Abstract P394: Associations of Street-View Greenspace Exposure With Cardiovascular Health (Life’s Essential 8) Among Women in Midlife

Abstract

Introduction: Many women experience suboptimal cardiovascular health (CVH) during midlife. Greenspace exposure has been inversely associated with cardiovascular disease outcomes and risk factors, but the majority of studies use satellite-based assessment of greenspace and limited evidence exists regarding overall CVH. Methods: We performed a longitudinal analysis of women in the Project Viva cohort based in Massachusetts. We applied deep learning algorithms to Google street view images to derive metrics of visible greenspace within 500m of participant’s residential addresses in 2012-2016 (median 46y). We calculated %greenspace from trees, grass, and other greenspace (i.e. plants, fields, flowers) within each image, separately and total combined, and transformed each into z-scores. At the Mid-Life Visit ~5 years later (51y), participants completed questionnaires and research staff measured their height, weight, blood pressure, and blood levels of glucose, HbA1c and cholesterol. We calculated CVH score using Life’s Essential 8 (LE8) algorithm (0-100 points, higher=better), that includes 8 components of CVH, including 4 behavioral measures (heart healthy levels of diet, physical activity, sleep, avoidance of nicotine) and 4 biomedical measures (body mass index (BMI), blood pressure, blood lipids, blood glucose). We used linear regression models adjusted for age, race and ethnicity, and an index score of socioeconomic status comprising both individual and neighborhood metrics. Results: Among 678 participants, 68% were non-Hispanic White and 74% were college grads. In midlife, mean (SD) LE8 score was 76 (12) points. In unadjusted models, greater total %greenspace (per SD) was associated with higher overall CVH score (β 3.09; 95% CI 2.01, 4.17), as well as higher behavioral (β 2.71; CI 1.69, 3.73) and biomedical (β 3.09; CI 1.56, 4.62) sub-scores. Covariate adjustment attenuated these associations but they remained statistically significant for overall (β 1.88; CI 0.58, 3.19), behavioral (β 1.61; CI 0.44, 2.77) and biomedical (β 2.20; CI 0.31, 4.10) CVH scores. When disentangling the components of greenspace, %trees (adjβ 1.83; CI 0.54, 3.13) was associated with overall CVH, but %grass was non-significant (adjβ 0.75; CI -0.53, 2.02). Greater %trees was associated with higher scores for each CVH component in unadjusted models (except lipids): diet (β 1.64; CI 0.41, 2.86), physical activity (β 4.19; CI 1.61, 6.76), nicotine avoidance (β 2.07; CI 0.40, 3.75), sleep (β 3.50; CI 1.93, 5.08), BMI (β 6.51; CI 3.89, 9.14), blood pressure (β 2.39; CI 0.30, 4.48) and glucose (β 3.09; 1.36, 4.83). Covariate adjustment attenuated these associations to non-significance except for diet (β 2.33; CI 0.90, 3.75), BMI (β 3.08; 0.10, 6.06) and glucose (β 2.82; CI 0.73, 4.91). Conclusions: Greater street-view greenspace exposure, especially %trees, was associated with better CVH among women in midlife.