SERUM CAROTENOID CONCENTRATIONS AND THEIR REPRODUCIBILITY IN CHILDRENIN BELIZE

Citation
J. Apgar et al., SERUM CAROTENOID CONCENTRATIONS AND THEIR REPRODUCIBILITY IN CHILDRENIN BELIZE, The American journal of clinical nutrition, 64(5), 1996, pp. 726-730
Citations number
25
Categorie Soggetti
Nutrition & Dietetics
ISSN journal
00029165
Volume
64
Issue
5
Year of publication
1996
Pages
726 - 730
Database
ISI
SICI code
0002-9165(1996)64:5<726:SCCATR>2.0.ZU;2-O
Abstract
Suggestions that carotenoid-containing foods are beneficial in maintai ning health have led to Several studies of circulating carotenoid conc entrations of adults. Because few data are available for children, we report serum carotenoid concentrations of 493 children in Belize. Caro tenoid concentrations were determined as part of a survey of vitamin A status of children, most between 65 and 89 mo of age. Reproducibility was tested by collecting a second blood sample 2 wk after the first c ollection from a subset of children (n = 23) who consumed their habitu al diet with no treatment during the interim. Predominant serum carote noids were lutein/zeaxanthin and beta-carotene; which ac counted fur 2 6% and 24% of median total carotenoids, respectively. The three provit amin A carotenoids, alpha- and beta-carotene and beta-cryptoxanthin, c onstituted 51% of median total carotenoid concentrations. Partial corr elations of each carotenoid with lasting retinol concentration indicat ed that beta-carotene had the highest correlation. Concordance correla tion coefficients (r(c)) for lasting carotenoid concentrations determi ned 2 wk apart were greater than or equal to 0.89 for lycopene, beta-c ryptoxanthin, and alpha- and beta-carotene. The r(c) for lutein/zeaxan thin and total carotenoids was lower, 0.59 and 0.68 respectively, beca use of higher lutein/zeaxanthin concentrations at the second sampling than at the first. The reproducibility of the concentrations suggests both that individuals have characteristic profiles and that serum caro tenoid concentrations ran be measured randomly over greater than or eq ual to 2 wk without significant bias.