Key points are not available for this paper at this time.
In their recent article, Plana-Ripoll and colleagues investigate temporal trends in mortality rates among individuals treated for mental disorders over a 14-year period in Denmark 1. This well-conducted study, which included nationwide data from over 7 million individuals, compared standardised mortality rates (SMRs) for individuals with mental disorders and members of the general population across five time-periods, including the COVID-19 pandemic (2010–2012 vs. 2013–2015 vs. 2016–2018 vs. 2019–2021 vs. 2022–2023). The authors reported that whilst SMRs reduced over time for most mental disorders (including depression, anxiety and personality disorders) or remained stable (bipolar disorder), those for schizophrenia spectrum disorders (SSDs), organic disorders and (to a lesser extent) substance use disorders increased. The worsening outcomes observed for people with SSDs are in line with findings from a recent study examining temporal trends in life expectancy among individuals with severe mental illness in Scotland 2: In that study, the life expectancy gap relative to the general population was unchanged for individuals with major depression and bipolar disorder between 2000 and 2019, yet the gap increased over the study period for those with schizophrenia. In contrast, a study conducted in Hong Kong observed that SMRs for people with schizophrenia showed minimal change over time, whilst those for other SSDs were slightly lower in 2012–2016 compared to 2006–2011 3. Together, these investigations provide evidence of a persistent mortality gap among individuals with SSDs, which in the case of Denmark and Scotland, has widened over time. The trends observed in Denmark are particularly concerning given the significant investment in early intervention (EI) services (via the OPUS programme) over the past three decades 4. Indeed, Denmark was among the first countries to adopt the EI model, which has been implemented nationally since 2003 5. This editorial discusses the implications of these findings for EI services and provides suggestions for future research to enable us to understand and address the persistent mortality gap for people with SSDs. By definition, SMRs are relative and therefore directly influenced by mortality rates in the general population as well as rates in the subpopulation of interest. The observed increase in SMRs for individuals with SSDs may therefore be due to a reduction in mortality rates in the general population which has not occurred (or at least not to the same extent) among individuals with SSDs, or an actual increase in mortality rates among individuals with SSDs. Thus, it is essential to consider both absolute and relative measures of mortality. A strength of the paper by Plana-Ripoll is that both outcome measures are presented: Here, we see that the crude (age-standardised) mortality rate per 1000 person-years among individuals with SSDs in fact decreased at each subsequent time-period within the first 12 years (falling from 21.99 in 2010–2012 to 18.83 in 2019–2021) but rose during the final 2-year period (increasing to 20.49 in 2022–2023). Importantly, this same pattern was observed for deaths due to natural and external causes and occurred in both men and women. Similarly, the crude rates provided for the general population (age-standardised to all individuals with mental disorders) showed a decrease in mortality rates during the first 12 years followed by an increase in the final 2 years, coinciding with the COVID-19 pandemic. These findings suggest that the improvements in population health in Denmark were not observed to the same degree among individuals with SSDs, but that the COVID-19 pandemic (and associated restrictions) may have had a more negative effect on the health of people with these disorders. Several European countries, including Denmark, experienced an increase in life expectancy in the three decades prior to the COVID-19 pandemic 6, which has been attributed to improvements in health behaviours (e.g., diet, physical activity, tobacco/alcohol consumption) and healthcare access. Clearly, there is an urgent need to understand why individuals with SSDs have not benefited from these population health improvements. One possible explanation for the increased mortality gap for people with SSDs is that the beneficial effects of EI, as seen in randomised controlled trials (RCTs), are diminished when implemented in real-world settings (i.e., the ‘efficacy-effectiveness gap’). However, the initial OPUS trial observed no significant group differences in suicide attempts or suicidal thoughts in individuals randomised to OPUS (N = 275) versus standard care (N = 272) at the 1-, 2-, or 5-year follow-ups 7, 8. Furthermore, a recent follow-up study observed that all-cause mortality rates did not differ among OPUS and standard care groups, but that individuals identified in the Danish registers who received OPUS in real-world settings (N = 3328) had significantly lower all-cause mortality rates (crude analyses only) when compared to those who received OPUS in the initial RCT 7. Thus, there was no evidence to suggest that any beneficial effect of OPUS on all-cause mortality was diminished in real-world settings. In contrast, a recent meta-analysis of controlled studies reported that EI was associated with a reduction in deaths due to suicide 9: Notably, all three of the five original studies which reported a significant reduction in suicide deaths evaluated the Early Assessment Service for Young people with psychosis (EASY 10) in Hong Kong. Disentangling which specific components of this population-based EI programme may have led to a reduction in suicide risk among individuals with SSDs is therefore a high priority. The extent to which interventions can improve physical health in this population is unclear. Indeed, the CHANGE trial—which included 428 individuals with SSDs who had abdominal obesity—observed no significant differences in the 10-year risk of cardiovascular disease in those randomised to receive 12 months of lifestyle coaching plus care coordination plus treatment as usual versus care coordination and treatment as usual versus treatment as usual alone 11. Clearly, further work is needed to develop more effective treatments for these outcomes in people with SSDs and ensure that these interventions can be feasibly implemented in real-world settings. Attempts to reduce the mortality gap among individuals with SSDs, either via public health initiatives or specialised healthcare, will be more likely to succeed if these interventions align with patients needs and preferences. Studies investigating patient-reported treatment priorities and unmet needs can therefore provide insights into the factors that may contribute to the persistent mortality gap. Unfortunately, previous large-scale surveys of perceived care needs 12 and treatment preferences 13 among individuals with SSDs have used structured tools which do not include domains pertaining to suicidality or physical health. Notably, one of these studies asked patients to list other areas (outside of domains listed in the questionnaire) where they would like psychological help 13 and found that ‘physical health’, ‘physical exercise’ and ‘weight management/diet’ were each reported by only ~1% of participants whilst help with ‘self-harm’ and ‘suicidal ideation’ were reported by 0.3% and 0.2% of respondents, respectively. Similarly, in a Delphi study where individuals with psychotic disorders rated 77 items pertaining to treatment priorities and preferences according to their importance, none of the 15 items rated as essential or important by ≥ 80% of the participants related to physical health or suicidality 14. These preliminary findings suggest that improving physical health and reducing suicidal behaviour may not be as important to individuals with SSDs as other aspects of their health. This may be because positive psychotic symptoms and other psychiatric disturbances (e.g., anxiety, stress, worry and insomnia) are more severe, frequent and distressing than concerns regarding physical health and suicidality. It is also possible that individuals with SSDs are not sufficiently aware of greatly increased risks of these outcomes, or that they are considered to be too distal to be of immediate concern. Understanding why these outcomes are not of greater importance to individuals with SSDs is needed to improve the success of targeted interventions. The persistent 3 and widening 1, 2 mortality gaps observed among individuals with SSDs indicate a clear need for novel, evidence-based approaches to improve outcomes for people living with these conditions. As argued above, these changes will be difficult to achieve if we focus only on the role of psychiatric services. Indeed, such approaches must integrate public health perspectives (ensuring that mortality improvements in the general population also benefit individuals with SSDs), clinical perspectives (developing and implementing treatments in EI services that can effectively reduce the risk of deaths due to both natural and unnatural causes), and patient perspectives (to understand why these outcomes may not be recognized as areas of need and how this impacts treatment engagement and response). To achieve this, significant investment in psychiatric research and health care is urgently needed. Whilst many countries are already taking steps towards this goal, more ambitious, adequately funded, long-term research programmes may be needed before we can see real change. Moving forward, we must integrate population health, clinical and patient perspectives in order to better understand why attempts to reduce the mortality gap for people with SSDs have not worked in order to improve the long-term outcomes for this population. A.E.C. drafted and edited the manuscript. The author has nothing to report. The author has nothing to report. The author declares no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.
Alexis E. Cullen (Fri,) studied this question.