Epinephrine

Limits on what this shows

• Null findings: DOSE BOUNDARY (stated explicitly): the red grade applies to exposure at plasma epinephrine of roughly 1.3 ng/mL or more, reached by infusion (Brown 0.1 ug/kg/min, about 1.74 ng/mL [9]; Darbar 160 ng/kg/min, 1.328 ng/mL [15]). In Darbar 1996 no fall occurred below 160 ng/kg/min [15], a dose-response bound rather than a contradiction. A 0.3 mg EpiPen thigh injection produced a peak plasma epinephrine (Cmax) of 0.52 ng/mL (n = 35) [7], about 2.5-fold below that, and [7] measured no potassium. At injection doses the only potassium data are the 1952 cohort [8], whose two tables disagree. A PubMed listing for subcutaneous or intramuscular epinephrine with plasma potassium in humans (10 records, listing level, including Simons 2001 [35]) contained no study measuring potassium after an autoinjector, intramuscular dose or in anaphylaxis. Whether an autoinjector dose lowers potassium is therefore unmeasured by any source opened.
• At-risk subgroups: People on potassium-depleting drugs (corticosteroids, diuretics, theophylline) per the label [2]; people given infusions in shock (the intravenous label indication) [2]; people prone to arrhythmia, since the authors of [15] conclude hypokalemia contributes to arrhythmic effects above 80 ng/kg/min. Beta-blocker users show blunted or reversed potassium responses [11][12][13].

Regulatory status

The intravenous Adrenalin label lists hypokalemia among metabolic adverse reactions and says potassium-depleting drugs (corticosteroids, diuretics, theophylline) "potentiate the hypokalemic effects of epinephrine" [2]. The EpiPen label (rev. 2/2023, first 100,000 of 102,662 characters read) does not mention potassium or hypokalemia [1].

Population evidence

Independent human cohorts with epinephrine itself, potassium measured directly: Brown 1983 [9], 6 volunteers on 0.1 ug/kg/min, circulating epinephrine about 1.74 ng/mL, plasma potassium fell about 0.82 mEq/L. Struthers and Reid 1984 [10], serum potassium 4.08 +/- 0.21 to 3.32 +/- 0.25 mmol/L (P < 0.002) during adrenaline infusion; sample size not stated on the record page. Struthers 1983 [11] and Reid 1986 [12] (one cohort of 9 volunteers, identical figures, counted once): 4.06 to 3.22 mmol/L. Whyte 1987 [16], 8 subjects, placebo-controlled: potassium fell and recovered within 30 minutes. Dury 1952 [8], 11 normal adults given 0.4 mg by injection (route read as intramuscular in the fetch summary; the text line quoted was not confirmed, so a subcutaneous route is not excluded; the body text writes 0.4 ml and the table footnote 0.4 mg): Table II, plasma potassium change from baseline at 20 minutes -16.5% (SE 1.3) and +1.0% (SE 1.4) at 60 minutes; Table V, 4.37 +/- 0.10 to 4.08 +/- 0.09 mEq/L at each subject's maximum change, mean -6.5%. The two tables disagree inside the paper itself (a per-subject maximum change cannot be smaller than the 20-minute mean), so the size of the fall at this dose is between about 0.3 and 0.7 mEq/L and is not resolved. The fall is transient and back to baseline by 60 minutes. Full PDF opened on a second attempt (OCR text, one figure garbled). DeFronzo 1981 [13] (6 subjects) and Rosa 1980 [14] (5 subjects given epinephrine with a potassium chloride load): epinephrine enhanced extrarenal potassium disposal, with the rise after the load about 0.1 mEq/L versus about 0.6 mEq/L for the load alone [14]. Darbar 1996 [15], dose-escalation infusion: serum potassium fell only at 160 to 320 ng/kg/min, with plasma epinephrine 1,328 +/- 902 pg/mL; sample size not in the abstract.

Mechanism

Named. Beta-2 receptor mediated cellular uptake: the fall was blocked by the selective beta-2 antagonist ICI 118551 (3.82 to 4.03 mmol/L, not significant) [10][9], reduced by the beta-1 selective atenolol (to 3.67 mmol/L) and reversed by timolol (4.25 mmol/L) [11][12], and reversed by propranolol [13]. Insulin contributed little to the extrarenal effect in [13]. Stimulation of skeletal-muscle Na+/K+-ATPase is the usual downstream step but was not named in the potassium abstracts opened.

Counter-evidence held on record

Counter-evidence and confounding No contradicting Tier A source was found. Rosa 1980 and DeFronzo 1981 show the same uptake effect from the opposite direction: epinephrine blunts the rise in potassium after a load [13][14]. Most cohorts are small healthy-volunteer infusion studies at pathophysiological plasma levels, and the 1952 injection cohort used an old assay and has internally conflicting tables. Struthers 1983 and Reid 1986 report one cohort twice. Re-examination of the red grade (2026-10-08) Held red. Criteria checked: at least 2 independent Tier A cohorts with epinephrine and potassium measured directly (Brown, Struthers 1984, DeFronzo, Rosa, Whyte, Darbar, Struthers/Reid), consistent direction, a fall to about 3.2 to 3.3 mmol/L (hypokalemic range) at infusion levels, a named beta-2 mechanism, no contradicting Tier A evidence, and the intravenous label lists hypokalemia. Hard Rule 5: this is a redistribution disturbance graded on the same tiers as an excess/shift entry, framed so that supplementation would be the hazard. The grade is scoped to infusion-level exposure; if an editor judges autoinjector relevance to govern, the defensible alternative is orange on the 1952 injection cohort alone. Framing caution This is a short-lived shift of potassium into cells after dosing, not a loss of potassium from the body. Nothing here supports adding potassium, and supplementing against a shift that reverses on its own is the hazard, not the remedy. Any potassium decision belongs to the prescriber who can see a measured value.

Limits on what this shows

• Null findings: CATS: no lasting harm by day 4 and no significant mortality difference [18][19]. No source measured lactate after an intramuscular autoinjector dose or in treated anaphylaxis; the evidence is infusion in critical illness and healthy volunteers. A topical-epinephrine burn study (Papp 2009) is reported in the HSDB excerpt as raising lactate; it was seen only through that excerpt and is not counted [4]. DOSE BOUNDARY (stated explicitly): the effect size rests on ICU infusion [17]; in healthy volunteers the rise was significant at a sustained-infusion plasma peak of about 0.33 ng/mL, below the EpiPen Cmax of 0.52 ng/mL [7][22], but with no magnitude retrieved and a 30-minute steady infusion rather than a bolus. A PubMed listing for subcutaneous or intramuscular epinephrine in humans (10 records, listing level) contained no lactate study after an injection dose or in anaphylaxis.
• At-risk subgroups: Septic shock, severe malaria and other critical illness during inotropic-dose infusion (the cohorts in [17][18][20]); people given the intravenous label indication of septic shock hypotension [2].

Regulatory status

The intravenous Adrenalin label lists "lactic acidosis" among metabolic reactions and says overdose can cause metabolic acidosis from elevated blood lactic acid [2]. The EpiPen label mentions metabolic acidosis only in the overdosage text and does not name lactate [1].

Population evidence

Day and colleagues [17], open randomised crossover, 23 critically ill patients (10 severe sepsis, 13 severe malaria), stepped adrenaline 0.1 to 0.5 ug/kg/min: 16 of 19 patients (84%) had the infusion stopped for lactic acidosis; mean lactate rose 3.2 mmol/L (95% CI 2.6 to 3.8), arterial pH fell 0.052 (0.035 to 0.068), base excess fell 3.8 mmol/L (2.8 to 4.7); dopamine lowered lactate 1.0 mmol/L (0.4 to 1.5). CATS trial [18][19], 330 septic shock patients in 19 French ICUs, epinephrine versus norepinephrine plus dobutamine: lactate significantly higher on day 1 only (P = 0.003), pH lower on days 1 to 4; 28-day mortality 40% versus 34% (RR 0.86, 95% CI 0.65 to 1.14). CAT trial [20][21], 280 ICU patients: epinephrine produced "significant but transient metabolic effects" and attending clinicians withdrew 18 of 139 epinephrine patients (12.9%); lactate values are not in the abstract. Healthy volunteers: Samra 1996 [22], 7 subjects on 25 ng/min/kg, plasma adrenaline rising from 0.15 +/- 0.03 to a peak of 1.82 +/- 0.15 nmol/L within 30 minutes (about 0.33 ng/mL, my conversion at 183.2 g/mol), blood lactate rose significantly (P < 0.001, no lactate values in the abstract); Kral 1969 [23], 4 subjects, adrenaline raised lactate and alprenolol abolished the rise. Massara and Camanni 1970 [33], adrenaline in man: propranolol prevented the normally observed rise in plasma lactate (abstract only, no n, dose or values; shared with the Phosphate entry).

Mechanism

Named. Beta-adrenergic stimulation of muscle glycolysis (the alprenolol block in [23]); the CATS authors propose increased aerobic glycolysis, reduced splanchnic and hepatic flow and higher oxygen consumption [19]. Na+/K+-ATPase driven muscle lactate production is shown in septic shock by Levy 2005 [25] (ouabain stopped the excess muscle lactate), but epinephrine was not named in the page read, so that source is mechanistic context only.

Counter-evidence held on record

Counter-evidence and confounding The ICU trials cannot separate a drug-driven rise from shock-driven lactate, but Day 1996 used a randomised crossover with dopamine as comparator [17], and healthy volunteers also show the rise [22][23]. CATS read through a trial summary and a review rather than the Lancet paper itself [18][19]; CAT lactate values were not retrieved [20]. Exposure is infusion; autoinjector exposure is unstudied. Re-examination of the red grade (2026-10-08) Held red. Criteria checked: independent Tier A sources with epinephrine and lactate measured directly (Day randomised crossover with CI, CATS, CAT, Samra, Kral, Massara), consistent direction, clinically meaningful size (3.2 mmol/L, 95% CI 2.6 to 3.8), named beta-adrenergic glycolysis mechanism, with healthy-volunteer data arguing against pure confounding by shock. Hard Rule 5 and Rule 2: lactate is a metabolite, not a nutrient, and the entry is an excess-direction drug-effect entry only (the intravenous label lists lactic acidosis as an adverse reaction); no nutrient deficiency is implied. If an editor weighs the absent injection-dose data, the defensible alternative is orange with an infusion boundary. Framing caution A rising lactate on epinephrine can be a drug effect as well as a sign of poor perfusion [17][18]. This is a metabolic disturbance of the drug, not a nutrient deficiency, and there is nothing to supplement. Interpretation belongs to the treating team.

Limits on what this shows

• Null findings: No source retrieved measured glucose after an autoinjector or anaphylaxis dose. The plasma glucose rise itself is reported as significant without values in three of the four abstracts [16][22][23].
• At-risk subgroups: People with diabetes (both labels) [1][2]; hyperthyroidism and the elderly are also cautioned in the EpiPen label [1].

Regulatory status

Intravenous label: "hypoglycemia, hyperglycemia, insulin resistance" among metabolic reactions [2]. EpiPen label: caution in diabetes (section 5.5), and "Patients with diabetes may develop increased blood glucose levels following epinephrine administration" in the patient-counseling section [1].

Population evidence

Deibert and DeFronzo 1980 [24], 9 normal-weight subjects, euglycemic insulin clamp with epinephrine 0.05 ug/kg/min: glucose metabolism fell 41% (5.49 +/- 0.58 to 3.26 mg/kg/min, P < 0.001) and epinephrine delayed and incompletely suppressed hepatic glucose production even with insulin above 100 uU/mL (P < 0.01). Whyte 1987 [16], 8 subjects, placebo-controlled: glucose rose during adrenaline infusion with no significant insulin change. Samra 1996 [22], 7 subjects: plasma glucose rose significantly (P < 0.001). Kral 1969 [23], 4 fasting men: adrenaline raised blood glucose, and the rise was not blocked by alprenolol.

Mechanism

Named: stimulation of hepatic glucose output and reduced insulin-mediated glucose disposal [24]. The alprenolol result suggests the glucose rise is not purely beta-1/2 driven [23].

Counter-evidence held on record

Counter-evidence and confounding The hypoglycemia listed in the intravenous label is the opposite direction and is not quantified in any source opened [2]. Hormone effects are not deficiencies (Hard Rule 2); this entry is an excess-risk entry only. Framing caution A transient glucose rise after dosing is not a nutrient problem and carries no dietary or supplement implication. People with diabetes should discuss glucose monitoring around use with their prescriber.

Limits on what this shows

• Null findings: No second independent human source with epinephrine and magnesium was found. A 1985 Clin Sci abstract from the same group ("Raised adrenaline levels cause hypomagnesaemia", DOI 10.1042/cs068019p) was seen only as a Crossref title, not opened, and is not counted.
• At-risk subgroups: Not defined by any source. Prolonged infusion is the exposure studied.

Regulatory status

Magnesium is not mentioned in the EpiPen or intravenous Adrenalin labels read [1][2], nor in the HSDB sections opened [4].

Population evidence

Whyte, Addis, Whitesmith and Reid 1987 [16]: placebo-controlled infusions of adrenaline and of intravenous salbutamol in 8 normal subjects. Both lowered plasma magnesium, adrenaline faster; magnesium did not recover within an hour; potassium fell and recovered within 30 minutes; insulin did not change significantly and glucose rose with adrenaline. The abstract (read as OpenAlex and Crossref records; the publisher page returned 403) gives no magnitude.

Mechanism

Named by the authors: intracellular movement of magnesium triggered by beta-adrenergic stimulation, not insulin mediated [16].

Counter-evidence held on record

Counter-evidence and confounding Plasma magnesium is a poor index of body magnesium, and no tissue or urinary magnesium was reported. The salbutamol arm in the same study shows the class effect but is not used as evidence for epinephrine. Framing caution A temporary fall in plasma magnesium after an infusion is not evidence that dietary magnesium is short. Do not read this as a reason to supplement magnesium.

Limits on what this shows

• Null findings: No magnitude, dose or route could be retrieved for either study (both closed access; OpenAlex records read), so the size and the relevance to an autoinjector dose are unknown. Parathyroid hormone did not change [34]. The cross-sectional finding in [34] (lower phosphate in hypertension) is an association and is not used. Two other PubMed hits seen only in a listing and not opened (a paediatric ICU study associating epinephrine above 0.3 ug/kg/min with hypophosphatemia, PMID 17327861; a 1979 chart review of hypophosphatemia, PMID 448890) are not used. Salbutamol and terbutaline data were not used (Hard Rule 6).
• At-risk subgroups: Not defined by either source. Infusion or injected adrenaline in adults was the exposure.

Regulatory status

Phosphate is not mentioned in the EpiPen or intravenous Adrenalin labels read [1][2], nor in the HSDB sections opened [4].

Population evidence

Massara and Camanni 1970 [33]: adrenaline-induced hypophosphataemia in man was almost completely abolished by propranolol (and insulin-induced hypophosphataemia significantly inhibited); propranolol also prevented the usual lactate rise. The record gives no sample size, adrenaline dose, route or values. Kjeldsen and colleagues 1988 [34]: men with untreated mild essential hypertension (n = 35) and normotensive controls (n = 44); during an epinephrine infusion that raised arterial plasma epinephrine into the lower pathophysiological range, serum phosphate fell significantly, with no change in serum parathyroid hormone. The infusion group size, dose and the size of the fall are not stated in the record read. Two different research groups, both with adrenaline given to people and phosphate measured directly.

Mechanism

Named: beta-adrenergic receptor mediated, since propranolol blocks the fall [33]. The authors of [34] attribute the lower phosphate of hypertensive men to higher plasma epinephrine. Cellular uptake is the probable route; urinary phosphate was not read.

Counter-evidence held on record

Counter-evidence and confounding Not graded red because no effect size, dose or sample size was retrieved for either study, so clinical size cannot be judged. Serum phosphate is a poor index of body phosphate. The combined shift pattern (potassium, magnesium, phosphate and lactate) is the same beta-2 mechanism and is disclosed in the Overlap map. Framing caution A temporary fall in serum phosphate after adrenaline is a redistribution disturbance, not evidence that dietary phosphate is short. Do not read this as a reason to supplement phosphate.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C source was opened, and no circulating claim of epinephrine calcium depletion was found. Re-check when PubMed access returns, including the cardiac-arrest epinephrine literature and the full text of [34] and [33] for calcium values.

What was searched

Crossref bibliographic queries ("adrenaline plasma phosphate calcium beta-adrenoceptor human", "epinephrine infusion phosphate calcium magnesium potassium volunteers plasma"; the last returned only veterinary, textbook and chemistry records); PubMed listings (2026-10-08, spaced 4 to 5 minutes apart): "adrenaline infusion plasma phosphate OR hypophosphataemia OR ionized calcium human" (ten refeeding, calcium-metabolism and rickets reviews, none with epinephrine), "epinephrine subcutaneous OR intramuscular plasma potassium humans" (ten records, none measuring calcium) and "adrenaline hypophosphataemia" (ten records; the two human epinephrine studies found concern phosphate, and the 1988 study reports parathyroid hormone unchanged without a calcium result in the record read); labels (neither mentions calcium); PubChem HSDB sections (no calcium text).

Channel outcomes

No source measuring total or ionized calcium with epinephrine was opened. Several further PubMed listing attempts returned 429 and the Europe PMC and Crossref calls for other records were rejected. Outcome: a genuine gap in what was retrieved plus tooling limits; the lead that exists is indirect (parathyroid hormone unchanged in [34]).

Limits on what this shows

• Why no grade: Dual condition met: the established COMT metabolism is a pharmacokinetic fact, not evidence of a nutrient effect, and no Tier B or C source on cofactor depletion was opened. A disclosed gap.

What was searched

Both DailyMed labels; PubChem pug_view metabolism section (DrugBank and HSDB text); Wikipedia; Drugs.com interactions index. Repair pass 2026-10-08: a PubMed listing for catechol-O-methyltransferase, S-adenosylmethionine and epinephrine methylation or depletion was attempted and rejected with 429, and was not retried on the same page.

Channel outcomes

The labels and HSDB confirm metabolism to metanephrine and vanillylmandelic acid by COMT and MAO, mainly in liver and kidney, with an intravenous half-life under 5 minutes [2][4]. The PubChem record states that S-adenosylmethionine and magnesium are not mentioned. The IV label notes that COMT inhibitors potentiate pressor effects [2]. No source on cofactor requirement or any measured SAMe or magnesium effect was opened. Direction check (Hard Rule 4): magnesium supporting COMT activity would be a nutrient acting on the drug, not the drug depleting the nutrient. Outcome: genuine gap in what was retrieved, plus tooling limits.