Author + information
- Received July 29, 2009
- Revision received November 18, 2009
- Accepted November 19, 2009
- Published online June 8, 2010.
- Brent T. Mausbach, PhD*,* (, )
- Susan K. Roepke, MS*,
- Michael G. Ziegler, MD†,
- Milos Milic, MD, PhD†,
- Roland von Känel, MD*,§,
- Joel E. Dimsdale, MD*,
- Paul J. Mills, PhD*,
- Thomas L. Patterson, PhD*,
- Matthew A. Allison, MD‡,
- Sonia Ancoli-Israel, PhD* and
- Igor Grant, MD*
- ↵*Reprint requests and correspondence:
Dr. Brent T. Mausbach, Department of Psychiatry (0680), University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093-0680
Objectives We examined the relationship between chronic caregiving stress and endothelial function.
Background Evidence suggests that caregiving stress is associated with pathophysiologic processes related to atherosclerosis. Endothelial dysfunction is a possible underlying mechanism explaining the relationship between caregiving stress and cardiovascular morbidity. We investigated the relationship between chronic caregiving stress and endothelial dysfunction assessed by reactive hyperemia–induced flow-mediated dilation (FMD).
Methods Seventy-eight elderly individuals participated in the study. Fifty-five were providing in-home care to a spouse with Alzheimer's disease, and 23 were married and living with a healthy, nondemented spouse. Analysis of covariance was used to examine the relationships between advancing dementia severity (Clinical Dementia Rating scores) and FMD and nitroglycerin-induced vasodilation of the brachial artery. Multiple linear regression was used to examine the relationship between years of caregiving and FMD.
Results Clinical Dementia Rating scale scores were significantly related to FMD (p = 0.033), with participants caring for a spouse with moderate to severe dementia showing significantly worse FMD than those caring for a spouse with mild dementia (p = 0.028) and noncaregivers (p = 0.032). Within the caregiver sample, the number of years of caregiving was significantly related to FMD (r = −0.465, p < 0.001).
Conclusions These results suggest that the chronic stress of caregiving is associated with impaired endothelial function, which may be a potential mechanistic link to the observed increased risk of cardiovascular disease in elderly caregivers.
The atherosclerotic disease process is mediated by a number of complex pathophysiologic processes resulting in thickening of the arterial walls nearest to the lumen. Dysfunction of the endothelial lining plays a key role in the development and progression of atherosclerosis (1,2). Among its numerous functions, the endothelium is involved in inhibiting platelet aggregation, contributing to formation and secretion of growth-regulatory molecules and cytokines, and the release of a number of chemical mediators such as nitric oxide (NO), a potent vasodilator released in response to shear stress (1,3). Bioavailability of NO is indicative of cardiovascular health (4); however, dysfunction of the endothelium may disrupt its ability to balance vasoconstriction and vasodilation, resulting in compensatory responses that interfere with the maintenance of vascular homeostasis (5). Endothelial injury results in a number of changes to its function that may promote procoagulant properties, inflammatory responses, and vasoconstriction that contributes to arterial thickening (5,6), atherosclerotic lesion formation (2), and ultimately to the development of atherosclerosis (1,3).
Brachial artery flow-mediated dilation (FMD) is a noninvasive method designed to assess endothelial function of the peripheral conduit artery in humans (7). FMD highly correlates with invasive quantification of the vasomotor responses of epicardial arteries to acetylcholine, which is adopted as the gold standard (8,9). The FMD technique uses upper arm occlusion to induce distal hypoxia followed by the reactive hyperemia and local vasodilation after cuff deflation. The distal vasodilation, in turn, will induce a large increase in the shear stress upstream in the brachial artery. In response to the increased shear, the brachial artery endothelial cells increase the production of NO, causing the vascular smooth muscle to relax and the artery to dilate (3). Impaired FMD has been prospectively associated with increased risk of cardiovascular events (10,11), such as post-operative events in patients undergoing vascular surgery. Furthermore, Shimbo et al. (12) found that impaired brachial FMD was predictive of incident cardiovascular events in asymptomatic, lower-risk individuals in a population-based study. However, the predictive value of FMD in this study was not independent of cardiovascular risk factors. Factors associated with worse FMD include increasing age (with declining function occurring earlier for men than women) (13,14), increased systolic blood pressure (SBP) (13), smoking (13), and total cholesterol to high-density lipoprotein cholesterol ratio (15). In contrast, use of cholesterol-lowering medication has been associated with improved FMD (16,17).
Importantly, it has been hypothesized that atherosclerosis may be exacerbated by repeated and sustained sympathetic nervous system activation resulting from exposure to environmental and psychological stressors (18,19). Indeed, the chronic stress of caring for a disabled loved one (e.g., with Alzheimer's disease [AD]), has been associated with increased cardiovascular risk (20–22). Chronic caregiving stress and factors associated with the stress process have also been associated with a number of physiological processes related to mechanisms associated with atherosclerosis including sympathetic reactivity (23,24), coagulation (25,26), and platelet activation (24). However, the association between chronic caregiving stress and endothelial dysfunction has yet to be demonstrated.
Given the evidence that caregiving stress is associated with pathophysiologic processes related to atherosclerosis, this study aimed to investigate the potential relationship between chronic caregiving stress and endothelial dysfunction (i.e., impaired FMD) in a population of 55 elderly AD caregivers and a control group of 23 age- and sex-equivalent noncaregivers. Endothelial dysfunction was conceptualized as a possible underlying mechanism explaining the relationship between caregiving stress and cardiovascular morbidity (27). Although caregiving stress has been conceptualized in a number of ways (e.g., number of care recipient problem behaviors, role overload, burden), these measures typically reflect snapshots of caregiver stress at specific moments in time rather than the accumulated wear and tear that caregivers may experience over time. Therefore, we assessed chronic stress with 2 measures: 1) the Clinical Dementia Rating (CDR) of participants' spouses; and 2) the number of years that participants were providing in-home care for their spouses. We hypothesized that higher CDR scores and more years of caregiving would be associated with impaired endothelial functioning, as indexed by reduced FMD.
Seventy-eight elderly individuals participated in this study. Of these, 55 were providing care to a spouse with AD and 23 were married to a healthy, nondemented spouse (i.e., healthy controls). All participants were enrolled in the Alzheimer's Caregiver Study at the University of California San Diego, which was designed to examine physiologic and psychological mechanisms of increased health risk in spousal caregivers. Participants were required to be free of major illnesses (e.g., cancer), at least 55 years of age, married, and living with their spouses at the time of enrollment. Participants were excluded if they had extreme hypertension (>200/120 mm Hg). A total of 127 participants were screened for the study. Of these, 88 (69.3%) were eligible to participate and 29 (22.8%) were ineligible. The remaining 10 (7.9%) participants were screened eligible but chose not to participate in the study. Of the 39 participants who were ineligible, the most common reasons for being excluded were: 1) the individual was previously, but not currently a caregiver (e.g., spouse passed away; n = 10); 2) nonspouse caregiver (e.g., caring for a parent with AD; n = 7); 3) current or recent serious medical condition (e.g., cancer requiring chemotherapy; n = 4); and 4) not currently living in San Diego or surrounding community (n = 2). Participants were recruited through referrals from the University of California San Diego Alzheimer's Disease Research Center, community agencies serving caregivers, local caregiver support groups, community health fairs, and referrals from other participants.
Participants were interviewed regarding several demographic and health variables. Because of their potential relationship to FMD, participants provided information regarding their age, smoking history, medication use over the past 30 days, and years of caregiving (years since spouse received a diagnosis of AD). For noncaregivers, a value of 0 was assigned for years of caregiving.
Brachial Artery FMD and Nitroglycerin-Mediated Dilation (NMD)
Tests of endothelium-dependent (FMD) and endothelium-independent (NMD) response of the right brachial artery to the increased blood flow and NO, respectively, were performed by a single technician with the modified method first described by Celermajer et al. (7). All measurements were done after 15 min of relaxation in the supine position, between 11:00 amand 1:00 pm, after fasting and without vasoactive medications. The occlusion cuff was placed on the right upper arm and the brachial artery was scanned, in longitudinal sections 4 to 10 cm proximal to the antecubital fossa, using an Acuson Cypress portable ultrasound system with a 5.4 to 6.6 MHz linear array transducer (Model 7L3, Siemens Medical Solutions, Mountain View, California).
After the brightest views of the anterior and posterior artery walls had been obtained, 3 baseline images were saved. Then, the occlusion cuff was inflated to 50 mm Hg above SBP, thereby producing distal hypoxia for 5 min. After the cuff was deflated, arterial images were saved every 15 s during the first minute post-occlusion and then once every 30 s for an additional 8 min. A single technician, blind to the caregiver status of the participant, measured artery diameters manually from the saved digital ultrasound images with the Acuson Cypress built-in vascular measurements software module (Siemens Medical Solutions). All measurements were done by placing electronic calipers on the anterior and posterior intima line (i-i line). FMD was calculated as the maximum percentage change in the brachial artery diameter, FMD%(max), from the average baseline diameter value DFMD(b)to the maximum diameter value after the cuff deflation DFMD(max):Fifteen minutes after the FMD test, the baseline brachial artery diameter was determined again by averaging 3 diameter measurements taken immediately before applying nitroglycerin. Then, a 400-μg nitroglycerin tablet (NitroQuick, ETHEX Corp, St. Louis, Missouri) was given sublingually to induce vasodilation. Ultrasound scans were performed continuously, and the brachial artery diameter was measured once every minute during the 7 min. NMD was calculated as the maximum percentage of change in the brachial artery diameter, NMD%NMD(max), from the baseline value DNMD(b)to the maximum value obtained with sublingual nitroglycerin DNMD(max):
Dementia Rating of Spouses
Participants were interviewed using the CDR scale (28), whereby participants indicated the extent to which their spouses exhibited symptoms of dementia in 6 domains: 1) memory; 2) orientation; 3) judgment and problem solving; 4) community affairs; 5) home and hobbies; and 6) personal care. Based on responses to these items, an overall dementia severity score is given; a score of 0 = no dementia, 1 = mild dementia, 2 = moderate dementia, and 3 = severe dementia. By study design, noncaregivers were required to be married to nondemented spouses, and so all spouses of noncaregivers were scored 0 (no dementia). In addition, all caregivers were required to have spouses with at least mild dementia, so spouses of caregivers had CDR scores of at least 1.
Each participant was administered the Role Overload Scale (29), which assesses overall stress experienced by the individual. This scale consists of 4 items (e.g., “You have more things to do than you can handle”) rated by the participant on a 4-point Likert scale ranging from 1 = not at all to 4 = completely.
Before FMD analysis, a total of 3 resting blood pressure measurements were collected by a research nurse over a 15-min resting period. The mean of the 3 measurements was taken as the participant's mean resting blood pressure.
Total cholesterol and high-density lipoprotein cholesterol were determined by standard methodology at the clinical chemistry laboratories at the University of California San Diego Medical Center. The total cholesterol to high-density lipoprotein cholesterol ratio was computed as an index of dyslipidemia.
CDR and FMD
Analysis of covariance was used to assess the relationship between the CDR group and FMD, in which FMD was our dependent variable and CDR score was our primary independent variable. Preliminary examination of CDR scores indicated that only 5 care recipients were classified as severely demented (i.e., CDR score = 3). Therefore, participants in this category were grouped with those with a CDR score of 2, thereby resulting in 3 CDR groups for analysis. These 3 groups corresponded to the CDR score for the care recipient (i.e., CDR0, CDR1, and CDR2). Because of their potential correlation with FMD, the following covariates were included in this analysis: age, sex, smoking history (yes/no), mean resting SBP, total cholesterol to high-density lipoprotein cholesterol ratio, current use of cholesterol-lowering medication (yes/no), and role overload. A significant omnibus test was followed by post hoc least significant difference tests to determine differences between the 3 CDR groups.
Because administration of nitroglycerin produces dilation of the brachial artery independent of the endothelium, NMD should not be dependent on chronic stress. To demonstrate this effect, a second analysis of covariance (ANCOVA) was conducted in which NMD was entered as our dependent variable and CDR group was our independent variable. Clinical covariates used in our first analysis were also entered in this analysis.
Years of Caregiving and FMD
We conducted a second set of analyses to examine the relationship between years of caregiving and FMD. A first analysis was conducted using ANCOVA, in which participants were divided into 3 groups: noncaregivers (n = 23), caregivers with “few” years of caregiving (i.e., <4 years; n = 26), and caregivers with “many” years of caregiving (i.e., ≥4 years; n = 28). Differentiation of few versus many years of caregiving was determined by a median split. Covariates were the same as in our previous analysis (see previous text).
In a second analysis, linear regression was used to examine the relationship between linear years of caregiving and FMD within our caregivers only. Covariates used in this analysis were the same as with our other analyses (described previously).
Demographic and clinical characteristics of caregivers and noncaregivers are presented in Table 1.Group comparisons of these characteristics, using ttests and chi-square analyses for linear and bivariate variables, indicated that the 2 groups were statistically similar on all variables except smoking history, for which caregivers were more likely to report a history of smoking (p = 0.044).
Relationship between CDR score and FMD
Mean baseline brachial artery diameter (mean ± SD) was 0.35 ± 0.07 for CDR0, 0.33 ± 0.04 for CDR1, and 0.37 ± 0.07 for CDR2. Analysis of variance indicated no significant differences by group (F = 2.41, df = 2, 75; p = 0.097).
Results of our ANCOVA indicated that CDR score was significantly related to FMD (F = 3.60, df = 2, 69; p = 0.033). Post hoc analyses indicated that participants in the CDR2 group (mean 12.52 ± 5.01) had significantly worse FMD compared with those in the CDR1 group (mean 16.01 ± 5.20; p = 0.028) and CDR0 group (mean 15.74 ± 6.09; p = 0.032). The CDR0 and CDR1 groups did not significantly differ in FMD (p = 0.918). Means and SE for covariate-adjusted FMD are presented in Figure 1.
Relationship between CDR score and nitroglycerin
As mentioned previously, we repeated our initial analysis using NMD as our dependent variable and CDR score as our primary independent variable. For this analysis, data were missing for 8 participants. The numbers of participants with missing data for CDR groups were as follows: CDR0 = 3, CDR1 = 2, and CDR2 = 3. The most common reasons for missing data were refusal of nitroglycerin (n = 6), history of negative reaction to nitroglycerin (n = 1), and a low resting pulse (n = 1).
Results of the NMD analysis indicated a nonsignificant effect of CDR group (F = 0.86, df = 2, 61; p = 0.427). Also, age was significantly related to NMD, with older participants having significantly worse NMD (p = 0.029). None of the other covariates was significant (all p values >0.05). Post hoc exploratory analyses for CDR group indicated no differences between CDR0 and CDR1 (p = 0.279), CDR0 and CDR2 (p = 0.980), or CDR1 and CDR2 (p = 0.223). The mean ± SD for CDR0, CDR1, and CDR2 were 25.87 ± 9.05, 29.06 ± 9.28, and 25.92 ± 9.55, respectively.
Relationship between years of caregiving and FMD
One caregiver had missing data for years of caregiving and was excluded from our analyses. The results of our ANCOVA indicated a significant omnibus test for group differences (F = 5.94, df = 2, 67; p = 0.004). The covariate-adjusted FMD mean (± SE) for noncaregivers was 15.5 ± 0.01. The mean (± SE) FMD for caregivers with <4 years of caregiving was 16.2 ± 0.01, whereas FMD for those with ≥4 years of caregiving was 11.7 ± 0.01. Post hoc least significant difference comparisons indicated that caregivers with <4 years of caregiving were not significantly different from noncaregivers (p = 0.690), whereas caregivers with ≥4 years of caregiving had significantly worse FMD than noncaregivers (p = 0.025).
The results of our multiple regression model (with covariates) are presented in Table 2.Overall, our model explained 36.7% of the variance in FMD (adjusted R2= 25.5%). Significant predictors in the model included taking cholesterol-lowering medication and years of caregiving; that is, caregivers who were taking cholesterol-lowering medication had improved FMD, whereas a greater number of years of caregiving was associated with worsened FMD (Table 2). The bivariate correlation between years of caregiving and FMD was −0.465 (p < 0.001) and is presented in Figure 2.
Independent predictability of CDR and FMD
A final ANCOVA (within caregivers) examined the effect of CDR group and years of caregiving in the same model. In this model, both CDR group (F = 4.35, df = 1, 44; p = 0.043) and years of caregiving (F = 13.67, df = 1, 44; p = 0.001) were significant predictors of FMD. Specifically, caregivers of moderately/severely demented patients had worse FMD than those caring for mildly demented patients, and the longer caregivers had provided care, the worse was their FMD. These results suggest that CDR and years of caregiving are independently predictive of FMD. Overall, this model explained 42.4% of the variance in FMD (adjusted R2= 30.6%).
This study of 78 elderly participants suggests that the chronic stresses of caring for a spouse with AD may be associated with worse endothelial functioning, as measured by brachial artery FMD. Specifically, advancing dementia (i.e., CDR) and years of caregiving were associated with impaired endothelial functioning, independent of endothelial risk factors including age, sex, cholesterol-lowering medication use, SBP, smoking history, and total cholesterol to high-density lipoprotein cholesterol ratio. These results are consistent with and expand on previous research showing increased cardiovascular risk in stressed caregivers (21,30), suggesting that cardiovascular risk in caregivers encompasses hemodynamic, inflammatory, and endothelial mechanisms. Noninvasive brachial FMD, measured by ultrasonography, is often used to assess endothelial dysfunction in peripheral conduit arteries and has been correlated with invasive measures of coronary artery endothelial dysfunction (8,9). The rapid flow of blood generates high shear stress at the endothelium causing the release of NO, which relaxes smooth muscle and dilates the blood vessel. We find no evidence of differences in the ability of the brachial artery smooth muscle to relax in long duration caregivers because the NO donor nitroglycerin gave equivalent vasodilation among groups. However, the diminished vasodilation in response to increase in flow indicates impaired endothelial NO release among long-duration caregivers. Impaired FMD is a first step in the development of atherosclerosis, but is partially reversible with angiotensin receptor blockers, statins, and exercise (31).
Although caregivers of patients in the early stages of dementia may be required to adapt to caregiving-related stresses, we conceptualized that these caregivers had not experienced the chronic buildup, or wear and tear, that mid- to late-stage caregivers had experienced. This wear and tear likely includes repeated hyperemia-induced shear stress on the endothelium, with this buildup of stress likely reaching threshold during the mid-to-late stages of dementia. The result is believed to be early signs of endothelial dysfunction. This hypothesis was supported by both of our analyses. Caregivers providing care to a spouse with mild dementia (CDR1) did not show significant impairment in FMD relative to noncaregivers (i.e., nonstressed). However, those providing care to a spouse with moderate to severe dementia demonstrated significantly impaired FMD relative to noncaregivers. The FMD difference between caregivers of those with moderate/severe dementia (CDR2) and noncaregivers (CDR0) was approximately 3.2%. Based on a large-scale FMD study of the elderly (32), this difference suggests that the odds of a participant in the CDR2 group having an earlier adverse cardiovascular event over a 5-year period are approximately 1.31 times greater than that of noncaregivers. Regarding our years of caregiving analysis, the magnitude of effect (i.e., effect size) (r = −0.465) was in the medium to large range when using Cohen's (33) definition of small, medium, and large effect sizes (i.e., r = 0.1, r = 0.3, and r = 0.5, respectively). These effect sizes suggest clinically meaningful relationships between chronic stress and endothelium functioning.
Although these results are promising, causal interpretation of these data is premature. However, if confirmed, these results suggest that the chronic caregiving stress may lead to early signs of atherosclerosis. As with any patient, clinicians are encouraged to monitor and curb clinical risk factors of atherosclerosis, including hyperlipidemia and blood pressure, particularly given that reversal of these factors has been shown to improve endothelial functioning (2). This is particularly relevant for caregiving populations given their increased risk of cardiovascular disease (CVD). Accordingly, clinicians would be encouraged to monitor caregiver-related stress and make appropriate recommendations for stress reduction such as referring caregivers to interventions that are efficacious for caregiver reducing distress. In this regard, a number of interventions for caregivers have been found efficacious, including cognitive-behavioral, psychoeducational, and multicomponent interventions (34), all of which teach caregivers specific, behavioral strategies for managing stress and improving well-being. In addition, given that disruptive patient behaviors are often the most stressful aspect of caregiving (35), clinicians might consider recommending efficacious interventions for reducing disruptive patient behaviors (36).
In addition to the cross-sectional nature of this study, other limitations should be noted. First, our sample was relatively small, which limited the number of covariates that we could include in our model. A related limitation is that our sample included only 5 participants caring for a spouse with severe dementia. This led us to group these participants with those in the moderate dementia group. It should be noted that FMD for participants caring for severely demented patients (12.42 ± 2.44) was similar to that of those in the moderate group (12.64 ± 0.95) and that our correlation analysis showed participants who had provided the longest care showed the worst FMD functioning. These analyses suggest that chronic wear and tear may indeed be associated with endothelial dysfunction. However, these results should be replicated in a larger sample that also includes more late-stage dementia caregivers.
We used least significant difference post hoc analyses to examine between-group differences in FMD, but these analyses do not control for multiple comparisons. However, our analyses can inform future researchers on developing planned comparisons that might include comparisons of caregivers of those mild, moderate, and severe dementia with noncaregivers. We strongly encourage this line of research.
A final limitation was that we included comparatively healthy participants. That is, our inclusion criteria limited our sample to healthy individuals without a history of major illnesses (e.g., severe hypertension, cancer, heart disease). Thus, although highly-distressed caregivers may be at greater risk of a CVD diagnosis (21) and mortality (37), participants in our study were likely not to have been diagnosed with CVD given our inclusion criteria. Nonetheless, the results provide preliminary data suggesting that caring for a spouse with AD may introduce circumstances that place caregivers at risk of future CVD.
We found that chronically-stressed AD caregivers demonstrated impaired endothelial functioning, as assessed by reactive hyperemia–induced FMD. Because caregiving (21,22) and worse endothelial function (10,11) have been associated with an increased risk of cardiovascular events, these results suggest a potential mechanism by which the chronic stresses of caregiving leads to negative health outcomes. Future studies should examine FMD, cardiovascular events, and stress among caregivers over time to determine the associations among these variables. If chronic stress is indeed associated with endothelial dysfunction, clinicians are encouraged to monitor chronic stress and make appropriate referrals to efficacious stress-reduction programs, including psychosocial treatments (34,36).
This study was supported by the National Institutes of Health/National Institute on Agingthrough award AG 15301(to Dr. Grant). Additional support was provided through award AG 03090(to Dr. Mausbach). Dr. Dimsdale receives research grant support from Sepracor Pharmaceuticalsfor work unrelated to that discussed in this paper. Dr. Ancoli-Israel serves as a consultant and is on the Scientific Advisory Board for Ferring Pharmaceuticals Inc., GlaxoSmithKline, Orphagen Pharmaceuticals, Pfizer, Respironics, Sanofi-Aventis, Sepracor, Inc., and Schering-Plough for work unrelated to this paper; she also has received grants/contracts from Sepracor, Inc.and Litebook, Inc.for work unrelated to this paper.
- Abbreviations and Acronyms
- Alzheimer's disease
- analysis of covariance
- Clinical Dementia Rating
- cardiovascular disease
- flow-mediated dilation
- nitroglycerin-mediated dilation
- nitric oxide
- systolic blood pressure
- Received July 29, 2009.
- Revision received November 18, 2009.
- Accepted November 19, 2009.
- American College of Cardiology Foundation
- Celermajer D.S.
- Anderson T.J.,
- Uehata A.,
- Gerhard M.D.
- Halcox J.P.,
- Schenke W.H.,
- Zalos G.,
- et al.
- Gokce N.,
- Keaney J.F.,
- Hunter L.M.,
- et al.
- Benjamin E.J.,
- Larson M.G.,
- Keyes M.J.,
- et al.
- Schnell G.B.,
- Robertson A.,
- Houston D.,
- Malley L.,
- Anderson T.J.
- Settergren M.,
- Bohm F.,
- Ryden L.,
- Pernow J.
- Parmar K.M.,
- Nambudiri V.,
- Dai G.,
- Larman H.B.,
- Gimbrone M.A. Jr..,
- Garcia-Cardena G.
- von Känel R.,
- Mills P.J.,
- Fainman C.,
- Dimsdale J.E.
- Widmaier E.P.,
- Raff H.,
- Strang K.T.
- Vitaliano P.P.,
- Scanlan J.M.,
- Zhang J.,
- Savage M.V.,
- Hirsch I.B.,
- Siegler I.C.
- Mausbach B.T.,
- von Känel R.,
- Aschbacher K.,
- et al.
- Vitaliano P.P.,
- Scanlan J.M.,
- Zhang J.,
- Savage M.V.,
- Hirsch I.B.,
- Siegler I.C.
- Pearlin L.I.,
- Mullan J.T.,
- Semple S.J.,
- Skaff M.M.
- von Känel R.,
- Dimasdale J.E.,
- Mills P.J.,
- et al.
- Yeboah J.,
- Crouse J.R.,
- Hsu F.C.,
- Burke G.L.,
- Herrington D.M.
- Cohen J.
- Schulz R.,
- O'Brien A.T.,
- Bookwala J.,
- Fleissner K.