This was a nationwide, population-based cohort study in Finland with follow-up from January 1, 2011, through December 31, 2023. The cohort comprised 5,526,599 residents aged 5 to 95 years who had no recorded prevalent mental disorder at cohort entry. During follow-up, 1,463,064 individuals (26.5%) received at least one diagnosis of a mental, behavioural, or neurodevelopmental disorder.
Data were drawn from primary and specialist health care registers and national population registers in Finland.
Exposures were defined using ICD-11 classifications for mental, behavioural, and neurodevelopmental disorders. Psychiatric care setting was treated as a time-varying variable, distinguishing diagnoses made in specialist psychiatric care versus those made outside specialist services. Substance use disorder (SUD) status was also modelled as time-varying to capture comorbid SUD during follow-up.
Primary outcomes included all-cause mortality, estimated mortality rate ratios (MRRs), and differences in restricted mean survival time (RMST) over a 10-year horizon. Analyses adjusted MRR estimates across care-setting and SUD strata and examined how associations changed with time since first diagnosis.
Of the 5.53 million residents in the cohort, 1.46 million (26.5%) were diagnosed with a mental disorder during the 2011–2023 follow-up. The study reports substantial heterogeneity in excess mortality across specific diagnoses, clinical subgroups defined by care setting, SUD comorbidity, and time since diagnosis.
Among individuals aged 5–64 years with any mental disorder, adjusted MRRs across combinations of care setting and SUD status ranged from 1.54 (95% CI, 1.39–1.71) to 8.97 (7.86–10.24). This range indicates that excess mortality is context-dependent: higher values were observed in some clinical subgroups (for example, those with SUD or those treated in specialist settings) while lower values occurred in other combinations.
MRRs were highest immediately following the first recorded diagnosis and declined over the first 2 to 3 years of follow-up. This temporal pattern was consistent across diagnostic groups, indicating elevated short-term mortality risk after initial diagnosis that attenuates but may remain elevated for some disorders.
At three years after diagnosis, the study reports distinct patterns:
For depressive, anxiety or fear-related, and stress-related disorders among individuals without SUD who were treated outside specialist psychiatric care, MRRs at 3 years ranged from 0.88 (0.79–0.98) to 1.20 (1.12–1.29) in men and from 0.81 (0.73–0.89) to 1.13 (1.04–1.22) in women. These estimates indicate that, for some common disorders diagnosed and managed outside specialist psychiatric services and in the absence of SUD, the excess mortality was small or near null by 3 years.
For schizophrenia and other primary psychotic disorders, MRRs at 3 years remained elevated: 1.84 (1.60–2.12) in men and 1.55 (1.37–1.75) in women, indicating persistent excess mortality for psychotic disorders even several years after diagnosis.
The 10-year survival loss across all mental disorders was reported as 0.53 years (95% CI, 0.53–0.54) in men and 0.34 years (0.33–0.34) in women. The authors note that survival loss was greater for natural causes of death than for external causes.
The analysis used linked registers maintained by Statistics Finland and the Finnish Institute for Health and Welfare (THL). Ethical review was granted by the THL Institutional Review Board, and data linkage was authorized by Statistics Finland and THL. The study notes that data access is restricted and requires authorization from the respective Finnish authorities (www.findata.fi and www.stat.fi).
The study concludes that excess mortality associated with mental, behavioural, and neurodevelopmental disorders is markedly heterogeneous by diagnosis, clinical context (including care setting and presence of SUD), and time since diagnosis. Some common disorders diagnosed and managed outside specialist psychiatric care without comorbid SUD showed only small excess mortality by 3 years, whereas disorders such as schizophrenia and other psychoses retained substantially elevated MRRs. The authors highlight that mortality estimates derived solely from specialist psychiatric populations or averaged across follow-up may not accurately reflect risk across the broader diagnosed population.
Competing interests are declared: one author reported lecture fees from a pharmaceutical company outside the submitted work; other authors declared no competing interests.
Note: All numerical estimates, confidence intervals, cohort definitions, and outcomes above are taken directly from the source report.