STUDY OBJECTIVES:GGIR is an R package for processing raw acceleration data to estimate sleep health parameters. We aimed to (1) assess the performance of three sleep algorithms within GGIR against PSG for detecting sleep/wake in clinically referred, typically-developing children (criterion validity); and (2) describe GGIR-derived sleep estimates from typically developing children enrolled in multiple cohort studies (face validity). METHODS:For criterion evaluation, children (8-16 years, N = 30) wore an actigraphy device for one night during in-lab polysomnography with performance assessed using epoch-by-epoch analyses. For face validity evaluation, four community/free living datasets were used: (1) Bone Mineral Accretion in Young Children (3-5 years, N = 310), (2) School Summer Sleep (5-8 years, N = 118), (3) Sleep and Growth Study 2 (12-13 years; N = 291), and (4) Early Life Exposure to Environmental Toxicants (9-18 years; N = 543). All raw acceleration data were processed using GGIR (v.3.0-0) with the Cole-Kripke (CK), Sadeh (S), and van Hees (vH) algorithm settings. RESULTS:Following the in-lab test, 60 per cent of children were diagnosed with mild to severe obstructive sleep apnea (OSA). For criterion evaluation, the 30-s epoch-by-epoch analyses revealed that average balanced accuracies were 0.80 (Sensitivity = 0.80; Specificity = 0.79), 0.76 (Sensitivity = 0.86; Specificity = 0.65), and 0.67 (Sensitivity = 0.95, Specificity = 0.39) for GGIR-CK, GGIR-vH, and GGIR-S, respectively. For face validity evaluation, sleep estimates mirrored the in-lab performance metrics (e.g. sleep duration estimates were similar when using GGIR-CK and GGIR-VH but approximately 1 h longer when using GGIR-S). CONCLUSIONS:The in-lab performance metrics from typically developing children with and without OSA and cohort-based descriptive statistics from samples of typically developing children provide benchmark data to guide investigators on the suitability of GGIR for automated processing of raw acceleration data for pediatric sleep estimation.

Abstract Introduction Delayed sleep timing has been associated with higher blood pressure among adolescents. However, rest-activity rhythms (RAR), which incorporate both sleep and physical activity, have been rarely explored. Methods The dataset included 558 adolescents (mean age 13.98, SD = 2.15) from the Mexico City-based ELEMENT cohort. Participants completed 7-day wrist actigraphy at the adolescent baseline (T1) visit. R package GGIR was used to derive RAR variables (Amplitude, IS, IV, Up Mesor, Down Mesor, Mesor, L5, L5 time, M10, M10 time, Acrotime, and Relative Amplitude). Blood pressure (systolic, SBP; and diastolic, DBP) was measured at two time-points approximately two years apart (T1 and T2). Linear regression models with RAR variables as continuous exposures and changes in SBP/DBP as outcomes, were adjusted for confounders. Sex stratified models were also tested. False detection rates (FDR) were controlled with the Benjamini Hochberg method. Results We found several associations related to timing of activity. Later L5 (“least active 5 hours”) starting time was associated with positive two year increases in DBP (Beta = 0.265 with SE 0.07 and P < 0.001) and higher SBP (Beta = 0.191, SE 0.08; P = 0.017). In contrast, later Up Mesor (transition from lower to higher activity) was associated with negative two-year changes in SBP and DBP (Beta=-0.37, SE 0.16; P=0.02; and Beta=-0.40, SE 0.14; P=0.003, respectively). Similarly, later M10 (most active 10 hours) was associated with negative two-year changes in SBP (Beta = -0.727, SE 0.20; P < 0.001) and DBP (Beta = -0.709, SE0.23; P = 0.002). All DBP associations remained significant with FDR adjustment. Finally, there were sex differences, such that the M10 association was more prominent in males and the L5 and Up Mesor associations were more prominent in females. Conclusion We found associations between timing of activity throughout the 24-hour day, but not other RAR metrics, with prospective changes in blood pressure. In line with prior research, later low activity periods (i.e., sleep) associated with higher blood pressure. In contrast to expectation, later timing of higher activity periods associated with relative reductions in blood pressure over time in a sex-specific manner. Support (if any) R01HL169893

Abstract Introduction Adequate sleep is essential for childhood development, yet wearable-based assessments often depend on proprietary algorithms. To enhance reproducibility, the field is moving toward open-source solutions. We developed the Modular Actigraphy Platform (MAP), an open-source tool that makes processing raw data easier and efficient. This study evaluates MAP’s performance using pediatric sensor data. Methods MAP is deployed on Google Cloud Platform and comprises four core modules: (1) data pre-processing, (2) non-wear detection, (3) sleep scoring, and (4) physical activity scoring. Each module is Docker-containerized, with Google Kubernetes Engine providing orchestration and Argo to manage workflow execution. The pipeline incorporated the GGIR algorithm and the Monitor Independent Movement Summary (MIMS) algorithm to process raw data for sleep and physical activity estimation. To demonstrate performance, we processed tri-axial acceleration wrist-worn data from ActiGraph and GENEActiv devices and compared MAP’s outputs to offline data processing benchmarks. Testing included unit, integration, and system checks, followed by alpha and beta testing. The goal of alpha testing was to evaluate MAP in a development environment. Alpha testing used 17 CHOP-generated files (4.9 GB, 1–14 days). Beta testing assessed functionality and scalability on real-world datasets from four cohort studies (686 files, each with up to 4 days of data at a 30–50 Hz sampling rate). Results MAP passed unit testing for stability, functionality, and security. Preprocessing was the most computationally demanding module, requiring 41–61% of the total processing time. Beta testing required up to 60 CPUs and 120 GiB memory for preprocessing and 40 CPUs/100 GiB for the other modules. The memory-to-CPU ratio was critical for avoiding crashes. Performance evaluations showed that MAP was faster than offline processing. The offline GGIR pre-processing was 1.6 to 2.9 times slower compared to MAP. MIMS processing was most resource-intensive, requiring up to 20 CPUs and 500 GiB of memory. For MIMS unit processing, MAP processed 243.3 GB in 444 minutes versus 6007 minutes offline (2.4–14× faster). Conclusion Raw actigraphy data increasingly support sleep estimation but require advanced expertise. MAP reduces barriers by offering an accessible, efficient cloud-based solution for deriving sleep metrics from raw sensor data. Support (if any)

Background:Clonal hematopoiesis (CH) is associated with increased risks of diverse cardiovascular diseases, hematologic malignancies and mortality, yet no preventive therapies are approved. As emerging data implicate lipid pathways in CH pathogenesis, we investigated the association of statin use and genetically proxied inhibition of HMG-CoA reductase (HMGCR) with CH risk, and validated findings using primary peripheral blood mononuclear cells (PBMCs). Methods:We performed an observational analysis of 416,118 UK Biobank participants of European ancestry using multivariable logistic regression to compare CH prevalence among statin users and nonusers. Mendelian randomization (MR) analyses evaluated the causal association of genetically proxied lowering of low-density lipoprotein cholesterol (LDL-C) with risk of CH using two instruments; (i) the lead HMGCR variant (rs12916) which proxied LDL-C lowering by statins, and, (ii) 303 genome-wide LDL-C-lowering variants representing polygenic mechanisms. Summary statistics were obtained from the Global Lipid Genetics Consortium genome-wide association study (N = 842,634). Experimentally, primary PBMCs from a DNMT3A R882 hotspot mutation carrier were cultured in methylcellulose with pravastatin or vehicle control to evaluate colony-forming dynamics. Results:Among 416,118 individuals, 20,488 had CH, including 11,550 with single DNMT3A-mutant and 4,375 with single TET2-mutant CH. Pre-recruitment statin users had reduced odds of DNMT3A-mutant CH (OR=0.93; 95% CI:0.88-0.98; P=0.009), driven primarily by associations with DNMT3A R882-mutant (OR=0.78; 95% CI:0.66-0.92; P=0.003), but not TET2-mutant CH (OR=1.05; 95% CI:0.97-1.14; P=0.20). Similarly, genetically predicted HMG-CoA-reductase inhibition equivalent to a 1 SD reduction in circulating LDL-C levels was associated with lower odds of DNMT3A-mutant CH (OR=0.66; 95% CI:0.45-0.95; P=0.03) but not TET2-mutant CH (OR=1.34; 95% CI:0.76-2.36; P = 0.31). By contrast, polygenic estimation of LDL-C lowering was not associated with DNMT3A-mutant CH (OR=1.05; 95% CI:0.97-1.14; P=0.20), suggesting protective effects were independent of LDL-C lowering per se. Genetically predicted HMG-CoA reductase inhibition had wide effects on blood cell counts and indices, suggesting effects on bone marrow cell dynamics. In vitro, pravastatin selectively suppressed colony formation of primary human DNMT3A R882-mutant relative to wild-type cells (P=0.031). Conclusions:Statin therapy and genetically predicted lifelong inhibition of HMG-CoA reductase were significantly associated with reduced risk of DNMT3A-mutant CH, likely via LDL-C-independent mechanisms, which may be specific to DNMT3A-mutant CH. This provides a strong rationale for prospective trials evaluating the effect of statins on risk of developing DNMT3A-mutant CH, subsequent clonal expansion, and associated clinical sequelae.

Abstract Introduction While growing evidence suggests that neighborhood safety and disorder may impact adolescent sleep, prior research has not accounted for travel outside the home neighborhood, which may misclassify environmental exposures. This study examined associations of neighborhood physical disorder and crime in adolescents’ daily activity spaces with sleep patterns. Methods In a micro-longitudinal study, 142 adolescents aged 15-18 were followed for 14 days using smartphone GPS tracking and actigraphy. GPS locations recorded during adolescents’ waking hours were linked to neighborhood physical disorder (number of vacant lots) and crime (annual count of part 1 crimes, e.g., homicide, assault) to derive daily average exposures in adolescents’ activity spaces. Multivariable linear mixed effects regression estimated associations of daily neighborhood exposures with nightly actigraphy-assessed adolescent sleep duration and timing, and mixed effects modified Poisson regression estimated associations with risk of insufficient sleep (< 8 hours) and very short sleep (< 6 hours). Exposures were decomposed to assess between-person differences (sleep differences based on adolescents’ average exposures) and within-person differences (whether sleep differed on days where exposure was higher than their individual average). Models adjusted for school night status, month (for seasonality), adolescent age, sex, race/ethnicity, household income, parent education, parent marital status, distance from home to school, percent of GPS points >250 meters from home, and daily activity space neighborhood economic deprivation, social fragmentation, and tree canopy cover. Results Among participants, the average sleep duration was 6.5 hours. On days when adolescents’ exposure to vacant lots was higher than their own average, they had shorter sleep duration, later sleep onset, earlier sleep offset, and higher risk of insufficient sleep (RR per IQR difference: 1.09, 95% CI: 1.02, 1.17) and very short sleep (RR: 1.23, 95% CI: 1.06, 1.44). Adolescents with higher average crime exposure had shorter sleep duration, earlier sleep offset, and higher risk of insufficient and very short sleep, but there were no within-person associations between crime and sleep outcomes. Conclusion Higher exposure to vacant lots and crime in adolescents’ activity spaces was associated with more adverse sleep patterns. While associations do not necessarily indicate causal relationships, the findings may suggest intervention targets to improve adolescent sleep. Support (if any) K01HL155860(NHLBI)

Abstract Introduction Brief Behavioral Treatment for Insomnia (BBTI) improves insomnia symptoms, but its effects on 24-hour rest–activity rhythms (RAR) are not well understood. We evaluated whether BBTI improved RAR parameters compared with an attention control (AC) group among breast cancer survivors (BCS) with insomnia. Methods This secondary analysis of a randomized controlled trial of insomnia in cancer survivors was restricted to BCS who completed both baseline and 1-month actigraphy assessments (N=52; BBTI n=27, AC n=25). Participants were randomized to three weekly sessions of BBTI or an AC healthy eating program and wore a wrist-actigraph (Philips Respironics) for one week at baseline and again 1 month after treatment. Activity count data (60 second epoch) were processed using ActCR and customized function from GGIR package to derive parametric (mesor, amplitude, acrophase) and non-parametric (interdaily stability, intradaily variability, relative amplitude, most active 10-hour midpoint [M10], least active 5-hour midpoint) rest–activity metrics. Linear mixed-effects models were used to assess changes in these metrics over time by treatment group, incorporating time, treatment, and their interaction in the fixed effects portion of the models. Results Participants were, on average, 60 years old, predominantly White and married or living with a partner. Time since breast cancer diagnosis averaged 9.6 and 7.4 years in the BBTI and attention control groups, respectively. Baseline insomnia severity was in the subthreshold–moderate range (mean ISI ≈13), and there were no significant between-group differences in demographic, clinical, or insomnia characteristics. Within groups, pre- to post-treatment changes in parametric and non-parametric rest activity metrics were small in the BBTI group. In the AC group, the M10 midpoint became modestly later (~0.4 hours; d≈0.30) with other parameters largely unchanged. Between groups, no significant treatment-by-time interactions were observed for any rest–activity rhythm parameter (all p≥0.09; |d| ≤ 0.42). Conclusion In this sample of BCS with insomnia, the BBTI intervention showed no significant group or time effects in actigraphic RAR parameters compared to the AC group over the 1-month period. Larger trials with longer follow-up and interventions that more directly target sleep and circadian health may be needed to meaningfully modify RAR in this population. Support (if any) NIH/NINR R01NR018215 (Dickerson); NIH/NHLBI T32HL007953 (Kwon).

Abstract Introduction Increasingly, actigraphy methods seek to leverage raw acceleration data and machine-learning scoring classification. However, much of the progress has been made in adults. We therefore trained machine-learning sleep-wake classifiers using pediatric data. We aimed to assess their sleep-wake scoring ability and benchmarked against an adult-trained classifier and algorithms in GGIR. Methods Sixty children (26 female, ages 5.3–17.7 years) completed in-lab overnight polysomnography at the Children’s Hospital of Philadelphia and wore a GENEActiv device (3-axis accelerometer, 50 Hz) on their non-dominant wrist. The acceleration data were converted into 30-second epochs and aligned with physician-scored sleep-wake data from electroencephalography. Six machine-learning models were trained using leave-one-subject-out cross-validation. Epoch-by-epoch analyses generated performance metrics: sensitivity and specificity, with balanced accuracy (BA) used to rank. Discrepancy analyses compared the overall sleep duration estimated. Results Overall, 560.1 hours of data were collected; 74.4% of epochs were scored as sleep with an average sleep duration of 7.1 hours (SD = 1.9). Of the six pediatric-trained machine learning models, the top two were random forest (BA = 0.78; Sensitivity = 0.87; Specificity = 0.69) and neural network (BA = 0.77; Sensitivity = 0.83; Specificity = 0.73). These performance metrics exceeded that of an adult-trained neural net classifier applied to our data (BA = 0.71; Sensitivity = 0.93; Specificity = 0.49), but were comparable to the GGIR Cole-Kripke (GGIR-CK: BA = 0.79; Sensitivity = 0.75; Specificity = 0.85) and GGIR van Hees algorithms (GGIR-vH: BA = 0.78; Sensitivity = 0.84; Specificity = 0.73). Overall, sleep duration was underestimated by an average of 15 minutes using the random forest classifier and by an average of 37 minutes using the neural network classifier. For comparison, both GGIR-CK and GGIR-vH underestimated sleep duration by an average of 33 minutes. Conclusion We trained pediatric sleep-wake classifiers that had a strong ability to detect sleep and a moderate-to-strong ability to detect wake. Based on epoch-by-epoch and discrepancy analyses, the random forest classifier was the most optimal, outperforming GGIR-CK, GGIR-vH, and an adult-trained neural network classifier. With larger samples used for training and validation, we may reduce variability and further improve pediatric sleep-wake classification using actigraphy. Support (if any)

Abstract Introduction Osteoporosis is one of the leading chronic diseases that can have origins in early life through suboptimal bone accretion. We know that diet and physical activity can enhance bone accretion, and there are emerging data suggesting sleep patterns may also be a determinant of bone accretion. We therefore aimed to evaluate the associations between actigraphic nighttime sleep measures with areal bone mineral density (aBMD) in adolescents. Methods Participants were enrolled in the Sleep and Growth Study 2 (S-Grow2), a cohort of typically developing adolescents in 7th grade (12-13y) recruited from the Philadelphia area. Participants wore a wrist actigraphic sleep monitor (GENEActiv, ActivInsights) for two weeks and the motion data were processed using GGIR to estimate sleep patterns. They also completed a dual energy X-ray (DXA) scan to estimate areal bone mineral density (aBMD) at multiple skeletal sites and expressed as age and sex specific Z-scores. Cross-sectional, multiple regression analyses tested the associations of average nighttime sleep measures (sleep duration, timing, and efficiency) with aBMD Z-scores (total body less head, lumbar spine, total hip, femoral neck, and the distal radius). Models included a quadratic term for sleep measures, and adjusted for sex, race, daily minutes of physical activity, body mass index z-score, and dietary calcium intake. Results Participants (n=302) were 65% white and 47% female. Sleep period duration demonstrated a U-shaped association with aBMD z-scores at the femoral neck (p=.023) and total hip (p=.008), with both shorter and longer sleep durations linked to higher areal bone mineral density. Sleep efficiency showed an inverse U-shaped relationship with aBMD z-scores at the distal radius (p=.01), such that both lower and higher sleep efficiency were associated with lower areal bone mineral density. Sleep timing was not associated with any aBMD measures. Conclusion Our cross-sectional results illustrate non-linear relationships between sleep and areal bone mineral density in adolescents aged 12-13y. This cohort is actively being followed up at 1 year intervals in 8th and 9th grades and future analyses will reveal how changes in sleep impact bone accretion during this developmental window. Support (if any) NIH/NICHD R01HD100421

To assess the contribution of rare coding germline genetic variants to prostate cancer risk and severity, we perform here a meta-analysis of 37,184 prostate cancer cases and 331,329 male controls from five cohorts with germline whole exome or genome sequencing data, and one cohort with imputed array data. At the gene level, our case-control collapsing analysis confirms associations between rare damaging variants in four genes and increased prostate cancer risk: SAMHD1, BRCA2 and ATM at the study-wide significance level (P < 1x10(-8)), and CHEK2 at the suggestive threshold (P < 2.6x10(-6)). Our case-only analysis, reveals that rare damaging variants in AOX1 are associated with more aggressive disease (OR = 2.60 [1.75-3.83], P = 1.35x10(-6)), as well as confirming the role of BRCA2 in determining disease severity. At the single-variant level, our study reveals that a rare missense variant in TERT is associated with substantially reduced prostate cancer risk (OR = 0.13 [0.07-0.25], P = 4.67x10(-10)), and confirms rare non-synonymous variants in a further three genes associated with reduced risk (ANO7, SPDL1, AR) and in three with increased risk (HOXB13, CHEK2, BIK). Altogether, this work provides deeper insights into the genetic architecture and biological basis of prostate cancer risk and severity, with potential implications for clinical risk prediction and therapeutic strategies.

Rest-activity rhythms (RAR), a behavioral manifestation of circadian rhythms, represent the patterns of activity and rest/sleep across a 24-hour period. However, little is known about how changes in RARs during adolescence may influence adiposity outcomes. We examined the associations between RAR and body composition measures in adolescents. We hypothesized that disrupted RAR patterns are associated with adverse adiposity outcomes. Baseline and follow-up data were collected from participants in the Sleep and Growth Study. Movement profiles were recorded using wrist-worn actigraphy devices (Philips Actiwatch 2); proprietary count data were analyzed using ActCR and GGIR packages to calculate RAR metrics. These included non-parametric (interdaily stability, intradaily variability, relative amplitude, most active 10-hour midpoint, least active 5-hour midpoint) and parametric (mesor, amplitude, acrophase) metrics. Body composition measures were assessed via dual-energy x-ray absorptiometry (DXA) scans: fat mass index (FMI, kg/m², and z-score) and visceral fat area (cm²). Quantile regression was used to test for associations between RAR and specific percentiles (i.e., 5th, 25th, 50th, 75th, 95th) of the body composition outcomes, that represent the lower, middle and upper ranges of the frequency distributions. The sample consisted of 107 adolescents (mean age: 13.9 years at baseline and 14.9 years at follow-up), with 54.2% male and predominantly non-Hispanic White participants. Associations between RAR metrics and FMI were first examined. No significant associations were observed between any RAR metric and FMI (kg/m2, and z-score) across the 5th, 25th, 50th, 75th, and 95th percentiles. We then examined the associations between RAR metrics and visceral fat area. Similarly, no significant associations were identified between RAR metrics and visceral fat area across any percentiles. In this relatively small longitudinal dataset with robust assessments of RAR and body composition, we found no evidence to support associations between RAR metrics and obesity-related outcomes in adolescents. Further research with larger sample sizes and additional cardiovascular data is needed to better understand the potential links between RAR and obesity-related health outcomes. NIH/NHLBI K01HL123612 (Mitchell), NIH/NHLBI T32HL007953 (Kwon)

The impact of genetic ancestry on the development of clonal hematopoiesis (CH) remains largely unexplored. Here, we compared CH in 136,401 participants from the Mexico City Prospective Study (MCPS) to 416,118 individuals from the UK Biobank (UKB) and observed CH to be significantly less common in MCPS compared to UKB (adjusted odds ratio = 0.59, 95% confidence interval (CI) = [0.57, 0.61], P = 7.31 × 10-185). Among MCPS participants, CH frequency was positively correlated with the percentage of European ancestry (adjusted beta = 0.84, 95% CI = [0.66, 1.03], P = 7.35 × 10-19). Genome-wide and exome-wide association analyses in MCPS identified ancestry-specific variants in the TCL1B locus with opposing effects on DNMT3A-CH versus non-DNMT3A-CH. Meta-analysis of MCPS and UKB identified five novel loci associated with CH, including polymorphisms at PARP11/CCND2, MEIS1 and MYCN. Our CH study, the largest in a non-European population to date, demonstrates the power of cross-ancestry comparisons to derive novel insights into CH pathogenesis.

Feathers are complex structures exhibiting structural/functional disparity across species and plumage. Flight was lost in >30 extant lineages from ~79.58 Ma-15 Ka. Effects of flight loss on senses, neuroanatomy, and skeletomusculature are known. To study how flightlessness affects feathers, we measured 11 feather metrics across the plumage of 30 flightless taxa and their phylogenetically closest volant taxa, with broader sampling of primaries across all orders of crown birds. Our sample includes 27 independent flight losses, representing nearly half of extant flightless species. Feather asymmetry measured by barb angle differences between trailing and leading vanes decreases in flightless lineages, most prominently in flight feathers and weakest in contour feathers. Greatest changes in feather anatomy occur in older flightless lineages (penguins, ratites). Comparative methods show that many microscopic feather traits are not dramatically modified after flightlessness compared to body mass increase and relative wing and tail fan reduction. Changes involved with greater vane symmetry show stronger shifts, however. Relaxing selection for flight does not rapidly modify feather flight adaptations, apart from asymmetry. Developmental constraints and relaxed selection for novel feather morphologies may explain some observed changes. Macroscopic changes to flight apparati (skeletomusculature, airfoil size) are more evident in recently flightless taxa and could more reliably detect flightlessness in fossils, with increased feather symmetry as a potential microscopic signal. We observed apical modification in later stages of feather development (asymmetric displacement of barb loci), while morphologies arising during early developmental stages are only altered after millions of years of flightlessness.

Background Telomere biology disorders (TBDs) are caused by pathogenic variants in telomerase reverse transcriptase (TERT) and other telomere-related genes and are defined by premature telomere shortening and multiorgan manifestations, including bone marrow failure, pulmonary fibrosis, liver disease, osteoporosis, and cancer. Although TBDs are clinically rare, TERT rare missense variants occur in ~1% of the population with 90% classified as variants of uncertain significance. This disparity suggests that classical TBDs represent the extreme end of a broader continuum of telomere dysfunction. To quantify this latent potential, we generated a population-scale functional atlas of TERT missense variants in the UK Biobank and assessed their association with telomere length (TL), clonal hematopoiesis (CH), and clinical phenotypes. Methods We identified TERT rare variants (max population allele frequency <0.001) in 462,666 UK Biobank participants. To define their functional impact, we developed a cell-based assay to quantify telomere extension relative to wild-type (WT) TERT. We performed molecular simulations and free energy (ΔΔG) calculations using the cryo-EM structure of telomerase bound to TPP1. Functional scores were integrated with a common variant-derived TL polygenic risk score (PRS) to define their impact on TL, hematologic traits, CH, and clinical phenotypes. Results To establish functional thresholds, we benchmarked our assay using 9 common variants and 120 TBD-associated variants. Common variants did not impact telomere extension, while all TBD-associated variants were impaired: 62% showed no extension ('severe') and 38% had reduced extension ('intermediate'). Among 601TERT missense variants, 69% were impaired (13% severe, 56% intermediate) in 1297 UK Biobank participants. Severe variants clustered in domains that form the active site (C-terminal extension domain, CTE: 29%; reverse transcriptase domain, RTD: 18%) or bind to TPP1 (telomerase-essential N-terminal domain, TEN: 25%). Variants in the telomerase RNA binding domain (TRBD) had fewer severe (7%) but more intermediate variants, yielding similar impairment rates (69% vs 74%, p=0.28). No severe variants occurred in the linker domain. To investigate mechanisms of impairment, we modeled 14 variants across five multiallelic positions (TEN:G135, TRBD:R358, RTD:R669/R756, CTE:R962) that had variants with divergent functional scores. Functional impairment correlated with increased ΔΔG (>3 kcal/mol, r = -0.75, p=0.007) and reduced side-chain contacts with TPP1, TERC, or the RNA:DNA duplex. Severe variants showed disrupted interactions and intermediate variants showed partial contact loss or altered flexibility, indicating domain-specific mechanisms of dysfunction and a gradient of severity across alleles. To determine whether TERT functional impairment correlated with in vivo telomere shortening, we analyzed leukocyte TL in UK Biobank participants. Severe variants were associated with shorter TL compared to controls (β = -1.55, p<0.0001), similar to protein-truncating variants (β= -1.27). Intermediate variants had a moderate effect (β = -0.55, p<0.0001), while preserved variants had minimal impact (β = -0.09). TL effects were strongest for variants in the TEN, RTD, and CTE domains. The PRS modified TL across functional groups, but its effect was attenuated with increasing TERT variant impairment (standardized interaction coefficients: preserved 1.7, intermediate 0.9, severe 0.3). Among intermediate carriers, TL ranged from preserved-like to severe-like across PRS quartiles. We examined whether TERT functional impairment was associated with clinical phenotypes. Intermediate and severe, but not preserved, variants were associated with bone marrow dysfunction (higher MCV, lower RBC count, lower platelet mass), TBD-related CH with somatic PPM1D or TERT promoter mutations (OR=22 and 118, both p<0.0001), interstitial lung disease (OR=7.9 and 66.1, both p<0.0001). Severe variants were associated with aplastic anemia (OR=10.4, p=0.012) and osteoporosis (OR=29.5, p<0.0001). Conclusion Most germline TERT rare missense variants are functionally impaired (69%) with domain-specific effects linked to shorter TL and manifestations of telomere dysfunction including TBD-related CH and clinical phenotypes. Polygenic background modulates – but does not offset – the effects of impaired TERT variants, revealing a continuum of inherited telomere dysfunction shaped by rare and common variants.