C-reactive protein levels and risk of dementia-Observational and genetic studies of 111,242 individuals from the general population

Abstract

Introduction: Increased plasma levels of C-reactive protein (CRP) in midlife are associated with increased risk of Alzheimer's disease (AD), whereas in older age the opposite association is observed. Whether genetically determined CRP is associated with AD remains unclear.

Methods: A total of 111,242 White individuals from the Copenhagen General Population Study and the Copenhagen City Heart Study were included. Plasma levels of CRP and four regulatory genetic variants in the CRP gene were determined.

Results: For CRP percentile group 1 to 5 (lowest plasma CRP) versus the 50 to 75 group (reference), the hazard ratio for AD was 1.69 (95% confidence interval 1.29-2.16). Genetically low CRP was associated with increased risk of AD in individuals with body mass index ≤25 kg/m² (P = 4 × 10^-6 ).

Discussion: Low plasma levels of CRP at baseline were associated with high risk of AD in individuals from the general population. These observational findings were supported by genetic studies.

Keywords: Alzheimer's disease; C-reactive protein; CRP; body mass index; gene-environment interaction.

FIGURE 1

Restricted cubic splines illustrating risk of Alzheimer's disease and all‐cause dementia as a function of plasma C‐reactive protein (CRP) on a continuous scale. Solid lines are multifactorially adjusted hazard ratios, whereas dashed lines indicate 95% confidence intervals (CIs) derived from restricted cubic spline regressions with three knots. Graphs are truncated at 22.0 g/L, due to statistically unstable estimates at extremely high levels, thus including 109,926 individuals in these analyses. Cox regression models were adjusted for age (time scale), sex, body mass index, hypertension, diabetes, smoking, alcohol consumption, physical inactivity, menopausal status and hormonal replacement therapy (women only), lipid‐lowering therapy, and education. APOE, apolipoprotein E

FIGURE 2

Plasma levels of C‐reactive protein in percentile groups and risk of Alzheimer's disease and all‐cause dementia. Hazard ratios were multifactorially adjusted for age (as time scale), sex, body mass index, hypertension, diabetes mellitus, smoking, alcohol consumption, physical inactivity, menopausal status and hormonal replacement therapy (women only), lipid‐lowering therapy, and education. The middle column was additionally adjusted for total cholesterol, low‐density lipoprotein cholesterol, high‐density lipoprotein cholesterol, and triglycerides. The right column was additionally adjusted for APOE genotype. APOE, apolipoprotein E

FIGURE 3

Interaction of potential confounders with weighted allele score groups in predicting risk of Alzheimer's disease. Potential confounders were dichotomized: Age (> 65 vs. ≤65 years); sex (men vs. women); study population (CGPS vs. CCHS); total cholesterol (≥ 50 pct vs. < 50 pct); LDL cholesterol (≥ 50 pct vs. < 50 pct); HDL cholesterol (≥ 50 pct vs. < 50 pct); triglycerides (≥ 50 pct vs. < 50 pct); plasma apoE (≥ 50 pct vs. < 50 pct); body mass index (> 25 vs ≤ 25); hypertension (use of antihypertensive medication, systolic blood pressure ≥140 mm Hg and/or diastolic blood pressure ≥90 mm Hg vs. systolic blood pressure < 140 mm Hg and diastolic blood pressure < 90 mm Hg and no antihypertensive medication); diabetes (self‐reported diabetes, use of insulin or oral hypoglycemic agents, and/or nonfasting plasma glucose level > 11 mmol/L vs. no diabetes); smoking (current smoking vs. no current smoking); lipid‐lowering therapy (use of lipid‐lowering therapy vs. no lipid‐lowering therapy); alcohol consumption (> 14/21 vs. ≤14/21 U per week for women/men, with 1 U = 12 g alcohol, equivalent to 1 glass of wine or spirit or 1 beer [33 cl]); physical inactivity (≤4 vs. > 4 hours per week of light physical activity in leisure time); education (< 8 vs. ≥8 years), APOE ε4 carrier (yes vs. no). APOE, apolipoprotein E; CCHS, Copenhagen City Heart Study; CGPS, Copenhagen General Population Study; CI, confidence interval; HDL, high density lipoprotein; LDL, low density lipoprotein; pct, percentile; WAS, weighted allele score

FIGURE 4

Plasma levels of C‐reactive protein and hazard ratios of Alzheimer's disease as a function of weighted/simple allele score groups stratified by BMI at or below 25 kg/m² or BMI above 25 kg/m². Hazard ratios were multifactorially adjusted for age (as time scale), sex, hypertension, diabetes mellitus, smoking, alcohol consumption, physical inactivity, menopausal status and hormonal replacement therapy (women only), lipid‐lowering therapy, and education. The right column was additionally adjusted for APOE genotype. APOE, apolipoprotein E; BMI, body mass index (kg/m²); CI, confidence interval; CRP, C‐reactive protein