Research paper
DeFronzo RA. The effect of insulin on renal sodium metabolism. A review with clinical implications. Diabetologia. 1981;21(3):165-171.
Why we cite it
Provides the physiological basis for the companion’s core mechanism: insulin, not calorie count per se, governs renal sodium balance. Physiological increases in plasma insulin stimulate sodium reabsorption in the distal nephron independently of other hormones or metabolites; conversely, when insulin falls (as during fasting or carbohydrate restriction), the kidney excretes sodium — the “natriuresis of starvation.” Refeeding raises insulin and produces sodium retention (“anti-natriuresis of refeeding”). This upgrades the corpus’s support for fasting-related electrolyte loss from a popular-science source (cfa10, Phinney & Volek) to a primary peer-reviewed review.
RE-REVIEWED 2026-09-10 (card MERNGY), for companion 406 — the figures hold, and the paper’s insulin unit does not.
– Every figure below is in the held full text: the >50% decline under the dog euglycaemic clamp, ~35% with hyperglycaemic hyperinsulinaemia, 35–50% in the micropuncture work, the ~50% single-kidney result, ~30–40 mmol/day over the first 3 days of fasting (Kolanowski), and up to one-third of the natriuresis attributed to acidosis (Stinebaugh and Schloeder).
– THE PAPER PRINTS INSULIN IN nmol/l, AND ITS OWN PARENTHETICALS SHOW THE UNIT IS pmol/l — a factor of a thousand. ”215-287 nmol/l (or 30-40 mU/1)”; a baseline of ”43 ± 7 nmol/1 or 6 ± 1 mU/l”; a clamp at ”1070 ± 108 nmol/1 or (149 ± 15 mU/1)”. Each pair converts at about 7.2 pmol per mU, the standard factor for insulin, so each printed number is a pmol/l value. The same unit appears in every insulin occurrence checked, so it is the paper’s own (or a systematic OCR p→n), not a one-off. This record copies it faithfully into its single-kidney bullet (“215–287 nmol/l”). Cite the mU/l figures, or give pmol/l with this note — never print nmol/l as an insulin concentration from this paper.
– ”Within 60 minutes” is in the paper, and a search will say it is not: the OCR renders min as rain — ”within 60 rain a highly significant reduction was documented.” Recorded so the next checker does not report the claim missing.
– One softening, noted and not repaired here: the second-to-last bullet says glucagon and aldosterone were ”ruled out as the primary drivers.” The paper is gentler: other studies ”suggest that changes in plasma glucagon levels within the physiological range have little effect on renal sodium handling”, and it argues against aldosterone because levels rise in early starvation while sodium excretion rises too, with normal tubular responsiveness. Argued against, not ruled out.
– The author note’s ”among the most highly cited researchers” was an unsurveyable ranking. It was removed the same afternoon (card E1AWST), along with ”internationally recognised for foundational work”.
About the author
Ralph A. DeFronzo, MD, is a diabetologist, Professor of Medicine, and Chief of the Diabetes Division at the University of Texas Health Science Center at San Antonio. His research includes insulin resistance and the euglycemic insulin clamp technique. At the time of this 1981 paper he was at the Divisions of Endocrine-Metabolism and Nephrology, Yale University School of Medicine — much of the primary clamp and micropuncture work reviewed here is his own.