Abstract
BACKGROUND: Suboptimal cardiovascular health (CVH) in pregnancy is associated with adverse maternal and offspring outcomes. To guide public health efforts to reduce disparities in maternal CVH, we determined the contribution of individual- and neighborhood-level factors to racial and ethnic differences in early pregnancy CVH. METHODS: We included nulliparous individuals with singleton pregnancies who self-identified as Hispanic, non-Hispanic Black (NHB), or non-Hispanic White (NHW) and participated in the nuMoM2b cohort study (Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be). First-trimester CVH was quantified using 6 routinely assessed factors in pregnancy included in the American Heart Association Life's Essential 8 score (0-100 points), in which higher scores indicate better CVH. Oaxaca-Blinder decomposition evaluated the extent to which racial and ethnic differences in CVH were explained by differences in individual- and neighborhood-level factors (age, socioeconomic characteristics, psychosocial factors, nativity, perceived racial discrimination, and area deprivation index). RESULTS: Among 9104 participants, the mean age was 26.8 years, 18.7% identified as Hispanic, 15.6% identified as NHB, and 65.8% identified as NHW. Mean (SD) CVH scores were 76.7 (14.1), 69.8 (15.1), and 79.9 (14.3) in the Hispanic, NHB, and NHW groups, respectively (P<0.01). The individual- and neighborhood-level factors evaluated explained all differences in CVH between Hispanic and NHW groups and 82% of differences between NHW and NHB groups. Racial and ethnic differences in educational attainment explained the greatest proportion of differences in CVH. If mean years of education among the Hispanic (14.0 [2.5]) and NHB (13.4 [2.4]) groups were the same as the NHW (15.8 [2.4]) group, mean CVH scores would be higher by 2.98 points (95% CI, 2.59-3.37) in the Hispanic and 4.28 points (95% CI, 3.77-4.80) in NHB groups. CONCLUSIONS: Racial and ethnic differences in early pregnancy CVH were largely explained by differences in individual- and neighborhood-level factors.