RACE DIFFERENCES IN REPRODUCIBILITIES - THE HERITAGE FAMILY STUDY

Citation
He. Elmoalem et al., RACE DIFFERENCES IN REPRODUCIBILITIES - THE HERITAGE FAMILY STUDY, American journal of human biology, 9(4), 1997, pp. 415-424
Citations number
19
Categorie Soggetti
Anthropology,Biology
ISSN journal
10420533
Volume
9
Issue
4
Year of publication
1997
Pages
415 - 424
Database
ISI
SICI code
1042-0533(1997)9:4<415:RDIR-T>2.0.ZU;2-Y
Abstract
The HERITAGE (HEalth, RIsk factors, exercise Training And GEnetics) Fa mily Study is a multicenter clinical trial conducted by five instituti ons in the United States and Canada, The overall objective of the stud y is to investigate the role of the genotype in cardiovascular, metabo lic, and hormonal responses to aerobic exercise training and the contr ibution of regular exercise to changes in cardiovascular disease and d iabetes risk factors in white and black families. Since the accuracy o f the assessment of the response to training depends on how repeatable or reproducible the measurements are, it is important to assess poten tial racial differences in reproducibilities, which may have implicati ons for pooling data across races, The sample studied consisted of 96 blacks and 304 whites, The black sample had 46 males with mean age 33. 6 +/- 14.2 years and 40 females with mean age 33.9 +/- 12.7 years, The white sample had 152 males with mean age 35.5 +/- 14.9 years, and 152 females with mean age 34.9 +/- 14.3 years. Reproducibilities, as meas ured by intraclass correlations among repeated measures, were comparab le between whites and blacks for variables in the anthropometry, i.e. lipid, exercise test, and blood pressure domains. Reproducibilities in both races exceeded 0.85 for most of the variables, When the within-r ace reproducibilities are very high, statistical significance of any o bserved racial difference in the reproducibilities may not be very mea ningful. There was a significant racial difference in the reproducibil ity for Apoprotein Al (0.73 in blacks, 0.89 in whites, P < 0.01). Howe ver, this is not a cause for concern, since only one among 37 comparis ons was significant. (C) 1997 Wiley-Liss, Inc.