Plasma Pro-Enkephalin A and Incident Cognitive Impairment: The Reasons for Geographic and Racial Differences in Stroke Cohort

Abstract

Background Cardiovascular disease is a risk factor for cognitive impairment. Evidence links both lower and higher concentration of the circulating opioid pro-enkephalin A (PENK-A) with stroke risk. We studied the association of plasma PENK-A with incident cognitive impairment. Methods and Results REGARDS (Reasons for Geographic and Racial Differences in Stroke) is a prospective cohort study of 30 239 adults enrolled from 2003 to 2007. Baseline PENK-A was measured in a nested case-control study of 462 participants who developed cognitive impairment over 4.7 years, and 556 controls. Logistic regression and spline plots adjusted for confounders estimated odds ratios (ORs) of cognitive impairment by baseline PENK-A. Interaction terms tested for differences in associations by age, sex, and race. Baseline PENK-A was comparable between cases and controls. There were significant differences in the association of PENK-A with cognitive impairment by sex and age (adjusted P=0.003 and 0.06, respectively). In women but not men, spline plots showed that higher and lower PENK-A were associated with decreased odds of cognitive impairment (ORs for 10th and 90th percentiles versus median, 0.65 [95% CI, 0.43-0.96] and 0.64 [95% CI, 0.41-0.99]), with no difference by age. In men ≥65 years of age but not younger men, higher PENK-A was associated with decreased odds for cognitive impairment (OR for fourth versus first quartile 0.47 [95% CI, 0.22-0.99]); this pattern was not confirmed with spline plotting. Conclusions High and low levels of circulating opioid PENK-A were associated with decreased odds of future cognitive impairment in specific subgroups. Additional research is warranted to understand the biology underlying this association and the observed differences by sex.

Keywords: cognitive impairment; cohort study; opioid; pro‐enkephalin; risk factor.

Figure 1. Spline plots for the association of plasma pro‐enkephalin A (PENK‐A) with incident cognitive impairment by sex.

Introduction of sex by PENK‐A terms into minimally and fully adjusted models showed significant interaction (P=0.001 and 0.003, respectively). Upper plots show odds (line) and 95% CIs (shading) for incident cognitive impairment relative to median plasma PENK‐A concentration. Splines are plotted with 5 knots. Lower plots show kernel density plots of PENK‐A distribution. Model 1 is unadjusted (A and C). Model 4 is fully adjusted for age, race, age×race interaction, diabetes, hypertension, coronary artery disease, atrial fibrillation, left ventricular hypertrophy, dyslipidemia, smoking status, and estimated glomerular filtration rate (B and D). PENK‐A values corresponding to the 10th percentile were 3.67 ln(pmol/L) for men and 3.74 ln(pmol/L) for women. PENK‐A values corresponding to the 90th percentile were 4.41 ln(pmol/L) for men and 4.57 ln(pmol/L) for women. ln indicates natural log transformation.

Figure 2. Fully adjusted spline plots for the association of plasma pro‐enkephalin A (PENK‐A) with incident cognitive impairment by age and sex.

Upper plots show odds (line) and 95% CIs (shading) for incident cognitive impairment relative to median plasma PENK‐A concentration. Splines are plotted with 5 knots. Analyses are presented by age–sex strata due to a substantial PENK‐A by age×sex interaction term (minimally adjusted P interaction=0.20, fully adjusted P interaction=0.29). Lower plots show kernel density plots of PENK‐A distribution. Each spline is fully adjusted for race, diabetes, hypertension, coronary artery disease, atrial fibrillation, left ventricular hypertrophy, dyslipidemia, smoking status, and estimated glomerular filtration rate. ln indicates natural log transformation.