Decreasing Sedentary Behaviors in Treating Pediatric Obesity | Lifestyle Behaviors | JAMA Pediatrics | JAMA Network
[Skip to Navigation]
Sign In
March 2000

Decreasing Sedentary Behaviors in Treating Pediatric Obesity

Author Affiliations

From the Departments of Psychology and Social and Preventive Medicine, University at Buffalo, The State University of New York.

Arch Pediatr Adolesc Med. 2000;154(3):220-226. doi:10.1001/archpedi.154.3.220

Background  Epidemiogical studies have shown television watching to be a risk factor for the development of obesity in children. The effect of reducing television watching and other sedentary behaviors as a component of a comprehensive obesity treatment program has not been thoroughly tested.

Objective  To compare the influence of targeting decreases in sedentary behavior vs increases in physical activity in the comprehensive treatment of obesity in 8- to 12-year-old children.

Design  Randomized, controlled outcome study.

Setting  Childhood obesity research clinic.

Design  Ninety families with obese 8- to 12-year-old children were randomly assigned to groups that were provided a comprehensive family-based behavioral weight control program that included dietary and behavior change information but differed in whether sedentary or physically active behaviors were targeted and the degree of behavior change required.

Results  Results during 2 years showed that targeting either decreased sedentary behaviors or increased physical activity was associated with significant decreases in percent overweight and body fat and improved aerobic fitness. Self-reported activity minutes increased and targeted sedentary time decreased during treatment. Children substituted nontargeted sedentary behaviors for some of their targeted sedentary behaviors.

Conclusion  These results support reducing sedentary behaviors as an adjunct in the treatment of pediatric obesity.

TELEVISION WATCHING has been cross-sectionally and prospectively related to childhood adiposity in several epidemiological studies.1-3 Television, along with other sedentary behaviors, may contribute to obesity by competing with more physically active behaviors,4 as well as setting the occasion for eating.2,3 Sedentary behaviors are reinforcing for children, and more reinforcing for obese than nonobese children.5 Reducing sedentary behaviors represents a potentially important goal in childhood obesity prevention and treatment.6

Sedentary and physically active behaviors can be considered alternative ways children can allocate some of their leisure time. The guidelines for moderate to vigorous physical activity are at least 30 minutes most days,7 which can be accomplished by reallocating less than 4 hours per week of sedentary behaviors to being physically active. Children may meet these guidelines by reallocating a percentage of their leisure time to physical activity, rather than exchanging all of their sedentary time for physical activity. Further reductions in sedentary behavior may not result in increased physical activity. Sedentary behaviors may also be so reinforcing for some obese children that reducing access to some sedentary activities may result in an increase in other sedentary behaviors, resulting in no net increase in physical activity.8

We have observed that reducing sedentary behavior is associated with increases in physical activity in the laboratory that are similar to those produced by targeting increases in physical activity.8,9 However, clinical research has shown that obese children targeted for reduction of sedentary behaviors have better weight loss at 1 year than children targeted for being more active.10 One purpose of this study was to provide another test of the hypothesis that decreasing sedentary behaviors produces equal or better weight loss and fitness change than increasing physical activity when included as part of a comprehensive pediatric obesity intervention. A second purpose was to test whether there was a dose-response relationship between the amount of reduction in sedentary behaviors and weight loss and fitness outcomes.

Participants and methods

Study participants

Ninety obese 8- to 12-year-old children and their parents were recruited in 2 cohorts that began 1 year apart through physician referrals, posters, and newspaper and television advertisements (Figure 1). Inclusionary criteria included child between 20% and 100% overweight, neither parent more than 100% overweight, 1 parent willing to attend treatment meetings, no family member participating in an alternative weight-control program, no parent or child with current psychiatric problems, and no dietary or exercise restrictions on the participating parent or child.

Figure 1. 
Screening, randomization, and participation flow of families in treatment.

Screening, randomization, and participation flow of families in treatment.

Randomization and treatment

Families were stratified by sex and randomly assigned to 1 of 4 groups that varied the targeted behaviors (sedentary behaviors vs physical activity), and treatment dose (low vs high). Low and high doses for the decrease sedentary or increase physical activity groups were 10 or 20 h/wk of targeted sedentary behaviors, or the equivalent energy expenditure of 16.1 or 32.2 km (10 or 20 mi) per week, respectively. These values bracketed the activity and sedentary goals used in our previous research.10

Common Components of Treatment

The 6-month treatment included 16 weekly meetings, followed by 2 biweekly and 2 monthly meetings. Families were seen at 12 and 24 months after randomization for follow-up. Families received parent and child workbooks, which included introduction to weight control and self-monitoring, the Traffic Light Diet,11 the specific activity program, behavior change techniques, and maintenance of behavior change. At treatment meetings participating family members were weighed, their weight was graphed, they met with an individual therapist for 15 to 30 minutes, and they attended separate 30-minute parent and child group meetings. During the individual meeting, therapists reviewed weekly weight change and diet, targeted physical or sedentary activities, and the behavioral contract.

The Traffic Light Diet11 was used to decrease energy intake and promote a balanced diet. Foods are categorized on the basis of their calorie and nutrient content. Green foods are very low in calories. Yellow foods are higher in calories and include the dietary staples needed for a balanced diet. Red foods are higher in calories with low nutrient density. While they were attempting to lose weight, children and overweight parents were instructed to consume between 4184 and 5021 kJ/d, to limit red foods to 10 or fewer per week, and to maintain nutrient balance by eating the recommended servings based on the food pyramid. When participants' weight decreased to the nonobese range, they were instructed how to develop a maintenance calorie level, which involved gradually increasing caloric intake in 418-kJ increments until weight gain occurred. Nonoverweight parents had no caloric restriction, but were asked to limit red foods. Families were provided additional nutritional information, including reading food labels and shopping, and were taught stimulus control to reduce access to high-calorie foods and increase access to healthy lower-calorie foods. Preplanning was taught to facilitate decision making and problem solving for difficult eating and activity situations, such as parties, holidays, and school or work functions.

Parents and children were taught positive reinforcement techniques including praise for targeted behaviors and reciprocal contracts in which parents and children set goals and reinforcers to be provided by the parent (or child) based on meeting the goal. Parents deposited $75 to be returned contingent on completing 75% of the treatment sessions and attending the 6- and 12-month follow-up. Families were paid $50 for their attendance at the 24-month follow-up.

Activity Programs

Participants assigned to the increase physical activity group were reinforced for increasing physical activities in addition to those engaged in at the onset of the program. Physical activities done as a required part of the work or school day were not counted in meeting activity goals. Participants assigned to the decrease sedentary activity group were reinforced for reducing sedentary behaviors that compete with being active or set the occasion for eating (watching television, watching videotapes, playing computer games, talking on the telephone, or playing board games). Not all sedentary activities were targeted, so that participants could substitute nontargeted sedentary activities for targeted sedentary activities. Academically relevant sedentary behaviors, such as homework or schoolwork, were not targeted for reduction.


Dependent measures were collected at baseline and 6, 12, and 24 months later. Height was measured in 0.32-cm (⅛-in) intervals using a stadiometer (Seca, Columbia, Md), and weight was measured in 0.55-kg (0.25-lb) increments using a balance beam scale (Healthometer, Bridgeview, Ill) that was calibrated daily. Body mass index (BMI [kg/m2]) was calculated and compared with population standards based on sex and age.12,13 Children and parents whose BMI was greater than the 85th percentile were considered obese. Percent overweight was established by comparing the subject's BMI to the 50th BMI percentile based on subject sex and age.12,13 Bioelectrical impedance was used to measure percent body fat, with equations based on sex and age (Model BIA-101A; RJL Systems, Clinton Township, Mich). Fat-free body mass (kg) was calculated based on weight and percent body fat measurements.

The majority of heights and weights were measured in the laboratory, with self-reported heights and weights used when families were unable to attend assessments. Because of underestimation of weight and overestimation of height,14 self-reported data was adjusted based on a data set of more than 1000 cases in which adult and child heights and weights were self-reported and then measured. Only 3 child observations were parent-reported at the 2-year assessment, while for the participating parents 1 observation at the 1-year assessment and 4 observations at the 2-year assessment were self-reported.

Fitness was assessed by monitoring heart rate (Polar Vantage SX HR Monitor; Polar CIC Inc, Port Washington, NY) during increasing workloads on a cycle ergometer adapted for children (Monark 868; Monark-Crescent AB, Varberg, Sweden). Children were seated for a 5-minute adaptation period, and began the test at 150 kiloponds per minute (kpm), with the workload increased by 75 to 150 kpm when the subject had completed 3 minutes at a workload and heart rate had stabilized. Each subject completed at least 3 workloads. Physical work capacity in kilopond meters per minute (kpm × min-1) at a heart rate of 150 bpm (PWC150) was calculated based on the regression between heart rate and workload. To control for the influence of increasing body mass on physical work capacity,15-17 PWC150 was divided by fat-free mass (PWC150× kgFFM-1). Maximal oxygen consumption was also estimated based on individual regression line between heart rate and the progressive workloads at a heart rate of 200 bpm, and adjusted for fat-free mass (mL × kgFFM-1× min-1).

Children and parents completed a physical activity questionnaire, based on the Minnesota Leisure Time Activity Survey,18 that assessed frequency and average time spent on each of 39 activities during the last month. Percent of time engaged in targeted and nontargeted sedentary behaviors (<2.9 metabolic equivalents [METs]), and moderate intensity or greater (>3 METs) physical activity was calculated. Self-report of percent of time in physical activity was significantly related to activity counts assessed by 3 days of accelerometer readings at baseline (r = 0.37, P<.001), 6 months (r = 0.38, P<.001), and 2 years (r = 0.26, P = .05). Socioeconomic status was assessed using the Hollingshead Four-Factor Index of Social Status.19

Statistical analysis

Group differences at baseline were assessed using 1-way analyses of variance. Treatment effects for body composition and fitness were assessed using mixed analysis of variance (ANOVA) with between-subject factors of cohort—increase physical activity or decrease sedentary behaviors and low or high behavior change—and a repeated-measures within-subjects factor. There were no differential rates of change × group for the 2 cohorts, and the results were presented across the 2 cohorts. Significant effects of time or the interaction of treatment × time observed in the ANOVA were probed using F tests for simple effects.

Percent overweight change analyses were conducted based on the subjects who attended the follow-up measurements, as well as intent-to-treat analyses. The percentage of the families providing height and weight at 6, 12, and 24 months was 88.9% (80/90), 87.8% (79/90), and 84.4% (76/90), respectively. In the intent-to-treat analysis missing values at 2 years were assumed to return to their baseline percent overweight. Missing data points at 6 months or 1 year were estimated using regression models for each treatment group.

Changes in the percent of time in physically active and targeted and nontargeted sedentary behaviors from baseline to 6 and 24 months after treatment were analyzed using mixed ANOVA. To reduce overestimates of physical activity at the 0-, 6-, and 24-month time points, children who reported more than 3 hours per day of moderate activity (3.0-4.9 METs) and more than 2 hours per day of activity per day greater than 5 METs were removed from the dataset. Activity and sedentary analyses were based on complete data for 48 children.


There were no significant baseline group differences for those completing the 2-year trial for anthropometric, body composition, fitness, and activity data, as presented in Table 1. Sixty-four percent (49/76) of the participating parents were obese. The 33 obese mothers were 39.7 ± 4.6 years of age and 51.5% ± 23.1% overweight, and the 16 obese participating fathers were 41.6 ± 4.6 years of age and 43.7% ± 15.7% overweight.

Table 1. 
Child and Parent Descriptive Data for Those Who Completed the Trial*
Child and Parent Descriptive Data for Those Who Completed the Trial*

There was no differential attrition across the 4 treatment groups from baseline to the 2-year assessment (P = .73). Significant (P<.001) decreases in percent overweight (Figure 2) were observed from baseline to 6 months through 2 years. Percent overweight decreases of 25.5% at the end of treatment represented a reduction of 41% from baseline, which was associated with average child growth of about 3.5 cm and average weight loss of 6.0 kg (Table 2). During 2 years the average child grew 11.4 cm and gained 9.0 kg, which was related to a decrease in percent overweight of 12.9%, and a reduction in percent overweight from baseline of 20.8%. Intent-to-treat analyses showed significant changes from baseline through 2 years (P<.001), with decreases in percent overweight of 22.7% at the end of 6 months and a decrease of 10.9% overweight at 2 years.

Figure 2. 
Change in percent overweight from baseline for obese children in the experimental groups at 6, 12, and 24 months. Changes from baseline were significant at 6, 12, and 24 months (P<.001).

Change in percent overweight from baseline for obese children in the experimental groups at 6, 12, and 24 months. Changes from baseline were significant at 6, 12, and 24 months (P<.001).

Table 2. 
Child Changes From Baseline to 6 and 24 Months*
Child Changes From Baseline to 6 and 24 Months*

Physical work capacity (Figure 3) also significantly (P<.001) improved, with increases of 33% (P<.001) from 341.1 ± 104.7 kpm × min-1 at baseline to 454.2 ± 119.9 kpm × min-1 at 6 months and to 530.0 ± 140.5 kpm × min-1 at 24 months, an increase of 55% (P<.001). The mL × kgFFM-1× min-1 changes represented significant (P<.001) increases from 41.3 mL × kgFFM-1× min-1 at baseline to 49 and 48.3 mL × kgFFM-1× min-1 at 6 and 24 months, respectively. Significant decreases were observed for percent body fat (P<.001) at both 6 and 24 months, but the decrease in percent body fat was reduced from 6.4% to 2.0% body fat. There are no large population-based samples to evaluate changes in body fat during development to better understand changes in relationship to normative data, which are available for BMI. No significant differences in the rate of change by group were observed for any anthropometric or fitness measure.

Figure 3. 
Changes in physical work capacity (kilopond meters per minute) for obese children in the experimental groups at 6, 12, and 24 months. Changes from baseline were significant at 6, 12, and 24 months (P<.001).

Changes in physical work capacity (kilopond meters per minute) for obese children in the experimental groups at 6, 12, and 24 months. Changes from baseline were significant at 6, 12, and 24 months (P<.001).

A significant increase in percent of time being active (Figure 4) was observed from baseline to 2 years (P<.05). Targeted sedentary behaviors showed a significant decrease from baseline at 6 (P<.001) and 24 (P<.05) months. Nontargeted sedentary behaviors were increased from baseline at 6 months (P<.05).

Figure 4. 
Changes (mean ± SEM) in percentage of time in physically active behaviors and targeted and nontargeted sedentary behaviors across groups for obese children at 6 and 24 months.

Changes (mean ± SEM) in percentage of time in physically active behaviors and targeted and nontargeted sedentary behaviors across groups for obese children at 6 and 24 months.

Obese participating parents also showed significant (P<.001) decreases in weight from baseline to 6 months through 2 years for those who completed the trial. The parents who completed the trial lost 12.0 kg at 6 months, 9.9 kg at 12 months, and 7.1 kg at 2 years. This represents a maintained decrease from initial weight of 7.8%.


The results provide experimental evidence that reducing access to sedentary behaviors is an alternative to targeting physical activity in the treatment of childhood obesity. The 2 approaches were associated with similar decreases in percent overweight and increases in fitness during the 2 years of observation. The use of a diet-only, no physical activity group would have been useful to provide a control group to contrast normal changes in physical activity or fitness when physical activity or sedentary behaviors were not targeted in any way.

Previous controlled research suggested that being less sedentary may improve child weight control more than reinforcing children for increasing their physical activity.10 In that study we found the influence of targeting a decrease in sedentary behaviors on percent overweight decrease at 12 months was similar (-26.2%) to the current study (-24.4%), but the effect for increasing physical activity was considerably lower in the previous compared with the current study (-10.1% vs -20.2%). Based on the enhanced changes observed in the current study for groups targeted for increasing physical activity, the results are similar across interventions, consistent with laboratory studies suggesting they are alternative ways to increase physical activity.9

Significant increases were observed in physical activity and significant decreases in targeted sedentary time. Children engaged in more of the targeted than nontargeted sedentary behaviors at baseline. When time in these sedentary behaviors was reduced, some was allocated to being more physically active, while some was allocated to other sedentary behaviors. Substitutability of sedentary for active behavior is based in part on the degree of preference for the sedentary behaviors and on availability of access to alternatives. Reducing access to sedentary behaviors may be very effective when there are a wide variety of acceptable physically active alternatives available. Children without access to enjoyable physical activities may not increase their activity when targeted sedentary behaviors are decreased, but rather switch to other sedentary behaviors.

Blair et al20 have argued that improvements in activity and fitness provide for larger health improvements than observed for other behavior changes, and fitness improvements can reduce health risks even for those who remain obese. The dietary and physical activity changes that comprise negative energy balance and weight loss may have different effects on the risk factors associated with being obese.21 The changes observed in physical work capacity at 6 months were maintained at 2 years, compared with a maintenance of 50.5% of treatment-influenced loss in relative weight, suggesting that physical activity changes were maintained better than dietary changes. The fitness changes were consistent with the changes in self-reported time being active, which increased at 6 and 24 months.

No differences in percent overweight or fitness change were observed for children randomized to decrease targeted sedentary behaviors to no more than 10 or 20 hours per week. There may be several reasons for this. Reducing targeted sedentary behaviors to 20 hours per week makes sufficient leisure time available to meet activity goals, and further reductions in sedentary behavior to 10 hours per week may not result in further increases in physical activity. It is also possible that while the groups were designed to be implemented at different doses, this may not have been accomplished. Similar to an exercise dose-response study in adults,22 families given the lower prescription may have implemented them at a slightly higher level, and those families given the higher dose implemented them at a slightly lower level, making the final dose very similar. It is also possible that as the dose increases, it may be harder to comply with treatment recommendations, reducing differences between groups.

Self-report of targeted sedentary behaviors decreased 13.4% during treatment, with allocation of 4.1% to physical activity and 9.3% to other sedentary behaviors. By the 2-year follow-up, self-report of being active had increased by 7.5%, which was associated with a decrease in percent time in targeted sedentary behaviors of 8.7% and an increase in other sedentary behaviors of only 1.2%. It is important in future research to obtain information on the pattern of changes children make in their physical activity. For example, it would be interesting to determine if there are specific types of sedentary behaviors that compete with being physically active, and are they the child's favorite sedentary behaviors or ones that are less preferred and probably easier to give up? Factors in the environment that support a sedentary lifestyle should be changed to reduce the probability of returning to a sedentary lifestyle.

Understanding the pattern of changes in active behavior and allocation of time to sedentary behaviors may be limited by self-report, and objective measurement of allocation of time to active and sedentary alternatives would provide better information on how children make choices to shift from being sedentary to becoming physically active. In addition, caloric intake was not measured in this study. Decreasing sedentary behaviors can reduce opportunities for eating, and thus make an important contribution to facilitating dietary adherence and caloric reduction.10 Further research should evaluate the influence of reductions in sedentary behavior on changes in intake and expenditure.

In summary, this study demonstrated that reducing sedentary behaviors as part of a comprehensive family-based weight control program is associated with similar weight, fitness, and psychological changes in comparison to a well-validated program to increase physical activity in children.23,24 This increases the flexibility of therapists who are treating obese children, and reinforces the idea that sedentary behavior is an important component in the acquisition and maintenance of excessive adiposity.

Editor's Note: I still like my idea of having TVs powered by electricity generated from viewers on stationary bicycles or treadmills.—Catherine D. DeAngelis, MD

Accepted for publication August 2, 1999.

This research was supported in part by grants HD 20829 and HD 25997 from the National Institute of Child Health and Human Development, Bethesda, Md (Dr Epstein).

Appreciation is expressed to Jodie O'Brien, MA, and Dominica Vito, MA, who assisted in coordination of the study, and Brian Saelens, PhD, and Michelle Myers, PhD, who served as therapists.

Corresponding author: Leonard H. Epstein, PhD, Department of Psychology, Park Hall, University at Buffalo, The State University of New York, Buffalo, NY 14260.

Andersen  RECrespo  CJBartlett  SJCheskin  LJPratt  M Relationship of physical activity and television watching with body weight and level of fatness among children: results from the Third National Health and Nutrition Examination Survey.  JAMA. 1998;279938- 942Google ScholarCrossref
Dietz  WHGortmaker  SL Do we fatten our children at the television set? obesity and television viewing in children and adolescents.  Pediatrics. 1985;75807- 812Google Scholar
Gortmaker  SLMust  ASobol  AMPeterson  KColditz  GADietz  WH Television watching as a cause of increasing obesity among children in the United States, 1986-1990.  Arch Pediatr Adolesc Med. 1996;150356- 362Google ScholarCrossref
Epstein  LH Application of behavioral economic principles to treatment of childhood obesity. Allison  DBPi-Sunyer  FXeds. Obesity Treatment Establishing Goals, Improving Outcomes and Reviewing the Research Agenda. New York, NY Plenum Press1995;113- 119Google Scholar
Epstein  LHSmith  JAVara  LSRodefer  JS Behavioral economic analysis of activity choice in obese children.  Health Psychol. 1991;10311- 316Google ScholarCrossref
Gortmaker  SLPeterson  KWiecha  J  et al.  Reducing obesity via a school-based interdisciplinary intervention among youth: Planet Health.  Arch Pediatr Adolesc Med. 1999;153409- 418Google ScholarCrossref
Pate  RRPratt  MBlair  SN  et al.  Physical activity and public health: a recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine.  JAMA. 1995;273402- 407Google ScholarCrossref
Epstein  LHSaelens  BEMyers  MDVito  D Effects of decreasing sedentary behaviors on activity choice in obese children.  Health Psychol. 1997;16107- 113Google ScholarCrossref
Epstein  LHSaelens  BEO'Brien  JG Effects of reinforcing increases in active behavior versus decreases in sedentary behavior for obese children.  Int J Behav Med. 1995;241- 50Google ScholarCrossref
Epstein  LHValoski  AMVara  LS  et al.  Effects of decreasing sedentary behavior and increasing activity on weight change in obese children.  Health Psychol. 1995;14109- 115Google ScholarCrossref
Epstein  LHWing  RRValoski  A Childhood obesity.  Pediatr Clin North Am. 1985;32363- 379Google Scholar
Must  ADallal  GEDietz  WH Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness.  Am J Clin Nutr. 1991;53839- 846Google Scholar
Must  ADallal  GEDietz  WH Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2): a correction.  Am J Clin Nutr. 1991;53839- 846Google Scholar
Stewart  AWJackson  RTFord  MABeaglehole  R Underestimation of relative weight by use of self-reported height and weight.  Am J Epidemiol. 1987;125122- 126Google Scholar
Armstrong  NWelsman  JKirby  BJ Peak oxygen uptake and maturation in 12-yr-olds.  Med Sci Sports Exerc. 1998;30165- 169Google ScholarCrossref
Janz  KFMahoney  LT Three-year follow-up of changes in aerobic fitness during puberty: the Muscatine Study.  Res Q Exerc Sport. 1997;681- 9Google ScholarCrossref
Katzmarzyk  PTMalina  RMSong  TMKBouchard  C Physical activity and health-related fitness in youth: a multivariate analysis.  Med Sci Sports Exerc. 1998;30709- 714Google ScholarCrossref
Taylor  HLJacobs  DRSchucker  BKnudsen  JLeon  ASDeBacker  G A questionnaire for assessment of leisure time activities.  J Chronic Dis. 1978;31741- 755Google ScholarCrossref
Hollingshead  AB Four-Factor Index of Social Status.  New Haven, Conn Yale University1975;
Blair  SNKohl  HWBarlow  CEPaffenbarger  RSGibbons  LWMacera  CA Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men.  JAMA. 1995;2731093- 1098Google ScholarCrossref
Gutin  BCucuzzo  NIslam  SSmith  CStachura  ME Physical training, lifestyle education, and coronary risk factors in obese girls.  Med Sci Sports Exerc. 1996;2819- 23Google ScholarCrossref
King  ACHaskell  WLTaylor  CBKraemer  HCDeBusk  RF Group- vs home-based exercise training in healthy older men and women: a community-based clinical trial.  JAMA. 1991;2661535- 1542Google ScholarCrossref
Epstein  LHWing  RRKoeske  ROssip  DJBeck  S A comparison of lifestyle change and programmed aerobic exercise on weight and fitness changes in obese children.  Behav Ther. 1982;13651- 665Google ScholarCrossref
Epstein  LHWing  RRKoeske  RValoski  A A comparison of lifestyle exercise, aerobic exercise and calisthenics on weight loss in obese children.  Behav Ther. 1985;16345- 356Google ScholarCrossref