Metabolic markers for risk stratification of diabetic ketoacidosis and perinatal mortality in pregestational diabetes mellitus

OBJECTIVE: This study aimed to identify metabolic markers associated with diabetic ketoacidosis (DKA) in pregnant women with pregestational diabetes mellitus (PDM) and to evaluate their utility in risk stratification and the prediction of perinatal mortality. MATERIALS AND METHODS: A retrospective cohort study was conducted from 2000 to 2022, including singleton pregnancies in women with PDM. Maternal metabolic parameters such as glucose levels, hemoglobin A1c (HbA1c), lipid profiles (high-density lipoprotein [HDL] and low-density lipoprotein [LDL]), arterial pH, electrolytes, and anion gap were analyzed. Receiver operating characteristic (ROC) curves and logistic regression models were used to identify metabolic predictors of DKA and perinatal mortality. RESULTS: Among 219 pregnant women with PDM, 21 (9.6%) developed DKA. Affected women had significantly higher glucose levels (395.33 vs. 126.91 mg/dL, P < 0.001), higher HbA1c (8.45% vs. 6.73%, P = 0.001), lower HDL (38.71 vs. 57.96 mg/dL, P < 0.001), and higher LDL (152.86 vs. 119.25 mg/dL, P = 0.012) compared with those without DKA. Perinatal mortality occurred in 6 cases (28.6%) and was associated with metabolic acidosis, including lower arterial pH (7.07 vs. 7.20, P = 0.030), lower bicarbonate (3.90 vs. 9.96 mEq/L, P = 0.001), and a wider anion gap (22.50 vs. 15.17 mEq/L, P = 0.014). Additionally, higher HbA1c (10.00% vs. 7.83%, P = 0.026) and hyperkalemia (5.77 vs. 4.23 mEq/L, P = 0.022) were associated with perinatal mortality. The ROC analyses demonstrated strong predictive value for DKA (area under the ROC curve [AUC] = 0.99) and perinatal mortality (AUC = 0.98) using combined metabolic markers. The logistic regression analyses identified higher HbA1c (adjusted odds ratio [aOR] 2.32, 95% confidence interval [CI] 1.13-4.75), lower HDL (aOR 0.89, 95% CI 0.81-0.96), and higher glucose (aOR 1.03, 95% CI 1.02-1.05) as independent predictors of DKA. Lower bicarbonate (aOR 0.58, 95% CI 0.35-0.96) was independently associated with an increased risk of perinatal mortality. The Kaplan-Meier analysis indicated a significant association between bicarbonate <6.50 mEq/L and increased perinatal mortality (P < 0.001). CONCLUSIONS: Key metabolic markers can aid with risk stratification for DKA and perinatal mortality in women with PDM, supporting early identification and guiding targeted clinical interventions.