ORIGINAL ARTICLE Acute Changes in Vascular Function Among Welders Exposed to Metal-Rich Particulate Matter Shona C. Fang,* Ellen A. Eisen,* Jennifer M. Cavallari,* Murray A. Mittleman,†‡ and David C. Christiani*§ Background: Although welding fume exposure is associated with adverse cardiovascular outcomes, the mechanisms remain unclear. To investigate the role of vascular function, we assessed levels of the augmentation index (a correlate of arterial stiffness) after short-term exposure to welding-derived fine particulate matter (PM 2.5 ). Methods: In a panel study, we monitored 26 male welders over 24 hours on a welding day (n = 25), a nonwelding day (n = 15), or both (n = 14). Augmentation index (expressed as a percent) was obtained in the morning before exposure (baseline) and after expo- sure in the afternoon and the following morning. Personal PM 2.5 exposure was measured over 6 hours of welding or an equivalent nonwelding period. We used linear mixed models adjusting for baseline augmentation index, smoking, age, and time to evaluate the effects of welding (binary) and PM 2.5 (continuous) on augmentation index levels. We also assessed modification by welding exposure the day before monitoring (binary). Results: Welding was associated with a 2.8% increase in afternoon augmentation index (95% confidence interval =-1.4 to 7.0) and a 2.4% decrease (-6.9 to 2.2) in next-morning augmentation index. Additional exposure the day prior to monitoring was associated with a greater afternoon increase (5.1%; 0.8 to 9.5). Using PM 2.5 con- centration, a positive association was observed in the afternoon and an inverse association the next morning; results differed by previous day’s welding status after excluding outliers. Conclusions: Subsequent to welding fume exposure, there is an increase in afternoon augmentation index and a decrease in next- morning augmentation index, with greater changes after consecutive days of exposure. These results suggest that exposure to the PM 2.5 component of welding fume elicits acute adverse vascular responses. (Epidemiology 2008;19: 217–225) W elding fume is a complex mixture containing metal- rich airborne particulate matter (PM), generated when heated metal vaporizes and condenses in air. Exposure in- cludes high concentrations of respirable particles with an aerodynamic diameter 2.5 m (PM 2.5 ). Although the pul- monary health effects of welding fume exposures are well- known, 1 there is growing concern for the cardiovascular health effects. Several mortality studies report increased risk of death from ischemic heart disease among welders. 2–4 In addition, the prevalence of myocardial infarction and other cardiovascular events and symptoms are increased in weld- ers. 5 The underlying mechanisms responsible for these car- diovascular effects are largely unclear; however, a role for vascular function may be involved. 6,7 Several studies have demonstrated vasoconstriction and increases in blood pres- sure after short-term exposure to ambient PM. 8 –12 Vascular changes may be mediated by direct action of PM on lung receptors 13 or via inflammatory pathways. 14 In observational studies, vasoconstriction has been measured by vasoreactivity and brachial artery diameter. 8,9 These measurements are not feasible in workplace studies where practical restrictions warrant the use of simpler tech- niques. One component of vascular function that is easily measured with a portable device is the aortic augmentation index, a measure of wave reflections that relates to central aortic pulse pressure, and is a correlate of arterial stiffness. 15 It is derived from reflective properties of the arterial bed and is measured reproducibly 16,17 by the noninvasive technique of pulse wave analysis. Increased augmentation index is associated with advancing age, hypertension, diabetes, and hypercholesterolemia. 18 –20 The measurement is also a strong and independent marker of established coronary artery dis- ease 21 and correlates with risk of coronary artery disease development. 22 Acute increases in the augmentation index are observed in individuals after exposure to first- and sec- ond-hand cigarette smoking. 23,24 However, the relation be- tween the augmentation index and metal-rich occupational particulate exposures has not been studied. We conducted a panel study to investigate the relation- ship between short-term exposure to metal-rich PM 2.5 and the augmentation index over a 24-hour period in a group of welders on either or both a welding and a nonwelding day. The augmentation index was evaluated with respect to weld- Submitted 22 March 2007; accepted 23 August 2007; posted 29 January 2008. From the Departments of *Environmental Health, †Epidemiology, Harvard School of Public Health, ‡Cardiovascular Epidemiology Research Unit, Beth Israel Deaconess Medical Center, and §Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts. Supported by NIEHS grants ES009860 and ES00002 and a pilot project grant awarded under the Harvard-NIOSH Education and Research Center grant T42 OH008416. Supported by NIOSH training grant T42 OH008416 (to S.F. and J.C.) and NIEHS training grant T32 ES007069 (to J.C.). Correspondence: David C. Christiani, Environmental and Occupational Medicine and Epidemiology Program, Department of Environmental Health, Harvard School of Public Health, 665 Huntington Avenue, Building I-Room 1402, Boston, MA 02115. E-mail: dchris@hsph. harvard.edu. Copyright © 2008 by Lippincott Williams & Wilkins ISSN: 1044-3983/08/1902-0217 DOI: 10.1097/EDE.0b013e31816334dc Epidemiology • Volume 19, Number 2, March 2008 217