Association of Lp(a) With Stroke and Cerebral Injury on MRI: Insights From the HCHS/SOL (Hispanic Community Health Study/Study of Latinos) and Investigation of Neurocognitive Aging MRI (SOL-INCA MRI).

BACKGROUND: Lp(a) (lipoprotein[a]) is a risk factor for cardiovascular disease; however, its association with cerebrovascular disease is not as well established. METHODS: Data from a population-based cohort of Hispanics/Latinos included 16 333 individuals with baseline Lp(a) levels (nmol/L) and self-reported prevalent stroke or transient ischemic attack (TIA). A subset of 2642 individuals with brain magnetic resonance imaging was also included. Linear and multivariate logistic regression assessed the association of Lp(a) with (1) self-reported stroke or TIA, (2) cerebral injury defined as self-reported stroke or TIA or evidence of a stroke on brain magnetic resonance imaging, (3) white matter hyperintensity volume, and (4) silent brain infarcts. Sampling weights were utilized given the HCHS/SOL (Hispanic Community Health Study/Study of Latinos) complex sample design. RESULTS: Mean age&plusmn;SE was 41.1&plusmn;0.3 years, 52.0% female, and median interquartile range (Q₁, Q₃) Lp(a) level of 19.7 (7.3-60.6) nmol/L; brain magnetic resonance imaging subset mean age&plusmn;SE was 49.9&plusmn;0.4 years, 56.4% female, and median (interquartile range) Lp(a) level of 21.7 (8.1-62.9) nmol/L. Each unit increase in log-transformed Lp(a) was associated with higher odds of self-reported stroke or TIA (odds ratio, 1.13 [95% CI, 1.01-1.27]; <em>P</em>=0.03). Lp(a) levels in the highest quintile (&gt;77 nmol/L) were significantly associated with higher odds of prevalent stroke or TIA compared with Lp(a) &lt;6 nmol/L (first quintile: odds ratio, 1.74 [95% CI, 1.09-2.77]; <em>P</em>=0.02). The highest proportion of cerebral injury was noted in Q5, while the lowest proportion was noted in Q2. When comparing Lp(a) &gt;77 nmol/L with Lp(a) of 6 to &lt;13 nmol/L (second quintile), a significant association was found between Lp(a) and cerebral injury that persisted after fully adjusted models (odds ratio, 2.03 [95% CI, 1.05-3.93]; <em>P</em>=0.03). Each unit increase in log-Lp(a) was associated with a 0.10 increase in log-white matter hyperintensity (&beta;, 0.10; <em>P</em>=0.005). No significant association was found between Lp(a) and silent brain infarcts. CONCLUSIONS: Lp(a) is independently and significantly associated with prevalent stroke/TIA, and white matter hyperintensity, in a large diverse population of Hispanics/Latinos.