do about 1% of autopsies show wernicke's encephalopathy, mostly missed in life?
“around one percent of autopsy series have orna keys and cephalopathy found you know and most of the cases are not diagnosed in life”
a host, a psychiatrist said on Psychiatry & Psychotherapy Podcast, 2026-11-21, at 31:34. hear it in context → the speaker cited: autopsy series.
the episode: Wernicke's Encephalopathy: Why We Miss It, How to Diagnose It, and How to Dose Thiamine (at 31:34). we link the show's own page; we do not re-host audio.
first, the part that holds
Both parts of the claim hold up. Unselected autopsy series do find wernicke's encephalopathy lesions in roughly 1% to 2% of brains, and in every series we read most of those cases had not been diagnosed before death.
what we found when we went and looked
We read four necropsy reports. One Perth series found wernicke lesions in 1.7% of all necropsies, and only 7 of its 51 cases had been diagnosed in life. A 1991 review gives about 2% for unselected autopsy series from the US and Australia and 0.6%-0.8% for Oslo. It says only 1%-20% of autopsy-found cases had been diagnosed clinically. A Perth analysis of 131 necropsy cases found that 80% had not been diagnosed in life. A South African series found 6.6%, with no case diagnosed in life. "Around one percent" is the low end of that range, near the Oslo figure. The larger series cluster near 2%, so the claim, if anything, understates the problem.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| around 1% of autopsies in autopsy series show wernicke's encephalopathy lesions | undersold | Perth: 1.7% of all necropsies over four years [1]. Unselected US and Australian series: about 2%. Oslo: 0.6%-0.8% [3]. A South African series: 6.6% [4]. A figure near 1% is defensible, but it sits at the low end. Most unselected series from the US and Australia report about 2% (moderate confidence that 'around 1%' understates the typical figure). [1][3][4] |
| most autopsy-detected cases were not diagnosed during life | confirmed | In Perth, 7 of 51 cases (14%) were diagnosed in life, so 86% were missed [1]. In the 131-case analysis, 80% were not diagnosed in life [2]. The 1991 review: only 1%-20% of autopsy-found cases were diagnosed clinically [3]. In the South African series, none of the cases was diagnosed clinically [4]. High confidence. [1][2][3][4] |
| the classic clinical triad is usually absent in cases found at autopsy (context for why cases are missed) | confirmed | In the 131-case necropsy analysis, only 16% had the classical triad and 19% had no documented clinical signs [2]. This is a retrospective chart review, so some signs may have been present but never recorded. [2] |
both directions
for the speaker: The low-end figure is real: the review gives 0.6%-0.8% for Oslo [3]. On the second clause the record is consistent. Each series we read found that roughly 80% to 100% of autopsy-detected cases had no diagnosis in life [1][2][3][4]. The 131-case analysis suggests one reason: only 16% of these patients showed the textbook triad [2].
against: The magnitude runs mostly above the speaker's number, not below it. Perth found 1.7% [1], US and Australian unselected series about 2% [3], and a South African series 6.6% [4]. The speaker named no specific series, so "around one percent" cannot be traced to a single source. It reads as a conservative rounding of a range that varies by country and by how much heavy drinking the autopsied population had.
what this cannot say
These are mostly hospital and coroner series from the 1970s to the 1990s. They overrepresent people who died of alcohol-related or unexplained causes, so they are not general-population prevalence. Diagnosis rates in life may have improved since then with wider use of mri and routine thiamine in at-risk patients. We did not find a recent unselected autopsy series to test that. The 131-case paper's lack of signs comes from a retrospective chart review, which counts what was written down, not what was present. We read the abstracts, not the full texts.
sources, each one fetched and read
- Harper C. Wernicke's encephalopathy: a more common disease than realised. A neuropathological study of 51 cases. J Neurol Neurosurg Psychiatry, 1979. PMID 438830 What we read there: 51 cases diagnosed at necropsy over four years at a Perth hospital and by the Perth City coroner, 1.7% of all necropsies performed. Only seven had been diagnosed during life. Fetched 2026-11-21 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC490724/
- Harper CG, Giles M, Finlay-Jones R. Clinical signs in the Wernicke-Korsakoff complex: a retrospective analysis of 131 cases diagnosed at necropsy. J Neurol Neurosurg Psychiatry, 1986 What we read there: 80% of patients with the Wernicke-Korsakoff syndrome were not diagnosed as such during life. Only 16% had the classical clinical triad, and 19% had no documented clinical signs. Fetched 2026-11-21 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC1028756/
- Torvik A. Wernicke's encephalopathy--prevalence and clinical spectrum. Alcohol Alcohol Suppl, 1991;1:381-4. PMID 1845567 (PubMed abstract via NCBI E-utilities) What we read there: Unselected autopsy materials from the United States and Australia give prevalence figures of 2%. Oslo figures are 0.6%-0.8%. Only a fraction of the cases discovered at autopsy have been diagnosed clinically (1%-20%). Fetched 2026-11-21 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/1845567/
- Autopsy prevalence of Wernicke's encephalopathy in alcohol-related disease. S Afr Med J, 1996. PMID 8888781 (PubMed abstract via NCBI E-utilities) What we read there: A high prevalence of wernicke's encephalopathy was found in adult autopsies (6.6%). A clinical diagnosis was not made in any of the cases studied. Fetched 2026-11-21 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/438830/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam