点击显示 收起
From the Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pa.
ABSTRACT
Objective— We investigated associations between segment-specific carotid intima-medial thickness (IMT) and cardiovascular risk factors collected before menopause for insight into mechanisms of atherosclerosis development.
Methods and Results— Participants were 453 healthy women (aged 46 to 58 years) enrolled in a dietary and physical activity randomized clinical trial. Ultrasound scan measures were taken 2.7 years after baseline in the common carotid artery (CCA), bifurcation (bulb), and internal carotid artery (ICA) segments. When scanned, 84% remained premenopausal. In linear regression models adjusted for age, menopausal status, and intervention group, measures independently (P<0.05) and positively associated were as follows: baseline weight (?=0.007 per 5 kg), systolic blood pressure (SBP; ?=0.008 per 10 mm Hg), and age (?=0.02 per 5 years) with CCA IMT; smoking (?=0.08), weight (?=0.009), and SBP (?=0.02) with bulb IMT; and apoprotein B (?=0.01 per 0.1 g/L) with ICA IMT. Differential effects in a repeated measures model with all 3 IMT locations showed these risk factors to have segment-specific positive associations. The effect of weight was strongest in the CCA, smoking and SBP were specific to the bulb, and apoprotein B was strongest in the ICA segment.
Conclusion— Analyses indicate that cardiovascular risk factors may differentially affect IMT in the CCA, bulb, and ICA segments of healthy middle-aged women.
Associations between segment-specific intima-medial thickness and cardiovascular risk factors in middle-aged women were examined through ultrasound in the common, bifurcation, and internal carotid arteries. Segment-specific effects, indicating cardiovascular risk factors may differentially affect intima-medial thickness and were evident in regression models and a repeated measures model with all 3 locations.
Key Words: atherosclerosis ? women ? carotid artery ? ultrasound ? cardiovascular risk factors
Introduction
Carotid intima-medial thickness (IMT) is considered a marker of early atherosclerosis, it predicts future risk of cardiovascular disease, and it has been found to be high in individuals with coronary heart disease and myocardial infarction (IMT0.73 mm and 1 mm, respectively).1,2 Among women of all ages, cardiovascular risk factors, such as high blood pressure, body mass index (BMI), and cholesterol, are associated with average and maximum carotid IMT.3–7 However, associations with specific locations of thickening within the carotid artery are not as definitive.8–13 Considering risk factors in conjunction with specific segments of IMT may provide better insight into the mechanism of atherosclerotic development. It is likely that differences in hemodynamics, cellular processes, and the risk factors themselves affect the development and progression of disease.14–22 Additionally, each site may have its own predisposing factors,12,19 and biomechanical and biochemical pathways16,20,23,24 may be acting in combination to influence atherosclerosis.
The purpose of this report is to evaluate atherosclerotic risk factors and IMT by carotid artery segment in a sample of 453 healthy women aged 46 to 58 years. In addition to using standard linear regression models, we use a repeated measures model to provide evidence of segment-specific pathways for development of IMT. Investigating risk factor associations with all 3 segments in the same model has not been done in previous studies.
Methods
Participants
Participants were recruited from the 535 women (92% white) who participated in the Women’s Healthy Lifestyle Project, an intervention trial designed to evaluate the efficacy of diet and physical activity in preventing weight gain and elevations in other cardiovascular risk factors.25–27 Women were enrolled from 1992 to 1994, randomized to either lifestyle intervention or an assessment-only control, and followed over 54 months. Beginning in 1994, carotid ultrasound scans were offered to all participants, and were obtained in 453 women (85%).
To be eligible for enrollment, women had to be aged 44 to 50 years, premenopausal (defined as less than 3 months amenorrhea in the 6 months before the screening interview), and have a diastolic blood pressure (DBP) <95 mm Hg, BMI between 10 and 34, fasting glucose <7.77 mmol, low density lipoproteins (LDLs) between 2.07 and 4.14 mmol, and total cholesterol between 3.26 and 6.72 mmol. Exclusion criteria included hysterectomy, hormone therapy, and use of antihypertensive, lipid-lowering, insulin, thyroid, or psychotropic medications. Further details of the clinical trial are available in previously published articles.25–27
Both the clinical trial and the carotid exams were approved annually by the Institutional Review Board of the University of Pittsburgh. All participants provided written consent.
Clinical Measures
Baseline clinical measures included systolic blood pressure (SBP), DBP, weight, BMI (kg/m2), waist circumference, percent body fat (from dual energy x-ray absorptiometry scanner) and fasting total cholesterol, high density lipoproteins (HDLs), HDL2, HDL3, LDLs, triglycerides, apoprotein B, and glucose. Dietary intake (kilocalories per day) was assessed by questionnaire.28 Leisure time physical activity (kilocalories in the past week) was measured by interview.29 Serum insulin was assessed at the 6-month follow-up visit rather than at baseline. Standard enzymatic assays were used to measure total serum cholesterol, HDLs, triglycerides, and glucose; LDL was estimated by the Friedewald equation; apoprotein B was determined via turbidimetric measurement; and serum insulin was determined by radioimmunoassay. Smoking was categorized as current, former, or never. Additional details of the clinical measures are available in previous publications.25–27
At each follow-up visit and at the time of their scan, women were assessed for hormone therapy use (any in the year previous) and hysterectomy, and categorized as pre-, peri-, or postmenopausal (perimenopausal indicates missing a bleeding cycle or taking hormone therapy for 3 to 11 cycles in the last year; postmenopausal indicates no bleeding cycle or taking hormone therapy for 12 consecutive cycles, or hysterectomy). Because most women remained premenopausal at scan time, menopausal status was dichotomized as premenopausal versus peri/postmenopausal for analyses.
Carotid Ultrasound Measures
Carotid arteries were examined using an ultrasound scanner (Toshiba SSA-270A) equipped with a 5-MHz linear array imaging probe. With participant in supine position, arteries were viewed in transverse and longitudinal projections. For the common carotid artery (CCA) segment, both near and far walls were examined 2 cm proximal to the bifurcation (bulb). For the bulb area and internal carotid artery (ICA), measurements were taken of the far walls only (because near walls cannot be consistently visualized). Digitized images were used to trace the medial-adventitial and intima-lumen interfaces across 1-cm lengths and compute IMT for each segment.30 Average IMT was calculated from the mean of both the right and left carotid arteries of the 3 segments (8 locations total).
Our IMT reliability study (n=15)30 reported correlation coefficients of 0.96 between sonographers and 0.99 between readers for average IMT, and was >0.87 for individual segments across sonographers and readers.
Statistical Analysis
Descriptive measures are summarized as mean±SD unless otherwise noted. Intervention and control groups were compared using standard t tests, Wilcoxon tests, and 2 tests. Analyses are reported on the pooled population, given that risk factors were collected before the intervention and to maximize available data.
Pearson correlation coefficients were calculated between normally distributed cardiovascular risk factors and each IMT measure. Spearman correlations were calculated for nonnormal risk factors. To determine which traditional cardiovascular risk factors were independently associated with each IMT segment, variables with a P value of 0.15 in univariate correlations were included in multivariable linear regression models. More parsimonious multivariable models were identified with a stepwise selection procedure, using P<0.05 as the selection cutoff. Models were adjusted for age, intervention group, and menopausal status at the time of the scan. The 3 segment models (ie, CCA, bulb, and ICA) suggested that risk factor associations with IMT varied according to the IMT measurement location.
To formally test these observed differences, a single model using all 3 IMT measures was developed. Because the 3 IMT measures taken from each woman were not independent, a repeated measures model was used. This modeled each IMT value (ie, 3 lines of data per person: 1 for CCA IMT, 1 for bulb IMT, and 1 for ICA IMT) as a function of risk factors and a 3-level variable indicating location. The correlation between each woman’s IMT values, which is accounted for in the repeated measures model, was modeled using a compound symmetric correlational structure.31 Covariates in the repeated measures model included all significant multivariable regression predictors, IMT measurement location, and examined possible location by risk factor interaction effects. Significant interaction effects would signify segment-specific effects. All analyses were implemented using the SAS system for Windows (version 8.2; SAS Institute).
Results
The women scanned were similar in clinical characteristics to those not scanned (data not shown). Clinical characteristics for the 453 participants are presented in Table 1 and were similar across control and intervention groups.
TABLE 1. Participant Characteristics (n=453)*
Scans were done 2.7 years (range 0.7 to 7.6 years) after the baseline visit when 84% of all subjects were still premenopausal. There were no significant differences in IMT measures between treatment groups, thus the groups were combined for remaining analyses (Table 2). Mean average IMT was 0.68 mm in the pooled group; IMT was highest in the bulb and lowest in the ICA segment. For each measure, results include only participants for whom both the right- and left-sided IMT measurement was available. Those with missing IMT (n=9) were similar in available data to those included (data not shown).
TABLE 2. Intima Media Thickness
Univariate Associations
In unadjusted correlations, average IMT was positively and significantly (P<0.15) associated with typical cardiovascular risk factors of age, weight, BMI, waist circumference, percent body fat, SBP, DBP, lipid measures (ie, cholesterol, LDL, apoprotein B, and triglycerides), insulin, and physical activity, whereas HDL and HDL2 showed a negative association. Examining arterial segments individually, IMT at all 3 locations was positively associated with baseline weight, BMI, and waist circumference. Additionally, higher CCA IMT was associated with older age and higher SBP, DBP, percent body fat, apoprotein B, and insulin; higher bulb IMT was associated with higher SBP, DBP, percent body fat, lipid measures, insulin and physical activity, and lower HDL and HDL2; higher ICA IMT was associated with older age, higher lipid measures and insulin, and lower HDL. In unadjusted bivariate analyses, significant positive associations were found between average IMT and smoking, hormone therapy, and hysterectomy. By segment, significant positive associations were found between CCA IMT and smoking, hormone therapy and hysterectomy, and between bulb IMT and smoking. Correlations between IMT across the 3 locations were moderate (range r=0.32 to 0.43).
Multivariable Associations
Linear regression models were adjusted for intervention group, age, and menopausal status at scan time. Risk factors showing significant univariate correlations with IMT (listed above) were included in respective models for average IMT and each location. Significant independent predictors of greater average IMT were higher weight, SBP, age, and current smoking (Table 3). Higher weight, SBP, age, and hysterectomy were independently associated with higher CCA IMT. Current smoking, weight, and SBP were positively associated with bulb IMT. Higher apoprotein B was associated with higher ICA IMT. Comparing individual covariate percents of variation in each model (ie, adjusted R2), it was evident that weight was the largest contributor to average and CCA IMT, SBP was most prominent for bulb IMT, and apoprotein B was nearly the sole contributor to ICA IMT. Additionally, age at scan was 1.03% and 1.34% of the "other" category for average and CCA IMT, respectively (Figure).
TABLE 3. Multivariable Regression Models of Intima-Media Thickness*
Percent of variation in IMT by location. *Indicates intervention group, peri/postmenopausal status, and age at scan for IMT locations, except CCA IMT for which hysterectomy is also included in the other variance.
Thus, differential effects could be seen across arterial locations. For example, a 10mm Hg increase in SBP would produce a 0.008 mm increase in CCA IMT, a 0.022 mm increase in bulb IMT, and no significant change in ICA IMT, although these comparisons do not take into account variations in wall thickness. To test the statistical significance of these observed differences, a single repeated measures linear model using all 3 IMT measures (thus accounting for within-woman and between-women variability in IMT) was used to examine segment-specific associations.
Repeated Measures
The repeated measures model included significant multivariable predictors from the previous models (weight, SBP, smoking, apoprotein B, and hysterectomy), control variables (intervention group, menopausal status, and age), segment location (CCA, bulb, ICA), and risk factor by segment interaction terms. This was a single model with IMT as the dependent variable and multiple lines of data per woman, corresponding to each IMT segment. As expected, significant main effect associations with IMT were found across the 3 locations for weight (?=0.007 per 5 kg, P<0.001), SBP (?=0.008 per 10 mm Hg, P=0.008), current smoking (compared with never smoked ?=0.037, P=0.005), and apoprotein B (?=0.006 per 0.1 g/L, P=0.01) with a trend for age at scan (?=0.02 per 5 years, P=0.08). Results were the same for never and former smokers. Significant interaction effects (signifying segment-specific effects) were found between location and SBP (P=0.02), smoking (P=0.01), and apoprotein B (P=0.02). Specifically, simple effects estimates (Table 4) revealed a greater effect of weight and age in the CCA than in the ICA, and a greater effect for apoprotein B in the ICA than in the CCA. Smoking and SBP appeared to have an effect that was specific to the bulb segment. In addition, individual estimates showed significant segment associations between weight and apoprotein B and bulb location. There were no main or interaction effects for hysterectomy, which was likely a surrogate for age in this sample. Thus, significant and unique associations between baseline risk factors and segments exist even after accounting for the correlation in IMT at all 3 locations.
TABLE 4. Linear Repeated Measures Model of Intima-Media Thickness*
Discussion
Our research shows that cardiovascular risk factors may differentially influence atherosclerosis in carotid arterial segments in healthy middle-aged women. Further, repeated measures regression, which has not been used in previous studies of carotid segments, showed that these segment-specific associations were statistically significant when locations were directly compared.
Results of this study build on our previous findings of carotid atherosclerosis, where we reported on average IMT in a subset (n=292) of the study sample.7 Similar associations between cardiovascular risk factors and average IMT were found. Additionally, segment-specific analyses were possible with the larger sample size.
Segment-Specific Carotid IMT
Unique to our study were the examination of IMT by carotid segment in healthy, primarily premenopausal women and the examination of risk factor and location interactions. Variations in methodology, including use of older women or heart clinic patients, make others’ results of IMT by carotid segment challenging to compare.8–13 Nonetheless, similarities between our results and those previously published include positive associations of SBP and body composition with CCA and bulb IMT,10,11,13 lipids associated with bulb and ICA IMT,9,11,13 and smoking with bulb IMT.11 In contrast, neither lipids nor smoking were found to be associated with CCA IMT in our analyses.9,11,13 Nor did we find older age associated with higher IMT across segments,10 although this may be because of our narrow age range. Further, others have generally found the segments similarly associated with cardiovascular risk factors,10,11,13 whereas we found distinct differences in risk factor and segment associations. Examining more diseased populations may have made it harder for others to tease out unique associations.
Theories on Differential IMT
Various mechanical and biochemical pathways and the risk factors themselves are most certainly acting in conjunction and possibly synergistically to determine the development and progression of atherosclerosis at any given site.23,24 The moderate level of IMT we found in the CCA (compared with the ICA and bulb) is likely due to the lower sheer wall force18,19,32 that is consistent with the laminar blood flow33 in this location. Its association with higher weight is consistent with the positive relationship commonly found between body size and atherosclerosis and CCA IMT.34 Diabetes, which is associated with obesity, is thought to decrease wall shear stress and is positively associated with CCA IMT.35 Finally, aging, which itself is associated with increased CCA IMT,3,11,13,32 is also inversely related to shear stress.32
Our positive association between apoprotein B and ICA IMT is consistent with research on sheer stress and lipids regarding influences on IMT. In the ICA segment, the smaller wall thickness, relative to the other segments, is likely attributable to the mix of laminar and oscillatory shear stress around the arterial wall.19 Infiltration of lipids during extended contact of blood with the wall has been linked to areas of low shear stress,36 which portions of this segment experience. Additionally, shear stress influences cellular processes, such as increase in growth-regulatory factors,16,22 expression of inflammatory molecules,24 and response of prooxidant processes,33 which have also been linked to LDL and atherosclerosis37,38 although not specifically to the ICA segment.
Not surprisingly, the bulb, which connects the CCA and ICA segments, was associated with risk factors linked to each site. The bulb experiences the most oscillatory stress and turbulence as a result of reverse flow velocity components occurring during pulsatile flow19,32,33 and consequentially exhibited greater wall thickness and was associated with higher SBP in our sample. The higher bulb IMT may also be a function of plaque, which is commonly found in this location,39 and was evident in the bulb in 18.3% of our sample (posthoc; plaque defined as a distinct area protruding into the vessel lumen with >50% thickness than surrounding area). The influence of shear stress on the extracellular factors of hypertension, such as increased pressure,16 decreases in vasoactive agents,17 and inflammatory effects,14 would have the largest impact on the bulb segment. Research on increased residence time of blood36 and adherence of platelets and macrophages to the arterial wall,40 which influence lipid infiltration and plaque formation, is consistent with our positive association between apoprotein B and bulb IMT. The increased residence time might also make the bulb more susceptible to the toxic components of cigarette smoke.15
Several study limitations should be mentioned. First, the wide time range between baseline and scan visits may have influenced results, although 91% of participants had scans within 1 SD of the mean interval. Posthoc, we substituted "time interval" and "age at baseline" for the covariate "age at scan" in regression models. Longer time interval replaced older age at scan as a significant predictor of higher average and CCA IMT, thus the model adequately adjusted for time. Second, risk factors explained only 3% to 9% of IMT segment variance, so other atherosclerotic mechanisms and pathways should continue to be examined. However, given the healthy status, young age, and low levels of IMT in our sample, the link between specific IMT segments and risk factors are noteworthy and comparable to the 12% to 22% segment variance found by Wei et al,13 whose sample included diabetic subjects. Further, the enrollment exclusion criteria (ie, requiring normal to high-normal ranges for DBP, BMI, glucose, and lipids) likely reduced the effects of these risk factors and potentially increased the relative contribution of other risk factors (eg, hysterectomy). Third, there may have been some influence of the intervention that was not detected in our analyses because the clinical trial was successful in minimizing increased weight and LDL that traditionally occur with menopause.25–27 Lastly, we examined a healthy, homogeneous sample of women, who may not be representative of the general population and thus may have underestimated associations. Future research should include other groups for a better understanding of segment-specific associations and their influences on clinical events.
In conclusion, cardiovascular risk factors traditionally associated with overall mean carotid IMT may differentially influence wall thickening of the carotid artery segments in middle-aged healthy women, even after adjusting for other cardiovascular risk factors and wall thickness in all 3 segments.
Acknowledgments
This research was funded by American Heart Association Grant-in-Aid Award No. 0050236N and National Institutes of Health grant No. HL-45167. We gratefully acknowledge the important contributions of the staff and study participants in time, data collection, and dedication to research.
References
Chambless LE, Folsom AR, Clegg LX, Sharrett AR, Shahar E, Nieto FJ, Rosamond WD, Evans G. Carotid wall thickness is predictive of incident clinical stroke: the Atherosclerosis Risk in Communities (ARIC) study. Am J Epidemiol. 2000; 151: 478–487.
Hodis HN, Mack WJ, LaBree L, Selzer RH, Liu CR, Liu CH, Azen SP. The role of carotid arterial intima-media thickness in predicting clinical coronary events. Ann Intern Med. 1998; 128: 262–269.
Bonithon-Kopp C, Scarabin PY, Taquet A, Touboul PJ, Malmejac A, Guize L. Risk factors for early carotid atherosclerosis in middle-aged French women. Arterioscler Thromb. 1991; 11: 966–972.
Dobs AS, Nieto FJ, Szklo M, Barnes R, Sharrett AR, Ko WJ. Risk factors for popliteal and carotid wall thicknesses in the Atherosclerosis Risk in Communities (ARIC) Study. Am J Epidemiol. 1999; 150: 1055–1067.
Lassila HC, Tyrrell KS, Matthews KA, Wolfson SK, Kuller, LH. Prevalence and determinants of carotid atherosclerosis in healthy postmenopausal women. Stroke. 1997; 28: 513–517.
Matthews KA, Kuller LH, Sutton-Tyrrell K, Chang YF. Changes in cardiovascular risk factors during the perimenopause and postmenopause and carotid artery atherosclerosis in healthy women. Stroke. 2001; 32: 1104–1111.
Sutton-Tyrrell K, Lassila HC, Meilahn E, Bunker C, Matthews KA, Kuller LH. Carotid atherosclerosis in premenopausal and postmenopausal women and its association with risk factors measured after menopause. Stroke. 1998; 29: 1116–1121.
Crouse JR, III, Craven TE, Hagaman AP, Bond MG. Association of coronary disease with segment-specific intimal-medial thickening of the extracranial carotid artery. Circulation. 1995; 92: 1141–1147.
Crouse JR, Goldbourt U, Evans G, Pinsky J, Sharrett AR, Sorlie P, Riley W, Heiss G. Risk factors and segment-specific carotid arterial enlargement in the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke. 1996; 27: 69–75.
Espeland MA, Tang R, Terry JG, Davis DH, Mercuri M, Crouse JR, III. Associations of risk factors with segment-specific intimal-medial thickness of the extracranial carotid artery. Stroke. 1999; 30: 1047–1055.
O’Leary DH, Polak JF, Kronmal RA, Savage PJ, Borhani NO, Kittner SJ, Tracy R, Gardin JM, Price TR, Furberg CD. Thickening of the carotid wall. A marker for atherosclerosis in the elderly? Cardiovascular Health Study Collaborative Research Group. Stroke. 1996; 27: 224–231.
Tell GS, Howard G, McKinney WM. Risk factors for site specific extracranial carotid artery plaque distribution as measured by B-mode ultrasound. J Clin Epidemiol. 1989; 42: 551–559.
Wei M, Gonzalez C, Haffner SM, O’Leary DH, Stern MP. Ultrasonographically assessed maximum carotid artery wall thickness in Mexico City residents and Mexican Americans living in San Antonio, Texas. Association with diabetes and cardiovascular risk factors. Arterioscler Thromb Vasc Biol. 1996; 16: 1388–1392.
Chobanian AV, Alexander RW. Exacerbation of atherosclerosis by hypertension. Potential mechanisms and clinical implications. Arch Intern Med. 1996; 156: 1952–1956. Review.
Davis JW, Shelton L, Eigenberg DA, Hignite CE, Watanabe IS. Effects of tobacco and non-tobacco cigarette smoking on endothelium and platelets. Clin Pharmacol Ther. 1985; 37: 529–533.
Dzau VJ, Gibbons GH, Cooke JP, Omoigui N. Vascular biology and medicine in the 1990s: scope, concepts, potentials, and perspectives. Circulation. 1993; 87: 705–719. Review.
Glasser SP, Selwyn AP, Ganz P. Atherosclerosis: risk factors and the vascular endothelium. Am Heart J. 1996; 131: 379–384. Review.
Kornet L, Lambregts J, Hoeks AP, Reneman RS. Differences in near-wall shear rate in the carotid artery within subjects are associated with different intima-media thicknesses. Arterioscler Thromb Vasc Biol. 1998; 18: 1877–1884.
Ku DN, Giddens DP, Zarins CK, Glagov S. Pulsatile flow and atherosclerosis in the human carotid bifurcation. Positive correlation between plaque location and low oscillating shear stress. Arteriosclerosis. 1985; 5: 293–302.
Nerem RM, Alexander RW, Chappell DC, Medford RM, Varner SE, Taylor WR. The study of the influence of flow on vascular endothelial biology. Am J Med Sci. 1998; 316: 169–175. Review.
Ohno M, Gibbons GH, Dzau VJ, Cooke JP. Shear stress elevates endothelial cGMP. Role of a potassium channel and G protein coupling. Circulation. 1993; 88: 193–197.
Song RH, Kocharyan HK, Fortunato JE, Glagov S, Bassiouny HS. Increased flow and shear stress enhance in vivo transforming growth factor-?1 after experimental arterial injury. Arterioscler Thromb Vasc Biol. 2000; 20: 923–930.
Laurent S, Tropeano AI, Lillo-Lelouet A, Jondeau G, Laloux B, Boutouyrie P. Local pulse pressure is a major determinant of large artery remodelling. Clin Exp Pharmacol Physiol. 2001; 28: 1011–1014. Review.
Tegos TJ, Kalodiki E, Sabetai MM, Nicolaides AN. The genesis of atherosclerosis and risk factors: a review. Angiology. 2001; 52: 89–98. Review.
Simkin-Silverman L, Wing RR, Hansen DH, Klem ML, Pasagian-Macaulay AP, Meilahn EN, Kuller LH. Prevention of cardiovascular risk factor elevations in healthy premenopausal women. Prev Med. 1995; 24: 509–517.
Simkin-Silverman LR, Wing RR, Boraz MA, Meilahn EN, Kuller LH. Maintenance of cardiovascular risk factor changes among middle-aged women in a lifestyle intervention trial. Womens Health. 1998; 4: 255–271.
Kuller LH, Simkin-Silverman LR, Wing RR, Meilahn EN, Ives DG. Women’s Healthy Lifestyle Project: A randomized clinical trial: results at 54 months. Circulation. 2001; 103: 32–37.
Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol. 1986; 124: 453–469.
Paffenbarger RS Jr., Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978; 108: 161–175.
Thompson T, Sutton-Tyrrell K, Wildman R. Continuous quality assessment programs can improve carotid duplex scan quality. J Vasc Technol. 2001; 25: 33–39.
Verbeke G, Molenberghs, G. Linear Mixed Models for Longitudinal Data. New York: Springer-Verlag; 2000.
Carallo C, Irace C, Pujia A, De Franceschi MS, Crescenzo A, Motti C, Cortese C, Mattioli PL, Gnasso A. Evaluation of common carotid hemodynamic forces. Relations with wall thickening. Hypertension. 1999; 34: 217–221.
De Keulenaer GW, Chappell DC, Ishizaka N, Nerem RM, Alexander RW, Griendling KK. Oscillatory and steady laminar shear stress differentially affect human endothelial redox state: role of a superoxide-producing NADH oxidase. Circ Res. 1998; 82: 1094–1101.
Lakka TA, Lakka HM, Salonen R, Kaplan GA, Salonen JT. Abdominal obesity is associated with accelerated progression of carotid atherosclerosis in men. Atherosclerosis. 2001; 154: 497–504.
Irace C, Carallo C, Crescenzo A, Motti C, De Franceschi MS, Mattioli PL, Gnasso A. NIDDM is associated with lower wall shear stress of the common carotid artery. Diabetes. 1999; 48: 193–197.
Sprague EA, Steinbach BL, Nerem RM, Schwartz CJ. Influence of a laminar steady-state fluid-imposed wall shear stress on the binding, internalization, and degradation of low-density lipoproteins by cultured arterial endothelium. Circulation. 1987; 76: 648–656.
Hulthe J, Fagerberg B. Circulating oxidized LDL is associated with subclinical atherosclerosis development and inflammatory cytokines (AIR Study). Arterioscler Thromb Vasc Biol. 2002; 22: 1162–1167.
Raines EW, Ross R. Smooth muscle cells and the pathogenesis of the lesions of atherosclerosis. Br Heart J. 1993; 69: S30–S37. Review.
Zureik M, Ducimetiere P, Touboul PJ, Courbon D, Bonithon-Kopp C, Berr C, Magne C. Common carotid intima-media thickness predicts occurrence of carotid atherosclerotic plaques: longitudinal results from the Aging Vascular Study (EVA) study. Arterioscler Thromb Vasc Biol. 2000; 20: 1622–1629.
Parmentier E, Morton W, Petschek H. Platelet aggregate formation in a region of separated blood flow. J Biol Chem. 1981; 20: 2012–2021.