Secondary Logo

Journal Logo


Reducing Sedentary Behavior: Role in Modifying Physical Activity

Epstein, Leonard H.; Roemmich, James N.

Author Information
Exercise and Sport Sciences Reviews: July 2001 - Volume 29 - Issue 3 - p 103-108
  • Free



This review discusses the use of reducing sedentary behaviors as a way to increase physical activity. Behavioral choice theory is presented to understand the process of choosing to be physically active or sedentary, followed by a discussion of the relationship between sedentary behaviors and physical activity. We then provide ideas for new research on sedentary behavior and physical activity.


Behavioral economics provides a model with which to study whether people choose to be physically active or sedentary, as well as the type, intensity, and duration of activity (2). To date, applications of this model to the physical activity field have occurred primarily in children. As shown in Figure 1, engaging in physical activity usually involves choosing exercise over a concurrent and powerfully competing sedentary behavior. In the morning, do you exercise before breakfast or sleep in a little later and not exercise? At lunchtime, do you go for a swim and a fast lunch, or a more leisurely lunch without being physically active? On the weekend, do you watch sports or play sports? These decisions are based in part on differences in access to active and sedentary alternatives, which can be conceptualized as the work needed to obtain either type of activity, and on the motivation to engage in the activities, which can be conceptualized as the reinforcing value of the activities. If accesses to both activities are equal, people generally engage in the more reinforcing activity. If the reinforcing values of the activities are equal, people generally engage in the most accessible activity. In many situations, both the access to and the reinforcing value of the activities differ, and both factors are considered when making the decision to exercise or remain sedentary (10). A main reason for choosing sedentary behaviors is easy accessibility, whereas physically active alternatives require more work to engage in them and are often accompanied by modest discomfort and sweating. Physical activity, then, has the added cost of changing clothes, showering, and so on. It would be interesting to evaluate changes in television watching if you had to change clothes and shower after each half-hour of television viewed!

Figure 1
Figure 1:
Schematic of behavioral choice theory as applied to physical activity and the loss or gain in adiposity. Engaging in physical activity usually involves choosing exercise over a concurrent and competing sedentary behavior. This decision is based on environmental factors, including differences in access, availability (number and type of physically active and sedentary alternatives), reinforcing value (motivation to engage in the activities), and the timing of access to physically active and sedentary alternatives. Environmental factors are filtered through individual differences in impulsivity, which can be conceptualized as the preference for smaller immediate rewards (sedentary behavior) over larger delayed rewards (improved adiposity and later health benefits through exercise) and how sensitive people are to reinforcers. The combination of environmental factors and individual differences results in a choice of engaging in either physically active or sedentary behaviors. When choosing physically active behaviors, sedentary behaviors are reduced because they are replaced by physical activity and eating and energy intake is reduced due to fewer opportunities to eat. Choosing physical activity ultimately results in reduced adiposity. Choosing sedentary behaviors results in increased eating and food intake due to the complementary nature of these behaviors, and physical activity is reduced while adiposity increases.

The choice of being physically active or sedentary can be influenced by the work necessary to participate, but there also are large differences in the reinforcing value (motivation to participate) of physical activity. Some children find being sedentary highly reinforcing, so they are more likely to become sedentary and less likely to develop a regular physical activity habit (3). Very little is known with respect to what determines an individual’s motivation to be physically active, but it is obvious that some people really like to regularly walk, run, swim, or play a vigorous game of tennis or basketball, whereas others do not. The individual differences in the reinforcing value of physical activity are not surprising, as they also occur for many other behaviors as well (e.g., eating, drug use).

Changing the access to or options for certain behaviors in a choice situation often influences other behaviors. One way in which behaviors can be related is as substitutes. If behavior A substitutes for behavior B, then when behavior B is decreased, behavior A will increase to take the place of behavior B. For example, if an individual usually runs outdoors at lunch but the weather is poor, riding on an exercise bicycle or swimming indoors may be an adequate substitute. Likewise, based on the behavioral economics model, if an individual usually watches a lot of television after work and television is reduced, the individual may increase their physical activity. Behaviors can also be complements (i.e., the two behaviors occur together). If behavior A is a complement of behavior B, whenever behavior B increases, behavior A will increase. For example, eating may increase when television watching increases if in general these two behaviors tend to occur together. Behaviors can also be independent of each other, so that changes in one behavior do not influence the other behaviors. Understanding the relationships among behaviors is critical to making predictions about what happens when sedentary behaviors are modified.

Often the choice between being physically active or sedentary involves more than the immediateness of accessibility or reinforcing value of an activity. Many important health choices involve delayed outcomes. Decreased intake of fatty foods results in a loss of immediate pleasure but offers the potential for substantial delayed health benefits. There are large differences in the ability to delay gratification, and those who better control the impulse to choose the immediate, and often more reinforcing but less healthy, reward are more able to engage in current behaviors for the eventual larger benefits (1). Given the choice between immediately gratifying sedentary behaviors or physical activity that may have delayed health benefits, it would not be surprising to find those with poor impulse control choosing the sedentary alternative.


Pediatric obesity is related to increases in sedentary behavior, so it might be expected that obese children would find sedentary behaviors more reinforcing than physical activity, and it may be harder to shift the choice from being sedentary to physically active in obese children. We studied the relative reinforcing value of being sedentary versus being physical active in 8- to 12-y-old obese and nonobese boys and girls (7) (Figure 2). When provided the choice and equal access to sedentary or physical activities in a laboratory setting, both obese and nonobese children chose to be sedentary. As access to sedentary behaviors was reduced by increasing the work required to be sedentary, nonobese children quickly switched to working for the more easily obtained physical activity. After further increases in the amount of work required to obtain sedentary activity time, moderately obese children switched to working for physical activities but very obese children did not, even with large disparities in the work required to obtain access to sedentary versus physical activities. This study suggests that one factor that enhances the reinforcing value of sedentary behaviors in children is obesity.

Figure 2
Figure 2:
Allocation of time working for vigorous physical activity (mean ± S.E.) by nonobese, moderately obese, and very obese children in a choice situation in which schedules for vigorous activity remained at variable ratio (VR) 2, whereas the schedule for the higher-rated sedentary activity ranged from VR2 to VR32. The total time at each comparison was 180 s. The variable ratio defines the average number of responses needed to obtain a reinforcer. (Reprinted from Epstein, L.H., J.A. Smith, L.S. Vara, and J.S. Rodefer. Behavioral economic analysis of activity choice in obese children. Health Psychol. 10:311–316, 1991. Copyright © 1991 American Psychological Association. Used with permission.)

Accessibility to sedentary behaviors is much easier than physical activity for most people. In a laboratory setting, we varied the access of sedentary and active behaviors to explore the role of environmental access in replacing physically active for sedentary behaviors (10). Nonobese (not greater than 20% over ideal weight for height) sedentary male college students were randomly assigned to groups in which physical and sedentary activities were located either near or relatively far from the person. Near and far were operationalized as immediately accessible (i.e., in the same room) and a 5-min walk away, respectively. If sedentary behaviors were near, subjects were sedentary, but if sedentary activities were far and physical activities were near, subjects spent the entire session being physically active (Figure 3). Reducing access by requiring subjects to walk 5 min to obtain access to sedentary activities shifted the choice from being sedentary to being physically active. These data suggest that physical activity could increase if environments were changed to increase the proximity and convenience of physical activity and that decreased access to sedentary activities could increase physical activity, at least in certain population groups. However, most personal environments are changing in the opposite direction, with an increase in labor-saving devices and more convenient sedentary activities that decrease the need to be physically active.

Figure 3
Figure 3:
Time spent in exercise activities (mean ± S.E.) of young men under different experimental conditions. ExNEAR/SedNEAR indicates both physically active and sedentary activities located in the same room as the subject (proximally); ExNear/SedFAR, physically active alternatives located proximally, sedentary activities located 5 min walk away (distally); ExFAR/SedNEAR, physically active alternatives located distally, sedentary activities located proximally; ExFAR/SedNEAR, both physically active and sedentary alternatives located distally. [Adapted from Raynor, D.A., K.J. Coleman, and L.H. Epstein. Effects of proximity on the choice to be physically active or sedentary. Res. Q. Exerc. Sport. 69:99–103, 1998 (10). Copyright © 1998 American Alliance for Health, Physical Education, Recreation & Dance. Used with permission.]

Saelens and Epstein (13) identified the relationship between frequency of sedentary behaviors and the ability to substitute physically active behavior for sedentary behaviors (Figure 4). During a baseline session, sedentary adults chose between one physical activity (stationary bicycle) and three sedentary activities (playing video games, watching movies, listening to music). Reinforcing value (high, moderate, or low) of the sedentary activities was determined by ranking the amount of time spent engaged in that sedentary activity during the baseline session. During the next two laboratory sessions, subjects had to be physically active to gain access to selected sedentary behaviors. The groups were provided access to high, moderate, or low reinforcing sedentary activities contingent on being active. Less-desired sedentary activities that were not contingent on activity were freely available. Subjects increased their activity to gain access to high or moderately reinforcing sedentary activities but not to the low-preference sedentary activity or control. Thus, greatly valued sedentary activities such as watching television compete with physical activity more than do lower-valued sedentary activities such as reading, and low-preference sedentary behaviors may not substitute for high-preference sedentary activities.

Figure 4
Figure 4:
Change in physical activity during baseline (Bl) and intervention (Int 1, Int 2) days for young men and women who were in groups that provided high-rated, moderately rated, low-rated, or no sedentary activities contingent on physical activity. (Reprinted from Saelens, B.E., and L.H. Epstein. The rate of sedentary activities determines the reinforcing value of physical activity. Health Psychol. 18:655–659, 1999. Copyright © American Psychological Association. Used with Permission.)

A strategy that we have used to study the effect of reducing sedentary behavior on the choice of being sedentary or physically active is to reinforce children for reducing time being sedentary. We found that reinforcing children for being less sedentary increased physical activity similar to the changes that occurred when reinforcing increases in physical activity (5,6). There are alternative approaches to reducing sedentary activity, including restricting access to sedentary behaviors or punishing children for being sedentary. To test whether the method of reducing sedentary behavior influences physical activity in the laboratory, 8- to 12-y-old obese children were provided free access to sedentary behaviors and then randomized into groups in which they were either positively reinforced (earned points for rewards) for decreases in high-preference sedentary activity, punished (lost points) for high-preference sedentary activity, had access to high-preference sedentary activity restricted, or had no contingencies on activity (control group) (5). Children in the reinforcement and punishment groups were more physically active on intervention days than were children in the control group. Liking for targeted sedentary activities that included watching television, playing video games, and computer time decreased in the reinforcement group but increased in the restriction and control groups (Figure 5). These results suggest that reinforcing decreases in high-preference sedentary activity can both increase physical activity and decrease liking for targeted sedentary activities (5,6).

Figure 5
Figure 5:
Time spent in physical activities by obese children positively reinforced for not engaging in their two favorite sedentary activities (reinforcement), punished for engaging in their two favorite sedentary activities (punishment), access to their two favorite sedentary activities eliminated (restriction), or reinforced for attendance with no contingencies on activity (control) across the preintervention (Day 1), intervention (Days 2–4), and postintervention (Day 5) days (left). Liking ratings of the high preference sedentary activities on the adaptation and postintervention days in the reinforcement, punishment, restriction, and control groups (right). Liking of activities was measured with 100-mm visual analog scales anchored by did not like at all and like a lot. [Adapted from Epstein, L.H., B.E. Saelens, M.D. Myers, and D. Vito. Effects of decreasing sedentary behaviors on activity choice in obese children. Health Psychol. 16:107–113, 1997 (5). Copyright © 1997 American Psychological Association. Used with permission.]

In these studies, a small set of sedentary behaviors were targeted for change, which is similar to real-world situations in which a parent targets reducing a specific sedentary behavior such as television watching but children are free to spend time in other sedentary behaviors. The good news is that, in laboratory studies of children, sedentary activities do not readily substitute for each other (5,6), although currently there are few data to shed light on this issue in adults. If all sedentary activities were more substitutable for each other than physical activity, children would simply allocate all time spent in high-preference sedentary activities with time engaged in low-preference sedentary activities. When time in high-preference sedentary activities was reallocated, children chose to increase physical activity with some of the time and allocated the remainder to other sedentary behaviors. This is a worthwhile tradeoff, because it is not necessary to allocate all sedentary time to being active, particularly if there are large decreases in sedentary behavior. If children reduce television watching from 40 to 20 h/wk, they will have 20 h to reallocate to new activities. If they allocate one third of that time, or 1 h per day, to physical activity, there are still 13 h that they can allocate to other sedentary behaviors.

A common assumption is that children will engage in activities that they like, but this does not consider alternatives to the liked activity, or how liking of one type of activity relates to liking for another type. The most liked physical activity for a sedentary child may not rank very highly in comparison to highly liked sedentary activities. We tested how liking of physical activity can influence the choice between sedentary and active alternatives for sedentary children (7). Children were provided the choice of a highly liked sedentary activity and either a highly liked physical activity or a least liked physical activity. When the sedentary and physical activity options were equally available in both tasks, children chose to be sedentary. When the cost to be sedentary increased, children switched to become physically active, but the choice point for switching was similar whether the physical activity option was highly liked or least liked. Thus, liking or preference for an activity does not always predict what activities a person will choose.

A factor that may influence choice is the number of sedentary or physically active options available. Access to a wide variety of sedentary behaviors but a limited array of physical activities is quite common. Perhaps people would choose to be physically active if more options were available than sedentary options. To test this, in the laboratory, we evaluated the effects of the number of physically active alternatives and physical activity variety on the choice between physical and sedentary activities (15). When given a choice between physical and sedentary alternatives, nonobese women chose to be sedentary, even when multiple physical activities and only one sedentary behavior were available. However, when the choice was between a variety of physical activities versus one physical activity, subjects reliably were physically active for a longer time.

The implications of these studies for treatment outcome in obese children were tested in two clinical outcome investigations. In the first study (8), 8- to 12-y-old obese children were reinforced for being more active, reinforced for being less sedentary, or reinforced for the combination of the two. The basic comparison of reinforcing children for being more active or less sedentary was replicated in the second study of obese youth, with the addition of testing for a dose-response effect (4). Results across both studies suggest that with respect to weight or fitness changes, reinforcing children for being less sedentary is equal to or better than reinforcing children for being physically active (Figure 6). Children allocated approximately one third of the time freed-up from targeted sedentary behaviors for physical activity and two thirds for other sedentary behaviors (4). Improvements in weight control were due to increases in physical activity and decreases in energy intake (8) when targeted sedentary behaviors were reduced.

Figure 6
Figure 6:
Change in percentage overweight from baseline for children reinforced for increasing physical activity 10 h/wk (Low Exercise) or 20 h/wk (High Exercise) above baseline and children reinforced for reducing sedentary behaviors that compete with being active or set the occasion for eating (television, computer games, telephone, board games) by 10 h/wk (Low Sedentary) or 20 h/wk (High Sedentary) (left). Changes in physical work capacity (kilopond meters per min) of the experimental groups at 6, 12, and 24 months (right). For both variables, changes from baseline were significant (P < 0.001) at 6, 12, and 24 months. [Adapted from Epstein, L.H., R.A. Paluch, C.C. Gordy, and J. Dorn. Decreasing sedentary behaviors in treating pediatric obesity. Arch. Pediatr. Adolesc. Med. 154:220–226, 2000 (4). Copyright © 2000 American Medical Association. Used with permission.]

Reductions in sedentary behavior may also play a role in prevention of obesity. Robinson (11) found that a school-based program reduced the time spent watching television, videotaped movies, and playing video games of children (mean age 8.9 y). The reduction in television time was accompanied by fewer meals eaten in front of the television, and the intervention group had smaller increases in body mass index, triceps skinfold thickness, and waist girth compared with a control group.

Although we have made the argument that sedentary behaviors take the place of physical activity, if sedentary behaviors are more reinforcing than physical activity, perhaps they can be used to reinforce children for being more active. Within the laboratory, we provided 8- to 12-y-old obese children the choice among various sedentary activities and pedaling a stationary exercise bike. For one group, video game playing and watching movies were contingent on riding an exercise bicycle, whereas other sedentary activities of reading and drawing/coloring were freely available. In the other group, all sedentary activities and the exercise bike were freely available. Children in the contingent group increased their physical activity to gain access to the more highly reinforcing sedentary behavior (Figure 7), even when other sedentary activities were freely available (12). Similar results were obtained in another laboratory study when sedentary behaviors were not directly contingent on physical activity, but rather when obese children earned points based on their physical activity to be exchanged for subsequent access to sedentary behaviors (9). These studies suggest that sedentary behaviors can reinforce physical activity and that these sedentary behaviors are not completely substituted for by other sedentary activities.

Figure 7
Figure 7:
The change (mean ± S.E.) from baseline in minutes spent in physical activity (pedaling a stationary bicycle), targeted sedentary activities (video games, movies), and nontargeted sedentary activities (reading, drawing/coloring) in obese children. The choice to participate in targeted sedentary activities either was contingent on pedaling a stationary bicycle with nontargeted activities freely available (CONTINGENT group) or there were no contingencies on any activities (CONTROL group) on 2 separate days. (Reprinted from Saelens, B.E., and L.H. Epstein. Behavioral engineering of activity choice in obese children. Int. J. Obes. 22:275–277, 1998. Copyright © Macmillan Publishers Ltd. Used with permission.)


The behavioral economic model of comparing the reinforcing value of competing activities and manipulating the constraints on or environment of one activity or reinforcer provides valuable information about the efficacy of a reinforcer as well as the substitutability among different activities or reinforcers. Increased knowledge of how people choose to be active or sedentary should enhance clinical and population-based efforts to increase physical activity.

One important research area involves how to increase the reinforcing value of physical activity, so that more people choose to be active. Although research on reducing sedentary behaviors suggests that this can help increase physical activity among some groups, a more direct approach is to make physical activity more reinforcing, so that regardless of the sedentary options, people will opt for more physical activity. Could using physical activity as a reward or pairing physical activity with pleasant associations increase the reinforcing value of physical activity? As deprivation increases the reinforcing value of food, can deprivation increase the reinforcing value of activity? According to the disequilibrium theory of reinforcement (14), reducing physical activity below baseline levels would increase the reinforcing value of physical activity. Can brief shifts in reinforcing value produce more permanent shifts in the amount and reinforcing value of at least some types of physical activity?

A behavioral choice approach demonstrates the influence of the environment on the choice of being sedentary or active. Technological advances continue to promote sedentary behavior at work and during leisure times. A challenge will be to develop broad public health initiatives that increase access to physical activity while reducing access to sedentary behaviors.

The ability to facilitate the substitution of physical activities for sedentary activities requires further investigation. There are differences in the reinforcing value of sedentary behaviors for different individuals (13), implying that the choice of being physically active or sedentary depends in part on the characteristics and contexts of the sedentary behaviors that compete with being active as well as the characteristics and contexts of the physically active alternatives. It is also important to understand the relationships among health behaviors within a behavioral economics perspective. Although eating and physical activity are separate classes of behavior, they can be related. Physical activity can influence eating. Television watching and other sedentary activities may act as a complement to stimulate eating, and reducing these behaviors can reduce energy intake (8).

Behavioral economic research provides a conceptual and methodological framework in which to understand factors that influence food selection and consumption as well as types and amount of physical activity. Additional clinical and public health research is needed to identify new ways to use this information to influence the choice of healthy versus unhealthy behaviors and to translate this basic research into meaningful interventions to prevent and treat obesity as well as to promote physically active lifestyles.


The research presented in this article was supported in part by National Institutes of Health grants HD-25997, HD-20829, and HD-34284 (to L.H.E.).


1. Bickel, W.K., Madden, G.J., Petry N.M.. The price of change: the behavioral economics of drug dependence. Behav. Ther. 29: 545–565, 1998.
2. Epstein, L.H. Application of behavioral economic principles to treatment of childhood obesity. In: Obesity Treatment: Establishing Goals, Improving Outcomes and Reviewing the Research Agenda, edited by Allison, D.B. Pi-Sunyer, F.X.. New York: Plenum Press, 1995, p. 113–119.
3. Epstein, L.H., Kilanowski, C.K., Consalvi, A., Paluch R.. Reinforcing value of physical activity as a determinant of child activity level. Health Psychol. 18: 599–603, 1999.
4. Epstein, L.H., Paluch, R.A., Gordy, C.C., Dorn J.. Decreasing sedentary behaviors in treating pediatric obesity. Arch. Pediatr. Adolesc. Med. 154: 220–226, 2000.
5. Epstein, L.H., Saelens, B.E., Myers, M.D., Vito D.. Effects of decreasing sedentary behaviors on activity choice in obese children. Health Psychol. 16: 107–113, 1997.
6. Epstein, L.H., Saelens, B.E., O’Brien J.G.. Effects of reinforcing increases in active behavior versus decreases in sedentary behavior for obese children. Int. J. Behav. Med. 2: 41–50, 1995.
7. Epstein, L.H., Smith, J.A., Vara, L.S., Rodefer J.S.. Behavioral economic analysis of activity choice in obese children. Health Psychol. 10: 311–316, 1991.
8. Epstein, L.H., Valoski, A.M., Vara, L.S., McCurley, J., Wisniewski, L., Kalarchian, M.A., Klein, K.R., Shrager L.R.. Effects of decreasing sedentary behavior and increasing activity on weight change in obese children. Health Psychol. 14: 109–115, 1995.
9. Goldfield, G.S., Kalakanis, L.E., Ernst, M.E., Epstein L.H.. Open-loop feedback to increase physical activity in obese children. Int. J. Obes. 124: 888–892, 2000.
10. Raynor, D.A., Coleman, K.J., Epstein L.H.. Effects of proximity on the choice to be physically active or sedentary. Res. Q. Exerc. Sport. 69: 99–103, 1998.
11. Robinson, T.N. Reducing children’s television viewing to prevent obesity: a randomized control trial. JAMA 282: 1561–1567, 1999.
12. Saelens, B.E., Epstein L.H.. Behavioral engineering of activity choice in obese children. Int. J. Obes. 22: 275–277, 1998.
13. Saelens, B.E., Epstein L.H.. The rate of sedentary activities determines the reinforcing value of physical activity. Health Psych. 18: 655–659, 1999.
14. Timberlake, W., Farmer-Dougan V.A.. Reinforcement in applied settings: figuring out ahead of time what will work. Psych. Bull. 110: 379–391, 1991.
15. Vara, L.S., Epstein L.H.. Laboratory assessment of choice between exercise or sedentary behaviors. Res. Q. Exerc. Sport 64: 356–360, 1993.

obesity; behavioral choice theory; children; adults

© 2001 Lippincott Williams & Wilkins, Inc.