(355) Body composition vs. traditional weight and BMI for assessing hypertensive disorders of pregnancy (HDOP) risk

Clinical history (hx) and BMI are established risk factors for HDOP. We sought to evaluate if incorporation of body composition parameters improves HDOP prediction compared to traditional clinical hx and weight/BMI.

 

Study Design:

 Primary analysis of a prospective cohort. Pregnant individuals at high risk for medically-indicated or spontaneous PTB were recruited from a MFM outpatient clinic, 2021-2023. Direct segmental multi-frequency bioelectrical impedance analysis technology measured each participant’s body composition (% body fat, skeletal muscle mass, total body water, lean body mass) up to 4 times during pregnancy using the non-invasive InBody 270 scale-type device. Those with ≥ 2 InBody measurements were included in this analysis. The primary outcome was development of HDOP. Logistic regression models were constructed to select the top: (1) clinical hx (2) clinical hx + weight/BMI; (3) clinical hx + body composition parameters to predict HDOP.  Area under the receiver operating characteristics curves (AUC) were used to select the model that best predicted HDOP.


Results:

88 study participants were included; 18 (20.5%) developed HDOP. The cohort was racially and ethnically diverse (38% Black, 9% Latinx); most (93%) were multiparous. First InBodys were at a median 18.2 (IQR 14.7-22.4) wks’ and last Inbodys were at a median 29.1 (IQR 25.3, 32.7) wks.’ The most predictive clinical factor for HDOP was hx of preeclampsia (AUC 0.64, 95% CI 0.52, 0.76). The most predictive weight/BMI factor was BMI change between the first & last InBody; adding this to hx of preeclampsia improved model prediction (AUC 0.809, 95% CI 0.694, 0.923) vs. clinical hx, p=0.009. The most predictive body comp factor was total body water change between the first & last InBody; adding this to hx of preeclampsia also improved model prediction (AUC 0.816, 95% CI 0.706, 0.926) vs. clinical hx, p=0.003, Figure. The BMI and total body water models similarly predicted HDOP (p=0.93).


Conclusion:

Adding mid-pregnancy changes in BMI or total body water to clinical factors significantly improves prediction of HDOP beyond clinical factors.