Determinants of chromosome-specific telomere lengths among 2,573 All of Us participants

Telomeres are DNA-protein structures that protect chromosome ends. The DNA component of telomeres shortens as cells divide, and telomere length (TL) is a key biomarker of aging and disease risk. Most previous studies in humans of TL have analyzed average TL; thus, our knowledge of TL variability across chromosome arms remains limited. The availability of long-read whole-genome sequencing (lrWGS) data has enabled the development of computational methods to measure chromosome-specific TL (csTL). We generated lrWGS-based csTLs for > 2,500 All of Us participants and characterized variability in csTL attributable to individuals, chromosome arms, participant characteristics, and technical factors. We found that TL varies by chromosome arm (9.1% of the variance in csTL), mirroring patterns observed in prior studies and highlighting the potential for chromosome-specific mechanisms of TL regulation. Substantial variance in csTL (8.9%) was attributable to individual, independent of age, supporting the hypothesis that individuals are endowed with short or long TL in early development, which is maintained throughout life. While age is inversely associated with TL across all arms, the strength of the association varied, with longer arms showing stronger associations. We demonstrate that csTL estimates can be used to estimate disease associations at individual telomeres, including outlying values in the csTL distribution. Our work identifies lrWGS quality metrics that impact csTL estimation, providing a framework to guide future studies. This study demonstrates the utility of lrWGS data for csTL profiling in population cohorts. Larger studies of csTL are needed to advance our understanding of telomeres in aging and disease.