The affective beneficence of vigorous exercise revisted

Abstract

Objectives. High exercise intensity may be associated with reduced adherence to exercise programmes, possibly because it is perceived as aversive. However, several authors have suggested that an intensity as high as 60% or 70% of maximal aerobic capacity (VO 2max ) is necessary for exercise to elicit positive affective changes. To elucidate this discrepancy, the affective responses to increasing levels of exercise intensity were examined.
Design. In total, 30 volunteers rated their affect every minute as they ran on a treadmill while the speed and grade were progressively increased.
Method. The methodology was unique in three respects: (1) affect was assessed in terms of the dimensions of the circumplex model instead of distinct affective states, (2) affect was assessed repeatedly before, during, and after exercise, not only before and after, and (3) exercise intensity was standardized across participants in terms of metabolically comparable phases (beginning, ventilatory threshold, VO 2max ) instead of percentages of maximal capacity.
Results. Pre‐to‐post‐exercise comparisons indicated affective benefits in the form of increased energetic arousal and decreased tense arousal. During exercise, however, affective valence deteriorated beyond the ventilatory threshold and until VO 2max , a trend that reversed itself instantaneously during cool‐down.
Conclusions. Exercise intensity that requires a transition to anaerobic metabolism can have a transient but substantial negative impact on affect and this may, in turn, reduce adherence to exercise programmes.

Full text

The affective bene cence of vigorous
exercise revisited
Eric E. Hall, Panteleimon Ekkekakis* and Steven J. Petruzzello
Department of Kinesiology, University of Illinois at Urbana-Champaign, USA
Objectives. High exercise intensity may be associated with reduced adherence to
exercise programmes, possibly because it is perceived as aversive. However, several
authors have suggested that an intensity as high as 60% or 70% of maximal aerobic
capacity (VO2m a x) is necessary for exercise to elicit positive affective changes. To
elucidate this discrepancy, the affective responses to increasing levels of exercise
intensity were examined.
Design. In total, 30 volunteers rated their affect every minute as they ran on a
treadmill while the speed and grade were progressively increased.
Method. The methodology was unique in three respects: (1) affect was assessed
in terms of the dimensions of the circumplex model instead of distinct affective states,
(2) affect was assessed repeatedly before, during , and after exercise, not only before
and after, and (3) exercise intensity was standardized across participants in terms of
metabolically comparable phases (beginning, ventilatory threshold, VO2m ax ) instead
of percentages of maximal capacity.
Results. Pre-to-post-exercise comparisons indicated affective bene ts in the form of
increased energetic arousal and decreased tense arousal. During exercise, however,
affective valence deteriorated beyond the ventilatory threshold and until VO2max, a
trend that reversed itself instantaneousl y during cool-down.
Conclusions. Exercise intensity that requires a transition to anaerobic metabolism
can have a transient but substantial negative impact on affect and this may, in turn,
reduce adherence to exercise programmes.
Physical activity surveys indicate that the proportion of the population that engages in
regular physical activity is alarmingly low. Specically, in the United States, the 1998 –99
progress review of the Healthy People 2000 programme shows that only 23% of adults
engage in light-to-moderate physical activity ve times per week and only 16% engage
in light-to-moderate physical activity seven times per week (United States National
Center for Health Statistics, 1999). Both  gures have remained virtually unchanged since
47
British Journal of Health Psychology (2002), 7, 47 –66
2002 The British Psychologi cal Society
www.bps.org.uk
*Requests for reprints should be addressed to Panteleimon Ekkekakis, 253 Barbara E. Forker Building, Department of
Health and Human Performance, Iowa State University, Ames, IA 50011, USA (e-mail: ekkekaki@iastate.edu).

1985 (22% and 16%, respectively) and both fall short of the targets for the year 2000
(30% for both categories).
In an effort to increase physical activity participation and adherence rates, recently
issued physical activity recommendations have called for activities of moderate inten-
sity, such as walking (National Institutes of Health Consensus Development Panel on
Physical Activity and Cardiovascular Health, 1996; Pate et al., 1995; United States
Department of Health and Human Services, 1996). These recommendations were
inuenced by two main considerations. First, there is accumulating evidence that
some activity is better than no activity for accruing health benets. Second, the focus
on moderate intensity activities was motivated by the belief that such activities are
likely to be more enjoyable or at least more tolerable (Brewer, Manos, McDevitt,
Cornelius, & Van Raalte, 2000) and, as a result, they are also ‘more likely to be continued
than are high-intensity activities’ (National Institute of Health Consensus Development
Panel on Physical Activity and Cardiovascular Health, 1996, p. 243).
The notion of an inverse relationship between exercise intensity and adherence
has been supported by several studies (Epstein, Koeske, & Wing, 1984; Lee et al., 1996;
Sallis et al., 1986). However, it is still unknown whether this effect is moderated by
the participants’ affective responses to physical activities of different intensities. To
date, no published studies have examined the association between affective responses
to single bouts of activity performed at varying intensities and long-term adherence.
The few studies that have examined the effects of different exercise intensities on
adherence and have included assessments of psychological variables examined only
chronic psychological changes (i.e. changes over the course of the entire training
period, not in response to single bouts of activity). These studies either showed
signicant intensity effects on psychological outcomes but no effect on adherence
(Moses, Steptoe, Mathews, & Edwards, 1989) or no intensity effects on either psycho-
logical outcomes or adherence (Blumenthal, Emery, & Rejeski, 1988; King, Haskell,
Taylor, Kraemer, & DeBusk, 1991; King, Haskell, Young, Oka, & Stefanick, 1995; King,
Taylor, & Haskell, 1993). In summary, although direct empirical evidence is still
lacking, a causal chain from intense exercise to negative affect and, ultimately, to
reduced adherence remains a distinct possibility.
The assumption of high exercise intensity having a negative impact on affect
and adherence, however, appears to be in contrast to a long-held belief in exercise
psychology, namely that, in order to produce affective benets, physical activity must
be performed at an intensity characterized as ‘vigorous’ as opposed to ‘mild’. This
idea has remained popular since Morgan’s (1985) inuential review published under
the title ‘Affective benecence of vigorous physical activity’. In this oft-cited paper,
Morgan asserted that ‘improved a ffective states a ccompany. . . acute . . . physical activity
of a vigorous nature’ (p. 99). Furthermore, it has been suggested that a vigorous
exercise intensity is a necessary condition for affective benets. Specically, some
authors have proposed that, in order to achieve signicant reductions in state anxiety,
exercise intensity should exceed 60% (Raglin & Morgan, 1985) or 70% (Dishman, 1986)
of maximal aerobic capacity (VO2max). Similarly, for mood enhancement (Berger &
Motl, 2000), psychotherapeutic effects (Hays, 1999; Ojanen, 1994), or general psycho-
logical bene ts (Kirkcaldy & Shepherd, 1990), vigorous exercise intensity is believed
to be necessary.
Therefore, the exercise science literature appears to contain a puzzling contra-
diction. On the one hand, there is a belief in the ‘affective benecence’ of vigorous
physical activity. On the other hand, recent physical activity recommendations reect
48 Eric E. Hall et al.

the notion that high exercise intensity may be partly responsible for the high dropout
rates, presumably because it is experienced as aversive. These two views are difcult
to reconcile. Research in general psychology has shown that people are likely to do
what makes them feel good and avoid what makes them feel bad. For example, Emmons
and Diener (1986) showed that the positive affect experienced in a situation was a
good predictor of the amount of time people chose to spend in that situation. If vigo-
rous physical activity is associated with positive affective changes, then why are most
people avoiding vigorous physical activity?
One way to resolve this seeming paradox is to critically reexamine the relationship
between exercise intensity and affective responses. A recent review of this literature
concluded that the extant studies do not allow any denitive statements to be made
regarding the form of the dose-response relationship between exercise and affect
(Ekkekakis & Petruzzello, 1999). This can be attributed to several key conceptual and
methodological issues, including (a) the measurement of affect, (b) the timing of
assessment of affective responses, and (c) the selection of exercise intensity levels.
These issues are discussed in more detail next, along with the tentative solutions
incorporated in the present study.
The measurement of affect
Among the various problems that reviewers of the literature on the relationship
between exercise and affect have identied, the measurement of affect has probably
received most critical attention (Byrne & Byrne, 1993; Ekkekakis & Petruzzello, 1999;
McAuley & Rudolph, 1995; Mutrie & Biddle, 1995; Steptoe, 1992; Tuson & Sinyor, 1993).
In most studies of this type, affect has been operationalized in terms of distinct states,
such as anxiety, depression, or various sets of moods. This approach is termed
‘categorical’ because it reects the assumption that affective states are unrelated,
organized in conceptually distinct categories. In contrast, ‘dimensional’ approaches
are based on the assumption that affective states are systematically interrelated, such
that their relationships can be modelled by a parsimonious set of dimensions. Both
categorical and dimensional models have relative strengths and weaknesses, therefore
the decision to adopt one or the other depends on the nature of the research problem.
The main advantage attributed to dimensional models is their parsimony; that is, they
can account for a large portion of the variation in affective states in terms of only a few
basic dimensions (Larsen & Diener, 1992). Thus, they can provide a broad investigative
scope, which is particularly useful when the goal is to describe the nature of the
affective changes that occur in a given situation. Discussing the implications of this issue
for the study of the affective changes associated with physical activity, Gauvin and
Brawley (1993) noted that:
. . . (a dim ensional approach) see ms better suited to the study of exercise and affect because
the models stemming from it are intended to be broad, encompassing conceptualizations
of affective experience. Because the affective experience that accompanies exercise has not
been thoroughly described, a model of affect that has a wider breadth is more likely to
capture the essence of exercise-induced affect than a model that, at the outset, limits the
focus of investigation to specic emotions (p. 152).
Thus, since there is at present no evidence that changes in anxiety or depression (for
example) are the most salient affective changes associated with exercise, a narrow focus
on these distinct variables is unlikely to capture the impact of exercise on affect in
general (Ekkekakis, Hall, & Petruzzello, 1999). To address this problem in the present
Affect and vigorous exercise 49

study, in accordance with Gauvin and Brawley’s (1993) and Ekkekakis and Petruzzello’s
(1999) suggestions, affect was examined from a dimensional perspective.
Specically, the two-dimensional circumplex model of affect was used, as described
by Russell (1978, 1980, 1989, 1997). According to the circumplex, the affective space
is dened by two bipolar and orthogonal dimensions: an affective valence dimension
and an activation dimension. Affective states are construed as combinations of varying
degrees of these two constituent dimensions, such that they can be conceptualized
as located around the perimeter of a circle dened by the dimensions of valence and
activation. A division of the circle into quadrants produces the following meaningful
variants: (1) unactivated pleasant affect, characteristic of relaxation and calmness,
(2) unactivated unpleasant affect, characteristic of boredom, fatigue, or depression,
(3) activated unpleasant affect, characteristic of tension and distress, and (4) activated
pleasant affect, a state characteristic of excitement and enthusiasm (see Fig. 1).
Timing of assessment of affective responses
Typically, studies examining the effects of acute exercise on affect have included one
assessment of affect before and one or more assessments after the bout of exercise.
Assessments of affect during the exercise bout have been rare, mainly as a consequence
of measuring affect through multi-item questionnaires that are impractical to administer
repeatedly during exercise. However, affect changes in a continuous and dynamic
fashion. It is, therefore, unlikely that changes from pre- to post-exercise are linear.
Assessments of affect made after exercise may only reect the effects of the few seconds
or minutes of recovery from exercise rather than the effects of the entire preceding
50 Eric E. Hall et al.
Figure 1. The circumplex model of affect

exercise bout. Affective changes during exercise may present a different and diverse
pattern. Consistent with this possibility, Ekkekakis and Petruzzello (1999) noted that,
although the majority (54%) of the studies that involved assessments of affect only
before and after exercise showed no evidence of reliable dose-response effects, six
of the seven studies that involved repeated assessments of affective valence during
exercise showed a reliable dose-response pattern. Specically, as exercise intensity
increased, affective valence was consistently shown to deteriorate (Acevedo, Rinehardt,
& Kraemer, 1994; Hardy & Rejeski, 1989; Partt & Eston, 1995a; Partt, Eston, &
Connolly, 1996; Partt, Markland, & Holmes, 1994).
Furthermore, the studies that included repeated assessments of affect both during
and after exercise have shown that an instantaneous improvement takes place as soon
as the exercise bout is terminated (Par tt et al., 1994; Steptoe & Bolton, 1988; Tate &
Petruzzello, 1995). This pattern of responses is consistent with the predictions of
the opponent-process theory of affect (Solomon, 1980, 1991; Solomon & Corbit, 1974).
When applied to exercise, this theory suggests that the initial affective reaction to
vigorous exercise is driven by a so-called ‘a-process’, resulting in negative affect.
However, the a-process arouses an opponent process, the so-called ‘b-process’, which
is characterized by the opposite affective quality (i.e. positive affect). The interaction
of these processes over time controls the intensity and the quality of the resultant
affect. The b-process is hypothesized to be of longer latency, slower build-up, and
slower decay relative to the a-process. Thus, its effects persist after the termination of
the exercise bout and may be responsible for the feelings of lowered tension and
exhilaration that are typically found post-exercise. If the opponent-process theory is
correct, then what is typically referred to as the ‘exercise-associated feel-better
phenomenon’ should be more accurately described as the ‘exercise recovery-associated
feel-better phenomenon’, since the effects of exercise and exercise recovery may in
fact be different. To investigate this possibility, in the present study, affective responses
were assessed repeatedly during and after exercise.
Selection of exercise intensity levels
One of the main challenges in studies examining the effects of exercise on affect has
been the standardization of exercise loads across individuals. Until now, the preferred
solution has been to use percentages of maximal capacity (VO2max, maximal heart
rate, or heart rate reserve). This is because, although an absolute workload (e.g. 100 W)
may induce a different metabolic response in trained versus untrained individuals
(i.e. requiring purely aerobic effort in trained but possibly eliciting an anaerobic
component in untrained), a percentage of maximal effort (e.g. 70% VO2max ) is assumed
to represent a metabolically equivalent stimulus across individuals.
However, this assumption appears to be false (Katch, Weltman, Sady, & Freedson,
1978; Meyer, Gabriel, & Kindermann, 1999). For example, Katch et al. (1978) reported
that, in a sample of 31 participants, when exercise was performed at 80% of maximal
heart rate (62.5% VO2max), 17 participants were working at a level above, whereas 14
were working at a level below metabolic acidosis (a sign of anaerobiosis). Given the
signicant differences in ventilatory, biochemical, and endocrine parameters between
aerobic and anaerobic effort, the assumption that the exercise stimulus can be
effectively standardized across individuals by selecting percentages of maximal capacity
is untenable.
As a potential solution, Ekkekakis and Petruzzello (1999) proposed taking into
Affect and vigorous exercise 51

account individually determined metabolic landmarks such as the lactate or ventilatory
(gas exchange) threshold and the level of critical power or power-time asymptote
(Gaesser & Poole, 1996). These metabolic landmarks have a profound adaptational
signicance, as the metabolic prole changes dramatically when exercise is performed
slightly below or slightly above them. Below the lactate threshold, the intensity is
characterized as ‘moderate’. Within this range, one can exercise for prolonged periods
of time, as blood lactate concentration and oxygen uptake remain stable and the
capacity for sustained energy repletion is not exceeded. From the lactate threshold
to the level of critical power (the highest work rate at which blood lactate can be
stabilized), the intensity is characterized as ‘heavy’. In this domain, lactate appearance
and removal rates are balanced, but at elevated blood lactate concentration levels.
Finally, from the level of critical power to VO2max , the intensity is characterized as
‘severe’. In this range, neither oxygen consumption nor blood lactate can be stabilized.
Both rise inexorably until exhaustion. In the present study, to achieve a more effective
standardization of exercise intensity across individuals, exercise intensity was dened
relative to individually determined metabolic landmarks, namely the ventilatory thresh-
old (as a less intrusive index of the lactate threshold, which requires repeated blood
sampling for its determination) and VO2max.
Affective responses to vigorous exercise reconceptualized
After almost three decades of research on the affective changes associated with acute
exercise, the substrates of these changes remain elusive. Several hypotheses have
been proposed and tested, but none seems to offer a comprehensive interpretive frame-
work and none has received unequivocal empirical support (for reviews, see Hateld,
1991; La Forge, 1995; Morgan, 1997a; Petruzzello, Landers, Hateld, Kubitz, & Salazar,
1991). This may be due to the fact that researchers have been searching for an
explanation of changes believed to be exclusively positive. If the possibility of negative
affective responses is acknowledged, as the empirical evidence suggests it should
(Ekkekakis & Petruzzello, 1999), then the phenomenon is placed under a new light.
According to evolutionary conceptualizations, affective responses represent adaptive
responses or, in other words, responses that have evolved to promote survival within
a specic context (Nesse, 1990, 1998). Thus, the affective responses to vigorous
exercise may be seen as adaptive responses that signal survival-critical metabolic
changes in the body. As many theorists have noted, affect represents the primary
means by which information about critical disruptions of bodily homeostasis enters
consciousness (Cabanac, 1995; Damasio, 1995; Panksepp, 1998a, b; Schulze, 1995).
With this as a backdrop, consider the implications of the typology of exercise
intensity described above from an adaptational standpoint. Within both the moderate
and the heavy domains, the maintenance of a physiological steady state is possible and,
therefore, there is no immediate threat to survival. However, the transition to anaerobic
metabolism constitutes a signicant challenge for the energy system as its performance
becomes dependent upon the nite pool of anaerobic energy sources. Above the level
of critical power, the energy supply system is overwhelmed and the maintenance of
a steady state is no longer possible. For successful adaptation, the importance of the
situation must enter consciousness, both when the organism rst experiences a
challenge (i.e. in the heavy domain) and most certainly when the organism is faced
with the impending exhaustion of metabolic resources (i.e. in the severe domain).
This conceptualization leads to the formulation of the following hypotheses. When
52 Eric E. Hall et al.

exercise intensity is in the moderate range, the maintenance of homeostasis is not
threatened and, consequently, affective responses, if any, are largely independent of
physiological changes. In the heavy domain, affective responses serve the function
of ‘alerting’ consciousness to the strain placed upon the metabolic system. In this
domain, affective responses depend partly on physiological changes and partly on
various individual-difference and cognitive factors related to coping. Thus, affective
responses are likely to vary from individual to individual and may be positive or nega-
tive. Finally, in the severe domain, affective responses represent an evolutionarily
primitive ‘alarming’ function, which, much like pain, is aimed to stop and withdraw
from the activity that is causing the severe homeostatic perturbation. In this domain,
affective responses are, therefore, likely to be driven mainly by physiological changes
and be mostly negative.
The present study
The primary purpose of the present study was to examine the pattern of affective
responses to increasing levels of exercise intensity through the prism of the conceptual
and methodological elements described in the previous sections. Thus, (a) affect was
examined from a dimensional, rather than a categorical, perspective, (b) affective
responses were assessed repeatedly, throughout the exercise bout and for several
minutes into recovery, and (c) exercise intensity was dened relative to the ventilatory
threshold and VO2max. A graded treadmill protocol was used as the methodological
platform because it allowed the examination of affective responses through the entire
range of exercise intensities. Because of the relatively short total duration of the
sessions, accumulated fatigue over increasing levels of intensity was not expected to
inuence the responses.
Method
Participants
A total of 30 healthy university students (13 women, 17 men; mean age 6SD 523.963.6years;
mean weight 6SD 571 .7611.2kg; mean VO2m ax 6SD 549.666.1ml kg21´min21) volunteered
to participate in the study. They were paid $10 each in compensation for their time. Prior to their
involvement in the study, all participants had read and signed an informed consent form approved
by the University’s Institutional Review Board.
Measures
Affect was measured from the perspective of the circumplex model, using both multi-item and
single-item instruments, the latter being more appropriate for repeated assessments during
exercise. The Feeling Scale (FS; Hardy & Rejeski, 1989) was used as a single-item measure of
affective valence and the Felt Arousal Scale (FAS) of the Telic State Measure (Svebak & Murgatroyd,
1985) was used as a single-item measure of perceived activation. The Activation Deactivation
Adjective Check List (AD ACL; Thayer, 1989) was used as a multi-item measure of the four
quadrants of circumplex affective space (see Fig. 1). All self-report measures were administered
with the standard instructions provided by their developers.
The FS (Hardy & Rejeski, 1989) is an 11-point, single-item, bipolar measure of pleasure-
displeasure, which is commonly used for the assessment of affective responses during exercise
(Ekkekakis & Petruzzello, 1999). The scale ranges from 25to 15. Anchors are provided at zero
(‘Neutral’) and at all odd integers, ranging from ‘Very Good’ (15) to ‘Very Bad’ (25). In previous
Affect and vigorous exercise 53

work in our laboratory, the FS has exhibited correlations ranging from .51 to .88 with the valence
scale of the Self Assessment Manikin (SAM; Lang, 1980) and from .41 to .59 with the valence scale
of the Affect Grid (AG; Russell, Weiss, & Mendelsohn, 1989).
The FAS (Svebak & Murgatroyd, 1985) is a 6-point, single-item measure of perceived activation.
The scale ranges from 1 to 6, with anchors at 1 (‘Low Arousal’) and 6 (‘High Arousal’). In previous
work in our laboratory, the FAS has exhibited correlations ranging from .45 to .70 with the arousal
scale of the SAM and from .47 to .65 with the arousal scale of the AG. The FAS has been used
extensively in the context of reversal theory research, including exercise-related studies (Kerr &
Vlaswinkel, 1993; Kerr & Van den Wollenberg, 1997).
The AD ACL is a multi-item measure of the bipolar dimensions of Energetic Arousal (EA) and
Tense Arousal (TA), as described by Thayer (1989). Each dimension is represented by ten 4-point
Likert-type items. The EA dimension ranges from Energy to Tiredness, whereas the TA dimension
ranges from Tension to Calmness. Thayer (1978, 1986, 1989) has provided extensive validity and
reliability information on the AD ACL. In the present study, the AD ACL was used within a
circumplex framework. As shown in Fig. 1, the Energy pole (high EA) is theorized to map the high-
activation pleasure quadrant of the circumplex, Tension (high TA) maps the high-activation
displeasure quadrant, Tiredness (low EA) maps the low-activation displeasure quadrant, and
Calmness (low TA) maps the low-activation pleasure quadrant (also see Thayer, 1989, pp. 133–
134, p. 164; Yik, Russell & Feldman-Barrett, 1999, for an empirical demonstration).
Finally, the Rating of Perceived Exertion (RPE; Borg, 1998) was used as a measure of perceived
effort during exercise. The RPE is a 15-point scale ranging from 6 to 20, with anchors ranging from
‘Very, very light’ to ‘Very, very hard’.
Procedures
On arrival at the laborator y, each participant was greeted, given an overview of the procedures to
be followed, and asked to read and sign an informed consent form. The purpose of the study was
described as an investigation of ‘some physiological and psychological responses to vigorous
exercise’. This was followed by the tting of a heart-rate monitor (model Vantage XL; Polar Electro
Oy, Finland). Once the integrity of the signal from the monitor was established, the participants
were asked to complete a pre-exercise battery of questionnaires that included the FS, FAS, and AD
ACL. Next, the participants were shown to a treadmill, were presented with a description of the
exercise protocol, and were tted with a face mask equipped with a low-resistance one-way valve
(Hans Rudolph, Kansas City, MO) for the collection of expired gases, ensuring that respiration was
unobstructed and comfortable.
The graded exercise protocol was as follows. First the oxygen and carbon dioxide analysers
(models S-3A/I and CD-3A, respectively; Ametek Applied Electrochemistry, Pittsburgh, PA) were
calibrated using a gas with a known mixture of oxygen and carbon dioxide and room air. Then the
participants’ expired gases were analysed for 2 min while they were seated on a stool, to ensure
the proper functioning of the various components of the metabolic analysis system. This was
followed by a 3-m in walk at 4.8 km hr21(0% grade), which served as a warm-up. Once the warm-
up was completed, the speed of the treadmill was increased to 8 km hr21(0% grade). Beyond this
point, the workload was increased every 2 min by alternatin g between increases in speed by
1.6 km hr21and increases in grade by 2%. This procedure was continued until each participant
reached the point of volitional exhaustion. This was veried by at least two of the standard criteria
for reaching VO2max , namely (a) reaching a peak or plateau in oxygen consumption (changes of
less tha n 2 ml kg21min21) followed immediately by a decrease in consumption with increasing
workloads; (b) attaining a respiratory exchange ratio equal to or higher than 1.1; and (c) reaching
or exceeding age-predicted maximal heart rate (i.e. 220 2age). The participants then cooled
down by walking on the treadmill for 2 min at 4.8 kmhr21and 0% grade. Finally, the participants
sat in a chair doing nothing for a recovery period of 20 min.
From the beginning of the incremental phase (8 km hr21; 0% grade) until the end of the cool-
down, the participants gave self-ratings on the RPE, FS, and FAS every minute by pointing out their
54 Eric E. Hall et al.

selections on a poster-size version of the scales which was placed directly in front of them. After
each response, a research assistan t repeated the participant’s selection out loud to ensure that the
information would be recorded correctly. Immediately after the cool-down, as well as 10 and 20
min later, the participants completed the FS, FAS, and AD ACL again.
Data reduction and analysis
Given that the duration of the graded treadmill protocol varied between individuals, exercise
intensity was standardized using the following times, which were considered to reect metabo-
lically comparable conditions across all participants: (a) the beginning of exercise, (b) the
ventilatory threshold (VT), and (c) the end of exercise. Two methods were followed to determine
the VT, one graphical and one quantitative. The graphical method, described by Davis, Frank,
Whipp, and Wasserman (1979), entails plotting the ventilatory equivalents for oxygen (VE/VO2)
and carbon dioxide (V
E/VCO2) across work rates and visually identifying the point at which there
is a systematic increase in VE/VO2without a corresponding increase in V
E/VCO2. The quantitative
method involves an iterative regression procedure, described by Jones and Molitoris (1984),
that can identify the ‘break point’ between two regression lines running through the data points
dened by oxygen consumed and carbon dioxide produced. These techniques have been shown
to lead to a determination of the VT with satisfactory accuracy (Ahmaidi et al., 1993; Caiozzo et al.,
1982).
Following the identication of the VT, the FS and FAS ratings made at the following ten time
points during exercise were retained: the rst 2 min, the minute before the VT, the minute of the
VT, 2 minutes following the VT, the last 2 min, and the 2 min of the cool-down period. In order to
reduce the number of data points in the repeated-measures analyses, the patterns in the data were
examined and the following  ve time points were entered in statistical analyses: minute 2, the
minute of the VT, the second minute after the VT, the last minute, and the second cool-down
minute (i.e. every alternate one of the above 10 time points).
For all multivariate analyses, the self-report scales were organized in two pairs, with each pair
considered to provide an independent representation of the circumplex affective space. One pair
consisted of EA and TA and the second pair consisted of the FS and FAS. As noted earlier, the
affective dimensions represented by EA and TA were theorized to be 458rotational variants of the
dimensions represented by FS and FAS (see Fig. 1).
A central purpose of the present investigation was to contrast the patterns of affective change
that emerge from protocols limited to pre- and post-exercise assessments as opposed to repeated
assessments during exercise. To address this issue, two sets of analyses were performed. The  rst
set involved an examination of change across four time points: pre-exercise, post-09, post-109, and
post-209. These analyses involved both EA/TA and FS/FAS. The second set examined changes in
affect during exercise using the FS/FAS.
Analyses of change across time began with repeated-measures multivariate analyses of
variance (MANOVAs) on each of the two sets of self-report scales. Statistically signicant  ndings
were followed up by univariate analyses and an examination of simple effects within each scale,
using Fisher-Hayter tests (qFH ) for pairwise comparisons. Moreover, effect sizes were computed
[d5(Mi2Mj)/SDpooled] to assess the magnitude of the differences.
Results
The average duration of exercise until the point of volitional exhaustion was 11.3
minutes (SD 52.29 minutes; range 8–17 minutes). The average terminal RPE was 17.77
(SD 51.89; between ‘Very hard’ and ‘Very, very hard’).
To ensure that the AD ACL scales performed satisfactorily, their internal consistency
was examined. Indeed, the acoefcients ranged from .84 to .91 for EA and from .70 to
.78 for TA.
Affect and vigorous exercise 55

The rst set of analyses examined changes in affect from pre-exercise to post-09, post-
109, and post-209. A positive effect was found, evidenced by both EA/TA and FS/FAS. A
repeated-measures MANOVA (pre, post-09, post-109, post-209) on EA and TA showed a
signicant main effect of Time, Wilks’ l5.167,F(6,23)519.169,p< .001, attributable
to both EA (F(3,84)515.134,p< .001) and TA (F(3,84)521.808,p< .001). The results
of the multiple comparisons for EA and TA are shown in Table 1. Immediately after the
exercise bout, there was an increase in EA compared to baseline, followed by decreases
in TA at post-109and post-209(see Fig. 2).
56 Eric E. Hall et al.
Table 1. Descriptive statistics (means, SD) and results of statistical comparisons (Fisher-Hayter
tests, effect sizes) of EA and TA scores at pre- and post-exercise time points
M6SD Post-09Post-109Post-209
EA Pre 26.41 65.61 1.13* * 0.44 0.08
Post-0932.41 64.95 20.73 21.10* *
Post-10928.76 65.03 20.38
Post-20926.83 65.20
TA Pre 21.41 63.81 0.38 20.81* 21.02*
Post-0922.86 63.77 21.19* * 21.44* *
Post-10918.31 63.87 20.14
Post-20917.79 63.27
*: p<.05; * *: p<.01 (based on Fisher-Hayter tests).
Figure 2. Responses to the EA and TA scales of the AD ACL plotted in circumplex space. The scales
were rotated 458for plotting using trigonometric procedures

Similarly, a repeated-measures MANOVA on FS and FAS showed a signicant main
effect of Time, Wilks’ l5.203,F(6,23)515.030,p< .001, attributable to both the FS
(F(3,84)516.174,p< .001) and the FAS (F(3,84)510.409,p< .001). The results of the
multiple comparisons for FS and FAS are shown in Table 2. Compared to baseline, there
was a signicant improvement in affective valence (i.e. FS) across all post-exercise time
points. Perceived activation (i.e. FAS) showed a signicant decrease at post-109and post-
209compared to baseline and post-09.
The second set of analyses examined the affective responses during exercise. A
repeated-measures MANOVA on FS and FAS showed a signicant main effect of Time,
Wilks’ l5.120,F(8,21)519.327,p< .001, attributable to both the FS (F(4,112)5
57.052,p< .001) and the FAS (F(4,112)530.203,p< .001). The results of the multiple
comparisons are shown in Table 3. Affective valence, as indexed by FS, showed a
marked decline from the second minute after the VT and until exhaustion. Immediately
after exercise was terminated and the cool-down began, however, there was an instan-
taneous improvement (see Fig. 3). Perceived activation, as indexed by the FAS, increased
throughout the exercise bout and decreased during the cool-down (see Fig. 3).
Discussion
The purpose of the present study was to examine the immediate affective responses
to increasing levels of exercise intensity. Three conceptual and methodological inno-
vations were incorporated into this experiment.
First, affect was examined from a dimensional perspective as opposed to the
commonly employed categorical approaches that focus on a few, distinct affective
variables. The broad and balanced scope afforded by the circumplex model was
expected to enable the identication of any salient affective changes, regardless of
direction. Consistent with this expectation, mapping the affective responses on the
two-dimensional circumplex space revealed a diversity of patterns and dynamic changes
in response to the stages of the exercise protocol. Specically, it was found that, in the
early incremental stages, affective change was characterized primarily by an increase
in activation, with little change in affective valence. After the transition to anaerobic
Affect and vigorous exercise 57
Table 2. Descriptive statistics (means, SD) and results of statistical comparisons (Fisher-Hayter tests,
effect sizes) of FS and FAS scores at pre- and post-exercise time points
M6SD Post-09Post-109Post-209
FS Pre 2.07 61.58 0.63* * 0.81* * 0.96* *
Post-093.00 61.34 0.20 0.32
Post-1093.31 61.49 0.07
Post-2093.41 61.18
FAS Pre 3.00 61.13 0.10 20.46* * 20.83* *
Post-093.10 61.11 20.55* * 20.94* *
Post-1092.48 61.15 20.33
Post-2092.14 60.92
*: p<.05; * *: p<.01 (based on Fisher-Hayter tests).

metabolism, however, the continued increase in perceived activation was coupled with
a substantial shift toward affective negativity, leading to an activated unpleasant state
presumably indicative of effort-related tension. Once the strenuous exercise was ter-
minated and the cool-down began, there was an instantaneous drop in activation in
conjunction with a marked improvement in valence, leading, in the course of 1 min, to a
deactivated pleasant state presumably associated with calmness and relaxation. Further-
more, when the pre-exercise state was compared to the post-exercise state using AD
ACL, a transient increase in energetic arousal was found, followed by a drop in tension
and an increase in calmness. Tapping into this phenomenological diversity would have
arguably been impossible had affect been measured through ‘categorical’ instruments
(i.e. by assessing only a few, distinct affective variables).
Second, affective responses were assessed repeatedly during and after the exercise
bout, as opposed to assessment protocols that only examine changes from pre- to post-
exercise. In previous research examining the effects of graded exercise protocols on a
variety of affective variables from pre- to post-exercise, the results have been perplexing
in their inconsistency, running the gamut of possible outcomes (Goldfarb, Hateld,
Sforzo, & Flynn, 1987; Hateld, Goldfarb, Sforzo, & Flynn, 1987; Koltyn, Lynch, & Hill,
1998; Morgan, Horstman, Cymerman, & Stokes, 1980; O’Connor, Petruzzello, Kubitz, &
Robinson, 1995; Pronk, Crouse, & Rohack, 1995; Steptoe, Kearsley, & Walters, 1993). In
the present study, by using single-item scales and, thus, minimizing the burden placed
on the participants, it was possible to track the dynamics of affect throughout the
58 Eric E. Hall et al.
Table 3. Descriptive statistics (means, SD) and results of statistical comparisons (Fisher-Hayter
tests, effect sizes) of FS and FAS scores during exercise and cool-down
M6SD VT VT 12 End Cool 2
FS Min 1 2.24 61.43
Min 2 2.34 61.40 20.27 20.67* * 22.08* * 0.38*
VT-1 2.14 61.55
VT 1.93 61.62 20.42* * 21.77* * 0.61* *
VT 11 1.79 61.90
VT 12 1.14 62.12 21.22* * 0.95* *
End-1 0.07 61.94
End 21.48 62.18 2.32* *
Cool 1 2.76 61.53
Cool 2 2.90 61.54
FAS Min 1 2.83 61.04
Min 2 2.97 60.94 1.03** 1.53* * 1.81* * 0.13
VT-1 3.69 60.60
VT 3.79 60.62 0.72* * 1.15* * 20.83* *
VT 11 4.00 60.65
VT 12 4.31 60.81 0.53* * 21.32* *
End-1 4.65 60.97
End 4.83 61.10 21.63* *
Cool 1 3.59 61.09
Cool 2 3.10 61.01
*: p<.05; * *: p<.01 (based on Fisher-Hayter tests).

exercise bout and recovery with sufcient temporal resolution to ensure that no mean-
ingful changes could escape detection. A comparison of the patterns that emerged
from pre-to-post versus repeated assessments (see Figs 2 and 3) shows that taking
into consideration only the pre- and post-exercise time points would have led to a
misrepresentation of the impact of exercise on affect. It is also apparent that affect
continued to change during the cool-down and recovery. Therefore, the timing of
affect assessments can substantially inuence the results from studies of the exercise–
affect relationship and this factor should, therefore, be considered in future research.
Third, the intensity of exercise was standardized in terms of phases considered to be
metabolically equivalent across individuals, such as the beginning of exercise, the VT,
and VO2max . This was based on the premise that affective responses to acute exercise,
particularly under conditions of vigorous effort, are tied to adaptationally important
metabolic events. The results of the present study revealed a clear dose-response
pattern, thus providing indications of a causal link between exercise-induced homeo-
static perturbations and affective responses. The decline in affective valence 2 minutes
after the occurrence of the VT suggests that the transition to anaerobic metabolism
represents an inherently negatively-charged stimulus. It is possible that this response
reects an evolutionarily primitive mechanism that has remained functional in con-
temporary humans due to its adaptational value. Much like pain, this negative affective
reaction has a strong impact on consciousness and reinstates survival as the top priority;
if the heavy demand on metabolic resources continues, exhaustion, and possibly death,
are inevitable.
Although words like ‘pain’ and ‘struggle’ are often used by sedentary people to
Affect and vigorous exercise 59
Figure 3. Responses to FS and FAS across the stages of exercise, cool-down, and recovery plotted
in circumplex space

describe the experience of vigorous exercise and the popular media abound with
messages about products that promise weight loss and health benets without having
to ‘suffer through exercise’, the possibility of acute exercise having a negative affective
impact has yet to be widely acknowledged in the exercise science literature. When
negative effects are identied, they are typically dismissed as being transient and,
instead, emphasis is placed on the fact that they are succeeded by longer-lasting and
more robust positive changes. For example, Morgan and Ellickson (1989) discounted
the ndings of increased state anxiety scores during and immediately after vigorous
exercise, characterizing the response as reecting ‘eustress rather than stress’ (p. 172).
This is because the increase is followed by ‘a sudden decrease in state anxiety during
the post-exercise recovery period’ (p. 172). Likewise, Partt and Eston (1995b), based
on ndings of decreasing FS scores with increasing exercise intensities among low-
active individuals (Par tt & Eston, 1995a; Partt et al., 1994, 1996), recommended that
these individuals simply ‘be encouraged to focus upon how they feel after exercise has
ended in order to facilitate them interpreting their physiological cues more positively’
(p. 876). However, it should be emphasized that there is presently no evidence to
support the notion that the positive post-exercise effect can outweigh any negative
during-exercise effects in shaping motivational tendencies. To ensure that the exercise
stimulus will not be paired to negative affect in memory, perhaps a more reasonable
choice would be to avoid any negative effects altogether.
As noted earlier, the transition to anaerobic metabolism presents the exerciser with
a challenge, as a multitude of salient interoceptive cues charged with negative affect
enter consciousness. This, however, does not necessarily mean that all individuals are
likely to respond with negative affect. This is because individual differences in prefer-
ence for (Berger, 1994; Morgan, 1997b) or tolerance of (Mogil, 1999) intense stimulation
or relevant cognitive variables such as self-efcacy (McAuley, Talbot, & Martinez, 1999)
may play a role in modulating affective experiences. In the present study, it was only 2
min after the occurrence of the VT that average self-ratings of affective valence exhibited
a signicant decline compared to the beginning of exercise. This generalized decline
may reect what Edwards (1983) called the ‘straw that breaks the camel’s back’ (p. 21).
According to Edwards, there is a point during exercise where vertically organized (i.e.
across levels of organizations, from the muscle cell to the brain) protective mechanisms
cause a failure in neuronal or muscular excitation in order to avoid the depletion of
energy stores. Edwards speculated that this work rate may trigger a ‘conscious or
unconscious need to cease the bombardment of afferent signals’ (p. 21). It is possible
that beyond this point the intensity of interoceptive cues becomes such that it
overpowers the effects of individual-difference and cognitive variables, leading to
uniformly negative affective responses.
Consistent with previous research (see Ekkekakis & Petruzzello, 1999, for a
review), the results of the present study showed a pronounced and instantaneous
rebound from affective negativity to affective positivity as soon as the vigorous
exercise stimulus was terminated. As discussed in the introduction, this pattern is
consistent with the predictions of the opponent-process theory of affect (Solomon
1980, 1991; Solomon & Corbit, 1974). Although the substrates of this phenomenon
remain elusive, it is theoretically and practically interesting and warrants further
research attention. For example, it is unclear at this point whether, as Solomon
(1991) suggested, the post-exercise improvement will only occur if the during-
exercise affective deterioration exceeds a certain threshold. It is also unknown
whether there is a relationship between the a- and b-processes, such that a more
60 Eric E. Hall et al.

pronounced negative affective response during exercise will be followed by a
stronger or longer-lasting improvement post-exercise.
An overarching question is whether vigorous exercise can indeed produce positive
affective changes, as is commonly believed. The data from the present study showed
that, on average and within the range of intensities examined, this was not the case.
Only a few individuals showed improvements in affective valence during exercise.
Moreover, these changes were transient and inconsistent, and were limited to intensities
below and up to a minute beyond the VT. The improvements in affective valence that
were found, although large and shared by all participants, only occurred once the
vigorous exercise stimulus was terminated. Therefore, it is more accurate to state that
this positive response accompanied the cessation of exercise and the beginning of
recovery rather than the vigorous exercise itself. It is interesting to contrast these
ndings to those of a recent study examining affective responses to 10–15 min bouts of
walking at intensities much lower than those used in the present study (i.e. 14%–22% of
maximal age-predicted heart rate reserve, approximately equivalent to 14% – 22%
VO2max ). In this study, Ekkekakis, Hall, VanLanduyt, and Petruzzello (2000) found
signicant, albeit short-lived, improvements in affective valence both during and after
the activity. This suggests that ‘affective benecence’ may be achieved with physical
activities performed at lower intensities and for shorter durations than originally
thought.
Despite the effort that was made in the present study to improve the standardization
of exercise intensity across individuals by taking into account the balance between
aerobic and anaerobic metabolism, it would be imprudent to attempt any generaliza-
tion of the  ndings to other populations without further study. The fact that only
young, healthy, and mostly physically active individuals participated in the present study
raises the possibility that part of the ndings were inuenced by these characteristics. It
is reasonable to assume, however, that if older, less  t, or medically vulnerable popu-
lations exhibit a different pattern of responses, the differences would probably be on
the side of accentuated negativity rather than positivity. It is important that, even in the
present sample, exercise intensity that required substantial anaerobic effort (i.e.
exceeded the VT) led to a signicant decline in affective valence during exercise.
This nding has signi cant practical implications. Exercise practitioners should
consider these ndings in conjunction with the following two points. First, in sedentary
adults and elderly individuals, the lactate and ventilatory thresholds can occur at 50%
VO2max or less. Given that these populations are typically characterized by a low
absolute level of tness, this means that exceeding these thresholds could be brought
about by activities generally considered as mild. Second, exercise performed at
intensities above the lactate and ventilatory thresholds has not been shown to confer
any additional  tness benets compared with exercise performed at or slightly below
these thresholds in previously untrained individuals (Belman & Gaesser, 1991; Weltman
et al., 1992). Therefore, considering the potential for a negative impact on affect and
long-term adherence that exercise intensity above these thresholds entails, it may be
benecial to emphasize careful self-regulation, particularly to novice exercisers. Inter-
estingly, research has shown that the lactate and ventilatory thresholds correspond to
stable ratings of perceived exertion that are unaffected by gender, training, or exercise
modality (DeMello, Cureton, Boineau, & Singh, 1987; Hetzler et al., 1991; Hill, Cureton,
Grisham, & Collins, 1987; Purvis & Cureton, 1981). Similarly, Acevedo et al. (1998)
reported that a running velocity only 10% higher than the onset of blood lactate
accumulation (i.e. 4 mmol l21) led to a signicant decline in affective valence. This
Affect and vigorous exercise 61

nding, which is consistent with the ndings of the present study, suggests that, in
addition to perceptual cues associated with perceived exertion, affective cues, such as
a decline in affective valence, could be used to aid exercisers in recognizing the tran-
sition to anaerobic metabolism.
In conclusion, the nding that ‘vigorous’ exercise has a transient but signicant
negative impact on affect might shed some light on the prevalence of inactivity and the
high dropout rates from exercise programmes. Although negative affective responses
are but one possible explanation, they might prove to be an important one. Thus,
exercise practitioners should be sensitized to the importance of prudent and conserva-
tive selection of workloads and teaching novice exercisers to self-monitor and self-
regulate the intensity of their efforts.
References
Acevedo, E. O., Kraemer, R. R., Haltom, R. W., Tryniecki, J. L., Powell, K., & Gaydos, P. (1998).
Perceptual responses at running velocities proximal to the onset of blood lactate accumula-
tion. Medicine and Science in Sports and Exercise,30, S6.
Acevedo, E. O., Rinehardt, K. F., & Kraemer, R. R. (1994). Perceived exertion and affect at varying
intensities of running. Research Quarterly for Exercise and Sport,65, 372– 376.
Ahmaidi, S., Hardy, J. M., Varray, A., Collomp, K., Mercier, J., & Pre´faut, C. (1993). Respiratory gas
exchange indices used to detect the blood lactate accumulation threshold during an incre-
mental exercise test in young athletes. European Journal of Applied Physiology,66, 31–36.
Belman, M. J., & Gaesser, G. A. (1991). Exercise training below and above the lactate threshold
in the elderly. Medicine and Science in Sports and Exercise,23, 562 –568.
Berger, B. G. (1994). Coping with stress: The effectiveness of exercise and other techniques.
Quest,46, 100–119.
Berger, B. G., & Motl, R. W. (2000). Exercise and mood: A selective review and synthesis of
research employing the Prole of Mood States. Journal of Applied Sport Psychology,12,
69– 92.
Blumenthal, J. A., Emery, C. F., & Rejeski, W. J. (1988). The effects of exercise training on
psychosocial functioning after myocardial infarction. Journal of Cardiopulmonary Rehabi-
litation,8, 183– 193.
Borg, G. (1998). Borg’s perceived exertion and pain scales. Champaign, IL: Human Kinetics.
Brewer, B. W., Manos, T. M., McDevitt, A. V., Cornelius, A. E., & Van Raalte, J. L. (2000). The effect
of adding lower intensity work on perceived aversiveness of exercise. Journal of Sport and
Exercise Psychology,22, 119– 130.
Byrne, A., & Byrne, D. G. (1993). The effect of exercise on depression, anxiety, and other mood
states: A review. Journal of Psychosomatic Research,37, 565– 574.
Cabanac, M. (1995). What is sensation? In R. Wong (Ed.), Biological perspectives on motivated
activities (pp. 399– 418). Norwood, NJ: Ablex.
Caiozzo, V. J., Davis, J. A., Ellis, J. F., Azus, J. L., Vandagriff, R., Prietto, C. A., & McMaster, W. C.
(1982). A comparison of gas exchange indices used to detect the anaerobic threshold. Journal
of Applied Physiology,53, 1184– 1189.
Damasio, A. R. (1995). Toward a neurobiology of emotion and feeling: Operational concepts
and hypotheses. Neuroscientist,1, 19 –25.
Davis, J. A., Frank, M. H., Whipp, B. J., & Wasserman, K. (1979). Anaerobic threshold altera-
tions caused by endurance training in middle-aged men. Journal of Applied Physiology,46,
1039– 1046.
DeMello, J. J., Cureton, K. J., Boineau, R. E., & Singh, M. M. (1987). Ratings of perceived exertion
at the lactate threshold in trained and untrained men and women. Medicine and Science in
Sports and Exercise,19, 354– 362.
62 Eric E. Hall et al.

Dishman, R. K. (1986). Mental health. In V. Seefeldt (Ed.), Physical activity and well-being
(pp. 303–341). Reston, VA: AAHPERD.
Edwards, R. H. T. (1983). Biochemical bases of fatigue in exercise performance: Catastrophe
theory of muscular fatigue. In H. G. Knuttgen, J. A. Vogel, & J. Poortmans (Eds.), Biochemistry
of exercise (pp. 3–28). Champaign, IL: Human Kinetics.
Ekkekakis, P., Hall, E. E., & Petruzzello, S. J. (1999). Measuring state anxiety in the context of
acute exercise using the State Anxiety Inventory: An attempt to resolve the brouhaha.
Journal of Sport and Exercise Psychology,21, 205– 229.
Ekkekakis, P., Hall, E. E., VanLanduyt, L. M., & Petruzzello, S. J. (2000). Walking in (affective)
circles: Can short walks enhance affect? Journal of Behavioral Medicine,23, 245– 275.
Ekkekakis, P., & Petruzzello, S. J. (1999). Acute aerobic exercise and affect: Current status,
problems, and prospects regarding dose-response. Sports Medicine,28, 337–374.
Emmons, R. A., & Diener, E. (1986). A goal-affect analysis of everyday situational choices. Journal
of Research in Personality,20, 309– 326.
Epstein, L. H., Koeske, R., & Wing, R. R. (1984). Adherence to exercise in obese children. Journal
of Cardiac Rehabilitation,4, 185–195.
Gaesser, G. A., & Poole, D. C. (1996). The slow component of oxygen uptake kinetics in humans.
Exercise and Sport Sciences Reviews,24, 35 – 70.
Gauvin, L., & Brawley, L. R. (1993). Alternative psychological models and methodologies
for the study of exercise and affect. In P. Seraganian (Ed.), Exercise psychology: The
inuence of physical exercise on psychological processes (pp. 146–171). New York: John
Wiley.
Goldfarb, A. H., Hateld, B. D., Sforzo, G. A., & Flynn, M. G. (1987). Serum beta-endorphin levels
during a graded exercise test to exhaustion. Medicine and Science in Sports and Exercise,19,
78– 82.
Hardy, C. J., & Rejeski, W. J. (1989). Not what, but how one feels: The measurement of affect
during exercise. Journal of Sport and Exercise Psychology,11, 304–317.
Hateld, B. D. (1991). Exercise and mental health: The mechanisms of exercise-induced
psychological states. In L. Diamant (Ed.), Psychology of sports, exercise, and tness: Social
and personal issues (pp. 17–49). New York: Hemisphere.
Hateld, B. D., Goldfarb, A. H., Sforzo, G. A., & Flynn, M. G. (1987). Serum beta-endorphin and
affective responses to graded exercise in young and elderly men. Journal of Gerontology,42,
429– 431.
Hays, K. F. (1999). Working it out: Using exercise in psychotherapy. Washington, DC: American
Psychological Association.
Hetzler, R. K., Seip, R. L., Boutcher, S. H., Pierce, E., Snead, D., & Weltman, A. (1991). Effect of
exercise modality on ratings of perceived exertion at various lactate concentrations. Medicine
and Science in Sports and Exercise,23, 88 –92.
Hill, D. W., Cureton, K. J., Grisham, S. C., & Collins, M. A. (1987). Effect of training on the rating
of perceived exertion at the ventilator y threshold. Europe an Journal of Applied Physiology,
56, 206– 211.
Jones, R. H., & Molitoris , B. A. (1984). A statistical method for determining the breakpoint of two
lines. Analytical Biochemistry,141, 287– 290.
Katch, V., Weltman, A., Sady, S., & Freedson, P. (1978). Validity of the relative percent concept
for equating training intensity. European Journal of Applied Physiology,39, 219– 227.
Kerr, J. H., & Vlaswinkel, E. H. (1993). Self-reported mood and running under natural conditions.
Work and Stress,7, 161–177.
Kerr, J. H., & van den Wollenberg, A. (1997). High and low intensity exercise and psychological
mood states. Psychology and Health,12, 603 – 618.
King, A. C., Haskell, W. L., Taylor, C. B., Kraemer, H. C., & DeBusk, R. F. (1991). Group- vs home-
based exercise training in healthy older men and women: A community-based clinical trial.
Journal of the American Medical Association,266, 1535–1542.
King, A. C., Haskell, W. L., Young, D. R., Oka, R. K., & Stefanick, M. L. (1995). Long-term effects
Affect and vigorous exercise 63

of varying intensities and formats of physical activity on participation rates,  tness, and
lipoproteins in men and women aged 50 to 65 years. Circulation,91, 2596 –2604.
King, A. C., Taylor, C. B., & Haskell, W. L. (1993). Effects of differing intensities and formats of 12
months o f exerc ise trainin g on psychological ou tcomes in older adu lts. Health Psychology,12,
292–300.
Kirkcaldy, B. C., & Shephard, R. J. (1990). Therapeutic implications of exercise. International
Journal of Sport Psychology,21, 165– 184.
Koltyn, K. F., Lynch, N. A., & Hill, D. W. (1998). Psychological responses to brief exhaustive
cycling exercise in the morning and the evening. International Journal of Sport Psychology,
29, 145– 156.
La Forge, R. (1995). Exercise-associated mood alterations: A review of interactive neurobiologic
mechanisms. Medicine, Exercise, Nutrition, and Health,4, 17– 32.
Lang, P. J. (1980). Behavioral treatment and bio-behavioral assessment: Computer applications.
In J. B. Sodowski, J. H. Johnson, & T. A. Williams (Eds.), Technology in mental health care
delivery systems (pp. 119–137). Norwood, NJ: Ablex.
Larsen, R. J., & Diener, E. (1992). Promises and problems with the circumplex model of emotion.
In M. S. Clark (Ed.), Review of personality and social psychology (vol. 13, pp. 25–59).
Newbury Park, CA: Sage.
Lee, J. Y., Jensen, B. E., Oberman, A., Fletcher, G. F., Fletcher, B. J., & Raczynski, J. M. (1996).
Adherence in the Training Levels Comparison trial. Medicine and Science in Sports and
Exercise,28, 47–52.
McAuley, E., & Rudolph, D. (1995). Physical activity, aging, and psychological well-being. Journal
of Aging and Physical Activity,3, 67– 96.
McAuley, E., Talbot, H. M., & Martinez, S. (1999). Manipulating self-efcacy in the exercise
environment in women: Inuences on affective responses. Health Psychology,18, 288– 294.
Meyer, T., Gabriel, H. H. W., & Kindermann, W. (1999). Is determination of exercise intensities as
percentages of VO2m ax of HRm a x adequate? Medicine and Science in Sports and Exercise,31,
1342– 1345.
Mogil, J. S. (1999). The genetic mediation of individual differences in sensitivity to pain and its
inhibition. Proceedings of the National Academy of Sciences,96, 7744 –7751.
Morgan, W. P. (1985). Affective benecence of vigorous physical activity. Medicine and Science
in Sports and Exercise,17, 94– 100.
Morgan, W. P. (Ed.) (1997a). Physical activity and mental health. Washington, DC: Taylor and
Francis.
Morgan, W. P., & Ellickson, K. A. (1989). Health, anxiety, and physical exercise. In D. Hackfort &
C. D. Spielberger (Eds.), Anxiety in sports: An international perspective (pp. 165–182). New
York: Hemisphere.
Morgan, W. P., Horstman, D. H., Cymerman, A., & Stokes, J. (1980). Exercise as a relaxation
technique. Primary Cardiology,6, 48–57.
Moses, J., Steptoe, A., Mathews, A., & Edwards, S. (1989). The effects of exercise training on
mental well-being in the normal population: A controlled trial. Journal of Psychosomatic
Research,33, 47–61.
Mutrie, N., & Biddle, S. J. H. (1995). The effects of exercise on mental health in nonclinical
populations. In S. J. H. Biddle (Ed.), European perspectives on exercise and sport psychology
(pp. 50– 70). Champaign, IL: Human Kinetics.
National Institutes of Health Concensus Development Panel on Physical Activity and Cardio-
vascular Health (1996). Physical activity and cardiovascular health. Journal of the American
Medical Association,276, 241–246.
Nesse, R. M. (1990). Evolutionary explanations of emotions. Human Nature,1, 261– 289.
Nesse, R. (1998). Emotional disorders in evolutionary perspective. British Journal of Medical
Psychology,71, 397–415.
O’Connor, P. J., Petruzzello, S. J., Kubitz, K. A., & Robinson, T. L. (1995). Anxiety responses to
maximal exercise testing. British Journal of Sports Medicine,29, 97–102.
64 Eric E. Hall et al.

Ojanen, M. (1994). Can the true effects of exercise on psychological variables be separated from
placebo effects? International Journal of Sport Psychology,25, 63–80.
Panksepp, J. (1998a). Affective neuroscience: The foundations of human and animal emotions.
New York: Oxford University Press.
Panksepp, J. (1998b). The periconscious substrates of consciousness: Affective states and the
evolutionary origins of the self. Journal of Consciousness Studies,5, 566–582.
Partt, G., & Eston, R. (1995a). Changes in ratings of perceived exertion and psychological affect
in the early stages of exercise. Perceptual and Motor Skills,80, 259 –266.
Partt, G., & Eston, R. (1995b). How to help low-active women feel better about exercising. In
R. Vanfraechem-Raway & Y. Vanden Auweele (Eds.), Proceedings of the IXth European
Congress on Sport Psychology (pp. 877– 882). Brussels: European Federation of Sports
Psychology and Belgian Federation of Sport Psychology.
Partt, G., Eston, R., & Connolly, D. (1996). Psychological affect at different ratings of perceived
exertion in high- and low-active women: A study using a production protocol. Perceptual and
Motor Skills,82, 1035 –1042.
Partt, G., Markland, D., & Holmes, C. (1994). Responses to physical exertion in active and
inactive males and females. Journal of Sport and Exercise Psycholog y,16, 178–186.
Pate, R. R., Pratt, M., Blair, S. N., Haskell, W. L., Macera, C. A., Bouchard, C., Buchner, D., Ettinger,
W., Heath, G. W., King, A. C., Kriska, A., Leon, A. S., Marcus, B. H., Morris, J., Paffenbarger, R. S.
Jr, Patrick, K., Pollock, M. L., Rippe, J. M., Sallis, J., and Wilmore, J. H. (1995). Physical activity
and public health: A recommendation from the Centers for Disease Control and Prevention
and the American College of Sports Medicine. Journal of the American Medical Association,
273, 402– 407.
Petruzzello, S. J., Landers, D. M., Hateld, B. D., Kubitz, K. A., & Salazar, W. (1991). A meta-analysis
on the anxiety-reducing effects of acute and chronic exercise: Outcomes and mechanisms.
Sports Medicine,11, 143 – 182.
Pronk, N. P., Crouse, S. F., & Rohack, J. J. (1995). Maximal exercise and acute mood response
in women. Physiology and Behavior,57, 1–4.
Purvis, J. W., & Cureton, K. J. (1981). Ratings of perceived exertion at the anaerobic threshold.
Ergonomics,24, 295 – 300.
Raglin, J. S., & Morgan, W. P. (1985). Inuence of vigorous exercise on mood state. Behavior
Therapist,8, 179–183.
Russell, J. A. (1978). Evidence of convergent validity on the dimensions of affect. Journal of
Personality and Social Psychology,36, 1152–1168.
Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology,
39, 1161–1178.
Russell, J. A. (1989). Measures of emotion. In R. Plutchik & H. Kellerman (Eds.), Emotion: Theory,
research, and experience (vol 4: The m easurement of emotions; pp. 83 –111). San Diego, CA:
Academic Press.
Russell, J. A. (1997). How shall an emotion be called? In R. Plutchik & H. R. Conte (Eds.),
Circumplex models of personality and emotions (pp. 205– 22). Washington, DC: American
Psychological Association.
Russell, J. A., Weiss, A., & Mendelsohn, G. A. (1989). Affect Grid: A single item scale of pleasure
and arousal. Journal of Personality and Social Psychology,57, 493– 502.
Sallis, J. F., Haskell, W. L., Fortmann, S. P., Vranizan, K. M., Taylor, C. B., & Solomon, D. S. (1986).
Predictors of adoption and maintenance of physical activity in a community sample.
Preventive Medicine,15, 331–341.
Schulze, G. (1995). Motivation: Homeostatic mechanisms may instigate and shape adaptive
behaviors through the generation of hedonic states. In R. Wong (Ed.), Biological perspectives
on motivated activities (pp. 265 – 288). Norwood, NJ: Ablex.
Solomon, R. L. (1980). The opponent-process theory of acquired motivation: The costs of pleasure
and the bene ts of pain. American Psychologist,35, 691–712.
Affect and vigorous exercise 65

Solomon, R. L. (1991). Acquired motivation and affective opponent-processes. In J. Madden (Ed.),
Neurobiology of learning, emotion, and affect (pp. 307– 347). New York: Raven.
Solomon, R. L., & Corbit, J. D. (1974). An opponent-process theory of motivation: I. Temporal
dynamics of affect. Psychological Review,18, 119– 145.
Steptoe, A. (1992). Physical activity and psychological well-being. In N. G. Norgan (Ed.), Physical
activity and health (pp. 207–229). Cambridge: Cambridge University Press.
Steptoe, A., & Bolton, J. (1988). The short-term inuence of high and low intensity physical
exercise on mood. Psychology and Health,2, 91–106.
Steptoe, A., Kearsley, N., & Walters, N. (1993). Acute mood responses to maximal and submaximal
exercise in active and inactive men. Psychology and Health,8, 89 – 99.
Svebak, S., & Murgatroyd, S. (1985). Metamotivational dominance: A multimethod validation
of reversal theory constructs. Journal of Personality and Social Psychology,48, 107– 116.
Tate, A. K., & Petruzzello, S. J. (1995). Varying the intensity of acute exercise: Implications
for changes in affect. Journal of Sports Medicine and Physical Fitness,35, 295 –302.
Thayer, R. E. (1978). Factor analytic and reliability studies on the Activation-Deactivation Adjective
Check List. Psychological Reports,42, 747– 756.
Thayer, R. E. (1986). Activation-Deactivation Adjective Check List: Current overview and
structural analysis. Psychological Reports,58, 607–614.
Thayer, R. E. (1989). The biopsychology of mood and arousal. New York: Oxford University
Press.
Tuson, K. M., & Sinyor, D. (1993). On the affective benets of acute aerobic exercise: Taking
stock after twenty years of research. In P. Seraganian (Ed.), Exer cise psychology: The inuence
of physical exercise on psychological processes ( pp. 80– 121). New York: John Wiley.
United States Department of Health and Human Services (1996). Physical activity and health: A
report of the Surgeon General. US Department of Health and Human Services, Centers for
Disease Control and Prevention, National Center for Chronic Disease Prevention and Health
Promotion, Atlanta, GA.
United States National Center for Health Statistics (1999). Healthy People 2000 review,
1998– 1999. Hyattsville , MD: Public Health Service.
Weltman, A., Seip, R. L., Snead, D., Weltman, J. Y., Haskvitz, E. M., Evans, W. S., Veldhuis, J. D., &
Rogol, A. D. (1992). Exercise training at and above the lactate threshold in previously
untrained women. International Journal of Sports Medicine,13, 257– 263.
Yik, M. S. M., Russell, J. A., & Feldman-Barrett, L. (1999). Structure of self-reported current affect:
Integration and beyond. Journal of Personality and Social Psychology,77, 600–619.
Received 9 March 2000; revised version received 22 September 2000
66 Eric E. Hall et al.