Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
August 2003

Physical Activity and Body Mass Index Among US AdolescentsYouth Risk Behavior Survey, 1999

Author Affiliations

From the Divisions of Nutrition and Physical Activity (Drs Levin, Brown, and Dietz) and Adolescent and School Health (Dr Lowry), National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Ga.


Copyright 2003 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2003

Arch Pediatr Adolesc Med. 2003;157(8):816-820. doi:10.1001/archpedi.157.8.816

Objective  To investigate associations of underweight and overweight with physical activity among high school students in the United States.

Methods  A nationally representative sample of 15 349 US high school students participated in the 1999 Youth Risk Behavior Survey; 13 295 were included in these analyses. Five measures of physical activity were examined as dichotomous variables: (1) vigorous-intensity physical activity (≥3 vs <3 sessions lasting at least 20 minutes each per week); (2) moderate-intensity physical activity (≥5 vs <5 sessions lasting at least 30 minutes each per week); (3) strength training (≥3 vs <3 sessions per week); (4) enrollment in physical education (yes or no); and (5) sports participation (yes or no). Using body mass indexes, students were categorized by percentiles as underweight (≤5th percentile), at risk for underweight (>5th to ≤15th percentiles), normal weight (>15th to <85th percentiles), at risk for overweight (≥85th to <95th percentiles), or overweight (≥95th percentile). Potential associations between physical activity and body mass index were examined using logistic regression.

Results  On several measures, adolescent boys who were underweight or overweight were less likely to be physically active than boys of normal weight (eg, odds ratio [OR], 0.23; 95% confidence interval [CI], 0.12-0.45; and OR, 0.75; 95% CI, 0.61-0.93; for boys who were underweight and overweight, respectively, for strength training). Adolescent girls who were overweight or at risk for overweight were less likely (OR, 0.62; 95% CI, 0.50-0.78; and OR, 0.63; 95% CI, 0.46-0.85; respectively) to be involved with sports than girls of normal weight; and girls who were underweight were less likely (OR, 0.44; 95% CI, 0.22-0.91) to be enrolled in physical education.

Conclusions  Weight status among high school students is correlated with selected physical activity behavior, especially among adolescent boys. Interventions to increase physical activity for high school students should target adolescents of all shapes and sizes, and may best be achieved by school policies requiring physical education or after-school sports.

IN THE United States1 and in many other advanced nations, overweight and obesity among youth are increasing at an alarming rate. Correspondingly, the extent to which young people engage in physical activity is of great public health interest. The levels of physical activity begin to decline before high school and continue to drop through the high school grade levels,2,3 but reports on the relationship between engaging in physical activity and overweight have been inconsistent.

One of the difficulties in establishing such a relationship has been the variation in approaches used to measure physical activity. In a national sample, reported participation in vigorous physical activity was not associated with weight status among youth, but boys aged 8 to 16 years who were overweight were less likely to participate in sports or exercise programs than boys who were not overweight.4 The same was true for a subset of girls, aged 14 to 16 years.4 In a small study5 in South Carolina among 187 middle school students, overweight students (body mass index [BMI] ≥95th percentile) were less physically active than their nonoverweight counterparts, as measured by an accelerometer. Also in this study, overweight children were involved in fewer community physical activities than those who were not overweight. To date, the published literature has failed to include any studies that have compared levels of participation in physical activity of children and adolescents who are underweight with those of their nonunderweight peers.

Since 1991, the Youth Risk Behavior Surveillance System has monitored behavioral trends among high school–age adolescents, including their physical activity.6 In the 1999 Youth Risk Behavior Survey (YRBS), self-report of height and weight permitted determination of the relationship between weight status and physical activity behaviors for the first time.

The objective of this study was to determine whether an association would be found between physical activity and weight status (underweight and overweight) among a nationally representative sample of high school students. Because of the complexity of measuring physical activity, we selected 5 indicators for examination: vigorous-intensity physical activity, moderate-intensity physical activity, strength training, enrollment in physical education, and participation on a sports team.


The design of the 1999 YRBS has been described in detail elsewhere.7 Briefly, a 3-stage cluster sampling procedure was used to produce a nationally representative sample of 9th through 12th grade students in public and private schools in the 50 states and District of Columbia. The 3 levels of sampling were large counties or groups of smaller adjacent counties that made up the primary sampling units, schools, and classes. Schools with many African American and Hispanic students were sampled at a higher rate to ensure reliable prevalence estimates for these subgroups. An institutional review board at the Centers for Disease Control and Prevention reviewed and approved the YRBS; parental consent was obtained before survey administration, following local school procedures. Students who responded to the survey did so on a voluntary basis, and safeguards are in place to protect against the identification of the students. Response rates were 77% for schools and 86% for students, resulting in an overall response rate of 66%.7


The 92-item 1999 YRBS was administered in the classroom by trained personnel. Data used in the present analyses included the demographic variables of grade level, sex, and ethnicity; height and weight as determined by self-report; and 5 items to assess physical activity. Physical activity variables were dichotomized as physically active or insufficiently active, using the following criteria for "active": 3 or more sessions of at least 20 minutes per week of vigorous physical activity, 5 or more sessions of at least 30 minutes per week of moderate physical activity, 3 or more sessions per week of strength training, enrollment in physical education, and participation on 1 or more sports teams. These criteria mirror recommended standards for physical activity, where available. For each behavior, a single question was used for assessment: "On how many of the past 7 days did you exercise or participate in physical activity for at least 20 minutes that made you sweat and breathe hard, such as basketball, soccer, running, swimming laps, fast bicycling, fast dancing, or similar aerobic activities?" "On how many of the past 7 days did you participate in physical activity for at least 30 minutes that did not make you sweat or breathe hard, such as fast walking, slow bicycling, skating, pushing a lawn mower, or mopping floors?" "On how many of the past 7 days did you do exercises to strengthen or tone your muscles, such as push-ups, sit-ups, or weight lifting?" "In an average week when you are in school, on how many days do you go to physical education (PE) class?" "During the past 12 months, on how many sports teams did you play?"

Self-reported height and weight were used to compute BMI, calculated as weight in kilograms divided by the square of height in meters. Weight status categories (by percentiles) were created using age- and sex-specific cut points from the revised growth charts produced by the Centers for Disease Control and Prevention.8 Adolescents were categorized as underweight (≤5th percentile), at risk for underweight (>5th to ≤15th percentiles), normal weight (>15th to <85th percentiles), at risk for overweight (≥85th to <95th percentiles), or overweight (≥95th percentile).

Of 15 257 survey respondents with data for height and weight, 13 295 (87.1%) were included in these analyses. The students with a BMI of less than 11 (n = 1) or greater than 55 (n = 8) were excluded. In addition, students with missing data for ethnicity (n = 113), grade level (n = 42), or any of the 5 physical activity variables were excluded. Most missing data were derived from the physical education enrollment question (n = 1081). An analysis of the demographics (sex, grade level, and ethnicity) for this subsample showed no significant differences compared with those of the students who did answer the question.

The YRBS has acceptable reliability in measuring health-risk behaviors in adolescents,9 and self-reported height and weight among a national sample of adolescents correlated well with measured BMI (r = 0.92).10 In addition, the specificity of self-reported overweight is high.11


All analyses were performed using an SAS version of SUDAAN (Research Triangle Park, NC). The weighted percentages of students classified as physically active by each of the 5 indicators were calculated and compared between demographic subgroups (adolescent boys vs girls, ethnicity groups, and grade levels), as were the weighted percentages of students classified in each weight status category. Differences in prevalence estimates were considered statistically significant if 95% confidence intervals did not overlap.

Multiple logistic regression analyses were performed to obtain adjusted odds ratios (ORs) and 95% confidence intervals, using 1 physical activity variable at a time as the dependent variable and weight status categories, with normal weight as the referent group, as the independent variables. Models were run separately by sex and adjusted for grade level and ethnicity (non-Hispanic white, African American, Hispanic, or other). Odds ratios were considered statistically significant if 95% confidence intervals did not include 1.0.


Of the 13 295 high school students, 78.4% were classified as active on 1 or more of the 5 physical activity measures. Just over 9% were classified as active on all 5 measures (Table 1), and 11.9% were not sufficiently active by any measure. Students were least likely to be active (26.8%) when the criterion was moderate-intensity physical activity. On all 5 measures of physical activity, adolescent boys were more likely than girls to be classified as active (P<.05 for all).

Table 1. 
Image not available
Weighted Prevalence Estimates and 95% Confidence Intervals of Physically Active Students by Sex

Estimates for vigorous physical activity and sports participation were higher for whites than for African Americans or Hispanics (Figure 1). Participation in physical education (Figure 2) declined with each increasing grade level (79.3% in 9th grade to 36.2% in 12th grade). Vigorous-intensity physical activity, strength-training exercise, and sports team participation also declined with increasing grade level, but not to the degree seen with physical education.

Figure 1.
Image not available

Weighted prevalence estimates of physically active students by ethnicity.

Figure 2.
Image not available

Weighted prevalence estimates of physically active students by grade level.

Among adolescent girls, being underweight was associated with decreased odds of physical activity as measured by physical education enrollment (Table 2), and being overweight and at risk for overweight were associated with decreased odds when sports participation was the measure. Among boys, being underweight was associated with decreased odds of physical activity as measured by vigorous-intensity physical activity, strength training, and sports participation. Overweight was associated with decreased odds of physical activity when moderate-intensity physical activity and strength training were the measures.

Table 2. 
Image not available
Prevalence and Adjusted Odds Ratios (ORs) for Being Physically Active by Weight Status

Many high school students are insufficiently active. For example, we found that 44.3% of high school students are not enrolled in physical education, and almost half do not participate on any sports team or engage in strength-training exercises often enough to reap their physical and psychological health benefits.

On the other hand, 65.4% of students are meeting the recommendation for vigorous physical activity as proposed by an international consensus panel (3 times per week or more for at least 20 minutes).12 Based on a review of 26 published studies, Epstein and colleagues13 found that youth attained approximately 30 minutes per day of moderate-intensity physical activity (using 50%-60% of their heart rate reserve) and almost 15 minutes per day of vigorous physical activity (using 60%-70% of their heart rate reserve). Taken together, the self-reported data from the YRBS and the heart rate data presented in the review article by Epstein et al suggest that most youth are meeting the physical activity recommendations for vigorous activity. Given the obesity epidemic among youth, it seems incongruent to suggest that youth are getting enough physical activity. In fact, the latest recommendations call for children and teenagers to engage in 1 hour of activity on most days of the week.14

Our analyses demonstrate that participation in physical activity is associated with weight status among US high school students, although not by all measures of physical activity. Similar results were recently reported among a sample of 187 middle school students (54 overweight and 133 nonoverweight); overweight students were less active than their nonoverweight counterparts as measured by an objective measure of physical activity (computer science and applications accelerometer), and they reported less involvement in community-based physical activity organizations.5 Our findings that the association between physical activity and weight status was more consistent among adolescent boys than girls are not unique to this study. Nationally representative data from the Third National Health and Nutrition Examination Survey revealed sex differences in associations between overweight and physical activity.4

Adolescent boys who were overweight were no less likely to participate in vigorous physical activity than boys of normal weight, but less likely to participate in moderate physical activity. One explanation is that overweight for this study is defined by BMI, which accounts for height and weight, not body composition. Some of the boys classified as overweight may actually be muscular and not overweight in terms of body fat. These muscular boys may be more likely to engage in vigorous activities than in moderate activities, which may be viewed as too easy. Far fewer boys of any weight category indicated that they engaged in moderate activity compared with vigorous activity.

Unique to our study is the finding that adolescent students, especially boys, who are underweight or at risk for underweight are less likely to be physically active than students of normal weight. Unfortunately, we could not control for health status, which may be compromised among underweight students and thereby account for the association. Girls who perceive themselves as underweight are more likely to smoke cigarettes,15 which may correspond with a cluster of behaviors that are incompatible with physical activity. Boys who are underweight are more likely to report disliking school,16 which may also extend to disliking school-related opportunities for physical activity (eg, strength training and sports participation).

At least 3 limitations apply to our study. First, the physical activity behavior, height, and weight data used in this study were based on self-report, and thus bias may be present, especially at the extremes, as students may tend to respond in a socially desirable manner. In addition to potential bias, there is potential for misclassification because of interpretation of the questionnaire items in regard to physical activity. As for the self-report of height and weight, adolescent girls and boys may respond differently. In a sample of college students, men reported their height to be taller than measured values, while women reported their weight to be less than measured values.17 Second, neither energy expenditure nor energy intake can be calculated by the YRBS, and the cross-sectional nature of these analyses precludes the establishment of causation. For example, emotional factors related to overweight or underweight may have led to physical inactivity, or physical inactivity may have led to overweight. Falkner et al16 reported that overweight girls, and overweight and underweight boys, have social and psychological risks compared with adolescents of normal weight. A third limitation is the unmeasured confounding and mediating variables. For example, illness or disease may be related to physical activity and weight status, but we could not control for this factor in our analyses.

The opportunity to investigate several measures of physical activity and weight status offered a unique strength to the present investigation. Weight status, for the most part, was not related to physical education enrollment; only adolescent girls who were underweight were less likely to be enrolled in physical education than girls of normal weight. Although opportunities for participation on sports teams and other modes of obtaining vigorous activity may vary by setting, physical education is available almost uniformly and thereby provides an important public health opportunity. According to the 2000 School Health Policies and Programs Study, 88.8% of school districts require that physical education be offered in high school.18 On the other hand, when the data are aggregated at the school level in regard to student requirements, only 13.3% of high school freshmen (9th grade) and 5.4% of high school seniors (12th grade) are required to take physical education. In addition, 79.6% of the high schools that require physical education allow students to be exempted for 1 or more grading periods for reasons such as religious beliefs or practices, participation in school sports, and participation in other school activities. These data highlight the disparity between district-wide policy and school-level enforcement. Mandating physical education across the grade levels might increase the physical activity level of high school students, regardless of body shape or size.

Schools have long been recognized as an important setting in which to promote national health objectives,1,19,20 including lifetime physical activity.21 Daily physical education from kindergarten through 12th grade is recommended by the Centers for Disease Control and Prevention,19 the American Heart Association,22 and the National Association for Sport and Physical Education.23 Nonetheless, the nation is far from implementing this recommendation. According to our data, close to half of all high school students are not enrolled in physical education even 1 day per week. A significant public health opportunity is missed by not requiring enrollment across the high school years. The need to put into practice effective school policies to increase enrollment in physical education or after-school sports, and thereby increase physical activity levels among students of all shapes and sizes, is emphasized by our findings. The psychological and physical health benefits of physical activity should be enjoyed by all.

What This Study Adds

Overweight, obesity, and physical activity have been identified as leading health indicators in the national health promotion initiative Healthy People 2010. Some evidence suggests that adolescent youth who are overweight are less likely to participate in physical activity; however, the data are inconclusive.

Our study was the first, to our knowledge, to investigate the relationship between weight status, including overweight and underweight, and physical activity behaviors among a representative sample of US adolescents. The relationship between physical activity and weight status differed by form of activity (eg, sports participation, strength training, and physical education enrollment) and by sex. Physical activity interventions must target and embrace adolescents who are underweight and overweight, as well as those of normal weight.

Back to top
Article Information

Corresponding author and reprints: Sarah Levin, PhD, Division of Nutrition and Physical Activity, Centers for Disease Control and Prevention, 4770 Buford Highway NE, Mailstop K-46, Atlanta, GA 30341 (e-mail:

Accepted for publication November 7, 2002.

Troiano  RPFlegal  KMKuczmarski  RJCampbell  SMJohnson  CL Overweight prevalence and trends for children and adolescents: the National Health and Nutrition Examination Surveys, 1963 to 1991. Arch Pediatr Adolesc Med. 1995;1491085- 1091
Sallis  JF Epidemiology of physical activity and fitness in children and adolescents. Crit Rev Food Sci Nutr. 1993;33403- 408
Sallis  JF Age-related decline in physical activity: a synthesis of human and animal studies. Med Sci Sports Exerc. 2000;321598- 1600
Dowda  MAinsworth  BEAddy  CLSaunders  RRiner  W Environmental influences, physical activity, and weight status in 8- to 16-year-olds. Arch Pediatr Adolesc Med. 2001;155711- 717
Trost  SGKerr  LMWard  DSPate  RR Physical activity and determinants of physical activity in obese and non-obese children. Int J Obes Relat Metab Disord. 2001;25822- 829
Kolbe  LJ An epidemiological surveillance system to monitor the prevalence of youth behaviors that most affect health. Health Educ. 1990;2144- 48
Centers for Disease Control and Prevention, CDC Surveillance Summaries, Youth Risk Behavior Surveillance–United States, 1999. MMWR Morb Mortal Wkly Rep. 2000;49(SS-5)1- 32
Centers for Disease Control and Prevention, CDC Growth Charts: United States.  Hyattsville, Md National Center for Health Statistics, US Dept of Health and Human Services2000;
Brener  NDCollins  JLKann  LWarren  CWWilliams  BI Reliability of the Youth Risk Behavior Survey Questionnaire. Am J Epidemiol. 1995;141575- 580
Goodman  EHinden  BRKhandelwal  S Accuracy of teen and parental reports of obesity and body mass index. Pediatrics. 2000;106 ((pt 1)) 52- 58
Brener  NDMcManus  TGaluska  DALowry  RWechsler  H Reliability and validity of self-reported height and weight among high school students. J Adolesc Health. 2003; Apr32 ((4)) 281- 7Article
Sallis  JFPatrick  K Physical activity guidelines for adolescents: consensus statement. Pediatr Exerc Sci. 1994;6302- 314
Epstein  LHPaluch  RAKalakanis  LEGoldfield  GSCerny  FJRoemmich  JN How much activity do youth get? a quantitative review of heart-rate measured activity. Pediatrics [serial online]. 2001;108e44Available at: November 15, 2001
Not Available, Nutrition and Your Health: Dietary Guidelines for Americans. 5th ed. Washington, DC US Dept of Agriculture and US Dept of Health and Human Services2000;
Boles  SMJohnson  PB Gender, weight concerns, and adolescent smoking. J Addict Dis. 2001;205- 14
Falkner  NHNeumark-Sztainer  DStory  MJeffery  RWBeuhring  TResnick  MD Social, educational, and psychological correlates of weight status in adolescents. Obes Res. 2001;932- 42
Jacobson  BHDeBock  DH Comparison of body mass index by self-reported versus measured height and weight. Percept Mot Skills. 2001;92128- 132
Burgeson  CRWechsler  HBrener  NDYoung  JCSpain  CG Physical education and activity: results from the School Health Policies and Programs Study 2000. J Sch Health. 2001;71279- 293
Not Available, Guidelines for school and community programs to promote lifelong physical activity among young people. Centers for Disease Control and Prevention. MMWR Recomm Rep. 1997;46(RR-6)1- 36
McGinnis  JMDeGraw  C Healthy Schools 2000: creating partnerships for the decade. J Sch Health. 1991;61292- 297
Sallis  JFMcKenzie  TL Physical education's role in public health. Res Q Exerc Sport. 1991;62124- 137
Not Available, Strategic Plan for Promoting Physical Activity.  Dallas, Tex American Heart Association1995;
Not Available, Sport and Physical Education Advocacy Kit.  Reston, Va National Association for Sport and Physical Education1994;