Abstract Body: Background: Studies have suggested that height shortening over a few decades may be associated with adverse health outcomes, but its association with subsequent risk of cardiovascular disease (CVD) is unknown.

Hypothesis: Height shortening over ~25 years is heterogeneous across individuals and greater shortening is independently associated with incident CVD.

Methods: We characterized height shortening from age 44 to 90 with mixed-effects models for 15,672 men and women using data from ARIC visit 1 (N=15,774, age 44-66 years in 1987-89) through visit 5 (N=6,520, age 66-90 years in 2011-13) after excluding 102 participants with outliers of height shortening (defined as >3 SD ). Then, we used multivariable Cox proportional hazards models to quantify associations between measured height shortening from visit 1 to visit 5 with subsequent risk of CVD (coronary heart disease [178 cases], stroke [173 cases], and heart failure [299 cases]) and all-cause mortality (1,012 deaths) over a mean follow-up of 8.3 years among 3,858 participants free of CVD at visit 5. CVD outcomes were adjudicated by physician panels.

Results: Estimated height shortening from age 44 to 90 was greater in women than men (-6.7 cm vs. -6.2 cm), despite their lower average height (162.3 cm vs 176.1 cm) at visit 1. Height shortening from visit 1 to visit 5 over ~25 years ranged from +3 (increase) to -9 cm across our study population. There was a dose-response relationship between greater height shortening and all outcomes of interest except stroke, independent of demographics and traditional CVD risk factors (Table). Coronary heart disease had the strongest association with height shortening (hazard ratio in Model 2, 2.11 [95%CI 1.40-3.10] for Q4, 1.71 [1.12-2.61] for Q3, and 1.19 [0.78-1.80] for Q2 vs. Q1).

Conclusions: Height declines on average ~6-7 cm from age 44 to 90 in men and women, and greater height shortening is independently associated with higher risk of CVD and mortality. Our results support the use of height shortening as a health parameter for gauging CVD risk.