Pii: s0021-9150(02)00246-0

Apolipoproteins and carotid artery atherosclerosis in an elderly multiethnic population: the Northern Manhattan stroke study Jiann-Shing Jeng a, Ralph L. Sacco a,b, Douglas E. Kargman a, Bernadette Boden- Albala a,b, Myunghee C. Paik b, Jeffrey Jones c, Lars Berglund c, a Departments of Neurology, Columbia University, New York, NY, USA b School of Public Health, Columbia University, New York, NY, USA c Department of Medicine, Columbia University, New York, NY, USA Received 20 December 2001; received in revised form 2 April 2002; accepted 10 June 2002 The association of apolipoproteins A-I and B (apo A-I and apo B) with cardiovascular disease has been studied in younger populations, but there is sparse information in the elderly. We determined whether apo A-I and apo B were associated with carotidartery atherosclerosis (CAA) in 507 stroke-free elderly community residents (mean age 70.19/11.7 years, 60% women, 41%Hispanics, 30% African American, 28% Caucasian). CAA severity was normal (no plaque or carotid stenosis) in 39%, mild(maximum plaque thickness 0/1.8 mm or carotid stenosis B/40%) in 25%, and moderate/severe (maximum plaque thickness /1.8mm or carotid stenosis E/40%) in 36%. CAA severity increased with age in all race/ethnic groups (P B/0.01). CAA was similaramong African Americans and Caucasians, but less in Hispanics (age adjusted OR: 0.5, CI: 0.4 Á/0.8). apo A-I B/1.2 g/l (OR: 2.0, CI:1.0 Á/3.3) and apo B E/1.4 g/l (OR: 2.0, CI: 1.1 Á/3.6) were associated with moderate Á/severe CAA. An apo B/apo A-I ratio E/1 wasassociated with moderate Á/severe CAA (OR: 2.4, CI: 1.3 Á/4.4), and the association varied by race (Hispanics OR: 4.3, CI: 1.8 Á/10;non-Hispanics, OR: 1.4, CI: 0.6 Á/3.2). Total cholesterol, triglycerides and low density lipoprotein cholesterol were not associatedwith moderate Á/severe CAA, while high density lipoprotein cholesterol was protective (OR: 0.4, CI: 0.2 Á/0.8). Thus, in an elderlypopulation, apo A-I and B were determinants of moderate Á/severe CAA, and the degree of association varied by race/ethnicity# 2002 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Apolipoprotein; Atherosclerosis; Carotid arteries; Racial differences; Risk factors progressively attenuated with age, the absolute riskattributable to blood lipids is greater among the elderly Extracranial carotid artery atherosclerosis (CAA) is a Further, intervention studies have shown that major cause of cerebral infarction and transient ischemic lowering of low density lipoprotein (LDL) cholesterol attacks, a disease primarily of the elderly. In population- levels results in similar benefits in older and younger based epidemiological studies, atherosclerotic risk fac- subjects, as well as in a reduction in cerebrovascular tors such as age, hypertension, systolic blood pressure, events The lack of a consistent association between diabetes mellitus, and smoking, but not lipids and lipids and CAA is therefore puzzling. In addition, most lipoproteins, have been consistently associated with population-based studies of carotid atherosclerosis have CAA It is notable that for coronary artery so far been performed among Caucasians and in a few disease, although the magnitude of the association is cases African Americans, but there is little informationavailable for Hispanics. We recently demonstrated thatHispanics had significantly less carotid plaque than * Corresponding author. Present address: Department of Medicine, Caucasians or African Americans, and found a signifi- University of California, Davis UCD Medical Center, 4150 V Street, cant relationship between race/ethnicity, carotid plaque Sacramento, CA 95817, USA. Tel.: '/1-916-734-7504; fax: '/1-916- and LDL cholesterol, suggesting that the impact of lipid 0021-9150/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 0 2 1 - 9 1 5 0 ( 0 2 ) 0 0 2 4 6 - 0 J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 Apolipoproteins A-I and B (apo A-I and B) are the major protein constituents of the high density lipopro-tein (HDL) and LDL fractions, respectively. Previous Race/ethnicity was categorized based on self-report, studies addressing the association of apo A-I and apo B following definitions used in the US census of 1990, as with cardiovascular disease have yielded mixed results.
Hispanic, Black but not of Hispanic heritage, non- In some studies, although not universally found, apo B Hispanic White, or others. Standardized questionnaires was reported to give better information than lipid levels, regarding sociodemographic characteristics, stroke risk while in general, apo A-I levels have not been found to factors and comorbid atherosclerotic diseases were be independently predictive . However, most of adapted from the Behavioral Risk Factor Surveillance these studies were performed in younger and middle- System by the Centers for Disease Control and Preven- aged populations. Among the elderly, although total tion Hypertension was defined as a systolic blood cholesterol levels tend to decline, the HDL fraction of pressure recording E/160 mmHg or a diastolic blood plasma cholesterol (HDL-C), and its ratio with total pressure recording E/95 mmHg based on the mean of cholesterol (THR) remain important protective factors two blood pressure measurements, a patient’s self-report against CHD among the elderly . Under the new of a history of hypertension, or use of antihypertensive National Cholesterol Education Program guidelines, a therapy. Diabetes mellitus was defined by a patient’s high proportion of the elderly are candidates for self-report of a history of diabetes mellitus, insulin use, additional evaluation and possibly lipid-lowering inter- oral hypoglycemic use, or fasting blood glucose E/126 ventions aimed at the primary or secondary prevention mg/dl ( E/7.0 mmol/l). Body mass index was calculated of coronary heart disease . However, few studies as weight (kilograms) divided by height (meters) have characterized the distributions of lipids and squared, and obesity was defined as body mass index apolipoproteins in elderly populations, particularly E/27.8 for men, and E/27.3 for women. Leisure Á/time/ among minority groups, and even fewer have studied recreation physical activity was assessed by a question- the association of these parameters with measurements naire adapted from the National Health Interview of atherosclerosis in this context. The aim of this study Survey of the National Center for Health Statistics was therefore to investigate the association of apolipo- . The questionnaire records the frequency and proteins and lipids with carotid atherosclerosis in a duration of 14 different recreational activities during multiethnic population-based sample of stroke-free the 2-week period before the interview. Leisure Á/time physical inactivity was defined as subjects with noleisure Á/time recreational physical activity. Cigarettesmoking was characterized by amount (packs per day) and duration (number of years smoked). Smoking wascategorized as no to mild smoking (none or any smoking for B/20 years, or smoking B/1 pack per day for 20 Á/39years), and heavy smoking (any smoking E/40 years, or The Northern Manhattan Stroke Study (NOMASS), smoking at least 1 pack per day for 20 Á/39 years).
recruiting stroke subjects and controls, is a prospective Alcohol use was collected as drinks per day, week, or community-based study designed to determine incidence month. Comorbid conditions included coronary artery rates, risk factors and outcome of stroke The disease (myocardial infarction, angina) and peripheral cohort for the present study was derived from the stroke-free subjects enrolled in NOMASS. Stroke-freesubjects E/40 years were randomly selected and re-cruited into this study through random digit dialingtechniques After informed consent was obtained, 2.3. Measurement of serum lipids, lipoproteins and sociodemographic characteristics, stroke risk factors, other medical conditions, diet, and functional statuswere evaluated, an electrocardiogram was obtained and Plasma levels of cholesterol and triglycerides were an echocardiography was performed. Fasting blood specimens for glucose, lipids, lipoproteins and apolipo- (Boehringer Mannheim, Germany). HDL cholesterol proteins were drawn. A neurological examination was levels were measured after precipitation of plasma apo completed. Approximately two thirds of the recruited B-containing lipoproteins with phosphotungstic acid stroke-free subjects were randomly selected to undergo and LDL cholesterol levels were calculated by the carotid duplex ultrasonographic studies. In all, 507 Friedewald formula. Serum levels of apo A-I and B stroke-free subjects, who had completed enrollment, were determined by commercially available immunone- including duplex ultrasonography and blood drawing, phelometric procedures. The interassay coefficients of variation for these measurements were 3 Á/5% J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 2.4. Category of CAA by duplex ultrasonography models were developed by separately adding lipidparameters to a model containing other sociodemo- CAA was assessed by a Siemens Quantum 2000 graphic and risk factors variables. Statistical interac- duplex ultrasound system with 7.5-MHz scanning fre- tions were evaluated in the final models.
quency in B-mode and 5.0-MHz frequency in pulsedDoppler mode as described in detail elsewhere .
Briefly, with the subject lying in a supine position, the extracranial carotid arteries (including the common,internal, and external carotid arteries and the carotid The mean age of the study group was 70.19/11.7 years bulb area) were imaged in the longitudinal and trans- (median age 71 years; range 40 Á/99 years); 40% were men verse planes. The presence, morphology and thickness of and 60% were women. The race-ethnic distribution was carotid plaques, defined as an area of focal hyperechoic 41% Hispanic (n 0/206), 28% non-Hispanic whites (n 0/ wall thickening, and the extent of internal carotid artery 141), 30% non-Hispanic blacks (n 0/153) and 1% others stenosis were recorded. Based on maximum carotid (n 0/7). Of the subjects, 196 (39%) had no detectable plaque thickness and extent of internal carotid artery CAA, while 128 subjects (25%) had mild CAA (max- stenosis bilaterally, CAA was categorized as normal (no imum carotid plaque thickness 0/1.8 mm or maximum carotid plaque or internal carotid artery stenosis), mild internal carotid artery stenosis B/40%), and 183 subjects (maximum carotid plaque thickness E/1.8 mm or (36%) showed signs of moderate Á/severe CAA (max- maximum internal carotid artery stenosis B/40%) and imum carotid plaque thickness /1.8 mm or maximum moderate Á/severe (maximum carotid plaque thickness internal carotid artery stenosis E/40%). Sociodemo- /1.8 mm or maximum internal carotid artery stenosis graphics, atherosclerotic risk factors and comorbid E/40%). The use of a carotid plaque thickness of 1.8 mm atherosclerotic diseases for each of the CAA categories to define the above groups was based on the CAA distribution in 1050 stroke-free subjects where the mean mellitus, hypertension, heavy smoking and presence of maximum plaque thickness was 1.31 mm (SD:1.32) and coronary artery disease or peripheral vascular disease the 66.6% percentile was 1.8 mm . Internal carotid was greater in the group with moderate Á/severe CAA artery stenosis E/40% was defined by standard criteria compared to no CAA. The group with mild CAA had requiring a ratio of internal carotid artery: carotid artery intermediate frequencies of these parameters between velocities greater than 2, or peak internal carotid artery the other two groups. Among the race/ethnicity groups, systolic flow velocity of E/110 cm/s. In previous studies the proportion of African Americans and Caucasians at our center, the interrater reliability of the estimation were higher among moderate Á/severe CAA compared to of plaque morphology based on duplex Doppler sono- no CAA, while the relative frequency of Hispanics graphy had a k value of 0.05 for plaque surface structure, well below the value of 0.40 suggested for For all subjects, CAA severity increased with age.
Thus, while any form of CAA was present in only 14%of the subjects between 40 Á/54 years, this proportion increased to 80% of the subjects E/75 years of age (datanot shown). This increase was most prominent among Distribution of mild and moderate Á/severe CAA was subjects with moderate Á/severe CAA, which increased evaluated and stratified by different age ranges. AN- from 3% of the youngest cohort to 51% of the oldest OVA was used to compare mean age in the different cohort. Concomitantly, the proportion of subjects with groups of CAA in all and among each race-ethnic no detectable CAA decreased sequentially from 86% of subgroup. The association of sociodemographic para- the subjects between 40 and 54 years to 20% of subjects meters, risk factors and comorbid conditions with mild E/75 years. When dichotomizing age at 65 years, there CAA and moderate Á/severe CAA were compared sepa- was a significant association with presence of CAA, with rately using those with absent CAA as a reference. The OR of 3.7 for mild CAA and 6.0 for moderate Á/severe odds ratios (OR), 95% confidence intervals (CI) and CAA (When analyzing the three race/ethnicity significance were judged based on the x2 test for groups separately, CAA severity increased with age in categorical variables. Multivariate analyses were per- all groups (P B/0.001) (data not shown).
formed using a logistic regression model to identify We next analyzed univariate associations for socio- independent factors. Variables were selected for entry if demographic factors, atherosclerotic risk factors and probability value was less than 0.10 after univariate comorbid atherosclerotic diseases with the presence of testing. Modeling was done using the selected variables CAA. As seen in , mild CAA was less common as the independent variables and CAA as the dependent among African Americans than Caucasians and both variable. Odds ratios and 95% CI were calculated.
mild and moderate Á/severe CAA was seen less frequently Because of the co-linearity of various lipid parameters, among Hispanics compared to Caucasians. No differ- J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 Table 1Frequencies of sociodemographics, atherosclerotic risk factors and comorbid atherosclerotic diseases in cohort Comorbid conditionsCoronary artery disease ence was seen in sociodemographic factors other than which increased significantly with increasing severity of age and race/ethnicity (gender, education and Medicaid CAA. Of the ethnic groups, Hispanics had the highest status) for subjects with and without CAA. Other apo B/apo A-I ratio, resulting from slightly higher mean variables, such as obesity, physical inactivity and apo B and slightly lower mean apo A-I levels compared alcohol drinking were not associated with the presence to Caucasians, while African Americans had the lowest of CAA. Among atherosclerotic risk factors, there was ratio, largely due to the higher apo A-I levels.
no significant difference between subjects with mild Categorical univariate analysis of lipids, lipoproteins CAA compared to subjects without CAA, except for and apolipoproteins for subjects with moderate Á/severe heavy smoking. More prominent differences were found for subjects with moderate Á/severe CAA, where hyper- The levels for dichotomization were chosen to represent tension, diabetes mellitus, and heavy smoking were clinically relevant cut-off levels. As seen in the table, a significantly associated with presence of CAA ( total cholesterol/HDL-C ratio E/5, apo A-I levels B/1.2 Further, history of coronary artery disease or g/l and apo B/apo A-I ratio E/1 were significantly peripheral vascular disease was significantly associated associated with moderate Á/severe CAA levels. In con- trast, no association with moderate Á/severe CAA was The frequency distribution of apo A-I, apo B and the found for total cholesterol levels E/5.92 mmol/l ( E/240 apo B/apo A-I ratio in the study cohort is shown in mg/dl), triglyceride levels E/2.03 mmol/l ( E/200 mg/dl) As seen in the figure, the apo A-I distribution was and LDL cholesterol levels E/3.95 mmol/l ( E/160 mg/ somewhat skewed with a tail towards higher levels, while dl). In univariate analysis, apo B levels E/1.4 g/l showed the apo B distribution had a more Gaussian pattern.
a borderline association with moderate Á/severe CAA.
Levels of apo A-I were higher among women than men, As gender and race/ethnicity influenced apolipopro- and higher among African Americans than Caucasians tein levels, a multiple logistic regression model was used or Hispanics in agreement with previous population to assess which apolipoprotein and lipid variables were studies (). No differences in apo B levels were independently associated with CAA. As seen in found between these groups. Further, apo B levels low apo A-I levels ( B/1.2 g/l), and an apo B/apo A-I increased gradually with increasing CAA, while a ratio E/1 remained significantly associated with mod- decreasing trend was seen for apo A-I levels. These erate Á/severe CAA after adjustment for age, hyperten- differences were reflected in the apo B/apo A-I ratio, sion, diabetes mellitus, heavy smoking and race- J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 logistic regression model. An apo B/apo A-I ratio E/1 Univariate odds ratio and 95% CI For variables associated with mild significantly predicted mild carotid atherosclerosis (OR: 3.2, CI: 1.04 Á/10.0). Also age E/65 years (OR: 3.3, CI:1.9 Á/5.6) and heavy smoking (OR: 1.6, CI: 1.0 Á/2.7) were predictors of mild CAA, while Hispanics were less prevalent than non-Hispanics (OR: 0.6, CI: 0.3 Á/0.9).
SociodemographicsAge ( E 65 years vs B 65 years) A number of studies, including our own, have shown that the prevalence of CAA and atherosclerotic risk factors as well as the mortality rates of cardiovascular diseases differ significantly among Hispanics, non- Although stroke incidence rates were higher among Hispanics than among Caucasians in NOMASS, the severity of CAA was less in stroke-free Hispanics than Caucasians and African Americans The race Á/ ethnic differences in CAA became more prominent in advanced (moderate Á/severe) atherosclerosis. Our pre- vious study also showed the age-adjusted maximal plaque thickness was less in Hispanics than Caucasians and African Americans These findings are con- cordant with The Insulin Resistance Atherosclerosis Study which revealed that Hispanics have significantly less CCA intima-media thickness than Caucasians after adjustment for conventional cardiovascular risk factors Values are OR (95% CI). *P B 0.005; **P B 0.05 subjects with CAA and insulin sensitivity . Other population-based studies of racial differences in CAA, especially theCardiovascular Health Study and Atherosclerotic Risk ethnicity (Hispanics versus non-Hispanics). In addition, in Communities study , have mainly focused on apo B levels E/1.4 g/l were associated with moderate Á/ African Americans and Caucasians and rarely on severe CAA. In contrast, higher total cholesterol, Hispanics. According to the findings from the Third triglyceride and LDL cholesterol levels were not asso- National Health and Nutrition Examination Survey, the ciated with moderate Á/severe CAA, while a higher HDL prevalences of cardiovascular risk factors, including cholesterol level was protective (OR: 0.4, CI: 0.2 Á/0.8).
systolic blood pressure, body-mass-index, physical in- In addition, a total/HDL cholesterol ratio E/5 was activity and diabetes, are greater in Hispanic and associated with moderate Á/severe CAA. This ratio African American women than in Caucasian women correlated significantly with the apo B/apo A-I ratio . However, the United States vital statistics analysis (r 0/0.56, P 0/0.02). Of note, the association between the showed the mortality rates of coronary artery disease apo B/apo A-I ratio and moderate Á/severe carotid and stroke were lower among Hispanics than among atherosclerosis remained significant also when the Caucasians . Heterogeneity in Hispanic populations total/HDL cholesterol ratio was included in the multiple could explain some of these differences, and genetic and regression model (OR 2.1, CI: 1.3 Á/4.4). A significant other environmental risk factors not measured in this interaction was detected between the apo B/apo A-I study could also contribute to these race-ethnic dispa- ratio and Hispanic race-ethnicity, but not among African Americans and Caucasians (Hispanics, OR:4.3, CI:1.8 Á/10; non-Hispanics, OR:1.4, CI:0.6 Á/3.2).
4.2. Relation of apo B and apo A-I to CAA No interaction between HDL cholesterol and Hispanicrace-ethnicity was detected.
The total apo B plasma concentration comprise the Of the lipid or apolipoprotein parameters, only the total amount of apo B in triglyceride-rich lipoproteins apo B/apo A-I ratio differed significantly between the (including very-low-density lipoproteins (VLDL) and groups with mild CAA versus no CAA using a multiple intermediate-density lipoproteins (IDL)) as well as in J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 Fig. 1. Relative frequency of apo A-I levels (top), apo B levels (middle) and the apo A-I/apo B ratio (bottom) in the studied population (n 0/507).
Results for apo A-I and apo B are given as g/l.
cholesterol-rich particles (mainly LDLs) Recent within the LDL spectrum there is particle heterogeneity, studies have disclosed that triglyceride-rich apo B- and smaller, dense LDL particles with a high apo containing lipoproteins, primarily VLDL remnants B:cholesterol ratio have been suggested to be particu- and IDL are significantly related to ischemic heart larly atherogenic . Studies have shown that hyper- disease Lately, it has been hypothesized that apobetalipoproteinemia is an important risk factor for preferentially smaller particles within the triglyceride- patients with ischemic heart disease as well as for rich lipoprotein range, with a higher apo B: lipid ratio, young adults with a parental history of premature may be associated with cardiovascular disease Also coronary heart disease So far, only a few studies J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 tion between apo B levels and carotid atherosclerosis is Mean apolipoprotein levels (9SD) by age, gender, race-ethnicity and not limited to elderly subjects, as apo B was found to be a significant indicator of CAA also in middle-aged menand women in the Bruneck Ischemic Heart Disease and Stroke Prevention Study Interestingly, we found aninteraction among apolipoproteins, Hispanic (versus non-Hispanic) race-ethnicity, and CAA, as the associa- tion between apolipoprotein levels and CAA was particularly pronounced among Hispanics. It is note- worthy that a high apo B level is a characteristic of the metabolic syndrome, present in a comparatively high frequency in Hispanics. Our previous study showed an interaction among Hispanic (versus non-Hispanic) race- ethnicity, LDL cholesterol and maximal ICA plaque thickness However, in the present analysis, LDL cholesterol was not associated with CAA and there was no interaction between LDL cholesterol and race- ethnicity. Although the methods of estimating CAA applied to these two studies were not identical, it is suggested that apo B may be more powerful than LDL-C in determination of CAA for Hispanics.
Results for apo A-I and apo B levels are given as g/l.
Both apo A-I and HDL cholesterol were similarly and have addressed the important role of apo B on CAA, significantly negatively associated with moderate Á/severe and there is virtually no information in elderly .
CAA in the present study. We have previously demon-strated that HDL cholesterol has significantly protective Our study clearly showed that an apo B concentration properties in ischemic stroke . In the Physicians’ E/1.4 g/l and an apo B/apo A-I ratio E/1 were Health Study, there was little predictive value of apo A-I significantly related to moderate Á/severe CAA, even for myocardial infarction after considering conventional after controlling for traditional risk factors. These result risk factors including HDL cholesterol and the total/ suggest that the plasma concentration of apo B, broadly HDL cholesterol ratio . Our study results revealed mirroring the number of potentially atherogenic parti- that apo A-I and HDL cholesterol had similar protective cles, was a better determinant of CAA than total effects for CAA among the elderly. Further, the ratio of cholesterol and LDL cholesterol in our elderly popula- apo B to apo A-I was a significant determinant of tion. This concept is supported by previous findings, moderate Á/severe CAA, underscoring that analysis of such as the Systolic Hypertension in the Elderly these apolipoprotein levels may provide important Program study, where a high apo B concentration was information in risk prevention among elderly. Notably, significantly related to carotid stenosis The associa- the apo B/apo A-I ratio remained significantly asso- Table 4Association of lipids and lipoproteins with moderate Á/severe CAA in univariate and multivariate models In the multivariate model, adjustments were made for age, hypertension, diabetes mellitus, heavy smoking and race-ethnicity (Hispanics versus J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 [2] Fine-Edelstein JS, Wolf PA, O’Leary DH, Poehlman H, Belanger when the total/HDL cholesterol ratio was included in AJ, Kase CS, et al. Precursors of extracranial carotid athero-sclerosis in the Framingham Study. Neurology 1994;44:1046 Á/50.
[3] ARIC Investigators, Heiss G, Sharrett R, Barnes R, Chambless LE, Szklo M, Alzola C. Carotid atherosclerosis measured by B-mode ultrasound in population: associations with cardiovascular risk factors in the ARIC Study. Am J Epidemiol 1991;134:250 Á/6.
[4] Kuller L, Borhani N, Furberg C, Gardin J, Manolio T, O’Leary Although the present study is the first study to D, et al. Prevalence of subclinical atherosclerosis and cardiovas- demonstrate an association of apolipoproteins and cular disease and association with risk factors in the Cardiovas- CAA across ethnicity, it has several limitations. First, cular Health Study. Am J Epidemiol 1994;139:1164 Á/79.
[5] Kuller L, Fisher L, McClelland R, Fried L, Cushman M, Jackson this study had a cross-sectional design limiting our S, et al. Differences in prevalence of and risk factor for subclinical possibilities to draw inferences about a causal pathway.
vascular disease among black and white participants in the Although age, hypertension, diabetes, smoking, HDL Cardiovascular Health Study. Arterioscler Thromb Vasc Biol cholesterol, apo A-I and apo B were independently related to CAA, we cannot with certainty establish [6] Sutton-Tyrrell K, Alcorn HG, Wolfson SK, Kelsey SF, Kuller LH. Predictors of carotid stenosis in older adults with and whether these factors predict development of CAA. In without isolated systolic hypertension. Stroke 1993;24:355 Á/61.
addition, it is likely that there are complex interactions [7] Kronmal RA, Cain KC, Ye Z, Omenn GS. Total serum among these risk factors. Second, our stroke-free cholesterol levels and mortality risk as a function of age: a report subjects were sampled from northern Manhattan and based on the Framingham data. Arch Intern Med 1993;153:1065 Á/ may be not representative of race-ethnicity variations in [8] Krumholz HM, Seeman TE, Merrill SS, Mendes de Leon CF, other areas of United States. As mentioned above, Vaccarino V, Silverman DI, et al. Lack of association between stroke incidence and mortality rates vary in different cholesterol and coronary heart disease mortality and morbidity Hispanic areas. The Hispanic residents in northern and all-cause mortality in persons older than 70 years. JAMA Manhattan are mainly from the Dominican Republic and other Caribbean islands and differ in several aspects [9] Scandinavian Simvastatin Study Group, Miettinen TA, Pyorala K, Olsson AG, Musliner TA, Cook TJ, Faergeman O, et al.
from Mexican Americans. On the other hand, the setting Cholesterol-lowering therapy in women and elderly patients with of the study also had offered important strengths.
myocardial infarction or angina pectoris. Findings from the Subjects were recruited using random digit dialing, Scandinavian Simvastatin Survival Study (4S). Circulation helping to insure that the sample was population based and race-ethnic comparisons were made using sub- [10] Sacco RL, Roberts JK, Boden-Albala B, Gu Q, Lin IF, Kargman DE, et al. Race-ethnicity and determinants of carotid athero- groups drawn from the same underlying community to sclerosis in a multiethnic population. The Northern Manhattan minimize sociodemographic differences.
Stroke Study. Stroke 1997;28:929 Á/35.
In conclusion, Hispanics had less significant CAA [11] Lamarche B, Moorjani S, Lupien PJ, Cantin B, Bernard P-M, et than blacks and whites from the Northern Manhattan al. Apolipoprotein A-I and B levels and the risk of ischemic heart community. Age, hypertension, diabetes, smoking, race- disease during a five-year follow-up of men in the Que´becCardiovascular Study. Circulation 1996;94:273 Á ethnicity, HDL cholesterol and apolipoprotein A-I and [12] Sharett AR, Patsch W, Sorlie PD, Heiss G, Bond MG, Davis CE.
B levels were independently related to moderate Á/severe Associations of lipoprotein cholesterols, apolipoproteins A-I and CAA. Apolipoproteins, in particular the apo B/apo A-I B, and triglycerides with carotid atherosclerosis and coronary ratio, may be a useful determinant of moderate Á heart disease. The Atherosclerosis Risk in Communities (ARIC) Study. Arterioscler Thromb Vasc Biol 1994;14:1098 Á/104.
[13] Sharett AR, Ballantyne CM, Coady SA, Heiss G, Sorlie PD, Catellier D, et al. Coronary heart disease prediction fromlipoprotein cholesterol levels, triglycerides, lipoprotein(a), apoli-poproteins A-I and B, and HDL density subfractions. The Atherosclerosis Risk in Communities (ARIC) Study. Circulation2001;104:1108 Á/13.
This study was supported by grants NS 9993, HL [14] Wilson PW, Anderson KM, Harris T, Kannel WB, Castelli WP.
62705, RR 00645 from the NIH. J.S.J. is a recipient of a Determinants of change in total cholesterol and HDL-C with age: research fellowship from the Department of Neurology, the Framingham Study. J Gerontol 1994;49:M252 Á/7.
[15] Expert Panel on Detection, Evaluation, and Treatment of High National Taiwan University, Taipei, Taiwan.
Blood Cholesterol in Adults. Executive Summary of the thirdreport of the National Cholesterol Education Program ExpertPanel on Detection, Evaluation, and Treatment of High BloodCholesterol in Adults (Adult Treatment Panel III). JAMA 2001; [16] The Northern Manhattan Stroke Study Collabrators, Sacco RL, [1] Burchfiel CM, Laws A, Benfante R, Goldberg RJ, Hwang LJ, Boden-Albala B, Gan R, Chen X, Kargman DE, Shea S, et al.
Chiu D, et al. Combined effects of HDL cholesterol, triglyceride, Stroke incidence among white, black, and Hispanic residents of an and total cholesterol concentrations on 18-year risk of athero- urban community. The Northern Manhattan Stroke Study. Am J sclerotic disease. Circulation 1995;92:1430 Á/6.
J.-S. Jeng et al. / Atherosclerosis 165 (2002) 317 Á/325 [17] Gentry EM, Kalsbeek WD, Hogelin GC, Jones JT, Gaines KL, [24] Liao Y, Cooper RS, Cao G, Kaufman JS, Long AE, McGee DL.
Forman MR, et al. The Behavioral Risk Factor Surveys, II: Mortality from coronary heart disease and cardiovascular disease design, methods and estimates from combined state data. Am J among adult US Hispanics: findings from the National Health Interview Survey (1986 Á/1994). J Am Coll Cardiol 1997;30:1200 Á/ [18] Sacco RL, Gan R, Boden-Albala B, Lin IF, Kargman DE, Hauser WA, et al. Leisure Á/time physical activity and ischemic [25] Lamarche B, Tchernof A, Maurie`ge P, Cantin B, Dagenais GR, stroke risk. The Northern Manhattan Stroke Study. Stroke Lupien PJ, et al. Fasting insulin and apolipoprotein B levels and low-density lipoprotein particle size as risk factors for ischemic [19] Tuck CH, Holleran S, Berglund L. Hormonal regulation of heart disease. JAMA 1998;279:1955 Á/61.
lipoprotein(a) levels: Effect of estrogen replacement therapy on [26] Reardon MF, Nestel PJ, Craig IH, Harper RW. Lipoprotein lipoprotein (a) and acute phase reactants in postmenopausal predictors of the severity of coronary artery disease in men and women. Arterioscler Thromb Vasc Biol 1997;17:1822 Á/9.
women. Circulation 1985;71:881 Á/8.
[20] Hartman A, Mohr JP, Thompson JL, Ramos O, Mast H.
[27] Rosseneu M, Fruchart JC, Bard JM, Nicaud V, Vinaimont N, Interrater reliability of plaque morphology classification in Cambien F, et al. Plasma apolipoprotein concentrations in young patients with severe carotid artery stenosis. Acta Neurol Scand adults with a parental history of premature coronary heart disease [21] Sacco RL, Kargman DE, Zamanillo MC. Race-ethnicity in stroke risk factors among hospitalized patients with cerebral infarction: The Northern Manhattan Stroke Study. Neurology 1995;45:659 Á [28] Willeit J, Kiechl S. Prevalence and risk factors of asymptomatic extracranial carotid artery atherosclerosis. A population-based [22] D’Agostino RB, Burke G, O’Leary D, Rewers M, Selby J, Savage study. Arterioscler Thromb 1993;13:661 Á/8.
PJ, et al. Ethnic differences in carotid wall thickness. The Insulin [29] Sacco RL, Benson RT, Kargman DE, Boden-Albala B, Tuck C, Resistance Atherosclerosis Study. Stroke 1996;27:1744 Á Lin I-F, et al. The protective effects of high density lipoprotein [23] Winkleby MA, Kraemer HC, Ahn DK, Varady AN. Ethnic and cholesterol for ischemic stroke in the elderly: The Northern socioeconomic differences in cardiovascular disease risk factors.
Manhattan Stroke Study. JAMA 2001;285:2729 Á/35.
Findings for women from the Third National Health and [30] Stampfer MJ, Sacks FM, Salvini S, Willett WC, Hennekens CH.
A prospective study of cholesterol, apolipoproteins, and the risk of myocardial infarction. N Engl J Med 1991;325:373 Á/81.

Source: http://library.hsc.pku.edu.cn/pbl/sdxm/evident/artery/17.pdf