Abstract
Introduction:
Markers of hemostasis such as procoagulant factors and peak thrombin generation are associated with cardiovascular outcomes, but their associations with dementia risk are unclear. We aimed to evaluate prospective associations of selected procoagulant factors and peak thrombin generation with dementia risk.
Methods:
We measured levels of 7 hemostatic factors (fibrinogen, factor VII coagulant activity [FVIIc], activated factor VII [FVIIa], factor VIIa-antithrombin [FVIIa-AT], factor XI antigen [FXI], peak thrombin generation, and platelet count) among participants in the Cardiovascular Health Study, a cohort of older adults free of dementia in 1992/1993 (n=3,185). Dementia was adjudicated and classified by DSM-IV criteria through 1998/1999. Cox proportional hazards models estimated hazard ratios (HRs) for any dementia associated with 1-standard deviation (SD) differences, adjusting for sociodemographic and clinical factors and APOE genotype. Secondary analyses separately evaluated the risk of vascular dementia, Alzheimer’s disease, and mixed dementia.
Results:
At baseline, participants had a median age of 73 years. Over 5.4 years of follow-up, we identified 448 dementia cases. There was no evidence of linear associations between levels of these hemostatic factors with any dementia risk (HRs per 1-SD difference ranged from 1.0–1.1; 95% confidence intervals included 1.0). Results of secondary analyses by dementia subtype were similar.
Conclusions:
In this prospective study, there was no strong evidence of linear associations between levels of fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI, peak thrombin generation, or platelet count with dementia risk. Despite their associations with cardiovascular disease, higher levels of these biomarkers measured among older adults may not reflect dementia risk.
Keywords: Aged, Dementia, Fibrinolysis, Hemostasis, Cognition
Introduction
Circulating levels of hemostatic factors involved in coagulation and fibrinolysis are associated with the risk of several cardiovascular disease (CVD) outcomes, including stroke.1,2 Despite the known overlap between cardiovascular and dementia risk factors,3 it remains unclear whether hemostatic factors are also associated with risk of dementia.4 In case-control studies, persons with dementia have higher mean levels of factor VIII (FVIII), prothrombin fragment 1.2 (F1.2), plasminogen activator inhibitor (PAI-1), D-dimer, von Willebrand factor (vWF), and fibrinogen than persons without dementia.4 However, few prospective studies of hemostatic factor levels and dementia risk have been conducted, and their results are inconsistent.5–8 For example, in the Rotterdam and Caerphilly studies, plasma fibrinogen levels were positively associated with vascular dementia risk5,6 but not with non-vascular dementia risk,6 and in the Three-City cohort study, there was no association of fibrinogen levels with risk of either Alzheimer’s disease (AD) or vascular dementia (VaD).9 Many hemostatic factors known to be associated with CVD outcomes have never been evaluated in relation to dementia risk.
In response to inconsistency in reported associations between hemostatic factor levels and dementia risk, and to add to our understanding of other hemostatic factors in relation to dementia risk, we prospectively evaluated dementia risk in relation to existing measures of hemostatic factors selected a priori (fibrinogen, factor VII coagulant activity [FVIIc], activated factor VII [FVIIa], factor VIIa-antithrombin [FVIIa-AT], factor XI antigen [FXI], peak thrombin generation, and platelet count) among men and women enrolled in the Cardiovascular Health Study (CHS), a cohort study of adults aged 65 years and older. We hypothesized that higher levels of these procoagulant factors and of peak thrombin generation would be associated with greater risk of dementia.
Methods
The Cardiovascular Health Study
Detailed methods regarding CHS, an ongoing prospective cohort study, were previously published.10 In 1989–1990, CHS recruited 5,201 community-dwelling adults ≥65 years of age, randomly selected from Medicare eligibility lists at four field centers (Washington County, MD; Pittsburgh, PA; Forsyth County, NC; Sacramento County, CA). In 1992/1993, an additional 687 predominantly African American participants were recruited (total n=5,888). At annual visits conducted between 1989–1990 until 1998–1999, demographic, medical history, and cognitive function data were collected. Institutional review board approval was received at all participating sites, and participants provided informed consent.
Dementia Assessment and Eligible Participants
Dementia was identified from 1992/1993 to 1998/1999 as part of the CHS Cognition Study, which included 3,602 CHS participants from the cohorts recruited in 1989/1990 and 1992/1993 who had completed a cerebral MRI and the Modified Mini-Mental State Examination (3MSE) in 1992–1993 and who did not have dementia at that time.11–13 Participants completed annual neuropsychiatric tests at in-clinic visits through 1998/1999.11,12 Medical records were reviewed for participants who were unable to attend in-clinic visits and informants were interviewed for deceased participants.
A committee of neurologists and psychiatrists reviewed results, and dementia was classified using DSM-IV criteria, as described previously.12,13 MRI data were used to classify dementia subtypes into VaD, AD, mixed dementia, and Parkinson’s disease and other types of dementia.14 For our analysis, any dementia, defined as the presence of any of these dementia subtypes, was the primary outcome of interest. Secondary analyses separately evaluated the risk of VaD, AD, and mixed dementia; however, these analyses were secondary given that dementia in old age is most often of mixed vascular and AD pathology, making it difficult to distinguish between types of dementia.15
From 3,602 participants of the CHS Cognition Study, we further excluded persons without at least one measure of fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, or platelet count available in 1992/1993 (n=139), persons using warfarin in 1992/1993 (n=74), or with prevalent dementia (n=204), leaving 3,185 persons eligible for analyses.
Hemostatic Factor Level Measurement
In this study, we determined a priori to evaluate as exposures all hemostatic factors that had been measured in the CHS Cognition Study population in 1992/1993 or at CHS baseline in 1989/1990. Previously measured hemostatic factors were fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, and platelet count, which had been measured in fasting blood samples collected from the full cohort in 1992/1993.16–18 In addition, peak thrombin generation was measured at baseline, in 1989/1990 for the original CHS cohort, and in 1992/1993 for the predominantly African American cohort. Previous publications describe our hemostatic factor measurement methods in detail. Briefly, fibrinogen was measured with a BBL fibrometer (Becton Dickinson, Cockeysville, Maryland) by the Clauss method with Dade fibrinogen calibration reference (Baxter-Dade, Bedford, Massachusetts) and bovine thrombin (Parke-Davis, Lititz, Pennsylvania) (coefficient of variation [CV]: 3.1%).16,17 FVIIc was measured on the Coag-A-Mate X2 (Organon-Teknika, Durham, NC, USA) (CV: 5.9%),18 FVIIa was measured using the Staclot VIIa-rTF assay (Diagnostica Stago, Inc., Mount Olive, NJ, USA) (CV: 8.7%),18,19 FVIIa-AT was measured using the Asserachrom VIIa-AT ELISA (Diagnostica Stago, Inc.z) (CV: 8.0%),18 and FXI antigen was measured using sandwich ELISA (Affinity Biologicals Inc., Ancaster, Ontario, Canada) (CV: 9.6%).20 Peak thrombin generation was determined from the peak height on the thrombin generation curve; thrombin generation was measured in three batches with the Thrombin Generation Assay (TGA) (Technoclone, Vienna, Austria) (CV: 10.9–29.0%).19 Platelet counts were measured at field center laboratories by Coulter counters.17,21
Covariates
Demographic and health characteristics including sex, age, educational level, and race were self-reported by participants at CHS study entry in 1989/1990,10 with all other covariates collected at baseline for this analysis, 1992/1993. Participants self-reported cigarette smoking, alcohol consumption, and use of hormone therapy. Total kcal/week from leisure-time physical activity was calculated using self-reported type, frequency, and duration of 15 different physical activities during the prior two weeks.22 Prevalent coronary heart disease was confirmed by medical record review, fasting glucose ≥126 mg/dl or use of diabetes medications was used to define prevalent diabetes, and systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg was used to define hypertension. Fasting plasma samples were collected and used to measure total cholesterol.10 Mild cognitive impairment (MCI) status was defined by <88 on the 3MSE.23 Apolipoprotein E (APOE) e4 genotyping was performed using the methods of Hixson and Vernier.24
Statistical Analysis
Analyses were conducted in Stata 15 (Stata Corp, College Station, TX, USA). We used Cox proportional hazards models to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association of a 1-standard deviation (SD) difference in levels of fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, and platelet count with the risk of any dementia, with each hemostatic factor included in a separate model. Person-time began in 1992/1993, until first of dementia diagnosis, death, loss to follow-up, or end of follow-up in 1998/1999. Two multivariable models were adjusted for potential confounders, identified a priori; model 1 included age, sex, education, race, and field center, and model 2 included those five covariates and body mass index, physical activity, alcohol intake, current and former smoking, hormone therapy use, prevalent diabetes, prevalent coronary heart disease, systolic blood pressure, hypertensive medication use, total cholesterol level, and the presence of the APOE-e4 allele (including a category for missing APOE-e4 data). In a sensitivity analysis we further adjusted for prevalent stroke. To evaluate whether non-linear associations might exist, sensitivity analyses evaluated any dementia risk associated with quintiles of each hemostatic factor. Primary analyses evaluated the risk of any dementia and secondary analyses separately evaluated the risk of VaD, AD, and mixed dementia. Additional secondary analyses evaluated interaction by continuous age, MCI status in 1992/1993, and APOE-e4 carrier status. Evidence of significant interaction was determined using an alpha level of 0.05.
Results
Among eligible participants (n=3,185), the median age was 73 at baseline in 1992/1993. Participants were predominantly white and nearly one-quarter were APOE-e4 allele carriers (Table 1).
Table 1.
Baseline (1992/1993) characteristics of CHS Cognition Study participants for analyses of dementia risk, by sex (n=3,185)
| Characteristica | Men (n=1,312) | Women (n=1,873) |
|---|---|---|
|
| ||
| Age, median (IQR), years | 74 (7) | 73 (6) |
| Race, % | ||
| American Indian/Alaskan Native | 0.2 | 0.1 |
| Asian or Pacific Islander | 0.0 | 0.1 |
| Black | 13.5 | 15.8 |
| Other | 0.3 | 0.3 |
| White | 86.0 | 83.7 |
| Beyond high school education, % | 52.9 | 43.5 |
| Current smoking, % | 9.1 | 9.3 |
| BMI, median (IQR), kg/m2 | 26.2 (4.5) | 26.2 (6.1) |
| Physical activity, median (IQR), kcal/week | 1228.8 (1995.0) | 805.0 (1440.0) |
| Number of alcoholic drinks per week, median (IQR) | 0.3 (2.8) | 0 (0.5) |
| Current hormone therapy use, % | - | 16.1 |
| Diabetes mellitus, % | 17.6 | 11.7 |
| Hypertension, % | 36.8 | 41.1 |
| Total cholesterol, median (IQR), mg/dL | 194.0 (47.0) | 216.0 (45.0) |
| Antihypertensive medication use, % | 45.6 | 49.0 |
| Lipid-lowering medication use, % | 3.1 | 5.8 |
| History of coronary heart disease, % | 25.2 | 15.3 |
| History of stroke, % | 5.8 | 2.2 |
| APOE-e4 allele carrier, % | 23.3 | 25.0 |
| Mild cognitive impairmentb, % | 20.4 | 17.9 |
| Cognitive scores, median (IQR) | ||
| 3MSE | 93.0 (8.0) | 94.0 (8.0) |
| DSST | 39.0 (17.0) | 41.0 (17.0) |
| Hemostatic factors, median (IQR) | ||
| Fibrinogen, mg/dL | 313.0 (77.0) | 321.0 (77.0) |
| FVIIc, % | 100.0 (23.0) | 117.0 (29.0) |
| FVIIa, mU mL−1 | 46.0 (25.0) | 57.0 (30.0) |
| FVIIa-AT, pM | 121.1 (51.0) | 145.1 (56.8) |
| FXI antigen, % | 105.3 (34.9) | 119.2 (36.9) |
| Peak thrombin generation, nM | 470.7 (168.8) | 487.0 (167.9) |
| Platelet count, ×109 L−1 | 214.0 (72.0) | 248.0 (78.0) |
All percentages presented are among non-missing values. Characteristics were missing in <1% of participants except for diabetes mellitus (1.2% missing), APOE-e4 carrier status (7.8%), DSST (1.5%), fibrinogen (1.5%), FVIIc (1.7%), FVIIa (25.7%), FVIIa-AT (25.1%), FXI antigen (3.8%), peak thrombin generation (7.2%), platelet count (1.9%).
Baseline mild cognitive impairment defined as a score <88 on 3MSE
BMI = body mass index; DSST = digit symbol substitution test; FVIIc = factor VII coagulant activity; FVIIa = activated factor VII; FVIIa-AT = factor VIIa-antithrombin; FXI = factor XI antigen; kcal = kilocalories; 3MSE = modified mini-mental state; SD = standard deviation.
During an average of 5.4 years of follow-up (17,309 person-years), we documented 448 incident cases of dementia (AD n=233; VaD n=54; mixed dementia n=139; other n=22). We found no evidence of a linear association between a 1-SD difference in levels of fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, or platelet count with risk of any dementia in minimally or fully-adjusted models (Table 2). Associations remained null when we separately evaluated the risk of the AD, VaD, and mixed dementia subtypes in secondary analyses, other than some evidence of an association between a 1-SD difference in levels of FVIIa-AT and risk of AD (HRadj = 1.18; 95% CI: 1.01, 1.38). Results remained similar when we adjusted for stroke risk.
Table 2.
Dementia risk associated with a 1-SD difference in hemostatic factor levels in 1992/1993 among eligible CHS Cognition Study participants (followed from 1992/1993–1998/1999).
| Minimally-Adjusted | Fully-Adjusted | |||||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Model 1a | Model 2b | |||||||
|
| ||||||||
| SD | Person-Years | n Dementia | Rate/1000-Person-Years | HR | 95% CI | HR | 95% CI | |
|
| ||||||||
| Any Dementia | ||||||||
| Fibrinogen, mg/dL | 62.8 | 17,199 | 446 | 25.9 | 1.05 | (0.95, 1.15) | 1.05 | (0.95, 1.16) |
| FVIIc, % | 23.3 | 17,024 | 440 | 25.8 | 1.01 | (0.91, 1.12) | 1.03 | (0.92, 1.16) |
| FVIIa, mU mL−1 | 23.1 | 12,920 | 322 | 24.9 | 1.04 | (0.93, 1.17) | 1.03 | (0.91, 1.16) |
| FVIIa-AT, pM | 46.7 | 13,029 | 327 | 25.1 | 1.11 | (0.99, 1.24) | 1.10 | (0.98, 1.24) |
| FXI antigen, % | 30.3 | 16,699 | 428 | 25.6 | 1.03 | (0.94, 1.14) | 1.01 | (0.92, 1.12) |
| Peak thrombin generation, nM | 131.9 | 16,071 | 400 | 24.9 | 1.00 | (0.90, 1.11) | 1.05 | (0.94, 1.18) |
| Platelet count, ×109 L−1 | 63.4 | 16,998 | 439 | 25.8 | 1.01 | (0.92, 1.12) | 1.02 | (0.93, 1.13) |
| Alzheimer’s Disease | ||||||||
| Fibrinogen, mg/dL | 62.8 | 17,199 | 233 | 13.5 | 1.09 | (0.95, 1.24) | 1.14 | (0.99, 1.30) |
| FVIIc, % | 23.3 | 17,024 | 231 | 13.6 | 1.00 | (0.86, 1.15) | 1.03 | (0.88, 1.20) |
| FVIIa, mU mL−1 | 23.1 | 12,920 | 170 | 13.2 | 0.99 | (0.85, 1.16) | 0.98 | (0.83, 1.16) |
| FVIIa-AT, pM | 46.7 | 13,029 | 172 | 13.2 | 1.17 | (1.01, 1.36) | 1.18 | (1.01, 1.38) |
| FXI antigen, % | 30.3 | 16,699 | 223 | 13.4 | 1.07 | (0.93, 1.23) | 1.05 | (0.91, 1.21) |
| Peak thrombin generation, nM | 131.9 | 16,070 | 206 | 12.8 | 0.92 | (0.80, 1.07) | 0.97 | (0.83, 1.13) |
| Platelet count, ×109 L−1 | 63.4 | 16,998 | 229 | 13.5 | 1.09 | (0.95, 1.24) | 1.12 | (0.98, 1.27) |
| Vascular Dementia | ||||||||
| Fibrinogen, mg/dL | 62.8 | 17,199 | 52 | 3.0 | 1.04 | (0.79, 1.37) | 0.92 | (0.69, 1.22) |
| FVIIc, % | 23.3 | 17,024 | 53 | 3.1 | 0.96 | (0.70, 1.30) | 0.82 | (0.58, 1.17) |
| FVIIa, mU mL−1 | 23.1 | 12,920 | 35 | 2.7 | 1.11 | (0.79, 1.57) | 1.02 | (0.71, 1.48) |
| FVIIa-AT, pM | 46.7 | 13,029 | 36 | 2.8 | 1.02 | (0.72, 1.44) | 0.86 | (0.58, 1.26) |
| FXI antigen, % | 30.3 | 16,699 | 48 | 2.9 | 0.95 | (0.70, 1.28) | 0.88 | (0.64, 1.21) |
| Peak thrombin generation, nM | 131.9 | 16,071 | 48 | 3.0 | 1.04 | (0.76, 1.41) | 1.13 | (0.83, 1.55) |
| Platelet count, ×109 L−1 | 63.4 | 16,998 | 51 | 3.0 | 1.27 | (0.97, 1.66) | 1.19 | (0.91, 1.57) |
| Mixed Dementia | ||||||||
| Fibrinogen, mg/dL | 62.8 | 17,199 | 139 | 8.1 | 1.01 | (0.85, 1.20) | 0.99 | (0.83, 1.19) |
| FVIIc, % | 23.3 | 17,024 | 135 | 7.9 | 1.00 | (0.83, 1.20) | 1.07 | (0.87, 1.31) |
| FVIIa, mU mL−1 | 23.1 | 12,920 | 106 | 8.2 | 1.07 | (0.88, 1.31) | 1.09 | (0.88, 1.34) |
| FVIIa-AT, pM | 46.7 | 13,029 | 107 | 8.2 | 1.06 | (0.87, 1.30) | 1.06 | (0.86, 1.32) |
| FXI antigen, % | 30.3 | 16,699 | 136 | 8.1 | 1.02 | (0.85, 1.21) | 1.01 | (0.84, 1.21) |
| Peak thrombin generation, nM | 131.9 | 16,071 | 128 | 8.0 | 1.08 | (0.89, 1.31) | 1.12 | (0.92, 1.36) |
| Platelet count, ×109 L−1 | 63.4 | 16,998 | 137 | 8.1 | 0.84 | (0.70, 1.01) | 0.85 | (0.71, 1.03) |
CI = confidence interval; FVIIc = factor VII coagulant activity; FVIIa = activated factor VII; FVIIa-AT = factor VIIa-antithrombin; FXI = factor XI antigen; HR = hazard ratio; SD = standard deviation
Model 1 adjusted for age, sex, education, race, and field center
Model 2 adjusted for model 1 covariates + BMI, physical activity, alcohol intake, current and former smoking, hormone therapy use (in women only), prevalent diabetes, prevalent coronary heart disease, systolic blood pressure, hypertensive medication use, total cholesterol, APOE genotype.
In sensitivity analyses, we observed limited evidence of a U-shaped association of (separately) FVIIa and FXI antigen quintiles in relation to dementia risk, and an inverse U-shaped association of platelet count with any dementia risk. (Table 3) Quintile analyses suggested that FVIIa and FXI antigen levels in the second quintile (and third quintile, for FVIIa) were associated with a lower dementia risk than levels of FVIIa and FXI antigen in the lowest quintile. Analyses of platelet count quintiles suggested that platelet counts in the fourth quintile were associated with a greater dementia risk than platelet counts in the lowest quintile. There was no evidence of non-linear associations between fibrinogen, FVIIc, FVIIa-AT, or peak thrombin generation with dementia risk (Table 3).In sensitivity analyses evaluating quintiles of hemostatic factors in relation to the separate secondary outcomes of AD, VaD, and mixed dementia (Supplemental Tables 1, 2, and 3, respectively), we observed approximately U-shaped associations between FVIIa quintiles and AD risk (Supplemental Table 1), and of FXI antigen quintiles and VaD and mixed dementia risk (Supplemental Tables 2 and 3), and an inverse U-shaped association between platelet count and AD risk (Supplemental Table 1). In secondary analyses, we observed a significant interaction between platelet count and MCI status at 1992/1993 baseline (p-interaction = 0.03). Among persons with baseline MCI, every 1-SD higher platelet count was associated with an 18% greater dementia risk (HR=1.18; 95% CI: 1.00, 1.39). Among persons without baseline MCI, there was no evidence of an association of platelet count with dementia risk (HR=0.97; 95% CI: 0.85, 1.10). Other interaction terms of hemostatic factors with continuous age, baseline MCI, or APOE-e4 carrier status were not significant (p>0.05).
Table 3.
Risk of any dementia associated with quintiles of hemostatic factor levels in 1992/1993 among eligible CHS Cognition Study participants (followed from 1992/1993–1998/1999).
| Minimally-Adjusted | Fully-Adjusted | ||||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| Model 1a | Model 2b | ||||||||
|
| |||||||||
| Range | n | Person-Years | n Dementia Classifications | Rate/1000-Person-Years | HR | 95% CI | HR | 95% CI | |
|
| |||||||||
| Fibrinogen, md/dL | |||||||||
| Q1 | ≤274 | 649 | 3608 | 85 | 23.6 | 1.00 | reference | 1.00 | reference |
| Q2 | 275 to ≤304 | 643 | 3565 | 87 | 24.4 | 0.99 | (0.73, 1.34) | 0.92 | (0.68, 1.26) |
| Q3 | 305 to ≤331 | 622 | 3402 | 85 | 25.0 | 1.00 | (0.74, 1.35) | 0.99 | (0.73, 1.34) |
| Q4 | 332 to ≤372 | 626 | 3432 | 96 | 28.0 | 1.13 | (0.84, 1.52) | 1.11 | (0.82, 1.49) |
| Q5 | ≥373 | 623 | 3192 | 93 | 29.1 | 1.11 | (0.82, 1.50) | 1.08 | (0.79, 1.46) |
| FVIIc, % | |||||||||
| Q1 | ≤93 | 653 | 3404 | 97 | 28.5 | 1.00 | reference | 1.00 | reference |
| Q2 | 94 to ≤104 | 617 | 3370 | 91 | 27.0 | 1.00 | (0.74, 1.33) | 1.09 | (0.81, 1.47) |
| Q3 | 105 to ≤115 | 618 | 3387 | 82 | 24.2 | 0.86 | (0.64, 1.16) | 0.96 | (0.70, 1.31) |
| Q4 | 116 to ≤130 | 636 | 3501 | 84 | 24.0 | 0.94 | (0.69, 1.28) | 1.01 | (0.74, 1.39) |
| Q5 | ≥131 | 607 | 3363 | 86 | 25.6 | 1.05 | (0.77, 1.43) | 1.17 | (0.84, 1.64) |
| FVIIa, mU mL−1 | |||||||||
| Q1 | ≤36 | 499 | 2671 | 81 | 30 .3 | 1.00 | reference | 1.00 | reference |
| Q2 | 37 to ≤47 | 466 | 2558 | 49 | 19.2 | 0.63 | (0.44, 0.90) | 0.61 | (0.42, 0.88) |
| Q3 | 48 to ≤58 | 495 | 2738 | 56 | 20.4 | 0.73 | (0.52, 1.03) | 0.67 | (0.47, 0.96) |
| Q4 | 59 to ≤72 | 455 | 2473 | 68 | 27.5 | 0.97 | (0.69, 1.35) | 1.05 | (0.75, 1.47) |
| Q5 | ≥73 | 452 | 2478 | 68 | 27.4 | 0.96 | (0.69, 1.34) | 0.93 | (0.66, 1.32) |
| FVIIa-AT, pM | |||||||||
| Q1 | ≤103.1 | 478 | 2605 | 66 | 25 .3 | 1.00 | reference | 1.00 | reference |
| Q2 | 103.2 to ≤−124.2 | 477 | 2653 | 61 | 23.1 | 0.92 | (0.65, 1.31) | 0.97 | (0.68, 1.39) |
| Q3 | 124.3 to 145.6 | 478 | 2631 | 58 | 22.0 | 0.91 | (0.64, 1.31) | 0.93 | (0.65, 1.35) |
| Q4 | 145.7 ≤ 175.5 | 477 | 2658 | 59 | 22.2 | 0.97 | (0.67, 1.40) | 0.97 | (0.66, 1.41) |
| Q5 | ≥175.6 | 477 | 2492 | 83 | 33.3 | 1.37 | (0.98, 1.93) | 1.38 | (0.97,1.97) |
| FXI antigen, % | |||||||||
| Q1 | ≤91.7 | 613 | 3238 | 105 | 32.4 | 1.00 | reference | 1.00 | reference |
| Q2 | 91.7 to ≤106.7 | 613 | 3408 | 76 | 22.3 | 0.69 | (0.51, 0.92) | 0.69 | (0.51, 0.93) |
| Q3 | 106.8 to ≤120.3 | 613 | 3341 | 82 | 24.5 | 0.85 | (0.63, 1.13) | 0.87 | (0.64, 1.17) |
| Q4 | 120.4 to ≤ 138.8 | 613 | 3357 | 77 | 22.9 | 0.83 | (0.61, 1.12) | 0.86 | (0.63, 1.17) |
| Q5 | ≥138.9 | 612 | 3355 | 88 | 26.2 | 0.99 | (0.74, 1.33) | 0.92 | (0.67, 1.25) |
| Peak thrombin generation, (nM) | |||||||||
| Q1 | ≤387 | 592 | 3177 | 84 | 26.4 | 1.00 | reference | 1.00 | reference |
| Q2 | 388 to ≤440 | 593 | 3242 | 75 | 23.1 | 0.90 | (0.66, 1.24) | 0.98 | (0.70, 1.36) |
| Q3 | 451 to ≤513 | 589 | 3230 | 80 | 24.8 | 0.99 | (0.73, 1.35) | 1.20 | (0.87, 1.66) |
| Q4 | 514 to ≤598 | 591 | 3223 | 86 | 26.7 | 1.03 | (0.76, 1.41) | 1.17 | (0.85, 1.61) |
| Q5 | ≥599 | 591 | 3198 | 75 | 23.5 | 0.93 | (0.67, 1.29) | 1.14 | (0.81, 1.61) |
| Platelet count, ×109 L−1 | |||||||||
| Q1 | ≤187 | 629 | 3290 | 88 | 26.7 | 1.00 | reference | 1.00 | reference |
| Q2 | 188 to ≤ 220 | 621 | 3440 | 78 | 22.7 | 0.92 | (0.68, 1.26) | 0.94 | (0.69, 1.29) |
| Q3 | 221 to ≤ 250 | 643 | 3551 | 100 | 28.2 | 1.20 | (0.90, 1.61) | 1.27 | (0.94, 1.70) |
| Q4 | 251 to ≤ 289 | 609 | 3342 | 95 | 28.4 | 1.22 | (0.91, 1.64) | 1.36 | (1.01, 1.85) |
| Q5 | ≥289 | 623 | 3375 | 78 | 23.1 | 1.00 | (0.73, 1.36) | 1.02 | (0.73, 1.41) |
CI = confidence interval; FVIIc = factor VII coagulant activity; FVIIa = activated factor VII; FVIIa-AT = factor VIIa-antithrombin; FXI = factor XI antigen; HR = hazard ratio; SD = standard deviation
Model 1 adjusted for age, sex, education, race, and field center
Model 2 adjusted for model 1 covariates + BMI, physical activity, alcohol intake, current and former smoking, hormone therapy use (in women only), prevalent diabetes, prevalent coronary heart disease, systolic blood pressure, hypertensive medication use, total cholesterol, APOE genotype.
Discussion
In this prospective cohort study of older adults, we found no evidence of a linear association between levels of fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, or platelet count and risk of any dementia, nor did we find strong evidence of associations with dementia subtypes (AD, VaD, or mixed dementia) in secondary analyses. Sensitivity analyses suggested a U-shaped association between quintiles of FVIIa and FXI antigen and an inverse U-shaped association of platelet count with any dementia risk. Furthermore, in secondary analyses there was some suggestion of a linear association between higher FVIIa-AT levels and a greater risk of AD, but given that we evaluated a large number of associations in this study and that some evidence of an association was present for AD only and not for other dementia subtypes, it is possible that this association was by chance. In sensitivity analyses of secondary outcomes, we observed trends toward U-shaped associations between FVIIa quintiles and AD risk, between FXI antigen quintiles and VaD and mixed dementia risk, and between platelet count and AD risk; however, these results are necessarily preliminary as they were results of sensitivity analyses of outcomes of secondary interest.
In prior studies, fibrinogen was inconsistently associated with dementia risk, with the Rotterdam Study of men and women ≥55 years of age reporting a positive association with any dementia and VaD risk,5 the Caerphilly Study of men with blood drawn between 45 to 59 years of age reporting a positive association with VaD but not with non-VaD,6 and with two other studies reporting no association with any dementia risk,25 AD,9 nor VaD.9 We are aware of only one study that evaluated FVIIc levels in relation to dementia risk; in the Caerphilly Study, there was no association of FVIIc levels with VaD or non-VaD.6 We are unaware of other studies that have evaluated FVIIa, FVIIa-AT, FXI antigen levels, peak thrombin generation, or platelet count in relation to dementia risk. While we found no evidence of continuous linear associations between levels of FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, or platelet count, sensitivity analyses suggested a U-shaped association between quintiles of FVIIa and FXI antigen and an inverse U-shaped association of platelet count with any dementia risk. A priori, we had hypothesized linear associations between these hemostatic factor levels and dementia risk, with higher levels being associated with greater dementia risk. This preliminary suggestion of potentially non-linear associations between quintiles of FVIIa, FXI antigen, and platelet count warrants confirmation in other cohorts with data available to evaluate this association.
Study population and design differences may contribute to the differences in results to date. One such difference may be the age of study participants; CHS participants had a median age of 73 years (range 65–98) at baseline in this analysis, which makes them an older population than the Rotterdam or Caerphilly studies. It is plausible that levels of midlife factors, such as fibrinogen,5,6 may be associated with dementia risk while levels at an older age may not be associated. Although we were unable to evaluate associations between hemostatic factors measured at younger ages in relation to dementia risk in this study (given the age of CHS participants), we found no evidence of an interaction between age and hemostatic factor levels within the age range of CHS participants. Our results suggest that knowledge of these factors in late life is unlikely to be useful in identifying subjects at elevated risk for dementia.
Our analyses of dementia risk leveraged existing measures of hemostatic factors that had been previously collected in the CHS Cognition Study population: fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, and platelet count. However, our results do not exclude the possibility that other hemostatic factors may be associated with risk of dementia. Other observational studies regarding dementia risk reported positive associations with levels of FVIII6 and PAI-1,6 inconsistent associations with tissue plasminogen activator activity,6,25 thrombin-antithrombin complex,6,25 D-dimer,6,9,25 and F1.2,6,25 and no evidence of associations with vWF,6,9 factor XIIa,6 activated partial thromboplastin time,6 activated protein C ratio,6 fibrin clotting time,6 or plasmin-α2 -antiplasmin complex.25
In a previous smaller CHS-based study of hemostatic factors in relation to cognitive decline (n=400 eligible CHS participants), we evaluated associations of 20 hemostatic factors measured in 1989/1990 in relation to cognitive decline.26 An assessment of all 20 of these hemostatic factors in relation to dementia risk was not possible in our current study, because these factors were not measured in the larger CHS Cognition Study population (n=3,602), in which our dementia outcome was identified. In our prior study of cognitive decline, among these 20 hemostatic factors evaluated as individual exposures, only higher levels of plasmin-α2-antiplasmin complex (PAP), tissue factor pathway inhibitor (TFPI), and lower factor X (FXc) levels were associated with faster cognitive decline, as estimated by annual change in 3MSE points.26 Furthermore, an analysis using derived principal components suggested that one of four principal components, which loaded positively on D-dimer, prothrombin fragment 1.2 and PAP was significantly associated with change in 3MSE.26 Of the 7 hemostatic factors evaluated here in relation to dementia risk, 3 were also included in our prior study of cognitive decline (fibrinogen, FVIIc, and platelet count); similar to our current results which suggest no evidence of an association between these 3 factors with dementia risk, there was also no evidence that these 3 factors were associated with changes in cognition in CHS. Together, these two studies suggest that although levels of fibrinogen, FVIIc, and platelet count are associated with cardiovascular event risk, they are unlikely to be associated with changes in cognition nor dementia risk among older adults.
Levels of some, but not all, hemostatic factors included in our study have previously been associated with cardiovascular outcomes in CHS1,2,18–20, supporting a priori study hypotheses that higher levels of these procoagulant factors and of peak thrombin generation could also be associated with greater dementia risk in this population of older adults. In previous CHS studies, fibrinogen levels were associated with incident CVD risk in men (HR per 1-SD: 1.12; 95% CI: 1.04, 1.22) but not women.1 Levels of FVIIa were associated with stroke risk (HR per 1-SD: 1.12; 95 CI: 1.01–1.23)18, and peak thrombin generation was associated with ischemic stroke risk (HR per 1 SD: 1.09; 95% CI: 1.01–1.17).19 Furthermore, levels of FVIIa-AT were associated with all-cause mortality (HR per 1-SD: 1.09; 95% CI:1.03, 1.12).18 Prior CHS publications did not find evidence of associations of FVIIc levels or platelet count with CVD risk1, nor between FXI antigen levels and incident coronary heart disease nor stroke risk20. Many risk factors for cardiovascular disease and dementia overlap, providing biologic rationale for a hypothesis that hemostatic factor levels may also be associated with dementia risk. However, risk factors for these outcomes are not identical; it is plausible that levels of these factors among older adults may be associated with cardiovascular outcomes but not dementia risk.
Hemostatic factor measures included in this study are from one time only, and as a limitation, these measures do not reflect within-person variability that may have led to a null bias. In addition, measurement of the thrombin generation assay included in our study had a wide range of variability in the measurement of peak thrombin generation (10.9–29.0%)19. Such variation could be due to measurement methods, or to variability in participant characteristics. As such, the lack of evidence of an association between peak thrombin generation and any dementia risk in our study should not be interpreted as an exclusion of the possibility of a true association. Although adjudication of dementia incorporated regular and systematic cognitive screening, confirmatory neuropsychiatric testing and physician committee review,12 the potential remains for misclassification that may have led to null findings; this is even more so for analyses of specific subtypes. In addition, some participants may have died before living long enough to develop dementia. However, the Cox proportional hazards model is the best suited for addressing etiological questions, like those posed here, and provides results of greatest clinical relevance. Results from our study cannot be extrapolated to younger individuals or to Latino or Asian populations, who were not meaningfully represented in the CHS sample. Strengths of this study are the population-based and prospective cohort study design of the CHS, the study’s large sample size, and adjudicated dementia outcome. An additional strength was our ability to evaluate hemostatic factors that are less often studied (for example: FVIIa, FVIIa-AT, peak thrombin generation) in relation to dementia risk.
Conclusion
In conclusion, we found no evidence of linear associations between levels of fibrinogen, FVIIc, FVIIa, FVIIa-AT, FXI antigen, peak thrombin generation, or platelet count measured in older adulthood in relation to risk of dementia. Although levels of these hemostatic factors are associated with both arterial and venous thrombotic event risk, physiology unique to the brain may not be reflected in higher levels of these biomarkers measured among older adults.
Supplementary Material
Highlights.
The association of hemostatic factor levels with dementia risk is unclear.
We evaluated hemostatic factors from 1992/1993 and dementia risk over 5.4 years.
No evidence of linear associations between hemostatic factors and dementia risk.
Higher levels of factors measured at median age 73 may not reflect dementia risk.
Acknowledgements
This research was supported by contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, and grants U01HL080295 and U01HL130114 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided by R01AG023629, R01AG15928, and R01AG20098 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at CHS-NHLBI.org. This study was further supported by NINDS grant R01NS089638 and NHLBI grants T32-HL098048 and K01HL139997. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Conflict of Interest Disclosure
L.B. Harrington, A.N. Ehlert, E.L. Thacker, N.S. Jenny, O. Lopez, M. Cushman, N.C. Olson, A. Fitzpatrick, K.J. Mukamal, and M.K. Jensen have no conflicts of interest to declare.
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Addendum
L.B. Harrington, A.N. Ehlert, K.J. Mukamal, and M.K. Jensen contributed to the analysis concept and design. A.N. Ehlert performed statistical analyses and all authors (L.B. Harrington, A.N. Ehlert, E.L. Thacker, N.S. Jenny, O. Lopez, M. Cushman, N.C. Olson, A. Fitzpatrick, K.J. Mukamal, and M.K. Jensen) contributed to the interpretation of data, provided substantial scientific contributions to the revision of the manuscript, and approved the final version of the manuscript for submission to Thrombosis Research.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Zakai NA, Katz R, Jenny NS, et al. Inflammation and hemostasis biomarkers and cardiovascular risk in the elderly: the Cardiovascular Health Study. J Thromb Haemost. Jun 2007;5(6):1128–35. doi: 10.1111/j.1538-7836.2007.02528.x [DOI] [PubMed] [Google Scholar]
- 2.Cushman M, Yanez D, Psaty BM, et al. Association of fibrinogen and coagulation factors VII and VIII with cardiovascular risk factors in the elderly: the Cardiovascular Health Study. Cardiovascular Health Study Investigators. Am J Epidemiol. Apr 01 1996;143(7):665–76. [DOI] [PubMed] [Google Scholar]
- 3.Baumgart M, Snyder HM, Carrillo MC, Fazio S, Kim H, Johns H. Summary of the evidence on modifiable risk factors for cognitive decline and dementia: A population-based perspective. Alzheimers Dement. Jun 2015;11(6):718–26. doi: 10.1016/j.jalz.2015.05.016 [DOI] [PubMed] [Google Scholar]
- 4.Quinn TJ, Gallacher J, Deary IJ, Lowe GD, Fenton C, Stott DJ. Association between circulating hemostatic measures and dementia or cognitive impairment: systematic review and meta-analyzes. J Thromb Haemost. Aug 2011;9(8):1475–82. doi: 10.1111/j.1538-7836.2011.04403.x [DOI] [PubMed] [Google Scholar]
- 5.van Oijen M, Witteman JC, Hofman A, Koudstaal PJ, Breteler MM. Fibrinogen is associated with an increased risk of Alzheimer disease and vascular dementia. Stroke. Dec 2005;36(12):2637–41. doi: 10.1161/01.str.0000189721.31432.26 [DOI] [PubMed] [Google Scholar]
- 6.Gallacher J, Bayer A, Lowe G, et al. Is sticky blood bad for the brain?: Hemostatic and inflammatory systems and dementia in the Caerphilly Prospective Study. Arterioscler Thromb Vasc Biol. Mar 2010;30(3):599–604. doi: 10.1161/atvbaha.109.197368 [DOI] [PubMed] [Google Scholar]
- 7.Barber M, Tait RC, Scott J, Rumley A, Lowe GD, Stott DJ. Dementia in subjects with atrial fibrillation: hemostatic function and the role of anticoagulation. J Thromb Haemost. Nov 2004;2(11):1873–8. doi: 10.1111/j.1538-7836.2004.00993.x [DOI] [PubMed] [Google Scholar]
- 8.Stott DJ, Robertson M, Rumley A, et al. Activation of hemostasis and decline in cognitive function in older people. Arterioscler Thromb Vasc Biol. Mar 2010;30(3):605–11. doi: 10.1161/atvbaha.109.199448 [DOI] [PubMed] [Google Scholar]
- 9.Carcaillon L, Gaussem P, Ducimetiere P, et al. Elevated plasma fibrin D-dimer as a risk factor for vascular dementia: the Three-City cohort study. J Thromb Haemost. Dec 2009;7(12):1972–8. doi: 10.1111/j.1538-7836.2009.03603.x [DOI] [PubMed] [Google Scholar]
- 10.Fried LP, Borhani NO, Enright P, et al. The Cardiovascular Health Study: design and rationale. Ann Epidemiol. Feb 1991;1(3):263–76. [DOI] [PubMed] [Google Scholar]
- 11.Fitzpatrick AL, Kuller LH, Ives DG, et al. Incidence and prevalence of dementia in the Cardiovascular Health Study. J Am Geriatr Soc. Feb 2004;52(2):195–204. [DOI] [PubMed] [Google Scholar]
- 12.Lopez OL, Kuller LH, Fitzpatrick A, Ives D, Becker JT, Beauchamp N. Evaluation of dementia in the cardiovascular health cognition study. Neuroepidemiology. Jan-Feb 2003;22(1):1–12. doi:67110 [DOI] [PubMed] [Google Scholar]
- 13.Fitzpatrick AL, Kuller LH, Lopez OL, et al. Midlife and late-life obesity and the risk of dementia: cardiovascular health study. Arch Neurol. Mar 2009;66(3):336–42. doi: 10.1001/archneurol.2008.582 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fitzpatrick AL, Irizarry MC, Cushman M, Jenny NS, Chi GC, Koro C. Lipoprotein-associated phospholipase A2 and risk of dementia in the Cardiovascular Health Study. Atherosclerosis. Aug 2014;235(2):384–91. doi: 10.1016/j.atherosclerosis.2014.04.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Snyder HM, Corriveau RA, Craft S, et al. Vascular contributions to cognitive impairment and dementia including Alzheimer’s disease. Alzheimers Dement. Jun 2015;11(6):710–7. doi: 10.1016/j.jalz.2014.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Jiang R, Burke GL, Enright PL, et al. Inflammatory markers and longitudinal lung function decline in the elderly. Am J Epidemiol. Sep 15 2008;168(6):602–10. doi: 10.1093/aje/kwn174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Cushman M, Cornell ES, Howard PR, Bovill EG, Tracy RP. Laboratory methods and quality assurance in the Cardiovascular Health Study. Clin Chem. Feb 1995;41(2):264–70. [PubMed] [Google Scholar]
- 18.Olson NC, Raffield LM, Lange LA, et al. Associations of activated coagulation factor VII and factor VIIa-antithrombin levels with genome-wide polymorphisms and cardiovascular disease risk. J Thromb Haemost. Jan 2018;16(1):19–30. doi: 10.1111/jth.13899 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Olson NC, Butenas S, Lange LA, et al. Coagulation factor XII genetic variation, ex vivo thrombin generation, and stroke risk in the elderly: results from the Cardiovascular Health Study. J Thromb Haemost. Oct 2015;13(10):1867–77. doi: 10.1111/jth.13111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Appiah D, Fashanu OE, Heckbert SR, Cushman M, Psaty BM, Folsom AR. Relation of coagulation factor XI with incident coronary heart disease and stroke: the Cardiovascular Health Study. Blood Coagul Fibrinolysis. Dec 22 2016;doi: 10.1097/mbc.0000000000000616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.van der Bom JG, Heckbert SR, Lumley T, et al. Platelet count and the risk for thrombosis and death in the elderly. J Thromb Haemost. Mar 2009;7(3):399–405. doi: 10.1111/j.1538-7836.2008.03267.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Taylor HL, Jacobs DR, Jr., Schucker B, Knudsen J, Leon AS, Debacker G. A questionnaire for the assessment of leisure time physical activities. Journal of chronic diseases. 1978;31(12):741–55. [DOI] [PubMed] [Google Scholar]
- 23.Sink KM, Craft S, Smith SC, et al. Montreal Cognitive Assessment and Modified Mini Mental State Examination in African Americans. Journal of aging research. 2015;2015:872018. doi: 10.1155/2015/872018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. Journal of lipid research. Mar 1990;31(3):545–8. [PubMed] [Google Scholar]
- 25.Bots ML, Breteler MM, van Kooten F, et al. Coagulation and fibrinolysis markers and risk of dementia. The Dutch Vascular Factors in Dementia Study. Haemostasis. May-Aug 1998;28(3–4):216–22. doi:22433 [DOI] [PubMed] [Google Scholar]
- 26.Harrington LB, Ehlert AN, Thacker EL, et al. Hemostatic Factor Levels and Cognitive Decline in Older Adults: The Cardiovascular Health Study. J Thromb Haemost. May 2021;19(5):1219–1227. doi: 10.1111/jth.15300 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.