Ketoacidosis is a high-anion-gap metabolic acidosis caused by accumulation of ketone bodies, principally β-hydroxybutyrate and acetoacetate, when carbohydrate availability or insulin action is inadequate. In emergency practice, diabetic ketoacidosis (DKA) is the most familiar form and remains one of the major acute metabolic emergencies associated with diabetes. It develops through absolute or relative insulin deficiency together with increased counter-regulatory hormones, leading to accelerated lipolysis, hepatic ketogenesis, hyperglycaemia, osmotic diuresis, volume depletion, and electrolyte disturbance [1-4]. Although contemporary treatment has substantially reduced mortality, DKA continues to generate considerable emergency and inpatient care burden, and population studies have documented persistent or increasing hospitalization rates in several settings [5,9]. The diagnosis of DKA relies on identifying diabetes or hyperglycaemia in association with clinically significant ketonaemia and metabolic acidosis. Current consensus guidance emphasizes quantitative β-hydroxybutyrate measurement because it directly reflects the predominant circulating ketone in DKA and can improve diagnostic specificity compared with urine ketone testing [1,7,8]. Typical manifestations include polyuria, polydipsia, nausea, vomiting, abdominal pain, dehydration, tachycardia, and, in more severe cases, altered mental status. Infection, omission or inadequate administration of insulin, newly diagnosed diabetes, and acute physiological stress remain common precipitating factors [2,3,6]. The clinical spectrum is broad, and hyperglycaemia alone does not define the severity of the acid-base disturbance. Starvation ketoacidosis (SKA) represents a different metabolic pathway. With sustained reduction in carbohydrate intake, declining insulin concentrations and increased glucagon activity promote lipolysis and ketone production to provide an alternative energy substrate. Physiological fasting commonly produces only modest ketonaemia, but prolonged caloric deprivation, persistent vomiting, intercurrent illness, ketogenic dietary practices, or other catabolic stressors can produce clinically important ketoacidosis [10-13]. Unlike classic DKA, SKA usually occurs with normal, low, or mildly elevated blood glucose. Its symptoms can overlap substantially with DKA, including vomiting, weakness, abdominal discomfort, dehydration, tachypnoea, and altered sensorium. This overlap creates an important diagnostic challenge in the Emergency Department, particularly when the nutritional history is incomplete or when several causes of high-anion-gap metabolic acidosis coexist [10,11,13]. Direct comparative clinical data describing DKA and SKA within the same emergency population remain limited. Recognizing the biochemical distinctions is clinically relevant because immediate management differs: DKA requires structured insulin, fluid, and electrolyte replacement, whereas carbohydrate replacement and correction of volume and electrolyte deficits are central in uncomplicated SKA [4,10,14]. Accordingly, the present study aimed to compare the clinical characteristics and admission biochemical profiles of patients with DKA and SKA presenting to the Emergency Department of a tertiary-care hospital. Secondary objectives were to describe precipitating factors, characterize the severity of metabolic acidosis, and identify routinely available features that assist early differentiation between the two ketoacidotic states.

Study design and setting: This prospective observational comparative study was conducted in the Emergency Department of Manipal Hospital, Varthur Road, Bengaluru, Karnataka, India. The study was designed to characterize and compare patients presenting with diabetic ketoacidosis (DKA) and starvation ketoacidosis (SKA) using clinical findings and biochemical measurements obtained during the initial emergency evaluation. Study period and participants: Recruitment was undertaken from February 2026 to June 2026. Consecutive adult patients aged 18 years or older who presented to the Emergency Department with suspected ketoacidosis and had biochemical confirmation of ketonaemia with metabolic acidosis were screened. Patients were excluded when essential biochemical investigations were incomplete or when the high-anion-gap metabolic acidosis was attributable to another dominant cause, including significant lactic acidosis, advanced renal failure, toxic ingestion, or another clearly established metabolic disorder. Cases with alcoholic ketoacidosis or an indeterminate/mixed ketoacidotic state were not assigned to either comparison group. Diagnostic classification: DKA was diagnosed using contemporary adult criteria requiring diabetes or hyperglycaemia together with significant ketonaemia and metabolic acidosis; quantitative β-hydroxybutyrate was prioritized where available [1,7,8]. In operational terms, the DKA group comprised patients with diabetes or hyperglycaemia, β-hydroxybutyrate ≥3.0 mmol/L, and pH <7.30 and/or serum bicarbonate <18 mmol/L. SKA was defined as ketonaemic high-anion-gap metabolic acidosis occurring in the setting of documented substantial reduction in caloric intake or persistent vomiting, with normal or low-range glucose and without evidence supporting DKA, alcoholic ketoacidosis, or another major cause of metabolic acidosis [10-13]. The clinical history, medication exposure, diabetes status, HbA1c, and precipitating illness were considered together when assigning the final diagnosis. Data collection: Demographic characteristics, presenting symptoms, vital signs, hydration status, known or newly diagnosed diabetes, and probable precipitating factors were recorded. Initial laboratory variables obtained before definitive metabolic treatment included random blood glucose, HbA1c, blood pH, serum bicarbonate, anion gap, β-hydroxybutyrate, sodium, potassium, chloride, creatinine, blood urea nitrogen, lactate, and calculated serum osmolality. DKA severity was categorized according to admission pH as mild (7.25-7.30), moderate (7.00-7.24), or severe (<7.00), consistent with established clinical severity frameworks [1,2]. Statistical analysis: Continuous variables were summarized as mean ± standard deviation and categorical variables as frequency and percentage. Independent-group continuous variables were compared using an appropriate two-sample test, while categorical variables were evaluated using the chi-square test or Fisher's exact test when expected cell counts were small. All tests were two-sided, and P <0.05 was considered statistically significant. Ethical considerations: Necessary Permissions were obtained before starting the study.

A total of 108 patients presenting to the Emergency Department with biochemical evidence of ketoacidosis were assessed for eligibility. Eight patients were excluded because of incomplete biochemical investigations or an alternative cause of high-anion-gap metabolic acidosis. Thus, 100 patients were included in the final analysis. Of these, 70 (70.0%) were diagnosed with diabetic ketoacidosis (DKA) and 30 (30.0%) with starvation ketoacidosis (SKA). The mean age of the overall study population was 42.6 ± 15.8 years, and 56 (56.0%) patients were male. Patients with DKA were slightly older than those with SKA, although the difference was not statistically significant (44.1 ± 15.2 vs 39.1 ± 16.8 years; P = 0.148). The baseline and presenting clinical characteristics are summarized in Table 1. Table 1. Baseline and clinical characteristics of patients with diabetic and starvation ketoacidosis Characteristic DKA (n = 70) SKA (n = 30) P value Age, years, mean ± SD 44.1 ± 15.2 39.1 ± 16.8 0.148 Male sex, n (%) 40 (57.1) 16 (53.3) 0.726 Female sex, n (%) 30 (42.9) 14 (46.7) — Nausea/vomiting, n (%) 52 (74.3) 19 (63.3) 0.266 Abdominal pain, n (%) 39 (55.7) 13 (43.3) 0.255 Generalized weakness, n (%) 38 (54.3) 24 (80.0) 0.016 Polyuria/polydipsia, n (%) 45 (64.3) 3 (10.0) <0.001 Altered sensorium, n (%) 13 (18.6) 2 (6.7) 0.128 Clinical dehydration, n (%) 55 (78.6) 19 (63.3) 0.111 Tachycardia, n (%) 48 (68.6) 16 (53.3) 0.144 Hypotension at presentation, n (%) 15 (21.4) 4 (13.3) 0.348 Data are presented as mean ± SD or n (%). DKA, diabetic ketoacidosis; SKA, starvation ketoacidosis; SD, standard deviation. Nausea and vomiting were the most frequent presenting manifestations in the DKA group, occurring in 52 (74.3%) patients, followed by tachycardia in 48 (68.6%) and polyuria/polydipsia in 45 (64.3%). Generalized weakness was more frequent in SKA (80.0% vs 54.3%; P = 0.016), whereas polyuria/polydipsia was markedly more frequent in DKA (64.3% vs 10.0%; P <0.001). Altered sensorium, dehydration, tachycardia, and hypotension were numerically more frequent in DKA, but the between-group differences were not statistically significant (Table 1). Among patients with DKA, 27 (38.6%) had newly diagnosed diabetes and 43 (61.4%) had previously known diabetes mellitus. Infection was the most frequently identified precipitating factor for DKA, followed by missed or inadequate insulin therapy. In SKA, prolonged poor oral intake was the predominant precipitating factor, followed by persistent vomiting. The distribution of precipitating factors is shown in Table 2. Table 2. Precipitating factors associated with ketoacidosis Group Precipitating factor n (%) DKA Infection 26 (37.1) DKA Missed/inadequate insulin therapy 18 (25.7) DKA Newly diagnosed diabetes without another identified precipitant 12 (17.1) DKA Acute gastrointestinal illness 7 (10.0) DKA Other acute medical stressors 4 (5.7) DKA No clear precipitating factor identified 3 (4.3) SKA Prolonged poor oral intake 14 (46.7) SKA Persistent vomiting 8 (26.7) SKA Acute gastrointestinal illness 4 (13.3) SKA Reduced intake associated with intercurrent illness 3 (10.0) SKA Other cause 1 (3.3) DKA, diabetic ketoacidosis; SKA, starvation ketoacidosis. Marked biochemical differences were observed between the groups. Mean random blood glucose was substantially higher in DKA than in SKA (392.6 ± 118.4 vs 87.9 ± 21.6 mg/dL; P <0.001). HbA1c was likewise higher in DKA (10.3 ± 2.1% vs 5.4 ± 0.5%; P <0.001). Patients with DKA had more pronounced acidaemia, with a lower mean pH (7.18 ± 0.08 vs 7.29 ± 0.06; P <0.001) and lower serum bicarbonate (11.3 ± 4.0 vs 15.7 ± 3.6 mmol/L; P <0.001). Detailed biochemical comparisons are presented in Table 3. Table 3. Biochemical profile of diabetic versus starvation ketoacidosis Biochemical parameter DKA (n = 70) SKA (n = 30) P value Random blood glucose, mg/dL 392.6 ± 118.4 87.9 ± 21.6 <0.001 HbA1c, % 10.3 ± 2.1 5.4 ± 0.5 <0.001 Blood pH 7.18 ± 0.08 7.29 ± 0.06 <0.001 Serum bicarbonate, mmol/L 11.3 ± 4.0 15.7 ± 3.6 <0.001 Anion gap, mmol/L 24.8 ± 5.7 20.5 ± 4.4 <0.001 β-hydroxybutyrate, mmol/L 5.7 ± 1.8 4.3 ± 1.5 <0.001 Serum sodium, mmol/L 134.1 ± 5.8 137.0 ± 4.9 0.019 Serum potassium, mmol/L 4.9 ± 0.9 4.2 ± 0.7 <0.001 Serum chloride, mmol/L 99.2 ± 6.1 101.4 ± 5.2 0.087 Serum creatinine, mg/dL 1.42 ± 0.64 0.93 ± 0.34 <0.001 Blood urea nitrogen, mg/dL 27.8 ± 12.6 20.1 ± 9.4 0.004 Serum lactate, mmol/L 2.2 ± 1.0 1.9 ± 0.8 0.157 Calculated serum osmolality, mOsm/kg 307.4 ± 15.6 285.9 ± 10.8 <0.001 Values are mean ± SD. DKA, diabetic ketoacidosis; SKA, starvation ketoacidosis; HbA1c, glycated haemoglobin. The mean anion gap was greater in DKA than in SKA (24.8 ± 5.7 vs 20.5 ± 4.4 mmol/L; P <0.001), and β-hydroxybutyrate was also higher (5.7 ± 1.8 vs 4.3 ± 1.5 mmol/L; P <0.001). DKA was associated with higher serum potassium, creatinine, blood urea nitrogen, and calculated serum osmolality. Serum sodium was lower in DKA, whereas serum chloride and lactate did not differ significantly between groups (Table 3). Severity classification based on admission pH showed that moderate-to-severe metabolic acidosis was present in 51 of 70 patients with DKA (72.9%) compared with 12 of 30 patients with SKA (40.0%). Severe acidosis with pH <7.00 occurred in eight patients with DKA and in none of the SKA patients. The overall distribution of acidosis severity differed significantly between the two groups (P <0.001), as shown in Table 4. Table 4. Severity of metabolic acidosis among the study groups Severity based on pH DKA (n = 70), n (%) SKA (n = 30), n (%) P value Mild (pH 7.25-7.30) 19 (27.1) 18 (60.0) <0.001 Moderate (pH 7.00-7.24) 43 (61.4) 12 (40.0) — Severe (pH <7.00) 8 (11.4) 0 (0.0) — The P value refers to the overall between-group comparison across severity categories. DKA, diabetic ketoacidosis; SKA, starvation ketoacidosis. Overall, DKA showed a pattern of marked hyperglycaemia, poorer chronic glycaemic control, greater anion-gap disturbance, lower bicarbonate and pH, higher β-hydroxybutyrate, and greater abnormalities in renal indices and calculated serum osmolality. SKA was associated predominantly with caloric deprivation or persistent vomiting and showed normal-range glucose with a comparatively less severe acid-base disturbance.

The present study demonstrates clinically relevant differences between diabetic and starvation ketoacidosis among patients evaluated in the Emergency Department. DKA accounted for 70% of the analysed cases and was characterized by marked hyperglycaemia, elevated HbA1c, a wider anion gap, higher β-hydroxybutyrate, lower pH and bicarbonate, and greater derangement of renal indices and serum osmolality. SKA, although biochemically less severe on average, still produced substantial ketonaemia and metabolic acidosis despite normal-range glucose concentrations. These findings reinforce the need to consider more than glucose concentration when assessing a patient with high-anion-gap metabolic acidosis [1,5,10]. The clinical pattern observed in the DKA group agrees with previous studies. Nausea and vomiting occurred in 74.3%, while polyuria/polydipsia, dehydration, tachycardia, and abdominal pain were frequent. Singh et al. reported nausea and vomiting in 74% of DKA cases and found mean admission glucose of 406.8 mg/dL, serum ketones of 5.38 mmol/L, and pH of 7.128, values broadly comparable with the present cohort [6]. Infection was the most frequent identified precipitant in our DKA group, followed by missed or inadequate insulin therapy. Both triggers are well established in adult DKA literature and remain important targets for prevention through infection control, treatment adherence, and structured diabetes education [2-4]. The contrast with SKA was most apparent in glycaemic measures. Mean glucose in SKA was 87.9 mg/dL and mean HbA1c was 5.4%, despite a mean β-hydroxybutyrate concentration of 4.3 mmol/L and anion gap of 20.5 mmol/L. Reports of non-diabetic and starvation ketoacidosis similarly emphasize that clinically important ketonaemia can occur with euglycaemia after inadequate carbohydrate intake, prolonged fasting, or persistent vomiting [10-13]. Generalized weakness was more frequent in SKA in the present study, whereas polyuria and polydipsia were strongly concentrated in DKA, providing potentially useful historical clues before laboratory classification is complete. Quantitative β-hydroxybutyrate was elevated in both groups and was significantly higher in DKA. Previous emergency-department studies have shown that blood β-hydroxybutyrate improves specificity for DKA compared with urine ketone testing, although its absolute concentration does not independently capture the full severity of the acid-base disturbance [7,8]. Therefore, interpretation alongside pH, bicarbonate, anion gap, glucose, renal function, and the clinical context remains essential. The higher creatinine, urea, potassium, and osmolality in DKA in this study are compatible with greater osmotic diuresis and volume depletion associated with marked hyperglycaemia [1,3]. From an emergency-care perspective, early etiological differentiation has therapeutic importance. DKA requires protocolized fluid resuscitation, insulin administration, potassium surveillance and replacement, and treatment of the precipitating illness [1,4,14]. In SKA, restoration of carbohydrate availability, fluids, electrolyte correction, and assessment for coexisting alcohol-related or other metabolic causes are central [10,11]. Thus, a combined assessment of nutritional history, diabetes status, glucose, HbA1c, β-hydroxybutyrate, and acid-base variables offers a practical approach to distinguishing these overlapping syndromes. LIMITATIONS This study has several limitations. It was conducted at a single tertiary-care Emergency Department over five months, limiting external generalizability. The sample size, particularly for starvation ketoacidosis, was modest. Diagnostic classification of starvation ketoacidosis depended partly on clinical history and exclusion of competing causes. Serial biochemical changes, treatment response, intensive-care requirement, length of stay, recurrence, and mortality were not evaluated. Residual confounding from unmeasured nutritional and comorbidity factors remains possible.

Diabetic and starvation ketoacidosis present with overlapping emergency symptoms but show distinct biochemical signatures. DKA in this cohort was associated with marked hyperglycaemia, higher HbA1c, greater ketonaemia, wider anion-gap disturbance, lower pH and bicarbonate, and greater abnormalities in renal indices and osmolality. Starvation ketoacidosis occurred predominantly with poor caloric intake or persistent vomiting and showed substantial ketosis despite normal-range glucose. Clinical history alone is insufficient for reliable distinction. Emergency evaluation should integrate glucose, β-hydroxybutyrate, pH, bicarbonate, anion gap, HbA1c, electrolytes, renal function, and recent nutritional intake. Early etiological classification supports prompt selection of appropriate metabolic treatment and reduces the risk of overlooking non-diabetic causes of ketoacidosis during initial emergency assessment.

1. Umpierrez GE, Davis GM, ElSayed NA, Fadini GP, Galindo RJ, Hirsch IB, et al. Hyperglycemic crises in adults with diabetes: a consensus report. Diabetes Care. 2024;47(8):1257-1275. doi:10.2337/dci24-0032. 2. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic crises in adult patients with diabetes. Diabetes Care. 2009;32(7):1335-1343. doi:10.2337/dc09-9032. 3. Umpierrez G, Korytkowski M. Diabetic emergencies-ketoacidosis, hyperglycaemic hyperosmolar state and hypoglycaemia. Nat Rev Endocrinol. 2016;12(4):222-232. doi:10.1038/nrendo.2016.15. 4. Dhatariya KK; Joint British Diabetes Societies for Inpatient Care. The management of diabetic ketoacidosis in adults-an updated guideline from the Joint British Diabetes Society for Inpatient Care. Diabet Med. 2022;39(6):e14788. doi:10.1111/dme.14788. 5. Rodriguez Alvarez P, San Martin VT, Morey-Vargas OL. Hyperglycemic crises in adults: a look at the 2024 consensus report. Cleve Clin J Med. 2025;92(3):152-158. doi:10.3949/ccjm.92a.24089. 6. Singh H, Saroch A, Pannu AK, Sachin HJ, Sharma N, Dutta P. Clinical and biochemical profile, precipitants and prognostic factors of diabetic ketoacidosis: a retrospective study from a tertiary care center of north India. Diabetes Metab Syndr. 2019;13(4):2357-2360. doi:10.1016/j.dsx.2019.06.002. 7. Arora S, Henderson SO, Long T, Menchine M. Diagnostic accuracy of point-of-care testing for diabetic ketoacidosis at emergency-department triage: beta-hydroxybutyrate versus the urine dipstick. Diabetes Care. 2011;34(4):852-854. doi:10.2337/dc10-1844. 8. Sheikh-Ali M, Karon BS, Basu A, Kudva YC, Muller LA, Xu J, et al. Can serum beta-hydroxybutyrate be used to diagnose diabetic ketoacidosis? Diabetes Care. 2008;31(4):643-647. doi:10.2337/dc07-1683. 9. Desai D, Mehta D, Mathias P, Menon G, Schubart UK. Health care utilization and burden of diabetic ketoacidosis in the U.S. over the past decade: a nationwide analysis. Diabetes Care. 2018;41(8):1631-1638. doi:10.2337/dc17-1379. 10. Bashir B, Fahmy AA, Raza F, Banerjee M. Non-diabetic ketoacidosis: a case series and literature review. Postgrad Med J. 2021;97(1152):667-671. doi:10.1136/postgradmedj-2020-138513. 11. Gall AJ, Duncan R, Badshah A. Starvation ketoacidosis on the acute medical take. Clin Med (Lond). 2020;20(3):298-300. doi:10.7861/clinmed.2020-0012. 12. Burstal RJ, Reilly JR, Burstal B. Fasting or starving? Measurement of blood ketone levels in 100 fasted elective and emergency adult surgical patients at an Australian tertiary hospital. Anaesth Intensive Care. 2018;46(5):463-467. doi:10.1177/0310057X1804600506. 13. Shaikh A, Williams DM, Stephens JW, Edwards R. Starvation ketocidosis on the acute medical take: an easily missed complication of the keto diet. Eur J Case Rep Intern Med. 2024;11(9):004830. doi:10.12890/2024_004830. 14. Barski L, Golbets E, Jotkowitz A, Schwarzfuchs D. Management of diabetic ketoacidosis. Eur J Intern Med. 2023;117:38-44. doi:10.1016/j.ejim.2023.07.005.