- Email: [email protected]
Temporal stability of acute stressor-induced changes in cellular immunity
INTERNATIONAL JOURNAL OF PSYCHOPHYSIOLOGY
ELSEVIER
International
Journal
of Psychophysiology
19 (1995) 287-290
Short Communication
Temporal stability of acute stressor-induced immunity Paul J. Mills Department
of Psychiatry, Received
*,
changes in cellular
Soraya L. Haeri, Joel E. Dimsdale
University of California, San Diego, 200 West Arbor Dr, La Jolla, CA 92103-0804, USA 6 December
1994; revised
14 February
1995; accepted
16 February
1995
Abstract This study examined the temporal stability of enumerative immune responses to acute psychosocial stress. Lymphocyte subsets were measured in 24 healthy male subjects at rest and following a speaking stressor on two occasions approximately six weeks apart. The speaking task caused significant increases in T-suppressor/cytotoxic cells, natural killer cells, T-cells, and total WBC and decreases in the T-helper/suppressor ratio. Baseline test-retest correlation’s were statistically significant for all variables (r values = 0.40-0.90). With two exceptions (T-cells and T-suppressor/cytotoxic cells), speaking task values (absolute reactivity scores) were also statistically significant (r values = 0.48-0.92). Baseline adjusted test-retest correlations were however generally less reliable, with only natural killer cells (r values > 0.40), the T-helper/suppressor ratio (r = 0.601, and total WBC (r = 0.48) showing statistical significance. The findings suggest that certain but not all cellular immune responses to acute stress are moderately stable over time. Keywords:
Immune
system;
Stress;
Temporal
stability
Research on the effects of acute stressors continues to be a focus of psychophysiologic research. An important consideration in the stress reactivity paradigm is the temporal stability of the stress response, i.e., whether a given individual’s pattern of reactivity represents a stable characteristic. Studies do support the notion that cardiovascular (Turner et al., 1986; Allen et al. 1987; Kasprowicz et al., 1990; Light, 1991; Durel et al., 1993) and hemodynamic (Sherwood et al., 1990) responses are stable over time. Evidence for tem-
* Corresponding 543-5462.
author.
Tel.:
(619)
0167-8760/95/$09.50 0 1995 Elsevier SSDI 0167-8760(95)00012-7
543-2506;
Science
Fax:
(619)
B.V. All rights
poral stability is of fundamental importance to the validity of the controversial notion that cardiovascular reactivity is a marker or risk factor for cardiovascular illness (Fredrikson and Matthews, 1990; Pickering and Gerin, 1990). Increasingly, studies are focusing on the responses of the immune system to acute behavioral stress. Like the cardiovascular system, the immune system is highly responsive to acute stress (for review see Kiecolt-Glaser et al., 1992; Herbert and Cohen, 1993). Given the growth of this research field, and the speculation that immune reactivity has possible clinical implications for long-term health, we thought it important to examine whether cellular immune responses to reserved
288
P.J. Mills et al. /International
Journal of Psychophysiology 19 (1995) 287-290
stress are also stable over time. To date, few if any reports have examined this issue. This study therefore, examined the temporal stability of enumerative immune responses to acute laboratory stress. Twenty-four healthy male volunteers were located through advertisement or word-of-mouth referral (mean age 30.2 years (SD 7); range 1943). Subjects were studied after obtaining written informed consent. The protocol was approved by the University of California, San Diego, Institutional Review Board. All subjects were studied at the same time of day (approximately 9:00 a.m.) on two separate occasions approximately 6 weeks apart. Subjects were instructed to eat a light breakfast and refrain from caffeinated beverages and smoking 12 h prior to study. Upon arrival to the laboratory, subjects were seated and a 19-gauge catheter was inserted into a forearm vein. Subjects then rested quietly for 30 min. The subjects were then given instructions for a speaking task. The task required preparing (3 min) and presenting (3 min) a speech in front of a video camera. Subjects were told that the speech would be evaluated and rated by experts. The topic of the task was different on each testing occasion, once involving defending oneself from being falsely accused of shoplifting and once dealing with a dishonest automobile service shop. If subjects stopped speaking before the 3 min were up, they were reminded to continue to talk by reiterating and summarizing their points. We have used this task in a different group of individuals in previous studies and it reliably results in significant cardiovascular and sympathetic nervous system activation (Nelesen et al., 1995). Following the resting baseline and at the end of the speaking task, 10 ml of blood was drawn for leukocyte determination. Flow cytometry (FAC-Scan, Becton Dickinson) on whole blood with CD45 gating was used to quantify white cell populations (Mills et al., 1995). We chose a panel of cells typically reported in studies examining the effects of acute psychologic stressors on cellular immune responses. The following cell types were determined (included are the antibodies employed): T-cells (CD3, anti-leu 41, T-helper cells (CD4,
P.J. Mills et al. /International
Journal of Psychophysiology
anti-leu 3a), T-suppressor/cytotoxic cells (CDS, anti-leu 2a), three natural killer (NK) cells, CD16 (anti-leu lib), CD56 (anti-leu 191, and CD57 (anti-leu 71, and total WBC. Blood was processed within 5 h of collection. Cells were detected by mouse monoclonal antibodies conjugated with either fluorescein or phycoerythrin. two-way repeated measures Separate ANOVAs (stress X session) were used to examine the effects of the stressor on cell population values. Pearson correlation coefficients were calculated between corresponding baseline and post-task measures across the two testing sessions. Previous research on temporal stability of reactivity has taken two approaches to examining change score reactivity (Manuck et al., 1989; Mills et al., 1993; Llabre et al., 1991). One approach utilizes an arithmetically-derived baseline-corrected value (simple change scores, task minus baseline) and the other utilizes a statistically-derived baseline-corrected value (residualized change scores), the latter being derived from the regression of the task values of each measure onto the corresponding pre-task baselines. We chose to examine the data using both of these methods. All analyses were performed using BMDP Statistical Software (Los Angeles, CA). The mean cell values as well as the results of the correlational and regression analyses are presented in Table 1. The stressor resulted in significant increases over baseline (main effect for task) in WBC (F,,,, = 8.12, p < 0.00, T-suppressor/ cytotoxic cells (F,,,, = 14.3, p < O.OOl), NK cells (CD16 F,,,, = 18.6, p < 0.001; CD56, F,,,, = 14.9, p < 0.001; CD57, F,,,, = 19.1, p < 0.001) and Tcells (F,,,, = 6.22, p < 0.031, and a significant decrease in the T-helper/suppressor ratio (F,,,3 = 9.4, p < 0.01). There were no significant task effects for T-helper cells nor significant main effects or interactions for session for any variable. For all cell types the baseline values showed statistically significant temporal stability, with CD16 natural killer cells showing the lowest stability (r = 0.40) and total WBC showing the highest (r = 0.82) (the combined helper/suppressor ratio (CD4/CD8) showed the highest value of r = 0.90). For the non-baseline-adjusted post-task values (absolute reactivity scores), with the excep-
I9 (I 99.5) 287-290
289
tion of the T-cells and the T-suppressor/cytotoxic cells, all cells showed significant correlation coefficients, with CD56 natural killer cells showing the lowest stability (r = 0.48) and total WBC showing the highest (r = 0.85). As with the baseline data, the absolute reactivity score helper/suppressor ratio (CD4/CD8) showed the highest value of r = 0.92. In contrast, the baseline adjusted residualized correlation coefficients were consistently lower, with only the natural killer (CD16, r = 0.48, p < 0.05 and CD56, r = 0.40, p < 0.051, total WBC (r = 0.48, p < 0.05), and the T-helper/suppressor ratio (CD4/CD8; r = 0.60, p < 0.001) showing statistical significance. The simple change score data were comparable to the residualized correlation coefficients. The findings suggest that a moderate intra-individual stability is a characteristic of some but not all measures of cellular immunity. In general, the baseline-adjusted immune test retest correlation coefficients were smaller as compared to cardiovascular test retest correlation coefficients typically reported in the literature, which generally range from r = 0.60 to 0.80 (Turner et al., 1986; Allen et al. 1987; Kasprowicz et ai., 1990; Light, 1991; Durel et al., 1993; Mills et al., 1993). The lower immune system coefficients may be the result of an inherently greater variability of these particular measures or may simply reflect measurement error. We utilized only a single blood draw (one at the baseline and one at the post-task time point) to determine these cell counts; cardiovascular studies typically employ measures in triplicate. The findings are similar however to the cardiovascular literature in that the correlation coefficients of the absolute reactivity scores were greater than the baseline-adjusted scores (Mills et al., 1993). Studies demonstrate that acute sympathetic nervous system activation drives such stress-induced changes in leukocyte cell numbers (Manuck et al., 1991; Mills et al., 1995). It may be that the relative stability of the sympathetic nervous system response to acute stress (Mills et al., 1993) underlies the moderate stability of these immune responses. Regarding generalizability, immune reactivity
290
P.J. Mills et al. /International
Journal of Psychophysiology 19 (1995) 287-290
studies typically examine either enumerative and/or functional measures of the immune system (such as natural killer cell cytotoxicity) (Irwin et al., 1991). We focused only on enumerative measures of immune reactivity. Additional studies would be needed to determine if these enumerative findings of temporal stability generalize to more functional immune measures. Finally, despite increasing evidence of the effects of behavioral stress on immune function, the exact nature of the relationship between immune reactivity to acute stress and long-term health has yet to be established. Given the importance of temporal stability in establishing a reactivity-risk relationship, these findings lend some support to continued investigation into the possible clinical implications of immune reactivity to acute stress.
Acknowledgements This work was supported by grants MOlRR00827, HL-36005, HL-40102, and HL-47074 from the National Institutes of Health. The authors are grateful to Ann Rearden, M.D. and the staff of the HLA and Immunogenetics Laboratory at UCSD for determination of the lymphocyte subsets.
References Allen, M.T., Sherwood, A., Obrist, P.A., Crowell, M. and Grange, L.A. (1987) Stability of cardiovascular reactivity to laboratory stressors: a 2 l/2 year follow-up. J. Psychosom. Res., 31: 639-645. Durel, L., Kus, L., Anderson, N., McNeilly, M., Llabre, M., Spitzer, S., Saab, P., Efland, .I., Williams, R. and Schneiderman, N. (1993) Patterns of stability of cardiovascular responses to variations of the cold pressor test. Psychophysiology, 30: 39-46. Fredrikson, M. and Matthews, K. (1990) Cardiovascular responses to stress and hypertension: A meta-analytic review. Ann. Behav. Med., 12: 30-39. Herbert, T.B. and Cohen, S. (1993) Stress and immunity in
humans: a meta-analytic review. Psychosom. Med., 55: 364-379. Irwin, M., Brown, M., Patterson, T., Hauger, R., Mascovich, A., and Grant, I. (1991) Neuropeptide Y and natural killer cell activity: findings in depression and AJzheimer caregiver stress. FASEB J., 5: 3100-3107. Kasprowicz, A.L., Manuck, S.B., Malkoff, S.B. and Krantz, D.S. (1990) Individual differences in behaviorally evoked cardiovascular response: temporal stability and hemodynamic patterning. Psychophysiology, 27/6: 605-619. Kiecolt-Glaser, J.K., Cacioppo, J., Malarkey, W. and Glaser, R. (1992) Acute psychological stressors and short-term immune changes: what, why, for whom, and to what extent? Psychosom. Med., 54: 680-685. Light, K.C. (1991) Cardiovascular responses to effortful active coping: implications for the role of stress in hypertension development. Psychophysiology, 18: 216-225. Llabre, M.M., Spitzer, S.B., Saab, P.G., Ironson, G.H. and Schneiderman, N. (1991) The reliability and specificity of delta versus residualized change as measures of cardiovascular reactivity to behavioral challenges. Psychophysiology, 28: 701-711. Manuck, S.B., Kasprowicz, A.L., Monroe, S.M., Larkin, K.T. and Kaplan, J.R. (1989) Psychophysiologic reactivity as a dimension of individual differences. In: Schneiderman, N., Weiss, SM. and Kaufmann, P. (Eds.), Handbook of Research Methods in Cardiovascular Behavioral Medicine, Plenum Press, New York, NY, pp. 365-382. Manuck, S.B., Cohen, S., Rabin, B.S., Muldoon, M.F. and Bachen, E.A. (1991) Individual differences in cellular immune response to stress. Psycbol. Sci., 2: 111-115. Mills, P., Berry, C., Dimsdale, J., Nelesen, R. and Ziegler, M. (1993) Temporal stability of task-induced cardiovascular, adrenergic, and psychological responses: the effects of race and hypertension. Psychophysiology, 30: 197-204. Mills, P.J., Berry, C., Dimsdale, J., Nelesen, R., Ziegler, M. and Kennedy, B. (1995) Lymphocyte subset redistribution in response to acute experimental stress: effects of gender, ethnic&y and the sympathetic nervous system. Brain Behav. Immun., in press. Nelesen, R., Dimsdale, J., Mills, P., Dillon, E. and Ziegler, M. 11995) Endogenous insulin influences cardiovascular reactivity to a laboratory stressor. Am. J. Hypertens., in press. Pickering, T.G. and Gerin, W. (1990) Cardiovascular reactivity in the laboratory and the role of behavioral factors in hypertension: a critical review. Ann. Behav. Med., 12: 3-16. Sherwood, A., Turner, J., Light, K. and Blumenthal, J. (1990) Temporal stability of the hemodynamics of cardiovascular reactivity. Int. J. Psychophysiol., 10: 95-98. Turner, J., Carroll, D., Sims, J., Hewitt, J. and Kelly, K. (1986) Temporal and inter-task consistency of heart rate reactivity during psychological challenge: a twin study. Physiol. Behav., 38: 641-644.
ELSEVIER
International
Journal
of Psychophysiology
19 (1995) 287-290
Short Communication
Temporal stability of acute stressor-induced immunity Paul J. Mills Department
of Psychiatry, Received
*,
changes in cellular
Soraya L. Haeri, Joel E. Dimsdale
University of California, San Diego, 200 West Arbor Dr, La Jolla, CA 92103-0804, USA 6 December
1994; revised
14 February
1995; accepted
16 February
1995
Abstract This study examined the temporal stability of enumerative immune responses to acute psychosocial stress. Lymphocyte subsets were measured in 24 healthy male subjects at rest and following a speaking stressor on two occasions approximately six weeks apart. The speaking task caused significant increases in T-suppressor/cytotoxic cells, natural killer cells, T-cells, and total WBC and decreases in the T-helper/suppressor ratio. Baseline test-retest correlation’s were statistically significant for all variables (r values = 0.40-0.90). With two exceptions (T-cells and T-suppressor/cytotoxic cells), speaking task values (absolute reactivity scores) were also statistically significant (r values = 0.48-0.92). Baseline adjusted test-retest correlations were however generally less reliable, with only natural killer cells (r values > 0.40), the T-helper/suppressor ratio (r = 0.601, and total WBC (r = 0.48) showing statistical significance. The findings suggest that certain but not all cellular immune responses to acute stress are moderately stable over time. Keywords:
Immune
system;
Stress;
Temporal
stability
Research on the effects of acute stressors continues to be a focus of psychophysiologic research. An important consideration in the stress reactivity paradigm is the temporal stability of the stress response, i.e., whether a given individual’s pattern of reactivity represents a stable characteristic. Studies do support the notion that cardiovascular (Turner et al., 1986; Allen et al. 1987; Kasprowicz et al., 1990; Light, 1991; Durel et al., 1993) and hemodynamic (Sherwood et al., 1990) responses are stable over time. Evidence for tem-
* Corresponding 543-5462.
author.
Tel.:
(619)
0167-8760/95/$09.50 0 1995 Elsevier SSDI 0167-8760(95)00012-7
543-2506;
Science
Fax:
(619)
B.V. All rights
poral stability is of fundamental importance to the validity of the controversial notion that cardiovascular reactivity is a marker or risk factor for cardiovascular illness (Fredrikson and Matthews, 1990; Pickering and Gerin, 1990). Increasingly, studies are focusing on the responses of the immune system to acute behavioral stress. Like the cardiovascular system, the immune system is highly responsive to acute stress (for review see Kiecolt-Glaser et al., 1992; Herbert and Cohen, 1993). Given the growth of this research field, and the speculation that immune reactivity has possible clinical implications for long-term health, we thought it important to examine whether cellular immune responses to reserved
288
P.J. Mills et al. /International
Journal of Psychophysiology 19 (1995) 287-290
stress are also stable over time. To date, few if any reports have examined this issue. This study therefore, examined the temporal stability of enumerative immune responses to acute laboratory stress. Twenty-four healthy male volunteers were located through advertisement or word-of-mouth referral (mean age 30.2 years (SD 7); range 1943). Subjects were studied after obtaining written informed consent. The protocol was approved by the University of California, San Diego, Institutional Review Board. All subjects were studied at the same time of day (approximately 9:00 a.m.) on two separate occasions approximately 6 weeks apart. Subjects were instructed to eat a light breakfast and refrain from caffeinated beverages and smoking 12 h prior to study. Upon arrival to the laboratory, subjects were seated and a 19-gauge catheter was inserted into a forearm vein. Subjects then rested quietly for 30 min. The subjects were then given instructions for a speaking task. The task required preparing (3 min) and presenting (3 min) a speech in front of a video camera. Subjects were told that the speech would be evaluated and rated by experts. The topic of the task was different on each testing occasion, once involving defending oneself from being falsely accused of shoplifting and once dealing with a dishonest automobile service shop. If subjects stopped speaking before the 3 min were up, they were reminded to continue to talk by reiterating and summarizing their points. We have used this task in a different group of individuals in previous studies and it reliably results in significant cardiovascular and sympathetic nervous system activation (Nelesen et al., 1995). Following the resting baseline and at the end of the speaking task, 10 ml of blood was drawn for leukocyte determination. Flow cytometry (FAC-Scan, Becton Dickinson) on whole blood with CD45 gating was used to quantify white cell populations (Mills et al., 1995). We chose a panel of cells typically reported in studies examining the effects of acute psychologic stressors on cellular immune responses. The following cell types were determined (included are the antibodies employed): T-cells (CD3, anti-leu 41, T-helper cells (CD4,
P.J. Mills et al. /International
Journal of Psychophysiology
anti-leu 3a), T-suppressor/cytotoxic cells (CDS, anti-leu 2a), three natural killer (NK) cells, CD16 (anti-leu lib), CD56 (anti-leu 191, and CD57 (anti-leu 71, and total WBC. Blood was processed within 5 h of collection. Cells were detected by mouse monoclonal antibodies conjugated with either fluorescein or phycoerythrin. two-way repeated measures Separate ANOVAs (stress X session) were used to examine the effects of the stressor on cell population values. Pearson correlation coefficients were calculated between corresponding baseline and post-task measures across the two testing sessions. Previous research on temporal stability of reactivity has taken two approaches to examining change score reactivity (Manuck et al., 1989; Mills et al., 1993; Llabre et al., 1991). One approach utilizes an arithmetically-derived baseline-corrected value (simple change scores, task minus baseline) and the other utilizes a statistically-derived baseline-corrected value (residualized change scores), the latter being derived from the regression of the task values of each measure onto the corresponding pre-task baselines. We chose to examine the data using both of these methods. All analyses were performed using BMDP Statistical Software (Los Angeles, CA). The mean cell values as well as the results of the correlational and regression analyses are presented in Table 1. The stressor resulted in significant increases over baseline (main effect for task) in WBC (F,,,, = 8.12, p < 0.00, T-suppressor/ cytotoxic cells (F,,,, = 14.3, p < O.OOl), NK cells (CD16 F,,,, = 18.6, p < 0.001; CD56, F,,,, = 14.9, p < 0.001; CD57, F,,,, = 19.1, p < 0.001) and Tcells (F,,,, = 6.22, p < 0.031, and a significant decrease in the T-helper/suppressor ratio (F,,,3 = 9.4, p < 0.01). There were no significant task effects for T-helper cells nor significant main effects or interactions for session for any variable. For all cell types the baseline values showed statistically significant temporal stability, with CD16 natural killer cells showing the lowest stability (r = 0.40) and total WBC showing the highest (r = 0.82) (the combined helper/suppressor ratio (CD4/CD8) showed the highest value of r = 0.90). For the non-baseline-adjusted post-task values (absolute reactivity scores), with the excep-
I9 (I 99.5) 287-290
289
tion of the T-cells and the T-suppressor/cytotoxic cells, all cells showed significant correlation coefficients, with CD56 natural killer cells showing the lowest stability (r = 0.48) and total WBC showing the highest (r = 0.85). As with the baseline data, the absolute reactivity score helper/suppressor ratio (CD4/CD8) showed the highest value of r = 0.92. In contrast, the baseline adjusted residualized correlation coefficients were consistently lower, with only the natural killer (CD16, r = 0.48, p < 0.05 and CD56, r = 0.40, p < 0.051, total WBC (r = 0.48, p < 0.05), and the T-helper/suppressor ratio (CD4/CD8; r = 0.60, p < 0.001) showing statistical significance. The simple change score data were comparable to the residualized correlation coefficients. The findings suggest that a moderate intra-individual stability is a characteristic of some but not all measures of cellular immunity. In general, the baseline-adjusted immune test retest correlation coefficients were smaller as compared to cardiovascular test retest correlation coefficients typically reported in the literature, which generally range from r = 0.60 to 0.80 (Turner et al., 1986; Allen et al. 1987; Kasprowicz et ai., 1990; Light, 1991; Durel et al., 1993; Mills et al., 1993). The lower immune system coefficients may be the result of an inherently greater variability of these particular measures or may simply reflect measurement error. We utilized only a single blood draw (one at the baseline and one at the post-task time point) to determine these cell counts; cardiovascular studies typically employ measures in triplicate. The findings are similar however to the cardiovascular literature in that the correlation coefficients of the absolute reactivity scores were greater than the baseline-adjusted scores (Mills et al., 1993). Studies demonstrate that acute sympathetic nervous system activation drives such stress-induced changes in leukocyte cell numbers (Manuck et al., 1991; Mills et al., 1995). It may be that the relative stability of the sympathetic nervous system response to acute stress (Mills et al., 1993) underlies the moderate stability of these immune responses. Regarding generalizability, immune reactivity
290
P.J. Mills et al. /International
Journal of Psychophysiology 19 (1995) 287-290
studies typically examine either enumerative and/or functional measures of the immune system (such as natural killer cell cytotoxicity) (Irwin et al., 1991). We focused only on enumerative measures of immune reactivity. Additional studies would be needed to determine if these enumerative findings of temporal stability generalize to more functional immune measures. Finally, despite increasing evidence of the effects of behavioral stress on immune function, the exact nature of the relationship between immune reactivity to acute stress and long-term health has yet to be established. Given the importance of temporal stability in establishing a reactivity-risk relationship, these findings lend some support to continued investigation into the possible clinical implications of immune reactivity to acute stress.
Acknowledgements This work was supported by grants MOlRR00827, HL-36005, HL-40102, and HL-47074 from the National Institutes of Health. The authors are grateful to Ann Rearden, M.D. and the staff of the HLA and Immunogenetics Laboratory at UCSD for determination of the lymphocyte subsets.
References Allen, M.T., Sherwood, A., Obrist, P.A., Crowell, M. and Grange, L.A. (1987) Stability of cardiovascular reactivity to laboratory stressors: a 2 l/2 year follow-up. J. Psychosom. Res., 31: 639-645. Durel, L., Kus, L., Anderson, N., McNeilly, M., Llabre, M., Spitzer, S., Saab, P., Efland, .I., Williams, R. and Schneiderman, N. (1993) Patterns of stability of cardiovascular responses to variations of the cold pressor test. Psychophysiology, 30: 39-46. Fredrikson, M. and Matthews, K. (1990) Cardiovascular responses to stress and hypertension: A meta-analytic review. Ann. Behav. Med., 12: 30-39. Herbert, T.B. and Cohen, S. (1993) Stress and immunity in
humans: a meta-analytic review. Psychosom. Med., 55: 364-379. Irwin, M., Brown, M., Patterson, T., Hauger, R., Mascovich, A., and Grant, I. (1991) Neuropeptide Y and natural killer cell activity: findings in depression and AJzheimer caregiver stress. FASEB J., 5: 3100-3107. Kasprowicz, A.L., Manuck, S.B., Malkoff, S.B. and Krantz, D.S. (1990) Individual differences in behaviorally evoked cardiovascular response: temporal stability and hemodynamic patterning. Psychophysiology, 27/6: 605-619. Kiecolt-Glaser, J.K., Cacioppo, J., Malarkey, W. and Glaser, R. (1992) Acute psychological stressors and short-term immune changes: what, why, for whom, and to what extent? Psychosom. Med., 54: 680-685. Light, K.C. (1991) Cardiovascular responses to effortful active coping: implications for the role of stress in hypertension development. Psychophysiology, 18: 216-225. Llabre, M.M., Spitzer, S.B., Saab, P.G., Ironson, G.H. and Schneiderman, N. (1991) The reliability and specificity of delta versus residualized change as measures of cardiovascular reactivity to behavioral challenges. Psychophysiology, 28: 701-711. Manuck, S.B., Kasprowicz, A.L., Monroe, S.M., Larkin, K.T. and Kaplan, J.R. (1989) Psychophysiologic reactivity as a dimension of individual differences. In: Schneiderman, N., Weiss, SM. and Kaufmann, P. (Eds.), Handbook of Research Methods in Cardiovascular Behavioral Medicine, Plenum Press, New York, NY, pp. 365-382. Manuck, S.B., Cohen, S., Rabin, B.S., Muldoon, M.F. and Bachen, E.A. (1991) Individual differences in cellular immune response to stress. Psycbol. Sci., 2: 111-115. Mills, P., Berry, C., Dimsdale, J., Nelesen, R. and Ziegler, M. (1993) Temporal stability of task-induced cardiovascular, adrenergic, and psychological responses: the effects of race and hypertension. Psychophysiology, 30: 197-204. Mills, P.J., Berry, C., Dimsdale, J., Nelesen, R., Ziegler, M. and Kennedy, B. (1995) Lymphocyte subset redistribution in response to acute experimental stress: effects of gender, ethnic&y and the sympathetic nervous system. Brain Behav. Immun., in press. Nelesen, R., Dimsdale, J., Mills, P., Dillon, E. and Ziegler, M. 11995) Endogenous insulin influences cardiovascular reactivity to a laboratory stressor. Am. J. Hypertens., in press. Pickering, T.G. and Gerin, W. (1990) Cardiovascular reactivity in the laboratory and the role of behavioral factors in hypertension: a critical review. Ann. Behav. Med., 12: 3-16. Sherwood, A., Turner, J., Light, K. and Blumenthal, J. (1990) Temporal stability of the hemodynamics of cardiovascular reactivity. Int. J. Psychophysiol., 10: 95-98. Turner, J., Carroll, D., Sims, J., Hewitt, J. and Kelly, K. (1986) Temporal and inter-task consistency of heart rate reactivity during psychological challenge: a twin study. Physiol. Behav., 38: 641-644.