High-Sensitivity Cardiac Troponin T and Cognitive Function Over 12 Months After Stroke-Results of the DEMDAS Study

Abstract

Background: Subclinical myocardial injury in form of hs-cTn (high-sensitivity cardiac troponin) levels has been associated with cognitive impairment and imaging markers of cerebral small vessel disease (SVD) in population-based and cardiovascular cohorts. Whether hs-cTn is associated with domain-specific cognitive decline and SVD burden in patients with stroke remains unknown.

Methods and results: We analyzed patients with acute stroke without premorbid dementia from the prospective multicenter DEMDAS (DZNE [German Center for Neurodegenerative Disease]-Mechanisms of Dementia after Stroke) study. Patients underwent neuropsychological testing 6 and 12 months after the index event. Test results were classified into 5 cognitive domains (language, memory, executive function, attention, and visuospatial function). SVD markers (lacunes, cerebral microbleeds, white matter hyperintensities, and enlarged perivascular spaces) were assessed on cranial magnetic resonance imaging to constitute a global SVD score. We examined the association between hs-cTnT (hs-cTn T levels) and cognitive domains as well as the global SVD score and individual SVD markers, respectively. Measurement of cognitive and SVD-marker analyses were performed in 385 and 466 patients with available hs-cTnT levels, respectively. In analyses adjusted for demographic characteristics, cardiovascular risk factors, and cognitive status at baseline, higher hs-cTnT was negatively associated with the cognitive domains "attention" up to 12 months of follow-up (beta-coefficient, -0.273 [95% CI, -0.436 to -0.109]) and "executive function" after 12 months. Higher hs-cTnT was associated with the global SVD score (adjusted odds ratio, 1.95 [95% CI, 1.27-3.00]) and the white matter hyperintensities and lacune subscores.

Conclusions: In patients with stroke, hs-cTnT is associated with a higher burden of SVD markers and cognitive function in domains linked to vascular cognitive impairment.

Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT01334749.

Keywords: acute stroke; cardiac troponin; cognitive impairment; heart and brain axis.

Figure 1. Log‐transformed hs‐cTnT and cognitive domains (continuous) across 12 months of follow‐up according to generalized linear regression models using generalized estimating equations.

The figure displays the respective regression coefficients and 95% CIs. Model 1: unadjusted. Model 2: adjusted for age, sex, and years of education. Model 3: additional adjustment for hypertension, diabetes, coronary artery disease, atrial fibrillation, cognitive impairment at baseline, baseline NIHSS score, and prestroke mRS score. *P _corr<0.05. After full adjustment, hs‐cTnT was associated with a decline in performance in the cognitive domain “attention” between 6 and 12 months after stroke. Hs‐cTnT indicates high‐sensitivity cardiac troponin T; mRS, modified Rankin Scale; and NIHSS, National Institutes of Health Stroke Scale.

Figure 2. Log‐transformed hs‐cTnT and cognitive domains (continuous) at 6 months of follow‐up according to linear regression models.

The figure displays the respective regression coefficients and 95% CIs. Model 1: unadjusted. Model 2: adjusted for age, sex, and years of education. Model 3: additional adjustment for hypertension, diabetes, coronary artery disease, atrial fibrillation, cognitive impairment at baseline, baseline NIHSS score, and prestroke mRS score. *P _corr<0.05. After full adjustment, hs‐cTnT was negatively associated with performance in the cognitive domain “attention” in the cross‐sectional analyses 6 months after stroke. Hs‐cTnT indicates high‐sensitivity cardiac troponin T; mRS, modified Rankin Scale; and NIHSS, National Institutes of Health Stroke Scale.

Figure 3. Log‐transformed hs‐cTnT and cognitive domains (continuous) at 12 months of follow‐up according to linear regression models.

The figure displays the respective odds ratios and 95% CIs. Model 1: unadjusted. Model 2: adjusted for age, sex, and years of education. Model 3: additional adjustment for hypertension, diabetes, coronary artery disease, atrial fibrillation, cognitive impairment at baseline, baseline NIHSS score, and prestroke mRS score. *P _corr<0.05. After full adjustment, hs‐cTnT was negatively associated with performance in the domains “attention” and “executive function” in the cross‐sectional analyses 6 months after stroke. Hs‐cTnT indicates high‐sensitivity cardiac troponin T; mRS, modified Rankin Scale; and NIHSS, National Institutes of Health Stroke Scale.

Figure 4. Log‐transformed hs‐cTnT and global cerebral small vessel disease score as well as 4 constituent SVD subscores.

The figure displays odds ratios and 95% CIs derived from ordinal logistic regression models for the global SVD score and binary logistic regression models for each constituent subscore, respectively. Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: additional adjustment for hypertension, diabetes, hyperlipidemia, coronary artery disease, atrial fibrillation, smoking status, and baseline NIHSS score. After full adjustment, hs‐cTnT was associated with higher global SVD scores as well as the WMH and lacune subscores. *P _corr<0.05. CMB indicates cerebral microbleeds; hs‐cTnT, high‐sensitivity cardiac troponin T; NIHSS, National Institutes of Health Stroke Scale; PVS, perivascular spaces; SVD, small vessel disease; and WMH, white matter hyperintensities.

Figure 5. Hs‐cTnT and individual cerebral small vessel disease markers in their entire range.

The figure displays odds ratios derived from ordinal regression models for periventricular WMH grade, deep WMH grade, and PVS grade. The figure displays rate ratios calculated using negative binomial regression models for lacune count and CMB count. Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: additional adjustment for hypertension, diabetes, hyperlipidemia, coronary artery disease, atrial fibrillation, smoking status, and baseline NIHSS score. After full adjustment, hs‐cTnT remained associated with higher deep WMH grade. *P _corr<0.05. CMB indicates cerebral microbleeds; DWM, deep white matter; hs‐cTnT, high‐sensitivity cardiac troponin T; NIHSS, National Institutes of Health Stroke Scale; PVS, perivascular spaces; PVWM, periventricular white matter; SVD, small vessel disease; and WMH, white matter hyperintensities.