PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Prev Cardiol. Author manuscript; available in PMC Feb 9, 2012.
Published in final edited form as:
PMCID: PMC3276243
NIHMSID: NIHMS348271
Risk Factors for Subclinical Carotid Atherosclerosis among Current Smokers
Heather M. Johnson, MD, Megan E. Piper, PhD, Douglas E. Jorenby, PhD., Michael C. Fiore, MD, MPH, Timothy B. Baker, PhD, and James H. Stein, MD
University of Wisconsin School of Medicine and Public Health; Madison, WI
Address for Correspondence: James H. Stein, MD University of Wisconsin School of Medicine and Public Health 600 Highland Avenue, G7/341 CSC (MC 3248) Madison, WI 53792 Phone: (608) 263-9648 Fax: (608) 263-0405 ; jhs/at/medicine.wisc.edu
This study characterized the determinants of carotid atherosclerosis in a large, contemporary sample of current smokers. Associations between risk factors, carotid intima-media thickness (CIMT) and carotid plaque presence were determined by multivariable regression. Subjects included 1,504 current smokers (58% female) who were a median (interquartile range) of 44.7 (38–53) years old and smoked 25 (15–40) pack-years; 55% had plaque. Pack-years, age, male sex, non-white race, body-mass index, systolic blood pressure, small low-density lipoproteins (LDL), and total high-density lipoproteins were independently associated with CIMT (model R2=0.434, p<0.001). Pack-years (OR 1.14 per 10 pack-years, p=0.001), age (OR 1.75 per 10 years, p<0.001), body-mass index (OR 0.91 per 5 kg/m2, p =0.035), and small LDL (OR 1.11 per 100 nmol/L, p<0.001), were independently associated with carotid plaque presence (model X2=210.7, p<0.001). The association between pack-years and carotid plaque was stronger in women (OR 1.09 per 10 pack-years, pinteraction=0.018).
Keywords: atherosclerosis, carotid, arteries, lipoproteins, smoking
Epidemiological studies have established a strong, direct relationship between cigarette smoking and the development of cardiovascular disease (CVD).1 Vasomotor dysfunction, vascular inflammation, and lipid oxidation are key pathways by which cigarette smoke influences the initiation and progression of atherosclerotic vascular disease.2,3 Carotid intima-media thickness (CIMT) and carotid plaque are non-invasive measures of atherosclerotic vascular disease that also predict CVD events.4,5 Smoking is associated with higher CIMT.68 Cigarette smoking independently increases atherosclerotic burden and acts synergistically with other risk factors to increase CVD risk.1 However, there is a paucity of data as to which risk factors are the most important predictors of atherosclerosis among smokers and how they interact with each other. It also is not known if smoking-related parameters such as the duration smoking, degree of nicotine dependence, and patient's personality and affect (mood) influence atherosclerotic burden in current smokers.
The purpose of this study was to identify and characterize the determinants of carotid atherosclerosis in a large cross-sectional sample of current smokers participating in a longitudinal study of the effects of continued smoking and smoking cessation on CVD risk factors and atherosclerosis progression.
Study Participants and Design
The institutional review board at the University of Wisconsin School of Medicine and Public Health approved this study. All subjects provided written informed consent. Subjects were participants in a longitudinal, randomized, double-blind, placebo-controlled smoking cessation trial to evaluate the efficacy of several smoking cessation pharmacotherapies and examine the natural history of continued smoking and smoking cessation.9 Each subjects' consent for the randomized clinical trial included permission to evaluate the physiological effects of active cigarette use and smoking cessation on atherosclerosis by ultrasound evaluation of CIMT and carotid plaque presence, at baseline and after three years. As the long-term study is ongoing, this manuscript describes the independent risk factors for CIMT and plaque burden in current smokers prior to randomization to the smoking cessation intervention. Inclusion and exclusion criteria were based on participation in the clinical trial. Inclusion criteria were: age ≥18 years old, current smoking of ≥10 cigarettes/day for the previous 6 months, an expired carbon monoxide level of >9 ppm, and stated motivation to try to quit smoking. Major exclusion criteria were uncontrolled hypertension (>160/100 mmHg) and myocardial infarction in the previous 4 weeks. Details of other exclusion criteria are described in the smoking cessation efficacy study.9
All subjects were recruited from communities in and around Madison and Milwaukee, Wisconsin. Adult smokers were recruited via television, radio and newspaper advertisements, flyers, and earned media including press conferences and interviews from January, 2005 to June, 2007. At the baseline visit, anthropometric data, fasting laboratory tests, validated questionnaires, and interviews were completed. Blood pressures were taken with calibrated mercury sphygmomanometers after subjects were seated for a minimum of five minutes. The bare upper right arm was used unless contraindicated. The appropriate size cuff was placed and blood pressures were obtained by trained personnel using standardized procedures. The waist circumference was measured with subjects standing, feet shoulder-width apart. A plastic measuring tape was placed horizontally around the bare waist, at the level of the iliac crest, with measurements taken at the end of normal exhalation by trained technicians using a standardized protocol. Advanced lipoprotein testing was performed via nuclear magnetic resonance spectroscopy by LipoScience, Inc. (Lipoprofile-2, LipoScience, Inc., Raleigh, NC).10 A total of 9 subclasses were measured. Low-density lipoproteins (LDL) and high-density lipoproteins (HDL) were defined as follows: large LDL (21.2–23.0 nm), small LDL (18.0–21.2 nm), large HDL (8.8–13.0 nm), medium HDL (8.2–8.8 nm), and small HDL (7.3–8.2 nm).
Carotid Ultrasonography
Before the initial quit attempt, baseline digital images of the right and left common carotid, bulb, and internal carotid artery segments were acquired using a high resolution linear array transducer (L10–5) and a cardiovascular ultrasound system (CV70, Siemens Medical Solutions, Mountain View, WA). The CIMT imaging protocol has been described previously.1112 Images were transferred via the Internet to a secure web server at the University of Wisconsin Atherosclerosis Imaging Program, the core ultrasound laboratory. All scanners were trained and certified by the core lab.
The mean far wall CIMT of the distal 1 cm of each common carotid artery was measured in triplicate at the ECG R-wave by a single reader using a semi-automated border detection program. Images were obtained on each side from 3 angles of interrogation.13 Longitudinal and cross-sectional images of the common carotid, bulb, and internal carotid artery segments were evaluated for the presence or absence of plaque (focal thickness of the intimal-medial layer of ≥1.2 cm).5 The coefficient of variation for repeatability of mean common carotid artery measurements in this study was 1.7%.
Statistical Analysis
Analyses were performed with SPSS software (SPSS, Inc., Chicago, IL). The average of the mean right- and left-sided far wall CIMT measurements of the common carotid artery was used to define the mean composite CIMT (“mean-mean”). Cardiovascular risk factors and CIMT were described by medians and interquartile ranges, unless noted otherwise. Multivariable linear regression models were created to determine associations between CVD risk factors and CIMT. Multivariable logistic regression models were used to identify independent risk factors for carotid plaque presence and to calculate odds ratios (OR) and 95% confidence intervals. Age, sex, race, and smoking burden (pack-years) were included in all models of CIMT and plaque presence. Other variables that were evaluated included body-mass index (BMI), waist circumference, total cholesterol, high-density lipoprotein cholesterol (HDL-c), triglycerides, low-density lipoprotein cholesterol (LDL-c), total/HDL-c ratio, glucose, systolic blood pressure, diastolic blood pressure, hemoglobin A1C, high-sensitivity C-reactive protein, serum creatinine, total LDL particles, small LDL particles, mean LDL particle size, total HDL particles, small HDL particles, mean HDL particle size, current use of anti-hypertensive and/or lipid-lowering medication, serum cotinine level, alcohol consumption (number of drinks per month), and three psychometric evaluations: the Fagerström Test of Nicotine Dependence, Positive and Negative Affect Schedule and the Multidimensional Personality Questionnaire.1416 Because of collinearity, separate models were used to determine which smoking-related parameters and which LDL- and HDL-related cholesterol and lipoprotein measures were most strongly associated with CIMT and carotid plaque presence; the strongest associations were included in the final models. Interactions between smoking burden and the significant variables in the final models for CIMT and plaque presence were formally tested. Evaluation of histograms of residuals, residual plots versus predicted values, and normal quartile-quartile plots (Q-Q plots) did not demonstrate significant violations of linear regression assumptions of errors and linearity.
Subject Characteristics
Subject characteristics are provided in Table 1. There were 1,504 current smokers in this study (58% female, 84% white, 14% African-American, 2% American Indian/Alaskan/Asian/Pacific). The median age was 45 (37.8–53) years old. Subjects smoked approximately one pack of cigarettes daily with a total smoking burden of 25 (15–40) current pack-years. Men [30 (16.2–45) pack-years] had a significantly greater smoking burden than women [23 (14.2–35) pack-years] (p<0.001). The median CIMT was 0.700 (0.633–0.778) mm which was higher than expected compared to the general population.17 55% had carotid plaque. The cohort had low HDL-c and an elevated total/HDL-c ratio, with increased small LDL particles. In this cohort, 529 (35.6%) subjects had metabolic syndrome. The presence of plaque among subjects with metabolic syndrome (39.1%) was significantly higher than subjects without metabolic syndrome (31.4%), [X2=9. 1, p=0.002]. The mean (standard deviation) Framingham Risk Score was 5.6 (0.1) %/10 years. The FRS in subjects with plaque [6.7%/10 years] was significantly higher than subjects without plaque [4.2%/10 years], p<0.001.
Table 1
Table 1
Subject Baseline Characteristics
Risk Factors Associated with Higher CIMT
The baseline and all subsequent models included age, sex, and race. Separate multivariable regression models were created to determine which smoking variables were most strongly associated with CIMT. The smoking-related variables evaluated were: number of cigarettes smoked per day (p=0.266), peak pack-years (peak cigarette packs per day * years smoked; p=0.660), current pack-years (current cigarette packs per day * years smoked; p<0.001), Fagerström Test of Nicotine Dependence score (p=0.212) and serum cotinine (p=0.170). Of the smoking-related variables, current pack-years had the strongest independent associations with the markers of subclinical atherosclerosis and was used in subsequent models. Male sex (p<0.001) and non-white race (p<0.001) were associated with CIMT, as was BMI (p<0.001), as opposed to waist circumference (p=0.115). Fasting glucose (p=0.119) and hemoglobin A1C (p=0.368) were not independently associated with CIMT. Systolic and diastolic blood pressures were analyzed in multivariable models and demonstrated a collinear relationship (r=0.68, p<0.001). Systolic blood pressure had a stronger association with CIMT (r=0.34, p<0.001) than diastolic blood pressure (r=0.15, p<0.001), so systolic blood pressure was used in all models. Alcohol consumption was not associated with CIMT (p=0.325). The Positive and Negative Affect Schedule (positive p=0.694; negative p=0.072) and the Multidimensional Personality Questionnaire analyses (positive emotionality p=0.696; negative emotionality p=0.565; constraint p=0.894) were not associated with CIMT.
In regard to lipids, HDL-c (p<0.001), triglycerides (p=0.005), and the total/HDL-c ratio (p<0.001) were independently associated with CIMT, but total cholesterol (p=0.835) and LDL-c (p=0.090) were not. Among the lipoprotein measurements, small LDL (p=0.006) and total HDL particles (p=0.002) were more associated with CIMT than total LDL particles (p=0.620) and small HDL particles (p=0.389). Mean LDL (p=0.077) and HDL particle diameters (p=0.094) were not independently associated with CIMT. Triglycerides were not independently associated with CIMT in any model that included HDL-c or HDL particles. The final model for CIMT is in Table 2. A model that included HDL-c rather than total HDL particles had the same contributors and an equivalent adjusted R2. The models were not notably changed after adjustment for use of lipid-lowering (n = 168, 11.1%) and anti-hypertensive (n = 205, 13.6%) medications. Interaction testing was performed between each of the significant variables and smoking burden (current pack-years). However, no significant interactions were identified.
Table 2
Table 2
Independent Risk Factors for Carotid Intima-Media Thickness (model R2 = 0.434, p<0.001)
Risk Factors for Carotid Plaque Presence
In the best multivariable model (model χ2 =210.7, p<0.001), variables that were independently associated with carotid plaque presence were age (OR 1.75 [1.53–2.00] per 10 years, p<0.001), BMI (OR 0.91 [0.83–0.99] per 5 kg/m2, p =0.035), small LDL particles (OR 1.11 [1.08–1.13] per 100 nmol/L, p<0.001), and smoking burden (OR 1.14 [1.05–1.23] per 10 pack-years, p=0.001) (Table 3a).
Table 3a
Table 3a
Independent Risk Factors for Carotid Plaque Presence (model X2 = 210.7, p<0.001)
Additional analyses were performed to evaluate the apparent, but unsuspected inverse relationship between BMI and carotid plaque presence. We observed that the prevalence of carotid plaque was consistent across quintiles of BMI (data not shown). BMI was significantly correlated with small LDL particles (r = 0.215, p<0.001), which also predicted plaque presence. These findings suggest that the weak, inverse relationship between BMI and plaque presence most likely is confounded by small LDL particles. The presence of metabolic syndrome (p=0.969) and the Framingham Risk Score were not independent predictors of carotid plaque presence (p=0.103).
Interaction testing between smoking burden and the other significant variables (model χ2 = 216.3, p<0.001) demonstrated that the effect of pack-years on carotid plaque presence was stronger in women (OR: 1.09 [1.01–1.18] per 10 pack-years, pint= 0.018) than men (Table 3b). There was no interaction between age and sex (pint = 0.257). Addition of an interaction term for age and sex in the model did not significantly affect the observed interaction between pack-years and female sex, demonstrating that the sex difference in the effect of plaque of sex on carotid plaque was not due to age differences. No other significant interactions were identified.
Table 3b
Table 3b
Significant Interactions of Carotid Plaque Risk Factors (model X2 = 216.3, p<0.001)
In this study of over 1,500 active smokers, both carotid plaque and CIMT were associated with increasing age, BMI, small LDL, and pack-years of smoking. The association between smoking burden and carotid plaque presence was stronger in women than in men. Increasing CIMT additionally was associated with male sex, non-white race, systolic blood pressure, and total HDL. The relationship between smoking burden (pack-years) and higher CIMT is consistent with a previous report from the Atherosclerosis Risk in Communities (ARIC) study.6 However, data regarding smoking in ARIC were ascertained two decades ago, participants in the ARIC study on average were a decade older than in our study, and were in a narrower age range (45–65 years old). We studied a modern set of active smokers across a wide range of ages (18–79 years old), and looked at the effect of smoking burden after controlling for lipoprotein measurements, which are more strongly associated with CIMT than lipids.18 Furthermore, we evaluated physiological markers of smoking such as cotinine and measures of nicotine dependence, alcohol consumption, personality, and affect disorders, which may be more common among current smokers.
Several of the findings in this study are unique. Because of the large sample size and evaluation of multiple descriptors of smoking burden, we were able to determine that of the smoking-related variables, current pack-years was most strongly associated with both carotid plaque presence and CIMT, even after controlling for the effects of age. In this cohort, current pack-years and age were the strongest predictors of both CIMT and carotid plaque presence, supporting the dose-response relationship previously observed between cigarette smoking and CVD incidence, suggesting that increased atherosclerotic burden contributes to the increased CVD risk observed in smokers.1,3
Several studies have supported the hypothesis that smoking interacts with other risk factors in a multiplicative fashion, further increasing CVD risk, even in young adults.1,19 In our study, another important finding was the interaction between smoking burden and female sex. Although men had a significantly higher smoking burden, the effect of increasing pack-years of smoking on the development of carotid plaque was stronger in women than in men. The observation that female smokers are at increased CVD risk is supported by previous research which demonstrated that smoking had a stronger association with coronary heart disease incidence in women compared to men.20 This study emphasizes the importance of targeting smoking cessation interventions and avoidance messages towards women.
Smoking is associated with low HDL-c and hypertriglyceridemia.21 However, this study demonstrated that direct lipoprotein measurements were more strongly associated with subclinical carotid atherosclerosis than their corresponding lipid measurements and lipid ratios. Small LDL and total HDL particle concentrations were independently associated with higher CIMT, and small LDL particles were independently associated with carotid plaque presence, a more advanced stage of subclinical atherosclerosis. LDL-c levels tend not to be increased in smokers, but smoking does shift the LDL particle size to the smaller, more atherogenic lipoprotein subtype.21 As expected, small LDL, rather than LDL-c, total LDL particles, or LDL size had the strongest association with carotid atherosclerosis of the LDL-related variables. In this context, BMI also was associated with increasing CIMT and carotid plaque presence. The relationship of BMI to CIMT is an expected finding, since obesity also is associated with low HDL and small LDL particles.22 Although BMI was an inverse predictor of carotid plaque presence in the multivariable model, the relationship was weak, of borderline statistical significance, and added only a minimal amount of predictive value to the regression model (adjusted R2=0.184 without BMI, adjusted R2=0.188 with BMI). Furthermore, the area under the curve for BMI was only 0.499. Because the presence of plaque was consistent across all quintiles of BMI and BMI was positively correlated with small LDL particles (r=0.215, p<0.001) which strongly predicted carotid plaque presence, we believe that the relationship between BMI and carotid plaque in smokers is confounded by small LDL particles. If small LDL particles are not present, then the negative metabolic effects of increased BMI on atherosclerosis are less likely to be seen.
Inventories regarding personality and affect were not associated with carotid atherosclerosis in our study. Previous research has not shown a direct relationship between positive and negative affect and coronary artery disease.23 However, certain emotions may be confounders of atherosclerosis development and CVD events.24
Limitations
This was a cross-sectional study of individuals who chose to participate in a smoking cessation intervention study; therefore, our findings may not be generalizable to all smokers. However, over 70% of current smokers plan to quit smoking in the next year.25 Although significant efforts were made to target recruitment of non-white subjects, non-whites comprised only approximately 16% of the study cohort. It is possible that there are racial differences in associations with carotid atherosclerosis among smokers. Data about family history of premature CVD and physical activity were not available for analysis. Uncontrolled hypertension was an exclusion criterion, which restricted the range to evaluate blood pressure and its relationship to carotid atherosclerosis. Inclusion of individuals with higher blood pressures may have demonstrated a stronger association of blood pressure with CIMT and/or plaque. Menopausal status was not known; however, 69% of the women were under 50 years old, and <1% were on estrogen therapy. Therefore, the use of hormone replacement therapy should not affect the results of this study. Finally, data were not available for biomarkers that may affect atherosclerosis among smokers, such as fibrinogen and inflammatory cytokines.
Conclusions
Among current smokers, increasing smoking burden, systolic blood pressure, BMI, small LDL, and total HDL particles are associated with the presence and extent of subclinical carotid atherosclerosis. Smoking burden was more strongly associated with carotid plaque presence among women than men. In addition to supporting the importance of smoking cessation for the prevention of CVD, this study highlights the importance of other risk factors for atherosclerosis and CVD among current smokers.
Acknowledgments
MC Fiore – Over the last three years, Dr. Fiore has served as an investigator in research studies at the University of Wisconsin that were funded by Pfizer, GlaxoSmithKline and Nabi Biopharmaceuticals. In 1998, the University of Wisconsin (UW) appointed Dr. Fiore to a named Chair funded by an unrestricted gift to UW from Glaxo Wellcome.
TB Baker – Research grants from Pfizer, GlaxoSmithKline, Nabi Biopharmaceuticals, and Sanofi.
JH Stein – Research grants from Siemens Medical Solutions and Sonosite, Inc.
Funding Source
This research was supported in part by grant P50 DA019706 from the National Institute on Drug Abuse to the University of Wisconsin Center for Tobacco Research and Intervention and grant T32 HL07936 from the National Heart Lung and Blood Institute to the University of Wisconsin Cardiovascular Research Center.
Abbreviations
ARICAtherosclerosis Risk in Communities
BMIbody-mass index
CIMTcarotid intima-media thickness
CVDcardiovascular disease
HDL-chigh-density lipoprotein cholesterol
LDL-clow-density lipoprotein cholesterol
ORodds ratio

Footnotes
Statement of Financial Disclosure HM Johnson, DE Jorenby, ME Piper – No conflicts to disclose.
1. Burns DM. Epidemiology of smoking-induced cardiovascular disease. Prog Cardiovasc Dis. 2003;46(1):11–29. [PubMed]
2. Ambrose JA, Barua RS. The pathophysiology of cigarette smoking and cardiovascular disease. J Am Coll Cardiol. 2004;43(10):1731–1737. [PubMed]
3. Benowitz N. Cigarette smoking and cardiovascular disease: Pathophysiology and implications for treatment. Progress in Cardiovascular Disease. 2003;46(1):91–111. [PubMed]
4. Lorenz MW, Markus HS, Bots ML, et al. Prediction of clinical cardiovascular events with carotid intima-media thickness. Circulation. 2007;115(4):459–467. [PubMed]
5. Wyman RA, Mays ME, McBride PE, Stein JH. Ultrasound-detected carotid plaque as a predictor of cardiovascular events. Vascular Medicine. 2006;11:123–130. [PubMed]
6. Howard G, Burke GL, Szklo M, et al. Active and passive smoking are associated with increased carotid wall thickness. The Atherosclerosis Risk in Communities Study. Arch Intern Med. 1994;154(11):1277–1282. [PubMed]
7. Johnson HM, Douglas PS, Srinivasan SR, et al. Predictors of carotid intima-media thickness progression in young men and women: The Bogalusa Heart Study. Stroke. 2007;38(3):900–905. [PubMed]
8. Tzou WS, Douglas PS, Srinivasan SR, et al. Distribution and predictors of carotid artery intima-media thickness in young adults: The Bogalusa Heart Study. Preventive Cardiology. 2007;10:181–189. [PubMed]
9. Piper ME, Smith SS, Schlam TR, et al. A randomized placebo-controlled clinical trial of five smoking cessation pharmacotherapies. Archives of General Psychiatry. In press. [PMC free article] [PubMed]
10. Jeyarajah EJ, Cromwell WC, Otvos JD. Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. Clin Lab Med. 2006;26:847–870. [PubMed]
11. Tzou WS, Korcarz CE, Aeschlimann SE, Stein JH. Use of handheld ultrasound by a nonsonographer clinician to measure carotid intima-media thickness. J Am Soc Echocardiogr. 2006;19(10):1286–1292. [PubMed]
12. Stein JH, Korcarz CE, Hurst RT, et al. Use of carotid ultrasound to identify subclinical vascular disease and evaluate cardiovascular disease risk: A consensus statement from the American Society of Echocardiography Carotid Intima-Media Thickness Task Force. J Am Soc Echocardiogr. 2008;21(2):93–111. [PubMed]
13. Stein JH, Korcarz CE, Mays ME, et al. A semi-automated border detection program that facilitates clinical use of ultrasound carotid intima-media thickness measurements. J Am Soc Echocardiogr. 2005;18(3):244–251. [PubMed]
14. Heatherton TF, Kozlowski LT, Frecker RC, Fagerstrom KO. The Fagerstrom test for nicotine dependence: A revision of the Fagerstrom Tolerance Questionnaire. Addiction. 1991;86(9):1119–1127. [PubMed]
15. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1998;54:1063–1070. [PubMed]
16. Patrick CJ, Curtin JJ, Tellegen A. Development and validation of a brief form of the Multidimensional Personality Questionnaire. Psychol Assess. 2002;14(2):150–163. [PubMed]
17. Howard G, Sharett A, Heiss G, et al. The ARIC Investigators Carotid artery intimal-media thickness distribution in general populations as evaluated by B-mode ultrasound. Stroke. 1993;24:1297–1304. [PubMed]
18. Mora S, Skzlo M, Otvos JD, et al. LDL particle subclasses, LDL particle size, and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA) Atherosclerosis. 2007;192(1):211–217. [PubMed]
19. Berenson GS, Srinivasan SR, Bao W, et al. Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults: The Bogalusa Heart Study. N Engl J Med. 1998;338(23):1650–1656. [PubMed]
20. Sharrett AR, Coady SA, Folsom AR, et al. Smoking and diabetes differ in their associations with subclinical atherosclerosis and coronary heart disease – the ARIC Study. Atherosclerosis. 2004;172:143–149. [PubMed]
21. Campbell SC, Moffatt R, Stamford B. Smoking and smoking cessation - The relationship between cardiovascular disease and lipoprotein metabolism: A review. Atherosclerosis. 2008;201:225–235. [PubMed]
22. Grundy SM. Obesity, metabolic syndrome, and cardiovascular disease. J Clin Endocrin Metab. 2004;89(6):2595–2600. [PubMed]
23. Nabi H, Kivimaki M, De Vogli R, Marmot, et al. Positive and negative affect and risk of coronary heart disease: Whitehall II Prospective Cohort Study. BMJ. 2008;337:a118. [PMC free article] [PubMed]
24. Chang PP, Ford DE, Meoni LA, et al. Anger in young men and subsequent premature cardiovascular disease. The Precursors Study. Arch Int Med. 2002;162:901–906. [PubMed]
25. Etter JF, Perneger TV, Ronchi A. Distributions of smokers by stage: International comparison and association with smoking prevalence. Prev Med. 1997;26:580–585. [PubMed]