Nonlinear and Episodic Growth in the Natural History of Ascending Aortic Dilation

BACKGROUND: Predicting ascending aortic growth is challenging. Linear, continuous assumptions may oversimplify biology. We evaluated the following: (i) whether baseline body size-indexed diameter has a nonlinear association with subsequent growth; and (ii) whether patient-level trajectories are continuous or episodic. METHODS: For this single-centr, retrospective study (2012-2024), the primary cohort (n = 3315; ≥ 2 computed tomography/magnetic resonance scans) modelled the association between baseline indexed size (z score) and annualized growth using multivariable linear regression and generalized additive models, adjusting for clinical covariates. A sub-cohort (n = 1055; ≥ 4 scans) was classified into the following longitudinal phenotypes: stable (total growth < 2.0 mm); stable-with-noise (total growth < 2.0 mm with alternating small changes); continuous growth (total growth ≥ 2.0 mm without an episodic event); and discontinuous/episodic growth (total growth ≥ 2.0 mm with ≥ 1 growth episode, defined as a diameter increase ≥ 2.0 mm within 1 interval or across 2 adjacent intervals). RESULTS: In the primary cohort, baseline z score showed a significant nonlinear (U-shaped) association with subsequent growth in generalized additive models (P < 0.001), with higher growth for small (z < 0) and severely dilated (z > 5) aortas. The size-growth relationship was significantly moderated by age, sex, and blood pressure. In the sub-cohort, the phenotype distribution was as follows: stable, 50.4% (532 of 1055); stable-with-noise, 21.6% (228 of 1055); continuous, 5.4% (57 of 1055); and discontinuous/episodic, 22.6% (238 of 1055). Among patients with total growth ≥ 2.0 mm (n = 295), 81% (238 of 295) were episodic, and 58.3% (172 of 295) had a nondilated baseline aorta (z < 2). CONCLUSIONS: The cohort-level size-growth relationship is nonlinear, and episodic growth behavior dominates among those that grow larger. These findings support reassessing surveillance intervals, risk communication, and threshold-based decision pathways in thoracic aortic disease.