A lthough sympathetic stressors can trigger acute coronary events,1,2 the growth of coronary thrombosis depends on hypercoagulable conditions.3 Psychological stress and catecholamine surge may affect regular hemostatic activity.4,5 We investigated the effects of standardized mental stress in a group of elderly subjects on markers of a hypercoagulable state—thrombin-antithrombin III complex (TAT) and fibrin D-dimer (DD)—and on the von Willebrand factor (vWF), a marker of endothelial dysfunction in cardiovascular disorders.6,7 To verify that the stressor provoked sympathetic activation, we also measured hemodynamics and plasma catecholamines. • • • The study participants were 53 community-dwelling elderly spousal caregivers (35 women and 18 men; mean age 73 years, range 59 to 82) of patients with Alzheimer’s disease participating in a longitudinal study of biologic and psychosocial adaptation to caregiving.8 Data reported in this report were obtained at study entry. Participants were recruited through community support groups or referred by the University of California, San Diego Alzheimer Disease Research Center or their treating physicians. The University of California, San Diego institutional review board approved the protocol; all subjects gave written consent. A research nurse obtained a structured medical history including information on coronary artery disease and cerebrovascular disorders (referred to hereafter as “vascular disease”), systemic hypertension, and type II diabetes previously diagnosed by a physician. The nurse also obtained information on cardiovascular risk factors (former/current smoking, hypercholesterolemia [positive self-report of “elevated blood cholesterol”], and overweight [body mass index .25 kg/m) in women and .26.5 kg/m in men9 as well as on the current use of nonsteroidal anti-inflammatory and antihypertensive drugs. Based on the history it was possible to classify subjects (“yes/no”) with respect to several variables (i.e., cardiovascular disorders, risk factors, drugs) that could influence hemostasis. Subjects taking b blockers and oral anticoagulants were excluded. Subjects were tested between 8:00 and 10:00 A.M. in their homes. After the research nurse had inserted an indwelling 22-gauge venous forearm catheter, the subject sat quietly for 15 minutes to acclimatize. Thereafter, subjects engaged in 2 standardized speech tasks about being wrongly accused of shoplifting and about an upsetting situation related to their caregiving experience. The order of exposure to the 2 tasks was randomly assigned. Both tasks were structured into a 5-minute instruction phase immediately followed by a 3-minute preparation phase and a 3-minute speaking phase. Heart rate and blood pressure (Finapres Ohmeda 2300, Louisville, Colorado) as well as blood samples for hemostasis and catecholamine measurements were obtained immediately before the instruction phase of the first task and in the middle of the second talking phase. Plasma epinephrine and norepinephrine levels were not available from 9 subjects. Venous blood was drawn and processed as per standard techniques. Plasma TAT, DD, and vWF levels were measured by enzyme-linked immunosorbent assays (TAT, Enzyme Research Lab [Dade Behring Inc., Deerfield, Illinois]; DD, Diagnostica Stago; and vWF [American Biproducts Co., Parsippany, New Jersey] as described by Short et al10). Antibodies for vWF antigen assay were from DAKO Corp. and substrate was from Bio-Rad Lab. Interassay coefficients of variation for vWF antigen, TAT, and DD were 4.6%, 5.9%, and 5.5%, respectively. The respective intra-assay coefficients of variation were 2.6%, 4.9%, and 2.2%. Plasma epinephrine and norepinephrine levels were measured by a radioenzymatic method.11 Data were analyzed using SPSS, Version 9.0 statistical software (SPSS Inc., Chicago, Illinois). Results were considered statistically significant at the p #0.05 level; all tests were 2-tailed. To approximate a normal distribution, hemostasis and catecholamine measurements were log10 transformed. We report original units (median, interquartile range) throughout. Student’s t test, and simple correlation and univariate analyses were used. To test for differences in stressFrom the Departments of Psychiatry and Medicine, University of California, San Diego; and Veterans Affairs San Diego Healthcare System, La Jolla, California. This report was supported by Grants AG15301, HL36005, and RR00827 from the National Institutes of Health, Bethesda, Maryland, to Drs. Grant, Dimsdale, and Ziegler, by fellowship 81BE-56155 from the Swiss National Science Foundation, and by an educational grant from Novartis Foundation, Bern, Switzerland, to Dr. von Kanel. Dr. Grant’s address is: Department of Psychiatry, UCSD, 9500 Gilman Drive, La Jolla, California 92093-0680. E-mail: igrant@ ucsd.edu. Manuscript received November 28, 2000; revised manuscript received and accepted January 23, 2001.