RESEARCH ARTICLE
Physical Activity of Fifth to Sixth Graders During School Hours According to School Race/Ethnicity: Suburban Cook County, Illinois SOYANG KWON, PhDa MARYANN MASON, PhDb SARAH WELCH, MPHc
ABSTRACT BACKGROUND: We compared moderate-to-vigorous intensity physical activity (MVPA) and inactivity levels among fifth and sixth graders during school hours according to school-level race/ethnicity and income attributes to inform school-based obesity interventions in Illinois Suburban Cook County (SCC). METHODS: Fifth- and sixth-grade students (N = 591) in 14 SCC public schools were asked to wear an accelerometer during school hours for 4 consecutive school days. School-level race/ethnicity was assigned based on the majority (≥70%) race/ethnicity of the student population: white (N = 4), black (N = 2), Hispanic (N = 5), or ‘‘mixed’’ (N = 3). All minority schools (N = 7) fell within a low-income category. RESULTS: A total of 538 students (91%) had valid accelerometry data. Average MVPA time during school hours was 20 minutes/day for boys and 11 minutes/day for girls. Multilevel linear regression models showed that compared to peers in white schools, boys, and girls in Hispanic schools engaged in 14 and 10 minutes less MVPA, respectively (p < .01). Girls in black and Hispanic schools had greater inactivity time than girls in white schools (p < .01). CONCLUSIONS: Objectively measured physical activity (PA) data support the need for school-level PA interventions during school hours, particularly in predominantly Hispanic schools. Keywords: children and adolescents; health disparities; minority youth; physical activity; sedentary; accelerometer. Citation: Kwon S, Mason M, Welch S. Physical activity of fifth to sixth graders during school hours according to school race/ethnicity: Suburban Cook County, Illinois. J Sch Health. 2015; 85: 382-387. Received on May 14, 2013 Accepted on December 19, 2014
P
hysical activity (PA) is a modifiable health behavior associated with a range of mental and physical health benefits as well as enhanced academic performance.1-3 Racial and ethnic disparities in moderate-to-vigorous intensity physical activity (MVPA) levels among adolescents have been documented with racial/ethnic minorities reporting lower MVPA levels.4-6 However, national accelerometermeasured PA data have shown less time spent in MVPA (MVPA time) among non-Hispanic white adolescents, compared with their non-Hispanic black or Hispanic peers.7,8 Household income also has been shown to be
positively associated with adolescent MVPA time.5 In addition to these individual attributes, environments have been suggested to be important determinants of PA behavior9 based on an ecological theoretical model.10 Schools are well-defined environments where school-aged children spend a significant portion of their weekdays. School-level attributes such as school racial/ethnic composition and the distribution of household income level in the student body may play a role in shaping student PA behaviors during school hours. Richmond et al4 found that although white
a
Research Assistant Professor, ([email protected]), Stanley Manne Children’s Research Institute, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605. bResearch Assistant Professor, ([email protected]), Stanley Manne Children’s Research Institute, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605. c Evaluation Manager, ([email protected]), Consortium to Lower Obesity in Chicago Children, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605. Address correspondence to: Soyang Kwon, Research Assistant Professor, ([email protected]), Stanley Manne Children’s Research Institute, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605.
382 •
Journal of School Health
•
June 2015, Vol. 85, No. 6
•
© 2015, American School Health Association
adolescents reported higher PA levels than black or Hispanic adolescents, PA levels did not differ among black, white, and Hispanic adolescent girls within a school. Furthermore, black and Hispanic adolescent boys were more active than white adolescent boys within a school. These evaluation results suggest that the racial/ethnic composition of a school may partly explain the association between individuals’ race/ethnicity and PA. The study by Richmond et al4 also reported that school-level median household income was positively associated with PA, and individual household income was not associated with PA after controlling for school-level income. However, study by Richmond et al4 used self-reported PA data which are subject to reporting bias for a child population.11 In addition, few studies have addressed whether PA differences by school racial/ethnic composition are present, particularly during school hours. Improved understanding of school-level attributes associated with student PA during school hours would inform school-based PA promotion strategies for school-aged children. Illinois Suburban Cook County (SCC) received Communities Putting Prevention to Work (CPPW) funding for obesity prevention and control efforts focused on policy, systems, and environmental change.12 To help guide the SCC CPPW school-based obesity prevention strategies for school-aged children, we evaluated accelerometry-measured MVPA and inactivity time during school hours among fifth and sixth graders in SCC schools based on school-level race/ethnicity and income attributes.
METHODS Participants The target population was school-aged students in SCC public schools. SCC, which refers to Cook County excluding the City of Chicago, includes 144 public school districts with approximately 380,000 public school students. This report used baseline data from the SCC CPPW school evaluation. The evaluation sample consisted of fifth- and sixth-grade students attending 15 public elementary schools participating in the school-based evaluation component of the SCC CPPW program. SCC CPPW is described elsewhere.12 Of the 15 participating schools, 10 schools were randomly selected from a stratified list (by geographical location and income level) of 21 PA-focused intervention schools serving fifth- and/or sixth-grade students. Each of these 10 schools consented to participate in the evaluation (0% refusal rate). A total of 20 comparison public elementary schools were selected from a sampling frame of 370 nonintervention public schools (schools not implementing a SCC CPPW PA intervention) with grades 5 and/or 6 in SCC to match with the intervention schools based on the Journal of School Health
•
stratification factors. Of the 20 comparison schools selected, 5 consented to participate (overall 75% refusal rate). The student sample included students from 2 randomly selected fifth- and/or sixth-grade classrooms in each of the 15 participating schools. A total of 627 students consented to wear accelerometers at School Year (SY) 2010 to 2011 (89% student participation rate). Parent/guardian passive consent (assumed participation unless they opted out through completion of a declination form) and students’ verbal assent were obtained. A stipend was provided to participating schools. Instruments and Procedures School-level data. SY 2010 to 2011 school race/ ethnicity composition and the proportion of students entitled to free/reduced-price lunch were obtained from the Illinois Interactive Report Card website http://iirc.niu.edu/. School race/ethnicity was assigned according to the student population’s majority race/ethnicity: non-Hispanic white (≥70% White students), non-Hispanic black (≥70% Black students), Hispanic (≥70% Hispanic students), and no majority race/ethnicity (mixed). On the basis of proportion of students entitled to free/reduced-price lunch, schools were classified into 2 income categories: high-income (0% to 50% of students eligible for free/reduced-price lunch) and low-income (51% to 100% of students eligible for free/reduced-price lunch). Accelerometry data. Actigraph accelerometers (model GT1M and GT3X; Pensacola, FL) were used to measure student PA. The accelerometer movement count data epoch was set at 5 seconds. Two research assistants (RAs) visited sampled school classrooms during the period from April to June 2011. Classroom teachers and students were informed about the evaluation and accelerometry data collection. RAs directly put an accelerometer belt on students’ waists with the accelerometer resting on the right hip on the first measurement day (usually Mondays). Students were asked to wear the accelerometers during school hours for 4 consecutive days. At the conclusion of each school day, students returned their assigned accelerometer to the teacher. Upon the start of school the following day, teachers assisted students in putting on their assigned accelerometer. RAs revisited the classrooms to collect accelerometers on the last measurement day (usually Thursdays). Records of school start time and end time on each of the accelerometer measurement days were obtained from classroom teachers. Accelerometry Data Reduction We only used accelerometry data recorded during a period of start time to end time of the school day. An accelerometer was considered as not worn if a
June 2015, Vol. 85, No. 6
•
© 2015, American School Health Association • 383
Table 1. Majority Race/Ethnicity and Income Level of 14 Participating Schools
Percentage of majority race/ethnicity (mean± SD) Income categories† High (N) Low (N)
Non-Hispanic White (N = 4)
Non-Hispanic Black (N = 2)
Hispanic (N = 5)
No Majority Race/ Ethnicity (N = 3)
73.5± 13.1
74.9± 0.1
95.6± 1.0
—∗
4 0
0 2
0 5
3 0
∗ Three schools had no single race/ethnicity that made up >70%. The means of the percentages of non-Hispanic Whites, Asians, and Hispanics were 38.6%, 31.5%, and 23.1%,
respectively. † On the basis of percentage of students entitled to free/reduced-price lunch, schools were classified into 2 income categories: 51-100% = low-income and 0-50% = high
income.
period of 60 minutes of consecutive zeros, allowing for 2 minutes of nonzero interruptions, was encountered anywhere in the data array. Accelerometry data were considered valid if wear time was 6 hours per school day for ≥3 wear days. MVPA was defined as ≥2296 accelerometer counts per minute.13,14 Inactivity was defined as
Physical Activity of Fifth to Sixth Graders During School Hours According to School Race/Ethnicity: Suburban Cook County, Illinois SOYANG KWON, PhDa MARYANN MASON, PhDb SARAH WELCH, MPHc
ABSTRACT BACKGROUND: We compared moderate-to-vigorous intensity physical activity (MVPA) and inactivity levels among fifth and sixth graders during school hours according to school-level race/ethnicity and income attributes to inform school-based obesity interventions in Illinois Suburban Cook County (SCC). METHODS: Fifth- and sixth-grade students (N = 591) in 14 SCC public schools were asked to wear an accelerometer during school hours for 4 consecutive school days. School-level race/ethnicity was assigned based on the majority (≥70%) race/ethnicity of the student population: white (N = 4), black (N = 2), Hispanic (N = 5), or ‘‘mixed’’ (N = 3). All minority schools (N = 7) fell within a low-income category. RESULTS: A total of 538 students (91%) had valid accelerometry data. Average MVPA time during school hours was 20 minutes/day for boys and 11 minutes/day for girls. Multilevel linear regression models showed that compared to peers in white schools, boys, and girls in Hispanic schools engaged in 14 and 10 minutes less MVPA, respectively (p < .01). Girls in black and Hispanic schools had greater inactivity time than girls in white schools (p < .01). CONCLUSIONS: Objectively measured physical activity (PA) data support the need for school-level PA interventions during school hours, particularly in predominantly Hispanic schools. Keywords: children and adolescents; health disparities; minority youth; physical activity; sedentary; accelerometer. Citation: Kwon S, Mason M, Welch S. Physical activity of fifth to sixth graders during school hours according to school race/ethnicity: Suburban Cook County, Illinois. J Sch Health. 2015; 85: 382-387. Received on May 14, 2013 Accepted on December 19, 2014
P
hysical activity (PA) is a modifiable health behavior associated with a range of mental and physical health benefits as well as enhanced academic performance.1-3 Racial and ethnic disparities in moderate-to-vigorous intensity physical activity (MVPA) levels among adolescents have been documented with racial/ethnic minorities reporting lower MVPA levels.4-6 However, national accelerometermeasured PA data have shown less time spent in MVPA (MVPA time) among non-Hispanic white adolescents, compared with their non-Hispanic black or Hispanic peers.7,8 Household income also has been shown to be
positively associated with adolescent MVPA time.5 In addition to these individual attributes, environments have been suggested to be important determinants of PA behavior9 based on an ecological theoretical model.10 Schools are well-defined environments where school-aged children spend a significant portion of their weekdays. School-level attributes such as school racial/ethnic composition and the distribution of household income level in the student body may play a role in shaping student PA behaviors during school hours. Richmond et al4 found that although white
a
Research Assistant Professor, ([email protected]), Stanley Manne Children’s Research Institute, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605. bResearch Assistant Professor, ([email protected]), Stanley Manne Children’s Research Institute, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605. c Evaluation Manager, ([email protected]), Consortium to Lower Obesity in Chicago Children, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605. Address correspondence to: Soyang Kwon, Research Assistant Professor, ([email protected]), Stanley Manne Children’s Research Institute, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E. Chicago Ave., Box 157, Chicago, IL 60611-2605.
382 •
Journal of School Health
•
June 2015, Vol. 85, No. 6
•
© 2015, American School Health Association
adolescents reported higher PA levels than black or Hispanic adolescents, PA levels did not differ among black, white, and Hispanic adolescent girls within a school. Furthermore, black and Hispanic adolescent boys were more active than white adolescent boys within a school. These evaluation results suggest that the racial/ethnic composition of a school may partly explain the association between individuals’ race/ethnicity and PA. The study by Richmond et al4 also reported that school-level median household income was positively associated with PA, and individual household income was not associated with PA after controlling for school-level income. However, study by Richmond et al4 used self-reported PA data which are subject to reporting bias for a child population.11 In addition, few studies have addressed whether PA differences by school racial/ethnic composition are present, particularly during school hours. Improved understanding of school-level attributes associated with student PA during school hours would inform school-based PA promotion strategies for school-aged children. Illinois Suburban Cook County (SCC) received Communities Putting Prevention to Work (CPPW) funding for obesity prevention and control efforts focused on policy, systems, and environmental change.12 To help guide the SCC CPPW school-based obesity prevention strategies for school-aged children, we evaluated accelerometry-measured MVPA and inactivity time during school hours among fifth and sixth graders in SCC schools based on school-level race/ethnicity and income attributes.
METHODS Participants The target population was school-aged students in SCC public schools. SCC, which refers to Cook County excluding the City of Chicago, includes 144 public school districts with approximately 380,000 public school students. This report used baseline data from the SCC CPPW school evaluation. The evaluation sample consisted of fifth- and sixth-grade students attending 15 public elementary schools participating in the school-based evaluation component of the SCC CPPW program. SCC CPPW is described elsewhere.12 Of the 15 participating schools, 10 schools were randomly selected from a stratified list (by geographical location and income level) of 21 PA-focused intervention schools serving fifth- and/or sixth-grade students. Each of these 10 schools consented to participate in the evaluation (0% refusal rate). A total of 20 comparison public elementary schools were selected from a sampling frame of 370 nonintervention public schools (schools not implementing a SCC CPPW PA intervention) with grades 5 and/or 6 in SCC to match with the intervention schools based on the Journal of School Health
•
stratification factors. Of the 20 comparison schools selected, 5 consented to participate (overall 75% refusal rate). The student sample included students from 2 randomly selected fifth- and/or sixth-grade classrooms in each of the 15 participating schools. A total of 627 students consented to wear accelerometers at School Year (SY) 2010 to 2011 (89% student participation rate). Parent/guardian passive consent (assumed participation unless they opted out through completion of a declination form) and students’ verbal assent were obtained. A stipend was provided to participating schools. Instruments and Procedures School-level data. SY 2010 to 2011 school race/ ethnicity composition and the proportion of students entitled to free/reduced-price lunch were obtained from the Illinois Interactive Report Card website http://iirc.niu.edu/. School race/ethnicity was assigned according to the student population’s majority race/ethnicity: non-Hispanic white (≥70% White students), non-Hispanic black (≥70% Black students), Hispanic (≥70% Hispanic students), and no majority race/ethnicity (mixed). On the basis of proportion of students entitled to free/reduced-price lunch, schools were classified into 2 income categories: high-income (0% to 50% of students eligible for free/reduced-price lunch) and low-income (51% to 100% of students eligible for free/reduced-price lunch). Accelerometry data. Actigraph accelerometers (model GT1M and GT3X; Pensacola, FL) were used to measure student PA. The accelerometer movement count data epoch was set at 5 seconds. Two research assistants (RAs) visited sampled school classrooms during the period from April to June 2011. Classroom teachers and students were informed about the evaluation and accelerometry data collection. RAs directly put an accelerometer belt on students’ waists with the accelerometer resting on the right hip on the first measurement day (usually Mondays). Students were asked to wear the accelerometers during school hours for 4 consecutive days. At the conclusion of each school day, students returned their assigned accelerometer to the teacher. Upon the start of school the following day, teachers assisted students in putting on their assigned accelerometer. RAs revisited the classrooms to collect accelerometers on the last measurement day (usually Thursdays). Records of school start time and end time on each of the accelerometer measurement days were obtained from classroom teachers. Accelerometry Data Reduction We only used accelerometry data recorded during a period of start time to end time of the school day. An accelerometer was considered as not worn if a
June 2015, Vol. 85, No. 6
•
© 2015, American School Health Association • 383
Table 1. Majority Race/Ethnicity and Income Level of 14 Participating Schools
Percentage of majority race/ethnicity (mean± SD) Income categories† High (N) Low (N)
Non-Hispanic White (N = 4)
Non-Hispanic Black (N = 2)
Hispanic (N = 5)
No Majority Race/ Ethnicity (N = 3)
73.5± 13.1
74.9± 0.1
95.6± 1.0
—∗
4 0
0 2
0 5
3 0
∗ Three schools had no single race/ethnicity that made up >70%. The means of the percentages of non-Hispanic Whites, Asians, and Hispanics were 38.6%, 31.5%, and 23.1%,
respectively. † On the basis of percentage of students entitled to free/reduced-price lunch, schools were classified into 2 income categories: 51-100% = low-income and 0-50% = high
income.
period of 60 minutes of consecutive zeros, allowing for 2 minutes of nonzero interruptions, was encountered anywhere in the data array. Accelerometry data were considered valid if wear time was 6 hours per school day for ≥3 wear days. MVPA was defined as ≥2296 accelerometer counts per minute.13,14 Inactivity was defined as
