11.06.2013 Views

Thursday-Abstracts

Thursday-Abstracts

Thursday-Abstracts

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Thursday</strong>, May 30, 2013<br />

S274 Vol. 45 No. 5 Supplement<br />

indoor, home/outdoor, transportation, and community). The proportion of time spent<br />

in moderate PA (MPA) and vigorous PA (VPA) was segmented by both purpose and<br />

location to understand the context of PA behavior.<br />

rEsuLTs: Approximately 83% of participants reported participating in some MPA<br />

(3-6 METS) on the previous day (mean = 156.4 +/- 32.6 min) while 21% reported<br />

participating in some VPA (&gt; 6 METS) on the previous day (mean = 67.1 +/- 9.1<br />

min). Participants reported that they performed 42% of their MPA while working,<br />

37% for home needs, 14% for leisure, 5% for sports, and about 2% for volunteer<br />

/ education tasks. Corresponding allocations for VPA were 33%, 27%, 14%, 25%,<br />

and 1%, respectively. The majority of MPA was performed at work (41%) followed<br />

by home indoor (26%), community settings (18%), home outdoor (13%), and in<br />

transportation (2%). The majority of VPA was also performed at work (33%),<br />

followed by community settings (32%), home indoor (21%), home outdoor (13%) and<br />

transportation (1%). Differences in allocations for purpose and location were noted<br />

when the sample was stratified by gender, age, and education<br />

CONCLusION: The results demonstrate the diverse nature of PA behavior and reveal<br />

that the majority of MPA and VPA is accumulated as part of daily living rather than for<br />

leisure or recreation. These results provide new insights about the settings and context<br />

of PA behavior in adults.<br />

1433 Board #25 May 30, 2:00 PM - 3:30 PM<br />

Pedometer accuracy in Lower-Limb amputees<br />

John D. Smith 1 , Gary Guerra 2 . 1 Texas A&M University-San<br />

Antonio, San Antonio, TX. 2 Loma Linda University, Loma Linda,<br />

CA.<br />

(No relationships reported)<br />

The accuracy of the pedometer on lower limb amputees has yet to be explored.<br />

PurPOsE: To assess the accuracy of pedometers in lower limb amputees.<br />

METhOds: Thirty-nine participants (age = 48.0±14.1 yrs, ht = 172.7±9.3 cm, wt<br />

= 76.2±14.8 kg) had their height and weight measured and were then fitted with a<br />

pedometer belt around the waistline at the hip. They were fitted with three pedometers:<br />

a SW-701 spring-levered pedometer on the anterior mid-line of the right hip, an NL-800<br />

pedometer just laterally to the SW-701, and an Omron HJ-112 on the outside of the NL-<br />

800. They then engaged in one over-ground level walk for 200m at a self-selected pace at<br />

an indoor facility. At the end of each trial pedometer counts and actual step counts from<br />

a hand tally (AC) were recorded. A repeated measures ANOVA was used to determine<br />

significant differences between the counts. Single measure intraclass correlation (ICC)<br />

from a two-way random effects ANOVA was used to assess the agreement between<br />

AC and pedometer counts. Bland-Altman plots of AC vs. pedometer counts were also<br />

used to provide an indication of agreement between the measures. Pedometer error was<br />

calculated as [(pedometer steps-actual steps)/actual steps] * 100.<br />

rEsuLTs: Counts from the SW were significantly lower than AC (255.6±12.6 and<br />

314.7±6.0 steps, p = .001) and counts from the NL and Omron were not significantly<br />

different from AC (p > .05). Agreement according to ICC was very low (0.26) for the<br />

SW, low for the NL (0.71) and high for the Omron (.98). Bland-Altman plots indicated<br />

agreement was strongest with the Omron and weakest with the SW. Percent error was<br />

greatest with SW at 18.0%, 2.1% for the NL, and least in the Omron (0.7%).<br />

CONCLusION: The results indicate that of the three pedometers tested, the Omron<br />

HJ-112 is the best pedometer for a person with a lower limb amputation wearing a<br />

prosthesis.<br />

1434 Board #26 May 30, 2:00 PM - 3:30 PM<br />

association Between Cardiorespiratory Fitness and<br />

Metabolic risk Factors: an application of Quantile<br />

regression<br />

Wonwoo Byun1 , Jonathan A. Mitchell2 , Mitchell H.<br />

Whaley, FACSM1 , Leonard A. Kaminsky, FACSM1 . 1Ball State University, Muncie, IN. 2University of Pennsylvania,<br />

Philadelphia, PA.<br />

(No relationships reported)<br />

Higher cardiorespiratory fitness (CRF) has been associated with improved metabolic<br />

risk factors (MRFs). However, the evidence that exists largely assessed the mean<br />

values of MRFs and it is not known if the association between CRF and MRFs is<br />

uniform across the distributions of individual MRFs.<br />

PurPOsE: To determine the association between CRF and MRFs, using quantile<br />

regression, in a cross-sectional sample of men and women.<br />

METhOds: Participants were 429 adults (Female: 51%, Age: 49 ± 13 yrs), who<br />

enrolled in Adult Physical Fitness Program at Ball State University. CRF was assessed<br />

via maximal exercise testing with measured VO max (ml/kg/min). Participants were<br />

2<br />

also screened for fasting glucose (GLU), triglycerides (TG), high-density lipoprotein<br />

cholesterol (HDL-C), mean arterial blood pressure (MAP), and waist circumference<br />

(WC). Quantile regression was used to examine the associations between CRF<br />

and MRFs at the 10th , 25th , 50th , 75th , and 90th percentiles of each MRF. The models<br />

were adjusted for age, gender, education level, smoking status, and medications for<br />

hypertension, diabetes, and dyslipidemia.<br />

rEsuLTs: CRF was inversely associated with GLU at the 50th (β= -0.34, 95% CI:<br />

MEDICINE & SCIENCE IN SPORTS & EXERCISE ®<br />

-0.52, -0.17), 75th (β= -0.53, 95% CI: -0.74, -0.32), and 90th (β= -0.70, 95% CI: -0.99,<br />

-0.40) percentiles; and inversely associated with MAP at the 50th (β= -0.21, 95% CI:<br />

-0.35, -0.07), 75th (β= -0.25, 95% CI: -0.44, -0.06), and 90th (β= -0.29, 95% CI: -0.49,<br />

-0.09) percentiles. Significant inverse associations between CRF and TG and WC were<br />

found at the 10th, 25th, 50th, 75th, and 90th percentiles; the magnitude of associations<br />

were gradually stronger at upper tails of TG distribution (from 10th percentile β=<br />

-1.14, 95% CI: -1.62, -0.66 to 90th percentile β= -3.74, 95% CI: -6.75, -0.72) and WC<br />

distribution (from 10th percentile β= -0.72, 95% CI: -1.10, -0.33 to 90th percentile β=<br />

-1.25, 95% CI: -1.48, -1.02). CRF was positively associated with HDL-C at 25th (β=<br />

0.33, 95% CI: 0.21, 0.45), 50th (β= 0.28, 95% CI: 0.11, 0.45), 75th (β= 0.39, 95% CI:<br />

0.07, 0.71), and 90th (β= 0.43, 95% CI: 0.09, 0.85) percentiles.<br />

CONCLusION: CRF was significantly associated with better MRFs profiles.<br />

The associations were stronger at upper tails of MRFs distributions indicating that<br />

improving CRF is especially important for those with poor MRF profiles.<br />

1435 Board #27 May 30, 2:00 PM - 3:30 PM<br />

Type I and Type II Errors in Correlation analyses of Various<br />

sample sizes<br />

Duane Knudson, FACSM1 , Crawford Lindsey2 . 1Texas State<br />

University, San Marcos, TX. 2Tennis Warehouse, San Louis<br />

Obispo, CA.<br />

(No relationships reported)<br />

Studies are designed to balance pragmatic and theoretical issues. A common problem<br />

in sport and exercise science research is small sample sizes, which was recently noted<br />

in ACSM abstracts (Pyne & Hopkins, 2012).<br />

PurPOsE: This study documented the likely rates of type I and type II errors in<br />

correlation analyses in sample sizes common in sport and exercise science.<br />

METhOds: Data from mechanical testing of most all polyester tennis strings<br />

commercially available in 2012 using the protocol by Cross et al. (2000) were analyzed<br />

in this study. Variables measured were gauge, impact duration, string deflection, peak<br />

force, string stiffness, and energy return for 198 different strings tested at 276 N of<br />

tension. Three additional sample sizes were created by randomly down sampling to 99,<br />

50, and 25. Correlations and partial correlations were calculated and compared to those<br />

observed for the complete sample (N=198).<br />

rEsuLTs: For the whole sample, 14 of 15 zero-order correlations were significant<br />

(p

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!