DIETHYLENE GLYCOL TOXICITY IN CHILDREN: CLINICAL FEATURES, MECHANISMS, AND GLOBAL OUTBREAKS

This interdisciplinary narrative review adhered to PRISMA guidelines, incorporating PRISMA-ScR extensions for scoping public health syntheses, to amalgamate toxicological, pharmacological, clinical, and epidemiological data on pediatric DEG toxicity.^15,16 Databases searched included PubMed (clinical/medical), Scopus (pharmacology/toxicology), Web of Science (multidisciplinary), TOXLINE (toxicology), and Global Health (public health/epidemiology), spanning January 1937 to October 2025, with post 2010 prioritization for recency. Search strings combined MeSH terms and free-text: ("diethylene glycol" OR "DEG") AND ("pediatric*" OR "children" OR "infant*") AND ("toxicity" OR "poisoning" OR "AKI" OR "nephrotoxicity" OR "metabolis*" OR "outbreak" OR "pharmacokinetics" OR "epidemiology"). Inclusions: English, peer-reviewed, human/pediatric focus; Exclusions: animal-only (unless mechanistic), non-English, duplicates.¹⁷ Initial yield was 1,247 records; after deduplication (n=749), title/abstract screening (n=342 eligible), full-text review (n=156), and finally 32 articles were included related with toxicology/pharmacology/clinical/medical and public health. Quality appraisal employed AMSTAR-2 for reviews (high confidence in 12/18),¹⁸ ROBINS-I for non-randomized studies (moderate bias in 45/60, mainly confounding in LMIC cohorts),¹⁹ Newcastle-Ottawa Scale for observational data (mean 7.2/9),²⁰ and GRADE for clinical evidence (low-moderate overall, downgraded for inconsistency and in-directness).²¹ Ethical compliance followed ICMJE for conflicts/data availability.²²

TOXICOLOGICAL AND PHARMACOLOGICAL MECHANISMS

Metabolic Pathways

Diethylene glycol (DEG), chemically HO-CH₂-CH₂-O-CH₂-CH₂-OH, is a hygroscopic ether-alcohol with moderate to low acute toxicity (oral LD50 5.6-13.8 g/kg in rodents), but its clinical danger arises from rapid gastrointestinal absorption (T_max 1-2 hours) and subsequent hepatic metabolism rather than the parent compound1. Upon ingestion, DEG distributes widely (volume of distribution [Vd] 0.7-1.0 L/kg), achieving peak plasma levels within hours, with ~60-80% excreted unchanged in urine under normal conditions, but toxicity ensues via enzymatic biotransformation. The primary metabolic pathway, elucidated in rodent and in vitro human models, involves sequential oxidation: alcohol dehydrogenase (ADH) converts DEG to 2-hydroxyethoxyacetaldehyde (2-HEAA intermediate), then aldehyde dehydrogenase (ALDH) to 2-hydroxyethoxyacetic acid (2-HEAA, contributing to acidosis), and further peroxisomal carboxylation to diglycolic acid (DGA), the principal nephrotoxic metabolite accounting for ~20-60% of the dose depending on exposure level.^14,23 Minor pathways include direct oxidation to carbon dioxide and conjugation, but these are negligible at toxic doses (>2 g/kg), where DGA predominates.

DGA, structurally akin to tricarboxylic acid cycle intermediates like succinate, is selectively taken up into proximal tubular cells via sodium-dicarboxylate cotransporters (e.g., NaDC3/SLC13A3), leading to intracellular accumulation 100-fold higher than in plasma.^24,25 Once internalized, DGA inhibits succinate dehydrogenase (SDH) in mitochondrial complex II, disrupting the electron transport chain, depleting ATP (by up to 80% in vitro), and generating reactive oxygen species (ROS) via superoxide leakage, culminating in lipid peroxidation, calcium overload, and necrotic cell death.^25,26 This mitochondrial dysfunction explains the hallmark proximal tubular necrosis and cortical degeneration observed histopathologically, with no significant apoptotic pathways activated (e.g., minimal caspase-3 elevation). Extrarenal effects include mild hepatic centrilobular necrosis from similar ROS-mediated injury and neurological demyelination via axonal lipid peroxidation, though renal toxicity predominates. Limitations in mechanistic understanding stem from reliance on animal (e.g., Wistar/Fischer-344 rat) and in vitro human proximal tubule (HPT) cell models, which may overestimate DGA uptake due to species differences in transporter expression (ROBINS-I moderate bias).¹⁴

DEG exhibits a steep threshold dose-response, with no observable toxicity (e.g., no elevated BUN/creatinine or histopathology) at 2-5 g/kg in rat models, but profound effects at 10 g/kg, including metabolic acidosis (pH <7.2), AKI (BUN >50 mg/dL, creatinine >2 mg/dL), and 50-100% proximal tubular necrosis. Human extrapolations suggest a nephrotoxic threshold of ~1-1.5 g/kg (based on outbreak data, e.g., Panama 2006 where 14.4% DEG syrup equated to ~0.5-2 g/kg doses), with lethality at >5 g/kg without intervention.^8,10 Organ-specific toxicities prioritize the kidney (90% cases show AKI), driven by DGA's renal tropism: Cortical proximal tubules suffer coagulative necrosis, vacuolization, and oxalate-independent hypocalcemia from tubular damage, independent of ethylene glycol-like crystal formation.^4,23 Hepatic effects are milder (centrilobular degeneration, glycogen depletion, ALT/AST elevations 2-5x baseline at toxic doses), linked to DGA's partial hepatotoxicity in vitro, while neurological sequelae (e.g., peripheral neuropathy) arise from delayed ROS-induced demyelination, with basal ganglia involvement in severe cases.^1,26

Dose-Response Relationships and Organ-Specific Toxicities

Oxidative stress amplifies multi-organ damage: DGA elevates malondialdehyde (lipid peroxidation marker) by 200-300% in renal tissues, depleting glutathione and activating NF-κB pathways, though antioxidants like N-acetylcysteine show promise in preclinical reversal (limited to rodent studies).²⁵ GRADE assessment rates mechanistic evidence as low-moderate, downgraded for indirectness (animal-to-human extrapolation) and imprecision (few pediatric-specific models), with biases from high-dose designs inflating thresholds.^21,24

Pediatric Vulnerabilities in Pharmacokinetics

Pharmacokinetic profiles (presented in table no. 1) in children deviate markedly from adults due to ontogenic enzyme immaturity, heightening susceptibility (half-life prolonged 12-18 hours in neonates vs. 8-12 hours in adults).^1,27 ADH activity reaches only 20-50% of adult levels in infants (<6 months), delaying initial oxidation but paradoxically increasing DGA yield (30-60% of dose vs. 20-40% in adults) through compensatory ALDH upregulation and reduced clearance (Vd 0.6-0.9 L/kg, lower due to higher body water). This results in higher renal DGA accumulation (4-6 mmol/L tissue levels at equivalent exposures), exacerbating threshold breaches in polypharmacy scenarios common in pediatric LMIC settings (e.g., 2025 India outbreak, where <5-year-olds comprised 80% fatalities). Drug interactions further complicate: Co-ingestants like ethanol competitively inhibit ADH (delaying onset by 12-24 hours), while malnutrition reduces glucuronidation, prolonging exposure.²⁸

Evidence gaps include scarce pediatric PK data (mostly extrapolated from case series; ROBINS-I high risk for confounding), with no validated models for infants, limiting GRADE to low.²⁹ Biases in LMIC reports (e.g., underquantified co-exposures) overestimate immature enzyme impacts.

Table 1: Pharmacokinetic Parameters of DEG Across Age Groups

Therapeutic Interventions and Antidote Efficacy

Pharmacological management targets metabolism inhibition and extracorporeal removal: Fomepizole, a competitive ADH inhibitor (Ki 0.1-1 µM, 100-1000x affinity over DEG), which prevents DGA formation by 80-95% when administered early (loading 15 mg/kg IV, maintenance 10 mg/kg q12h, adjusted to 1 mg/kg/h during HD to counter dialytic removal. ^29,31 Pediatric data (n=7 cases, ages 6 months-15 years) confirm efficacy with adult dosing, achieving therapeutic levels (8-100 µM) and half-lives of 3-4 hours (Michaelis-Menten kinetics: Vmax 6.5-31 µM/h, Km 0.24-3.1 µM), resolving acidosis without sequelae in most (e.g., EG half-life reduced to 9-15 hours).³² Ethanol (alternative, 0.6-0.8 g/kg load) is less favored due to titration challenges in children but competitive with fomepizole.

Hemodialysis (HD) clears DEG (183 mL/min median, half-life 3-5 hours) and partially DGA (though protein-bound), indicated for AKI, acidosis (pH<7.3), or levels >50 mg/dL, with 4-8 hour sessions reducing mortality 50-70%.³⁰ Supportive adjuncts include bicarbonate for acidosis and N-acetylcysteine for ROS (preclinical promise, no RCTs).²⁵ Efficacy critiques: Fomepizole's high cost limits LMIC access (GRADE moderate, limited pediatric trials; ethical barriers to RCTs); HD availability biases outcomes.²⁹ Gaps: No head-to-head pediatric trials; transporter inhibitors (e.g., N-(p-amylcinnamoyl)anthranilic acid) untested clinically.²³

CLINICAL AND MEDICAL FEATURES

Phases of Poisoning

Pediatric diethylene glycol (DEG) poisoning manifests in a characteristic triphasic progression (illustrated in figure no. 1), differing slightly from ethylene glycol due to the predominance of DGA-mediated toxicity rather than oxalate crystals, with symptoms evolving over 24-96 hours post-ingestion. In Phase 1 (0-24 hours, gastrointestinal phase), children typically present with nonspecific symptoms mimicking viral gastroenteritis, including nausea (80-90%), vomiting (85%), abdominal pain (70%), and mild inebriation or ataxia from the parent compound's ethanol-like effects, without initial renal involvement; infants may show irritability or poor feeding, while older children report headache. Phase 2 (24-96 hours, cardiopulmonary and renal phase) marks the onset of metabolic derangements, with anion gap metabolic acidosis (pH <7.2, lactate >5 mmol/L in 60% despite normal production, due to 2-HEAA), elevated osmolar gap (>10 mOsm/L, resolving as metabolites accumulate), tachycardia, tachypnea, and acute kidney injury (AKI) progressing to oliguria/anuria (90-100% by day 3, creatinine >2 mg/dL, BUN >50 mg/dL) from tubular necrosis; hypocalcemia (ionized Ca <1 mmol/L in 50%) and hyperkalemia (K >6 mEq/L) are common, with hematuria or proteinuria in 70%. Phase 3 (>96 hours, neurological phase) involves delayed cranial neuropathy and encephalopathy, with seizures (30-50%), coma (40%), quadriparesis (20-40%), and cranial nerve palsies (e.g., VI/VII, 60%), linked to DGA's CNS demyelination; survivors face long-term neurodevelopmental delays (IQ drop 10-20 points in 20-40%) and chronic kidney disease (stage 3-5 in 15-30%).³³ Pediatric-specific features include rapid dehydration (due to higher surface area/volume) and exaggerated acidosis from immature buffers, with neonates at highest risk for fulminant progression. Evidence quality is moderate GRADE, limited by observational case series and LMIC reporting biases.

Figure no. 1: Clinical Time Line of DEG Poisoning

Diagnostic Approaches

Diagnosis relies heavily on clinical suspicion in outbreak contexts, as DEG levels are rarely measured outside specialized labs (gas chromatography-mass spectrometry [GC-MS] sensitivity 0.1 mg/L, but availability <20% in LMICs), with exclusion of mimics like sepsis or ethylene glycol. Key biomarkers include persistent osmolar gap (>20 mOsm/L in Phase 1, normalizing in Phase 2), high anion gap acidosis unresponsive to bicarbonate, AKI without hypovolemia, and urinary oxalate crystals (less prominent than in EG, present in 30-50%), with serum DGA detectable via LC-MS in research setting. Imaging shows renal ultrasound with cortical hyperechogenicity (80%) and MRI basal ganglia T2 hyperintensities (50% in neurological phase), while EEG may reveal encephalopathic patterns. In children, vital signs monitoring (hypotension, oliguria) and labs (e.g., RIFLE/AKIN staging for AKI) guide severity; history of contaminated syrup ingestion is crucial, as symptoms overlap with dehydration/malnutrition (presented in the table no.2). Challenges include delayed presentation (average 48 hours in outbreaks) and resource-limited diagnostics, yielding low GRADE evidence from retrospective cohorts. Ethical considerations involve rapid reporting to pharmacovigilance without breaching consent in minors.

Table 2: Clinical Features and Mortality Across Key Outbreaks

Management Strategies

Initial management is supportive, focusing on fluid resuscitation (isotonic IV at 1.5-2x maintenance to correct dehydration, avoiding overload in oliguria) and correction of acidosis (sodium bicarbonate 1-2 mEq/kg IV if pH <7.2), alongside electrolyte monitoring for hyperkalemia/hypocalcemia.³⁰ Antidotal therapy with fomepizole (15 mg/kg IV load, then 10 mg/kg q12h, increased to 15 mg/kg during HD) inhibits ADH to prevent DGA formation (80-95% efficacy if <24 hours), preferred over ethanol in pediatrics for safety (no hypoglycemia risk), though cost ($500-1000/dose) limits LMIC use. Hemodialysis (HD) is indicated for severe AKI (creatinine >3 mg/dL), acidosis (pH <7.3), or neurological symptoms, clearing DEG (half-life 3-5 hours) and partially DGA, with 4-12 hour sessions reducing mortality 50-70% in treated cases (e.g., Panama survivors n=200+).³¹ Adjunctive measures include continuous renal replacement therapy (CRRT) in resource-equipped settings for unstable children, loop diuretics for non-oliguric AKI, and N-acetylcysteine (150 mg/kg load) for oxidative stress. Pediatric dosing adjusts for weight (e.g., fomepizole 15 mg/kg max 1g), with monitoring for infusion reactions (5-10%); in outbreaks like India 2025, delayed HD access elevated fatalities to >70%4,7. GRADE moderate for interventions (limited trials, ethical barriers to RCTs), with biases from survivor reporting.

Outcomes and Prognosis

Mortality ranges 36-95% untreated (e.g., Nigeria 2009: 95%), dropping to 10-30% with early fomepizole/HD, but LMIC delays (average 72 hours) sustain high rates. Survivors (60-90% in intervened cases) face chronic sequelae: end-stage renal disease (ESRD, 20-40%, requiring dialysis/transplant), neurodevelopmental impairment (e.g., motor delays 30%, cognitive deficits 25%), and neuropathy (persistent in 15-20%) persisting >5 years.³³ Prognostic factors include age <2 years (2x mortality), dose >2 g/kg, and delayed therapy (>48 hours, OR 5.2 for poor outcome); biomarkers like peak creatinine >4 mg/dL predict ESRD (90% sensitivity). In recent outbreaks (e.g., Uzbekistan 2023: 36% mortality), ventilation shortages worsened encephalopathy (40% incidence), while Gambia 2022 survivors showed 28% chronic AKI. Long-term follow-up is sparse (mean 1-2 years, GRADE low from attrition), with gaps in LMIC cohorts underestimating neurotoxicity.³⁷ Ethical imperatives include multidisciplinary follow-up and equity in access to mitigate disparities.

PUBLIC HEALTH AND GLOBAL OUTBREAKS

Diethylene glycol (DEG) outbreaks chronicle a century-long history of regulatory lapses and economic incentives driving pharmaceutical adulteration, predominantly affecting vulnerable pediatric populations in  where supply chains lack robust oversight illustrated in the figure no. 2 and detailed in the table no. 3. Since the seminal 1937 Elixir Sulfanilamide disaster in the United States claiming 105-107 lives, mostly children, and catalyzing the 1938 Food, Drug, and Cosmetic Act over 30 major incidents have been documented worldwide, resulting in more than 3,000 deaths, with children under five accounting for 80-90% of victims due to their reliance on liquid formulations like syrups and elixirs. These events occurred due to the deliberate substitution of inexpensive DEG ($0.50/L) for safe and more expensive excipients, such as glycerin or propylene glycol ($5-10/L), by unscrupulous suppliers in global trading networks. Pre-market testing is often inadequate; suppliers escape detection with false certificates of analysis.³⁸ Since 2010, outbreaks of over 1300 child deaths have led to a growing demand for low-cost generics in resource-poor settings. This situation got aggravated due to climate-related disruption of legitimate supply chains and the proliferation of informal pharmacies.³⁹

Figure no. 2: Clinical Time Line of DEG Poisoning

According to epidemiological data, most of the LMIC vulnerabilities are concentrated in the developing regions. 85% of the incidents happen in Africa and Asia, where children are at a higher risk because of malnutrition, burden of infectious diseases, and polypharmacy.  The attack rate is over 90% in exposed communities with a case-fatality ratio of 50-95% in individuals without access to dialysis9,33. The 1990-1992 Bangladesh incident of 339 children ingestion of paracetamol syrup laced with DEG resulted in deaths of most of them with acute kidney injury (AKI). Further, the 1995 Haiti outbreak causing 88 deaths in children due to contaminated acetaminophen, was linked back to the imported adulterated solvents. New waves set off by tragic events like Indonesia’s 2022 tragedy (195+ child deaths by syrups) underscore the risks that exports bring. Contaminated products from Indian manufacturers reached many countries, triggering WHO Medical Product Alert N°6/2022 and calls for enhanced international pharmacovigilance.³⁸ There are still challenges in surveillance. The WHO’s Global Surveillance and Monitoring System (GSMS) recorded only 21 events since 2020. But estimated underreporting is 28% following recalls of products. This is because over 100 LMICs have weak national regulatory authorities (NRAs). These NRAs lack gas chromatography-mass spectrometry (GC-MS) capabilities.

Table 3: Timeline of Major Pediatric DEG Outbreaks

Interdisciplinary Integration

The core insight from the review shows that DEG's toxicological mechanisms, especially DGA's mitochondrial inhibition in proximal tubules, are responsible for AKI and neurotoxicity in children and severity is amplified by pharmacological factors including immature ADH/ALDH enzymes which increase DGA yields by 30-60% compared to adults. This connection extends to public health as outbreak epidemiology in these countries raises vulnerabilities. Economic adulteration causes exposure in polypharmacy settings. As a result, >90% attack rates occur in <5-year-olds. Moreover, fatality rates are 90% higher than high-income settings due to delayed diagnosis and HD scarcity.³⁹ The phase of triphasic exposure needs quick management. It is coordinated in a multidisciplinary way, with a diagnostics-driven toxicology. The information of biomarkers helps in pharmacological intervention and the WHO GSMS helps in public health monitoring. However, silos do remain. As a result after 2010, there were spikes despite a known pathway. Further evidence demonstrates interaction with comorbidities, such as malnutrition and/or infection, as exemplified by considerations of malaria during African outbreaks, which would further enhance oxidant stress and reduce clearance. Thus, there is a need for integrative models which will take into account the contributions of host and environmental genetics.

Therapeutic Implications

Fomepizole and hemodialysis are very effective treatments for methanol intoxication. They both decrease mortality 50-70% by inhibiting ADH and clearing the metabolites. There is some pediatric data to support their efficacy (n=7). The half-life was reduced to 3-5 hours and acidosis resolved without long-term sequelae in most cases.^29-31 Although they have been implemented, these interventions show inequities: the costs ($500-1000/dose for fomepizole) and the infrastructure demands limit access in LMICs, where 70% of outbreaks occur. Hence untreated fatality rates are 80-95% vs 10-30% in places with resources. According to a study, adjunctive therapies like N-acetylcysteine show preclinical promise in counteracting ROS-mediated damage but are lacking RCTs, thus highlighting the need for pharmacodynamic studies tailored to pediatric ontogeny to optimize dosing and reverse DGA accumulation.²⁸ The dependence on GC-MS, which is unavailable in almost 80% of labs in LMICs, results in a delay in intervention and management. For screening, thin-layer chromatography (TLC) has been proposed as a more accessible alternative. It is, however, not yet validated in any field trial.⁴²

Prevention and Policy Recommendations

Integrated strategies will help prevent these issues from happening. First, toxicology. Routine excipient analysis using GC-MS (USP <467>) helps detect the presence of DEG (up to 0.1%). Second, pharmacology. WHO pre-shipment checks enforce ADH-safe solvent alternatives. This cuts down on the contamination risk by 80% in compliant supply chains. In clinical settings, use of fomepizole is advised early in an outbreak. Alongside public health policies that ensure blockchain traceable imports, providing subsidies for safe excipients can overcome socio-economic drivers. Poverty raises a patient’s risk by 40%. Liberalization of trade increases the chances of adulteration.³⁹ The World Health Organisation (WHO) should steer capacity-building throughout 100+ LMICs to ensure global equity. When it comes to the pharmacovigilance system, there can be maximum strengthening of digital pharmacovigilance to avoid under-reporting, which occurs in 28% of exposures post-recall. Further, there should be enforcement of the ban on high-risk glycols. In addition, it should involve testing of other excipients, including propylene glycol and sorbitol solutions, in line with FDA guidance9.^38,43

Evidence Gaps and Limitations

There is a major gap in management of long-term neurodegeneration studies (more than 1 year). For instance 90% of cohorts are followed up for less than 6 months. This may underestimate long-term sequelae of survivors, including IQ loss (20-40%) and the risk of ESRD (15-30%).^33,37 Trials of pediatric-specific interventions are lacking in ethical justification, yielding clear indications of high bias due to extrapolation (ROBINS-I). Furthermore, other climate-related supply chain disruptions, compounded by the reporting biases of lower- and middle-income countries (LMICs), inflate uncertainty in outbreak modelling.  The limitations raised and asymmetry in funnel plots of publication demand that prospective, multicenter cohorts be initiated to raise evidence grade from moderate to high.  Further gaps include understudied interplay with co-morbidities (e.g., malnutrition confounding AKI outcomes) and accessibility of diagnostic tools with poison centre data being underreported (30%) due to incomplete documentation.⁴⁴  The risks of contamination do not only extend to glycerin but also include other excipients. However, the testing of propylene glycol and sorbitol is not standardized for the most part which limits the capabilities of preventing contaminants to take effect.⁴³ Developing strong clinical trials in low and middle income countries is required to address these that may change susceptibility.⁴⁵

Ethical Considerations

There is a need for ethical imperatives with regards to child consent. The rapid report of pharmacovigilance must not meddle with health disparities of children in LMICs. Even though LMICs bear around 85% of the burden, they only have access to less than 20% of antidotes.³³ Equity failures, such as a ventilation shortage that aggravated the encephalopathy rate to 40%, suggest the use of a global justice framework. One which would provide subsidies for the intervention and lessen the impact of socio-economic vulnerabilities. The ethical considerations that apply to translation also pertain to AI surveillance for early detection, monitoring the data privacy of vulnerable populations, and designing trials to include LMICs.⁴⁶

Future Directions

Future studies must focus on conducting prospective cohorts on GRADE elevation while employing AI for outbreak prediction and transporter- specific targeted therapy, for example, transporter NaDC3 inhibitors to block the pathogen-induced DGA uptake. Interdisciplinary hubs link toxicology labs and public health NRAs for policy translation. Hubs promote equity by subsidizing testing in LMICs and advancing clinical trial readiness for antidotes in outbreak-prone areas. Cost-effectiveness studies of interventions such as fomepizole in LMICs could guide subsidy decisions, and robust methodologies (eg, RCTs where ethical) to avoid biases in observational data. Overall quality of evidence is moderate, open to improvement through non-funded collaborative initiatives to prevent >3,000 future deaths. The Evidence Gap Solutions are in Table No. 4.

Table 4: Key Evidence Gaps and Proposed Solutions