As evidenced from the weak correlation (r = 0.08, p = 0.64) between the RRV (i.e., motivating value) and liking of physical activity they are independent constructs in normally active children. Research has demonstrated that both RRV [9
] and liking [14
] of physical activity are associated with the amount of physical activity that children engage in. This study extends previous research by including measures of both RRV and liking of physical activity and demonstrating that RRV and liking are independently associated with time in MVPA. The results also demonstrate that children who have both a high RRV and liking of physical activity engage in greater MVPA than other children. Liking of a food or drug is thought to be primarily related to opioid system while the appetite to consume a reinforcer as well as the instrumental performance to obtain the reinforcer is primarily regulated by the dopamine system [13
], though the activity of both systems is interrelated [10
The hedonic and motivational aspects of rewards usually go together, as people usually like what they want [11
]. Liking of physical activity may be more related to the types of physical activity, sports, or games children choose to engage in. Reinforcing value, on the other hand, may be more related to the behavioral effort or how hard a child will work to engage in their most liked physical activities. It makes sense that children are more likely to engage in behaviors that they like and that they are motivated to obtain. Many children may report that they like a physical activity, but do not put forth the effort to engage in that activity.
This identification of unique although overlapping, neural systems suggests that motivation for physical activity may also be influenced by multiple neurobiological systems. Central dopamine release, dopamine D2 receptor (DRD2) number, and synaptic dopamine transporters have been investigated for their role in physical activity motivation given the importance of dopamine for controlling central motivational responses [10
] and motor movement [31
]. DRD2 antagonists reduce total motor activity by up to 41% in healthy adults [34
] and DNA sequence variations in the DRD2 gene are associated with reduced total motor activity levels in women [35
]. Imaging studies in adults have shown that DRD2 [36
] and dopamine transporters [37
] are reduced with age and that age-related reductions in DRD2 availability are correlated with reduced motor function [38
]. However, 30 minutes of running was found to have no effect on their brain dopamine activity [39
]. Wheel running by rats and other small animal models occurs spontaneously and is a reinforcing behavior in that they will perform lever presses to engage in wheel running [40
]. Moreover, wheel running produces a reinforcing aftereffect as measured by conditioned place preference after the exercise has stopped [43
]. In small animal models increased synaptic dopamine concentrations through inhibition of dopamine transporter action increases locomotor activity [44
], whereas DRD2 receptor deficiency or pharmacologic blockade or knockout of DRD2 reduces locomotor activity [31
Research regarding alterations in the opioid system during exercise has also been conducted. Exercise increases peripheral endogenous opioids in humans [45
] and central opioids in animal models [46
]. Exercise increases central beta-endorphin release [48
] and dynorphin mRNA [49
] in small animal models. Although opioid agonists generally increase locomotion in laboratory animals [50
], opioid antagonists have generally had little effect on the amount of locomotion in open-field tests [51
]. In contrast, voluntary wheel running [53
] and conditioned place preference produced by wheel running [54
] in animals that have established the behavior, as well as the initial acquisition of wheel running behavior [55
] are all attenuated by opioid receptor antagonists.
Running and other MVPA-type exercises produce discomfort originating from the active muscles [56
]. One way opioids may be related to the hedonics of physical activity and participation in MVPA is by reducing the pain that typically occurs during exercise [57
]. Peripheral endogenous opioids may also help to maintain blood glucose concentrations [62
] and muscle contractile function during vigorous exercise [65
], which would also help to reduce muscular discomfort. Another way endogenous opioids may influence the hedonic aspects of MVPA is by promoting exercise-induced increases in mood [66
Based on the above findings and dopamine system physiology [68
] some combination of greater DRD2 receptor number, ample synaptic dopamine release, and slower synaptic dopamine reuptake would be expected to promote a higher reinforcing value of MVPA. Characteristics of the central opioid system such as an increased number of opioid mu receptors and greater neuronal release of endogenous peptides including enkephalins, dynorphins and endorphin would be expected to promote a greater hedonic value or sensory pleasure during MVPA [69
]. In contrast, children who have lower central dopamine or opioid signaling may not engage in as much MVPA due to a lower reinforcing value or liking of more vigorous types of physical activity.
Additional research is needed to better understand the role of the central dopamine and opioid systems for engaging in MVPA. Central dopamine and opioid systems have been primarily studied for their roles in eating and drugs of abuse [10
]. Some evidence suggests that there may be differences in the neural mechanisms for the motivation to eat and to be physically active. Studies of children (Temple, Legierski, Giacomelli, Salvy & Epstein, unpublished results) and adults [19
] have found that the RRV of food was more strongly associated with energy intake in ad libitum
eating tasks than liking for food. In contrast, the present study found that both RRV and liking of physical activity were independently associated with physical activity participation and that MVPA was greatest only when the individual had relatively high RRV and liking of physical activity. These results suggest that RRV and liking may have different amounts of influence on MVPA than on eating.
The present results can be generalized to normally active youth in that time in MVPA of the children in the present study is very similar to previous studies that used accelerometers to measure MVPA in 8 to 12 year old children [23
]. Although some studies have reported much lower time in MVPA [73
], this appears to be a function of the method used to define MVPA [70
]. As shown by Guinhouya and colleagues [70
] using the cutoff regression equations of Freedson and colleagues [74
] produces MVPA minutes that agree quite well with the current study while using the cutoff values of Puyau [75
] results in 5-fold lower estimates of MVPA. The current study has the advantage of having each subject walk and jog on a treadmill while collecting oxygen consumption data and wearing the accelerometer that they wore during the week of monitoring to determine individualized 3-MET MVPA cutoffs for each subject. This method produced MVPA data that not only agree with the method of Freedson and colleagues [74
], but also agree with time in MVPA calculated from an independent method, heart-rate monitoring, measured across 26 studies that included 1883 youth [3
However, other aspects of the study limit generalization. All the participants were normal weight, and it is possible that different results would be obtained if overweight children were studied. The subjects' history of physical activity or participation in sports was not assessed. History of physical activity may have influenced the subject liking and/or reinforcing value of physical activity. In addition, the measurements of liking and RRV were limited to the activities that were studied, and different results might be obtained if different sedentary and physically active alternatives were studied. One major advantage of the questionnaire is that it greatly expands the types of activities that can be assessed in future studies. The questionnaire can be used to ask children to choose between more activities, such as soccer and tennis that cannot be replicated in the laboratory. Another limitation is that there are no data regarding the validity of VAS methods for assessing the liking of activities.
An alternative interpretation of the data is that the allocation of responding for the sedentary versus the active alternative could be due to avoidance of physical activity due to aversive characteristics of physical activity, rather than low reinforcing value of being active. There are characteristics of physical activity that may be aversive and lead to avoidance, such as poor performance, sweating, discomfort or pain, and later muscle soreness. It would be interesting to test whether low response rates for physical activity were due to low reinforcing value of this alternative, or avoidance of being active due to aversive qualities of physical activity. The most direct way to test this would be to arrange an experimental paradigm to assess whether physical activity serves as an aversive stimulus which can reduce the rate of behavior that it follows.
In conclusion, we have examined the association of RRV and liking of physical activity with the usual participation in physical activity of children. The RRV and liking of physical activity were independently associated with time in MVPA. This is the first example of a separation between the reinforcing value and liking of physical activity in children. Children who have both a high reinforcing value and liking of physical activity engage in greater MVPA than other children. The reinforcing value and liking of physical activity may play an important role in influencing the amount of MVPA children perform. A better understanding of how the reinforcing value and liking of activity develops, as well as methods to make physical activity more reinforcing and liked may improve our efforts to help youth be more physically active and derive the health benefits associated with an active lifestyle.