Pregnancy and postpartum dynamics revealed by millions of lab tests

Abstract

Pregnancy and delivery involve dynamic alterations in many physiological systems. However, the physiological dynamics during pregnancy and after delivery have not been systematically analyzed at high temporal resolution in a large human population. Here, we present the dynamics of 76 lab tests based on a cross-sectional analysis of 44 million measurements from over 300,000 pregnancies. We analyzed each test at weekly intervals from 20 weeks preconception to 80 weeks postpartum, providing detailed temporal profiles. About half of the tests take 3 months to a year to return to baseline postpartum, highlighting the physiological load of childbirth. The precision of the data revealed effects of preconception supplements, overshoots after delivery and intricate temporal responses to changes in blood volume and renal filtration rate. Pregnancy complications-gestational diabetes, preeclampsia, and postpartum hemorrhage-showed distinct dynamical changes. These results provide a comprehensive dynamic portrait of the systems physiology of pregnancy.

Fig. 1.. Dataset of lab tests over preconception, pregnancy, and postpartum.

(A) Schematic overview of the dataset. (B) Alkaline phsophatase (ALP) test (mean of the quantile-transformed values; see Methods) over 140 weeks. Error bars are SEM. Units are the standard for the test (IU/liter). m, months; y, year. (C) Seventy-six test values over 140 weeks; the period of pregnancy is in gray. For units and full test names, see table S1. Created with BioRender.

Fig. 2.. Lab test dynamics vary during pregnancy and postpartum and can show overshoots and undershoots.

(A) Relative dynamic range across the study period (Methods). Dynamic range (max-min)/(preconception average) is roughly proportional to the coefficient of variation (CV) of the test value in the reference population. r_p is the Pearson correlation. The dashed line within the 95% confidence interval has slope 1, R² = 0.68. (B) Histogram of the dynamic range of quantile scores of the 76 tests. (C to F) Four clusters of ranked test dynamical profiles. In gray are ranked individual tests; colored lines are cluster means. (G and H) Theoretical profiles for first-order response. (I and J) Theoretical profiles for a system with a slow compensatory mechanism. (K) Physiological circuits governing response to load and recovery. Full model found in Supplementary Text. (a) Circuit in which the load of pregnancy affects test X as a first-order system. (b) Circuit in which the load affects test Y with X as a compensatory system. (L) Pregnancy load in the theoretical model rises in pregnancy and drops abruptly at delivery. (M) First-order system X responds with no undershoot. (N) Compensated system Y shows an undershoot and a rebound effect.

Fig. 3.. Postpartum recovery times of tests range between days and a year.

(A) Dimensionality reduction of test mean values as a function of time using PCA shows that the trajectory during pregnancy (blues) differs from the postpartum trajectory (orange and red). Each point is a week interval, progression clockwise from 60 weeks before delivery, through conception and delivery, to the end point 80 weeks after delivery. (B) Settling time is defined as the time after which the test remains within 0.2 SDs of its baseline. (C) Settling time of the tests in the dataset.

Fig. 4.. Tests affected by health behaviors show preconception dynamics.

(A to C) Folic acid, CRP, and albumin are examples of a strong preconception change. Mean conception time is indicated by a dashed line. The inset highlights the preconception period, red is regression with 95% confidence intervals. (D) Temporal slope of test values during preconception from linear regression. Tests are arranged by physiological system; tests in bold color have significant nonzero slope (>0.11 in absolute value and P < 0.05 adjusted for multiple comparisons, Benjamini-Hochberg). Error bars are 95% confidence intervals of slope.

Fig. 5.. Dynamics of lab tests in complications of pregnancy.

Tests with significant differences from healthy pregnancy (green) are marked by Pe for preeclampsia (blue), PH for PPH (purple), and GD for gestational diabetes (orange). The y axis is different between tests; for units, see table S1. For the same figure with quantile scores, see fig. S5. Created with BioRender.

Fig. 6.. Tests with significant deviations from healthy pregnancies in preeclampsia, gestational diabetes, and PPH.

(A to I) Volcano plots of −log P value (FDR corrected) versus effect size for each test at each time point. See full test names in table S1. Significant tests are marked with a red dot and their name (Methods). (J to L) Venn diagrams showing the significant tests in the preconception (blue), gestation (green), and postpartum (red) periods for each complication. Created with BioRender.