determinants of physical activity in rural, african american adolescents. mike bamman, phd
TRANSCRIPT
Determinants of Physical Activity in Rural, African American Adolescents.
Mike Bamman, PhD
Health risks in children and adolescents - overweight and obesity 13% of adolescents obese, 15% overweight
(CDC, 2008)
60% of overweight (<85%, BMI) children had at least one cardiovascular risk factor compared to 10% of normal weight. 25% had two or more risk factors. (Freedman, Dietz, Srinivasan, and
Berenson, 1999)
Independent predictor for developing HTN, NIDDM, CHOL, beginning in childhood (Berenson et al., 1998)
Health risks in children and adolescents HTN – Obese children have a 9x risk (Lauer, Connor, & Leaverton,
1975)
NIDDM – Relationship between PA and insulin resistance in obese youth. (Pinhas-Hamiel, Dolan, Daniels, Standiford, Khoury, & Zeitler, 1996; Schmitz, Jacobs, Hong, Steinberger, Moran, & Sinaiko, 2002; Quarry-Horn, et al, 2003).
CHOL 90 percent of the children with elevated TG were also
overweight. A strong positive relationship exists between physical
activity (PA) levels and HDL cholesterol. (Armstrong & Simmons-Morton,
1994; Riddoch & Boreham, 1995; Freedman, Dietz, Srinivasan, and Berenson, 1999; Twisk, 2000).
Adolescent PA levels (2007 YRBS)
Nationwide adolescent PA levels Only 37% of students grade 9-12 meet PA guidelines Over 22% of students are sedentary Only 30% (32%, 2003) receive daily physical education
Mississippi adolescent PA levels 36% meet PA guidelines 23% are sedentary 23% receive daily PE
Determinants of PA in adolescents Self efficacy – Reynolds, et al, 1990;Zakarian, Hovell, Hofstetter,
Salles, & Keating, 1994 & Trost, et al, 1997; Dwyer et al, 1998; Allison, Dwyer, Makin, May 1999.
Barriers to PA - Allison, Dwyer, Makin, August 1999
Sex differences - Aaron, et al, 1993; Garcia, et al, 1995; Trost, et al, 2000
Age differences - Aaron, 1993; Saris, Elvers, Van’t Hof & Binkhorst, 1986; Verschuur & Kemper, 1985; Allison, Dwyer, & Makin, May, 1999; Trost, et al, 2002
BMI - Berkowitz, 1984
Purpose
The purpose of this study was to identify the determinants of PA in African-American schoolchildren ages 12-19 in Tunica County, Mississippi. A secondary function of this study was to determine the relationships among and identify any differences between recognized determinants of PA (body mass index, age, perceived barriers to exercise, and self-efficacy) of the students.
Significance
The results of this study will allow researchers, school officials, community leaders, and parents to better understand the possible determinants of PA in African-American adolescents in Tunica County, Mississippi. This research will also make a significant contribution to the existing literature relating to determinants of PA in African American adolescents
Tunica County, MS (2000, US Census)
Predominantly African American - 70%
Tunica County US
Median Household Income (per year) $19,000 $37,000
Poverty (% of families living at or below the HHS poverty threshold)
27% 13%
Undereducated (% of persons >25 yrs with a HS diploma)
24% 54%
Unemployed (February, 2009) 17.8% 8.9%
Hypotheses
To address the purpose of the study the following null hypotheses were tested.
There is no significant relationship between level of PA and perceived self-efficacy, perceived barriers to exercise, age, and BMI for males.
There is no significant relationship between level of PA and perceived self-efficacy, perceived barriers to exercise, age, and BMI for females.
Instrumentation Physical Activity
MTI (Manufacturing Technology Inc) accelerometer (Trost, Pate, Freedson, Sallis & Taylor, 2000)(r = 0.87)
7-day physical activity recall – PAR (Wallace, McKenzie & Nader, 1985; Sallis, Buono, Roby, Micale & Nelson, 1993; Blair, et al, 1998; Dunn, et al, 1997)(r = 0.81)
Self-efficacy (SE)– subset of SAHHS questionnaire (Reynolds, et al, 1990)(r = 0.89)
Barriers to PA – Campbell’s well-being questionnaire (Stephens, Craig, 1990; Allison, Dwyer & Makin, 1999)(r = 0.76)
BMI, Age
Procedures Permission – IRB (#04-081), Tunica County
Schools, NMDBGC Session 1 - Program description, handouts
(program, MTI), parental consent form Session 2 - Student verbal/written assent, study
procedures review, instruments (medical history/demographic data form, SAHHS, Campbell, BMI), MTI and log book, contact information
Intermediate - Phone calls, incentive distribution Session 3 - MTI and logbook collection, feedback
information, incentive distribution
Statistics
Means and standard errors Correlation of PA instruments Stepwise multiple regression
Equations for both males and females Chow test Alpha level set at 0.05 a priori
Subjects AA Adolescents - Tunica County, MS
Age 12-19; Grades 8-12 (975 enrolled)
Rosa Fort High School Physical Education Classes (Total Enrollment 168 – 5 classes)
North MS Delta Boys and Girls Club (Total Enrollment 78)
Sample size: > 40 males and >40 females Based upon power analysis = alpha 0.05, power of 0.80 and
an effect size of 0.25 Total Students Eligible - 246 Total Parental Consents Returned - 172 Total Subjects Initiating Protocol - 141 Total Subjects Completing Protocol – 84
Females n = 43; Males n = 41
Results – DescriptivesMean SE Min Max
M F M F M F M F
Age 15.4 16.2 0.25 0.21 12 14 18 19
BMI 24.9 25.1 0.93 0.97 16.6 18.2 46.4 54.8
SE 32.9 31.7 1.31 1.32 8 8 48 48
Barriers to PA 38.8 47.3 2.17 1.47 15 28 70 66
PAR 6894 6303 328.2 351.2 3645 3548 12942 14046
MTI 4593 2818 410.8 182.7 1148 1127 14002 6352
males n = 41, females n = 43
Results – FrequenciesFrequency Percent of Total
Free Lunch F M F MYes 35 29 81.4 70.7
No 8 12 18.6 29.3
Current Sport F M F MYes 0 17 0 41.5
No 43 24 100.0 58.5
Usual Act. Level F M F MMore 9 10 20.9 24.4
Less 11 6 25.6 14.6
About the same 23 25 53.5 61
females n = 43, males n = 41
Correlation of PA measures
Instrument N Mean SE Pearson r
PAR (kcal/wk) 84 6,592 241.6 0.90a
PAR (METs/wk) 84 10,137 382.9
MTI (kcal/wk) 84 3,685 240.5 0.67a
MTI (METs/wk) 84 5,582 367.7
ap<0.05
Results – Validation of PA measures
Instrument N Mean SE Pearson r
PAR (kcal/wk) 84 6892 241.54 0.57a
MTI (kcal/wk) 84 3685 240.47
ap<0.05
Results – Correlations, FemalesAge BMI SAHHP Barriers
to PAPAR MTI
Age 1.0 0.372a -0.056 0.070 0.101 0.151
BMI 0.372a 1.0 0.148 0.154 0.508a 0.203
SE -0.056 0.148 1.0 0.179 0.411a 0.137
Barriers to PA
0.070 0.154 0.179 1.0 0.241 0.064
PAR 0.101 0.508a 0.411a 0.241 1.0 0.412a
MTI 0.151 0.203 0.137 0.064 0.412a 1.0
n = 43; ap<0.05
Results – Regression, Females βa βb R2 F p
Step 1
BMI 0.508 184.39 0.258 14.23 0.001
Step 2
βa βb R2 F p R2∆
BMI 0.457 165.98
SE 0.343 91.54 0.373 11.89 0.001 0.115
n = 43; aStandardized Beta Coefficient; bUnstandardized Beta Coefficient
Results – Correlation, MalesAge BMI SAHHP Barriers
to PAPAR MTI
Age 1.0 0.289 -0.014 0.226 0.426a 0.387a
BMI 0.289 1.0 0.085 -0.137 0.543a 0.360a
SE -0.014 0.085 1.0 0.050 0.102 0.096
Barriers to PA
0.226 -0.137 0.050 1.0 -0.005 0.093
PAR 0.426a 0.543a 0.102 -0.005 1.0 0.735a
MTI 0.387a 0.360a 0.096 0.093 0.735a 1.0
n = 41; ap<0.05
Results – Regression, Males βa βb R2 F p
Step 1
BMI 0.543 192.55 0.295 16.31 0.000
Step 2
βa βb R2 F p R2∆
BMI 0.458 162.44
Age 0.294 394.56 0.347 11.349 0.000 0.097
n = 41; aStandardized Beta Coefficient; bUnstandardized Beta Coefficient
Results – Chow test formula
Used traditionally in economics. Assesses the equality between sets of coefficients in two
linear regression equations.
F = [SSe(M/F)a – SSe(M)b – SSe(F)c ] / pd
[SSe(M)b + SSe(F)c ] / (ne +mf -2pd)
aStandard regression analysis, both groups bStandard regression analysis, males cStandard regression analysis, females dParameters (#IV +1) eFemale n fMale n
Results – Chow test
Group R2
Standardized Beta Coefficients
Chow Test
FcvAge BMI SE Barriers
to PA
Males 0.38 0.30a 0.45a
Females 0.39 0.47a 0.32a 1.849 2.37
Males n= 41; females n = 43ap<0.05
Discussion – Female PA
Significant Predictors of PA level BMI accounts for 26% of the variance in PA
level SE and BMI account for 34% of the variance in
PA level.
PARBMI R2 = 0.26
BMI
PARSE
R2 = 0.34
Discussion – Females
Nonsignificant results were found for PAR and barriers to PA (r = 0.101, p = 0.518)
Barriers to PA Found to be predictive in 9th grade female students.
Expected r is negative. Specific barriers – Internal vs External Recent research has assessed barriers in:
Urban setting Minorities – Hispanic and African American
Discussion – Females
Nonsignificant results were found for PAR and age (r = 0.241, p = 0.119)
Age Found to decrease sig. after 12 years. Expected r is
negative. 30% (6 of 18) of subjects ages 17-19 had PAR scores
at or above +1 SEM. The most active subjects were 17 and 18 years old
Discussion – Males PA
Significant Predictors of PA level BMI accounts for 29.5% of the variance in PA
level Age and BMI account for 34% of the variance
in PA level.
PARBMI R2 = 0.295
BMI
PARAge
R2 = 0.34
Discussion – Males
Nonsignificant results were found for PAR and SE (r = 0.102, p = 0.527)
Maximum possible range 8-48, both min. and max. score were reached
Outliers removed (5 subjects) - (r = 0.305, p<0.05)
Discussion – Males
Nonsignificant results were found for PAR and barriers to PA (r = -0.005, p = 0.977)
Barriers to PA Found to be predictive in 9th grade male students.
Expected r is negative. Specific barriers – Internal vs External Recent research has assessed barriers in:
Urban setting Minorities – Hispanic and African American
Discussion – Chow test Results
F (5, 74) = 1.85, Fcv = 2.37
Significance Males /Females – BMI (β= - 0.45, p<0.05; β = 0.47,
p<0.05) Males – Age (β = 0.49, p<0.05) Females – SE (β = 0.32, p<0.05)
Nonsignificance Males/Females – barriers to PA (β= -0.01, p = 0.92; β =
0.12, p = 0.37) Males – SE (β= 0.07, p = 0.61) Females – Age (β= -0.06, p = 0.65)
Conclusions
Females PA predicted by BMI and SE. Nonsignificant relationships found between PA
and barriers to PA and age. Ho1: There is no significant relationship between
level of PA and perceived self-efficacy, perceived barriers to exercise, age, and BMI for females.
Ha1: Females’ level of PA will be influenced by their level of perceived self-efficacy, followed by the perceived barriers to pa, followed by age, and BMI.
Conclusions
Males PA predicted by BMI and age. Nonsignificant relationships found between PA
and barriers to PA and SE. Ho2: There is no significant relationship between
level of PA and perceived self-efficacy, perceived barriers to exercise, age, and BMI for males.
Ha2: Males’ level of PA will be influenced by their level of perceived self-efficacy, followed by the perceived barriers to pa, followed by age, and BMI.
Recommendations
Random selection of subjects Interview process for 7-day PAR Additional measures assessing barriers to PA
and SE Comparison of internal vs external barriers to
PA