Life Family Study Research Articles

ALL-CAUSE MORTALITY AND CAUSE OF DEATH IN US LONG-LIVED SIBLINGS: DATA FROM THE LONG LIFE FAMILY STUDY

Abstract This study compared the mortality risk of long-lived siblings with the general population in the US and investigated the leading causes of death and the correspondence in age at death and causes of death between long-lived siblings. In the Long Life Family Study (LLFS), 1,264 siblings (Mean age 90.1, SD 6.4) were recruited from the US and followed over 12 years. Their survival function, based on the Kaplan-Meier estimator, was compared with an age-, race-, and sex-matched sample from US census data. To examine underlying and contributing causes, we examined in detail 338 deaths with complete death adjudication using death certificates, family interview, hospital and nursing home records at the Pittsburgh field center through the year 2018. Concordance between siblings was examined using correlation coefficients, Bland-Altman plots, percent agreement, and Kappa statistics. The LLFS proband siblings had better survival compared with US general population matched on birth cohort and age at enrollment. Age at death ranged from 75-104, mean 91.4 (SD 4.8). The leading causes of death were cardiovascular disease (33.1%), dementia (22.2%), and cancer (10.7%). In addition, we did not find significant correspondence in the age at death, the underlying cause of death, or conditions directly contributing to death between siblings. Our findings demonstrate further exceptional longevity in the LLFS proband generation compared to the general population, but do not show significant correspondence in age at death and causes of death between siblings.

TRANSCRIPTOMIC SIGNATURES FOR AGING AND EXTREME OLD AGE IDENTIFIED IN LONG LIFE FAMILY STUDY

Abstract Transcriptional studies of blood have previously identified changes in gene expression profiles associated with age. Centenarians, a rare population of individuals that reach 100 years of age, experience delays in aging-related diseases and mortality. Thus, the analysis of their blood transcriptome may help shed light on mechanisms contributing to healthy aging and extreme longevity, as well as enable the discovery of candidate molecular therapeutic targets. In this project, we analyzed RNAseq-based blood transcriptomic data and genetic data from 1,286 participants in the Long Life Family Study age ranged from 25 to 108. To identify transcripts with different abundances at different ages, we used linear mixed-effect models and included relevant covariates to control for genetic, socioeconomic, and technical confounders. We also estimated a genetic relationship matrix (GRM) from the genotype data to account for familial relatedness. We used the same models and covariates to compare transcripts’ abundance between the extreme old and younger participants to discover markers of extreme old age. Among the 11k RNA transcripts tested, we identified 5,249 aging markers (FDR ≤ 0.01) and 761 extreme old age markers (FDR ≤ 0.01). We successfully replicated our aging signature with previously published aging signature in another cohort with comparable Z-scores. Currently, further investigation is ongoing to distinguish markers that contribute to extreme longevity from markers that represent extreme old age. This study found large signatures of age and extreme old age and will serve as a valuable reference for future research to identify molecular mechanisms of healthy aging.

MULTI OMICS LANDSCAPE OF INSULIN SENSITIVITY IN HEALTHY OLDER PEOPLE FROM THE LONG LIFE FAMILY STUDY COHORT

Abstract Insulin sensitivity (HOMA-S), which measures the body’s response to insulin, is a crucial indicator of glucose homeostasis, particularly in healthy people. The present study utilizes MultiOmics profiling and integration to explore molecular processes crucial for maintaining healthy HOMA-S. To this end, 3783 non-diabetic individuals from 567 pedigrees in Long Life Family Study Cohort were selected (Average age = 70±16). Upon adjustment of HOMA-S for relevant covariates, linear mixed models were employed to analyze the genome sequence, blood transcriptome, and metabolome associations with HOMA-IS. Moreover, using STAAR and Pascal packages, rare and common variants were collapsed to genomic co-ordinates. The results highlight 10 genes, 2 soluble metabolites, and 2 lipids (phosphatidylcholine 35:1 and 40:6), significantly associated with HOMA-S. Integrative module enrichment analysis with PascalX package, identified multiple significant networks for HOMA-S. Notably, immune response in asthma is one of the processes with most genes (57%) displaying a positive correlation with HOMA-S (module p-val = 1.89 × 10E-6). The network contains 2 novel genes from transcriptome analysis, FCER1A and MS4A2 (p-vals = 1.32 × 10E-8 and 1.42 × 10E-6). Interestingly, one of the key metabolites, N-acetylglycine, also has a strong association with MS4A2 (p-val = 0.001034). Furthermore, SNP colocalization for gene expression, metabolome peak intensities, and HOMA-S, revealed multiple significant signals simultaneously associated with PC 35:1 and two members of this network, HLA-DQA1 and HLA-DQB1. The results of the study underscore the intricate interplay of genetic and metabolic factors for healthy HOMA-S with aging and their potential protective role for glycemic implications.

Examination of Genetic Risk Factors for Alzheimer’s Disease in a Cohort of Familial Longevity: the Long Life Family Study

AbstractBackgroundGenetic risk factors have been explored extensively for late onset Alzheimer’s Disease (LOAD). The largest meta‐analysis GWAS studies to date (Bellenguez et al., 2022) identified 83 single nucleotide polymorphisms (SNPs) associated with LOAD in the general population. The Long Life Family Study (LLFS) is aimed to identify factors associated with familial longevity and healthy aging by examining US and Danish families. Our earlier studies have shown that LLFS families are at lower risk of several age‐related diseases, including LOAD. Here, we will: (1) compare allele frequencies and effect sizes of the identified meta‐GWAS variants against those in LLFS; and (2) assess whether a polygenic risk score (PRS) based on meta‐GWAS variants differs between LOAD vs. cognitively healthy in LLFS.MethodWe examined LLFS participants who were Non‐Hispanic Whites, had available cognitive and whole genome sequencing (WGS) (n = 4,220). Allele frequencies were estimated based on founders only. Association analyses were performed via logistic regression using Wald test and adjusting for sex, age, site, education, first principal components, and the genetic relationship matrix. PRS was calculated using a weighted formula, and to account for family structure, means for each status per family were obtained.ResultIndependent of LOAD status, 13 SNPs had significantly different allele frequencies between the two studies. Minor allele frequencies of 10 of the SNPs were lower in LLFS when compared with Bellenguez et al. When examining the relations between LOAD and the 77 available SNPs, only five variants appeared to be associated in LLFS. The direction of association differed for two of the SNPs. Lastly, estimated PRS scores did not differ by LOAD status (affected vs unaffected: ‐0.059 vs. ‐0.054, respectively).ConclusionHere we show that variants associated with LOAD in the general population were not necessarily associated with LOAD in in the LLFS cohort, a cohort ascertained for familial healthy aging. When estimating disease risk using genetic markers, as expected, ascertainment needs to be taken into consideration.

95 Delving Beyond the Test Score: Linguistic Markers of Cognitive Impairment on Paragraph Recall

Objective:Cognitive tests requiring spoken responses, such as paragraph recall, are rich in cognitive-related information that is not captured using traditional scoring methods. This study aimed to determine if linguistic features embedded in spoken responses may differentiate between individuals who are and are not cognitively impaired.Participants and Methods:Participants in the Long Life Family Study completed a neuropsychological assessment which included the WMS-R Logical Memory I paragraph recall. For a subset of participants (N=709), test responses were digitally recorded and manually transcribed. We used Linguistic Inquiry Word Count, a text analysis program, to quantify word counts, grammatical features (e.g, prepositions, verb tenses), and the use of content words related to specific semantic categories (e.g., work-related, numbers) for immediate (IR) and delayed recall (DR). We used regression models with Generalized Estimating Equations adjusted by age, sex, education, and within-family correlation to select features associated with cognitive status (normal cognition [NC] versus cognitive impairment [CI]; Bonferroni-corrected threshold p<0.001). Next, we developed a “polyfeature score” (PFS) for both immediate and delayed recall, each calculated as a weighted sum of the selected linguistic features. We then built a logistic regression model to evaluate the predictive value of each PFS for identifying cognitively impaired individuals. In secondary analyses, we used regression models as above to identify features associated with mild cognitive impairment subtype (amnestic [aMCI] versus nonamnestic [naMCI]; threshold p< .05).Results:The sample included 599 participants with NC and 110 with CI (mean age = 72.3 ± 11.0 years, 54% female). The regression identified 8 linguistic features for IR and 7 for DR that significantly predicted cognitive status. Decreased use of content words related to work (e.g., employed, school, police) and biological processes (e.g., cook, cafeteria, eat) and the use of negations (e.g., no, not, can’t) were predictive of cognitive impairment in both recall conditions. In contrast, the use of other content word categories were predictive of cognitive status in only one recall condition (IR: leisure, cognitive processes, space; DR: drives, number). The use of fewer prepositions in IR, more first-person pronouns in DR, and fewer words in the past tense in DR were each associated with cognitive impairment. Word count was not predictive of cognitive status. Both PFSs were highly associated with cognitive status (PFS_IR ß= 0.74, p< 0.001; PFS_DR ß= 0.86, p= 0.001) with high discriminative value (PFS_IR AUC= 0.93, sensitivity = 0.81, specificity= 0.91; PFS_DR AUC= 0.95, sensitivity= 0.77, specificity= 0.88). In the CI subset, linguistic features differed between those classified as aMCI (n= 24) and naMCI (n= 40). Two function word categories predicted aMCI in IR whereas decreased word count, two function word categories, and two content word categories predicted aMCI in DR (all p< .05)Conclusions:Linguistic features from paragraph recall provide high predictive value for classifying cognitive status increasing its potential as a cognitive screener in clinical settings. Additionally, each recall condition identified unique linguistic features associated with cognitive impairment which may aid differentiation of cognitive impairment subtypes and elucidate processes underlying deficits in learning and recall.

The association between frailty and perceived physical and mental fatigability: The Long Life Family Study.

Higher levels of frailty, quantified by a frailty index (FI), may be linked to fatigue severity as tasks become more physically and mentally demanding. Fatigue, a component of frailty research, has been ambiguous and inconsistent in its operationalization. Fatigability-the quantification of vulnerability to fatigue in relation to specific intensity and duration of activities-offers a more sensitive and standardized approach, though the associationbetween frailty and fatigability has not been assessed. Using cross-sectional data from the Long Life Family Study at Visit 2 (2014-2017; N = 2524; mean age ± standard deviation (SD) 71.4 ± 11.2 years; 55% women; 99% White), we examined associations between an 83-item FI after excluding fatigue items (ratio of number of health problems reported (numerator) out of the total assessed (denominator); higher ratio = greater frailty) and perceived physical and mental fatigability using the Pittsburgh Fatigability Scale (PFS) (range 0-50; higher scores = greater fatigability). Participants had mean ± standard deviation FI (0.08 ± 0.06; observed range: 0.0-0.43), PFS Physical (13.7 ± 9.6; 39.5% more severe, ≥15), and PFS Mental (7.9 ± 8.9; 22.8% more severe, ≥13). The prevalence of more severe physical and mental fatigability was higher across FI quartiles. In mixed effects models accounting for family structure, a clinically meaningful 3%-higher FI was associated with 1.9 points higher PFS Physical score (95% confidence interval (CI) 1.7-2.1) and 1.7 points higher PFS Mental score (95% CI 1.5-1.9) after adjusting for covariates. Frailty was associated with perceived physical and mental fatigability severity. Understanding this association may support the development of interventions to mitigate the risks associated with greater frailty and perceived fatigability. Including measurements of perceived fatigability, in lieu of fatigue, in frailty indices has the potential to alleviate the inconsistencies and ambiguity surrounding the operationalization of fatigue and provide a more precise and sensitive measurement of frailty.

Interactions between genes involved in physiological dysregulation and axon guidance: role in Alzheimer's disease.

Dysregulation of physiological processes may contribute to Alzheimer's disease (AD) development. We previously found that an increase in the level of physiological dysregulation (PD) in the aging body is associated with declining resilience and robustness to major diseases. Also, our genome-wide association study found that genes associated with the age-related increase in PD frequently represented pathways implicated in axon guidance and synaptic function, which in turn were linked to AD and related traits (e.g., amyloid, tau, neurodegeneration) in the literature. Here, we tested the hypothesis that genes involved in PD and axon guidance/synapse function may jointly influence onset of AD. We assessed the impact of interactions between SNPs in such genes on AD onset in the Long Life Family Study and sought to replicate the findings in the Health and Retirement Study. We found significant interactions between SNPs in the UNC5C and CNTN6, and PLXNA4 and EPHB2 genes that influenced AD onset in both datasets. Associations with individual SNPs were not statistically significant. Our findings, thus, support a major role of genetic interactions in the heterogeneity of AD and suggest the joint contribution of genes involved in PD and axon guidance/synapse function (essential for the maintenance of complex neural networks) to AD development.

Signatures of Neuropsychological Test Results in the Long Life Family Study: A Cluster Analysis.

Discovering patterns of cognitive domains and characterizing how these patterns associate with other risk factors and biomarkers can improve our understanding of the determinants of cognitive aging. To discover patterns of cognitive domains using neuropsychological test results in Long Life Family Study (LLFS) and characterize how these patterns associate with aging markers. 5,086 LLFS participants were administered neuropsychological tests at enrollment. We performed a cluster analysis of six baseline neuropsychological test scores and tested the association between the identified clusters and various clinical variables, biomarkers, and polygenic risk scores using generalized estimating equations and the Chi-square test. We used Cox regression to correlate the clusters with the hazard of various medical events. We investigated whether the cluster information could enhance the prediction of cognitive decline using Bayesian beta regression. We identified 12 clusters with different cognitive signatures that represent profiles of performance across multiple neuropsychological tests. These signatures significantly correlated with 26 variables including polygenic risk scores, physical and pulmonary functions, and blood biomarkers and were associated with the hazard of mortality (p < 0.01), cardiovascular disease (p = 0.03), dementia (p = 0.01), and skin cancer (p = 0.03). The identified cognitive signatures capture multiple domains simultaneously and provide a holistic vision of cognitive function, showing that different patterns of cognitive function can coexist in aging individuals. Such patterns can be used for clinical intervention and primary care.

Rare genetic variants correlate with better processing speed

We conducted a genome-wide association study of Digit Symbol Substitution Test scores administered in 4207 family members of the Long Life Family Study (LLFS). Genotype data were imputed to the HRC panel of 64,940 haplotypes resulting in ∼15M genetic variants with a quality score > 0.7. The results were replicated using genetic data imputed to the 1000 Genomes phase 3 reference panel from 2 Danish twin cohorts: the study of Middle Aged Danish Twins and the Longitudinal Study of Aging Danish Twins. The genome-wide association study in LLFS discovered 18 rare genetic variants (minor allele frequency (MAF) < 1.0%) that reached genome-wide significance (p-value < 5 × 10−8). Among these, 17 rare variants in chromosome 3 had large protective effects on the processing speed, including rs7623455, rs9821776, rs9821587, rs78704059, which were replicated in the combined Danish twin cohort. These SNPs are located in/near 2 genes, THRB and RARB, that belonged to the thyroid hormone receptors family that may influence the speed of metabolism and cognitive aging. The gene-level tests in LLFS confirmed that these 2 genes are associated with processing speed.

GENOME-WIDE ASSOCIATION STUDY OF EXTREME LONGEVITY USING WHOLE-GENOME SEQUENCING DATA

Abstract Extreme longevity (EL) runs in families which supports the hypothesis that this is a genetically regulated trait. However, with the exception of APOE, genome-wide association studies (GWAS) of EL have not identified many genetic variants that replicate in independent studies. The majority of GWAS of EL have used imputed genotype data. Recently, the Long Life Family Study has generated the largest whole-genome sequencing data of centenarians and matched controls. We perform single-variant and gene-based tests of EL in these data using the nf-gwas-pipeline with the saddle point approximation adjustment of the p-values. These analyses suggest that, in addition to the APOE/TOMM40 region, some uncommon variants of GRM7 (chr3), AUTS2 (chr7), KIF13B (chr8), SLC2A14 (chr12), and ADCY9 (chr16) genes, and other intergenic SNPs in chromosomes 5, 10, and 20 may be implicated with EL.

NOVEL GENOMIC AND EXCEPTIONAL LONGEVITY FINDINGS FROM THE LONG LIFE FAMILY STUDY

Abstract The Long Life Family Study (LLFS), funded by the National Institute on Aging, is an international collaborative study of the genetics and familial components of exceptional longevity and healthy aging. We phenotyped 4,953 individuals from 539 two-generational families (1,727 proband; 3,226 offspring) at baseline (2006-2009). A second visit (2014-2017) was conducted for 2,904 (478 proband; 2,426 offspring) participants. The longitudinal, comprehensive in-person visits measured domains of healthy aging, including physical performance, cognition, and blood markers. Extensive genetic analyses were performed using the baseline blood draw, including genotyping with the Illumina 2.5M Human Omni array, linkage analyses with the families, whole genome sequencing using the TopMED protocol, and metabolomic assays. Collectively, this symposium will present novel findings that examined differences in end of life events and cause of death between non-exceptional and exceptional long-lived women, elucidate potential rare variants associated with exceptional longevity, and examine new potential metabolomics pathways involved in gait speed and cardiovascular disease. Specifically, Dr. Galvin will share results from three Danish nationwide studies (including LLFS) on different end of life events in long-lived female siblings. Then, Dr. Gurinovich will share findings on new uncommon variants associated with extreme longevity. Next, Dr. Kuipers will discuss associations between lipid metabolomics and vascular health. Lastly, Dr. Santanasto will discuss lipid metabolomics associated with lower odds of slow gait speed. Dr. Evan Hadley, NIA, will be the discussant and will share insights and propose future directions for LLFS.

ENERGY METABOLISM RELATED CANDIDATE GENE ASSOCIATION STUDY OF PERCEIVED PHYSICAL FATIGABILITY

Abstract Mitochondrial energy production decreases while fatigability increases with age. Genes associated with energy metabolism may contribute to fatigability. Using Long Life Family Study (LLFS), we initially assessed variants (SNPs) in 272 candidate autosomal genes involved in energy metabolism (previously associated with mitochondrial dysfunction disease) with perceived physical fatigability. Two generations of LLFS enrollees (N=2342, mean age=73.7, range 60-108 years) completed the Pittsburgh Fatigability Scale (PFS, 0-50, higher=greater fatigability) at Visit 2 (2014-2017). Physical fatigability prevalence was 42.1% (PFS≥15). Generalized linear mixed models assessed the association of each SNP with continuous PFS (GENESIS R package) adjusted for age, sex, field center, and family relatedness. We found no associations with perceived physical fatigability, all p&amp;gt;2.5E-7 (Bonferroni multiple comparison corrected p-value). Next steps will examine variants in the mitochondrial genome and BTBD3, another promising candidate gene recently discovered. Genetic biomarker(s) may identify individuals susceptible to greater fatigability to target for early intervention.

PERCEIVED AND PERFORMANCE FATIGABILITY: NOVEL FINDINGS ON CORRELATES, CONSEQUENCES, AND CANDIDATE GENES

Abstract Characterizing fatigability enables quantification of an individual’s susceptibility to fatigue in the context of standardized physical task(s). This approach eliminates self-pacing, and is a less-biased, more sensitive approach to assess the presence and severity of fatigue, a critical prodromal symptom of chronic and acute health conditions. Collectively, we have established the Pittsburgh Fatigability Scale (PFS) and Borg Rating of Perceived Exertion at the end of a standardized 5-minute treadmill walk (RPE Fatigability) as prognostic indicators of aging-related outcomes. Recently, we extended our methodological work in developing a novel accelerometry-based performance fatigability measure of slowing down during in-lab walking (Pittsburgh Performance Fatigability Index, PPFI). This symposium presents findings from the Study of Muscle, Mobility, and Aging (SOMMA), Long Life Family Study (LLFS), and Baltimore Longitudinal Study of Aging (BLSA). Specifically, using SOMMA, Ms. Qiao will share development and validation of the PPFI using a usual-paced 400m walk, Ms. Moffit will report the association of fitness (i.e., VO2peak) with PFS physical and mental fatigability, and extending the SOMMA findings, Dr. Moored will detail the dual roles of fitness and fatigability on life-space mobility. Using LLFS data, Ms. Gay will report associations of variants (SNPs) in 272 candidate autosomal genes involved in energy metabolism (previously associated with mitochondrial dysfunction disease) with PFS physical fatigability. Lastly, using BLSA data, Dr. Wanigatunga will compare the discriminatory power of RPE Fatigability versus general fatigue symptoms for predicting mortality. Our Discussant, Dr. Simonsick, will provide critical review in the context of new directions for fatigability research.

MEGA-ANALYSIS OF A CONSORTIUM OF FOUR CENTENARIAN STUDIES IDENTIFIES NOVEL EXTREME LONGEVITY VARIANTS

Abstract The meta-analysis of extreme longevity (EL) conducted in 2017 using a consortium of four longevity studies—the New England Centenarian Study (NECS), the Long Life Family Study (LLFS), the Southern Italian Centenarian Study (SICS), and the Longevity Gene Project (LGP)—confirmed previous associations with APOE alleles and identified additional candidate genes. The current study builds upon this work. Using an aggregated set of the four studies (mega-analysis), we sought to identify additional longevity variants. The current study includes 234 additional cases with a total of 2304 cases (median age=103, age range=[96, 119]), defined as living past the age at which less than 1% individuals from the 1900 - 1920 birth year cohorts survived and approximately 6,000 controls. Approximately 10 million genotyped and imputed SNPs that passed imputation quality score threshold of 0.7 and that had a minor allele count of 3 or greater were analyzed. We ran a mixed effects logistic model for EL adjusting for sex, top principal components, indicator for living in Denmark, indicator for living in Italy, and the genetic relationship matrix. The analysis identified 21 novel loci that are genome-wide significant, both common and rare, and these loci were not previously published. The genome-wide significant loci include genes such as ADGRL2 (p=6.46-15), HLA-DPA1 (p=1.06-8), GRK5 (p=2.01-8), TECTB (p=3.38-8), KCNQ1 (p=7.53-10), TEAD2 (p=9.53-10), and several intergenic regions. We are currently in the process of seeking replication of the top results in independent cohorts.

IMPACT OF COMBINATIONS OF THE APOE Ɛ4 ALLELE AND TOMM40-APOC1 VARIANTS ON SURVIVAL TO OLDER AGES AND ALZHEIMER’S RISK

Abstract Multifactorial diseases, health-related traits, and lifespan are polygenic phenotypes with complex genetic architectures. Polygenicity implies that multiple variants can impact the risks of these phenotypes independently or jointly. Recently, we showed that carriers of minor alleles of rs429358 (APOE ɛ4 encoding polymorphism), rs2075650 (TOMM40), and rs12721046 (APOC1) polymorphisms have up to 4.4 times higher risk of Alzheimer’s disease (AD) than APOE ɛ4 carriers without the minor alleles of rs2075650 and rs12721046. Here, we examined the chances of living to older ages—85 years and above—for carriers of compound genotypes combining these three polymorphisms using data from the Long Life Family Study, Framingham Heart Study, Cardiovascular Health Study, and UK Biobank. Consistent with their higher AD risk, carriers of the APOE ɛ4 allele with one or two minor alleles of rs2075650 and rs12721046 had 24.3% lower log odds of living to 85+ years (β=-0.243, p=2.22×10-2) than APOE ɛ4 carriers without either minor allele. Counterintuitively, AD did not mediate this risk. With AD-affected subjects excluded from the analysis, the effect size for the log odds of living to 85+ years (β=-0.352, p=2.35×10-3) was 1.45 times larger for APOE ɛ4 carriers with one or two minor alleles of rs2075650 and rs12721046. The chances of survival could be associated with lipid- and immunity-related mechanisms, whereas the risk of AD may be associated with amyloid-β-related mechanisms, among others. Targeting heterogeneous polygenic profiles of individuals at higher risks of multifactorial phenotypes may be a promising strategy for translating genetic discoveries to health care.

ETHNICITY-SPECIFIC EXTREME LONGEVITY VARIANTS IN A CONSORTIUM OF FOUR CENTENARIAN STUDIES

Abstract It has been shown that some longevity variants including the APOE variants have ethnicity-specific effects on EL within European ethnicities. The goal of the present study is to identify genetic variants whose effects vary by ethnicity by conducting a genome-wide association study of extreme longevity (EL: defined as living past the age at which less than 1% individuals from the 1900 - 1920 birth year cohorts survived) that includes the SNP-by-ethnicity interaction terms using a consortium of four centenarian studies: the New England Centenarian Study, the Long Life Family Study, the Southern Italian Centenarian Study, and the Longevity Gene Project. We used the Uniform Manifold Approximation and Projection (UMAP), a non-linear dimension reduction technique, to identify distinct ethnic clusters. The UMAP analysis revealed four distinct ethnic groups (Danish, Italian, Ashkenazi Jewish, and middle European) in the aggregated data set with 2223 cases and 5673 controls. Using a mixed effects logistic model with SNP-by-ethnicity interaction terms, we found 29 loci, in which the test of any interaction effect produced p&amp;lt; 10-5. The results showed that some variants had ethnicity-specific effects on EL. We sought for replication in two independent studies of longevity: the Danish Longevity Study and the Italian Longevity Study. In the Italian Longevity Study, rs7907949 (PFKP), rs11667516 (intergenic: RAB11B;MARCHF2), rs13245505 (intergenic: ZC3HC1;KLHDC10) replicated with a nominal significance level of 0.05. In the Danish Longevity Study, rs79853795 (PLCB1) and rs4072601 (SLC39A11) replicated. Future drug development should account for ethnic-specific differences in the genetic effects for higher efficacy for more diverse populations.

THE ASSOCIATION BETWEEN FRAILTY AND PERCEIVED FATIGABILITY IN THE LONG LIFE FAMILY STUDY

Abstract Higher levels of frailty, quantified by a frailty index (FI), may be linked to fatigue severity as tasks become more physically and mentally demanding. However, the association between frailty and fatigability—quantification of vulnerability to fatigue in relation to specific intensity and duration of activities—has not been assessed. Using cross-sectional data from the Long Life Family Study Visit 2 (2014–2017; n=2,524; mean age +/- standard deviation 71.4+/-11.2 years; 55% women; 99% White), we examined the association between a 79-item FI (ratio of number of health problems reported (numerator) out of the 79 (denominator); higher percentage=greater frailty) and perceived physical and mental fatigability using the Pittsburgh Fatigability Scale (PFS) (range 0–50; higher scores=greater fatigability). Mean+/-SD FI scores were 0.08+/-0.06 and mean+/-SD PFS Physical and Mental scores were 13.7+/-9.6 (39.5% more severe, &amp;gt;=15) and 7.9+/-8.9 (22.8% more severe, &amp;gt;=13), respectively. Both PFS subscale scores were higher for each 0.10 increment in FI. Mean PFS scores were 10.7 and 34.2 (Physical) and 5.7 and 28.8 (Mental) for FI scores of &amp;lt; 0.10 (non-frail) and ≥0.30 (moderate-severely frail), respectively. In mixed effects models, a 0.03 higher FI score (accepted clinically meaningful increase in FI) was associated with 1.9-point higher PFS Physical (95% confidence interval (CI) 1.7–2.1) and 1.7-point higher PFS Mental (95% CI 1.5–1.9) scores after accounting for family structure and adjusting for age, sex, field center, body mass index, smoking status, education, and marital status. Individuals with higher FI scores may benefit from targeted interventions to mitigate further poor health outcomes.

Lexical markers from paragraph recall predict cognitive status

AbstractBackgroundCognitive tests that require a spoken response, such as paragraph recall, are rich in cognitive‐related information that is not captured using traditional scoring methods. This study aimed to determine whether linguistic features embedded in spoken responses can differentiate between individuals who are and are not cognitively impaired.MethodWe used linguistic text analysis software (Linguistic Inquiry Word Count) to identify lexical and grammatical features from manually transcribed recordings of Logical Memory immediate (IR) and delayed recall (DR) from a subset (N = 169) of Long Life Family Study participants. We used regression models with Generalized Estimating Equations adjusted by age, sex, education, and within‐family correlation to select features associated with cognitive status (normal cognition [NC] versus cognitive impairment [CI]; threshold p&lt;0.1). We developed a “polyfeature score” (PFS) for both immediate and delayed recall, each calculated as a weighted sum of the selected lexical features for each recall condition. We then built a logistic regression model to evaluate the predictive value of the PFSs for identifying cognitively impaired individuals.ResultThe sample included 133 participants with NC and 36 with CI (mean age = 75.8±11.7 years, 56% female). The regression identified 11 features for IR (e.g., percent of prepositions, words &gt; = 6 letters) and 13 features for DR (e.g., percent of perceptual process words, percent of conjunctions) that significantly predicted cognitive status. Only 45% of features selected in IR were also selected in DR. Both PFSs were highly associated with cognitive status (PFS_IR β = 1.48, p &lt; 0.001; PFS_DR β = 0.67, p = 0.001). The PFS yielded high discriminative value for cognitive status (PFS_IR AUC = 0.96, sensitivity = 0.73, specificity = 0.86; PFS_DR AUC = 0.90, sensitivity = 0.82, specificity = 0.86). Combining lexical features from both recall conditions did not increase the predictive value of the models.ConclusionLexical features from immediate paragraph recall alone provide sufficient information for classifying cognitive status increasing its potential as a cognitive screener in clinical settings. Additionally, each recall condition identified unique lexical features associated with cognitive impairment which may help elucidate processes underlying deficits in learning and recall.

Association of leukocyte telomere length with perceived physical fatigability

BackgroundLeukocyte telomere length (LTL) is a potential genomic marker of biological aging, but its relation to fatigability, a prognostic indicator of phenotypic aging (e.g., functional decline) is unknown. We hypothesized shorter LTL would predict greater perceived physical fatigability, but that this association would be attenuated by adjusting for chronological age. MethodsTwo generations of participants (N = 1997; 309 probands, 1688 offspring) were from the Long Life Family Study (age = 73.7 ± 10.4, range 60–108, 54.4 % women), a longitudinal cohort study of aging. LTL was assayed at baseline. Perceived physical fatigability was measured 8.0 ± 1.1 years later using the validated, self-administered 10-item Pittsburgh Fatigability Scale (PFS, 0–50, higher scores = greater fatigability). Generalized estimating equations were generated to model the association between LTL and PFS Physical scores. ResultsPrevalence of greater physical fatigability (PFS scores≥15) was 41.9 %. Using generalized estimating equations, a one kilobase pair shorter LTL was associated with higher PFS Physical scores (β = 1.8, p < .0001), accounting for family structure, and adjusting for field center, follow-up time, sex, and follow-up body mass index, physical activity, and chronic health conditions. When age was included as a covariate, the association was fully attenuated (β = 0.1, p = .78). ConclusionLTL may provide an alternative method for estimating an individual's lifetime exposure to chronic stressors, but does not appear to provide additional information not captured by chronological age. Further research is needed to characterize the interaction between age, LTL, and perceived fatigability, and develop a method of identifying individuals at risk for deleterious aging.

Evaluation of GENESIS, SAIGE, REGENIE and fastGWA-GLMM for genome-wide association studies of binary traits in correlated data.

Performing a genome-wide association study (GWAS) with a binary phenotype using family data is a challenging task. Using linear mixed effects models is typically unsuitable for binary traits, and numerical approximations of the likelihood function may not work well with rare genetic variants with small counts. Additionally, imbalance in the case-control ratios poses challenges as traditional statistical methods such as the Score test or Wald test perform poorly in this setting. In the last couple of years, several methods have been proposed to better approximate the likelihood function of a mixed effects logistic regression model that uses Saddle Point Approximation (SPA). SPA adjustment has recently been implemented in multiple software, including GENESIS, SAIGE, REGENIE and fastGWA-GLMM: four increasingly popular tools to perform GWAS of binary traits. We compare Score and SPA tests using real family data to evaluate computational efficiency and the agreement of the results. Additionally, we compare various ways to adjust for family relatedness, such as sparse and full genetic relationship matrices (GRM) and polygenic effect estimates. We use the New England Centenarian Study imputed genotype data and the Long Life Family Study whole-genome sequencing data and the binary phenotype of human extreme longevity to compare the agreement of the results and tools’ computational performance. The evaluation suggests that REGENIE might not be a good choice when analyzing correlated data of a small size. fastGWA-GLMM is the most computationally efficient compared to the other three tools, but it appears to be overly conservative when applied to family-based data. GENESIS, SAIGE and fastGWA-GLMM produced similar, although not identical, results, with SPA adjustment performing better than Score tests. Our evaluation also demonstrates the importance of adjusting by full GRM in highly correlated datasets when using GENESIS or SAIGE.