Skip to main content

Genetically predicted telomere length and Alzheimer’s disease endophenotypes: a Mendelian randomization study

Abstract

Telomere length (TL) is associated with biological aging, consequently influencing the risk of age-related diseases such as Alzheimer’s disease (AD). We aimed to evaluate the potential causal role of TL in AD endophenotypes (i.e., cognitive performance, N = 2233; brain age and AD-related signatures, N = 1134; and cerebrospinal fluid biomarkers (CSF) of AD and neurodegeneration, N = 304) through a Mendelian randomization (MR) analysis. Our analysis was conducted in the context of the ALFA (ALzheimer and FAmilies) study, a population of cognitively healthy individuals at risk of AD. A total of 20 single nucleotide polymorphisms associated with TL were used to determine the effect of TL on AD endophenotypes. Analyses were adjusted by age, sex, and years of education. Stratified analyses by APOE-ɛ4 status and polygenic risk score of AD were conducted. MR analysis revealed significant associations between genetically predicted longer TL and lower levels of CSF Aβ and higher levels of CSF NfL only in APOE-ɛ4 non-carriers. Moreover, inheriting longer TL was associated with greater cortical thickness in age and AD-related brain signatures and lower levels of CSF p-tau among individuals at a high genetic predisposition to AD. Further observational analyses are warranted to better understand these associations.

Graphical Abstract

Introduction

Telomere length (TL) is a candidate biomarker of biological aging and aging-related outcomes [1, 2]. Telomeres are tandem deoxyribonucleic acid (DNA) repeats of a short sequence unit located at the end of the eukaryotic linear chromosomes that undergo attrition every time a somatic cell divides [3]. Shorter telomeres are associated with mortality [4], increased rates of age-related diseases [5,6,7,8], progression to dementia [9], and higher overall risk of developing Alzheimer’s disease (AD) [10, 11].

Previous studies have reported associations between TL and cognitive performance [12] as well as cognitive decline in cognitively unimpaired individuals [13]. Furthermore, TL is thought to affect brain structure; shorter TL has been associated with cortical atrophy in healthy subjects [14]. However, both longer and shorter TL have been associated with lower hippocampal volumes in the general population [15,16,17,18,19]. To date, no studies have evaluated the impact of TL in cortical or hippocampus regions differentially affected by AD or aging-related processes. Moreover, the potential interplay between TL and AD pathology remains to be elucidated. TL has been shown to be independent of amyloid-β (Aβ) deposition in AD patients [20, 21]. In contrast, Mahoney et al. [22] described an interaction between longer TL, cognitive decline, and abnormal levels of cerebrospinal fluid (CSF) Aβ and tau in participants of the AD Neuroimaging Initiative.

TL and AD are affected by common lifestyle-related risk factors [23, 24]. Among other reasons, these common factors may be confounders and lead to reverse causation in observational studies, partially explaining the heterogeneity of results found in the literature. The Mendelian randomization (MR) approach was developed to overcome these limitations. Unlike observational studies, the MR approach does not explore the association directly but through genetic instrumental variables (e.g., single nucleotide polymorphisms, SNPs) whose estimation of TL is not influenced by confounding or reverse causation [25, 26]. Previous MR studies have described potential causal effects of genetically predicted TL on the risk of AD [27,28,29,30]. However, other MR studies showed inconsistent results regarding TL and cognitive traits [31, 32].

The aim of the present study was to evaluate whether TL may play a causal role in the underlying pathophysiological mechanisms of AD by influencing its endophenotypes measured as cognitive performance, brain structure, and CSF biomarkers of AD and neurodegeneration, in a sample of cognitively unimpaired individuals at increased risk of AD.

Methods

Study participants

The present is a cross-sectional analysis that was conducted in the context of an existing study, the ALFA (ALzheimer and FAmilies) study [33]. Briefly, the ALFA study (Clinicaltrials.gov Identifier: NCT01835717) includes a total of 2573 cognitively unimpaired participants, aged 45–74, many of them kindred of AD patients (86% had at least one parent with dementia regardless of age at onset, 48% of the participants had at least one parent diagnosed with AD before the age of 75).

Participants were characterized at their baseline visit in 2013–2014 at multiple levels (sociodemographic, anthropometric, clinical, epidemiological, cognitive, and neuroimaging). For a subset of the ALFA study participants, we additionally collected blood samples for further genetic analysis. Moreover, a subset of the ALFA study participants (ALFA+ study) underwent lumbar puncture for CSF biomarker analysis. ALFA exclusion criteria were [1] cognitive performance falling outside established cutoffs: Mini-Mental State Examination [34, 35] < 26, Memory Impairment Screen [36, 37] < 6, Time-Orientation subtest of the Barcelona Test II [38] < 68, semantic fluency [39, 40] (animals) < 12, and Clinical Dementia Rating scale [41] > 0; [2] major psychiatric disorders (DSM-IV-TR) or diseases that could affect cognitive abilities; [3] severe auditory and/or visual, neurodevelopmental, and/or psychomotor disorders, significant diseases that could interfere with cognition; [4] neurological disorders; [5] brain injury that could interfere with cognition; and [6] family history of AD with a suspected autosomal dominant pattern. Further details of ALFA participants can be found in Molinuevo et al. [33].

The study was conducted in accordance with the directives of the Spanish Law 14/2007, of 3rd of July, on Biomedical Research (Ley 14/2007 de Investigación Biomédica). All participants accepted the study procedures by signing an informed consent form.

Genetic data

Genotyping

DNA samples were obtained from whole blood samples by applying a salting-out protocol. DNA was eluted in 800μl of H2O (Milli-Q) and quantified using Quant-iTT PicoGreen® dsDNA Assay Kit (Life Technologies) at the Cancer Epigenetics and Biology Program (PEBC; IDIBELL, Barcelona, Spain). The DNA concentration for each sample was additionally normalized. Genome-wide genotyping was performed using the Illumina Infinim NeuroChip [42] backbone version 1.0 and 1.2, based on the genome-wide genotyping array (Infinium HumanCore-24 v1.0). PLINK was used for the quality control (QC) of the genetic data [43]. We applied the following sample QC thresholds: sample missingness rates > 2% and heterozygosity 3 standard deviations (SD). Individuals showing sex discordances and high percentage of shared alleles (i.e., identity-by-descent (IBD) > 0.185) [44] were also excluded. A total of 2280 individuals remain after the genetic QC procedure.

Imputation

Genetic imputation was done using the Michigan imputation server, according to established guidelines (https://imputationserver.sph.umich.edu) and the European ancestry haplotype reference panel (HRC r1.1 2016) [45, 46]. The genetic datasets were phased using EAGLE software v.2.4, and data were rechecked using the default QC in the Michigan imputation server. After imputation, 15,805,054 variants were obtained with an imputation quality of 90% (r-squared parameter > 0.1).

Genetic variants associated with longer telomere length

From the final genetic sample of the ALFA study, 20 SNPs independently associated with mean TL (p-val < 5 × 10−8) in a genome-wide meta-analysis of 78,592 individuals were selected [47]. Proxies for those SNPs were included through LDlink [48] when the genetic information was not available in our sample. Proxies were considered as those SNPs with r2 ≥ 0.9, except for the genetic variants rs55749605 (r2 = 0.84) and rs7705526 (r2 = 0.79). See the list of r2 in Additional file 1: Supplementary Table 1. Departures from the Hardy-Weinberg equilibrium and allele frequencies were inspected using the compareGroups R package [49]. Effect sizes (i.e., SD change in TL per copy of the effect allele) and standard errors (SE) were obtained from summary statistics from Li et al. [47]. Detailed information of those SNPs is presented in Additional file 1: Supplementary Table 1.

Apolipoprotein (APOE) genotyping

The APOE allelic variants were obtained from allelic combinations of the rs429358 and rs7412 polymorphisms [50]. According to the genotypes of these polymorphisms, subjects were classified depending on the status of the ε4 allele (non-carriers, carriers of at least one ε4 allele). For the purpose of the present study, APOE-ε2/ε4 individuals (N = 47) were excluded.

Polygenic risk score of Alzheimer’s disease

Genetic variants that passed the post-imputation QC were used to compute the polygenic risk score (PRS) for AD (PRS-AD) using PRSice version 2 [51]. PRSice combined AD GWAS hits (p-value < 5 × 10−8), including the APOE region, by summing up all the SNP alleles carried by the participants weighted by the SNP allele effect size estimated in a previous GWAS [52]. PRS-AD was Z-standardized. Individuals with a PRS-AD above percentile 75th were classified within the high genetic predisposition group, whereas those with lower PRS were classified within the low genetic predisposition group.

Cognitive performance evaluation

During a neuropsychological evaluation at baseline, all participants were administered a cognitive test battery for the detection of early decline in longitudinal follow-up. This battery assesses episodic verbal memory measured by means of the Memory Binding Test (MBT) [53, 54], as well as executive and reasoning functions assessed by the Wechsler Adult Intelligence Scale (WAIS)-IV including psychomotor speed, visual processing, executive function (EF), and non-verbal and verbal reasoning (coding, visual puzzles, digit span, matrix reasoning, and similarities) [55]. A composite to assess cognitive performance was created based on the Preclinical Alzheimer Cognitive Composite (PACC) [56]. For our study, a modified-PACC (mPACC) composite was created by averaging the Z-scores of the following variables: MBT immediate total paired recall, MBT delayed free recall, WAIS-IV coding, and semantic fluency. Moreover, two additional cognitive composites to assess global episodic memory (EM) and EF were calculated by creating Z-scores for the cognitive measures from MBT and from WAIS-IV subtests, respectively. These global measures were calculated by averaging normalized raw scores of all subtests in each domain [57]. As with the individual cognitive tests, higher scores in the different composites represent better cognitive performance, while lower scores correspond to worse cognitive performance. EM, EF, and mPACC were used as the main outcomes to test the cross-sectional association between genetically predicted TL and cognitive performance. Data on cognitive performance was available for a total of 2233 participants.

Neuroimaging features: image acquisition and signature calculation

The acquisition of neuroimaging data was performed for a subset of our participants through magnetic resonance imaging (MRI). MRI scans were obtained with a 3-T scanner (Ingenia CX, Philips, Amsterdam, Netherlands). The MRI protocol was identical for all participants and included a high-resolution 3D T1-weighted turbo field echo (TFE) sequence (voxel size 0.75 × 0.75 × 0.75 mm, TR/TE: 9.90/4.6 ms, flip angle = 8°). Structural T1-weighted images were segmented using FreeSurfer version 6.0 [58]. The average of the cortical thickness between hemispheres of specific brain regions was used to calculate the AD and aging brain signatures. AD brain signature was calculated as the average of the cortical thickness of specific regions known to be affected in AD (i.e., AD vulnerable brain regions): medial temporal, inferior temporal, temporal pole, superior parietal, precuneus, angular, supramarginal, superior frontal, and middle frontal [59]. Aging brain signature was calculated as the average of the cortical thickness of specific brain regions known to be affected by aging: calcarine, caudal insula, cuneus, caudal fusiform, dorsomedial frontal, lateral occipital, precentral, and inferior frontal [60]. We used AD and aging brain signatures as the main outcomes to test the association between genetically predicted TL and brain structure. Higher values in these signatures represent a thicker cortex of the areas included in the signature. AD and aging brain signatures values were available for a total of 1134 participants.

CSF biomarkers: collection and measurement

CSF biomarkers were acquired from lumbar punctures for a subset of participants. In brief, CSF total-tau (t-tau) and phosphorylated-tau (p-tau) were measured using the electrochemiluminescence immunoassays Elecsys® phospho-tau(181P) CSF and total-tau CSF, respectively, on a fully automated cobas e 601 instrument (Roche Diagnostics International Ltd.). The rest of the CSF biomarkers Aβ42, Aβ40 and neurofilament light (NfL), were measured with in vitro diagnostic assays and robust prototype assays as part of the NeuroToolKit (Roche Diagnostics International Ltd, Rotkreuz, Switzerland) on both cobas e 601 and e 411 instruments. All measurements were performed at the Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden.

The CSF Aβ42/40 ratio was used as a biomarker of Aβ pathology [61]. The aforementioned biomarkers were used as the main outcomes to test the cross-sectional association between genetically predicted TL and core AD and neurodegeneration biomarkers in a total of 304 participants.

Statistical analysis

Alleles were harmonized to take as reference the effect allele associated with longer TL. In this study, genetic associations with longer TL were obtained from the summary statistics of a previously computed meta-GWAS [47]. Furthermore, genetic associations with the outcomes (i.e., AD endophenotypes) were assessed in our sample by linear regression models including age, sex, years of education, and APOE-ε4 status as covariates (see Additional file 2: Supplementary Tables 1-15; Fig. 1). We used inverse-variance weighted (IVW) [62], maximum likelihood [63], weighted median [64], and weighted mode [65] methods with summarized data to estimate the causal effect of genetically predicted longer TL on the outcomes of the study in a two-sample MR design.

Fig. 1
figure 1

Workflow diagram of the Mendelian randomization method. (1) Selection of single nucleotide polymorphisms (SNPs) associated with telomere length (TL). (a) TL is measured in peripheral blood leukocytes through qPCR. (b) Genome-wide association analysis (GWAS) is performed to identify genome-wide hits (Z) associated with mean TL (X) through linear regression models. (2) Mendelian randomization procedure is performed using summary statistics from the GWAS and individual-level data including genotypic information and the outcomes of interest (Y). \({\beta}_{X_j}\), SNP association with TL from the GWAS summary statistics; \({\beta}_{Y_j}\), SNP association with each outcome is calculated through linear regression models using individual-level data for genotypes and outcomes; \({\beta}_{X{Y}_j}\), causal effect of each SNP on one given outcome; βIVW, inverse-variance weighted (IVW) method is applied to obtain a global estimate representing the causal effect of genetically predicted TL on one given outcome while avoiding confounding (C)

Leave-one-SNP-out and Cochran Q statistic [66] were used as ad hoc sensitivity analyses for evaluating the robustness of significant results. MR-Pleiotropy RESidual Sum and Outlier (MR-PRESSO) was also used to identify horizontal pleiotropic outliers [67]. MR-Egger regression intercept test was used as ad hoc sensitivity analysis for evaluating directional pleiotropy [68]. Effect sizes were reported in the SD change (βIVW) per copy of the allele associated with longer TL. Causal associations with p-values below 0.05 were considered to be nominal significant. False discovery rate (FDR) corrections for multiple comparisons were performed at a significant level of 0.05 [69, 70]. p-value correction was performed within AD endophenotypes and separately in the whole sample and stratified analyses. Models were assessed in the entire sample, as well as stratified by APOE-ε4 status and PRS-AD. All analyses were conducted under R software, version 4.1.0 [71]. MR analyses were performed using the MendelianRandomization package [72]. Description of study participants was performed using the compareGroups R package [49].

Results

Descriptive of study participants

Characteristics of ALFA study participants included in our study can be found in Additional file 3: Supplementary Tables 1-3. The final sample size was determined by the number of participants with available genotype information (N = 2233) and the number of participants with available information for each one of the main outcomes included in the study, that is, cognitive performance (N = 2233), neuroimaging signatures (N = 1134), and CSF biomarkers (N = 304).

Mendelian randomization results

Cognition outcomes

No statistically significant associations were identified between genetically predicted longer TL and any of the cognitive outcomes in the entire sample, nor in stratified analyses by APOE-ɛ4 status or by the genetic predisposition to AD based on high/low categories of PRS-AD (Additional file 4: Supplementary Tables 1-5).

Neuroimaging outcomes

IVW estimates revealed statistically significant associations between genetically predicted longer TL and greater cortical thickness for both AD and aging signatures only in individuals at high genetic predisposition to AD [AD signature: βIVW = 0.064, SE = 0.029, q-value = 0.032; aging signature: βIVW = 0.062, SE = 0.029, q-value = 0.032] (Fig. 2A, B; Table 1). Maximum likelihood, weighted median, and weighted mode methods suggested similar patterns of effect on cortical thickness. The MR-Egger intercept test did not provide evidence for directional pleiotropy [AD signature: p-value = 0.998; aging signature: p-value = 0.939]. Furthermore, according to both Cochran Q and MR-PRESSO, there was no evidence of increased heterogeneity (see Table 1). IVW leave-one-out analyses showed MR estimates did not fluctuate after sequentially removing each of the 20 SNPs used as instrumental variables (Additional file 5: Supplementary Figs. 1-2). No other statistically significant associations were found on individuals at low genetic predisposition to AD, as well as in the entire sample, or in stratified analyses by APOE-ɛ4 status (Additional file 4: Supplementary Tables 1-5).

Fig. 2
figure 2

Scatter plot representing the significant causal effects of genetically predicted longer telomere length (TL) on neuroimaging outcomes among individuals at high genetic predisposition to AD. A Causal effect of genetically predicted TL on AD signature. B Causal effect of genetically predicted TL on aging signature. Each dot represents the genetic association of one given SNP with TL in the X-axis and the genetic association of this SNP with either AD or aging signature in the Y-axis. The slope of the line represents the inverse-variance weighted causal estimate

Table 1 Statistically significant results reflecting the effect of genetically predicted longer telomere length on neuroimaging outcomes

Core AD and neurodegeneration CSF biomarkers

Statistically significant associations were found between genetically predicted longer TL and lower levels of CSF Aβ ratio [βIVW = −0.007, SE = 0.002, q-value = 0.001] as well as higher levels of CSF NfL [βIVW = 13.267, SE = 2.604, p-value < 0.0001] only in APOE-ɛ4 non-carriers (Fig. 3A, B; Table 2). Robust MR methods suggested significantly analogous patterns of effect on both CSF Aβ ratio and NfL. There was no evidence of unbalanced directional pleiotropy according to the MR-Egger intercept test [CSF Aβ ratio: p-value = 0.170 ; CSF NfL: p-value = 0.742]. Cochran Q and MR-PRESSO also indicated the paucity of heterogeneity in MR estimates (see Table 2). Nevertheless, the leave-one-out analysis showed IVW estimates lost significance when excluding rs35640778 from both MR analyses (Additional file 5: Supplementary Figs. 3-4). Among APOE-ɛ4 carriers, MR-PRESSO indicated the presence of increased heterogeneity due to the inclusion of the outlier rs228595 in the analyses (Additional file 4: Supplementary Table 2). However, the MR estimate remained null after correction of pleiotropy via outlier removal.

Fig. 3
figure 3

Scatter plot representing the significant causal effects of genetically predicted longer telomere length (TL) on CSF biomarkers in APOE-ε4 non-carriers. A Causal effect of genetically predicted TL on Aβ ratio. B Causal effect of genetically predicted TL on CSF NfL. Each dot represents the genetic association of one given SNP with TL in the X-axis and the genetic association of this SNP with CSF Aβ ratio in the Y-axis. The slope of the line represents the inverse-variance weighted causal estimate

Table 2 Statistically significant results reflecting the effect of genetically predicted longer telomere length on CSF biomarkers

In stratified analysis by PRS-AD, a nominal statistically significant association was found between SNPs predicting longer TL and lower levels of CSF p-tau only in individuals classified as high genetic predisposition to AD [βIVW = −9.186, SE = 4.534, p-value = 0.043] (Fig. 4; Table 2). No violations of MR assumptions were found according to the MR-Egger intercept test, Cochran Q, and MR-PRESSO. Furthermore, MR showed similar estimates when sequentially leaving out genetic variants from the analyses (Additional file 5: Supplementary Fig. 5). No other statistically significant associations were identified in the entire sample in relation to genetically predicted TL and CSF biomarkers.

Fig. 4
figure 4

Scatter plot representing the significant causal effect of genetically predicted longer TL on CSF p-tau among individuals at high genetic predisposition to AD. Each dot represents the genetic association of one given SNP with TL in the X-axis and the genetic association of this SNP with an aging signature in the Y-axis. The slope of the line represents the inverse-variance weighted causal estimate

Discussion

In this study, we explored the potential causal role of genetically predicted longer TL on AD endophenotypes measured as cognitive performance, brain structure (i.e., AD signature and aging signature), and CSF biomarkers of AD and neurodegeneration, in cognitively unimpaired individuals from the ALFA study. We performed stratified analysis by APOE-ɛ4 status and the estimated genetic predisposition of developing AD (i.e., PRS-AD). We observed statistically significant associations between genetically predicted longer TL, lower levels of CSF Aβ ratio, and higher levels of CSF NfL only in APOE-ɛ4 non-carriers. Additionally, stratified analysis by the PRS-AD revealed statistically significant associations between SNPs predicting longer TL, greater cortical thickness in AD and aging signatures and lower levels of p-tau only among individuals at high genetic predisposition of developing the disease.

Genetically predicted TL and cognitive performance

We did not observe any statistically significant association between genetically predicted longer TL and cognition outcomes. Previous literature showed inconsistent findings regarding TL effect on cognitive performance. Similar to our approach, Hägg et al. [31] reported an association between genetically predicted longer TL and better scores on EF, but not for global cognition or EM, in a large cohort of cognitively healthy individuals (N= 17, 052). Most recently, Demanelis et al. [32] found genetically predicted TL to be associated with several aging-related traits but not cognitive function among participants of the UK Biobank (N = 337, 522).

Regarding observational studies, Yaffe et al. [73] reported an association between longer TL and better performance at baseline on the EF domain among community-dwelling elders. In contrast, longer TL at baseline has also been associated with a decline in EF only among cognitively healthy APOE-ε4 carriers [22]. Albeit, Takata et al. [74] did not find any association between TL and cognitive decline. In the Rotterdam study (N = 1961) a U-shaped association between TL and risk of AD was found, meaning that both, extremely shorter TL as well as extremely longer TL, were associated with an increased risk of AD, with stronger associations seen for APOE-ε4 carriers [75]. Alternatively, shorter TL has been shown to predict AD incidence only among APOE-ε4 non-carriers [76].

In the present study, the lack of statistically significant associations may be due to its cross-sectional nature as well as the idiosyncrasies of the ALFA cohort, which is composed of middle-aged (i.e., mean age of 56.3 years) and cognitively unimpaired individuals.

Genetically predicted TL and brain structure

Our results showed a significant association between genetically predicted longer TL and greater cortical thickness in both the AD and aging brain signatures, only among individuals at high genetic predisposition of AD. To our knowledge, no previous MR analyses have evaluated the causal effect of TL on brain structure, nor in regions primarily affected by AD or normal aging-related atrophy. Longer TL has been previously associated with larger hippocampal volumes in non-demented individuals in observational cross-sectional studies [15, 17]. Furthermore, a population-based study (i.e., Dallas Heart Study) found TL to be associated with total cerebral volume and specific subsegmental regions of the cortex [14].

Mechanisms explaining how TL could affect brain structure during the life-course are not completely understood. TL is known to play a crucial role in brain development with implications in embryonic and adult neurogenesis [77]. However, there is still a scientific debate over the existence of telomere attrition in neurons due to its post-mitotic nature. Even so, telomerase reactivation (i.e., the enzyme that maintains TL in the cell) reversed brain tissue degeneration among aged telomerase-deficient mice [78]. Therefore, inherited longer TL might protect brain structure through multiple mechanisms, either in regions primarily affected by AD-related processes or aging itself.

Stratified analyses did not show significant differences in the association between genetically predicted TL and brain structure in relation to APOE-ɛ4, nor PRS-AD subgroups. Similarly, King et al. [14] described that associations between TL and cortical volumes remained significant after adjusting by APOE-ε4 status.

Genetically predicted TL and biomarkers of AD and neurodegeneration

Our results showed significant associations between SNPs predicting longer TL, lower levels of CSF Aβ ratio, and higher levels of CSF NfL in APOE-ɛ4 non-carriers. These results suggested an unexpected adverse effect of longer TL on CSF biomarkers related to both core AD pathology and neurodegeneration. However, stratified analysis by PRS-AD suggested a protective effect of inherited longer TL on core AD pathology, with longer TL nominally significant associated with lower levels of p-tau among individuals at high genetic predisposition of AD.

To our knowledge, no previous MR studies have tested the effect of genetically predicted TL on AD biomarkers and observational studies are scarce. TL has been previously associated with AD pathology with implications on cognitive function in participants of the AD Neuroimaging Initiative [22]. Moreover, Flanary et al. [79] showed that senescence of isolated human microglia related to telomere shortening enhanced Aβ accumulation and induced AD progression. However, underlying biological mechanisms explaining the association are not fully understood yet.

Given the complexity of microglia homeostasis [80], we might speculate that unexplored roles of TL on microglia activation could lead to both beneficial or detrimental effects on AD pathology and neurodegeneration processes, which would partially explain the aforementioned disparity of results. However, it remains unclear how APOE-ɛ4 carriership and the genetic predisposition of AD would affect this putative association. Interestingly, some of the genetic variants which were used as instrumental variables to estimate genetically predicted telomere length in the present study were primarily related to the risk of cardiovascular disease [81], and they were later found to be causally associated with AD risk [27]. Cardiovascular risk factors comprise important targets for preventing AD especially during midlife [24]. Indeed, midlife vascular risk factors have been previously associated with elevated amyloid burden [82], tau pathology, and cognition [83], supporting their role on the etiology of AD. Thus, the TL effect on cardiovascular health outcome could be mediating or contributing to the causal associations described in our cohort. Nevertheless, further studies with larger sample sizes are required to better understand these results since a variety of undisclosed pathological mechanisms could be responsible for this complex interplay.

Limitations and strengths

Our study is not free of limitations. Potential limitations could be due to MR assumptions. For instance, IVW estimates could be biased in the presence of pleiotropy. To ensure estimates were reliable, sensitivity analyses through the MR-Egger method were performed. Nevertheless, we cannot completely ensure whether other unexplored factors are confounding these effects [62]. In addition, our sample size is small when compared to other MR studies, which can result in a relative loss of significance when exploring multiple MR methods with different assumptions. However, we ensured the inclusion of strong instruments by applying a two-sample MR approach in which the estimates of the genetic associations with the exposure of interest (i.e., longer TL) were calculated in a previous genome-wide meta-analysis of 78,592 individuals.

It is noteworthy that our cohort is rather selected and composed of middle-aged cognitively unimpaired individuals. This could entail a recruitment bias, particularly in individuals at high genetic predisposition of AD. Individuals at high genetic predisposition might have already shown signs of cognitive decline at recruitment, which was an exclusion criterion of the present study. Furthermore, this also implies that our participants showed a low prevalence of other common comorbidities and that our results might not be generalizable to individuals at more advanced stages in the AD continuum. Indeed, given its cross-sectional nature, endophenotypes included in the present study might reflect a transient stage within the AD pathological process. Finally, those associations that did not survive FDR multiple comparisons correction should be interpreted with caution.

Despite the aforementioned limitations, the strengths of our study include the robust and well-characterized cohort of cognitively unimpaired middle-aged individuals. This included the extensive characterization of cognitive outcomes, the use of high-resolution brain scans, the quantification of AD biomarkers and neurodegeneration biomarkers in CSF, and the genome-wide genotyping. Another important strength is also the higher proportion of APOE-ε4 carriers in the ALFA participants compared to the general population (19% vs. 14% in the general population; p < 0.001) [84] and the available PRS of AD. Although carriership of the APOE-ɛ4 allele is the largest genetic risk factor of AD, PRS of AD has been also associated with cognitive decline [85], hippocampal function [86], and core AD CSF biomarkers [87]. In this sense, this study also has evidenced the benefits of using PRS-AD to characterize individuals at high genetic predisposition of AD (besides APOE) in population-based cohort studies and elucidate specific biological patterns associated with telomere length homeostasis.

To our knowledge, no prior MR studies have explored the potential causal association between TL and these AD endophenotypes (i.e., cognitive performance, AD and aging brain signatures based on brain structure and biomarkers of AD, and neurodegeneration in CSF) in a similar cohort of cognitively unimpaired individuals at increased risk of developing AD. Thus, our findings should be replicated in larger cohorts including AD patients at more advanced stages of the disease. Similarly, the follow-up of ALFA participants and further observational analyses are required to further understand such observations.

Conclusions

In conclusion, the results of this cross-sectional study show that genetically predicted longer TL may protect brain structure in cognitively unimpaired individuals. In addition, APOE alleles and the genetic predisposition to AD could modulate TL effects on CSF biomarkers of AD and neurodegeneration. This result opens new avenues to study the role of telomeres on AD progression.

Availability of data and materials

The data that support the findings of this study are available from the corresponding authors (NVT, MCB), upon reasonable request.

Abbreviations

Aβ:

Amyloid-β

AD:

Alzheimer’s disease

APOE:

Apolipoprotein E

CSF:

Cerebrospinal fluid

DNA:

Deoxyribonucleic acid

EF:

Executive function

EM:

Episodic memory

GWAS:

Genome-wide association study

IVW:

Inverse-variance weighted

MBT:

Memory Binding Test

MCI:

Mild cognitive impairment

MR:

Mendelian randomization

MRI:

Magnetic resonance imaging

NfL:

Neurofilament light

PACC:

Preclinical Alzheimer Cognitive Composite

PRESSO:

Pleiotropy RESidual Sum and Outlier

PRS:

Polygenic risk score

QC:

Quality control

SE:

Standard error

SD:

Standard deviation

SNP:

Single nucleotide polymorphism

TL:

Telomere length

WAIS:

Wechsler Adult Intelligence Scale

References

  1. López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Cell. 2013;153(6):1194–217 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0092867413006454.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Demanelis K, Jasmine F, Chen LS, Chernoff M, Tong L, Delgado D, et al. Determinants of telomere length across human tissues. Science (80- ). 2020;369(6509):eaaz6876 Available from: https://www.sciencemag.org/lookup/doi/10.1126/science.aaz6876.

    Article  Google Scholar 

  3. Blackburn EH. Structure and function of telomeres. Nature. 1991;350(6319):569–73 Available from: http://www.nature.com/articles/350569a0.

    Article  CAS  PubMed  Google Scholar 

  4. Wang Q, Zhan Y, Pedersen NL, Fang F, Hägg S. Telomere length and all-cause mortality: a meta-analysis. Ageing Res Rev. 2018;48:11–20 Available from: https://linkinghub.elsevier.com/retrieve/pii/S1568163718301235.

    Article  CAS  PubMed  Google Scholar 

  5. Blasco MA. Telomeres and human disease: ageing, cancer and beyond. Nat Rev Genet. 2005;6(8):611–22 Available from: http://www.nature.com/articles/nrg1656.

    Article  CAS  PubMed  Google Scholar 

  6. Calado RT, Young NS. Telomere diseases. N Engl J Med. 2009;361(24):2353–65 Available from: http://www.nejm.org/doi/abs/10.1056/NEJMra0903373.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Haycock PC, Heydon EE, Kaptoge S, Butterworth AS, Thompson A, Willeit P. Leucocyte telomere length and risk of cardiovascular disease: systematic review and meta-analysis. BMJ. 2014;349:g4227. Available from: https://www.bmj.com/lookup/doi/10.1136/bmj.g4227.

  8. Blackburn EH, Epel ES, Lin J. Human telomere biology: a contributory and interactive factor in aging, disease risks, and protection. Science (80- ). 2015;350(6265):1193–8.

    Article  CAS  Google Scholar 

  9. Honig LS, Kang MS, Schupf N, Lee JH, Mayeux R. Association of shorter leukocyte telomere repeat length with dementia and mortality. Arch Neurol. 2012;69(10):1332–9.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Forero DA, González-Giraldo Y, López-Quintero C, Castro-Vega LJ, Barreto GE, Perry G. Meta-analysis of telomere length in Alzheimer’s disease. J Gerontol Ser A Biol Sci Med Sci. 2016;71(8):1069–73 Available from: https://academic.oup.com/biomedgerontology/article-lookup/doi/10.1093/gerona/glw053.

    Article  Google Scholar 

  11. Koh S-H, Choi SH, Jeong JH, Jang J-W, Park KW, Kim E-J, et al. Telomere shortening reflecting physical aging is associated with cognitive decline and dementia conversion in mild cognitive impairment due to Alzheimer’s disease. Aging (Albany NY). 2020;12(5):4407–23 Available from: https://www.aging-us.com/lookup/doi/10.18632/aging.102893.

    Article  CAS  Google Scholar 

  12. Valdes AM, Deary IJ, Gardner J, Kimura M, Lu X, Spector TD, et al. Leukocyte telomere length is associated with cognitive performance in healthy women. Neurobiol Aging. 2010;31(6):986–92 Available from: https://linkinghub.elsevier.com/retrieve/pii/S019745800800256X.

    Article  CAS  PubMed  Google Scholar 

  13. Martin-Ruiz C, Dickinson HO, Keys B, Rowan E, Kenny RA, Von Zglinicki T. Telomere length predicts poststroke mortality, dementia, and cognitive decline. Ann Neurol. 2006;60(2):174–80 Available from: http://doi.wiley.com/10.1002/ana.20869.

    Article  PubMed  Google Scholar 

  14. King KS, Kozlitina J, Rosenberg RN, Peshock RM, McColl RW, Garcia CK. Effect of leukocyte telomere length on total and regional brain volumes in a large population-based cohort. JAMA Neurol. 2014;71(10):1247 Available from: http://archneur.jamanetwork.com/article.aspx?doi=10.1001/jamaneurol.2014.1926.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Grodstein F, van Oijen M, Irizarry MC, Rosas HD, Hyman BT, Growdon JH, et al. Shorter telomeres may mark early risk of dementia: preliminary analysis of 62 participants from the Nurses’ Health Study. Khoury J El, editor. PLoS One. 2008;3(2):e1590 Available from: https://dx.plos.org/10.1371/journal.pone.0001590.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Wikgren M, Karlsson T, Lind J, Nilbrink T, Hultdin J, Sleegers K, et al. Longer leukocyte telomere length is associated with smaller hippocampal volume among non-demented APOE ε3/ε3 subjects. Uddin M, editor. PLoS One. 2012;7(4):e34292 Available from: https://dx.plos.org/10.1371/journal.pone.0034292.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Jacobs EG, Epel ES, Lin J, Blackburn EH, Rasgon NL. Relationship between leukocyte telomere length, telomerase activity, and hippocampal volume in early aging. JAMA Neurol. 2014;71(7):921 Available from: http://archneur.jamanetwork.com/article.aspx?doi=10.1001/jamaneurol.2014.870.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Nilsonne G, Tamm S, Månsson KNT, Åkerstedt T, Lekander M. Leukocyte telomere length and hippocampus volume: a meta-analysis. F1000Research. 2015;4:1073 Available from: https://f1000research.com/articles/4-1073/v1.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Staffaroni AM, Tosun D, Lin J, Elahi FM, Casaletto KB, Wynn MJ, et al. Telomere attrition is associated with declines in medial temporal lobe volume and white matter microstructure in functionally independent older adults. Neurobiol Aging. 2018;69:68–75 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0197458018301556.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Franco S, Blasco MA, Siedlak SL, Harris PLR, Moreira PI, Perry G, et al. Telomeres and telomerase in Alzheimer’s disease: epiphenomena or a new focus for therapeutic strategy? Alzheimers Dement. 2006;2(3):164–8 Available from: https://onlinelibrary.wiley.com/doi/10.1016/j.jalz.2006.03.001.

    Article  CAS  PubMed  Google Scholar 

  21. Movérare-Skrtic S, Johansson P, Mattsson N, Hansson O, Wallin A, Johansson JO, et al. Leukocyte telomere length (LTL) is reduced in stable mild cognitive impairment but low LTL is not associated with conversion to Alzheimer’s disease: a pilot study. Exp Gerontol. 2012;47(2):179–82.

    Article  PubMed  Google Scholar 

  22. Mahoney ER, Dumitrescu L, Seto M, Nudelman KNH, Buckley RF, Gifford KA, et al. Telomere length associations with cognition depend on Alzheimer’s disease biomarkers. Alzheimer’s Dement Transl Res Clin Interv. 2019;5(1):883–90 Available from: https://onlinelibrary.wiley.com/doi/10.1016/j.trci.2019.11.003.

    Article  Google Scholar 

  23. Lin J, Epel E, Blackburn E. Telomeres and lifestyle factors: roles in cellular aging. Mutat Res Mol Mech Mutagen. 2012;730(1–2):85–9 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0027510711002223.

    Article  CAS  Google Scholar 

  24. Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413–46.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Katan M. Apoupoprotein E isoforms, serum cholesterol, and cancer. Lancet. 1986;327(8479):507–8 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0140673686929727.

    Article  Google Scholar 

  26. Smith GD, Ebrahim S. Mendelian randomization: prospects, potentials, and limitations. Int J Epidemiol. 2004;33(1):30–42.

    Article  PubMed  Google Scholar 

  27. Zhan Y, Song C, Karlsson R, Tillander A, Reynolds CA, Pedersen NL, et al. Telomere length shortening and Alzheimer disease—a Mendelian randomization study. JAMA Neurol. 2015;72(10):1202 Available from: http://archneur.jamanetwork.com/article.aspx?doi=10.1001/jamaneurol.2015.1513.

    Article  PubMed  Google Scholar 

  28. Guo Y, Yu H. Leukocyte telomere length shortening and Alzheimer’s disease etiology. J Alzheimers Dis. 2019;69(3):881–5.

    Article  CAS  PubMed  Google Scholar 

  29. Scheller Madrid A, Rasmussen KL, Rode L, Frikke-Schmidt R, Nordestgaard BG, Bojesen SE. Observational and genetic studies of short telomeres and Alzheimer’s disease in 67,000 and 152,000 individuals: a Mendelian randomization study. Eur J Epidemiol. 2020;35(2):147–56 Available from: http://link.springer.com/10.1007/s10654-019-00563-w.

    Article  CAS  PubMed  Google Scholar 

  30. Yu G, Lu L, Ma Z, Wu S. Genetically predicted telomere length and its relationship with Alzheimer’s disease. Front Genet. 2021;12:595864. Available from: https://www.frontiersin.org/articles/10.3389/fgene.2021.595864/full.

  31. Hägg S, Zhan Y, Karlsson R, Gerritsen L, Ploner A, van der Lee SJ, et al. Short telomere length is associated with impaired cognitive performance in European ancestry cohorts. Transl Psychiatry. 2017;7:e1100. Available from: http://www.nature.com/articles/tp201773.

  32. Demanelis K, Tong L, Pierce BL. Genetically increased telomere length and aging-related traits in the U.K. Biobank. Le Couteur D, editor. J Gerontol Ser A. 2021;76(1):15–22 Available from: https://academic.oup.com/biomedgerontology/article/76/1/15/5585929.

    Article  CAS  Google Scholar 

  33. Molinuevo JL, Gramunt N, Gispert JD, Fauria K, Esteller M, Minguillon C, et al. The ALFA project: a research platform to identify early pathophysiological features of Alzheimer’s disease. Alzheimer’s Dement Transl Res Clin Interv. 2016;2(2):82–92 Available from: http://doi.wiley.com/10.1016/j.trci.2016.02.003.

    Article  Google Scholar 

  34. Folstein MF, Folstein SE, McHough PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189–98.

    Article  CAS  PubMed  Google Scholar 

  35. Blesa R, Pujol M, Aguilar M, Santacruz P, Bertran-Serra I, Hernández G, et al. Clinical validity of the “mini-mental state” for Spanish speaking communities. Neuropsychologia. 2001;39(11):1150–7.

    Article  CAS  PubMed  Google Scholar 

  36. Buschke H, Kuslansky G, Katz M, Stewart WF, Sliwinski MJ, Eckholdt HM, et al. Screening for dementia with the Memory Impairment Screen. Neurology. 1999;52(2):231 Available from: http://www.neurology.org/cgi/doi/10.1212/WNL.52.2.231.

    Article  CAS  PubMed  Google Scholar 

  37. Böhm P, Peña-Casanova J, Gramunt N, Manero RM, Terrón C, Quiñones-Ubeda S. Spanish version of the Memory Impairment Screen (MIS): normative data and discriminant validity. Neurologia. 2005;20(8):402–11 Available from: http://www.ncbi.nlm.nih.gov/pubmed/16217689.

    PubMed  Google Scholar 

  38. Quiñones-Ubeda S. Desenvolupament, normalització i validació de la versió estàndard de la segona versió del Test Barcelona. Barcelona: Ramon Llull University; 2009.

    Google Scholar 

  39. Ramier AM, Hécaen H. Respective rôles of frontal lesions and lesion lateralization in “verbal fluency” deficiencies. Rev Neurol (Paris). 1970;123(1):17–22 Available from: http://www.ncbi.nlm.nih.gov/pubmed/5516326.

    CAS  Google Scholar 

  40. Pena-Casanova J, Quinones-Ubeda S, Gramunt-Fombuena N, Quintana-Aparicio M, Aguilar M, Badenes D, et al. Spanish Multicenter Normative Studies (NEURONORMA Project): norms for verbal fluency tests. Arch Clin Neuropsychol. 2009;24(4):395–411 Available from: https://academic.oup.com/acn/article-lookup/doi/10.1093/arclin/acp042.

    Article  PubMed  Google Scholar 

  41. Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology. 1993;43(11):2412 Available from: http://www.neurology.org/cgi/doi/10.1212/WNL.43.11.2412-a.

    Article  CAS  PubMed  Google Scholar 

  42. Blauwendraat C, Faghri F, Pihlstrom L, Geiger JT, Elbaz A, Lesage S, et al. NeuroChip, an updated version of the NeuroX genotyping platform to rapidly screen for variants associated with neurological diseases. Neurobiol Aging. 2017;57:247.e9–247.e13 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0197458017301641.

    Article  CAS  Google Scholar 

  43. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–75 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0002929707613524.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Anderson CA, Pettersson FH, Clarke GM, Cardon LR, Morris AP, Zondervan KT. Data quality control in genetic case-control association studies. Nat Protoc. 2010;5(9):1564–73 Available from: http://www.nature.com/articles/nprot.2010.116.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Das S, Forer L, Schönherr S, Sidore C, Locke AE, Kwong A, et al. Next-generation genotype imputation service and methods. Nat Genet. 2016;48(10):1284–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. the Haplotype Reference Consortium. A reference panel of 64,976 haplotypes for genotype imputation. Nat Genet. 2016;48(10):1279–83 Available from: http://www.nature.com/articles/ng.3643.

    Article  Google Scholar 

  47. Li C, Stoma S, Lotta LA, Warner S, Albrecht E, Allione A, et al. Genome-wide association analysis in humans links nucleotide metabolism to leukocyte telomere length. Am J Hum Genet. 2020;106(3):389–404 [cited 2022 Mar 3]. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0002929720300483.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Myers TA, Chanock SJ, Machiela MJ. LDlinkR: an R package for rapidly calculating linkage disequilibrium statistics in diverse populations. Front Genet. 2020;11:157. Available from: https://www.frontiersin.org/article/10.3389/fgene.2020.00157/full.

  49. Subirana I, Sanz H, Vila J. Building bivariate tables: the compareGroups package for R. J Stat Softw. 2014;57(12):1–16.

    Article  Google Scholar 

  50. Radmanesh F, Devan WJ, Anderson CD, Rosand J, Falcone GJ. Accuracy of imputation to infer unobserved APOE epsilon alleles in genome-wide genotyping data. Eur J Hum Genet. 2014;22(10):1239–42.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Choi SW, O’Reilly PF. PRSice-2: polygenic risk score software for biobank-scale data. Gigascience. 2019;8(7):giz082. Available from: https://academic.oup.com/gigascience/article/doi/10.1093/gigascience/giz082/5532407.

  52. Jansen IE, Savage JE, Watanabe K, Bryois J, Williams DM, Steinberg S, et al. Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimer’s disease risk. Nat Genet. 2019;51(3):404–13 Available from: http://www.nature.com/articles/s41588-018-0311-9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Buschke H. In: Nilsson LG, Ohta N, editors. The rationale of the memory binding test. New York: Taylor and Francis; 2013. p. 55–71. Available from: https://www.taylorfrancis.com/books/9781315851730.

    Google Scholar 

  54. Gramunt N, Buschke H, Sánchez-Benavides G, Lipton RB, Peña-Casanova J, Diéguez-Vide F, et al. Reference data of the Spanish Memory Binding Test in a midlife population from the ALFA STUDY (Alzheimer’s and Family). Garre-Olmo J, editor. J Alzheimers Dis. 2015;48(3):613–25 Available from: https://www.medra.org/servlet/aliasResolver?alias=iospress&doi=10.3233/JAD-150237.

    Article  PubMed  Google Scholar 

  55. Wechsler D. WAIS-IV. Escala de inteligencia de Wechsler para adultos-IV. Madrid: Pearson; 2012.

    Google Scholar 

  56. Donohue MC, Sperling RA, Salmon DP, Rentz DM, Raman R, Thomas RG, et al. The preclinical Alzheimer cognitive composite. JAMA Neurol. 2014;71(8):961 Available from: http://archneur.jamanetwork.com/article.aspx?doi=10.1001/jamaneurol.2014.803.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Brugulat-Serrat A, Salvadó G, Sudre CH, Grau O, Falcon C, Sánchez-Benavides G, et al. Regional distribution of white matter hyperintensity correlates with cognition in the ALFA cohort. Alzheimers Dement. 2018;14(7S_Part_17):P925 Available from: http://doi.wiley.com/10.1016/j.jalz.2018.06.1199.

    Article  Google Scholar 

  58. Fischl B. FreeSurfer. Neuroimage. 2012;62(2):774–81 Available from: https://linkinghub.elsevier.com/retrieve/pii/S1053811912000389.

    Article  PubMed  Google Scholar 

  59. Dickerson BC, Feczko E, Augustinack JC, Pacheco J, Morris JC, Fischl B, et al. Differential effects of aging and Alzheimer’s disease on medial temporal lobe cortical thickness and surface area. Neurobiol Aging. 2009;30(3):432–40 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0197458007002953.

    Article  PubMed  Google Scholar 

  60. Bakkour A, Morris JC, Wolk DA, Dickerson BC. The effects of aging and Alzheimer’s disease on cerebral cortical anatomy: specificity and differential relationships with cognition. Neuroimage. 2013;76:332–44 Available from: https://linkinghub.elsevier.com/retrieve/pii/S1053811913002140.

    Article  PubMed  Google Scholar 

  61. Milà-Alomà M, Suárez-Calvet M, Molinuevo JL. Latest advances in cerebrospinal fluid and blood biomarkers of Alzheimer’s disease. Ther Adv Neurol Disord. 2019;12:1–23.

    Article  Google Scholar 

  62. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–65 Available from: https://onlinelibrary.wiley.com/doi/10.1002/gepi.21758.

    Article  PubMed  PubMed Central  Google Scholar 

  63. Thompson JR, Minelli C, Abrams KR, Tobin MD, Riley RD. Meta-analysis of genetic studies using Mendelian randomization - a multivariate approach. Stat Med. 2005;24(14):2241–54.

    Article  PubMed  Google Scholar 

  64. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14 Available from: https://onlinelibrary.wiley.com/doi/10.1002/gepi.21965.

    Article  PubMed  PubMed Central  Google Scholar 

  65. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 2017;46(6):1985–98 Available from: https://academic.oup.com/ije/article/46/6/1985/3957932.

    Article  PubMed  PubMed Central  Google Scholar 

  66. Bowden J, Del Greco MF, Minelli C, Zhao Q, Lawlor DA, Sheehan NA, et al. Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol. 2019;48(3):728–42 Available from: https://academic.oup.com/ije/article/48/3/728/5251908.

    Article  PubMed  Google Scholar 

  67. Verbanck M, Chen C-Y, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8 Available from: http://www.nature.com/articles/s41588-018-0099-7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  68. Slob EAW, Burgess S. A comparison of robust Mendelian randomization methods using summary data. Genet Epidemiol. 2020;44(4):313–29 Available from: https://onlinelibrary.wiley.com/doi/10.1002/gepi.22295.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B. 1995;57(1):289–300 Available from: https://onlinelibrary.wiley.com/doi/10.1111/j.2517-6161.1995.tb02031.x.

    Google Scholar 

  70. Yekutieli D, Benjamini Y. Resampling-based false discovery rate controlling multiple test procedures for correlated test statistics. J Stat Plan Inference. 1999;82(1–2):171–96 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0378375899000415.

    Article  Google Scholar 

  71. R Core Team. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2022. Available from: https://www.r-project.org/.

  72. Yavorska OO, Burgess S. MendelianRandomization: an R package for performing Mendelian randomization analyses using summarized data. Int J Epidemiol. 2017;46(6):1734–9.

    Article  PubMed  PubMed Central  Google Scholar 

  73. Yaffe K, Lindquist K, Kluse M, Cawthon R, Harris T, Hsueh W-C, et al. Telomere length and cognitive function in community-dwelling elders: findings from the Health ABC Study. Neurobiol Aging. 2011;32(11):2055–60 Available from: https://linkinghub.elsevier.com/retrieve/pii/S0197458009003984.

    Article  CAS  PubMed  Google Scholar 

  74. Takata Y, Kikukawa M, Hanyu H, Koyama S, Shimizu S, Umahara T, et al. Association between ApoE phenotypes and telomere erosion in Alzheimer’s disease. J Gerontol Ser A. 2012;67A(4):330–5 Available from: https://academic.oup.com/biomedgerontology/article-lookup/doi/10.1093/gerona/glr185.

    Article  CAS  Google Scholar 

  75. Fani L, Hilal S, Sedaghat S, Broer L, Licher S, Arp PP, et al. Telomere length and the risk of Alzheimer’s disease: the Rotterdam study. J Alzheimers Dis. 2020;73(2):707–14 Available from: https://www.medra.org/servlet/aliasResolver?alias=iospress&doi=10.3233/JAD-190759.

    Article  PubMed  PubMed Central  Google Scholar 

  76. Hackenhaar FS, Josefsson M, Adolfsson AN, Landfors M, Kauppi K, Hultdin M, et al. Short leukocyte telomeres predict 25-year Alzheimer’s disease incidence in non-APOE ε4-carriers. Alzheimers Res Ther. 2021;13(1):130 Available from: https://alzres.biomedcentral.com/articles/10.1186/s13195-021-00871-y.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  77. Liu M-Y, Nemes A, Zhou Q-G. The emerging roles for telomerase in the central nervous system. Front Mol Neurosci. 2018;11:160. Available from: http://journal.frontiersin.org/article/10.3389/fnmol.2018.00160/full.

  78. Jaskelioff M, Muller FL, Paik J-H, Thomas E, Jiang S, Adams AC, et al. Telomerase reactivation reverses tissue degeneration in aged telomerase-deficient mice. Nature. 2011;469(7328):102–6 Available from: http://www.nature.com/articles/nature09603.

    Article  CAS  PubMed  Google Scholar 

  79. Flanary BE, Sammons NW, Nguyen C, Walker D, Streit WJ. Evidence that aging and amyloid promote microglial cell senescence. Rejuvenation Res. 2007;10(1):61–74 Available from: https://www.liebertpub.com/doi/10.1089/rej.2006.9096.

    Article  CAS  PubMed  Google Scholar 

  80. Hemonnot A-L, Hua J, Ulmann L, Hirbec H. Microglia in Alzheimer disease: well-known targets and new opportunities. Front Aging Neurosci. 2019:11:233. Available from: https://www.frontiersin.org/article/10.3389/fnagi.2019.00233/full.

  81. Codd V, Wang Q, Allara E, Musicha C, Kaptoge S, Stoma S, et al. Polygenic basis and biomedical consequences of telomere length variation. Nat Genet. 2021;53(10):1425–33.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  82. Gottesman RF, Schneider ALC, Zhou Y, Coresh J, Green E, Gupta N, et al. Association between midlife vascular risk factors and estimated brain amyloid deposition. JAMA. 2017;317(14):1443 Available from: http://jama.jamanetwork.com/article.aspx?doi=10.1001/jama.2017.3090.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  83. Bos I, Vos SJB, Schindler SE, Hassenstab J, Xiong C, Grant E, et al. Vascular risk factors are associated with longitudinal changes in cerebrospinal fluid tau markers and cognition in preclinical Alzheimer’s disease. Alzheimers Dement. 2019;15(9):1149–59 Available from: http://doi.wiley.com/10.1016/j.jalz.2019.04.015.

    Article  PubMed  PubMed Central  Google Scholar 

  84. Ward A, Crean S, Mercaldi CJ, Collins JM, Boyd D, Cook MN, et al. Prevalence of apolipoprotein E4 genotype and homozygotes (APOE e4/4) among patients diagnosed with Alzheimer’s disease: a systematic review and meta-analysis. Neuroepidemiology. 2012;38(1):1–17 Available from: https://www.karger.com/Article/FullText/334607.

    Article  PubMed  Google Scholar 

  85. Daunt P, Ballard CG, Creese B, Davidson G, Hardy J, Oshota O, et al. Polygenic risk scoring is an effective approach to predict those individuals most likely to decline cognitively due to Alzheimer’s disease. J Prev Alzheimers Dis. 2021;8(1):78–83 Available from: https://link.springer.com/article/10.14283/jpad.2020.64.

  86. Xiao E, Chen Q, Goldman AL, Tan HY, Healy K, Zoltick B, et al. Late-onset Alzheimer’s disease polygenic risk profile score predicts hippocampal function. Biol Psychiatry Cogn Neurosci Neuroimaging. 2017;2(8):673–9 Available from: https://linkinghub.elsevier.com/retrieve/pii/S2451902217301441.

    PubMed  Google Scholar 

  87. Zettergren A, Lord J, Ashton NJ, Benedet AL, Karikari TK, Lantero Rodriguez J, et al. Association between polygenic risk score of Alzheimer’s disease and plasma phosphorylated tau in individuals from the Alzheimer’s Disease Neuroimaging Initiative. Alzheimers Res Ther. 2021;13(1):17 Available from: https://alzres.biomedcentral.com/articles/10.1186/s13195-020-00754-8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

This publication is part of the ALFA (ALzheimer and FAmilies) study. The authors would like to express their sincerest gratitude to the ALFA project participants, without whom this research would have not been possible. We thank Roche Diagnostics International Ltd. for providing the kits to measure CSF biomarkers of ALFA+ study participants. We thank Kaj Blennow and Henrik Zetterberg from the Clinical Neurochemistry Laboratory at Sahlgrenska University Hospital (Mölndal, Sweden) for performing the CSF biomarkers measurements.

Collaborators of the ALFA study are Müge Akinci, Annabella Beteta, Anna Brugulat-Serrat, Raffaele Cacciaglia, Alba Cañas, Irene Cumplido, Carme Deulofeu, Ruth Dominguez, Maria Emilio, Carles Falcon, Karine Fauria, Sherezade Fuentes, Oriol Grau-Rivera, José M. González-de-Echávarri, Laura Hernandez, Patricia Genius, Gema Huesa, Jordi Huguet, Eva M Palacios, Paula Marne, Tania Menchón, Marta Milà-Alomà, Cleofé Peña-Gomez, Albina Polo, Sandra Pradas, Aleix Sala-Vila, Gonzalo Sánchez-Benavides, Gemma Salvadó, Mahnaz Shekari, Anna Soteras, Laura Stankeviciute, and Marc Vilanova.

The measuring range of the assay is 200 (lower technical limit)–1700 pg/mL (upper technical limit). The performance of the assay beyond the upper technical limit has not been formally established. Therefore, the use of values above the upper technical limit, which are provided based on an extrapolation of the calibration curve, is restricted to exploratory research purposes and is excluded for clinical decision-making or for the derivation of medical decision points.

COBAS, COBAS E, and ELECSYS are trademarks of Roche.

We would like to thank Prof. Dickerson Laboratory (http://www.dickersonlab.org/) at Massachusetts General Hospital and Harvard Medical School for sharing relevant information to compute the neuroimaging signatures used in this manuscript. Schematic representations were created with Biorender.com.

Funding

The project leading to these results has received funding from the Alzheimer’s Association (Grant AARG-19-618265). This project has received funding from Instituto de Salud Carlos III (PI19/00119). Additional support has been received from “la Caixa” Foundation (ID 100010434), under agreement LCF/PR/GN17/50300004, the Alzheimer’s Association and an international anonymous charity foundation through the TriBEKa Imaging Platform project (TriBEKa-17-519007), the Health Department of the Catalan Government (Health Research and Innovation Strategic Plan (PERIS) 2016-2020 grant# SLT002/16/00201), and the Universities and Research Secretariat, Ministry of Business and Knowledge of the Catalan Government under the grant no. 2017-SGR-892. All CRG authors acknowledge the support of the Spanish Ministry of Science, Innovation and Universities to the EMBL partnership, the Centro de Excelencia Severo Ochoa, and the CERCA Programme/Generalitat de Catalunya. NV-T is funded by a post-doctoral grant, Juan de la Cierva Incorporación Programme (IJC2020-043216-I), Ministry of Science and Innovation–Spanish State Research Agency. EMAU is supported by the Spanish Ministry of Science, Innovation and Universities - Spanish State Research Agency (RYC2018-026053-I). MS-C receives funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement no. 948677). MS-C also receives funding from the Instituto de Salud Carlos III (PI19/00155) and from the Spanish Ministry of Science, Innovation and Universities (Juan de la Cierva Programme grant IJC2018-037478-I). HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532); the European Research Council (#681712); Swedish State Support for Clinical Research (#ALFGBG-720931); the Alzheimer Drug Discovery Foundation (ADDF), USA (#201809-2016862); the AD Strategic Fund and the Alzheimer’s Association (#ADSF-21-831376-C, #ADSF-21-831381-C, and #ADSF-21-831377-C); the Olav Thon Foundation; the Erling-Persson Family Foundation, Stiftelsen för Gamla Tjänarinnor, Hjärnfonden, Sweden (#FO2019-0228); the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860197 (MIRIADE); and the UK Dementia Research Institute at UCL. KB is supported by the Swedish Research Council (#2017-00915); the Alzheimer Drug Discovery Foundation (ADDF), USA (#RDAPB-201809-2016615); the Swedish Alzheimer Foundation (#AF-742881); Hjärnfonden, Sweden (#FO2017-0243); the Swedish state under the agreement between the Swedish government and the County Councils, the ALF-agreement (#ALFGBG-715986); the European Union Joint Program for Neurodegenerative Disorders (JPND2019-466-236); and the National Institute of Health (NIH), USA, (grant #1R01AG068398-01). JDG is supported by the Spanish Ministry of Science and Innovation (RYC-2013-13054). AS-V. is the recipient of an Instituto de Salud Carlos III Miguel Servet II fellowship (CP II 17/00029).

Author information

Authors and Affiliations

Authors

Consortia

Contributions

B.R-F contributed to sample preparation and harmonization, analyzed the data, interpreted the data, and took the lead in writing the manuscript and generated data visualizations. N.V-T supervised the statistical analysis. G. S-B, M. S-C, G.O, C.M, K.F, G.K, I.S, M.C.dM, D.P, M.E, KB, H.Z, and I.dV contributed in the acquisition and preparation of the data, provided critical feedback, and helped shape the manuscript. E.M.A-U, J.L.M, A.N, J.D.G, and A.S-V provided critical feedback and helped shape the research, analysis, and manuscript. N.V-T and M.C-B conceived the presented idea, developed the theoretical formalism, designed the models, supervised the project and the versions of the manuscript, provided critical feedback, and contributed and supervised manuscript writing. All authors read and approved the final version of the manuscript. All authors have agreed both to be personally accountable for the author’s own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which the author was not personally involved, are appropriately investigated, resolved, and the resolution documented in the literature.

Corresponding authors

Correspondence to Natalia Vilor-Tejedor or Marta Crous-Bou.

Ethics declarations

Ethics approval

The study was conducted in accordance with the directives of the Spanish Law 14/2007, of 3rd of July, on Biomedical Research (Ley 14/2007 de Investigación Biomédica). The ALFA and ALFA+ study protocols have been approved by the Independent Ethics Committee “Parc de Salut Mar,” Barcelona and registered at ClinicalTrials.gov (ALFA Identifier: NCT01835717; ALFA+ Identifier: NCT02485730). All participants accepted the study procedures by signing an informed consent form.

Consent for publication

Not applicable.

Competing interests

GK is a full-time employee of Roche Diagnostics GmbH. IS is a full-time employee and shareholder of Roche Diagnostics International Ltd. HZ has served at scientific advisory boards for Denali, Roche Diagnostics, Wave, Samumed, and CogRx; has given lectures in symposia sponsored by Fujirebio, Alzecure, and Biogen; and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program. KB has served as a consultant, at advisory boards, or data monitoring committees for Abcam, Axon, Biogen, JOMDD/Shimadzu, Julius Clinical, Lilly, MagQu, Novartis, Roche Diagnostics, and Siemens Healthineers, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program. JLM is currently a full-time employee of Lundbeck and priorly has served as a consultant or at advisory boards for the following for-profit companies, or has given lectures in symposia sponsored by the following for-profit companies: Roche Diagnostics, Genentech, Novartis, Lundbeck, Oryzon, Biogen, Lilly, Janssen, Green Valley, MSD, Eisai, Alector, BioCross, GE Healthcare, and ProMIS Neurosciences. MSC has served as a consultant and at advisory boards for Roche Diagnostics International Ltd and has given lectures in symposia sponsored by Roche Diagnostics, S.L.U, and Roche Farma, S.A. The other co-authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplementary Table 1.

Characteristics of Single Nucleotide Polymorphisms (SNPs) associated with longer telomere length. The effect allele refers to the allele that is associated with longer telomere length. Chromosomal position of the SNPs (genome assembly GRCh37 (hg19)) according to the public archive for genetic variation within and across different species developed and hosted by the National Center for Biotechnology Information (NCBI) in collaboration with the National Human Genome Research Institute (NHGRI) (dbSNP).

Additional file 2: Supplementary Table 1.

Linear regression estimates for cognition outcomes in the entire sample. All models are adjusted for covariates: age, sex, education, and APOE status. Supplementary Table 2. Linear regression estimates for neuroimaging outcomes (i.e., Alzheimer’s disease and aging signatures) outcome in the entire sample. All models are adjusted for covariates: age, sex, education, and APOE status. Supplementary Table 3. Linear regression estimates for CSF biomarkers outcomes in the entire sample. All models are adjusted for covariates: age, sex, education, and APOE status. Supplementary Table 4. Linear regression estimates for cognition outcomes in APOE-ɛ4 carriers. All models are adjusted for covariates: age, sex, and education. Supplementary Table 5. Linear regression estimates for neuroimaging outcomes (i.e., Alzheimer’s disease and aging signatures) in APOE-ɛ4 carriers. All models are adjusted for covariates: age, sex, and education. Supplementary Table 6. Linear regression estimates for CSF biomarkers outcomes in APOE-ɛ4 carriers. All models are adjusted for covariates: age, sex, and education. Supplementary Table 7. Linear regression estimates for cognition outcomes in APOE-ɛ4 non-carriers. All models are adjusted for covariates: age, sex, and education. Supplementary Table 8. Linear regression estimates for neuroimaging outcomes (i.e., Alzheimer’s disease and aging signatures) in APOE-ɛ4 non-carriers. All models are adjusted for covariates: age, sex, and education. Supplementary Table 9. Linear regression estimates for CSF biomarkers outcomes in APOE-ɛ4 carriers. All models are adjusted for covariates: age, sex, and education. Supplementary Table 10. Linear regression estimates for cognition outcomes among individuals at high genetic predisposition to AD. All models are adjusted for covariates: age, sex, and education. Supplementary Table 11. Linear regression estimates for neuroimaging outcomes (i.e., Alzheimer’s disease and aging signatures) among individuals at high genetic predisposition to AD. All models are adjusted for covariates: age, sex, and education. Supplementary Table 12. Linear regression estimates for CSF biomarkers outcomes (i.e., Alzheimer’s disease and aging signatures) among individuals at high genetic predisposition to AD. All models are adjusted for covariates: age, sex, and education. Supplementary Table 13. Linear regression estimates for cognition outcomes among individuals at low genetic predisposition to AD. All models are adjusted for covariates: age, sex, and education. Supplementary Table 14. Linear regression estimates for neuroimaging outcomes (i.e., Alzheimer’s disease and aging signatures) among individuals at low genetic predisposition to AD. All models are adjusted for covariates: age, sex, and education. Supplementary Table 15. Linear regression estimates for CSF biomarkers outcomes among individuals at low genetic predisposition to AD. All models are adjusted for covariates: age, sex, and education.

Additional file 3: Supplementary Table 1.

Characteristics of the study participants with information for cognition outcomes. Mean and SD are shown for continuous variables. Supplementary Table 2. Characteristics of the study participants with information for neuroimaging outcomes. Mean and SD are shown for continuous variables. Supplementary Table 3. Characteristics of the study participants with information for CSF biomarkers. Mean and SD are shown for continuous variables.

Additional file 4: Supplementary Table 1.

Results of the effect of genetically predicted longer telomere length on AD endophenotypes in the entire sample. Supplementary Table 2. Results of the effect of genetically predicted longer telomere length on AD endophenotypes among APOE-ɛ4 carriers. Supplementary Table 3. Results of the effect of genetically predicted longer telomere length on AD endophenotypes among APOE-ɛ4 non-carriers. Supplementary Table 4. Results of the effect of genetically predicted longer telomere length on AD endophenotypes among individuals at high genetic predisposition to Alzheimer's disease. Supplementary Table 5. Results of the effect of genetically predicted longer telomere length on AD endophenotypes among individuals at low genetic predisposition to Alzheimer's disease.

Additional file 5: Supplementary Figure 1.

Leave-one-out permutation analysis plot for AD signature among individuals at high genetic predisposition to AD obtained by leaving out the SNP indicated and repeating the Inverse-Variance Weighted method with the rest of the instrumental variables. Supplementary Figure 2. Leave-one-out permutation analysis plot for Aging signature among individuals at high genetic predisposition to AD, obtained by leaving out the SNP indicated and repeating the Inverse-Variance Weighted method with the rest of the instrumental variables. Supplementary Figure 3. Leave-one-out permutation analysis plot for Aβ ratio among APOE-ɛ4 non-carriers obtained by leaving out the SNP indicated and repeating the Inverse-Variance Weighted method with the rest of the instrumental variables. Supplementary Figure 4. Leave-one-out permutation analysis plot for NfL among APOE-ɛ4 non-carriers obtained by leaving out the SNP indicated and repeating the Inverse-Variance Weighted method with the rest of the instrumental variables. Supplementary Figure 5. Leave-one-out permutation analysis plot for p-tau among individuals at high genetic predisposition to AD, obtained by leaving out the SNP indicated and repeating the Inverse-Variance Weighted method with the rest of the instrumental variables.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Rodríguez-Fernández, B., Vilor-Tejedor, N., Arenaza-Urquijo, E.M. et al. Genetically predicted telomere length and Alzheimer’s disease endophenotypes: a Mendelian randomization study. Alz Res Therapy 14, 167 (2022). https://doi.org/10.1186/s13195-022-01101-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13195-022-01101-9

Keywords