The suicide drop in Russia's Northwest landed on men the clinic rarely sees
- Register-based observational study of completed suicides in the Northwestern Federal District of Russia, published open access in Psychiatry (Moscow) on 25 November 2025, volume 23, issue 4, pages 15–26, by Rozanov and five co-authors from St Petersburg State University, the Bekhterev National Medical Research Centre, the Mechnikov North-Western State Medical University, the St Petersburg Bureau of Forensic Medical Examination and the Pavlov First St Petersburg State Medical University. No conflict of interest declared.
- Two continuous databases from regional forensic medical bureaus: 2,270 deaths in St Petersburg from 1 January 2017 to 31 December 2023, of which 1,668 men (73.48%) and 602 women (26.52%), and 1,321 deaths in Leningrad Region from 1 January 2019 to 31 December 2023, of which 1,080 men (81.76%) and 241 women (18.24%). Combined population above 7.6 million. A wider table covers all ten subjects of the district from 2017.
- In 2023 registered suicides fell 43.95% in St Petersburg (176 against 314) and 29.33% in Leningrad Region (171 against 242), while Pskov Region rose 7.53% and Vologda Region 44.26%. The male rate in St Petersburg was 4.79 per 100,000 (95% CI 4.01–5.73) against 9.60 (8.45–10.89) a year earlier; in Leningrad Region, 14.50 (12.30–17.10) against 20.50 (17.80–23.60).
- The decline concentrates in two age bands. St Petersburg men aged 40–59 fell to 4.93 (3.58–6.79) from 8.94 (7.00–11.41), and men over 60 to 6.73 (4.79–9.45) from 13.52 (10.57–17.30). The authors report the 2023 fall as statistically confirmed among men in those two bands only; among older women rates declined inside overlapping intervals, and among women under 39 the paper reports a tendency in the other direction.
- More than 96% of deaths in both subjects belong to the three generations above 20, spread evenly between them at 29% to 36% each. People under 19 account for fewer than 4% of cases: 87 deaths in St Petersburg across seven years and 51 in Leningrad Region across five.
Everyone counted in this paper was already dead when the record was made. The units are completed suicides registered by two regional forensic medical bureaus – 2,270 cases in St Petersburg across the seven years from 1 January 2017, and 1,321 cases in Leningrad Region across the five years from 1 January 2019. Men are 1,668 of the St Petersburg cases (73.48%) and 1,080 of those in Leningrad Region (81.76%). Denominators came from Rosstat, rates were computed per 100,000 per year, and confidence intervals by the Wilson method. Nobody in this dataset was ever a patient in it. It holds a death, a sex, an age and a year, and carries no diagnosis, no service contact and no attempt history.
Where the fall landed
The year 2023 broke the series. St Petersburg registered 176 suicides against 314 the year before, a fall of 43.95%, and Leningrad Region 171 against 242, a fall of 29.33%. Karelia halved, from 46 cases to 23. The movement was not district-wide: Pskov Region rose 7.53% that year and Vologda Region 44.26%.
In the two subjects with continuous databases, the fall belongs to men. The male rate in St Petersburg was 4.79 per 100,000 in 2023 (95% CI 4.01–5.73) against 9.60 (8.45–10.89) in 2022, and the authors note that the upper bound of the 2023 interval crosses the lower bound of no earlier year in the series. Men in Leningrad Region went from 20.50 (17.80–23.60) to 14.50 (12.30–17.10).
By age the drop narrows further. St Petersburg men aged 40–59 fell to 4.93 (3.58–6.79) from 8.94 (7.00–11.41), below every previous year. Men over 60 fell to 6.73 (4.79–9.45) from 13.52 (10.57–17.30), an interval that still overlaps the 2020 figure of 12.14 (9.31–15.84). Female rates moved the same way and stayed inside overlapping intervals; the St Petersburg female rate of 1.72 (1.32–2.25) in 2023 sits below the 2021 peak of 4.04 (3.38–4.83). Rozanov and colleagues read the male fall through Durkheim and attribute it to social cohesion, optimism and military and economic mobilisation.
The part of the table a clinic works in
More than 96% of deaths in both subjects belong to the three generations above 20, divided evenly between them at 29% to 36% each. People under 19 account for fewer than 4% of cases – 87 deaths in St Petersburg across seven years, 51 in Leningrad Region across five.
The sex ratio swings with age. In St Petersburg it runs 1.35 to 1 below 19, then 4.01 at 20–39, 3.63 at 40–59 and 1.71 above 60; Leningrad Region gives 1.55, 7.23, 6.23 and 2.90, with an overall 4.48 against the city's 2.68. Among the under-19s the male and female curves overlap within their intervals in both subjects, and in Leningrad Region in 2023 the female rate sat above the male, which the authors record as non-significant. For St Petersburg women below 19 and aged 20–39 they describe a weak upward tendency across the whole period.
That arithmetic is where the two readings separate. The groups carrying the headline decline are men in their forties, fifties and sixties, the least likely here to have reached a psychiatrist before dying. The group that fills an adolescent ward sits under 4% of the register and therefore cannot shift the regional index whatever it does. A falling rate for the subject is not a statement about the people booked into next week's clinic, and this register was never built to make one: it carries deaths, not patients.
More than 96% of registered suicides in both subjects belong to the three generations above 20, and people under 19 account for fewer than 4% of cases – 87 deaths in St Petersburg across seven years – so whatever happens in the youngest group cannot move the regional rate in either direction.
The paper carries no limitations section, and the constraints follow from its design. This is a mortality register rather than a clinical study: it holds no diagnoses, no contacts with psychiatric services, no history of attempts and no individual-level exposures, so nothing in it identifies who is at risk. The authors note that some completed suicides in Russia may be recorded under other headings, such as deaths of undetermined intent, accidents or unintentional poisoning by alcohol or drugs, and they report comparing the trend against other external causes and finding similar movement, but those comparison figures are not printed. The paper confirms the 2023 fall statistically for men in two age bands only, and the counts beneath the female trends are small: 37 female deaths below 19 in St Petersburg across seven years and 20 in Leningrad Region across five. The two databases cover unequal periods, seven years against five. The explanation offered for the 2023 fall is sociological and untested in these data, and rurality, access to care and alcohol consumption are discussed as reasons for the gap between the two subjects without being measured here. Rates rest on annual average population figures, and the detailed analysis covers two subjects out of ten.