Abstract
Background
People with mental disorders have increased mortality rates and reduced life expectancies. We recently found that, compared to the general population, men and women with any mental disorder experienced 10 and 7 years, respectively, of life-years lost (LYLs), a new metric to estimate reduced life expectancy that takes into account the age of onset of the disorder. Our aim is to examine changes in mortality rate ratios (MRRs) and LYLs for both external and natural causes over twenty years for a comprehensive range of mental disorders, including schizophrenia spectrum disorder.
Methods
We conducted a cohort study comprising all 7,369,926 people living in Denmark in 1995–2015. Information on mental disorders and mortality was obtained from national registers. We looked at all mental disorders combined and specific groups of diagnoses as defined by the ICD-10 F-subchapters (substance use disorders, schizophrenia spectrum disorder, mood disorders, neurotic disorders, etc.) and classified causes of death into natural and external causes. We estimated MRRs using Poisson regression models, adjusting for sex and age and including an interaction term with calendar time. Differences in remaining life expectancy after disease diagnosis were estimated as excess LYLs (divided into LYLs due to natural and external causes of death) between those with each disorder and the general Danish population (matched on sex and age) for specific periods separately (1995–1999, 2000–2004, 2005–2009, 2010–2015).
Results
Over the period of observation, mortality rates decreased for those with any diagnosed mental disorder, as well as for those without a diagnosis. Despite these improvements, the MRRs between the two groups increased from 2.38 (95% CI: 2.32–2.44) in 1995 to 2.60 (95% CI: 2.55–2.65) in 2015. For external causes of death, MRRs decreased from 6.64 (95% CI: 6.15–7.17) to 5.27 (95% CI: 4.87–5.70), while MRRs for natural causes increased from 2.19 (95% CI: 2.14–2.25) to 2.52 (95% CI: 2.47–2.56). Remaining life expectancy after disease diagnosis increased 4.6 years from 32.0 to 36.6 years; however, remaining life expectancy increased also in the matched general population of same age and sex by 3.2 years (from 41.7 to 44.9 years). The life expectancy gap between the two periods was therefore shortened by 1.4 years; excess LYLs were 9.7 years in 1995–1999 (5.8/3.8 years due to natural/external causes) and 8.3 years in 2010–2015 (6.6/1.7 years due to natural/external causes). When looking at specific mental disorders, the life expectancy gap was reduced for mood disorders (0.8 years), neurotic disorders (1.7 years), and personality disorders (0.9 years); remained similar for schizophrenia spectrum disorder and substance use disorders; and increased for organic disorders (1.1 years).
Discussion
Mortality rates for people experiencing mental disorders decreased from 1995 to 2015. However, for natural causes of death, those with mental disorders did not reflect the benefits seen in the general population. Consequently, life lost due to natural causes increased. Overall, life expectancy increased an additional 1.4 years for those with mental disorders compared with the general population, thus reducing the gap. Nevertheless, for some disorders e.g. schizophrenia spectrum disorder and substance use disorders, life expectancy gap did not change. These findings support the hypothesis that service improvements have reduced mortality due to suicide and accidents, but similar benefits are not apparent in natural causes of death, which suggests that interventions related to promoting a healthier lifestyle and optimizing the general medical care of those with mental disorders warrants added investment.
High hopes: lower risk of death due to mental disorders and self-harm in a century-long US Olympian cohort compared with the general population
