ABSTRACT
BACKGROUND AND AIM: Tissue source has been shown to exert a significant effect on the magnitude of associations between telomere length and various health outcomes and exposures. The purpose of the present qualitative review and meta-analysis is to describe and investigate the impact of study design and methodological features on the correlation between telomere lengths measured in different tissues from the same healthy individual. METHODS: This meta-analysis included studies published from 1988 to 2022. Databases searched included PubMed, Embase, and Web of Science and studies were identified using the keywords "telomere length" and "tissues" or "tissue." A total of 220 articles of 7856 initially identified studies met inclusion criteria for qualitative review, of which 55 met inclusion criteria for meta-analysis in R RESULTS: Studies meeting inclusion criteria for meta-analysis tended to have enhanced demographic and methodological reporting relative to studies only included in the qualitative review. A total of 463 pairwise correlations reported for 4324 unique individuals and 102 distinct tissues were extracted from the 55 studies and subject to meta-analysis, resulting in a significant effect size z = 0.66 (p < 0.0001) and meta-correlation coefficient of r = 0.58. Meta-correlations were significantly moderated by sample size and telomere length measurement methodology, with studies of smaller size and those using hybridization-based analyses exhibiting the largest meta-correlation. Tissue source also significantly moderated the meta-correlation, wherein correlations between samples of a different lineage (e.g., blood vs. non-blood) or collection method (e.g., peripheral vs. surgical) were lower than correlations between samples of the same lineage or collection method. CONCLUSION: These results suggest that telomere lengths measured within individuals are generally correlated, but future research should be intentional in selecting a tissue for telomere length measurement that is most biologically relevant to the exposure or outcome investigated and balance this with the feasibility of obtaining the sample in sufficient numbers of individuals.
Subject(s)
Telomere , Humans , Databases, FactualABSTRACT
Use of telomere length (TL) as a biomarker for various environmental exposures and diseases has increased in recent years. Various methods have been developed to measure telomere length. Polymerase chain reaction (PCR)-based methods remain wide-spread for population-based studies due to the high-throughput capability. While several studies have evaluated the repeatability and reproducibility of different TL measurement methods, the results have been variable. We conducted a literature review of TL measurement cross-method comparison studies that included a PCR-based method published between January 1, 2002 and May 25, 2020. A total of 25 articles were found that matched the inclusion criteria. Papers were reviewed for quality of methodologic reporting of sample and DNA quality, PCR assay characteristics, sample blinding, and analytic approaches to determine final TL. Overall, methodologic reporting was low as assessed by two different reporting guidelines for qPCR-based TL measurement. There was a wide range in the reported correlation between methods (as assessed by Pearson's r) and few studies utilized the recommended intra-class correlation coefficient (ICC) for assessment of assay repeatability and methodologic comparisons. The sample size for nearly all studies was less than 100, raising concerns about statistical power. Overall, this review found that the current literature on the relation between TL measurement methods is lacking in validity and scientific rigor. In light of these findings, we present reporting guidelines for PCR-based TL measurement methods and results of analyses of the effect of assay repeatability (ICC) on statistical power of cross-sectional and longitudinal studies. Additional cross-laboratory studies with rigorous methodologic and statistical reporting, adequate sample size, and blinding are essential to accurately determine assay repeatability and replicability as well as the relation between TL measurement methods.
