Native T1 mapping by cardiovascular magnetic resonance discriminates physiological from pathological left ventricular hypertrophy: a systematic review and meta-analysis

Background Differentiating physiological left ventricular hypertrophy (LVH) in competitive athletes from hypertrophic cardiomyopathy (HCM) remains a major diagnostic challenge in sports cardiology, particularly within the 12–15 mm wall thickness “grey zone”. Given the clinical consequences of misclassification, quantitative myocardial tissue characterisation by cardiovascular magnetic resonance (CMR) may provide incremental diagnostic resolution beyond conventional morphological assessment. Purpose To quantitatively assess whether native T1 mapping and extracellular volume fraction (ECV) differentiate athlete’s heart from HCM. Methods A systematic search of PubMed, Embase, Scopus and Web of Science identified 152 records. After duplicate removal, 130 unique studies underwent structured screening according to predefined inclusion criteria. Four studies (n=247 participants) reporting quantitative native T1 and/or ECV data were included. Continuous outcomes were pooled using inverse-variance weighting under a random-effects model employing restricted maximum likelihood (REML) estimation with Hartung–Knapp adjustment. Mean differences (MD) with 95% confidence intervals (CI) were calculated. Between-study heterogeneity was quantified using the I² statistic and Cochran’s Q test. Sensitivity analyses evaluated robustness of pooled estimates to cohort-level variability. Results Three studies (n=197 participants; 115 HCM, 82 athletes) contributed to the pooled native T1 analysis. Native T1 values were significantly higher in HCM compared with athletes (MD 58.11 ms; 95% CI 5.01–111.20; p=0.042), with moderate between-study heterogeneity (I²=63.9%; Q p=0.062). All contributing studies demonstrated concordant directionality favouring higher native T1 in HCM, with intergroup differences ranging from 38 to 78 ms across cohorts. The pooled estimate remained directionally stable under sensitivity analyses. Exploratory ECV analysis demonstrated higher values in HCM; however, substantial heterogeneity was observed in the primary model (I²=91.5%), and effect estimates were sensitive to cohort-level characteristics. Upon exclusion of a phenotypically distinct cohort, heterogeneity resolved (I²=0%) and ECV remained higher in HCM (MD 2.56%; 95% CI 1.03–4.09). Conclusion Native T1 mapping demonstrates statistically significant and consistent quantitative intergroup separation between pathological and physiological LVH, supporting its potential role as an adjunctive diagnostic biomarker within the LVH grey zone. While ECV findings are directionally concordant, limited data availability and inter-study variability constrain definitive threshold derivation. Larger standardised head-to-head cohorts are required to refine diagnostic cut-offs and inform clinical decision algorithms.