What I could not verify

The largest increase in Europe

It is widely claimed that antidepressant consumption in Greece went from 19 to 66 defined daily doses per 1,000 inhabitants per day between 2000 and 2020 — an increase of 247%, the largest among European countries.

The arithmetic holds and the trend is confirmed by several converging sources. But I did not find the table: the study to which it is usually attributed covers 2010–2020 and mentions neither Greece nor this figure, and the report that gives the European average does not mention Greece. OECD.Stat did not yield the relevant table.

What is verified, for context: the average across 18 European countries rose from 30.5 to 75.3 DDD over the same twenty years, an increase of 147%. So even if the 66 is correct, in 2020 Greece stood below the European average in absolute terms. What would be notable is the rate, not the level.

[48] average across 18 European countries: analysis of OECD data, Euronews (Yanatma S.), 12.11.2022 — journalistic processing. The Greek figure remains unverified.

That Greeks consume two thirds more than the European average

The claim is attributed to a presentation of IQVIA data at a health economics conference. There is no publication: no journal article, no abstract, no institutional report, no posting in a university repository. The only source is a journalistic account of a spoken intervention.

It may be entirely correct. But a page that demands randomised-trial-level evidence from psychiatry cannot accept for itself a journalist's paraphrase of an unrecorded discussion.

Journalistic account of a conference intervention. No peer-reviewed publication or institutional report was located.

That the pharmaceutical companies silently abandoned the chemical imbalance theory

It circulates everywhere. No company has retracted or disavowed the theory in a statement, a regulatory filing, an internal document or a court record — no such primary source was located.

What is documented is narrower and must be stated as such: the advertising language softened over time — there is a peer-reviewed analysis of the advertisements themselves — and in one case a regulator imposed the withdrawal of a form of wording from a leaflet. Imposition is not abandonment.

No primary source was located. The existing references are secondary or non-scientific.

That Open Dialogue achieved 80% recovery without drugs, versus 5% with drugs

This is not merely unverified — it is wrong, and we report it because it circulates in the very community that is regarded as an ally of this page.

The “80%” and the “5%” come from two entirely unrelated studies: the first from Western Lapland in Finland (absence of residual psychotic symptoms, 82% at 5 years), the second from an American study in Chicago in a different era, with a different cohort, diagnosis and outcome definition. Neither of the two contains the other's figure.

And the “without drugs” does not hold: 29% of the Finnish cohort used neuroleptics at some point and 17% were taking them at the five-year follow-up.

We reject it explicitly. An argument does not become correct because it is convenient.

A splice of two unrelated studies. The claim is supported by neither of them.

That psychiatric drugs are the third leading cause of death, after heart disease and cancer

It is perhaps the most widely circulated number in this field, and you will have seen it in many forms: “prescription drugs are the third leading cause of death”, “psychiatric drugs kill 539,000 people a year in the US and EU”. We followed the chain to its source. It is not a count. It is an estimate by extrapolation — and that is how its own author describes it; the word simply gets lost on the way.

What it rests on. The version about drugs in general was published in a medical journal in 2014, in a controversies section, with no data of its own; it cites a book by the same author. The version about psychiatric drugs exists in peer-reviewed form once, as far as we found: in a BMJ head-to-head debate in 2015, where the author explains the method. He takes mortality rates from studies — 1% a year for antipsychotics, starting from a meta-analysis in elderly people with dementia; 1% for benzodiazepines; 2% for antidepressants — applies them to Danish consumption at ages 65 and over, arrives at 3,693 deaths a year in Denmark, and scales them up to 539,000 for the US and EU. The phrase “third leading cause of death” does not appear in the BMJ text; it appears in the books and in website articles, from comparing 539,000 with heart disease at 600,000 and cancer at 575,000 in the CDC list for 2010.

Why it cannot stand as a number. Three points, none of which requires doubting the input studies. First, the CDC list of “leading causes of death” classifies the underlying cause recorded on the death certificate under ICD-10; adverse drug reactions have codes (Y40–Y59) but are rarely certified as the underlying cause — a few thousand deaths a year — so the “third place” is a comparison of an estimate with a classification that measures something else. Second, the 1% rate for antipsychotics comes from dementia patients in trials of 10–12 weeks — the population with the highest known risk, the same one that produced the finding in Part B — and is applied to all users aged 65 and over. Third, the scaling from 3,693 to 539,000 is a simple population ratio — about 146 times Denmark — so whatever holds for the input rates is multiplied with them. A rebuttal by Cochrane researchers was published in the same journal a few weeks later: that the debate's recommendations are not supported by the reviews it invokes.

The author himself, from the 2015 book to the 2024 version, also notes the limitation that decides everything: “you can only die once, and many people receive polypharmacy”. We keep it as what it is — an argument that deaths from psychiatric drugs in older people are underestimated, which this page shares on the strength of Part B. Not as the third number in a list on which it does not appear.

[87] Gøtzsche PC. Pol Arch Med Wewn 2014;124(11):628–634 · [88] Gøtzsche PC. BMJ 2015;350:h2435 · [89] Tovey D et al. BMJ 2015;350:h2952 · [90] Gøtzsche PC. Deadly Psychiatry and Organised Denial, 2015, ch. 14 · [91] Gøtzsche PC, website article, 16.04.2024 · [92] Lazarou J et al. JAMA 1998;279(15):1200–1205 · [93] Kvasz M et al. MedGenMed 2000;2(2):E3 · [94] NCHS Data Brief 548, 2026