| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Brown University School of Medicine (R.N., S.M.B.), Providence, RI; University of Memphis Prevention Center (K.D.W.), Memphis, TN; Department of Psychology, Ohio State University (C.M.S.), Columbus, OH; Normative Aging Study, Boston VA Outpatient Clinic, and Department of Epidemiology and Biostatistics, Boston University School of Public Health (A.S.), Boston, MA; Department of Applied Behavioral Sciences, University of California at Davis (C.M.A.), Davis, CA; Department of Medicine, Evans Memorial Hospital, Northwestern University School of Medicine (L.L.), Northbrook, IL; and The Channing Laboratory, Brigham and Women’s Hospital and Harvard Medical School (S.T.W.), Boston, MA.
Address reprint requests to: Raymond Niaura, PhD, Center for Behavioral and Preventive Medicine, The Miriam Hospital, 164 Summit Avenue, Providence, RI 02906.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
METHODS: Subjects included men who completed the Minnesota Multiphasic Personality Inventory in 1986 and who participated in a subsequent laboratory examination within 1 to 4 years. Total and subscale Ho scores were computed, and associations with anthropometric data, cigarette smoking, dietary information, serum lipids, blood pressure, and fasting glucose and insulin levels were examined.
RESULTS: The total Ho score was positively associated with waist/hip ratio, body mass index, total caloric intake, fasting insulin level, and serum triglycerides. The Ho score was inversely related to education and high-density lipoprotein cholesterol concentration. Path analysis also suggested that the effects of hostility on insulin, triglycerides, and high-density lipoprotein cholesterol were mediated by its effects on body mass index and waist/hip ratio, which, in turn, exerted their effects on lipids and blood pressure through insulin.
CONCLUSIONS: The results are consistent with those of prior research and also suggest that, in older men, hostility may be associated with a pattern of obesity, central adiposity, and insulin resistance, which can exert effects on blood pressure and serum lipids. Furthermore, effects of hostility on the metabolic syndrome appear to be mediated by body mass index and waist/hip ratio.
Key Words: hostility • metabolic syndrome • men • lipids • insulin
Abbreviations: BMI = body mass index; CHD = coronary heart disease; DPB = diastolic blood pressure; HDL-C = high-densitylipoprotein cholesterol; Ho = Cook-Medley Hostility Scalescore; LDL-C = low-density lipoprotein cholesterol; MMPI = Minnesota Multiphasic Personality Inventory; SBP =systolic blood pressure; TRG = triglycerides; VLDL-C =very-low-density lipoprotein cholesterol; WHR = waist/hip ratio.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Not only has hostility been found to directly predict CHD and other adverse health outcomes, but there is also abundant evidence to suggest a relationship between hostility and various sociodemographic, behavioral, and physiological risk factors for these disease outcomes. For instance, studies investigating the relationship between hostility and sociodemographic variables have consistently found a pattern of relationships that mimics the pattern found between these variables and morbidity and mortality. That is, higher hostility scores have been found in nonwhites, men, and those of lower socioeconomic status (ie, lower education and income) (14, 15). Likewise, most studies investigating the associations between hostility and behavioral risk factors have found relationships in the expected direction. Hostility has been positively associated with alcohol consumption (8, 16–19), cigarette use (7, 9, 19), current smoking status and caffeine consumption (19), and caloric intake (18).
In addition to sociodemographic and behavioral risk factors, several physiological correlates of CHD, stroke, diabetes, and premature death have been investigated for their relationship with hostility. Findings in this area, however, have been less consistent. Positive relationships have been identified between hostility and WHR (18, 20), BMI (16, 19), hypertension (4, 19, 21), total cholesterol (22–24), and the ratio of total cholesterol divided by HDL-C (19). One study, however, failed to find a relationship with BMI, total cholesterol, HDL-C, or LDL-C (18). Yet, although there have been other studies that failed to find an association between hostility and one or more CHD risk factors (6, 8, 11, 12, 16, 25), no studies have reported that high levels of hostility are associated with reduced risk. Overall, the preponderance of data suggests that hostility is associated with many of the risk factors of CHD and other adverse health outcomes.
Observation of a statistical association among abdominal obesity/upper body fat distribution (usually measured by WHR), insulin resistance, hyperglycemia, dyslipidemia (ie, elevated VLDL-C and TRG levels and by low HDL-C levels), and hypertension on one hand, and their ability to independently predict atherosclerotic cardiovascular disease, stroke, non–insulin-dependent diabetes mellitus, and premature death on the other (26–31), has led to a recent medical hypothesis of a common pathogenic "metabolic syndrome" underlying these disease outcomes and premature death (26, 27, 32). The metabolic syndrome, or "Syndrome X," has come to refer to this cluster of metabolic disorders and disease end points. The metabolic syndrome has become the focus of much recent empirical investigation into the pathogenesis of cardiovascular disease and non–insulin-dependent diabetes mellitus. Yet, although hostility clearly seems to play an important role in the development of cardiovascular disease, little empirical attention has been given to the role that hostility may play in the development of the metabolic syndrome. Ravaja et al. (32) found that high baseline aggression in male adolescents and young adults predicted elevations of serum TRG and insulin concentrations and increased BMI at 3-year follow-up examination. Vitaliano et al. (33) found that women who had high anger-out/hostility and high hassles and men who had high anger-out/hostility or high hassles had elevated fasting insulin levels. Furthermore, anger-out/hostility was positively associated with elevated fasting glucose levels in both men and women.
The present study augments prior studies (32, 33) in a number of ways. First, hostility was measured using the Cook-Medley Hostility Scale (34). This scale is the most commonly used measure of hostility and has well-established psychometric properties, facilitating comparisons across studies. Second, because hostility has been conceptualized as a multidimensional construct (35), two different methods, one statistical (36) and the other conceptual (9), were used to divide this construct into its subcomponents. Third, upper-body fat distribution was measured by the WHR, which has been associated extensively with the metabolic syndrome (30), and a more comprehensive assessment of physiological correlates of the metabolic syndrome was included. Unique aspects of this study were the simultaneous examination of associations among Ho scores and variables representing aspects of the metabolic syndrome among older males and the application of path analysis to describe more completely the structure of these associations.
The aims of the present study, then, were twofold: 1) to comprehensively examine the relationship between hostility and the most important constituents of the metabolic syndrome, including insulin resistance, hyperglycemia, upper-body fat distribution, dyslipidemia, and hypertension; and 2) to clarify and extend previous findings of an association between hostility and sociodemographic and health behavior variables. To achieve these aims, we analyzed data obtained in the Normative Aging Study, which offers unique opportunities for investigation of the relationship between hostility and the metabolic syndrome because of its large sample size and extensive range of sociodemographic, behavioral, and physiological measures.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
For inclusion in the present analyses, participants had to have completed the MMPI (41). The MMPI was completed in 1986 by 1548 participants. In addition, only participants who were examined between the years 1987 and 1991, during which time serum insulin and WHR measures were collected, were included in the present analyses. The final sample consisted of 1081 men.
Procedures
On the night before examination, participants refrained from eating or drinking after midnight and refrained from smoking after 8:00 PM. The examination included blood pressure measurement, blood work (serum levels of glucose, insulin, and lipids), anthropometric evaluation, and assessment of health behaviors (diet, alcohol intake, and smoking) by standardized questionnaires. Blood was drawn at 8:00 AM while the participant was fasting. Sociodemographic data, including educational attainment, were obtained on entry into the study.
Measures
Blood lipids.
Serum samples were drawn the morning after an overnight fast and analyzed for total cholesterol, HDL-C, TRG, and (calculated) LDL-C. Serum cholesterol was assayed enzymatically (SCALVO Diagnostics, Wayne, NJ). The HDL-C fraction was measured in the supernatant after precipitation of the LDL-C and VLDL-C fractions with dextran sulfate and magnesium, using the Abbott Biochromatic Analyzer 100 (Abbott Laboratories, South Pasadena, CA). TRG concentration was measured using the Dupont ACA discrete clinical analyzer (Dupont Company, Biomedical Products Department, Wilmington, DE). LDL-C concentration was estimated using the formula of Friedewald et al. (42).
Fasting blood glucose.
Serum glucose concentration was measured in duplicate on an autoanalyzer by the hexokinase method (43).
Fasting insulin.
Serum insulin concentration was determined by a solid phase [125I]-radioimmunoassay (Diagnostic Products Corporation, Los Angeles, CA).
Blood pressure.
Blood pressure was measured using a standard mercury sphygmomanometer with a 14-cm cuff. SBP and fifth-phase DBP were measured to the nearest 2 mm Hg. Both left and right arm pressures were measured with the subject sitting; right arm pressures were then taken with the subject in a supine position, and 30 seconds later a second reading of right arm pressures was taken with the subject standing. The palpatory method was used to check auscultatory systolic readings. The means of all systolic and all diastolic readings were used in analyses. There were no methodological differences in assessing blood pressure from one examination to another.
Medication use.
During the laboratory examination, current use of prescription and over-the-counter medication was assessed by the examining physician. In addition, diagnosed conditions were also noted (eg, hypertension and hypercholesterolemia).
Body mass index.
Weight was taken on a standard hospital scale with the participant dressed in undershorts and socks. Weight was measured to the nearest 0.5 lb and then converted to kilograms. Height was measured with the participant standing in bare feet against a wall to the nearest 0.1 inch and then converted to meters. Body mass index was computed as kilograms per squared meters (kg/m2).
Waist-to-hip ratio.
With the participant standing, abdomen circumference was measured in centimeters at the level of the umbilicus, and hip circumference was measured in centimeters at the greatest protrusion of the buttocks. WHR was calculated as abdomen circumference divided by hip circumference.
Health behaviors.
Behavioral risk factors assessed included alcohol and tobacco consumption and diet. Dietary data were obtained by means of a semiquantitative food frequency questionnaire (44), which was mailed to each participant and completed before the examination. The food frequency questionnaire lists food items with serving sizes and elicits information on frequency of intake during the past year. Nutrient scores are computed by multiplying the frequency of intake by the nutrient content of the food items. Macronutrients examined in the present analyses were total energy intake (kcal/day) and alcohol (drinks/week). Information was obtained on number of cigarettes currently smoked per day. Smoking status was categorized into never or ex-smokers vs. current smokers (
1 cigarettes/day).
Hostility.
Hostility was measured with the Cook-Medley Hostility Scale (34) taken from the MMPI. Form AX (41) of the MMPI was administered, which includes items from both the MMPI and MMPI-2. Nine different scores were derived from the Cook-Medley scale, including a total hostility score (Ho), scores for paranoid alienation and cynicism based on the factor structure of Costa et al. (36), and scores for the six subcategories of hostility (cynicism, hostile attributions, hostile affect, aggressive responding, social avoidance, and other) based on the scheme of Barefoot et al. (9).
Demographic risk factors.
Education was divided into four categories: less than high school, high school graduate (including attainment of a general education diploma), some college or college graduate (2 years of technical school or 4 years of college), and postcollege (some postgraduate or postgraduate). Age (in years) was assessed at the time of the laboratory examination.
Data Analysis
Data analysis was conducted using the following strategy. First, bivariate relationships among the variables were examined using correlational procedures (Spearman’s r). Next, based on the initial bivariate results and hypotheses concerning whether hostility was directly related to variables representing the metabolic syndrome or whether the influence of hostility on the metabolic syndrome was mediated by BMI and WHR, multivariate relationships were explored using multiple linear regression. Finally, on the basis of these initial results, a path model was developed in which Ho indirectly predicts TRG, HDL-C, SBP, and DBP through the mediating effects of BMI, WHR, and fasting insulin. This model was then constructed and evaluated using structural equation modeling procedures (45). Structural equation modeling offers a unique opportunity to examine simultaneously the relationship among multiple variables, to estimate path coefficients or relative weights of the paths among the variables, and to test the directionality of such paths.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
1 drinks/year). Participants reported smoking an average of 2.7 cigarettes/day and drank a mean of 10.5 alcoholic beverages/week.
|
|
|
|
|
On the basis of these initial regression results, a path model was developed in which Ho indirectly predicts TRG, HDL-C, SBP, and DBP through the mediating effects of BMI, WHR, and fasting insulin (see Figure 1). 1 Fasting insulin was hypothesized, on the basis of prior empirical and theoretical work (30), to mediate the effects of BMI and WHR on HDL-C and TRG. An initial model was tested and then revised by dropping paths that were not significant. The final model is presented in Figure 1. Parameter estimates were calculated using maximum likelihood methods. F1 and F2 represent latent constructs that underlie two variable pairs (TRG with HDL-C and SBP with DBP) and indicate that the variables are correlated. The variances of the F1 and F2 factors were fixed at 1 so that the fixed path from each factor to its measured variable indicator could be set free to be estimated (45). The overall fit of the model was good (
2(12) = 50.23, p < .001; comparative fit index = 0.957).
|
140 mg/dl. In both instances, the parameter estimates and model fit did not differ materially from the final model, which included diabetic men. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Also consistent with prior studies were the findings that total Ho scores were associated positively with both BMI (16, 19) and WHR (18, 20). Furthermore, we were able to demonstrate that Ho was not significantly associated with serum total cholesterol or with serum LDL-C but was associated with higher serum TRG, a higher lipid ratio, and lower serum HDL-C. Although the relationships of hostility and serum lipid levels have been inconsistent in the literature as a whole, these findings are consistent with results from several earlier studies. For example, at least three studies found no relationships between Ho and total cholesterol or LDL-C (4, 8, 18). Siegler et al. (19) observed a positive association between Ho and the lipid ratio, and Ravaja et al. (32) found that a baseline measure of aggression predicted elevations in serum TRG 3 years later.
Another contribution of this study was the simultaneous examination of Ho subscales and their relationships with sociodemographic, behavioral, anthropometric, and physiological variables representing the metabolic syndrome. Consistent with the findings reported by Scherwitz et al. (15), we found inverse associations for Barefoot et al.’s subscales of cynicism, hostile attributions, and aggressive responding with education; however, we failed to find a significant relationship between hostile affect and education. The cynicism subscale derived by Costa et al. (36) was also inversely related to years of education. Regarding the behavioral variables, Costa et al.’s cynicism subscale and Barefoot et al.’s hostile affect and aggressive responding subscales were associated positively with caloric intake. However, Barefoot et al.’s cynicism and hostile attributions subscales were not associated with calories consumed. Of note, correlations of the five subscales with demographic and behavioral variables were largely of similar magnitude. No one subscale stood out as particularly associated with these variables except aggressive responding, which was positively associated with cigarette use and alcohol consumption, in addition to education and total caloric intake.
Except for social avoidance and other, the remaining hostility subscales behaved similarly to the total Ho score with regard to their relationships with BMI and WHR. That is, higher scores on these subscales were associated with greater BMI and greater WHR. Correlations of the subscales with anthropometric and physiological variables varied to some degree but were generally similar in magnitude and direction compared with the total Ho score. Only hostile affect stood out, because it was consistently unrelated to all physiological measures of the metabolic syndrome.
We also replicated earlier observations of a bivariate association between hostility and fasting insulin concentrations (32, 33). Our study, however, examined the hostility-insulin relationship using Ho, which has been shown to be related to CHD and total mortality in several studies (1, 2, 4, 6, 9, 14), whereas previous studies have used a measure of aggression (32) and a measure of anger-out proneness (33) as proxies for hostility. Moreover, in the latter study, the relationship between anger-out proneness and insulin remained significant after controlling for BMI. Thus, more work is needed to determine the degree to which related, but not entirely overlapping, constructs of hostility and anger expression may be associated with insulin and other aspects of the metabolic syndrome, independent of their relationships with obesity and body fat distribution. At least two studies have reported that higher hostility is associated with increased incidence of hypertension (4, 21). We failed, however, to find a relationship between hostility and either SBP or DBP. Similarly, although higher hostility was found to be associated with elevated fasting glucose levels in a recent study (33), we did not confirm this finding.
The most important contribution of this study was the construction of a multivariate path model, which demonstrated that Ho does not have direct effects on the physiological variables of the metabolic syndrome. Rather, its effects are mediated by BMI and WHR. Specifically, Ho was found to have indirect effects on fasting insulin by way of both BMI and WHR and to predict indirectly both serum TRG and HDL-C by way of BMI. Hostility, however, was more strongly associated with BMI compared with WHR. Therefore, BMI may be relatively more important in determining the indirect relationships between Ho and other variables which define the metabolic syndrome. Furthermore, BMI was found to exert its influence on TRG, HDL-C, SBP, and DBP both directly and indirectly by way of fasting insulin; WHR, on the other hand, exerted its effects on TRG, HDL-C, SBP, and DBP only indirectly through its effects on fasting insulin. These relationships were not influenced by age, educational level, caloric intake, use of antihypertensive medications, or diabetes or clinically elevated glucose concentrations. That the relationships among Ho and other variables in the path model were not affected by excluding diabetics suggests that these associations fall more along a continuum rather than being influenced by development of a diabetic state. Others have noted consistent, positive relationships among the variables that define the metabolic syndrome (fasting insulin, TRG, HDL-C, and blood pressure) among adults and even children who are not diabetic, dyslipidemic, or hypertensive (46).
The path model suggests that insulin is antecedent to and exerts its effects on blood pressure and serum lipid values. This directional relationship is consistent with what is currently known about the pathogenesis of the metabolic syndrome, namely, that fasting insulin concentrations, which may signal insulin resistance, drive hypertension and dyslipidemia (eg, Refs. 29, 32 and 47–49).
Some limitations to this study must be noted. First, although Ho was measured 1 to 4 years before the anthropometric and physiological variables, the data were analyzed cross-sectionally; therefore, causality cannot be inferred. Second, the sample used in this study was drawn from an older, male, predominantly white population. Therefore, generalizability to other populations, such as females, younger adults, or nonwhites may be limited. Third, the magnitude of the associations between Ho and demographic, behavioral, and physiological variables may be viewed as small; however, this is generally consistent with the results of other studies (eg, Ref. 20). Thus, the clinical implications of the findings (eg, whether hostile men who keep their BMI and WHR low may decrease their health risk) are not immediately apparent. It is possible, though, that there exist subgroups of high hostility men for whom these relationships are stronger and for whom CHD risk factors may be reduced in a clinically significant way by weight loss and reduction in the WHR. Finally, we did not include measures of perceived stress, stress hormones, other CHD risk factors, and morbidity or mortality outcomes, the significance of which is discussed below.
The results of this study may have implications for interpreting findings from other studies, past and future, which examine the association between hostility, CHD risk factors, and CHD morbidity and mortality. The finding that Ho does not have direct effects on metabolic dysfunction raises questions about the relationship of Ho with CHD risk factors and CHD. Our findings suggest that if one controls for the influence of BMI and WHR, the association observed between hostility and CHD risk factors may be attenuated or eliminated. However, this may not be true in every instance (eg, Ref. 33), and further work is needed to explore in which populations and under what conditions these relationships hold. One could also interpret the findings of the path model to suggest that Ho will not be related to CHD morbidity or mortality if one adjusts for CHD risk factors. This interpretation should be viewed with caution because the results of the present study portrayed cross-sectional associations, and we did not assess CHD morbidity or mortality. However, longitudinal studies that did control for CHD risk factors (eg, lipids) nevertheless found positive associations between Ho and CHD morbidity and mortality (4, 5, 8). Moreover, it is entirely possible that Ho predisposes toward CHD not only through its associations with risk factors but also independently through other mechanisms (eg, cardiac arrhythmia; imbalance in sympathetic and parasympathetic nervous system activity; cardiovascular, endocrine, and neuroendocrine responses to stress; coronary artery vasospasm; and clotting factors). The question also remains whether Ho is causally related to CHD or whether it is just a correlate of metabolic dysfunction. Prospective studies examining Ho in relation to the development of metabolic dysfunction, other CHD risk factors, and incidence of CHD are needed to answer these questions.
Future studies should focus on the mechanisms underlying the observed relationships of hostility with BMI and WHR. A number of studies have revealed associations of centrally distributed body fat with socioeconomic, psychosocial, and behavioral correlates of low socioeconomic status (20, 50–54). These findings have led Björntorp (28) to hypothesize that abdominal fat distribution is the end result of a series of physiological responses to psychosocial stress. The chronic stress associated with low socioeconomic status leads to chronic stimulation of the adrenal-cortical system, causing elevated levels of adrenal corticosteroids, which in turn direct the storage of fat to central adipose tissue depots. Related findings come from studies observing associations between WHR and depression and anxiety symptoms (55, 56). The observed relationship between hostility and WHR suggests that hostility may be part of the cognitive/emotional/behavioral response to the chronic stress of low socioeconomic status. Therefore, future studies may do well to incorporate measurements of socioeconomic, psychosocial, and behavioral correlates of low socioeconomic status, as well as measurement of perceived stress and stress hormones, in an attempt to elucidate psychosocial and physiological mechanisms for the relationship between hostility, obesity, and distribution of body fat. For example, stressful challenges may potentiate physiological reactions in individuals with high Ho scores. Examination of interactions with socioeconomic status, race, and sex is also in order, given some evidence that associations of BMI and WHR with psychosocial factors differ across race and sex groups (20). Research should also consider the possibility that hostility might be associated with clinical or subclinical eating disorders (57), increasing BMI and WHR, given that hostility appears to be associated with greater caloric intake (although, in this study, caloric intake did not mediate the association between Ho and BMI or WHR).
In conclusion, this study replicated previous findings of a negative association between Ho and education, a positive association between Ho and total caloric intake, and positive correlations between Ho and both BMI and WHR. Furthermore, Ho was found to be related to constituents of the metabolic syndrome, in particular higher serum TRG, a higher lipid ratio, and lower serum HDL-C. However, multivariate analyses demonstrated that Ho was associated with the metabolic syndrome variables only indirectly through its influence on BMI and WHR.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Received for publication July 21, 1997.
Revision received July 12, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. Ohira, A. Hozawa, C. Iribarren, M. L. Daviglus, K. A. Matthews, M. D. Gross, and D. R. Jacobs Jr. Longitudinal Association of Serum Carotenoids and Tocopherols with Hostility: The CARDIA Study Am. J. Epidemiol., January 1, 2008; 167(1): 42 - 50. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Zhang, R. Niaura, J. R. Dyer, B.-J. Shen, J. F. Todaro, J. M. McCaffery, A. Spiro III, and K. D. Ward Hostility and Urine Norepinephrine Interact to Predict Insulin Resistance: The VA Normative Aging Study. Psychosom Med, September 1, 2006; 68(5): 718 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Knox, G. Weidner, A. Adelman, C. M. Stoney, R. C. Ellison, and for the Investigators of the National Heart, Lung, Hostility and Physiological Risk in the National Heart, Lung, and Blood Institute Family Heart Study Arch Intern Med, December 13, 2004; 164(22): 2442 - 2447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Dougherty, S. L. Rauch, T. Deckersbach, C. Marci, R. Loh, L. M. Shin, N. M. Alpert, A. J. Fischman, and M. Fava Ventromedial Prefrontal Cortex and Amygdala Dysfunction During an Anger Induction Positron Emission Tomography Study in Patients With Major Depressive Disorder With Anger Attacks Arch Gen Psychiatry, August 1, 2004; 61(8): 795 - 804. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. Kaplan What's Wrong with Social Epidemiology, and How Can We Make It Better? Epidemiol. Rev., July 1, 2004; 26(1): 124 - 135. [Full Text] [PDF]
|
|
|
|
|
|
L. S. Kinder, M. R. Carnethon, L. P. Palaniappan, A. C. King, and S. P. Fortmann Depression and the Metabolic Syndrome in Young Adults: Findings From the Third National Health and Nutrition Examination Survey Psychosom Med, May 1, 2004; 66(3): 316 - 322. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Wilson, J. L. Bienias, C. F. Mendes de Leon, D. A. Evans, and D. A. Bennett Negative Affect and Mortality in Older Persons Am. J. Epidemiol., November 1, 2003; 158(9): 827 - 835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. B. Williams, J. C. Barefoot, and N. Schneiderman Psychosocial Risk Factors for Cardiovascular Disease: More Than One Culprit at Work JAMA, October 22, 2003; 290(16): 2190 - 2192. [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. Sofhauser Psychosocial Antecedents of Hostility in Persons with Coronary Heart Disease J Holist Nurs, September 1, 2003; 21(3): 280 - 300. [Abstract] [PDF]
|
|
|
|
|
|
I. C. Siegler, P. T. Costa, B. H. Brummett, M. J. Helms, J. C. Barefoot, R. B. Williams, W. G. Dahlstrom, B. H. Kaplan, P. P. Vitaliano, M. Z. Nichaman, et al. Patterns of Change in Hostility from College to Midlife in the UNC Alumni Heart Study Predict High-Risk Status Psychosom Med, September 1, 2003; 65(5): 738 - 745. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B.-J. Shen, J. F. Todaro, R. Niaura, J. M. McCaffery, J. Zhang, A. Spiro III, and K. D. Ward Are Metabolic Risk Factors One Unified Syndrome? Modeling the Structure of the Metabolic Syndrome X Am. J. Epidemiol., April 15, 2003; 157(8): 701 - 711. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Surwit, R. B. Williams, I. C. Siegler, J. D. Lane, M. Helms, K. L. Applegate, N. Zucker, M. N. Feinglos, C. M. McCaskill, and J. C. Barefoot Hostility, Race, and Glucose Metabolism in Nondiabetic Individuals Diabetes Care, May 1, 2002; 25(5): 835 - 839. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. P. Vitaliano, J. M. Scanlan, J. Zhang, M. V. Savage, I. B. Hirsch, and I. C. Siegler A Path Model of Chronic Stress, the Metabolic Syndrome, and Coronary Heart Disease Psychosom Med, May 1, 2002; 64(3): 418 - 435. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Harper, J. Lynch, W.-L. Hsu, S. A Everson, M. M Hillemeier, T. E Raghunathan, J. T Salonen, and G. A Kaplan Life course socioeconomic conditions and adult psychosocial functioning Int. J. Epidemiol., April 1, 2002; 31(2): 395 - 403. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Bardwell, M. G. Ziegler, and J. E. Dimsdale Influence of Cholesterol and Fasting Insulin Levels on Blood Pressure Reactivity Psychosom Med, July 1, 2000; 62(4): 569 - 575. [Abstract] [Full Text] [PDF]
|
|
| ||||||
