A 28-year-old woman arrived at the ED via emergency medical services sent by her primary care provider (PCP) due to unexplained asymptomatic tachycardia found during a regular follow-up for postpartum hypertension.
The patient was an otherwise healthy woman who was 9 weeks postpartum after a cesarean section for prolonged labor. She was breastfeeding her infant regularly without issue. Five days after delivery, she was admitted to the hospital for observation with preeclampsia concerns secondary to new-onset postpartum hypertension and tachycardia. She was discharged after 24 hours and placed on nifedipine 30 mg daily, which was tapered to every other day after 6 weeks of good compliance and close PCP follow-up. During a follow-up visit, she reported shortness of breath (SOB) and was found to have an elevated heart rate in the 120s, prompting her PCP to send her for emergency evaluation at the ED. She did not take any over-the-counter medications or supplements, had no surgical history other than cesarean delivery, no known allergies, or any history of tobacco or alcohol use. She also denied any significant family medical history.
On initial evaluation at this ED visit, she reported gradually worsening SOB and fatigue with everyday tasks such as putting away laundry. Her spouse added that she appeared to be increasingly tired without explanation. In the ED, she denied any current chest pain, SOB at rest, difficulty breathing, abdominal pain, headache, visual changes, or urinary symptoms.
Intake triage confirmed tachycardia with a heart rate of 120 beats/minute; her remaining vital signs, including a BP of 108/78 mm Hg, were within normal limits. Her cardiovascular examination was otherwise unremarkable. Her remaining physical examination was unremarkable, without any abnormalities of her head, eyes, ears, nose, throat, lungs, or abdomen.
- acute myocardial infarction
- pulmonary embolism (PE)
While laboratory tests were pending, the patient was started on 1 L of lactated Ringer solution for suspected hypovolemia. Her other vital signs remained unchanged. A complete blood cell count, thyroid-stimulating hormone, toxic ingestion panel (ethanol, salicylates, acetaminophen, urine drug screen), and high-sensitivity troponin were unremarkable. D-dimer was elevated to 827 ng/mL (normal range, less than 250 ng/mL). An initial ECG showed sinus tachycardia without signs of ischemia or other significant abnormalities (Figure 1).
The patient's urine was abnormally concentrated at more than 1.03 specific gravity, and urinalysis revealed high ketones (greater than 80 mg/dL), large hemoglobin, and trace protein. PE was considered given the patient's relatively recent hospitalization, period of immobilization, SOB, and elevated D-dimer. However, a subsequent CT pulmonary angiogram for possible PE found no evidence of thrombus in the pulmonary vasculature. Bedside venous blood gas analysis demonstrated a significant anion gap metabolic acidosis without respiratory compensation, revealing a dysregulation of blood pH due to a buildup of acidic products without correction from the increased respiratory rate.
Based on her current workup, no obvious reason could be identified as the cause of her metabolic acidosis. The patient was questioned further and revealed that she was on a new diet to “lose the baby weight.” This self-prescribed regimen included a severe calorie restriction over the previous 3 weeks, consisting of a small fruit shake for breakfast and only vegetables for all other meals. She estimated her daily intake to be less than 500 calories. Outside of her breakfast shake, she denied any protein intake. She reported losing 15 lb during this short time while continuing to breastfeed.
A hospitalist was consulted and suggested 1 L of 5% dextrose in 0.45% sodium chloride solution to normalize the anion gap. However, repeated metabolic laboratory tests showed a continued drop in the patient's pH and bicarbonate and increased plasma glucose to 358 mg/dL. In addition to an unimproved anion gap, the glucose result was concerning for the patient's inability to self-regulate glucose levels. Poor glucose control was discussed as secondary to insulin secretion or receptor dysfunction, possibly due to a broader refeeding syndrome-type response given the direct administration of IV dextrose. The ED and hospitalist teams agreed on admission for further workup, but the patient refused, left against medical advice, and was subsequently lost to follow-up.
Starvation ketoacidosis can occur in patients undergoing long fasts, and is rare and usually mild in nonpregnant patients.1,2 However, pregnancy-induced alterations in progesterone and cortisol exacerbate the starvation state, worsening the ketosis through increased insulin resistance, lipolysis, and free fatty acid mobilization.2,3 These changes can linger into the postpartum period, where women may seek to lose weight gained during pregnancy.2,3 An additional debt of up to 500 calories per day stemming from postpartum breastfeeding can exacerbate lipid breakdown from dietary restrictions.3,4 This lactation-induced ketoacidosis is sometimes referred to as bovine ketoacidosis in literature because it was first discovered in nursing cattle.5
Previous reports of postpartum starvation and bovine ketoacidosis note various nonspecific presentations with isolated complaints of mild dyspnea, tachypnea, or nausea with or without vomiting.5-7 Likewise, initial examination reveals isolated findings of tachypnea or tachycardia and is otherwise generally unremarkable. Laboratory testing is prompted by recent history of pregnancy and/or lactation with nonspecific examination findings. Dietary changes may not be inquired about often but are an additional critical input to the differential.
Early identification and optimal carbohydrate replacement therapy are critical for patients with postpartum starvation and bovine ketoacidosis.2 Limited literature supports the administration of low-concentrations of IV glucose (such as 5% dextrose in saline) to treat hypoglycemia and halt ketosis.7,8 Unlike diabetic ketoacidosis (DKA), postpartum starvation and bovine ketoacidosis generally are not conditions of hypoinsulinemia, and exogenous insulin administration is unnecessary if blood glucose levels are elevated. Exogenous insulin in a treatment style similar to DKA could cause or worsen electrolyte imbalances, specifically that of potassium and phosphate, leading to a refeeding syndrome response.8 Disposition of the patient between admission and outpatient treatment is multifactorial based on severity of presentation, response to initial treatment, and reliability for follow-up.
Starvation ketoacidosis and bovine ketoacidosis both are possible in postpartum patients due to lingering hormonal changes exacerbating significant caloric restrictions and expenditures. Although rare, starvation ketoacidosis and bovine ketoacidosis should be considered in breastfeeding postpartum patients presenting with nonspecific symptoms like fatigue, dyspnea, or nausea. History and specifically questions of postpartum diet and weight changes are critical to suspicion and diagnosis.
1. Frise CJ, Mackillop L, Joash K, Williamson C. Starvation ketoacidosis in pregnancy. Eur J Obstet Gynecol Reprod Biol
2. Hui L, Shuying L. Acute starvation ketoacidosis in pregnancy with severe hypertriglyceridemia: a case report. Medicine (Baltimore)
3. Greaney DJ, Benson P. Life-threatening lactation or “bovine” ketoacidosis: a case report. A A Case Rep
4. Tigas S, Sunehag A, Haymond MW. Metabolic adaptation to feeding and fasting during lactation in humans. J Clin Endocrinol Metab
5. Heffner AC, Johnson DP. A case of lactation “bovine” ketoacidosis. J Emerg Med
6. Hudak SK, Overkamp D, Wagner R, et al. Ketoacidosis in a non-diabetic woman who was fasting during lactation. Nutr J
7. von Geijer L, Ekelund M. Ketoacidosis associated with low-carbohydrate diet in a non-diabetic lactating woman: a case report. J Med Case Rep
8. Hammerbeck H, Holland MR. Starvation ketoacidosis: treatment pitfalls. J Intensive Care Soc