Caffeine Overdose-Induced Diabetic Ketoacidosis With Gastrointestinal Injury in Undiagnosed Type 2 Diabetes

Abstract

Caffeine is a widely consumed stimulant contained in beverages and supplements, and intentional overdose is becoming increasingly common. Although hyperglycemia is often seen in caffeine toxicity, diabetic ketoacidosis (DKA) is a rare complication, particularly in those without a known history of diabetes. We report the case of a 46-year-old man without a known medical history who ingested 10 g of caffeine to attempt suicide. He presented with impaired consciousness, arrhythmia, hyperglycemia, and metabolic acidosis. Laboratory testing confirmed DKA, and endoscopy revealed esophageal severe erosions and a duodenal ulcer. He was managed with IV fluids, insulin infusion, sedation, potassium-competitive acid blocker, and β-blocker therapy. After stabilization, he was diagnosed with type 2 diabetes mellitus. The condition was characterized by insulin resistance and preserved endogenous insulin secretion, which allowed discontinuation of insulin therapy. He recovered fully and was discharged without diabetic medications. This case highlights the importance of considering DKA in patients with caffeine overdose, especially those with unrecognized diabetes, and the potential for gastrointestinal injury. Clinicians should be aware of these rare but serious complications to ensure prompt diagnosis and management.

Introduction

Caffeine is a widely consumed central nervous system stimulant contained in coffee, tea, energy drinks, and over-the-counter supplements. While moderate consumption is generally safe, excessive intake can result in life-threatening toxicity, including arrhythmia, seizures, and metabolic derangements [1]. Caffeine exerts dose-dependent effects. In most adults, daily intake up to 400 mg is considered safe [2, 3]. Toxicity has been reported at doses of approximately 1.2 g, while the estimated lethal dose ranges from 10 to 14 g [4]. Plasma caffeine concentrations of 3 to 6 mg/L are regarded as safe, whereas levels above 15 mg/L are associated with severe intoxication, and concentrations exceeding 80 mg/L are considered life-threatening [1]. The accessibility of high-dose caffeine supplements through online markets has led to a rise in intentional overdoses, often associated with suicide attempts [5]. Hyperglycemia is reported in approximately 47% of caffeine intoxication cases [5]. However, diabetic ketoacidosis (DKA) as a complication is rare and has only been reported in individuals with type 1 diabetes [6, 7]. There are no previous reports of DKA following caffeine overdose in patients with type 2 diabetes mellitus (T2DM). We experienced a unique case of DKA triggered by caffeine toxicity in a patient with undiagnosed T2DM, accompanied by drug-induced gastrointestinal injury. The case underscores the need for awareness of serious metabolic and gastrointestinal complications in caffeine overdose.

Case Presentation

A 46-year-old man without a known medical history was brought to the emergency department with impaired consciousness and black vomit. He reported ingesting 100 caffeine tablets (totaling 10 g of caffeine) at midnight to attempt suicide due to private issues. He was found by an acquaintance at 7 Pm and transported to our hospital at 9 Pm. He had no history of diabetes, psychiatric history, medication use, or prior hospital visits. He was a nonsmoker with no history of alcohol or illicit drug use.

Diagnostic Assessment

On arrival, vital signs were as follows: temperature, 36.9 °C; heart rate, 112 beats per minute and irregular; blood pressure, 148/97 mmHg; and respiratory rate, 32 breaths per minute. His height was 183 cm, his body weight was 90.2 kg, and his body mass index was 26.9 kg/m². He was confused but responded to verbal stimuli. Abdominal examination was unremarkable, with a soft, nontender abdomen and normal bowel sounds. No other abnormalities were noted on physical examination.

An electrocardiogram revealed frequent premature ventricular contractions. Laboratory tests showed significant hyperglycemia, metabolic acidosis, positive ketonuria, and elevated hemoglobin A1c (Table 1). These findings confirmed the diagnosis of DKA. Measurement of serum caffeine concentrations was precluded owing to the unavailability of the equipment at the hospital. An abdominal computed tomography scan showed thickening of the esophageal wall, indicating esophageal inflammation (Fig. 1). A head computed tomography scan revealed no acute findings. We diagnosed caffeine intoxication and diabetes mellitus complicated by DKA, based on clinical presentation and laboratory findings.

Table 1.

Laboratory data on admission

	Value	Reference values
Albumin	5.7 g/dL (57 g/L)	4.1-5.1 g/dL (41-51 g/L)
Creatinine	1.12 mg/dL (99.8 μmol/L)	0.65-1.07 mg/dL (57.5-94.6 μmol/L)
Sodium	144 mEq/L (144 mmol/L)	138-145 mEq/L (138-145 mmol/L)
Potassium	3.1 mEq/L (3.1 mmol/L)	3.6-4.8 mEq/L (3.6-4.8 mmol/L)
Chlorine	107 mEq/L (107 mmol/L)	101-108 mEq/L (101-108 mmol/L)
Glucose	538 mg/dL (29.9 mmol/L)	73-109 mg/dL (4.0-6.0 mmol/L)
Hemoglobin A1c	11.6% (103 mmol/mol)	4.9-6.0% (29−42 mmol/mol)
C-peptide	0.92 ng/mL (0.30 nmol/L)	0.61-2.09 ng/mL (0.20-0.69 nmol/L)
Aspartate transaminase	65 U/L (65 U/L)	13-30 U/L (13-30 U/L)
Alanine aminotransferase	51 U/L (51 U/L)	10-42 U/L (10-42 U/L)
Amylase	407 U/L (407 U/L)	44-132 U/L (44-132 U/L)
Creatine kinase	8632 U/L (8632 U/L)	59-248 U/L (59-248 U/L)
C-reactive protein	6.05 mg/dL (60.5 mg/L)	0-0.14 mg/dL (0-1.4 mg/L)
White blood cell count	28.77 × 103/μL (28.77 × 109/L)	3.3-8.6 × 103/μL (3.3-8.6 × 109/L)
Hemoglobin	18.8 g/dL (188 g/L)	13.7-16.8 g/dL (137-168 g/L)
Platelet	31.0 × 104/μL (310 × 109/L)	15.8-34.8 × 104/μL (158-348 × 109/L)
Total ketone bodies	9391 μmol/L (9.391 mmol/L)	0-130 μmol/L (0-0.130 mmol/L)
Acetoacetate	1786 μmol/L (1.786 mmol/L)	0-55 μmol/L (0-0.055 mmol/L)
β-hydroxybutyrate	7605 μmol/L (7.605 mmol/L)	0-85 μmol/L (0-0.085 mmol/L)
Venous blood gas
pH	7.16	7.35-7.45
pCO2	22.8 mmHg (3.0 kPa)	35-48 mmHg (4.7-6.4 kPa)
HCO3−	11.8 mmol/L (11.8 mmol/L)	22-26 mmol/L (22-26 mmol/L)
Anion gap	25.2 mmol/L (25.2 mmol/L)	10-14 mmol/L (10-14 mmol/L)
Lactic acid	2.5 mmol/L (2.5 mmol/L)	0.5-1.9 mmol/L (0.5-1.9 mmol/L)
Immunological data
Antiglutamic acid decarboxylase antibody	<5.0 U/mL (<5000 U/L)	<5.0 U/mL (<5000 U/L)
Urinary test
Urinary glucose	4+	—
Urinary protein	2+	—
Urinary blood	3+	—
Urinary ketone	4+	—

Figure 1.

Axial image from a noncontrast computed tomography scan demonstrates thickening of the esophageal wall (circle).

Treatment

The patient was admitted to the intensive care unit. Treatment for DKA included aggressive IV fluid resuscitation and continuous insulin infusion. Due to agitation and restlessness, IV midazolam was administered for sedation over 3 days. β-blocker therapy with IV landiolol was initiated to control ventricular arrhythmia and continued for 4 days. On day 5, once the patient's consciousness improved and cardiac rhythm normalized, oral feeding was resumed. He subsequently developed epigastric pain, leading to discontinuation of oral intake. An oral potassium-competitive acid blocker, vonoprazan (20 mg/day), was initiated for presumed upper gastrointestinal injury. Esophagogastroduodenoscopy (EGD) was performed on day 8. The EGD findings on day 8 showed white circumferential severe erosions in the middle to lower esophagus. A part of esophageal epithelium was peeled off. Additionally, a massive ulcer in the duodenal bulb was found. However, the stomach was intact (Fig. 2A). These findings were consistent with drug-induced gastrointestinal injury by oral caffeine intake [8].

Figure 2.

Findings of esophagus, stomach, and duodenal bulb on day 8 (A) day 21 (B) by esophagogastroduodenoscopy.

Outcome and Follow-up

By day 3, DKA had resolved, with a blood glucose level of 107 mg/dL (SI: 5.94 mmol/L) (reference range, 73-109 mg/dL [SI: 4.0-6.0 mmol/L]), venous blood pH of 7.39 (reference range, 7.35-7.45), HCO₃⁻ of 18.0 mmol/L (reference range, 22-26 mmol/L), and an anion gap of 16 mmol/L (reference range: 10-14 mmol/L). The continuous IV insulin infusion was stopped and transitioned to subcutaneous insulin injection. Consciousness fully recovered after 3 days of sedation. Normal sinus rhythm was achieved after 4 days of landiolol infusion.

After continuous treatment with vonoprazan, follow-up EGD on day 21 showed improvement in the endoscopic findings. Severe esophagitis was cured, and duodenal ulcer was reduced. The stomach kept intact (Fig. 2B). Then oral intake could be resumed without pain.

When his general condition stabilized on day 20, insulin secretion was evaluated. Fasting blood glucose was 114 mg/dL (SI: 6.33 mmol/L), and C-peptide was 1.55 ng/mL (SI: 0.513 nmol/L) (reference range, 0.61-2.09 ng/mL [SI: 0.20-0.69 nmol/L]). Two hours postprandial blood glucose was 153 mg/dL (SI: 8.49 mmol/L), and C-peptide was 3.3 ng/mL (SI: 1.092 nmol/L), indicating insulin resistance. The patient had rarely visited a hospital and had no known history of diabetes mellitus prior to this event. He had a family history of diabetes in his father and tested negative for antiglutamic acid decarboxylase antibody. Therefore, we made a diagnosis of T2DM. No microvascular complications were detected. Insulin was tapered off, and the patient was discharged on day 23 without diabetic medications. At a 2-month follow-up, his hemoglobin A1c was 6.6% (SI: 49 mmol/mol) (reference range, 4.9-6.0% [SI: 29-42 mmol/mol]), and he remained symptom-free. EGD findings became normal.

Discussion

This case highlights a rare but serious metabolic complication of caffeine overdose: DKA in a patient with previously undiagnosed T2DM. While caffeine-induced hyperglycemia is known, progression to DKA has only been reported in individuals with type 1 diabetes [6, 7], typically due to interruption of insulin therapy. This appears to be the first reported case of caffeine-induced DKA in T2DM, based on a literature search using PubMed.

The mechanisms linking caffeine overdose to severe metabolic disturbances, including DKA, are primarily mediated by sympathetic surge. Caffeine (1,3,7-trimethylxanthine), a methylxanthine alkaloid, acts as a potent adenosine receptor antagonist and phosphodiesterase inhibitor. By blocking adenosine receptors, it enhances sympathetic outflow and increases catecholamine release [9]. Catecholamines promote hepatic gluconeogenesis, suppress insulin secretion, increase glucagon release, and reduce insulin sensitivity, thereby elevating blood glucose levels. Decreased insulin sensitivity following caffeine administration has been reported, and this effect is also attributed to elevated catecholamine levels [10, 11]. Furthermore, phosphodiesterase inhibition increases intracellular cyclic adenosine monophosphate, which stimulates glycogenolysis, gluconeogenesis, and hormone-sensitive lipase-mediated lipolysis, leading to an excess of circulating free fatty acids [12]. In the presence of reduced insulin activity, hepatic oxidation of free fatty acids leads to excessive ketone body production, culminating in DKA.

In this case, the patient had chronic hyperglycemia underlying untreated T2DM. The large dose of caffeine increased catecholamine release, leading to a rapid rise in blood glucose levels and enhanced lipolysis, ultimately triggering DKA. Given the increasing availability of caffeine-containing dietary supplements and the rising incidence of intentional overdose [5], this case underscores the importance for emergency physicians to consider DKA in caffeine-intoxicated patients who present with altered mental status or level of consciousness, particularly those with preexisting or unknown diabetes. Early recognition and management of DKA in this setting are crucial for improving outcomes.

In addition to DKA, this patient developed significant gastrointestinal complications, including esophageal erosions and a duodenal ulcer, following a caffeine overdose. While upper gastrointestinal symptoms such as nausea, emesis, and abdominal pain are frequently observed after caffeine overdose [5, 13], overt gastrointestinal injuries like erosions or ulcers are uncommon.

The mechanisms underlying caffeine-induced gastrointestinal injury are thought to involve both local and systemic effects. High-dose caffeine, especially in tablet form, can cause direct mucosal damage to the esophagus if tablets lodge or dissolve slowly, acting as a chemical irritant. Severe erosive esophagitis has been reported following ingestion of large quantities of caffeine pills [8, 14]. A case of acute esophageal necrosis caused by caffeine has been reported, which is thought to be caused by direct mucosal damage combined with vasoconstriction of mucosal blood flow and subsequent vomiting [15]. Furthermore, caffeine is known to stimulate gastric acid secretion and may lower esophageal sphincter tone, potentially exacerbating acid reflux [16]. Although necrotizing esophagitis, so-called “black esophagus”, is a known but uncommon complication associated with severe DKA [17], the endoscopic findings in this patient showed white esophageal erosions, which are inconsistent with necrotizing esophagitis. The patient reported no oral intake of supplements, medications, alcohol, or illicit drugs other than caffeine and had no history of mental illness. Based on the clinical course and the absence of other confirmed ingestions, caffeine remains the most likely contributor. Nevertheless, the patient was under extreme psychological stress that led to a suicide attempt, and stress-related gastrointestinal injury cannot be entirely excluded. Psychological stress increases risk for peptic ulcer disease, regardless of Helicobacter pylori infection or the use of nonsteroidal anti-inflammatory drugs. It may contribute to ulcer formation through mechanisms such as increased gastric acid load, impaired healing due to hypothalamic-pituitary-adrenal axis activation, altered blood flow, or cytokine-mediated disruption of mucosal defenses [18]. Gastric ulcers typically develop gradually in the setting of hypochlorhydria due to chronic mucosal inflammation [19]. In this case, however, the gastric mucosa was intact, and mucosal damage was confined to the esophagus and duodenum. These findings suggest that the acute hyperchlorhydric state induced by caffeine overdose selectively injured the esophageal and duodenal mucosa while sparing the stomach. Overall, caffeine overdose appears to have been the predominant cause of the mucosal injury, although psychological stress may have contributed.

A limitation in the evaluation of this case was the inability to measure serum caffeine concentrations at our hospital. This prevented a definitive correlation between the caffeine level and the severity of intoxication and DKA. In this case, the patient ingested a near-fatal dose of 10 g of caffeine, suggesting that such a dose may trigger DKA in patients with diabetes mellitus.

In conclusion, we report a rare and severe case of caffeine intoxication complicated by DKA and drug-induced gastrointestinal injury in a patient with previously undiagnosed T2DM. This case highlights the potential risk for patients with diabetes developing DKA when consuming large doses of caffeine. In cases of caffeine overdose in patients with preexisting or unknown history of diabetes mellitus, accompanied by symptoms of impaired consciousness, DKA should be considered in addition to caffeine intoxication. Prompt diagnosis and management of both caffeine intoxication and DKA are essential.

Learning Points

• Caffeine overdose can precipitate severe metabolic complications, including DKA, particularly in individuals with preexisting or undiagnosed diabetes mellitus.

• In patients presenting with altered mental status or metabolic disturbances following caffeine overdose, clinicians should consider complications of DKA in addition to caffeine intoxication, even in the absence of a history of diabetes.

• Acute high-dose caffeine ingestion can cause significant drug-induced gastrointestinal injuries, including esophageal erosions and duodenal ulcers.

• A thorough metabolic evaluation, including blood glucose, electrolytes, and acid-base status, is crucial in the assessment of patients with severe caffeine intoxication.

Contributors

All authors made individual contributions to authorship. H.K., S.U., and T.M. were involved in the diagnosis and management of the patient and manuscript submission. H.K. wrote the draft of the manuscript. Y.O., T.K., and T.A. contributed to the interpretation of clinical data and critical revision of the manuscript. All authors reviewed and approved the final manuscript.

Funding

No public or commercial funding.

Disclosures

None declared.

Informed Patient Consent for Publication

Signed informed consent was obtained directly from the patient.

Data Availability Statement

Original data generated and analyzed during this study are included in this published article.