J. (2005a). Variation and homogeneity
cognition, neurophysiology, and neurochemical pathways: a review. Brain in affective responses to physical activity of varying intensities: an alternative
Plast. 2, 127–152. doi: 10.3233/BPL-160040 perspective on dose-response based on evolutionary considerations. J. Sports
Batacan, R. B. Jr., Duncan, M. J., Dalbo, V. J., Tucker, P. S., and Fenning, A. S. Sci. 23, 477–500. doi: 10.1080/02640410400021492
(2017). Effects of high-intensity interval training on cardiometabolic health: a Ekkekakis, P., Hall, E. E., and Petruzzello, S. J. (2008). The relationship between
systematic review and meta-analysis of intervention studies. Br. J. Sports Med. exercise intensity and affective responses demystified: to crack the 40-year-
51, 494–503. doi: 10.1136/bjsports-2015-095841 old nut, replace the 40-year-old nutcracker! Ann. Behav. Med. 35, 136–149.
Binder, R. K., Wonisch, M., Corra, U., Cohen-Solal, A., Vanhees, L., Saner, H., et al. doi: 10.1007/s12160-008-9025-z
(2008). Methodological approach to the first and second lactate threshold in Ekkekakis, P., Hartman, M. E., and Ladwig, M. A. (2020). “Affective responses to
incremental cardiopulmonary exercise testing. Eur. J. Cardiovasc. Prev. Rehabil. exercise,” in Handbook Sport Psychology, eds G. Tenenbaum and R. C. Eklund
15, 726–734. doi: 10.1097/HJR.0b013e328304fed4 (Toronto, ON: John Wiley and Sons, Inc.), 231–253.
Bourke, M., Hilland, T. A., and Craike, M. (2020). Variance in the valenced Ekkekakis, P., Parfitt, G., and Petruzzello, S. J. (2011). The pleasure and displeasure
response during moderate-to-vigorous physical activity: a review of people feel when they exercise at different intensities: decennial update and
cognitive and contextual mechanisms. Int. Rev. Sport Exerc. Psychol. 1–32. progress towards a tripartite rationale for exercise intensity prescription. Sports
doi: 10.1080/1750984X.2020.1780626 Med. 41, 641–671. doi: 10.2165/11590680-000000000-00000
Box, A. G., Feito, Y., Zenko, Z., and Petruzzello, S. J. (2020). The affective Ekkekakis, P., Zenko, Z., Ladwig, M. A., and Hartman, M. E. (2018). “Affect as
interval: an investigation of the peaks and valleys during high- and moderate- a potential determinant of physical activity and exercise: critical appraisal of
intensity interval exercise in regular exercisers. Psychol. Sport Exerc. 49:101686. an emerging research field”, in Affective Determinants of Health Behaviour, eds
doi: 10.1016/j.psychsport.2020.101686 D. M. Williams, R. E. Rhodes, and M. T. Conner (New York, NY: Oxford
Cox, A. E., Ullrich-French, S., Hargreaves, E. A., and McMahon, A. K. (2020). University Press), 1–24.
The effects of mindfulness and music on affective responses to self-paced Engels, J., and Diehr, P. (2003). Imputation of missing longitudinal
treadmill walking. Sport Exerc. Perform. Psychol. 9, 571–584. doi: 10.1037/spy00 data: a comparison of methods. J. Clin. Epidemiol. 56, 968–976.
00192 doi: 10.1016/S0895-4356(03)00170-7
Deci, E. L., and Ryan, R. M. (1985). Intrinsic Motivation and Self-Determination in Finger, J. D., Tafforeau, J., Gisle, L., Oja, L., Ziese, T., Thelen, J., et al.
Human Behavior. New York, NY: Plenum. (2015). Development of the European health interview survey-physical activity
Decker, E. S., and Ekkekakis, P. (2017). More efficient, perhaps, but at questionnaire (EHIS-PAQ) to monitor physical activity in the European Union.
what price? Pleasure and enjoyment responses to high intensity interval Arch. Public Health 73, 1–11. doi: 10.1186/s13690-015-0110-z
exercise in low-active women with obesity. Psychol. Sport Exerc. 28, 1–10. Gillen, J. B., and Gibala, M. J. (2014). Is high-intensity interval training a time-
doi: 10.1016/j.psychsport.2016.09.005 efficient exercise strategy to improve health and fitness? Appl. Physiol. Nutr.
Ekkekakis, P. (2003). Pleasure and displeasure from the body: perspectives from Metab. 39, 409–412. doi: 10.1139/apnm-2013-0187
exercise. Cogn. Emot. 17, 213–239. doi: 10.1080/02699930302292 Haines, M., Broom, D., Gillibrand, W., and Stephenson, J. (2020). Effects of three
Ekkekakis, P. (2005). “The study of affective responses to acute exercise: the dual- low-volume, high-intensity exercise conditions on affective valence. J. Sports
mode model,” in New Approaches to Sport and Exercise Psychology, eds R. Stelter Sci. 38, 121–129. doi: 10.1080/02640414.2019.1684779
and K. K. Roessler (Oxford: Meyer and Meyer Sport), 119–146. Hardy, C. J., and Rejeski, W. J. (1989). Not what, but how one feels: the
Ekkekakis, P. (2009). The dual-mode theory of affective responses to exercise in measurement of affect during exercise. J. Sport Exerc. Psychol. 11, 304–317.
metatheoretical context: II. Bodiless heads, ethereal cognitive schemata, and doi: 10.1123/jsep.11.3.304

Hartman, M. E., Ekkekakis, P., Dicks, N. D., and Pettitt, R. W. (2019). Dynamics Rose, E. A., and Parfitt, G. (2007). A quantitative analysis and qualitative
of pleasure–displeasure at the limit of exercise tolerance: conceptualizing the explanation of the individual differences in affective responses to prescribed
sense of exertional physical fatigue as an affective response. J. Exp. Biol. and self-selected exercise intensities. J. Sport Exerc. Psychol. 29, 281–309.
222:jeb186585. doi: 10.1242/jeb.186585 doi: 10.1123/jsep.29.3.281
Hofmann, P., and Tschakert, G. (2017). Intensity- and duration- Rose, E. A., and Parfitt, G. (2010). Pleasant for some and unpleasant for others:
based options to regulate endurance training. Front. Physiol. 8:337. a protocol analysis of the cognitive factors that influence affective responses
doi: 10.3389/fphys.2017.00337 to exercise. Int. J. Behav. Nutr. Phys. Act. 7:15. doi: 10.1186/1479-5868-
Jung, M. E., Bourne, J. E., and Little, J. P. (2014). Where does HIT fit? 7-15
An examination of the affective response to high-intensity intervals in Russell, J. A., and Barrett, L. F. (1999). Core affect, prototypical emotional episodes,
comparison to continuous moderate- and continuous vigorous-intensity and other things called emotion: dissecting the elephant. J. Pers. Soc. Psychol. 76,
exercise in the exercise intensity-affect continuum. PLoS ONE 9:e114541. 805–819. doi: 10.1037/0022-3514.76.5.805
doi: 10.1371/journal.pone.0114541 Solomon, R. L. (1980). The opponent-process theory of acquired motivation:
Karageorghis, C. I., Jones, L., Howard, L. W., Thomas, R. M., Moulashis, P., the costs of pleasure and the benefits of pain. Am. Psychol. 35, 691–712.
and Santich, S. J. (2021). When it HIITs, you feel no pain: psychological and doi: 10.1037/0003-066X.35.8.691
psychophysiological effects of respite-active music in high-intensity interval Solomon, R. L. (1991). “Acquired motivation and affective opponent-processes,” in
training. J. Sport. Exerc. Psychol. 43, 41–52. doi: 10.1123/jsep.2019-0335 Neurobiology of Learning, Emotion, and Affect, ed. J. Madden (New York, NY:
Kellogg, E., Cantacessi, C., McNamer, O., Holmes, H., von Bargen, R., Ramirez, Raven), 307–347.
R., et al. (2019). Comparison of psychological and physiological responses to Stevens, C. J., Baldwin, A. S., Bryan, A. D., Conner, M., Rhodes, R. E.,
imposed vs. self-selected high-intensity interval training. J. Strength. Cond. Res. and Williams, D. M. (2020). Affective determinants of physical activity:
33, 2945–2952. doi: 10.1519/JSC.0000000000002528 a conceptual framework and narrative review. Front. Psychol. 11:568331.
Kilpatrick, M. W., Greeley, S. J., and Collins, L. H. (2015). The impact of doi: 10.3389/fpsyg.2020.568331
continuous and interval cycle exercise on affect and enjoyment. Res. Q. Exerc. Stork, M. J., Banfield, L. E., Gibala, M. J., and Martin Ginis, K. A. (2017). A scoping
Sport 86, 244–251. doi: 10.1080/02701367.2015.1015673 review of the psychological responses to interval exercise: is interval exercise
Maibach, M., Niedermeier, M., Sudeck, G., and Kopp, M. (2020). Erfassung a viable alternative to traditional exercise? Health Psychol. Rev. 11, 324–344.
unmittelbarer affektiver Reaktionen auf körperliche Aktivität. Z. Sportpsychol. doi: 10.1080/17437199.2017.1326011
27, 4–12. doi: 10.1026/1612-5010/a000291 Stork, M. J., Gibala, M. J., and Martin, G. K. (2018). Psychological and behavioral
Malik, A. A., Williams, C. A., Weston, K. L., and Barker, A. R. (2019). responses to interval and continuous exercise. Med. Sci. Sports Exerc. 50,
Perceptual and cardiorespiratory responses to high-intensity interval exercise 2110–2121. doi: 10.1249/MSS.0000000000001671
in adolescents: does work intensity matter? J. Sports Sci. Med. 18, 1–12. Strohacker, K., and Zakrajsek, R. A. (2016). Determining dimensionality of exercise
Martinez, N., Kilpatrick, M. W., Salomon, K., Jung, M. E., and Little, J. P. readiness using exploratory factor analysis. J. Sports Sci. Med. 15, 229–238.
(2015). Affective and enjoyment responses to high-intensity interval training Sudeck, G., and Conzelmann, A. (2014). Inter-individual variability of acute
in overweight-to-obese and insufficiently active adults. J. Sport Exerc. Psychol. affective responses in the course of leisure time and health-oriented
37, 138–149. doi: 10.1123/jsep.2014-0212 exercise programs. Z Gesundheitspsychol. 22, 89–103. doi: 10.1026/0943-8149/
Mattioni Maturana, F., Martus, P., Zipfel, S., and Nieß, A. M. (2021a). a000118
Effectiveness of HIIE versus MICT in improving cardiometabolic risk factors Thiel, A., Sudeck, G., Gropper, H., Mattioni Maturana, F., Schubert, T., Srismith,
in health and disease: a meta-analysis. Med. Sci. Sports Exerc. 53, 559–573. D., et al. (2020). The iReAct study – a biopsychosocial analysis of the individual
doi: 10.1249/MSS.0000000000002506 response to physical activity. Contemp. Clin. Trials Commun. 17:100508.
Mattioni Maturana, F., Schellhorn, P., Erz, G., Burgstahler, C., Widmann, M., doi: 10.1016/j.conctc.2019.100508
Munz, et al. (2021b). Individual cardiovascular responsiveness to work- Upton, G. J. G., and Cook, I. (1996). Understanding Statistics. Oxford: Oxford
matched exercise within the moderate- and severe-intensity domains. Eur. J. University Press.
Appl. Physiol.121, 2039–2059. doi: 10.1007/s00421-021-04676-7 Williams, D. M., Dunsiger, S., Jennings, E. G., and Marcus, B. H. (2012).
Niven, A., Laird, Y., Saunders, D. H., and Phillips, S. M. (2020). A systematic Does affective valence during and immediately following a 10-min walk
review and meta-analysis of affective responses to acute high intensity interval predict concurrent and future physical activity? Ann. Behav. Med. 44, 43–51.
exercise compared with continuous moderate- and high-Intensity exercise. doi: 10.1007/s12160-012-9362-9
Health Psychol. Rev. 25, 1–34. doi: 10.1080/17437199.2020.1728564
Niven, A., Thow, J., Holroyd, J., Turner, A. P., and Phillips, S. M. (2018). Conflict of Interest: The authors declare that the research was conducted in the
Comparison of affective responses during and after low volume high- absence of any commercial or financial relationships that could be construed as a
intensity interval exercise, continuous moderate- and continuous high- potential conflict of interest.
intensity exercise in active, untrained, healthy males. J. Sports Sci. 36,
1993–2001. doi: 10.1080/02640414.2018.1430984 Publisher’s Note: All claims expressed in this article are solely those of the authors
Oliveira, B. R., Deslandes, A. C., and Santos, T. M. (2015). Differences in exercise and do not necessarily represent those of their affiliated organizations, or those of
intensity seems to influence the affective responses in self-selected and imposed
the publisher, the editors and the reviewers. Any product that may be evaluated in
exercise: a meta-analysis. Front. Psychol. 6:1105. doi: 10.3389/fpsyg.2015.01105
this article, or claim that may be made by its manufacturer, is not guaranteed or
Oliveira, B. R. R., Slama, F. A., Deslandes, A. C., Furtado, E. S., and Santos, T.
M. (2013). Continuous and high-intensity interval training: which promotes endorsed by the publisher.
higher pleasure? PLoS ONE 8:e79965. doi: 10.1371/journal.pone.0079965
Rhodes, R. E., and Kates, A. (2015). Can the affective response to exercise predict Copyright © 2021 Dierkes, Mattioni Maturana, Rösel, Martus, Nieß, Thiel and
future motives and physical activity behavior? A systematic review of published Sudeck. This is an open-access article distributed under the terms of the Creative
evidence. Ann. Behav. Med. 49, 715–731. doi: 10.1007/s12160-015-9704-5 Commons Attribution License (CC BY). The use, distribution or reproduction in
Roloff, Z. A., Dicks, N. D., Krynski, L. M., Hartman, M. E., Ekkekakis, P., other forums is permitted, provided the original author(s) and the copyright owner(s)
and Pettitt, R. W. (2020). Ratings of affective valence closely track changes are credited and that the original publication in this journal is cited, in accordance
in oxygen uptake: application to high-intensity interval exercise. Perform. with accepted academic practice. No use, distribution or reproduction is permitted
Enhance. Health 7:100158. doi: 10.1016/j.peh.2020.100158 which does not comply with these terms.

Different Exercise Endurance Modalities

Uploaded by

Copyright:

Available Formats

Different Exercise Endurance Modalities

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Different Exercise Endurance Modalities

Uploaded by

Copyright:

Available Formats

ORIGINAL RESEARCH

published: 13 August 2021

Different Endurance Exercise

Affect experienced during an exercise session is supposed to predict future exercise

Frontiers in Psychology | www.frontiersin.org 1 August 2021 | Volume 12 | Article 686661

INTRODUCTION represent a challenge to the adaptive capacity of the body,

Frontiers in Psychology | www.frontiersin.org 2 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 3 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 4 August 2021 | Volume 12 | Article 686661

TABLE 1 | Demographic and anthropometric characteristics of participants at Cognitive Factors

Frontiers in Psychology | www.frontiersin.org 5 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 6 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 7 August 2021 | Volume 12 | Article 686661

TABLE 2 | Descriptive statistics for study variables.

Variable Between-person variability Within-person variability

ICC N M SD Range N mV SD Range

exercise modalities were within the targeted intensity domains

to be significantly higher in MICE than in VICE and HIIE.

Frontiers in Psychology | www.frontiersin.org 8 August 2021 | Volume 12 | Article 686661

MODEL 1: in-task affective valencea t1 vs. t0 0.250 0.107 0.119

For AOI (Model 3b), we observed significant main effects for

Frontiers in Psychology | www.frontiersin.org 9 August 2021 | Volume 12 | Article 686661

domain (VICE) and an interval exercise in the severe-intensity

Frontiers in Psychology | www.frontiersin.org 10 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 11 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 12 August 2021 | Volume 12 | Article 686661

first draft of the manuscript. FM provided the information for ACKNOWLEDGMENTS

FUNDING SUPPLEMENTARY MATERIAL

Frontiers in Psychology | www.frontiersin.org 13 August 2021 | Volume 12 | Article 686661

Frontiers in Psychology | www.frontiersin.org 14 August 2021 | Volume 12 | Article 686661

You might also like