Abstract
Background:
Age-related declines in cognitive function may begin in midlife.
Purpose:
To determine whether blood-based biomarkers of inflammation, metabolic dysregulation and neurotoxins are associated with risk of cognitive decline and impairment.
Methods:
Baseline blood samples from the longitudinal Beaver Dam Offspring Study (2005–2008) were assayed for markers of inflammation, metabolic dysregulation, and environmental neurotoxins. Cognitive function was measured at baseline, 5-(2010–2013) and 10-year (2015–2017) examinations. Participants without cognitive impairment at baseline and with cognitive data from at least one follow-up were included. Cox proportional hazards models were used to evaluate associations between baseline blood biomarkers and the 10-year cumulative incidence of cognitive impairment. Poisson models were used to estimate the relative risk (RR) of 5-year decline in cognitive function by baseline blood biomarkers. Models were adjusted for age, sex, education, and cardiovascular related risk factors.
Results:
Participants (N=2421) were a mean age of 49 years and 55% were women. Soluble vascular cell adhesion molecule-1(sVCAM-1Tertile(T)3 vs T1–2 Hazard Ratio (HR)=1.72, 95% Confidence Interval (CI)=1.05,2.82) and hemoglobin A1C (HR=1.75, 95% CI=1.18,2.59, per 1% in women) were associated with the 10-year cumulative incidence of cognitive impairment. sVCAM-1 (RRT3 vs T1–2=1.45, 95% CI=1.06,1.99) and white blood cell count (RR=1.10, 95% CI=1.02,1.19, per 103/μL) were associated with 5-year cognitive decline.
Conclusions:
Biomarkers related to inflammation and metabolic dysregulation were associated with an increased risk of developing cognitive decline and impairment. These results extend previous research in cognitive aging to early markers of cognitive decline in midlife, a time when intervention methods may be more efficacious.
Keywords: Vascular Cell Adhesion Molecule, Hemoglobin A1C, White Blood Cell Count, Epidemiology, Longitudinal
INTRODUCTION
Age-related declines in cognitive function may begin as early as the fourth decade and pathophysiological processes that can lead to cognitive impairment later in life are thought to begin decades before the onset of clinical dementia or Alzheimer’s disease [1]. Health status and behavioral and environmental factors in midlife may therefore have a substantial impact on cognitive health later in life. Identification of factors associated with cognitive changes in midlife or early old age may provide opportunities to encourage behavioral and lifestyle changes or initiate treatments, when available, to slow or delay cognitive decline.
Vascular health, metabolic dysregulation, inflammation, and exposure to environmental neurotoxins all have the potential to have adverse effects on the brain and cognitive function. Inflammation, as indicated by higher levels of circulating inflammatory markers such as C-reactive protein (CRP) and interleukin-6 (IL-6), has been associated with brain atrophy [2] and declines in cerebral blood flow [3]. Nevertheless, while there is evidence that inflammation in the brain is associated with neurodegeneration and Alzheimer’s disease and dementia, the relationship between peripheral inflammatory markers and the development of cognitive decline or impairment is less clear. Previous studies of circulating inflammatory markers have found associations between CRP and IL-6 and an increased risk of developing cognitive decline or impairment but, results have not been consistent across studies and were primarily conducted in older adults [4–6].
Another peripheral marker of inflammation which has been studied as a risk factor for cognitive decline is white blood cell count (WBC). One study using a composite inflammatory score that included WBC, found a higher score at midlife was associated with greater cognitive decline [7], but the association of WBC with cognitive decline or impairment is not yet understood. More recently, the ratio of specific types of white blood cells, the neutrophil to lymphocyte ratio (NLR), has been reported to be elevated in older adults with cognitive impairment or AD [8,9]. Although a couple of studies have found NLR predictive of increased risk for dementia [10,11], previous studies on NLR and cognitive outcomes have been either cross-sectional and/or in older adults and little is known about the association of NLR and cognitive decline in midlife in the general population. Likewise, few studies have evaluated soluble intercellular adhesion molecule-1 (sICAM-1) and soluble vascular cell adhesion molecule −1 (sVCAM-1), proteins associated with vascular health and inflammation [12], as risk factors for cognitive decline [13, 14]. In the Beaver Dam Offspring Study (BOSS), higher sICAM-1 was associated with the 5-year incidence of brain aging using a composite score of sensory and cognitive function but the long-term association with cognitive impairment was not assessed [15].
Metabolic dysregulation in the form of diabetes, has also been associated with brain atrophy [16]. Hemoglobin A1C (HbA1C) is a peripheral blood marker with higher levels indicative of glucose dysregulation and diabetes. Whereas having diabetes in midlife [16] has been associated with a higher risk for cognitive impairment later in life, it is unclear if midlife HbA1C levels alone are associated with increased risk for cognitive impairment and dementia later in life [17,18].
Cadmium and lead are known neurotoxins that may cause neurodegeneration through increased oxidative stress and neuronal apoptosis [19]. Although cadmium and lead are known to have detrimental effects on the central nervous system at high levels, less is known about their effects on cognitive function at the lower concentrations that are found in the general population. There is some evidence from cross-sectional studies [20,21] that cadmium and lead are associated with worse cognitive function, but prospective studies evaluating environmental neurotoxins with cognitive outcomes in the general population are lacking.
Much of the previous research on peripheral markers of inflammation, metabolic dysregulation, and neurotoxin exposure and cognitive outcomes has been conducted in older adults and longitudinal studies in midlife in the general population are limited. The aim of this study was to determine if midlife blood-based biomarkers of inflammation, metabolic function, and environmental neurotoxin exposure are associated with the 10-year cumulative incidence of cognitive impairment and 5-year incidence of cognitive decline in a general population cohort.
2. METHODS
Participants
The Beaver Dam Offspring Study (BOSS) is a longitudinal cohort study of sensory and cognitive aging in the adult children of participants in the population-based Epidemiology of Hearing Loss Study (EHLS;1993–2020) [6,22,23]. BOSS baseline data collection occurred in 2005–2008 (N=3298) with follow-up phases at 5 (2010–2013) and 10 years (2015–2017) [6,22,23]. At baseline, the mean age of BOSS participants was 49 years (range 21–84 years; standard deviation=9.9 years), 55% were women and about 70% had more than 12 years of education [23]. Participants with baseline blood biomarker and cognitive data and, with cognitive test data from at least one follow-up examination, were eligible for the analyses of 10-year cumulative incidence of cognitive impairment [23]. Participants with cognitive impairment at baseline were excluded. Individuals who participated at baseline and had complete cognitive function test data from both the 5- and 10-year phases were included in the analyses of 5-year decline on a cognitive summary score. Approval for this research was obtained from the University of Wisconsin Health Sciences Institutional Review Board and written informed consent was obtained from all participants at each phase prior to examination.
Cognitive Measures
The Mini-Mental State Examination (MMSE) and Trail Making Tests A (TMTA) and B (TMTB), were administered at all three phases [23]. The MMSE was administered to all participants 50 years and older; participants younger than 50 years of age were assigned a score of 29 based on normative data [23,24]. The Digit Symbol Substitution Test (DSST), Verbal Fluency Test (VFT) and modified Rey Auditory Verbal Learning Test (AVLT) were obtained at the 5 and 10- year examinations [23,24]. Medical and behavioral health history and demographic information were obtained by interview including self-or surrogate-reported history of physician diagnosed Alzheimer’s disease or dementia and, at the 10-year follow-up, self-report of memory problems [23].
Classification of Cognitive Impairment and Decline
Participants were classified as having incident cognitive impairment if they had a MMSE < 24, reported physician diagnosed AD or dementia, or were classified as having mild cognitive impairment or dementia at follow-up [23]. Using Principal component analysis, composite scores of cognitive function were constructed from the neurocognitive test data (TMT-A, TMT-B, DSST, AVLT, DSST) obtained at the 5- and 10-year phases. The 10% of participants with the most decline on the composite score between the 5- and 10-year phases were classified as having cognitive decline [23]. Details of the construction of the cognitive outcomes have been previously published. [23]
Blood Biomarker Measures
Non-fasting blood samples were obtained at the baseline examination. HbA1C and a complete blood count that included WBC and absolute counts for individual types of white blood cells, were measured in whole blood concurrent with the exam [22]. Additional whole blood and serum samples were stored at −80 degrees Celsius. IL-6 and sICAM-1 were measured in serum in 2009; high sensitivity CRP (hsCRP) and sVCAM-1 were measured in serum in 2013; and cadmium and lead were measured in whole blood in 2016 [22]. Laboratory methods and coefficients of variation (CV) for biomarker assays are shown in Table 1. The presence of an Apolipoprotein E epsilon 4 (APOE ε4) allele (Illumina IBC chip (Illumina, Inc., San Diego, CA, USA) was available for participants who were 45 years and older at baseline.
Table 1.
Distribution and Test Methods for Baseline Blood Biomarkers in the Beaver Dam Offspring Study
| Biomarker | N data | Mean (SD)(Range) | Method |
|---|---|---|---|
| HbA1C (%) | 2391 | 5.4 (0.6) (3.6–12.9) | High Performance Liquid Chromatography (HPLC) Tosoh A1c 2.2 Plus Glycohemoglobin Analyzer. Laboratory CV 1.4–1.9%. |
| WBC (k/μL) | 2401 | 7.3 (1.9) (3.0–16.6) | Automated hematology analyzer: Cell-Dyn 3200 (Abbott, Abbott Park, IL). |
| Abs. Neutrophils (k/μL) | 4.3 (1.5) (1.0–12.5) | ||
| Abs. Lymphocytes (k/μL) | 2.1 (0.7) (0.5–5.7) | ||
| Cadmium (μg/L) | 2203 | 0.44 (0.8) (0.1–21.6) | Inductively Coupled Plasma – Mass Spectrometry (ICP-MS). External quality evaluation through proficiency testing programs and quality control samples within ± 10% of the target. 10% of samples duplicated. |
| Lead (μg/dL) | 2203 | 1.53 (1.2) (0.15–25.3) | |
| hsCRP (mg/L) | 2350 | 2.8 (5.4) (0.12–125.8) | Latex-particle enhanced immunoturbidimetric assay kit/ Roche COBAS 800 Chemistry analyzer (Roche Diagnostics, Indianapolis, IN). Laboratory CV: 2.6%−4.3%. |
| IL-6 (pg/mL) | 2331 | 2.4 (5.09) (0.18–119.5) | High Sensitivity Human IL- 6 ELISA kit (R & D Systems, Minneapolis, MN, USA). Laboratory CV: 4.9–6.5% |
| sICAM-1 (ng/mL) | 2338 | 223.3 (71.7) (50.9–1055.8) | Human soluble ICAM-1 ELISA kit; (R & D Systems, Minneapolis, MN, USA). Laboratory CV: 7.8% |
| sVCAM-1 (ng/mL) | 2350 | 594.7 (198.6) (195.4–2920.9) | Human soluble VCAM-1/CD106 ELISA kit (R & D Systems, Minneapolis, MN, USA). Laboratory CV: 6.1% |
Abs: Absolute; CV: coefficient of variation; Hemoglobin A1C; hsCRP: high sensitivity C-reactive protein; IL-6: Interleukin-6; sICAM-1: soluble intercellular adhesion molecule-1; sVCAM-1: soluble vascular cell adhesion molecule-1; SD: Standard deviation; WBC: white blood cell count.
Statistical Analysis
Statistical analyses were conducted with SAS 9.4 (SAS Institute, Inc., Cary, NC, USA). Higher levels of inflammatory markers, cellular adhesion molecules, HbA1C, and neurotoxins levels were evaluated for their associations with the 10-year cumulative incidence of cognitive impairment and 5-year cognitive decline. WBC and HbA1C were analyzed as continuous variables. The NLR was calculated by dividing the absolute neutrophil count by the absolute lymphocyte count. IL-6, sICAM-1, sVCAM-1 were divided into tertiles (T) and analyzed as T3 vs T1–2 and hsCRP was divided into risk groups based on established clinical cut-points and analyzed as >3 vs ≤3 mg/L. Because of the highly skewed distribution of cadmium and lead, data were divided in quintiles(Q) and dichotomized with Q5 versus Q1–4. Cox proportional hazards with discrete ties (Proc Phreg) were used to evaluate associations between baseline blood biomarkers and the 10-year cumulative incidence of cognitive impairment. Poisson models (Proc Genmod) were used to estimate the relative risk of 5-year decline (binary outcome) between the 5- and 10-year examinations by baseline blood biomarker levels [25]. First, biomarkers were tested individually in models adjusted for age and sex and in sex-specific models adjusted for age. Biomarkers that were statistically significant (p<0.05) in age-sex adjusted models were then modelled together to determine their relative associations with cognitive decline and impairment. To adjust for relevant confounder variables, we included variables that have previously been associated with cognitive impairment (education, cardiovascular disease (CVD), alcohol, exercise, smoking (current), head injury) and decline (education, CVD, carotid plaque, alcohol) in our fully adjusted models in this cohort [23]. To avoid overfitting and multi-collinearity, other potential confounders such as hypertension, diabetes, body mass index, and waist circumference, which were not associated with the 10-year cumulative incidence of cognitive impairment or the 5-year incidence of cognitive decline in this cohort, were not included in these models. [23]
Sensitivity Analyses
ApoE ε4, available on a subset of participants, was added to the final models to determine if genetic status modified associations between biomarkers and cognitive outcomes. Because high hsCRP may be indicative of acute infection, sensitivity analyses were conducted by repeating the final models excluding participants with hsCRP >10 mg/L.
RESULTS
There were 2421 participants with baseline blood biomarker and cognitive data and, with cognitive test data from at least one follow-up examination. Participants were a mean age of 49 years (standard deviation: 9.8; range: 22–84 years) and 55% were women. The distribution of blood biomarkers is displayed in Table 1. Cadmium and lead levels were generally low (cadmium Q5: ≥0.53 μg/L; lead Q5: ≥2.07 μg/dL) and the mean NLR was 2.2 (SD=1.0).
10-year Cumulative Incidence of Cognitive Impairment
There were 88 (3.6%) cases of incident cognitive impairment over the 10 years. Higher levels of cadmium, sVCAM-1 and higher WBC were associated with the 10-year cumulative incidence of cognitive impairment in individual age-sex-adjusted models (Table 2). An interaction between sex and HbA1C was present and HbA1C was associated with an increased risk for developing cognitive impairment in women (Hazard Ratio (HR)=1.95, 95% Confidence Interval (CI)= 1.35, 2.80, per 1% increase), but not men (HR=1.01, 95% CI = 0.67, 1.52, per 1% increase). Lead, hsCRP, IL-6, sICAM-1 and NLR were not associated with the incidence of cognitive impairment.
Table 2.
Age-and Sex-Adjusted Associations of Individual Blood-based Biomarkers with Cognitive Impairment and Decline in the Beaver Dam Offspring Study
| 10-y Cumulative Incidence Cognitive Impairmenta | 5-y Cognitive Declineb | |||
|---|---|---|---|---|
| Baseline Biomarker | N | HR (95% CI) | N | RR (95% CI) |
| Menc | 1095 | 1.01 (0.67, 1.52) | 765 | 1.13 (0.91, 1.40) |
| Cadmium ≥0.53 μg/L | 2203 | 1.73 (1.04, 2.88) | 1629 | 1.21 (0.86, 1.71) |
| Lead ≥2.07 μg/dL | 2203 | 1.26 (0.75, 2.10) | 1629 | 1.28 (0.93, 1.76) |
| hsCRP >3.0 mg/L | 2350 | 1.12 (0.66, 1.91) | 1639 | 1.01 (0.73, 1.42) |
| IL-6 ≥2.28 pg/mL | 2331 | 1.09 (0.69, 1.73) | 1625 | 1.44 (1.08, 1.92) |
| sICAM-1 ≥ 238.5 ng/mL | 2338 | 1.27 (0.81, 1.98) | 1630 | 1.25 (0.93, 1.67) |
| sVCAM-1 ≥635.4 ng/mL | 2350 | 1.59 (1.02, 2.49) | 1639 | 1.39 (1.02, 1.89) |
| WBC, per 103/μL | 2401 | 1.15 (1.04, 1.28) | 1612 | 1.13 (1.05, 1.21) |
| NLR | 2401 | 0.99 (0.78, 1.25) | 1612 | 0.93 (0.79, 1.09) |
Cox Proportional hazards models adjusted for age and sex.
Poisson models adjusted for age and sex
Due to significant HbA1C by sex interaction, results are presented stratified by sex. And only adjusted for age
Note. HR: hazard ratio; RR: relative risk; CI: confidence interval; HbA1C: Hemoglobin A1C; hsCRP: high sensitivity C-reactive protein; IL-6: Interleukin-6; NLR: neutrophil to lymphocyte ratio; sICAM-1: soluble intercellular adhesion molecule-1; sVCAM-1: soluble vascular cell adhesion molecule-1; WBC: white blood cell count; y: year.
In the model including all biomarkers significant in the age-sex-adjusted models, sVCAM-1 and HbA1C were associated with the incidence of cognitive impairment (Table 3, model 1). The cadmium and WBC estimates were attenuated and no longer statistically significant when modelled together. A post-hoc t-test showed WBC was significantly higher in participants with cadmium levels in Q5 versus those with levels in Q 1–4 (mean WBC 8.1 k/μL vs 7.0 k/ μL, p<0.0001, respectively).
Table 3.
Associations of Biomarkers with Cognitive Impairment in the Beaver Dam Offspring Study
| 10-year Cumulative Incidence of Cognitive Impairment | ||
|---|---|---|
| Baseline Biomarker | Model 1a
HR (95% CI) |
Model 2b
HR (95% CI) |
| Men | 0.86 (0.52, 1.43) | 0.69 (0.40, 1.17) |
| Cadmium ≥0.53 μg/L | 1.55 (0.90, 2.66) | 1.24 (0.59, 2.58) |
| sVCAM-1 ≥635.4 ng/mL | 1.79 (1.10, 2.89) | 1.72 (1.05, 2.82) |
| WBC, per 103/μL | 1.11 (0.98, 1.25) | 1.09 (0.96, 1.23) |
Note. N=2143. Results are based on Cox proportional hazards models. HR: hazard ratio; CI: confidence interval; HbA1C: Hemoglobin A1C; sVCAM-1: soluble vascular cell adhesion molecule-1; WBC: white blood cell count.
Model 1: Includes HbA1C, Cadmium, sVCAM-1, WBC and age and sex.
Model 2: Includes HbA1C, Cadmium, sVCAM-1, WBC and age, sex, education, current smoking, exercise, weekly alcohol consumption, history of head injury and cardiovascular disease.
Due to significant HbA1C by sex interaction, results are presented stratified by sex.
sVCAM-1 (HRT3vs 1–2=1.72, 95% CI=1.05, 2.82) and HbA1C (HR=1.75, 95% CI=1.18, 2.59, in women) remained independent risk factors for the development of cognitive impairment in the full model adjusted for age, sex, education, current smoking, exercise, weekly alcohol consumption, history of head injury and cardiovascular disease. (Table 3, model 2).
Cognitive Decline
There were 1683 participants with complete cognitive function test data from the 5- and 10-year phases and of those, 166 (10%) were classified as having 5-year cognitive decline. Baseline biomarkers associated with an increased risk of cognitive decline in individual age-sex- adjusted models included higher IL-6, sVCAM-1, and WBC. (Table 2) HbA1C, cadmium, lead, hsCRP, sICAM-1 and NLR were not associated with the 5-year incidence of cognitive decline.
In a model including age, sex, and all the biomarkers significant in the individual biomarker models, sVCAM-1 and WBC remained associated with cognitive decline, but IL-6 did not (Table 4, model 1). sVCAM-1(Relative Risk (RR)T3vs 1–2=1.45, 95% CI= 1.06, 1.99) and WBC (RR=1.10, 95% CI=1.02, 1.19, per 103/μL) remained independent predictors of cognitive decline in the full model adjusted for age, sex, education, cardiovascular disease, carotid artery plaque, and a history of heavy alcohol use. (Table 4, model 2).
Table 4.
Associations of Biomarkers with 5-year Cognitive Decline in the Beaver Dam Offspring Study
| 5-year Cognitive Decline | ||
|---|---|---|
| Baseline Biomarker | Model 1a
RR (95% CI) |
Model 2b RR (95% CI) |
| IL-6 ≥2.28 pg/mL | 1.20 (0.88, 1.65) | 1.14 (0.83, 1.55) |
| sVCAM-1 ≥635.4 ng/mL | 1.46 (1.07, 2.00) | 1.45 (1.06, 1.99) |
| WBC, per 103/μL | 1.12 (1.03, 1.20) | 1.10 (1.02, 1.19) |
Note. N=1574. Results are based on Poisson models; CI: confidence interval; IL-6: interleukin-6; RR: relative risk; sVCAM-1: soluble vascular cell adhesion molecule-1; WBC: white blood cell count.
Model 1: Includes IL-6, sVCAM-1, WBC and age and sex.
Model 2: Includes IL-6, sVCAM-1, WBC and age, sex, education, cardiovascular disease, carotid artery plaque, and history of heavy alcohol use.
Sensitivity Analyses
The final models for cognitive decline and impairment were repeated after removing participants with a hsCRP >10 mg/L and results were similar. APOEε4 genotype was not significantly associated with either cognitive outcome; including it in the final models did not alter associations.
DISCUSSION
In this longitudinal study, blood-based biomarkers of inflammation and metabolic dysregulation were associated with decline in cognitive function and 10-year cumulative incidence of cognitive impairment in a primarily middle-aged cohort. sVCAM-1, WBC and HbA1C were independent predictors of cognitive decline and impairment in models including education, and cardiovascular-related risk factors and disease. These results are consistent with the hypothesis that inflammation and metabolic dysregulation, potentially modifiable conditions, may be important in cognitive aging and extends previous research in older adults to midlife and a general population cohort. Additionally, we determined the independent associations of the biomarkers with cognitive decline and impairment, by modeling the biomarkers together and further adjusting for a comprehensive set of potential confounders.
In the current study, we found higher blood levels of sVCAM-1 were associated with an increased risk for both cognitive decline and impairment. Previous studies of sVCAM-1 and cognitive dysfunction in older adults have been inconsistent [5,13]. A cross-sectional study reported elevated sVCAM-1 associated with poorer performance on cognitive function tests, worse general cognitive function, and higher cerebrovascular resistance [13] whereas a longitudinal study reported no association between sVCAM-1 level and development of dementia [5].
Our study extends previous research to midlife and suggests that sVCAM-1 may be an important biomarker of cognitive decline. Expression of the cell adhesion marker VCAM-1 is stimulated by cytokines and sVCAM-1 is associated with endothelial cell dysfunction and atherosclerosis and is therefore considered a marker for inflammation-associated endothelial damage [12,13]. Thus, elevated sVCAM-1 could potentially indicate the presence of inflammation or atherosclerosis, pathophysiological processes which have been associated with cognitive decline and impairment [12, 26]. Previously in this cohort, we found the presence of carotid plaque increased the risk of cognitive decline and a history of CVD increased the risk of incident cognitive impairment [23]. In our models, sVCAM-1 remained an independent predictor of 5-year cognitive decline and the 10-year cumulative incidence of cognitive impairment even after we adjusted for subclinical atherosclerosis and CVD, respectively, suggesting it might be an indicator of pathophysiological processes that are not captured by these vascular disease indicators.
We also found that WBC, a general marker of inflammation, was associated with an increased risk of cognitive decline. This is consistent with studies showing WBC associated with brain aging, atherosclerosis, and mortality [14,27,28]. The NLR was not associated with the 10-year cumulative incidence of cognitive impairment or with the 5-year incidence of cognitive decline. Findings from a longitudinal study by Ramos-Cejudo et. al. [11], showed NLR associated with increased risk of dementia during a median of 6 years of follow-up in the Framingham cohort. Our findings may have differed due to the large age difference in the cohorts. The BOSS participants were on average 20 years younger than those in the Framingham study (mean age 49 years versus 69 years, respectively) [11]. NLR may be a better predictor of incident dementia in older adults who may have more advanced pathology. This would be consistent with cross-sectional studies that have found the NLR higher in people with clinical AD or cognitive impairment than in healthy controls [8,9].
In this study, we evaluated HbA1C as our blood biomarker of metabolic dysregulation and found that in women, HbA1C level was associated with an increased risk for developing cognitive impairment. Although diabetes was not previously associated with the incidence of cognitive impairment in this cohort [23], evaluating HbA1C levels as a continuous measure may be a more sensitive indicator of glucose dysregulation than a binary variable which classifies people with diabetes who have good glycemic control in the same category as those who have poor control. Non-blood markers of metabolic dysregulation, waist circumference, hypertension, and diabetes have been evaluated previously in this cohort and were not associated with cognitive decline and impairment.[23]. HbA1C may be a better indicator of metabolic dysregulation in middle-age when people at higher risk for metabolic syndrome may be starting to have higher levels of HbA1C but may have not yet reached the threshold for diabetes. In line with this, other studies have found HbA1C levels associated with cognitive decline in people without diagnosed diabetes [29,30]. Sex-specific differences in metabolic regulation and vascular pathologies have been reported previously. For example, in a meta-analysis, women with diabetes had a greater risk of developing vascular dementia than did men with diabetes [31]. Additional studies are needed to clarify the sex-specific associations between HbA1C levels and cognitive impairment.
Several blood biomarkers which were statistically significant in the age-sex adjusted models, were no longer significant when modelled together with other biomarkers or, in models that included cardiovascular-related risk factors. This may be due to the complex interrelationships of the different biomarkers with each other [12,32]. IL-6, a cytokine produced by white blood cells, which has been associated with an increased risk for cognitive decline or impairment in other studies was previously associated with prevalent brain aging in the BOSS cohort. [4,6,12,15]. In the current study, IL-6 was associated with cognitive decline in an age-sex-adjusted model but not in the multivariable model including WBC which suggests IL-6 may be in the pathway between WBC and cognitive decline. Similarly, cadmium and WBC were both attenuated when modelled together suggesting they may be capturing the same underlying pathophysiological pathway. Cadmium has been previously reported to be associated with inflammation and atherosclerosis [33] and, in the BOSS cohort, an increased risk for sensorineural impairments [22,34,35]. Some of the other risk factors for cognitive impairment, such as smoking, exercise, and CVD, are also associated with many of the blood-based biomarkers and including them in the final full models may have led to over adjustment. [12,33,36].
There are a few limitations of this study that should be considered. The BOSS is a primarily non-Hispanic white cohort and results may not be generalizable to other populations. One-time measures of the biomarkers may not reflect long-term exposure to inflammation, metabolic dysregulation, or environmental neurotoxins. Still, relatively higher levels of these biomarkers in midlife might indicate a predisposition towards inflammation or metabolic dysregulation. Future studies with repeated measures of both biomarkers and cognitive outcomes may help elucidate the complex processes involved in brain aging. While the objective was to study the association of midlife risk factors with cognitive outcomes, in this primarily healthy middle-aged cohort, the incidence of cognitive impairment and the prevalence of high lead exposure were both low, which may have limited our power to detect some associations However, despite the younger age of the cohort and lower incidence of cognitive outcomes, WBC was associated with cognitive decline and sVCAM-1 with cognitive decline and 10-yr incidence of cognitive impairment.
The strengths of this study include the large, well characterized cohort with midlife measures of blood-based biomarkers and up to 10-years of follow-up of cognitive function. Additionally, we had standardized and repeated measures of cognitive function at each phase and extensive covariate information including measures of vascular health.
CONCLUSION
In conclusion, midlife inflammatory, vascular, and metabolic physiological profiles may be important indicators for cognitive health later in life. These results extend previous research on modifiable risk factors and early markers of cognitive decline in midlife, a time when intervention and prevention methods may be more efficacious.
ACKNOWLEDGEMENTS
FUNDING:
This work was supported by R01AG021917(Karen J. Cruickshanks) from the National Institute on Aging and an unrestricted grant from Research to Prevent Blindness, Inc., to the Department of Ophthalmology and Visual Sciences. The content is solely the responsibility of the authors and does not necessarily reflect the official views of the National Institute on Aging or the National Institutes of Health.
Footnotes
DECLARATIONS OF INTEREST: On behalf of all authors, the corresponding author states that there is no conflict of interest.
ETHICS APPROVAL: This study was performed in line with the principles of the Declaration of Helsinki. Approval for this research was obtained from the University of Wisconsin Health Sciences Institutional Review Board.
CONSENT: Written informed consent was obtained from all participants at each phase prior to examination.
DATA AVAILABILITY:
Data are available to researchers through data sharing (Data Use) agreements. All Data Use agreements require approval by an institutional review board.
REFERENCES
- 1.Sperling RA, Aisen PS, Beckett LA, Bennett DA, Craf S, Fagan AM, Iwatsubo T, Jack CR, Kaye J, Montine TJ, Park DC, Reiman EM, Rowe CC, Siemers E, Stern Y, Yaffe K, Carrillo MC, Thies B, Morrison-Bogorad M, Wagster MV, Phelps CH Toward defining the pre-clinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7(3):280–92. doi: 10.1016/j.jalz.2011.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gu Y Vorburger R, Scarmeas N, Luchsinger JA, Manly JJ, Schupf N, Mayeux R, Brickman AM Circulating inflammatory biomarkers in relation to brain structural measurements in a non-demented elderly population. Brain Behav Immun. 2017; 65: 150–160. doi: 10.1016/j.bbi.2017.04.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Warren KN, Beason-Held LL, Carlson O, Egan JM, An Y, Doshi J Davatzikos C, Ferrucci L, Resnick SM Elevated markers of inflammation are associated with longitudinal changes in brain function in older adults. J Gerontol A Biol Sci Med Sci. 2018; 73(6): 770–778. doi: 10.1093/Gerona/glx199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Koyama A, O’Brien J, Weuve J, Blacker D, Metti AL, Yaffe K The role of peripheral inflammatory markers in dementia and Alzheimer’s disease: A meta-analysis. J Gerontol A Biol Sci Med Sci. 2013;68(4): 433–440. doi: 10.1093/gerona/gls187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Engelhart MJ, Geerlings MI, Meijer J, Kiliaan A, Ruitenberg A, van Swieten JC, Stijnen T, Hofman A, Witteman JCM, Breteler MMB Inflammatory proteins in plasma and the risk of dementia. Arch Neurol. 2004;61: 668–672. doi: 10.1001/archneur.61.5.668. [DOI] [PubMed] [Google Scholar]
- 6.Wichmann MA, Cruickshanks KJ, Carlsson CM, Chappell R, Fischer ME, Klein BEK, Klein R, Tsai MY, Schubert CR Long-term systemic inflammation and cognitive impairment in a population-based cohort. J Am Geriatr Soc. 2014; 62(9):1683–91. doi: 10.1111/jgs.12994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Walker KA, Gottesman RG, Wu A, Knopman DS, Gross AL, Mosley TH, Selvin E, Windham BG Systemic inflammation during midlife and cognitive change over 20 years. The ARIC Study. Neurology. 2019;92(11):e1256–e1267. doi: 10.1212/SNL.0000000000007094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rembach A, Watt AD, Wilson WJ, Rainey-Smith S, Ellis KA, Rowe CC, Villemagne VL, Macaulay SL, Bush AI, Martins RN, Ames D, Masters CL, Doecke JD, the AIBL Research Group. An increased neutrophil-lymphocyte ratio in Alzheimer’s disease is a function of age and is weakly correlated with neocortical amyloid accumulation. J Neuroimmunol. 2014; 273(1–2):65–71. doi: 10.1016/j.jneuroim.2014.05.005. [DOI] [PubMed] [Google Scholar]
- 9.Liu JH, Zhang YJ, Ma QH, Sun HP, Xu Y, Pan CW. Elevated blood neutrophil to lymphocyte ratio in older adults with cognitive impairment. Arch Gerontol Geriatr. 2020;88:104041. doi: 10.1016/j.archger.2020.104041. [DOI] [PubMed] [Google Scholar]
- 10.van der Willik KD, Fani L, Rizopoulos D, Licher S, Fest J, Schagen SB, Ikram MK, Ikram MA. Balance between innate versus adaptive immune system and the risk of dementia: a population-based cohort study. J Neuroinflammation. 2019;16(1):68. doi: 10.1186/s12974-019-1454-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ramos-Cejudo J, Johnson AD, Beiser A, Seshadri S, Salinas J, Berger JS, Fillmore NR, Do N, Zheng C, Kovbasyuk Z, Ardekani BA, Bubu OM, Parekh A, Convit A, Betensky RA, Wsniewski TM, Osorio RS. The neutrophil to lymphocyte ratio is associated with the risk of subsequent dementia in the Framingham Heart Study. Front Aging Neurosci. 2021;13:773984. doi: 10.3389/fnagi.2021.773984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Stoner L, Lucero AA, Palmer BR, Jones LM, Young JM, Faulkner J Inflammatory biomarkers for predicting cardiovascular disease. Clin Biochem. 2013;46: 1353–1371. doi: 10.1016/j.clinbiochem.2013.05.070. [DOI] [PubMed] [Google Scholar]
- 13.Tchalla AE, Wellenius GA, Sorond FA, Gagnon M, Iloputaife I, Travison TG, Dantoine T, Lipsitz LA Elevated soluble vascular cell adhesion molecule is associated with cerebrovascular resistance and cognitive function. J Gerontol A Biol Sci Med Sci. 2017;72(4): 560–566. doi: 10.1093/gerona/glw099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yoon CY, Steffen LM, Gross MD, Launer LJ, Odegaard A Reiner A, Sanchez O, Yaffe K, Sidney S, Jacobs DR Circulating cellular adhesion molecules and cognitive function: The Coronary Artery Risk development in Young Adults Study. Front Cardiovasc Med. 2017. May 24;4: 37. doi: 10.3389/fcvm.2017.00037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Schubert CR, Fischer ME, Pinto AA, Chen Y, Klein BEK, Klein R, Tsai MY, Tweed TS, Cruickshanks KJ. Brain aging in midlife: The Beaver Dam Offspring Study. J Am Geriatr Soc. 2019;67(8):1610–6. doi: 10.1111/jgs.15886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Roberts RO, Knopman DS, Przybelski SA, Mielke MM, Kantarci K, Preboske GM Senjem ML, Pankratz VS, Geda YE, Boeve BF Ivnik RJ, Rocca WA, Petersen RC, Jack CR Association of type 2 diabetes with brain atrophy and cognitive impairment. Neurology. 2014. Apr 1;82(13):1132–41. doi: 10.1212/WNL.0000000000000269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yaffe K, Blackwell T, Whitmer RA, Krueger K, Barrett-Connor E Glycosylated hemoglobin level and development of mild cognitive impairment or dementia in older women. J Nutr Health Aging. 2006;10(4): 293–295. [PubMed] [Google Scholar]
- 18.Christma AL, Matsushia K, Gottesman RF, Mosley T, Alonso A, Coresh J Hill-Briggs F, Sharrett AR, Selvin E Glycated haemoglobin and cognitive decline: the Atherosclerosis Risk in Communities (ARIC) study. Diabetologia. 2011;54(7): 1645–1652. doi: 10.1007/s00125-011-2095-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Karri V, Schuhmacher M, Kumar V Heavy metals (Pb, Cd, As and MeHg) as risk factors for cognitive dysfunction: A general review of metal mixture mechanism in brain. Environ Toxicol Pharmacol. 2016;48: 203–213. doi: 10.1016/j.etap.2016.09.016. [DOI] [PubMed] [Google Scholar]
- 20.Pryzbyla J, Houseman EA, Smit E, Kile ML A path analysis of multiple neurotoxic chemicals and cognitive functioning in older US adults (NHANES 1999–2002). Environ Health. 2017;16(1):19. doi. 10.1186/s12940-017-0227-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Li H, Wang Z, Fu Z, Yan M, Wu N, Wu H, Yin P Associations between blood cadmium levels and cognitive function in a cross-sectional study of US adults aged 60 years or older. BMJ Open. 2018; 8(4): e020533. doi: 10.1136/bmjopen-2017-020533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Dalton DS, Schubert CR, Pinto A, Fischer ME, Huang G-H, Klein BEK, Klein R, Pankow JS, Paulsen AJ Tsai MY, Tweed TS, Cruickshanks KJ Cadmium, obesity, and education and the 10-year incidence of hearing impairment. Laryngoscope. 2020;130(6):1396–1401. doi: 10.1002/lary.28244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Schubert CR, Cruickshanks KJ, Fischer ME, Pinto AA, Chen Y, Huang GH, Klein BEK, Klein R, Pankow JS, Paulsen AJ, Dalton DS, Tweed TS Sensorineural impairments, cardiovascular risk factors, and 10-year incidence of cognitive impairment and decline in midlife: The Beaver Dam Offspring Study. J Gerontol A Biol Sci Med Sci. 2019; 74(11):1786–92. doi: 10.1093/gerona/glz011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Strauss E, Sherman EMS, & Spreen O (2006). A Compendium of Neuropsychological Tests: Administration, Norms, and Commentary (3rd ed.) Oxford University Press, New York, NY. [Google Scholar]
- 25.Zou G A Modified Poisson Regression Approach to Prospective Studies with Binary Data. Am J Epidemiol. 2004;159: 702–706. [DOI] [PubMed] [Google Scholar]
- 26.Cunningham C, Hennessy E Co-morbidity and systemic inflammation as drivers of cognitive decline: new experimental models adopting a broader paradigm in dementia research. Alzheimers Res Ther. 2015;7:33, doi 10.1186/s13195-015-0117-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Janowitz D, Habes M, Toledo JB, Hannemann A, Frenzel S, Terock J, Davatzikow C, Hoffmann W, Grabe HJ Inflammatory markers and imaging patterns of advanced brain aging in the general population. Brain Imaging Behav. 2020; 14(4), 1108–1117. doi: 10.1007/s11682-019-00058-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ruggiero C, Metter EJ, Cherubini A, Maggio M, Sen R, Najjar SS Windham GB, Ble A, Senin U, Ferrucci L White blood cell count and mortality in the Baltimore Longitudinal Study of Aging. J Am Coll Cardiol, 2007;49(18): 1841–1850. doi: 10.1016/j.jacc.2007.01.076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Rawlings AM, Sharrett AR, Schneider ALC, Coresh J, Albert M, Couper D, Griswold M, Gottesman RF, Wagenknecht LE, Windham BG, Selvin E Diabetes in midlife and cognitive change over 20 years: A cohort study. Ann Intern Med. 2014; 161(11):785–793. doi: 10.7326/M14-0737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Marden J, Mayeda ER, Tchetgen Tchetgen EJ, Kawachi I, Glymour MM High hemoglobin A1c and diabetes predict memory decline in the Health and Retirement Study. Alzheimer Dis Assoc Disord. 2017;31(1): 48–54. doi: 10.1097/WAD.0000000000000182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chatterjee S, Peters SAE, Woodward M, Arango SM, Batty GD, Beckett N, Beiser A, Borenstein AR, Crane PK Haan M, Hassing LB, Hayden KM, Kiyohara Y, Larson EB, Li C, Ninomiya T, Ohara T, Peters R, Russ TC, Seshadri S, Strand BH, Walker R, Xu W, Huxley RR Type 2 diabetes as a risk factor for dementia in women compared with men: A pooled analysis of 2.3 million people comprising more than 100,000 cases of dementia. Diabetes Care. 2016;39: 300–307. doi: 10.2337/dc15-1588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Yaffe K, Kanaya A, Lindquist K, Simonsick EM, Harris T, Shorr RI, Tylavsky FA, Newman AB The metabolic syndrome, inflammation, and risk of cognitive decline. JAMA. 2004;292(18): 2237–2242. doi: 10.1001/jama.292.18.2237. [DOI] [PubMed] [Google Scholar]
- 33.Tinkov AA, Filippini T, Ajsuvakova OP Skalnaya MG, Aaseth J, Bjorklund G, Gatiatulina ER, Popova EV, Nemereshina ON, Huang PT, Venceti M, Skalny AV Cadmium and atherosclerosis: A review of toxicological mechanisms and a meta-analysis of epidemiologic studies. Environ Res. 2018;162: 240–260. doi: 10.1016/j.envres.2018.01.008. [DOI] [PubMed] [Google Scholar]
- 34.Schubert CR, Pinto AA, Paulsen AJ, Cruickshanks KJ. Exposure to cadmium, lead, and tobacco smoke and the 10-year cumulative incidence of olfactory impairment: The Beaver Dam Offspring Study. JAMA Otolaryngol Head Neck Surg. 2021;147(6): 510–517. doi: 10.1001/jamaoto.2021.0079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Paulsen AJ, Schubert CR, Johnson LJ, Chen Y, Dalton DS, Klein BEK, Klein R, Pinto A, Cruickshanks KJ. Association of cadmium and lead exposure with the incidence of contrast sensitivity impairment among middle-aged adults. JAMA Ophthalmol. 2018;136(12):1342–1350. doi: 10.1001/jamaophthalmol.2018.3931. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Singh T, Newman AB Inflammatory markers in population studies of aging. Aging Res Rev. 2011;10: 319–329. doi: 10.1016/j.arr.2010.11.002. [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.
Data Availability Statement
Data are available to researchers through data sharing (Data Use) agreements. All Data Use agreements require approval by an institutional review board.