Association between serum alkaline phosphatase levels in late pregnancy and the incidence of venous thromboembolism postpartum: a retrospective cohort study.

Background: Two previous studies found alkaline phosphatase (ALP) levels were related with the development of venous thromboembolism (VTE) in hospitalised patients. VTE is a leading cause of death during pregnancy and postpartum. No prior study has investigated the associations of ALP levels and VTE postpartum, and the related mechanisms remain unclear. This study aimed to investigate the associations between ALP levels and VTE postpartum, and to reveal the potential mechanisms.

Methods: In this retrospective cohort study, we included pregnant women who planned to deliver at the Department of Obstetrics and Gynecology in the three designated hospitals in a multicentre cohort of pregnant women in Wuhan, China, during two recruitment periods of January 1, 2018 to December 31, 2019, and May 14, 2020 to March 25, 2022. A total of 10,044 participants with serum ALP and whole blood hemoglobin measurements in late pregnancy (median, 37 (35, 39) weeks) were enrolled. The participants' incidences of VTE (deep venous thrombosis and/or pulmonary embolism) postpartum were confirmed from the medical records. Pregnant women with new-onset VTE postpartum (within 6 weeks after delivery) were confirmed as VTE cases.

Findings: Approximately 0.8% (79/10,044) of the pregnant women were diagnosed with VTE postpartum. In the unadjusted model, pregnant women with the lowest quintile of serum ALP levels (≤116 U/L) in late pregnancy had higher risk of VTE postpartum compared with those with the highest quintile (≥199 U/L) (OR, 2.83 [1.32, 6.05]). After adjusting for covariates of demographic, life style, birth outcomes, and other liver enzymes, pregnant women with the lowest quintile of serum ALP levels (≤116 U/L) in late pregnancy had increased risk of VTE postpartum compared with those with the highest quintile (≥199 U/L) (OR, 2.48 [1.14, 5.40]). A one standard deviation decrease of ln-transformed ALP levels were associated with elevated risk of VTE postpartum (OR, 1.29 [1.02, 1.62]). Significant negative associations of ALP with VTE were found in the unadjusted and adjusted models. The negative associations between ALP and VTE remained consistent in sensitivity analyses among participants with non-GDM, single pregnancy, non-preeclampsia, non-postpartum hemorrhage, non-extremely/very preterm and cesarean delivery. Decreased serum ALP levels significantly ( < 0.05) related to decreased hemoglobin, which was significantly ( < 0.05) related to increased risk of VTE postpartum. Decreased hemoglobin significantly ( < 0.05) mediated 7.59% of ALP-associated VTE postpartum.

Interpretation: This study suggested that low serum ALP levels in late pregnancy were associated with increased risk of VTE postpartum, and the ALP-associated VTE risk may be partially mediated by hemoglobin, suggesting that serum ALP in late pregnancy could be a promising biomarker for the prediction of VTE postpartum.

Funding: The National Natural Science Foundation of China, and the Program for HUST Academic Frontier Youth Team.