点击显示 收起
the Division of Epidemiology and Community Health (T.O., P.J.S., A.R.F.), University of Minnesota, Minneapolis, Minn
the Osaka Medical Center for Health Science and Promotion (T.O.), Osaka, Japan
the Departments of Medicine and Biochemistry (J.D.M.), Baylor College of Medicine, Houston, Tex
the Departments of Biostatistics (L.E.C.) and Epidemiology (W.D.R.), University of North Carolina, Chapel Hill, NC.
Correspondence and reprint requests to Aaron R. Folsom, MD, Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, 1300 South 2nd Street, Suite 300, Minneapolis, MN 55454-1015. E-mail folsom@epi.umn.edu
Abstract
Background and Purpose— Numerous case-control and cross-sectional studies have reported higher median lipoprotein(a) [Lp(a)] levels among stroke patients than controls, but existing prospective studies have not consistently shown an association. We sought to examine the relationship between plasma Lp(a) levels and the incidence of ischemic stroke among blacks and whites.
Methods— Between 1987 and 1989, 14 221 men and women (3647 blacks and 10 574 whites) aged 45 to 64 years and free of clinical cardiovascular disease, took part in the first examination of the Atherosclerosis Risk in Communities (ARIC) study cohort. Lp(a) and other risk factors for cardiovascular disease were measured at baseline.
Results— During the 13.5-year follow-up, 496 ischemic strokes occurred. Participants with Lp(a) 300 μg/mL had a 79% higher age, sex, and race-adjusted rate ratio (RR) of ischemic stroke than did those with Lp(a) levels <100 μg/mL. Compared with Lp(a) <100 μg/mL, the multivariate adjusted RRs for Lp(a) 300 μg/mL were 1.84 (95% CI, 1.05 to 3.07) in black women, 1.72 (95% CI, 0.86 to 3.48) in black men, 2.42 (95% CI, 1.30 to 4.53) in white women, and 1.18 (95% CI, 0.47 to 2.90) in white men. There was no significant increment in the RRs for 100 to 199 μg/mL and 200 to 299 μg/mL groups.
Conclusions— A high Lp(a) concentration is associated with a higher incidence of ischemic stroke in blacks and white women, but not in white men.
Key Words: brain infarction epidemiology lipoprotein risk factors
Introduction
Numerous case-control and cross-sectional studies have reported higher median lipoprotein(a) [Lp(a)] levels among stroke/transient ischemic attack (TIA) patients than controls among whites, blacks, and Asians,1–5 but existing prospective studies have not consistently shown an association.6–12 Specifically, 4 case-control studies nested within prospective studies failed to demonstrate an association of Lp(a) and the future risk of stroke,7–9,11 but 3 non-nested prospective studies reported that elevated levels of Lp(a) independently predicted an increased risk of stroke/TIA.6,10,12 Because instability of Lp(a) during prolonged storage has been reported,13,14 analysis of Lp(a) should be performed on the baseline samples within a year of collection. However, prospective studies that satisfied the above conditions are scarce.6,10,12 Furthermore, no prospective study has examined the relationship between Lp(a) and the incidence of stroke among blacks, despite blacks having approximately twice the median plasma Lp(a) level of whites.15
A link between Lp(a) and ischemic stroke is biologically plausible. Apo(a), a unique apolipoprotein contained within Lp(a), has structural homology with plasminogen, so Lp(a) could have a thrombogenic effect by suppressing fibrinolysis.16 Lp(a) particles are susceptible to oxidative modification and avidly taken up by the scavenger receptor pathway, which leads to intracellular cholesterol accumulation and foam cell formation.16,17 Most prior prospective studies did not distinguish ischemic stroke from other types of stroke in their analyses.6,7,10–12
To examine the relationship between plasma Lp(a) levels and the incidence of ischemic stroke among blacks and whites, we used data from a 13.5-year follow-up of men and women in the Atherosclerosis Risk in Communities (ARIC) study.
Methods
Study Population
The populations surveyed included 15 792 men and women aged 45 to 64 years who participated in the ARIC study between 1987 and 1989 in 4 US communities: Forsyth County, North Carolina; Jackson, Mississippi; 8 northwestern suburbs of Minneapolis, Minnesota; and Washington County, Maryland. The study design is described in detail elsewhere.18
We first excluded participants in Forsyth County who were not white or black (n=21) and participants in Minneapolis and Washington County who were not white (n=82). We then excluded participants with missing Lp(a) data (n=570) at baseline, participants with a history of coronary heart disease (n=713) at baseline, and participants with a history of stroke or TIA at baseline (n=185). The remaining 14 221 participants were included in the analyses. Subjects were followed to determine the incidence of stroke through 2002. The average (interquartile range) follow-up time was 13.5 (13.3 to 14.9) years. The study protocol was approved by the institutional review boards of the collaborating institutions, and informed written consent was obtained from each participant.
Baseline Measurements
Methods for blood collection and processing in the ARIC study have been described in detail.19 Participants were asked to fast for 12 hours before their morning clinic appointments. Most of the blood samples were analyzed within 6 weeks of receipt. Lp(a) was measured as total protein component (apolipoprotein(a) plus apolipoprotein B) with a double-antibody ELISA technique for apo(a) detection.20 This protein moiety represents approximately one-third of total Lp(a) lipoprotein mass. Therefore, a value of 100 μg/mL Lp(a) protein is comparable to a total Lp(a) value of 300 μg/mL. For this assay at Lp(a) protein levels in the range 10 to 100 μg/mL, the contribution from plasminogen at physiological concentrations (2000 μg/mL) was negligible.20 In a 40-person subsample, the assay reliability (between-person component of the variance divided by the total variance) was 0.90, with essentially no within-person variability (indicative of a largely genetic measurement), and 9% of the total variance was associated with the assay method itself.21 Total cholesterol and triglycerides were measured by enzymatic methods. High-density lipoprotein (HDL) cholesterol was assayed after dextran sulfate–magnesium precipitation, and low-density lipoprotein (LDL) cholesterol was estimated from the Friedewald equation.
End Point Determination
For the present study, we included stroke events18 occurring between ARIC visit 1 and December 31, 2002. TIAs were not ascertained. All participants were contacted annually by phone and asked about all hospitalizations and deaths in the previous year. We also surveyed lists of discharges from local hospitals and death certificates from state vital statistics offices for potential cerebrovascular events. A nurse abstractor recorded from hospital records signs and symptoms and photocopied neuroimaging (CT or MRI) and other diagnostic reports, if the list of discharge diagnoses included a cerebrovascular disease code (International Classification of Diseases, 9th Revision, code 430 to 438), if a cerebrovascular condition or procedure was mentioned in the discharge summary, or if a cerebrovascular finding was noted on a CT or MRI report. Each eligible case was classified by computer algorithm and by expert reviewer, according to criteria adapted from the National Survey of Stroke. Details on quality assurance for ascertainment and classification of stroke are described elsewhere.22 Qualifying strokes were further classified into definite or probable hospitalized ischemic (cardioembolic or thrombotic), or hemorrhagic stroke on the basis of neuroimaging studies and autopsy, when available. A stroke was classified as ischemic if a brain CT or MRI revealed acute infarction or showed no evidence of hemorrhage. Cardioembolic stroke was defined as an ischemic stroke with establishment of a likely source of embolus, such as valvular heart disease, atrial fibrillation, cardiac or arterial procedure, and intracardiac thrombus. A small number of out-of-hospital fatal strokes (n=4) were not counted as end points.
Statistical Analysis
Because the distribution of Lp(a) was highly right-skewed in both race groups and differed by race and gender, incidence rates were calculated according to categories of Lp(a) (<100 μg/mL, 100 to 199 μg/mL, 200 to 299 μg/mL, and 300 μg/mL) stratified by gender and race.15 Differences among the categories of Lp(a) in age-adjusted mean values or prevalences of potential confounding factors at baseline were calculated using ANOVA or logistic regression models.23
We compared Kaplan-Meier survival plots according to Lp(a) levels stratified by gender and race. The rate ratios of stroke incidence and 95% CIs relative to Lp(a) levels <100 μg/mL were calculated with adjustment for age and other potential confounding factors using the Cox proportional hazards model. The proportional hazards assumption was confirmed. Covariates included age (years), race-field center, systolic blood pressure (mm Hg), antihypertensive medication (yes, no), smoking status (never, former, and current smokers), use of postmenopausal hormone therapy (yes, no), diabetes status (yes, no), LDL cholesterol (mg/dL), HDL cholesterol (mg/dL), fibrinogen (mg/dL), and von Willebrand factor (%).
Results
Mean (SD) and median values (interquartile range) of Lp(a) at baseline were 167 (128), 137 (74 to 228) μg/mL for black women, 146 (112), 119 (64 to 195) μg/mL for black men, 86 (97), 48 (20 to 118) μg/mL for white women, and 73 (86), 39(17 to 97) μg/mL for white men. Age- and gender-adjusted geometric mean values of Lp(a) were higher in blacks than whites (P<0.001), and age- and race-adjusted geometric mean values of Lp(a) were higher in women than men (P<0.001).
Table 1 shows age-, gender-, and race-adjusted mean values or prevalences of risk characteristics at baseline according to Lp(a) levels. The proportions of black women, white women, black men, and white men who had Lp(a) levels 300 μg/mL were 13.6%, 4.6%, 9.8%, and 2.8%. LDL cholesterol and fibrinogen levels were greater with increasing Lp(a) levels. Higher Lp(a) levels also tended to be associated with lower prevalence of taking estrogen or progesterone. BMI and smoking status were not associated with Lp(a) levels.
Among 14 221 men and women followed for an average 13.5 years, 83 incident hemorrhagic strokes and 496 incident ischemic strokes (103 lacunar, 302 nonlacunar, and 91 cardioembolic) occurred. Participants with Lp(a) 300 μg/mL had a 79% higher age-, sex-, and race-adjusted rate ratio (RR) of ischemic stroke than did those with Lp(a) levels <100 μg/mL; the RR was 1.79 (95% CI, 1.32 to 2.42). As the Figure shows, Kaplan-Meier plots for ischemic stroke were significantly different by Lp(a) level in black women (log-rank test, P=0.03) and white women (P=0.0005), but not black men (P=0.69) and white men (P=0.90).
Kaplan-Meier plots of incident ischemic stroke according to Lp(a) levels (μg/mL) stratified by gender and race.
Compared with Lp(a) <100 μg/mL, the multivariate-adjusted RRs for Lp(a) 300 μg/mL were 1.84 (95% CI, 1.05 to 3.07) in black women, 1.72 (95% CI, 0.86 to 3.48) in black men, 2.42 (95% CI, 1.30 to 4.53) in white women, and 1.18 (95% CI, 0.47 to 2.90) in white men (Table 2). There was no significant increment in the RRs for the 100 to 199 μg/mL and 200 to 299 μg/mL groups. When we excluded cardioembolic strokes, the associations of Lp(a) levels with ischemic stroke incidence in men grew stronger. The respective multivariate-adjusted RRs for Lp(a) 300 μg/mL versus Lp(a) <100 μg/mL were 1.78 (95% CI, 0.97 to 3.29) in black women, 2.08 (95% CI, 1.01 to 4.38) in black men, 2.25 (95% CI, 1.10 to 4.60) in white women, and 1.47 (95% CI, 0.59 to 3.65) in white men. There were no interactions between Lp(a) and race or gender (P for interaction were >0.10) for ischemic stroke.
Discussion
This prospective study found that increased levels of Lp(a) were associated positively and independently with the incidence of ischemic stroke in black and white women. The association of Lp(a) with ischemic stroke incidence was also seen among black men after excluding cardioembolic strokes. The multivariate-adjusted RRs for ischemic stroke were not very different between black women and men, but Lp(a) was not associated with ischemic stroke among white men.
Three previous prospective studies of stroke incidence, in which Lp(a) was measured concurrent with a baseline examination, reported that elevated Lp(a) independently predicted an increased risk of stroke/TIA.6,10,12 The Cardiovascular Health Study, conducted among 3972 elderly people, showed that higher Lp(a) levels were independently associated with increased risk of stroke in men but not in women.12 Another prospective study of 11 335 middle-aged whites reported significant positive associations between Lp(a) and stroke/TIA incidence in men but not in women.10 On the other hand, the Framingham Heart Study showed that elevated plasma Lp(a) was an independent predictor of stroke/TIA in 3103 middle-aged women.6 Although Lp(a) levels were positively associated with LDL cholesterol and fibrinogen levels, these previous prospective studies did not control the effect of these confounders. Further, these studies did not distinguish ischemic stroke from other type of strokes in their analyses. Because the pathogenesis is different between ischemic and hemorrhagic stroke, it is difficult to compare the results of these previous studies with our results. A previous ARIC study reported that plasma lipids, such as triglycerides and HDL cholesterol, were weakly associated with incidence of ischemic stroke in women but not men,24 which shows the possibility that effects of lipids on incidence of ischemic stroke may differ according to sex.
The present study showed no monotonic dose-response relationship between Lp(a) levels and ischemic stroke incidence, which suggests the existence of a threshold relationship. The Cardiovascular Health Study also reported that the association was not linear, with a multivariate-adjusted relative risk for the highest versus lowest quintile of Lp(a) in men being 2.92 (95% CI, 1.53 to 5.57).12 The other 2 prospective studies detected Lp(a) qualitatively as a sinking pre–-lipoprotein band on electrophoresis, and could not assess a dose-response relationship between Lp(a) levels and stroke incidence.6,10
The first strength of the present study is that measurement of Lp(a) was performed on the baseline samples within 6 weeks of collection to avoid effects of prolonged storage. Kronenberg et al measured Lp(a) levels after 3 and 28 months of storage at –80°C and reported that Lp(a) decreased markedly, more in subjects with low-molecular-weight isoforms than with high-molecular-weight apo(a) isoforms.13 Because low-molecular-weight apo(a) isoforms are more frequent in patients with atherothrombotic disease compared with control subjects,25 long storage of blood samples may create biased associations between Lp(a) levels and stroke incidence. Simo et al measured Lp(a) levels in 65 survivors of myocardial infarction and 95 age-matched controls before and after storage for 5 years at –70°C.14 During storage, mean Lp(a) levels decreased significantly in samples from patients (–23%) but not in samples from controls (–9%), which also supports the above hypothesis. Four previous case-control studies nested within prospective studies measured Lp(a) after storage for 7 to 14 years. This prolonged storage may be one reason why these studies failed to demonstrate an association of Lp(a) and the future risk of stroke.7–9,11 On the other hand, 3 non-nested prospective studies, which measured Lp(a) within a year after storage, reported that elevated levels of Lp(a) independently predicted an increased risk of stroke.6,10,12
A second strength was our study’s inclusion of both blacks and whites, with a relatively large numbers of events. Although the median plasma Lp(a) level in blacks is approximately twice as high as in whites,15 no prospective study has examined the relationship between Lp(a) and the incidence of stroke among blacks. The present study suggests that higher Lp(a) levels are associated with increased risk of ischemic stroke in women, with higher RRs in whites than blacks. However, it is not necessarily the case that Lp(a) confers less cardiovascular disease risk in blacks than whites. Because the frequency of high Lp(a) levels (300 μg/mL) was higher in blacks than whites, the population-attributable fractions of ischemic stroke for high Lp(a) levels are higher in blacks than in whites (6.2% versus 2.7% in women, 4.1% versus 0.4% in men).
Potential limitations of this study warrant consideration. First, we analyzed the associations between Lp(a) and ischemic stroke incidence using a single assessment of Lp(a) at baseline, which may lead to misclassification of the habitual Lp(a) levels of some individuals. However, Lp(a) is primarily genetically determined, and therefore is little influenced by age, diet and lifestyle.16 Second, many blacks with high plasma levels of Lp(a) have apo(a) isoforms of intermediate size,26 whereas most whites with high plasma levels of Lp(a) have a small apo(a) isoform.27 Because elevated levels of Lp(a) with small apo(a) isoforms independently predict risk of cardiovascular disease in blacks,28 it may be important to measure apo(a) isoforms in future studies.
Acknowledgments
The ARIC study was funded by National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022. The authors thank Prof Eyal Shahar and Lori L. Boland, University of Minnesota, for their valuable help in this study. The authors also acknowledge the staff and participants in the ARIC study for their important contributions.
References
Zenker G, Koltringer P, Bone G, Niederkorn K, Pfeiffer K, Jurgens G. Lipoprotein(a) as a strong indicator for cerebrovascular disease. Stroke. 1986; 17: 942–945.
Schreiner PJ, Chambless LE, Brown SA, Watson RL, Toole J, Heiss G. Lipoprotein(a) as a correlate of stroke and transient ischemic attack prevalence in a biracial cohort: the ARIC Study. Atherosclerosis Risk in Communities. Ann Epidemiol. 1994; 4: 351–359.
Nagayama M, Shinohara Y, Nagayama T. Lipoprotein(a) and ischemic cerebrovascular disease in young adults. Stroke. 1994; 25: 74–78.
Jurgens G, Taddei-Peters WC, Koltringer P, Petek W, Chen Q, Greilberger J, Macomber PF, Butman BT, Stead AG, Ransom JH. Lipoprotein(a) serum concentration and apolipoprotein(a) phenotype correlate with severity and presence of ischemic cerebrovascular disease. Stroke. 1995; 26: 1841–1848.
Kario K, Matsuo T, Kobayashi H, Asada R, Matsuo M. ‘Silent’ cerebral infarction is associated with hypercoagulability, endothelial cell damage, and high Lp(a) levels in elderly Japanese. Arterioscler Thromb Vasc Biol. 1996; 16: 734–741.
Bostom AG, Gagnon DR, Cupples LA, Wilson PW, Jenner JL, Ordovas JM, Schaefer EJ, Castelli WP. A prospective investigation of elevated lipoprotein (a) detected by electrophoresis and cardiovascular disease in women. The Framingham Heart Study. Circulation. 1994; 90: 1688–1695.
Alfthan G, Pekkanen J, Jauhiainen M, Pitkaniemi J, Karvonen M, Tuomilehto J, Salonen JT, Ehnholm C. Relation of serum homocysteine and lipoprotein(a) concentrations to atherosclerotic disease in a prospective Finnish population based study. Atherosclerosis. 1994; 106: 9–19.
Ridker PM, Stampfer MJ, Hennekens CH. Plasma concentration of lipoprotein(a) and the risk of future stroke. JAMA. 1995; 273: 1269–1273.
Glader CA, Stegmayr B, Boman J, Stenlund H, Weinehall L, Hallmans G, Dahlen GH. Chlamydia pneumoniae antibodies and high lipoprotein(a) levels do not predict ischemic cerebral infarctions. Results from a nested case-control study in Northern Sweden. Stroke. 1999; 30: 2013–2018.
Nguyen TT, Ellefson RD, Hodge DO, Bailey KR, Kottke TE, Abu-Lebdeh HS. Predictive value of electrophoretically detected lipoprotein(a) for coronary heart disease and cerebrovascular disease in a community-based cohort of 9936 men and women. Circulation. 1997; 96: 1390–1397.
Price JF, Lee AJ, Rumley A, Lowe GD, Fowkes FG. Lipoprotein (a) and development of intermittent claudication and major cardiovascular events in men and women: the Edinburgh Artery Study. Atherosclerosis. 2001; 157: 241–249.
Ariyo AA, Thach C, Tracy R. Lp(a) lipoprotein, vascular disease, and mortality in the elderly. N Engl J Med. 2003; 349: 2108–2115.
Kronenberg F, Trenkwalder E, Dieplinger H, Utermann G. Lipoprotein(a) in stored plasma samples and the ravages of time. Why epidemiological studies might fail. Arterioscler Thromb Vasc Biol. 1996; 16: 1568–1572.
Simo JM, Camps J, Vilella E, Gomez F, Paul A, Joven J. Instability of lipoprotein(a) in plasma stored at –70°C: effects of concentration, apolipoprotein(a) genotype, and donor cardiovascular disease. Clin Chem. 2001; 47: 1673–1678.
Schreiner PJ, Heiss G, Tyroler HA, Morrisett JD, Davis CE, Smith R. Race and gender differences in the association of Lp(a) with carotid artery wall thickness. The Atherosclerosis Risk in Communities (ARIC) Study. Arterioscler Thromb Vasc Biol. 1996; 16: 471–478.
Milionis HJ, Winder AF, Mikhailidis DP. Lipoprotein (a) and stroke. J Clin Pathol. 2000; 53: 487–496.
de Rijke YB, Jurgens G, Hessels EM, Hermann A, van Berkel TJ. In vivo fate and scavenger receptor recognition of oxidized lipoprotein(a) isoforms in rats. J Lipid Res. 1992; 33: 1315–1325.
The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989; 129: 687–702.
Papp AC, Hatzakis H, Bracey A, Wu KK. ARIC hemostasis study I. Development of a blood collection and processing system suitable for multicenter hemostatic studies. Thromb Haemost. 1989; 61: 15–19.
Gaubatz JW, Chari MV, Nava ML, Guyton JR, Morrisett JD. Isolation and characterization of the two major apoproteins in human lipoprotein(a). J Lipid Res. 1987; 28: 69–79.
Chambless LE, McMahon RP, Brown SA, Patsch W, Heiss G, Shen YL. Short-term intraindividual variability in lipoprotein measurements: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Epidemiol. 1992; 136: 1069–1081.
Rosamond WD, Folsom AR, Chambless LE, Wang CH, McGovern PG, Howard G, Copper LS, Shahar E. Stroke incidence and survival among middle-aged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke. 1999; 30: 736–743.
Wilcosky TC, Chambless LE. A comparison of direct adjustment and regression adjustment of epidemiologic measures. J Chron Dis. 1985; 38: 849–856.
Shahar E, Chambless LE, Rosamond WD, Boland LL, Ballantyne CM, McGovern PG, Sharrett AR. Plasma lipid profile and incident ischemic stroke: the Atherosclerosis Risk in Communities (ARIC) study. Stroke. 2003; 34: 623–631.
Sandholzer C, Saha N, Kark JD, Rees A, Jaross W, Dieplinger H, Hoppichler F, Boerwinkle E, Utermann G. Apo(a) isoforms predict risk for coronary heart disease. A study in six populations. Arterioscler Thromb. 1992; 12: 1214–1226.
Marcovina SM, Albers JJ, Wijsman E, Zhang Z, Chapman NH, Kennedy H. Differences in Lp(a) concentrations and apo(a) polymorphs between black and white Americans. J Lipid Res. 1996; 37: 2569–2585.
Utermann G. The mysteries of lipoprotein(a). Science. 1989; 246: 904–910.
Paultre F, Pearson TA, Weil HF, Tuck CH, Myerson M, Rubin J, Francis CK, Marx HF, Philbin EF, Reed RG, Berglund L. High levels of Lp(a) with a small apo(a) isoform are associated with coronary artery disease in African-American and white men. Arterioscler Thromb Vasc Biol. 2000; 20: 2619–2624.