Related Experiment Video
Updated: Sep 10, 2026
A Novel Method for Involving Women of Color at High Risk for Preterm Birth in Research Priority Setting
Published on: January 12, 2018
Investigating the associations between prepregnancy lipoprotein(a) and birth outcomes: The Hispanic Community Health
Ashkan Habib1, Anna F Ballou1, Ayana April-Sanders2
1Department of Epidemiology, University of North Carolina, Chapel Hill, USA (Habib, Ballou, Van Wickle, Meyer, Graff, and Avery).
Background:
Lipoprotein(a) [Lp(a)] is an atherosclerotic risk factor that may influence adverse pregnancy complications, although mechanisms remain poorly understood. Evidence linking Lp(a) with pregnancy outcomes relies on small cross-sectional studies, nonuniform Lp(a) measurement across pregnancy, and isoform-sensitive assays.
Methods:
We identified 508 women with singleton live birth >43 weeks after baseline (2008-2011). Self-reported outcomes at second visit include: gestational diabetes mellitus (GDM, n = 80 [16%]), hypertensive disorders of pregnancy (HDP, n = 55 [11%]), preterm birth (PTB, n = 55 [11%]), small for gestational age (SGA, n = 37 [8%]), and birth weight Z-score (BWZ, n = 483). Adverse pregnancy outcome (APO, n = 179 [37%]) was defined as the presence of GDM, HDP, PTB, or SGA. Confounder-adjusted (age, kidney function, Hispanic/Latino background) logistic regression examined linear, curvilinear, quartile-categorized, and binary risk-threshold (≥75 nmol/L) patterns.
Results:
Baseline mean age was 26.4 (SD = 5.8), almost half were of Mexican background, and the Lp(a) median = 16.9 nmol/L (range: 2.8-415.1; IQR: 6.7-52.8 nmol/L; n = 93 [18%] with Lp(a) >75 nmol/L). Lp(a) >75 nmol/L had an odds ratio of 0.77 for HDP (95%CI: 0.31-1.92) and 0.86 for PTB (95%CI: 0.32-2.32), both imprecise and did not reach statistical significance. The curvilinear model, though imprecise, suggested an inverted U-shaped dose-response pattern for odds of GDM and HDP, while PTB and BWZ showed downward and upward trends, respectively, across increasing Lp(a). Other associations showed no clear pattern.
Conclusions:
Prepregnancy Lp(a) may exhibit a U-shaped dose-response with HDP and GDM, and a downward trend in PTB; however, there were no definite associations. Future work with larger samples is needed to understand the physiological implications of Lp(a) for pregnancy.
