Jones · Journal of analytical toxicology 2017 · pooled case series review · n=51

Review of Caffeine-Related Fatalities along with Postmortem Blood Concentrations in 51 Poisoning Deaths.

Cited 53 times in the scientific literature.

Level 4 - case-series / case-control

Pooled case series review of published forensic fatality reports

PubMed 28334840 · doi:10.1093/jat/bkx011 · record verified 2026-08-30

What was done

Literature review of published case reports detailing postmortem blood caffeine concentrations in fatal poisonings/overdoses (N = 51). The authors analyzed age, sex, manner of death (suicide, accidental, or undetermined), co-ingested substances, and postmortem blood caffeine concentrations.

What was found

The 51 deceased individuals had a mean age of 39 ± 17.8 years (range 18–84) and were 61% female (N = 31). Postmortem blood caffeine concentrations had a mean of 187 ± 96 mg/L, a median of 180 mg/L, and a range of 33–567 mg/L. Blood concentrations did not differ significantly by sex (median 161 mg/L in males vs. 182 mg/L in females, P = 0.235) and did not correlate with age (R² = 0.026, P > 0.05). Manner of death was classified as suicide in 51% (median 185 mg/L), accidental in 16% (median 183 mg/L), and undetermined in 33% (median 113 mg/L, significantly lower than suicide or accidental cases, P = 0.023). Caffeine was consumed in tablet or powder form, and while co-ingestants like ethanol, benzodiazepines, antidepressants, antipsychotics, or ephedrine were common, caffeine was identified as the predominant lethal agent.

Why it matters

Provides a benchmark reference range for toxicologists and forensic pathologists investigating caffeine-related deaths, illustrating that lethal concentrations (median ~180 mg/L) generally require concentrated powdered or tablet forms rather than caffeinated beverages alone.

Limits

Data are derived from a small, heterogeneous collection of published case reports (N = 51) with inherent publication bias. Co-ingestion of other psychoactive substances was frequent and may have contributed to toxicity. Postmortem redistribution and postmortem interval variability were not controlled for in the abstract.

Cited by