EVIDENCE LIBRARY
Can a Blood Alcohol Result Prove Hypoglycemia? What an Oral-versus-IV Ethanol Trial Found
A 12-man human ethanol trial measured glucose, glucagon and FGF21. Learn why a measured BAC cannot diagnose hypoglycemia or prove causation.
Short answer: No. A laboratory ethanol result measures ethanol in a particular specimen; it does not measure the person’s glucose concentration or establish that alcohol caused a particular episode of low blood sugar. In a controlled crossover experiment in 12 healthy men, researchers matched plasma ethanol exposure after intragastric and intravenous ethanol administration and measured a series of metabolic hormones. Glucose initially rose, then fell below its starting concentration later in both sessions. However, the study had no ethanol-free placebo condition, so continued fasting could partly explain the later decline. Its strongest finding concerns the pattern of hormonal responses under two administration routes, not a diagnostic rule for hypoglycemia in an individual.
The primary source is Lanng and colleagues’ 2019 study, “Gluco-metabolic effects of oral and intravenous alcohol administration in men,” Endocrine Connections, DOI: 10.1530/EC-19-0317. The investigators measured circulating ethanol, glucose and hormones—not driving ability, eye movements, sobriety-test performance or injury causation.
Why does this distinction matter in a toxicology or injury case?
Three questions are easy to conflate. First, was ethanol present, and at what measured concentration in the tested specimen? Second, was blood glucose actually low at a clinically relevant time? Third, did alcohol materially contribute to the person’s symptoms or outcome? A positive or elevated blood alcohol concentration does not answer the second or third question. Conversely, a measured low glucose value does not, without temporal and clinical context, establish why it occurred.
The 2019 experiment helps explain the distinction because it measured several biological outcomes during controlled ethanol exposure. It did not establish a mathematical conversion from BAC to glucose, an individual risk score for hypoglycemia, or an ethanol-related impairment threshold. If a legal or medical opinion depends on a glucose event, the relevant evidence includes timed glucose measurements, symptoms, nutrition and fasting history, medications, medical conditions, treatment and the sequence of events. Neither a single ethanol result nor this small study replaces those records.
How was the oral-versus-IV experiment designed?
The investigators used a double-blind, randomized crossover design: the same 12 healthy male participants underwent two ethanol-exposure sessions on separate days. One delivered ethanol directly into the stomach (intragastric ethanol infusion, IGEI); the other delivered ethanol intravenously (intravenous ethanol infusion, IVEI). The study described a dose of 0.7 grams of ethanol per kilogram of body weight and adjusted the comparison to create broadly comparable plasma ethanol exposure across the routes. The reported route comparison for ethanol exposure was not statistically different (P = 0.38).
“Oral” in the publication’s title therefore describes the gastrointestinal route; it was not a naturalistic drinking session with participants choosing beverages, meal timing or drinking speed. Direct stomach infusion controls aspects of delivery that often vary in real-world alcohol consumption. The study compared two ethanol conditions, not ethanol against an ethanol-free placebo. This is particularly important for interpreting a late change in glucose during a prolonged, controlled session.
The investigators repeatedly sampled plasma glucose and ethanol, as well as insulin, C-peptide, glucagon, the incretin hormones GIP and GLP-1, and fibroblast growth factor 21 (FGF21). They also calculated an insulin secretion rate from physiological measurements. These are not interchangeable endpoints: a directly assayed concentration, a modeled secretion rate and a clinical diagnosis carry different evidentiary meanings.
What actually happened to glucose and the hormonal measurements?
During both routes, plasma glucose rose initially, reaching a peak approximately 30 minutes into the exposure, then later declined below baseline. In the authors’ discussion, that decline occurred after roughly 3.5 hours. Importantly, a below-baseline concentration is not synonymous with laboratory-confirmed hypoglycemia: an individual glucose value, diagnostic threshold and clinical symptoms are necessary to characterize a hypoglycemic event.
The hormonal measurements showed similarities and differences between routes. GIP and GLP-1 did not show the stimulation that would be expected from a conventional nutrient-associated incretin effect. Insulin concentration did not show the same pattern in each route: it was reported as unchanged with intragastric infusion and decreased with intravenous infusion. However, C-peptide and the inferred insulin secretion rate decreased during both routes, as did circulating glucagon. FGF21 rose markedly—approximately ninefold—after both interventions.
| Endpoint | Study observation | What the observation cannot establish |
|---|---|---|
| Plasma ethanol | Similar overall exposure was achieved between intragastric and IV sessions; route comparison P = 0.38. | That one person’s ethanol concentration predicts their glucose, impairment or symptoms. |
| Plasma glucose | Initial rise at about 30 minutes, later decline below baseline under both routes. | That alcohol alone caused the decline, or that participants experienced clinical hypoglycemia. |
| Insulin concentration | No clear increase during intragastric delivery; a decrease during IV delivery. | That peripheral insulin concentration by itself equals pancreatic secretion. |
| C-peptide and estimated insulin secretion rate | Decreased under both ethanol delivery conditions. | A validated individual diagnostic test for alcohol-induced hypoglycemia. |
| Glucagon | Decreased similarly under both routes. | That a particular patient’s counterregulatory response failed or caused an event. |
| GIP and GLP-1 | No overall incretin-hormone stimulation attributable to the ethanol intervention. | That every gut-related metabolic effect of alcohol is absent. |
| FGF21 | Marked rise, about ninefold, in both conditions. | That FGF21 independently diagnoses ethanol exposure, intoxication or impairment. |
The data above describe the investigators’ experiment, not an individual clinical cutoff. Source: Lanng et al., 2019.
Does the glucose decline prove that alcohol caused hypoglycemia?
No. The authors expressly identified continued fasting as a possible explanation for the later decline in glucose. They also acknowledged the absence of a placebo intervention. Even a randomized crossover experiment can answer only the questions its comparison groups permit. Matching two routes of ethanol exposure helps evaluate whether the route changes observed endocrine patterns; it does not, by itself, isolate the effect of ethanol against the effect of time, study procedures or fasting.
Nor does “below baseline” report a clinical diagnosis. Baseline is the participant’s earlier measured glucose, not necessarily the threshold used to define a hypoglycemic episode. A reduction from a normal baseline can remain within a normal range. The paper’s aggregate time course should not be transformed into the claim that the 12 men all developed symptomatic hypoglycemia, or that someone elsewhere with a measured ethanol concentration necessarily had low glucose.
In an actual case, the most probative evidence of glucose status is an appropriately timed, reliable glucose measurement. The causal question may additionally require consideration of missed meals, diabetes treatment or other medication, endocrine or hepatic disease, exercise, vomiting, comorbid illness, sample timing and corrective treatment. Those facts cannot be inferred from the blood ethanol result alone.
Did matching blood ethanol make the two conditions physiologically identical?
No. The researchers attempted to make circulating ethanol exposure comparable, which reduces an important source of variation. It does not mean intragastric and IV administration produce identical concentrations at every moment in every compartment, identical first-pass exposure or identical local gut and portal effects. A nonsignificant test for overall ethanol exposure is not proof of mathematical identity.
This design is valuable precisely because a difference in hormone responses between matched-ethanol routes can inform hypotheses about gastrointestinal versus systemic mechanisms. The similar responses for glucagon, C-peptide, calculated insulin secretion and FGF21 are consistent with effects that do not require direct gastrointestinal ethanol stimulation. Nevertheless, a pathway was not proven by directly blocking it, and the study did not evaluate the response to customary drinking with food under ordinary conditions.
For a broader explanation of why dose, route and sampling time affect interpretation, see the site’s pharmacokinetics and toxicology guide. A pharmacokinetic comparison is not an individualized reconstruction of a person’s clinical condition.
What do glucagon, C-peptide and FGF21 add beyond a glucose measurement?
Glucagon is among the hormones involved in glucose regulation. The study measured lower circulating glucagon after both ethanol delivery modes. That group-level observation supports a hypothesis about metabolic regulation, but it does not establish whether any participant failed to respond to an episode of actual low glucose, and it does not diagnose a patient’s glucose counterregulation in an unrelated case.
C-peptide is released along with endogenous insulin and can help researchers interpret secretion because circulating insulin also reflects hepatic extraction. The paper reported lower C-peptide and an insulin secretion rate calculated from its measurements under both routes, even though peripheral insulin concentration itself differed in pattern by route. The inferred secretion rate is derived; it should not be described as an independently measured insulin concentration.
FGF21 is a metabolic signaling hormone. Its approximately ninefold rise was a striking finding in the controlled sessions, including during IV infusion. That helps constrain hypotheses about whether gut exposure is necessary for the response. It does not create a validated forensic biomarker of how much alcohol a person drank or whether they could drive. An FGF21 response is also not equivalent to a proven clinical outcome attributable to that response.
Could these results establish intoxication, mental status or accident causation?
They cannot. No standardized field sobriety tests, simulated driving, attention tasks, crash outcomes or individual impairment classifications were measured in this experiment. The study was designed around metabolic physiology in healthy men, not around courtroom reconstruction of conduct. A person’s measured blood glucose and blood ethanol concentrations may both be relevant to differential diagnosis, but an opinion about mental status or an accident needs contemporaneous clinical, behavioral and chronological evidence.
The general distinction between measurement and inference also appears in this site’s therapeutic drug monitoring guide. Quantifying a chemical or hormone answers a narrower question than proving a downstream behavioral outcome. For additional context on specimens, analytical methods and reporting limits, consult the forensic toxicology evidence library.
Which study limitations matter most before citing it?
| Limit | Why it changes the inference |
|---|---|
| Only 12 healthy men | Responses may differ in women, older adults, people with diabetes, malnutrition, liver disease or medication exposure; no subgroup-specific rules were established. |
| Two active ethanol conditions; no ethanol-free placebo | Late glucose change cannot be cleanly assigned to ethanol rather than concurrent fasting or study time. |
| Direct intragastric versus IV infusion | Not the same as ordinary drinking with food, varied beverages or unsupervised intake. |
| Controlled single exposure | No basis for projecting chronic alcohol use, withdrawal, adaptation or future episodes. |
| Hormone assays and a calculated secretion rate | These describe selected physiological markers; they are not individual impairment tests or direct measures of clinical causation. |
| No driving or accident outcomes | No validated conversion from ethanol or FGF21 to actual task performance, capacity or legal responsibility. |
Limitations are not reasons to discard the study. They define where it is useful: testing mechanistic hypotheses under controlled ethanol exposure. They also mark where an expert should stop rather than extending a small metabolic trial to a different patient or legal question.
Frequently asked questions
Can someone have a measurable BAC and a normal glucose concentration?
Yes; these are separate laboratory measurements. A positive ethanol result does not establish that glucose is low. The 2019 experiment did not supply a BAC-to-glucose conversion formula.
Did everyone in the study become hypoglycemic?
The paper reports an aggregate glucose decline below baseline after approximately 3.5 hours, not a demonstrated episode of clinically significant hypoglycemia in every participant. The distinction matters because below-baseline and below a clinical diagnostic threshold are not the same.
Does the route of ethanol administration make no difference?
That is too broad. Several hormone responses were similar across intragastric and IV exposure, but insulin concentration differed in its reported pattern. Comparable circulating ethanol does not make the two delivery routes equivalent in every biological respect.
Can FGF21 tell an expert when someone last drank?
Not from this study. Researchers examined controlled FGF21 responses after ethanol exposure, not a validated timing algorithm or a forensic diagnostic threshold for an unknown person.
Can a low glucose result prove that ethanol caused a crash?
No. A reliable, time-relevant glucose test could establish glucose status at its sampling time. Establishing why glucose was low, whether it affected function at the material time and whether it contributed to a crash requires additional evidence.
Was the study an ordinary drinking experiment?
No. Ethanol was delivered by intragastric or intravenous infusion under laboratory conditions. The design enabled a route comparison but does not reproduce normal drinking behavior, food intake or every pattern of absorption.
Is this research enough to diagnose an individual?
No. It is a small physiological experiment in healthy male volunteers. Individual diagnosis and causal interpretation depend on the patient’s clinical history, measurements, timing, examination and alternative explanations.
Primary study and practical evidentiary takeaway
Primary source: Lanng AR, Gasbjerg LS, Bergmann NC, et al. Gluco-metabolic effects of oral and intravenous alcohol administration in men. Endocrine Connections. 2019;8(10):1372–1382. doi:10.1530/EC-19-0317.
Bottom line: Controlled ethanol exposure was accompanied by changes in measured glucose and several hormones, with important similarities between intragastric and IV delivery. The absence of a no-ethanol control prevents assigning the later glucose fall solely to alcohol, and the study offers no rule for inferring hypoglycemia, individual impairment or causation from a BAC. Keep specimen measurements, hormone assays, derived physiological quantities and case-specific clinical conclusions separate.