Psychiatrist Arthur Sackler created the ads that made benzodiazepines popular. His nephew, Richard Sackler, took that to a new level with opioids, resulting in nearly a million U.S. deaths and the removal of the family name from over 20 institutions.
At high doses, overdose deaths rise sixfold, but what is the cause?
STUDY: Särkilä H et al, Acta Psychiatr Scand 2026
STUDY TYPE: Nationwide cohort study
FUNDING: Turku Psychiatric Services; Turku University Central Hospital
Background
We know the risks of high dose benzos: Falls, accidents, impaired cognition, respiratory problems, and worse withdrawals. This Finnish study looks at the mortality risk.
The Study
- 48,124 new benzodiazepine users ages 18–65, who had not taken benzos for two years before enrollment.
- For five years, doses were tracked at each prescription fill and classified by Diazepam Daily Dose Equivalents (DDD) as low (<1.0 defined daily doses/day), medium to high (1.0–3.0 DDDs/day), or very high (≥3.0 DDDs/day). One DDD equals 10 mg of diazepam.
- Primary outcome: all-cause mortality. Secondary outcomes: overdose, suicide, and accidental death.
- Data adjusted for: sex, age, disability, education, receipt of social benefits, the use of gabapentinoids/opioids, number of psychiatric disorders, number of substance use disorders, Charlson’s comorbidity score, and the order of treatments.
Results
Nearly 5% of the cohort (2,294 people) died during the five-year follow-up. Before we get to the mortality links, remember that sicker patients are prescribed higher doses.
Overall, benzodiazepine use was associated with a 28% higher mortality risk compared to non-use periods. The risk rose sharply with dose:
- Low dose: no increased risk (aHR 1.00).
- Medium to high dose: 58% higher all-cause mortality (aHR 1.58), with overdose deaths tripling (aHR 3.34) and suicides more than doubling (aHR 2.43).
- Very high dose: all-cause mortality nearly tripled (aHR 2.68), with overdose deaths rising sixfold (aHR 6.18), suicides more than fourfold (aHR 4.46), and accidental deaths almost fourfold (aHR 3.77).
Most very-high-dose periods involved more than one benzodiazepine at once. Clonazepam and alprazolam were most likely to lead to dose escalation, while oxazepam was least likely (consistent with earlier research showing greater misuse and overdose liability with those high-potency benzos, and less with the slow-to-act oxazepam).
Using three or more benzodiazepines together carried the highest mortality risk of any combination (aHR 3.12).
Combining a benzodiazepine with a Z-drug like zolpidem or zopiclone also raised risk substantially (aHR 1.86). Z-drugs used alone did not increase mortality.
Clonazepam and alprazolam were the drugs most often seen in the very-high-dose group, consistent with their known misuse potential.
Limitations
This is correlation, not causation. Data was adjusted for number of comorbidities, but not for severity or past suicide attempts.
Practice Implications
- At low doses, mortality risk looks comparable to non-use, and the danger zone is above 10 mg diazepam per day.
- However, whether this is due to the benzo or to other factors isn’t clear.
- Most observational studies find a higher suicide risk with benzodiazepines, and while that could be due to other factors, it’s concerning that we don’t see a lower risk.
- This adds to other reports suggesting clonazepam and alprazolam are the riskiest benzos, while oxazepam is lower risk. Lorazepam is also lower risk.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
