Background-Passive smoking is associated with early arterial damage in adults, but its effect on endothelial function in children is unknown. Methods and Results-Serum cotinine concentration was measured annually in children between 8 and 11 years of age who had participated since infancy in a randomized, prospective atherosclerosis prevention trial (Special Turku Coronary Risk Factor Intervention Project for children [STRIP]). At age 11, endothelium-dependent flow-mediated vasodilatory responses of the brachial artery were examined with high-resolution ultrasound in 402 children. These children were divided into 3 groups according to serum cotinine concentrations: the noncotinine group (nondetectable cotinine, nϭ229), the low cotinine group (cotinine between 0.2 and 1.6 ng/mL, nϭ134), and the top decile cotinine group (cotinine Ն1.7 ng/mL, nϭ39). Longitudinal cotinine data in children aged 8 to 11 years and ultrasound studies were available in 327 children. At age 11, the increase in cotinine concentration was associated with attenuated peak flow-mediated dilation response (meanϮSD: the noncotinine group 9.10Ϯ3.88%, the low-cotinine group 8.57Ϯ3.78%, and the top-decile cotinine group 7.73Ϯ3.85%; Pϭ0.03 for trend). Similarly, total dilation response (the area under the dilation response versus time curve between 40 and 180 seconds after hyperemia) was affected by the cotinine level (Pϭ0.02 for trend). These trends were not explained by traditional atherosclerosis risk factors. Arterial measures and passive smoking showed even stronger associations when longitudinal cotinine data were used (peak flow-mediated dilation, Pϭ0.01 for trend; total dilation response, Pϭ0.008 for trend). Conclusions-Exposure

To characterize brachial artery flow-mediated dilatation (FMD) in children, we monitored arterial diameter changes with ultrasound between 40 and 180 s after a 4.5-min forearm cuff occlusion-induced hyperemia in 105 healthy children (mean age, 11 yr; range, 9-16 yr). The peak FMD was 7.7 +/- 4.0% and occurred 79 +/- 33 s after cuff release. FMD at 60 s (5.3 +/- 4.0%) was significantly lower than the peak FMD (P < 0.0001). Twenty-three percent of the children (n = 24) reached peak FMD first after 110 s of postocclusion. Compared with others, these late responders weighed less, had smaller vessel size, and were more often girls, but had similar peak FMD. In multivariate analysis, FMD responses were inversely associated with brachial artery baseline diameter and serum cholesterol concentration. We conclude that the time to reach the peak FMD response in children varies considerably. When studying endothelial function in children with the use of the noninvasive ultrasound method, several brachial artery diameter measurements up to 120 s after cuff release are needed to determine the true FMD peak response.