Abstract
The primary aim of this study was to examine the effects of exercise intensity and duration on choice reaction time with a secondary aim to examine the influence of aerobic fitness on choice reaction time. Fifteen college-aged men enrolled in a physical fitness activity course participated. We observed a cubic trend (r =.99) across exercise intensity that ranged from unloaded cycling to 90% of physical working capacity (PWC). However, no significant differences were noted across intensity (medium effect size) or duration (small effect size). The fastest choice reaction time occurred at 75% of PWC, which was near the ventilatory threshold (72.6% of VO2max). Aerobic fitness was average for college-aged men (39.3 ml/kg/min.) and was poorly correlated with choice reaction time across exercise intensity. Further research is needed to compare choice reaction time in high to low fitness groups across exercise intensity.
Yerkes and Dodson (1908) were the first to propose an inverted-U shaped hypothesis, which states that for any given performance variable an optimal level of arousal or activation exists (Yerkes-Dodson law). They proposed the hypothesis following laboratory experiments on mice. Basically, over-activation or under-activation of physiological stress may lead to poorer performance results such as a prolonged reaction time to a stimulus. Cognitive tasks are also influenced by physical exercise (Gutin, 1973). Researchers examining exercise and a response to cognitive tasks (such as reaction time) have demonstrated an inverted-U hypothesis phenomenon during exercise (Davey, 1973; Gupta, et al., 1974; Reilly & Smith, 1986; Salmela & Ndoye, 1986). Generally, moderate levels of arousal are considered most conducive to top level performance in sport (Cratty, 1973; Oxendine, 1984). However, the inverted-U hypothesis has some limitations. For example, the hypothesis could not describe the mutually incompatible states in high or low arousal, such as anxiety and excitement or relaxation and boredom (Apter, 1982). In spite of the inverted-U relationship, some investigators have found a facilitative effect of exercise on certain aspects of cognition (Tomporowski, 2003; Kashihara & Nakahara, 2005; Davranche, Audiffren, & Denjean, 2006), while others maintained a negative relationship (Gutin & Digennaro, 1968a; Malomske & Szmodis, 1970; Weingarten, 1973; Hancock & McNaughton, 1986; Fleury & Bard, 1987). Also, some researchers have reported mixed results, with both beneficial and negative effects or no effect on cognition (McMorris & Graydon, 2000; Collardeau, Brisswalter, & Audiffren, 2001; Lemmink & Visscher, 2005). In an excellent review, Tomporowski (2003) noted that information processing (including reaction time response) is most often enhanced by submaximal exercise performance up to about one hour of effort. Longer exercise bouts may adversely affect information processing and memory function, with dehydration a likely culprit.
Several important methodological factors might partly explain diverging reaction time results, including variance in experimental design (Naatanen, 1975). Some investigators have employed a simple reaction time task while others have used choice reaction time (Draper, McMorris, & Parker, 2010). Tomporowski and Ellis (1986) proposed four factors to account for the variance in results including: exercise intensity and duration, physical fitness, the psychological tasks employed, and the time at which the task was administered (i.e., during exercise or following exercise). Therefore, the primary aim of the current study was to examine the effects of exercise intensity and duration on choice reaction time. A secondary aim was to examine the influence of aerobic fitness as indicated by VO2max on choice reaction time.
Method
Participants
Fifteen college men enrolled in physical activity courses volunteered to participate (M age = 24.1 yr., SD = 3.3; M height = 176 cm, SD = 6; M body mass = 78.2 kg, SD = 11.9; M fat = 13.1%, SD = 4.4). The project was approved by the committee for the use of human participants. All participants completed a physical activity questionnaire and a written consent form.
Measures
Anthropometries.—Body mass was measured by a standard Detecto scale and height with a stadiometer. Relative fat was estimated from skinfold thicknesses after Jackson and Pollock (1978), using their three skinfold site (sum of chest, abdomen, and thigh) equation.
Cardiorespiratory fitness.—In order to determine cardiorespiratory fitness level and to ascertain relative workloads during reaction time trials, maximal oxygen uptake (VO2max) and related variables, including physical work capacity (PWC), were obtained. Assessment of VO2max and PWC was determined during leg cycling on a Monark friction resistance ergometer. Initial power output was set at 29.4 W (0.5 kg, 60 rpm) and increased by 29.4 W every 2 min. Pedal rate was held constant at 60 rpm. The participants continuously cycled until volitional termination or a drop in pedal rate > 5 rpm.
Maximal oxygen uptake (VO2 max) was measured indirectly via open circuit spirometry. Briefly, ventilatory volume (VE) was measured with a Pneumoscan S-301 spirometer. The fractional aspect of oxygen and carbon dioxide was measured with a Beckman OM-14 and LB-2 analyzer, respectively. Prior to each test, the analyzers were calibrated with gas samples of known composition. The ventilatory threshold was determined after Wasserman, Hansen, Sue, and Whipp (1987). Briefly, the ventilatory threshold was identified as the breakpoint in the ventilatory equivalent for oxygen uptake (ventilation/oxygen uptake) while the ventilatory equivalent for carbon dioxide production (ventilation/volume of carbon dioxide production) was not increasing. As a result of the initial data collection, the work capacity for each participant was determined. Participants were then given orientation and training on the exercise and reaction time apparatus.
Reaetion time.—The reaction time apparatus consisted of a Lafayette four choice reaction time device that was custom mounted and centered on the handle bars of a cycle ergometer. The modified apparatus allowed measurement of reaction time while the participants exercised at relative workloads.
The procedure for orientation and training consisted of 10 min. of cycling at 60 rpm at a resistance of 1 to 1.5 kg (∼59 to 88 W). Pedal rate was confirmed by use of photoelectric cells. Reaction time training consisted of forty practice trials. The participants were presented an auditory preparatory stimulus, which cued them to place the fingers of their dominant hand on the four buttons of the reaction time task. Following a variable time interval of 1.5, 3, or 4.5 sec., the obligatory stimulus was presented. The participants were taught to press the button that corresponded to the lighted stimulus. After the response, participants returned their hand to the handlebar grip. Reaction time was noted as the time to release the finger from the lighted stimulus button. A stimulus was presented every 15 sec. Familiarization was judged to have been complete when the participants could maintain a consistent 60 rpm pedal rate while smoothly moving the hand to and from the handle bar while performing the reaction time task. All participants demonstrated familiarity with the apparatus as a result of the 10-min. training session.
A counter-balanced design enabled the participants to perform under five intensity and five duration conditions. The intensity conditions consisted of the following: unloading cycling, 25, 50, 75, and 90% of PWC, respectively. Relative workloads were determined from the PWC test. The ergometer resistance was adjusted individually for each participant under each intensity condition.
In order to determine duration effect, reaction time data were collected four times per minute throughout the exercise bout. Each bout lasted 5 min. Reaction time for each minute was calculated by determining the mean of the four reaction time trials. To prevent a finger preference effect, each of the four fingers was represented by random order in each of the five trials.
At the beginning of each of the five data collection sessions, the participants were allowed 2 min. of physiological and psychological warm-up. Participants received 10 reaction time trials while cycling against mild resistance (approximately 20 watts). Following warm up, the participants began exercising at the prescribed level as determined by the counter-balanced design. A protocol data sheet was followed so that all participants received the same reaction time treatment conditions. To prevent the influence of fatigue, 24 hr. separated each work bout. Also, the participants performed each bout at the same time of the day to minimize circadian rhythm effect.
Statistical Analysis
A one-way analysis of variance (ANOVA) was used to examine reaction time across exercise intensity and a two-way ANOVA was employed to examine exercise intensity by exercise minute. Regression analysis was employed to examine data trends (linear, quadratic, cubic, or exponential) in choice reaction time across exercise intensity. Finally, bivariate correlations were employed to examine relationships among VO2max and reaction time at each exercise intensity. Statistical significance was set at p <.05 level.
Results
Maximal cardiorespiratory function and performance variables are presented in Table 1. Relative fat and VO2max show that the participants were normal weight and at an average cardiorespiratory fitness level. As noted in Table 2, no significant differences were found for reaction time across the five exercise intensities, although significance was approached (p =.08). Moreover, the effect size (Cohen's d) was moderate (0.47).
Maximal Cardiorespiratory Function and Performance
Choice Reaction Time Response
p <0.05.
As noted in Fig. 1, a cubic trend (r =.99) was observed across exercise intensity. As was the case with exercise intensity, no significant differences were noted when exercise minute (Minutes 1 to 5) were compared within and between exercise intensities (small effect size = 0.20).
Reaction time (mean±95% confidence interval) across exercise intensity.
Table 3 shows power output, heart rate, and perceived exertion across the five exercise intensities. Finally, cardiorespiratory fitness as indicated by VO2max (ml/kg/min.) was poorly correlated with reaction time across exercise intensity as correlations ranged from r2 =.02 to.04.
Means and Standard Deviations of Heart Rate and Perceived Exertion
Discussion
Choice Reaction Time and Submaximal Aerobic Exercise
As noted in the results of the current study, a moderate effect size was indicated for choice reaction time during exercise at low to heavy intensity. Also, a significant cubic trend was noted across exercise intensity. In previous research, Draper, et al. (2010) found a significant linear trend (improvement) in choice reaction time as exercise intensity increased from a resting state to severe exercise conditions. Interestingly, the authors did not find the same trend for simple reaction time. The choice reaction time response curve for our study was similar to the work of Brisswalter, Arcelin, Audiffren, and Delignieres (1997), who examined simple reaction time. In their work, high and low fitness groups exercised at workloads ranging from resting to 80% of VO2 max in a counter-balanced design. Although Brisswalter and colleagues did not examine data trends, when plotted their data trends were similar to our results. We suggest that our findings generally follow Easterbrook's cue utilization theory which predicted that high levels of arousal will lead to attention narrowing, defined as a decrease in the range of cues from the stimulus and its environment to which the organism is sensitive (Easterbrook, 1959). Exercising below maximum intensities has yielded a positive effect on cognitive performance, most probably as a result of an enhanced activation of a peripheral component (muscle contraction) and central nervous system integrity (Hogervorst, Riedel, Jeukendrup, & Jolles, 1996). Neural resources most probably increased the demand in some brain regions, which reduced essential metabolic resources (e.g., oxygen and glucose) availability in other brain regions (Dietrich, 2006). In addition, the beneficial effects of exercise might be affected by the debilitating effects of peripheral or central muscular fatigue (Kamijo, Nishihira, Higashiura, & Kuroiwa, 2007). Thus, arousal of the central nervous system and physical fatigue of the skeletal-motor system might have alternative influences on mental functioning.
In regard to high intensity exercise and reaction time, McMorris and Keen (1994) observed a significant (p <.05) prolonged reaction time response at 100% of maximal workload compared to resting or 70% of workload conditions for simple reaction time. The authors suggested that for an undemanding central nervous system task such as simple reaction time, heavy exercise stress adversely affected reaction time, perhaps caused by fatigue. Recently, Audiffren, Tomporowski, and Zagrodnik (2008) found choice reaction time to significantly improve by about 17 msec. during the mid-portion of 35 min. of cycle ergometer exercise at 90% of the ventilatory threshold. The authors speculated that the improved choice reaction time responses may have been due to energized motor outputs. They also suggested that several minutes of exercise may be necessary in order to see improved reaction time responses. Davranche, et al. (2006) found similar results during cycle ergometer exercise at 90% of the ventilatory thresh-old. During the later work stages, choice reaction time significantly improved when compared to rest. In earlier work, Chmura, Nazar, and Kaciuba-Uscilko (1994) observed choice reaction time significantly faster at a moderate workload (M = 227 W) when compared to pre-exercise (resting) conditions. The exercise regimen included 3-min. bouts of increasing intensity with a 1-min. rest period between bouts. The final workload (about 300 W) yielded a significantly prolonged reaction time. Interestingly, the blood lactate response was about 5 mmol for the fastest reaction time response and about 10 mmol for the final exercise bout. Thus, the fastest reaction time response most probably occurred at or near the lactate threshold (not measured). Although not significantly different, in the current study the fastest reaction time response occurred at 75% of PWC. The ventilatory threshold was reached at an average of 72.6% of VO2max, thus similar to the workload for the fastest reaction time. Thus, it appears that optimal choice reaction time responses are found at moderate exercise intensities below or near the lactate or ventilatory threshold.
Several factors have been postulated to account for improvements or decrements of choice reaction time during exercise. For example, increase of oxidative enzyme activity (Gerchman, Edgerton, & Carrow, 1975), the level of nervous system activation (Chmura, et al., 1994) McMorris & Graydon, 2000), blood acidosis or the accumulation of metabolic wastes and/or the involvement of different metabolic systems (Fleury & Bard, 1987), and an increase in core temperature (Ruch, Patton, Woodbury, & Towe, 1966; Audiffren, et al., 2008) have been suggested as factors. In addition, other factors postulated to effect reaction time include a generalized improvement of the whole response time distribution (Daveranche, et al., 2006), modification in humoral functioning (Brisswalter, Durand, Delignieres, & Legros, 1995), a greater efficiency of peripheral motor processes (i.e., better synchronization of the motor units discharge), peripheral sensorial processes (Davranche, Burle, Audiffren, & Hasbroucq, 2005, 2006), and muscle tension during moderate exercise (Tomporowski & Ellis, 1986).
Dual Task and Choice Reaction Time
In the current study, the reaction time response during the unloaded work level was perhaps influenced by maintaining a pedal rate of 60 rpm against limited resistance, thus demanding additional motor processing attention. Gutin (1973) suggested that participants had to inhibit their movement until they processed information to decide which direction to move in a five-choice reaction time task. Compared to simple reaction time, choice reaction time required both temporal and spatial inhibitions. The increases in exercise-induced activation above resting were harmful to performance of tasks that demanded great inhibition (Gutin, 1973). In later work, Brisswalter, et al. (1995) noted that the optimal pedal rate during cycle ergometer exercise for untrained participants was 50 rpm when performing a simple reaction time task. They included trials that varied pedal rate from 30 to 80 rpm and also a freely chosen pedal rate. No significant differences were found when comparing simple reaction time across the pedal rates noted. We acknowledge that the unloaded cycling condition in our paper may have adversely impacted the choice reaction time values. However, the slowest choice reaction time scores were found during the 25% PWC condition.
Influence of Aerobic Fitness on Choice Reaction Time
Aerobic fitness level (VO2max) has been shown by some researchers to positively affect cognitive performance during or following exercise stress (Gutin & DiGennaro, 1968a, 1968b; Sjoberg, 1980; Heckler & Croce, 1992). However, other investigators (Tomporowski, Ellis, & Stephens, 1987; Travlos & Marisi, 1995) found no significant difference between high and low fit individuals on mental performance. Brisswalter, et al. (1997) reported significantly faster simple reaction times in 10 participants with higher aerobic fitness (VO2max, 64.1 ml/kg/min.) when compared to 10 participants with lower values (42.2 ml/kg/min.). Simple reaction time was compared at several exercise intensities ranging from 20 to 80% of VO2max. In the present study, we observed low correlations between VO2max and choice reaction time across exercise intensity. Since the participants in the current study were considered average in regard to VO2 max and the range in values was limited (34.8 to 43.8 ml·kg−1·min.−1), this may have influenced our results. However, Travlos and Marisi (1995) also observed choice reaction time unaffected by aerobic fitness. Additional research is needed to compare choice reaction time in low to high VO2max groups.
Summary
We observed a cubic trend in choice reaction time across exercise intensities that ranged from unloaded cycling to 90% of PWC. Other researchers have found similar responses. Consequently, our results do not support the inverted U-shaped hypothesis. The effect size in comparing choice reaction time across exercise intensity was moderate, with the fastest choice reaction time occurring at 75% of PWC. Furthermore, exercise duration by exercise minute yielded a small effect size. Finally, we observed a very low relationship between VO2max and choice reaction time across exercise intensity.