Lipoprotein(a) and Risk of Ischemic Stroke in the REGARDS Study
Arteriosclerosis, Thrombosis, and Vascular Biology
Abstract
Objective—
Increased Lp(a) [lipoprotein(a)] is associated with coronary heart disease risk, but links with stroke are less consistent. Blacks have higher Lp(a) levels and stroke incidence than whites but have been underrepresented in studies. We hypothesized that Lp(a) is a risk factor for ischemic stroke and that risk differs by race.
Approach and Results—
REGARDS (Reasons for Geographic and Racial Differences in Stroke) recruited 30 239 black and white US adults aged ≥45 in 2003–2007 to study regional and racial differences in stroke mortality. We measured baseline Lp(a) by immunonephelometric assay in 572 cases of incident ischemic stroke and a 967-person cohort random sample. The hazard ratio of stroke by baseline Lp(a) was calculated using Cox proportional hazards models, stratified by race. Lp(a) was modeled in sex- and race-specific quartiles, given known differences in distributions by race and sex. Interactions were tested by including interaction terms in the proportional hazards models, with P<0.10 considered statistically significant. After adjustment for age, sex, and stroke risk factors, being in the fourth versus the first Lp(a) quartile was weakly associated with ischemic stroke overall, hazard ratio, 1.45 (95% CI, 0.96–2.19). In blacks, the hazard ratio was 1.96 (95% CI, 1.10–3.46), whereas in whites HR was 1.14 (95% CI, 0.64–2.04); P interaction=0.12. Lp(a) was lower in men than women, but associations with stroke in men and women were similar.
Conclusions—
We confirm that Lp(a) is a risk factor for ischemic stroke. Further research is needed to confirm the role of racial differences of the Lp(a) risk multiplier in ischemic stroke.
Graphical Abstract
Highlights
•
In this population-based cohort study, Lp(a) [lipoprotein(a)] concentrations were higher in black participants than white participants.
•
Elevated Lp(a) was associated with increased risk of ischemic stroke among all participants, and this relationship may be stronger for black than white participants. Further research is needed to define racial differences in stroke propensity based on Lp(a) levels.
•
Baseline statin usage did not alter the association between Lp(a) levels and stroke.
Introduction
Lp(a) [lipoprotein(a)] is a cardiovascular risk factor that has been under intense investigation in recent years. Proposed mechanisms for its associated proatherosclerotic effects include a role in foam cell formation, promotion of cholesterol deposition into atherosclerotic plaques, and alterations in immunogenic responses.1,2 Lp(a) levels are largely genetically determined and are thought to be only minimally affected by lifestyle.3 This suggests that there may be an inherent predisposition to cardiovascular disease and stroke in people with elevated Lp(a). Given its aforementioned properties, Lp(a) has been investigated as a risk factor for incident cardiovascular disease and stroke.
It is well established that stroke risk and mortality differ between blacks and whites patients in the United States.4 Alongside the conventional cardiovascular risk factors, such as LDL-C (low-density lipoprotein cholesterol), there is emerging interest in identifying biomarkers that can help to explain racial differences in ischemic stroke risk.5,6 Prior data from other large, population-based cohort studies have been conflicting on stroke, with some studies linking elevated Lp(a) levels to a higher incidence of ischemic stroke,7–10 whereas others have not found an association.11–13 This may be partly attributable to lack of differentiation between the subtypes of incident stroke13 or to racial or other differences in cohort composition. This may also be secondary to lack of adjustment for LDL-C levels, which is an important risk factor for stroke.5,6
To address these limitations, we examined the association of Lp(a) levels with ischemic stroke in a large biracial population-based cohort study. We specifically investigated whether Lp(a) levels differed by race and gender and the role of Lp(a) as a predictor of future stroke events in blacks compared with whites and in men compared with women.
Materials and Methods
The data that support the findings of this study are available from the corresponding author on reasonable request.
Subjects
The REGARDS (Reasons for Geographic and Racial Differences in Stroke) cohort is well-described population-based cohort study, that is, evaluating racial and geographic differences in cardiovascular disease.14 The study participants included 30 239 men and women, age 45 and older that were enrolled over 4 years across the contiguous United States. Participants from the region known as the stroke belt (the states of Alabama, Arkansas, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, and Tennessee) were oversampled, as were black participants with a goal of 50% in the stroke belt, 50% black, and 50% women. Computer-assisted telephone interviews were used to elicit verbal informed consent and information on demographics, medical history, and socioeconomic factors. Further data, such as ECG, medication reconciliation, blood samples for laboratory examination, and written informed consent, were obtained via in-home examination at baseline. The Institutional Review Boards of each participating institution reviewed and approved the study methods.
Measurements and Definitions
Age, race, sex, smoking history, and prebaseline stroke were determined by self-report. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic pressure ≥90 mm Hg, or self-reported hypertension treated with antihypertensive medications. Diabetes mellitus was defined by fasting glucose >126 mg/dL, nonfasting glucose >200 mg/dL, or self-report with the use of antidiabetic medications. Atrial fibrillation was defined as self-report or presence on ECG. Prevalent cardiovascular disease was defined as self-reported myocardial infarction, coronary artery bypass, percutaneous coronary intervention, or myocardial infarction on ECG. Left ventricular hypertrophy was determined based on ECG criteria.
Blood was collected in the morning after an overnight fast and transported to a local laboratory, where it was centrifuged and serum or plasma separated. Samples were shipped overnight on ice to the Laboratory for Clinical Biochemistry Research at the University of Vermont, where they were recentrifuged at 30 000g and either analyzed or stored at −80°C.15
Stroke Ascertainment
The primary outcome was first ischemic stroke through September 1, 2011. Incident stroke that occurred after study inclusion has been defined previously.16 Stroke occurrence was determined by contacting participants and proxies every 6 months via telephone and through review of medical records by 2 independent stroke expert physicians from the adjudication committee. All ischemic stroke events were adjudicated by an expert committee using predefined criteria.14 We classified strokes as ischemic or hemorrhagic then further defined the etiologic subtypes of ischemic stroke as large vessel disease, small vessel disease, cardioembolic, other, or unknown.
Case-Cohort Sample
A case-cohort sample with a mean follow-up of 5.4 years was selected from the REGARDS cohort to evaluate the hypotheses. Cases were 572 participants without any baseline stroke who developed ischemic stroke during follow-up. The cohort random sample was selected from the entire cohort, details of which have been published previously.17 Briefly, the cohort random sample for this analysis consisted of 967 participants. They were selected using age (20%, 45–54; 20%, 55–64; 25%, 65–74; 25%, 75–84; and 10%, ≥85), sex (50% women, 50% men), and race (50% black, 50% white) stratified sampling. Similar to the cases, participants with baseline stroke were excluded for this analysis.
Laboratory Methods
Cholesterol, HDL-C (high-density lipoprotein cholesterol), and triglycerides were measured in baseline serum samples as they arrived at the central laboratory using colorimetric reflectance spectrophotometry on the Ortho Vitros Clinical Chemistry System 950IRC instrument (Johnson & Johnson Clinical Diagnostics, Rochester, NY). LDL-C was calculated by the Friedwald equation for participants with triglycerides <400 mg/dL. The CVs for cholesterol, HDL, and triglyceride were <2%, 7%, and <2%, respectively. Lp(a) was measured in 2012 using baseline plasma retrieved from −80°C storage that had been thawed 1× to 2× before analysis. We used an automated particle-enhanced immunonephelometric assay (N Latex Lp(a), Siemens Healthcare Diagnostics, Inc) in which polystyrene particles are coated with monoclonal antibodies to Lp(a) that agglutinate in the presence of antigen to cause an increase in the intensity of scattered light. The increase in scattered light is proportional to the amount of Lp(a) in the sample, as compared with a standard (N Latex Lp(a) Standard SY). The CV was 2% to 5% at different concentrations.
Statistical Methods
Two models were created for proportional hazards regression to evaluate the relationship between Lp(a) and stroke. Model 1 included age, race, sex, and an age×race interaction term. Model 2 also included the Framingham stroke risk factors (atrial fibrillation, smoking, systolic blood pressure, prior cardiovascular disease, left ventricular hypertrophy, and use of any antihypertensive medication). For all modeling, gender- and race-specific quartiles were used for Lp(a) values. Hazard ratios (HRs) for stroke comparing the stroke incidence in the second, third, and fourth quartiles, to incidence in the first quartile were calculated. Interactions were tested by including interaction terms in the proportional hazards models, with P<0.10 considered statistically significant. A type I error rate of 0.10 was chosen a priori for tests of interactions, recognizing that the increased power provided by this comes at the expense of increased risk of false-positive findings. This is because we perceived a type II error to be more costly than a type I error in this area.18
Sensitivity Analyses
We conducted 2 sensitivity analyses. One was to determine the relationship of Lp(a) with ischemic stroke subtypes, with separate proportional hazards models fit for each stroke subtype. The second was to determine if the association of Lp(a) with stroke varied by baseline statin use, which was examined by including an Lp(a) by statin use interaction term in the models and by assessing if addition of statin use to the models attenuated the relationship between Lp(a) and stroke.
Results
Lp(a) values ranged from 1.0 to 217.0 mg/dL in the cohort sample of 967 study participants. There were 242 black women, 236 black men, 246 white women, and 243 white men in the cohort sample. Median (interquartile range) values for Lp(a) in the cohort random sample were 32.8 (16.8–55.0) mg/dL in black women, 26.7 (12.6–56.4) mg/dL for black men, 11.0 (3.5–39.5) mg/dL for white women, and 8.8 (3.3–30.6) mg/dL for white men (Table 1). In the highest quartile of Lp(a) concentrations, the mean (range) Lp(a) concentrations were 99.0 (72.4–125.6) mg/dL in black women, 72.2 (60–104.8) mg/dL in black men, 56.7 (49.6–75.3) mg/dL in white women, and 53.2 (39.5–61.7) mg/dL in white men.
| Group | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| N=215 | N=241 | N=265 | N=246 | |
| Lp(a), black men, mg/dL | <13 | 13–26 | 27–56 | ≥57 |
| Lp(a), black women, mg/dL | <17 | 17–32 | 33–55 | ≥56 |
| Lp(a), white men, mg/dL | <4 | 4–8 | 9–31 | ≥32 |
| Lp(a), white women, mg/dL | <4 | 4–10 | 11–39 | ≥40 |
Lp(a) indicates lipoprotein(a); Q, quartile; and REGARDS, Reasons for Geographic and Racial Differences in Stroke.
Table 2 presents the baseline characteristics of the cohort sample participants based on race-sex specific Lp(a) quartiles, weighted to reflect the entire at-risk cohort of 26 242. The only factors that differed by Lp(a) quartile were the prevalence of dyslipidemia and mean LDL-C, with higher levels of both dyslipidemia and LDL-C noted with increasing Lp(a) quartiles.
| 1 | 2 | 3 | 4 | P Value | |
|---|---|---|---|---|---|
| N | 215 | 241 | 265 | 246 | |
| Weighted N | 6108 | 6305 | 7178 | 6651 | |
| Age, mean (SD) | 63.5 (9.3) | 65.1 (9.4) | 65.1 (9.5) | 64.8 (9.0) | 0.06 |
| Atrial fibrillation, % | 6.5 | 7.2 | 13.3 | 8.4 | 0.09 |
| Body mass index, mean (SD) | 29.1 (5.3) | 29.7 (6.5) | 28.8 (5.6) | 29.6 (6.2) | 0.73 |
| Current smoking, % | 20 | 12 | 12 | 11 | 0.09 |
| Diabetes mellitus, % | 20 | 21 | 22 | 21 | 0.96 |
| Dyslipidemia, % | 56 | 56 | 54 | 69 | 0.01 |
| Stroke belt residence | 35.5 | 30.1 | 37.9 | 31.5 | 0.55 |
| HDL-C, mean (SD), mg/dL | 51 (19) | 51 (15) | 52 (16) | 52 (16) | 0.27 |
| Hypertension, % | 53 | 56 | 58 | 59 | 0.60 |
| Left ventricular hypertrophy, % | 9.0 | 6.8 | 6.2 | 8.4 | 0.74 |
| LDL-C, mean (SD), mg/dL | 111 (35) | 106 (33) | 113 (30) | 120 (34) | <0.001 |
| Prevalent CVD, % | 13 | 10 | 19 | 24 | 0.001 |
| Statin, % | 28 | 29 | 29 | 45 | 0.002 |
| SBP, mean (SD), mm Hg | 127 (16) | 130 (18) | 126 (17) | 126 (15) | 0.06 |
P value represents test of the null hypothesis that the means/percentages are equivalent for the quartiles.
CVD indicates cardiovascular disease; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; Lp(a), lipoprotein(a); and SBP, systolic blood pressure in millimeters of mercury.
There were 572 incident ischemic stroke cases at median 5.4 years of follow-up (range 1 day to 8.5 years), with 239 strokes in black participants and 333 strokes in white participants. The median Lp(a) for the cases was 24.0 mg/dL (interquartile range, 7.0–53 mg/dL). Of note, a total of 20 cohort sample participants had incident stroke during the follow-up.
Among all cases and controls with ischemic stroke, subjects in the third quartile of Lp(a) had a HR of 1.43 (95% CI, 1.01–2.03) of having ischemic stroke compared with subjects in the first Lp(a) quartile after adjustment for age, race, and sex (Table 3). There was also a trend towards significance in the fourth quartile. After adjustment for age, sex, and stroke risk factors, being in the fourth versus the first Lp(a) quartile was weakly associated with ischemic stroke overall with an HR of 1.45 (95% CI, 0.96–2.19). In model 1, the black participants with Lp(a) in the fourth quartile compared with the first quartile had an HR of 1.74 (95% CI, 1.03–2.94) for incident stroke (Table 3). There were no significant association between Lp(a) and incident stroke in white participants, although the interaction between Lp(a) quartiles and race was not statistically significant (P=0.37). After adjustment for the Framingham stroke risk factors in model 2, the association with stroke among black participants in the fourth quartile of Lp(a) was stronger with an HR of 1.96 (95% CI, 1.10–3.46) compared to black participants with the Lp(a) levels in the lowest quartile (Table 3). There was no significant association between Lp(a) and incident stroke in white study participants in model 2 (race×Lp(a) interaction P value=0.12). There was no difference in the association of Lp(a) and incident stroke by sex (Table 3).
| First Quartile | Second Quartile | Third Quartile | Fourth Quartile | Interaction P Value | |
|---|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | ||
| All | |||||
| N cases | 99 | 114 | 174 | 160 | |
| N controls | 215 | 241 | 265 | 246 | |
| Model 1 | 1.0 (ref) | 1.05 (0.72–1.53) | 1.43 (1.01–2.03) | 1.38 (0.96–1.98) | |
| Model 2 | 1.0 (ref) | 1.24 (0.81–1.92) | 1.49 (0.98–2.25) | 1.45 (0.96–2.19) | |
| Blacks | |||||
| N cases | 39 | 43 | 66 | 79 | |
| N controls | 104 | 124 | 122 | 128 | |
| Model 1 | 1.0 (ref) | 0.90 (0.51–1.59) | 1.47 (0.87–2.49) | 1.74 (1.03–2.94)† | |
| Model 2 | 1.0 (ref) | 0.95 (0.50–1.81) | 1.48 (0.82–2.67) | 1.96 (1.10–3.46)† | |
| Whites | |||||
| N cases | 60 | 71 | 108 | 81 | |
| N controls | 111 | 117 | 143 | 118 | |
| Model 1 | 1.0 (ref) | 1.14 (0.70–1.86) | 1.36 (0.86–2.14) | 1.16 (0.72–1.87) | |
| Model 2 | 1.0 (ref) | 1.59 (0.87–2.90) | 1.39 (0.79–2.47) | 1.14 (0.64–2.04) | |
| Race interaction, model 1/2 | 0.37/0.12 | ||||
| Women | |||||
| N cases | 49 | 57 | 86 | 73 | |
| N controls | 107 | 117 | 137 | 127 | |
| Model 1 | 1.0 (ref) | 0.97 (0.58–1.63) | 1.27 (0.78–2.07) | 1.17 (0.71–1.94) | |
| Model 2 | 1.0 (ref) | 1.10 (0.61–2.01) | 1.32 (0.75–2.31) | 1.23 (0.68–2.21) | |
| Men | |||||
| N cases | 50 | 57 | 88 | 87 | |
| N controls | 108 | 124 | 128 | 119 | |
| Model 1 | 1.0 (ref) | 1.11 (0.66–1.86) | 1.55 (0.95–2.52) | 1.63 (0.99–2.67) | |
| Model 2 | 1.0 (ref) | 1.33 (0.73–2.42) | 1.61 (0.91–2.86) | 1.62 (0.90–2.92) | |
| Sex interaction, model 1/2 | 0.88/0.96 | ||||
| Black women | |||||
| N cases | 24 | 25 | 29 | 47 | |
| N controls | 52 | 60 | 62 | 68 | |
| Model 1 | 1.0 (ref) | 0.94 (0.44–1.99) | 1.12 (0.54–2.32) | 1.55 (0.77–3.13) | |
| Model 2 | 1.0 (ref) | 1.00 (0.40–2.50) | 1.27 (0.56–2.92) | 1.90 (0.86–4.17) | |
| Black men | |||||
| N cases | 15 | 18 | 37 | 32 | |
| N controls | 52 | 64 | 60 | 60 | |
| Model 1 | 1.0 (ref) | 0.98 (0.42–2.28) | 2.21 (1.02–4.77)† | 2.17 (0.98–4.78) | |
| Model 2 | 1.0 (ref) | 0.66 (0.23–1.93) | 1.93 (0.81–4.65) | 2.04 (0.79–5.32) | |
| White women | |||||
| N cases | 25 | 32 | 57 | 26 | |
| N controls | 55 | 57 | 75 | 59 | |
| Model 1 | 1.0 (ref) | 1.01 (0.49–2.07) | 1.35 (0.69–2.64) | 0.84 (0.40–1.76) | |
| Model 2 | 1.0 (ref) | 1.79 (0.69–4.67) | 1.31 (0.54–3.20) | 0.78 (0.30–2.02) | |
| White men | |||||
| N cases | 35 | 39 | 51 | 55 | |
| N controls | 56 | 60 | 68 | 59 | |
| Model 1 | 1.0 (ref) | 1.22 (0.63–2.38) | 1.28 (0.68–2.39) | 1.41 (0.75–2.65) | |
| Model 2 | 1.0 (ref) | 1.76 (0.75–4.11) | 1.34 (0.61–2.96) | 1.53 (0.64–3.65) | |
HRs are compared with those in the first quartile (ref).
CI indicates confidence interval; HR, hazard ratio; Lp(a), lipoprotein(a); and ref, reference.
*Model 1: Adjusted for age, race, sex, and age×race. Interaction terms for Lp(a) quartiles, model 1: age=0.82, race=0.37, and sex=0.88.
Model 2: Adjusted for age, sex, and Framingham stroke risk factors (atrial fibrillation, smoking, prior cardiovascular disease, left ventricular hypertrophy, and use of any antihypertensive medication). Interaction terms for Lp(a) quartiles: age=0.77, race=0.12, and sex=0.96.
The sensitivity analysis by stroke subtype did not reveal a statistically significant association between elevated Lp(a) and specific ischemic stroke subtypes owing to small numbers (Table 4). However, the patterns of association suggested similar associations as for overall stroke except that there was no association with small vessel subtype. Furthermore, there was no difference in the association of Lp(a) with stroke risk by statin use (interaction P value=0.37 for model 1 and P=0.29 for model 2). Also, the HRs of stroke with higher Lp(a) were similar after adding statin use to the models (Table I in the online-only Data Supplement).
| Subtype | HR Stroke | |||
|---|---|---|---|---|
| Second Quartile | Third Quartile | Fourth Quartile | ||
| HR (95% CI) | HR (95% CI) | HR (95% CI) | ||
| Cardioembolic | N | |||
| All (107 cases) | ||||
| Model 1 | 1073 | 1.02 (0.50–2.07) | 1.55 (0.82–2.95) | 1.76 (0.93–3.34) |
| Model 2 | 984 | 1.43 (0.61–3.34) | 1.41 (0.64–3.14) | 1.63 (0.73–3.64) |
| Black (36 cases) | ||||
| Model 1 | 514 | 1.20 (0.38–3.84) | 1.61 (0.53–4.89) | 2.04 (0.70–6.00) |
| Model 2 | 472 | 1.71 (0.42–7.03) | 1.88 (0.49–7.19) | 3.25 (0.92–11.52) |
| White (71 cases) | ||||
| Model 1 | 559 | 0.92 (0.38–2.25) | 1.49 (0.68–3.25) | 1.65 (0.75–3.64) |
| Model 2 | 512 | 1.18 (0.36–3.89) | 0.98 (0.33–2.87) | 1.02 (0.33–3.16) |
| Large vessel disease | ||||
| All (74 cases) | ||||
| Model 1 | 1040 | 1.42 (0.61–3.30) | 1.98 (0.90–4.32) | 2.18 (1.00. 4.77) |
| Model 2 | 948 | 1.55 (0.62–3.88) | 1.63 (0.69–3.84) | 2.00 (0.84–4.75) |
| Black (26 cases) | ||||
| Model 1 | 504 | 1.75 (0.41–7.40) | 2.74 (0.70–10.74) | 2.35 (0.57–9.63) |
| Model 2 | 459 | 1.45 (0.29–7.25) | 1.90 (0.39–9.15) | 2.15 (0.44–10.43) |
| White (48 cases) | ||||
| Model 1 | 536 | 1.29 (0.45–3.71) | 1.69 (0.65–4.39) | 2.09 (0.81–5.39) |
| Model 2 | 489 | 1.81 (0.48–6.74) | 1.47 (0.44–4.97) | 2.30 (0.68–7.75) |
| Small vessel disease | ||||
| All (85 cases) | ||||
| Model 1 | 1051 | 1.13 (0.57–2.24) | 1.55 (0.82–2.92) | 0.64 (0.30–1.39) |
| Model 2 | 962 | 1.25 (0.56–2.77) | 1.99 (0.99–4.00) | 0.69 (0.30–1.57) |
| Black (42 cases) | ||||
| Model 1 | 520 | 0.65 (0.23–1.86) | 1.28 (0.50–3.26) | 1.15 (0.44–2.97) |
| Model 2 | 476 | 0.66 (0.18–2.37) | 1.50 (0.53–4.23) | 1.16 (0.41–3.33) |
| White (43 cases) | ||||
| Model 1 | 531 | 1.66 (0.66–4.18) | 1.64 (0.67–4.04) | 0.10 (0.01–0.86) |
| Model 2 | 486 | 2.75 (0.76–9.95) | 2.58 (0.83–7.97) | 0.15 (0.02–1.39) |
| Other | ||||
| All (35 cases) | ||||
| Model 1 | 1001 | 2.18 (0.64–7.38) | 2.37 (0.72–7.76) | 2.46 (0.76–8.02) |
| Model 2 | 913 | 2.30 (0.57–9.31) | 2.15 (0.60–7.72) | 1.98 (0.55–7.07) |
| Black (14 cases)* | ||||
| Model 1 | 492 | … | … | … |
| Model 2 | 450 | … | … | … |
| White (21 cases) | ||||
| Model 1 | 509 | 0.99 (0.23–4.24) | 1.40 (0.39–5.10) | 1.30 (0.35–4.82) |
| Model 2 | 463 | 0.81 (0.14–4.91) | 1.34 (0.29–6.19) | 1.11 (0.21–5.79) |
| Unknown | ||||
| All (243 cases) | ||||
| Model 1 | 1209 | 0.87 (0.54–1.39) | 1.18 (0.76–1.83) | 1.27 (0.82–1.98) |
| Model 2 | 1105 | 0.98 (0.57–1.67) | 1.30 (0.79–2.14) | 1.46 (0.88–2.40) |
| Black (108 cases) | ||||
| Model 1 | 586 | 0.66 (0.32–1.38) | 1.21 (0.63–2.34) | 1.68 (0.89–3.16) |
| Model 2 | 537 | 0.63 (0.27–1.43) | 1.24 (0.60–2.55) | 1.98 (1.00–3.93) |
| White (135 cases) | ||||
| Model 1 | 623 | 1.03 (0.56–1.90) | 1.12 (0.63–1.99) | 1.01 (0.55–1.84) |
| Model 2 | 568 | 1.45 (0.68–3.11) | 1.33 (0.65–2.69) | 1.10 (0.53–2.27) |
HR are compared to those in the first quartile (ref). Model definitions given in Table 3.
CI indicates confidence interval; HR, hazard ratio; and Lp(a), lipoprotein(a).
*
Other black: 0 cases in Q1, so HR undefined.
Discussion
In this population-based cohort, Lp(a) concentration was higher in black participants than white participants and in women compared with men. Elevated Lp(a) was associated with an increased risk of ischemic stroke after adjustment for age, gender, and race. This trend was stronger for black than white participants, although the racial difference was not statistically significant. This relationship was preserved after adjustment for traditional stroke risk factors. Men had significantly lower Lp(a) values compared with women, but the association of Lp(a) with ischemic stroke did not differ between men and women. Finally, baseline statin use did not alter the association between Lp(a) and stroke, and Lp(a) was not differentially related to specific stroke subtypes, with the possible exception of no association with small vessel subtype.
There are many possible mechanistic explanations for our results. Lp(a) appears to accelerate atherogenesis.19 Mounting epidemiologic20,21 and genetic data22 suggest a causal relationship between Lp(a) and stroke. This explains the stronger relationship between Lp(a) and stroke at higher Lp(a) concentrations that we observed in the analysis of all participants. The basis of racial differences in Lp(a) has been explored previously. There are well-recognized racial variations in the apo(a) [apolipoprotein(a)] particle that makes up Lp(a). The size of apo(a) size determines the size of Lp(a), which is it itself inversely correlated with overall concentrations of Lp(a).23 In a cohort from New York, black participants had smaller Lp(a) isoforms, a higher overall concentration of Lp(a), and there was a stronger association of Lp(a) with coronary artery disease on coronary angiography compared with white participants.24 Moreover, Lp(a) is increased by conditions like chronic kidney disease.25 This may further explain the racial difference in levels, as blacks have a higher incidence and prevalence of renal disease. However, chronic kidney disease was not associated with stroke risk in REGARDS, so we did not consider it as a potential confounder. Additionally, nonwhites have an acknowledged higher rate of statin nonadherence.26 Therefore, the cardiovascular disease risk that high levels of Lp(a) confer in blacks may be associated with a failure to address other established cardiovascular risk factors because of a disparity in statin usage. We conducted a sensitivity analysis adjusting for baseline statin usage but did not find any evidence for mediation by baseline statin use in the hazards related to Lp(a).
We reiterated prior findings that Lp(a) levels are higher in blacks,1 in participants with cardiovascular disease,27 and in those with higher LDL-C concentrations.28 Prior data from other large, population-based cohort studies have been conflicting on stroke, with some studies linking elevated Lp(a) levels to a higher incidence of ischemic stroke,7–10 whereas others have not found an association.11–13 This may be partly attributable to lack of differentiation between the subtypes of incident stroke13 or to racial or other differences in cohort composition. This may also be secondary to confounding from LDL-C levels, which is an important risk factor for stroke.5,6 We evaluated the association of Lp(a) concentration with stroke subtypes with inconclusive findings, but no association with the small vessel subtype. A study with a larger number of strokes would be needed to confirm this finding. To the best of our knowledge, most previous studies that have addressed the association of Lp(a) with stroke or subtypes have not included a large number of black participants. Though the statistical testing for racial differences in association did not reach statistical significance in our investigation, we consider our findings to be of clinical significance. The ARIC (Atherosclerosis Risk in Communities) cohort investigators studied the relationship between Lp(a) and ischemic stroke; Ohira et al10 reported an association with ischemic stroke in black participants and white women but not white men. Our findings, which included 76 more stroke cases compared with ARIC, are consistent with Ohira et al10 about the relationship in black participants. Additionally, REGARDS had a wider geographic catchment area than the ARIC cohort did. Virani et al27 later reported 20-year follow-up results on association of Lp(a) levels with the risk of both coronary heart disease events and ischemic stroke in the ARIC cohort. Virani et al27 confirmed the previous findings from the ARIC cohort that Lp(a) was associated with incident ischemic stroke in black participants. In that larger ARIC study with 663 ischemic stroke events and longer follow-up, the investigators did not find any differences by sex in association of Lp(a) with ischemic stroke. This is concordant with our results from REGARDS.
Our results have important clinical implications. We add to the growing body of literature from large, prospective cohort studies noting the correlation between Lp(a) and ischemic stroke. It is estimated that 1.5 billion people have Lp(a) >50 mg/dL.29 Additionally, the effect of lifestyle and dietary changes on Lp(a) levels remain somewhat unclear.3 This suggests that Lp(a) warrants serious consideration as a cardiovascular risk factor and provides justification for study of direct targeting of levels. Its current use in prognostication of cardiovascular disease remains largely limited to specialist centers and clinical trials. Lp(a) has also gained a greater clinical presence with the publication of the 2018 multisociety cholesterol management guidelines, where Lp(a) levels ≥50 mg/dL or 125 nmol/L were highlighted as an important risk modifier for patients with a history of atherosclerotic cardiovascular disease that is unexplained by major risk factors or with a family history of atherosclerotic cardiovascular disease.30 This is pertinent to our investigation, as over 25% of subjects (particularly black subjects), had Lp(a) levels higher than this threshold. Along with this, the concomitant approval of 2 International Classification of Diseases, Tenth Revision diagnostic codes by the Centers of Disease Control and Surveillance for elevated Lp(a) may increase the usage of Lp(a) testing in clinical practice.31 There is also a recognized heterogeneity in the measurement methods for Lp(a), with little consensus on a gold standard.19,32 In terms of pharmacological interventions, proprotein convertase subtilisin/kexin-9 inhibitors reduce Lp(a) to unprecedented levels.33 This therapeutic class offers an exciting option to address this stroke risk factor that is largely unaffected by statin therapy. Mipomersen and lopitamide are also potential therapeutic options targeting Lp(a).34 Finally, there is an antisense oligonucleotide in development that targets Lp(a).35
We acknowledge that our investigation has limitations. Cohort studies can be hindered by loss to follow-up. Cohort retention in REGARDS was excellent at 97.1% annually in whites and 96.2% annually in blacks, similar to other population-based cohort studies. The REGARDS cohort also did not include participants who were not white or black. However, the purpose of this cohort study was to examine the starkest racial disparity in stroke—affecting blacks in the United States. Although we observed differences by race in the relationship between Lp(a) and ischemic stroke, the interaction was not statistically significant.2 These findings should be confirmed in other studies or with longer follow-up in REGARDS. The association between Lp(a) and ischemic stroke in other racial groups warrants further investigation. We acknowledge the wide variety of assays available for measuring Lp(a) (Table II in the online-only Data Supplement). This may make it difficult to interpret study findings from studies using different methodology and has been identified as an area that needs to be addressed further.19 A strength of the current study is use of a single automated assay with a low analytical CV, which reduces imprecision in measurement. However, bias because of sample matrix effects in diluted samples cannot be ruled out. We also note that the HR for the cardioembolic stroke subtype trended towards significance. Prior data suggest that Lp(a) may be associated with upregulation of the coagulation pathways and impaired fibrinolysis,36 which could be a predisposing factor for the formation of cardiac thrombi that may lead to cardioembolic stroke. Unfortunately, this was a secondary analyses and our study power was limited to evaluate this association more fully; further confirmation and investigation are warranted. It is important to recognize the different distribution of Lp(a) among whites and blacks. In whites, the distribution is highly skewed, whereas the Lp(a) distribution follows a more normal distribution in blacks.2 In this analysis, we used race-specific quartile definitions so we could accurately determine the association of Lp(a) with stroke in blacks and whites separately.
In conclusion, elevated Lp(a) is an emerging risk factor for ischemic stroke. Elevated Lp(a) may also be associated with the racial disparity in stroke affecting blacks in the United States, and thus may be a race-specific risk factor for ischemic stroke. More research is needed to confirm the racial differences in Lp(a) and to evaluate apo(a) isoforms to identify the role of Lp(a) in the prognostication and prevention of ischemic stroke.
Acknowledgments
We thank investigators, staff, and participants of the REGARDS study (Reasons for Geographic and Racial Differences in Stroke) for their valuable contributions. A full list of investigators and institutions can be found at http://www.regardsstudy.org.
Footnote
Nonstandard Abbreviations and Acronyms
- apo(a)
- apolipoprotein(a)
- ARIC
- Atherosclerosis Risk in Communities
- HDL-C
- high-density lipoprotein cholesterol
- HR
- hazard ratio
- LDL-C
- low-density lipoprotein cholesterol
- Lp(a)
- lipoprotein(a)
- REGARDS
- Reasons for Geographic and Racial Differences in Stroke
Supplemental Material
File (atvb_atvb-2018-311857d_supp1.pdf)
- Download
- 32.47 KB
References
1.
Dubé JB, Boffa MB, Hegele RA, Koschinsky ML. Lipoprotein(a): more interesting than ever after 50 years. Curr Opin Lipidol. 2012;23:133–140. doi: 10.1097/MOL.0b013e32835111d8
2.
Tsimikas S. A test in context: lipoprotein(a): diagnosis, prognosis, controversies, and emerging therapies. J Am Coll Cardiol. 2017;69:692–711. doi: 10.1016/j.jacc.2016.11.042
3.
Milionis HJ, Winder AF, Mikhailidis DP. Lipoprotein (a) and stroke. J Clin Pathol. 2000;53:487–496.
4.
Sacco RL, Benjamin EJ, Broderick JP, Dyken M, Easton JD, Feinberg WM, Goldstein LB, Gorelick PB, Howard G, Kittner SJ, Manolio TA, Whisnant JP, Wolf PA. American Heart Association Prevention Conference. IV. Prevention and rehabilitation of stroke. Risk factors. Stroke. 1997;28:1507–1517.
5.
Anderson TJ, Grégoire J, Hegele RA, et al. 2012 update of the Canadian Cardiovascular Society guidelines for the diagnosis and treatment of dyslipidemia for the prevention of cardiovascular disease in the adult. Can J Cardiol. 2013;29:151–167. doi: 10.1016/j.cjca.2012.11.032
6.
Powers WJ, Rabinstein AA, Ackerson T, et al; American Heart Association Stroke Council. 2018 guidelines for the early management of patients with acute ischemic stroke: A guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2018;49:e46–e110. doi: 10.1161/STR.0000000000000158
7.
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.
8.
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.
9.
Ariyo AA, Thach C, Tracy R; Cardiovascular Health Study Investigators. Lp(a) lipoprotein, vascular disease, and mortality in the elderly. N Engl J Med. 2003;349:2108–2115. doi: 10.1056/NEJMoa001066
10.
Ohira T, Schreiner PJ, Morrisett JD, Chambless LE, Rosamond WD, Folsom AR. Lipoprotein(a) and incident ischemic stroke: the Atherosclerosis Risk in Communities (ARIC) study. Stroke. 2006;37:1407–1412. doi: 10.1161/01.STR.0000222666.21482.b6
11.
Ridker PM, Stampfer MJ, Hennekens CH. Plasma concentration of lipoprotein(a) and the risk of future stroke. JAMA. 1995;273:1269–1273.
12.
Glader CA, Stegmayr B, Boman J, Stenlund H, Weinehall L, Hallmans G, Dahlén 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.
13.
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.
14.
Howard VJ, Cushman M, Pulley L, Gomez CR, Go RC, Prineas RJ, Graham A, Moy CS, Howard G. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25:135–143. doi: 10.1159/000086678
15.
Gillett SR, Boyle RH, Zakai NA, McClure LA, Jenny NS, Cushman M. Validating laboratory results in a national observational cohort study without field centers: the reasons for geographic and racial differences in Stroke cohort. Clin Biochem. 2014;47:243–246. doi: 10.1016/j.clinbiochem.2014.08.003
16.
Howard VJ, Kleindorfer DO, Judd SE, McClure LA, Safford MM, Rhodes JD, Cushman M, Moy CS, Soliman EZ, Kissela BM, Howard G. Disparities in stroke incidence contributing to disparities in stroke mortality. Ann Neurol. 2011;69:619–627. doi: 10.1002/ana.22385
17.
Zakai NA, Judd SE, Alexander K, McClure LA, Kissela BM, Howard G, Cushman M. ABO blood type and stroke risk: the REasons for Geographic And Racial Differences in Stroke Study. J Thromb Haemost. 2014;12:564–570. doi: 10.1111/jth.12507
18.
Feise RJ. Do multiple outcome measures require p-value adjustment? BMC Med Res Methodol. 2002;2:8.
19.
Nordestgaard BG, Chapman MJ, Ray K, et al; European Atherosclerosis Society Consensus Panel. Lipoprotein(a) as a cardiovascular risk factor: current status. Eur Heart J. 2010;31:2844–2853. doi: 10.1093/eurheartj/ehq386
20.
Smolders B, Lemmens R, Thijs V. Lipoprotein (a) and stroke: a meta-analysis of observational studies. Stroke. 2007;38:1959–1966. doi: 10.1161/STROKEAHA.106.480657
21.
Erqou S, Kaptoge S, Perry PL, Di Angelantonio E, Thompson A, White IR, Marcovina SM, Collins R, Thompson SG, Danesh J. Lipoprotein(a) concentration and the risk of coronary heart disease, stroke, and nonvascular mortality. JAMA. 2009;302:412–423.
22.
Lee SR, Prasad A, Choi YS, Xing C, Clopton P, Witztum JL, Tsimikas S. LPA gene, ethnicity, and cardiovascular events. Circulation. 2017;135:251–263. doi: 10.1161/CIRCULATIONAHA.116.024611
23.
Gavish D, Azrolan N, Breslow JL. Plasma Ip(a) concentration is inversely correlated with the ratio of Kringle IV/Kringle V encoding domains in the apo(a) gene. J Clin Invest. 1989;84:2021–2027. doi: 10.1172/JCI114395
24.
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.
25.
Kronenberg F, König P, Neyer U, Auinger M, Pribasnig A, Lang U, Reitinger J, Pinter G, Utermann G, Dieplinger H. Multicenter study of lipoprotein(a) and apolipoprotein(a) phenotypes in patients with end-stage renal disease treated by hemodialysis or continuous ambulatory peritoneal dialysis. J Am Soc Nephrol. 1995;6:110–120.
26.
Lewey J, Shrank WH, Bowry AD, Kilabuk E, Brennan TA, Choudhry NK. Gender and racial disparities in adherence to statin therapy: a meta-analysis. Am Heart J. 2013;165:665.e1–678.e1. doi: 10.1016/j.ahj.2013.02.011
27.
Virani SS, Brautbar A, Davis BC, Nambi V, Hoogeveen RC, Sharrett AR, Coresh J, Mosley TH, Morrisett JD, Catellier DJ, Folsom AR, Boerwinkle E, Ballantyne CM. Associations between lipoprotein(a) levels and cardiovascular outcomes in black and white subjects: the Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 2012;125:241–249. doi: 10.1161/CIRCULATIONAHA.111.045120
28.
Jacobson TA. Lipoprotein(a), cardiovascular disease, and contemporary management. Mayo Clin Proc. 2013;88:1294–1311. doi: 10.1016/j.mayocp.2013.09.003
29.
Saeedi R, Frohlich J. Lipoprotein (a), an independent cardiovascular risk marker. Clin Diabetes Endocrinol. 2016;2:7. doi: 10.1186/s40842-016-0024-x
30.
Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol. A report of the American College of Cardiology/American Heart Association Task Force on clinical practice guidelines [published online ahead of print November 3, 2018]. Circulation. doi: 10.1161/CIR.0000000000000625. https://www.ahajournals.org/doi/abs/10.1161/CIR.0000000000000625. Accessed November 3, 2018.
31.
Tremulis, S. CDC grants ICD-10 codes of elevated lipoprotein(a). https://www.lipoproteinafoundation.org/news/407398/cdc-grants-icd-10-codes-of-elevated-lipoproteina.htm. Accessed November 20, 2018.
32.
Tsimikas S, Hall JL. Lipoprotein(a) as a potential causal genetic risk factor of cardiovascular disease: a rationale for increased efforts to understand its pathophysiology and develop targeted therapies. J Am Coll Cardiol. 2012;60:716–721. doi: 10.1016/j.jacc.2012.04.038
33.
Sabatine MS, Giugliano RP, Wiviott SD, Raal FJ, Blom DJ, Robinson J, Ballantyne CM, Somaratne R, Legg J, Wasserman SM, Scott R, Koren MJ, Stein EA; Open-Label Study of Long-Term Evaluation against LDL Cholesterol (OSLER) Investigators. Efficacy and safety of evolocumab in reducing lipids and cardiovascular events. N Engl J Med. 2015;372:1500–1509. doi: 10.1056/NEJMoa1500858
34.
Man LC, Kelly E, Duffy D. Targeting lipoprotein (a): an evolving therapeutic landscape. Curr Atheroscler Rep. 2015;17:502. doi: 10.1007/s11883-015-0502-0
35.
Graham MJ, Viney N, Crooke RM, Tsimikas S. Antisense inhibition of apolipoprotein (a) to lower plasma lipoprotein (a) levels in humans. J Lipid Res. 2016;57:340–351. doi: 10.1194/jlr.R052258
36.
Dahl T, Kontny F, Slagsvold CE, Christophersen B, Abildgaard U, Odegaard OR, Morkrid L, Dale J. Lipoprotein(a), other lipoproteins and hemostatic profiles in patients with ischemic stroke: the relation to cardiogenic embolism. Cerebrovasc Dis. 2000;10:110–117. doi: 10.1159/000016039
Information & Authors
Information
Published In
Copyright
© 2019 American Heart Association, Inc.
Versions
You are viewing the most recent version of this article.
History
Received: 18 August 2018
Accepted: 28 January 2019
Published online: 21 February 2019
Published in print: April 2019
Keywords
Subjects
Authors
Disclosures
None.
Sources of Funding
This research project is supported by a cooperative agreement U01 NS041588 cofunded by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute on Aging (NIA), National Institutes of Health, and Department of Health and Human Service. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINDS or the NIA. Representatives of the funding agency have been involved in the review of the article but not directly involved in the collection, management, analysis, or interpretation of the data. K.S. Alexander was supported by National Heart, Lung, and Blood Institute T32HL007594-27.
Metrics & Citations
Metrics
Citations
Download Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.
- Lipoprotein(a) and Risk of Incident Atherosclerotic Cardiovascular Disease: Impact of High-Sensitivity C-Reactive Protein and Risk Variability Among Human Clinical Subgroups, Nutrients, 17, 8, (1324), (2025).https://doi.org/10.3390/nu17081324
- Lp(a): A clinical review, Clinical Biochemistry, (110929), (2025).https://doi.org/10.1016/j.clinbiochem.2025.110929
- Rethinking cardiovascular risk: The emerging role of lipoprotein(a) screening, American Journal of Preventive Cardiology, 21, (100945), (2025).https://doi.org/10.1016/j.ajpc.2025.100945
- Lipoprotein (a) in the context of atherosclerosis: pathological implications and therapeutic perspectives in myocardial infarction. A narrative review, Romanian Journal of Morphology and Embryology, 65, 4, (609-616), (2025).https://doi.org/10.47162/RJME.65.4.07
- Association of lipoprotein(a) and coronary artery calcium with atherosclerotic cardiovascular disease, Journal of Clinical Lipidology, (2025).https://doi.org/10.1016/j.jacl.2025.02.007
- Lipoprotein(a) levels in a population with clinical atherosclerotic cardiovascular disease in the United States: A subanalysis from the Lp(a)HERITAGE study, Journal of Clinical Lipidology, 19, 1, (28-38), (2025).https://doi.org/10.1016/j.jacl.2024.11.007
- Lipoprotein(a) as a Risk Factor for Recurrent Acute Myocardial Infarction and Mortality: Insights from Routine Clinical Practice, Diagnostics, 14, 23, (2757), (2024).https://doi.org/10.3390/diagnostics14232757
- Association of white matter hyperintensities with lipoprotein (a) levels: insights from a cohort study, Frontiers in Neurology, 15, (2024).https://doi.org/10.3389/fneur.2024.1476005
- Lipoprotein (a): Underrecognized Risk with a Promising Future, Reviews in Cardiovascular Medicine, 25, 11, (2024).https://doi.org/10.31083/j.rcm2511393
- A Retrospective Analysis of the Underlying Health Status of Patients Treated for Stroke in the Emergency Department of a Community Hospital Situated in a Health Professional Shortage Area, Cureus, (2024).https://doi.org/10.7759/cureus.68150
- See more
Loading...
View Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginPurchase Options
Purchase this article to access the full text.
eLetters(0)
eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.
Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.