Normal reference values for thoracic and abdominal aorta and main pulmonary artery dimensions by cardiovascular magnetic resonance: the Framingham heart study

Enlargement of the aorta or main pulmonary artery( MPA) is associated with cardiopulmonary disease, and an increased MPA-to-ascending aorta ratio is associated with pulmonary hypertension. We sought to determine mean and upper 90th percentile (p90) diameters of the thoracic and abdominal aorta and MPA in a longitudinally-followed adult cohort without clinical cardiopulmonary disease.

1794 Framingham Heart Study Offspring cohort members (65±9 yrs, 844 men) underwent ECG-gated, free breathing T2-weighted black-blood TSE at 1.5T (Philips Gyroscan NT, TR=3RR, TE=45ms, trigger delay=75ms (thorax) or 125ms (abdomen), 1.03x0.64-mm² in-plane resolution, THK=5mm, Gap=5 (abdomen) or 10mm (thorax)). Ascending (ASC) and descending thoracic (DTA) aortic and MPA diameters were measured at MPA-bifurcation level, abdominal (ABD) aorta was measured 5 mm above renal artery origins. We determined sex-specific mean, SD and p90 values for vessel diameters and MPA/ASC ratio in a healthy referent group free of hypertension (SBP≥140 or DBP≥90 mmHg or on medication), obesity (body mass index≥30 kg/m²), emphysema, prevalent myocardial infarction or heart failure, and any smoking history. Men were compared to women using 2-sample t test. We also indexed vessel diameters to sex-and-vessel specific allometric powers of height (HT^β); β's were determined by linear regression of log(HT) to log(diameter). Indexation to HT^β was selected since indexation to HT or body surface area (BSA) resulted in significant inverse correlations of vessel diameters to HT and/or BSA.

370 Offspring (62±9 yrs) met referent-group criteria. Men had greater aortic (at all levels) and MPA diameters than women both before and after indexation to HT^β. The β values corresponding to ASC, DTA, ABD and MPA were 0.22, 0.30, 0.11 and 0.29, respectively, in men, and 0.10, 0.39, 0.37 and 0.48 in women. Vessel diameters indexed to HT^β were correlated with neither HT nor BSA. The MPA/ASC ratio did not differ between sexes. The mean±SD and upper limits (p90) for raw and indexed vessel diameters and MPA/ASC ratio are shown (Table).

We present CMR-derived sex-specific upper 90th percentile values for aortic and MPA diameters and MPA/ASC ratio derived from a cohort of longitudinally-followed, community dwelling adults free of clinical cardiac and pulmonary disease. These p90 thresholds may be useful for identification of cardiopulmonary pathology.

Supported in part by grant RO1 AG17509 and by subcontract N01-HC-38038 from the National Institutes of Health.