Blood biomarkers of Alzheimer's disease in the community: Variation by chronic diseases and inflammatory status

We explored the variations of blood biomarkers of Alzheimer's disease (AD) by chronic diseases and systemic inflammation.


Highlights
• Participants with a complex clinical profile (i.e., multiple co-occurring diseases or specific disease combinations) display elevated levels of AD blood-biomarkers.
• Anemia, heart, cerebrovascular, and kidney diseases are associated with variations is the levels of AD blood biomarkers in cognitively intact older adults.
• Systemic inflammation amplifies the association between several chronic diseases and AD blood biomarkers.

BACKGROUND
4][5][6] These promising results, coupled with the minimal invasiveness and lower cost of blood biomarker measurement, suggest the potential for their clinical implementation in early AD detection.However, most prior studies on blood biomarkers of AD were conducted in specialized clinical settings, and their findings may not be directly translated to the general population. 7Indeed, in the community individuals are more heterogeneous and often present with multiple chronic diseases that could potentially influence the blood concentration of these biomarkers, posing challenges to their clinical interpretation. 7,8It is therefore important to evaluate factors that may influence the levels of AD blood biomarkers in population-based studies, that are more representative of real-world settings.
Whether other conditions, as well as the co-occurrence of multiple diseases, may influence the blood levels of these biomarkers in a real-world setting is still under-investigated.Similarly, the potential role played in these associations by systemic low-grade inflammation requires proper exploration.In fact, systemic inflammation might be interpreted as a proxy of an accelerated aging process and is often associated with higher clinical complexity 16 in older adults.
Using data from a population-based cohort of 2366 dementia-free individuals, aged 60 years or older, from the Swedish National Study on Aging and Care in Kungsholmen (SNAC-K), this study aimed to explore (1) the association of demographics, apolipoprotein E (APOE) genotype and chronic diseases, alone and in combination, with the concentrations of several blood biomarkers of AD; and (2) whether the associations between chronic diseases and AD blood biomarkers varied by systemic inflammatory status.

Study population
We used data from the SNAC-K, 17

Blood biomarkers and APOE genotyping
Peripheral venous blood samples were collected at baseline (fasting was not compulsory) and upon centrifugation serum aliquots were stored at the Karolinska Institutet Bio Bank at −80

Chronic diseases and cardiovascular risk factors
Chronic diseases and cardiovascular risk factors were identified at baseline by the examining physician through a previously reported comprehensive assessment. 20Diagnoses were based on self-reports, medical examination, laboratory and instrumental tests, medication use, medical charts and data from the Swedish National Patient Register (inpatient and outpatient care).Diseases were coded following the International Classification of Diseases, 10 th revision (ICD-10).In the present study we investigated: hypertension, dyslipidemia, body mass index [BMI] ≥30, diabetes, chronic kidney disease, heart diseases (i.e., ischemic heart disease, atrial fibrillation, heart failure), cerebrovascular disease, cancer, and anemia.This selection was based on previous findings and knowledge gaps, as well as considering the high prevalence of such conditions in the older population.

Statistical analyses
We used Kruskal-Wallis test to compare biomarkers' concentrations between groups, and Spearman correlations to explore betweenbiomarker correlations.Blood biomarkers were transformed into zscores based on baseline mean and standard deviation of the study population, facilitating comparison between coefficients.
Quantile regression models on the 50 th (median) percentile were used to examine the associations of demographics, APOE

RESEARCH IN CONTEXT
1. Systematic review: The authors reviewed the literature (PubMed and Embase) and found few population-based studies reporting an association between some chronic diseases and altered levels of Alzheimer's disease (AD) blood biomarkers.However, data on the co-occurrence of multiple diseases and on the involvement of systemic low-grade inflammation in these associations are still lacking.

Interpretation:
In the community, the concentration of AD blood biomarkers varies in relation to multiple factors.
Among chronic diseases, the strongest associations were found for anemia, heart, cerebrovascular, and kidney diseases, alone and-even more-when combined.Systemic inflammation was associated with elevated levels of neurofilament light chain (NfL) and amplified the associations between several chronic diseases and elevated AD blood biomarkers.
3. Future directions: Future studies are needed to clarify the mechanisms underlying the associations between chronic diseases and AD blood biomarkers and ascertain whether these influences can affect the diagnostic and prognostic performance of blood biomarkers in detecting impending dementia.
genotype, chronic diseases, and IL-6 (predictors) with AD blood biomarkers (outcomes).The associations were tested using separate models (i.e., one for each predictor) adjusted for age, sex, and education.Analyses for the association between IL-6 and blood biomarkers of AD were further adjusted for the number of chronic diseases.
Additionally, we tested the interaction between IL-6 concentration (continuous values, centered around the median) and chronic diseases in relation to AD blood biomarkers and repeated the analyses for the association between chronic diseases and AD blood biomarkers stratifying by IL-6 concentration (third vs. first and second tertiles).
In a sensitivity analysis, we explored the associations between demographics, APOE genotype, chronic diseases, IL-6 (predictors) and AD blood biomarkers (outcomes), adjusted for age, sex, and education, excluding participants with MMSE score below 27 (n = 250) or missing data on MMSE (n = 4), 21 to further restrict the analysis to a cognitively intact population.
A two-tailed p-value < 0.05 was considered statistically significant in all analyses.The statistical analyses were performed using Stata

Characteristics of the study population
Characteristics of the study population are reported in Table 1.Overall, median age was 72.2 years, 61.7% were females, 35.8% had an educational level of university or above, and 29.4% were APOE-ε4 carriers.Older participants had higher concentrations of all the biomarkers, except for the Aβ42/40 ratio, which was lower in the older age groups (Figure 1).
Table S1 reports the comparison between dementia-free SNAC-K participants with and without AD biomarkers.Overall, participants without AD blood biomarkers were older, more likely to be female and had a lower educational level than those with available biomarkers; they also presented a higher number of chronic diseases and a lower MMSE score.

Chronic diseases and cardiovascular risk factors
We found an association between a higher number of co-occurring chronic diseases and higher concentrations of p-tau181, t-tau, NfL, and GFAP, following a dose-response relationship (p for trend < 0.01) (Figure 3).Considering individual chronic diseases, participants with cerebrovascular disease showed lower Aβ42/40 ratio and higher concentrations of NfL and GFAP, and those with heart diseases showed lower Aβ42/40 ratio and higher levels of p-tau181, t-tau, and NfL, compared with participants free from these diseases (Figure 4).Additionally, chronic kidney disease was associated with elevated levels of p-tau181, t-tau, NfL, and GFAP.Variations in the levels of all the examined biomarkers were observed in presence of anemia (Figure 4).
Looking at specific disease combinations, the concentrations of p-tau181, t-tau, NfL, and GFAP were even further elevated when two among chronic kidney disease, anemia, and heart diseases co-occurred in the same person (Figure S3).
We found an inverse association between cardiovascular risk factors and several blood biomarkers of AD.Participants with hypertension had lower concentrations of p-tau181 and t-tau compared with those with normal/low blood pressure, and those with BMI ≥30 had lower levels of NfL and GFAP than those with BMI < 30.We also found an association between dyslipidemia and lower concentrations of ttau and NfL (Figure 4).Diabetes was instead associated with elevated levels of p-tau181 (Figure 4).
In analyses stratified by IL-6 levels (Figure 5), the association between chronic kidney disease and concentration of NfL, t-tau, and GFAP was stronger in participants with high levels of IL-6 than in those with low IL-6.Participants in the high IL-6 subgroup also exhibited a stronger association between cerebrovascular disease and levels of NfL and GFAP than those in the low IL-6 subgroup.Finally, the association between anemia and the concentrations of all the biomarkers, except for the Aβ42/40 ratio, was stronger in participants with high IL-6 than in those with low IL-6 (Figure 5).

Sensitivity analysis
Most of the associations between demographics, APOE genotype, chronic diseases, IL-6 levels, and blood biomarkers of AD did not change after the exclusion of participants with a MMSE score below 27 (Figure S4 and S5).However, the associations between atrial fibrillation

DISCUSSION
In this large population-based cohort of dementia-free older adults, the concentration of blood biomarkers of AD varied in relation to multiple factors, including demographics, APOE genotype, medical conditions, and systemic inflammation.Identifying those factors that may influence the levels of blood biomarkers of AD is a critical step in the roadmap toward their implementation in clinical practice. 8,22ile previous studies on AD blood biomarkers were primarily conducted in specialized clinical settings, limited evidence arises from community-dwelling cohorts, which are more representative of the real-world setting.We not only confirm but also extend previous findings by assessing diseases that were not considered before, such as anemia, investigating the co-occurrence of multiple diseases and exploring the influence of systemic inflammation on AD blood biomarkers concentration.In this exploratory study, we were able to consider a large set of variables that -directly or indirectly -could alter the blood concentration of five different AD blood biomarkers.
Based on our findings, older participants exhibited lower Aβ42/40 ratio and higher concentrations of the other biomarkers than their younger counterparts and APOE-ε4 carriers had a lower Aβ42/40 ratio and higher p-tau181 concentration than non-carriers.These results align with previous studies 4,9,10,14 and may reflect the increased neuropathological burden associated with aging and APOE-ε4 carriership, which in turn influence the risk of dementia.
Beyond age and genetic predisposition, we explored several medical conditions in relation to the circulating level of blood biomarkers.
We observed a dose-response relationship between the number of cooccurring chronic diseases and the levels of all the examined biomarkers, except for the Aβ42/40 ratio.Similarly, two studies from the Mayo Clinic Study of Aging found elevated levels of blood biomarkers of AD in participants with higher Charlson comorbidity index. 9,10These results align with the concept that a higher disease burden may contribute to brain pathology and accelerate cognitive decline. 23Notably, previous studies found that a high chronic disease burden was associated with neuroimaging markers of neurodegeneration (i.e., brain atrophy and reduced metabolism) [24][25][26] but not with amyloid deposition. 25,27e association between chronic diseases and AD blood biomarkers may arise from various mechanisms.Certain diseases may exert an impact on brain structure and function, increasing neuropathology and dementia risk, while others may interfere peripherally with the distribution and metabolism of blood biomarkers.
Delving further into individual diseases, we found that cerebrovascular, heart and kidney diseases, as well as anemia, were those most frequently associated with variations in biomarker levels.The heart-brain connection is known to play a role in dementia development. 28Heart diseases are linked to cerebral hypoperfusion, 29 embolization, cerebral small vessels disease, 30 as well as brain 31,32 and hippocampal 33 atrophy.The variation in the levels of AD blood biomarkers associated with heart diseases might reflect these mechanisms.
Elevated levels of blood biomarkers of AD in relation to renal function have been repeatedly observed 9,10,13,34,35 and may be attributed to reduced renal clearance.In line with this hypothesis, Stocker et al.
found that impaired kidney function was associated with higher concentrations of blood biomarkers of AD but not with higher dementia risk. 15This suggests that kidney impairment may alter the level of AD blood biomarkers through a peripheral mechanism rather than through an effect on neuropathology.Nevertheless, a link between impaired kidney function and higher dementia risk has also been sparsely reported. 36Future studies should delve deeper into the link between kidney function, AD blood biomarkers and dementia risk and assess whether different biomarkers' cutoffs should be applied for individuals with impaired kidney function.
Ours is the first study that reports an association between anemia and variations in the levels of blood biomarkers of AD.Anemia is a highly prevalent condition in the older population, affecting 10% of individuals over age 65 37 and up to 30% of those over age 80. 38 There is some evidence that individuals with anemia are at higher risk of dementia 39,40 and that low hemoglobin is associated with cerebral hypoxia and neuroimaging pathological changes. 39At the same time, in the older population, anemia may be the result of multiple other conditions, including chronic diseases (e.g., CKD, cancer) and chronic inflammation, thus proxying a greater clinical complexity encountered in some older adults. 37The impact of anemia on blood biomarkers of AD and on dementia risk needs to be further investigated.
Older adults are known to suffer from multiple chronic conditions, 20 which tend to cluster in the same individual following shared underlying biological mechanisms and/or risk factors. 41We explored how the levels of AD blood biomarkers varied in relation to combinations of anemia, heart, and chronic kidney diseases and observed that biomarkers' concentrations were even higher when two of these diseases, instead of one alone, co-occurred in the same individual.This finding points to the need to further investigate the impact of co-occurring chronic diseases on AD pathology.
In our study, cardiovascular risk factors, apart from diabetes, were associated with lower concentrations of several biomarkers.Previous studies reported an inverse association between BMI and p-tau181, p-tau217, and NfL, 9,10,14,42 which is hypothesized to be due to higher blood volume and fat mass, leading to a different distribution of the biomarkers in obese individuals. 7,42On the other hand, our findings on the relation between hypertension and AD biomarkers are less clear, as in another population-based study high blood pressure was associated with higher levels of t-tau and amyloid markers. 10The relationship between blood pressure and brain aging is intricate; in fact, hypertension -mid-life hypertension in particular-is well-known to contribute to AD pathology, but there is evidence that late-life hypotension is also a risk factor for AD pathology, 43,44 and some studies suggest that antihypertensive medications may reduce AD risk. 43Beyond the effects of blood pressure on neuropathology and dementia risk, the association between hypertension and biomarker levels may also reflect peripheral mechanisms (e.g., higher blood volume in individuals with hypertension).Future studies are needed to disentangle the complex link between blood pressure, blood biomarkers of AD, and dementia risk.
For the first time, our study highlighted the role of low-grade systemic inflammation in the association between chronic diseases and AD blood biomarkers concentration.Systemic inflammation is a driver of the aging process, 45 as well as both trigger and manifestation of several diseases, 16,46 and can be interpreted as a marker of higher clinical complexity in older adults.Also, the imbalance between proinflammatory and anti-inflammatory agents that occurs with aging (i.e., inflammaging) can ultimately contribute to neurodegeneration and dementia development. 47,48Our study revealed an association between systemic inflammation and elevated levels of NfL.In addition, we found an interplay between certain chronic diseases and inflammation, ultimately influencing blood biomarkers' concentration.
Specifically, systemic inflammation amplified the association of chronic kidney disease, cerebrovascular disease and anemia with elevated blood biomarkers of AD.Conversely, systemic inflammation did not modify the relationship between heart diseases and blood biomarkers of AD, suggesting that other factors may drive this association.
Taken as a whole, these findings highlight that many features need to be considered in the interpretation of these biomarkers; special attention must be given to the clinical complexity that some individuals might display.

Limitations and strengths
Some aspects concerning the study must be considered.First, its crosssectional design restricts our possibility to speculate on the temporal and causal relation between the examined factors and the levels of the biomarkers.Moreover, AD biomarkers were measured in serum, while plasma biomarkers are more widely used.The concentrations of the biomarkers tend to be lower in serum than in plasma, however, serum biomarkers of AD have been already used in other studies to predict dementia development 49,50 and have shown a high correlation with plasma biomarkers and a similar diagnostic accuracy. 1,51rther, SNAC-K includes older adults who live in a central district in Stockholm and who are generally wealthy, fit, and healthy.This might limit the generalizability of our findings to other populations.However, similar results were reported in other population-based studies conducted in different countries. 10,15Additionally, among SNAC-K participants, those with available AD blood biomarkers at baseline were younger and had fewer chronic diseases compared to those for whom blood biomarkers were not available.It is plausible to believe that this selection bias towards a relatively healthier sample may have resulted in an underestimation of the observed associations.Our study has some important strengths.We were able to measure five different blood biomarkers of AD and one marker of inflammation in a large and well-characterized population-based sample of more than 2000 dementia-free participants.Furthermore, SNAC-K participants undergo a comprehensive assessment that provides information on many variables of interest including demographics, APOE genotype, risk factors, and chronic diseases.

Conclusions
In the community, the concentration of blood biomarkers of AD varies in relation to multiple factors.Understanding these influences is crucial for their optimal interpretation.Future studies should delve deeper into the mechanisms underlying these associations and ascertain whether these factors can affect the diagnostic and prognostic performance of AD blood biomarkers.

18 2.2 Data collection
19 ongoing longitudinal populationbased study.At baseline (2001-2004), 3363 individuals aged 60 years or older from the Kungsholmen district of Stockholm were randomly enrolled (73.3% participation rate).Participants were recruited based on their age across 11 different age cohorts (i.e., age 60, 66, 72, 78, 81, 84, 87, 90, 93, 96, 99+) and were followed-up at intervals of 6 (<78 years old) or 3 (≥78 years old) years.From baseline SNAC-K participants, we excluded those with dementia (n = 240).Further, we excluded 10 participants missing information on dementia, 665 without available blood samples, and 82 missing at least one of the measured AD biomarkers, leaving a final analytical sample of 2366 participants.All phases of SNAC-K were approved by the ethical committee at Karolinska Institutet and the Regional Ethical Review Board in Stockholm, Sweden (ethical permit for baseline: KI 01-114).All participants provided written informed consent to participate in the study.The results of this study are reported in keeping with the STROBE recommendations.At baseline, participants underwent medical examination, nurse interview, and cognitive assessment conducted by trained staff.Data on demographics were obtained during nurse interviews.Educational level was categorized into elementary, high school, or university and above.Mini-Mental State Examination (MMSE) was used as a measure of global cognition.19 Baseline characteristics of the study population.