Recent trends in longevity promise longer life spans for each new generation in many countries of the world (Oeppen & Vaupel, Reference Oeppen and Vaupel2002). However, reaching higher ages also implies an elevated risk of developing dementia (Corrada et al., Reference Corrada, Brookmeyer, Paganini-Hill, Berlau and Kawas2010). Also, suicide rates have consistently been found to be high among older adults, especially at very high ages (Shah et al., Reference Shah, Bhat, Zarate-Escudero, DeLeo and Erlangsen2016). It is, thus, of relevance to consider whether dementia may be linked to elevated risks of suicide. Given that suicide is a relatively rare event, the research findings, which have sought to answer this question, have been of a lower level of evidence, and data on large samples have only been available in few countries (Erlangsen et al., Reference Erlangsen, Zarit and Conwell2008; Haw et al., Reference Haw, Harwood and Hawton2009).
The rate of suicide deaths is highest among older adults in many high-income countries when compared to younger age groups (Shah et al., Reference Shah, Bhat, Zarate-Escudero, DeLeo and Erlangsen2016). It has, however, been more complicated to identify the true rate of non-fatal events, i.e. suicide attempts, as these might escape detection among older adults. For instance, it might be challenging to determine whether an overdose of a prescribed medication originated from an intentional or accidental act (Qin & Mehlum, Reference Qin and Mehlum2020). As with suicide deaths, older adults who have a suicide attempt seem to be more determined when compared to younger adults. This is also supported by the finding that risk factors associated with serious suicide attempts have been shown to resemble those, which are linked to suicide deaths (Beautrais, Reference Beautrais2001). In other words, older adults who have a suicide attempt might be comparable to those who die by suicide in terms of what sets them at elevated risks of suicidal behaviors and it can be meaningful to examine the two outcomes jointly, especially when studying rare events.
Obtaining representative data for individuals diagnosed with dementia is difficult. In countries where linkage or administrative data are available, hospital registers may provide a helpful resource for identifying individuals diagnosed with dementia, albeit only being available for those individuals who were given diagnosis in a hospital setting (Erlangsen et al., Reference Erlangsen, Qin and Mittendorfer-Rutz2018). A substantial proportion of individuals with dementia are likely to only be diagnosed in the primary care sector. Clinical databases, such as the one used by Hedna et al. (Reference Hedna, Sigström, Johnell and Waern2024), provide therefore a unique resource in terms of having data on virtually all individuals who had been diagnosed with dementia either in a primary care office or a memory clinic and were residing in Sweden. Further, the database was based on validated diagnoses and included other relevant and detailed information.
By analyzing data on a large and community-based population of individuals with dementia, valuable insights were secured regarding the possible association between dementia and suicide for individuals aged 75 years and over (Hedna et al., Reference Hedna, Sigström, Johnell and Waern2024). Having data on newly recorded cases in the Swedish Dementia Registry over a 10-year period, the authors had access to detailed information regarding type of dementia and cognitive status at the time of diagnosis, as measured by the Mini-Mental State Examination (MMSE). Using a unique personal identifier, data from the dementia database were linked with national Swedish register data (Erlangsen et al., Reference Erlangsen, Qin and Mittendorfer-Rutz2018). This provided individual-level information regarding in-home care, prescribed medication for mental disorders, and in- and outpatient hospital care for mental disorders, such as depression. Among these data, were several covariates, such as MMSE scores and in-home care, which not previously have been examined for national samples and in relation to dementia. Likewise, information on suicide attempts and suicides was available via national register data and examined jointly as suicidal behavior.
Dementia was found to be associated with suicidal behavior (Hedna et al., Reference Hedna, Sigström, Johnell and Waern2024). While suicidal behavior may be considered an extreme act of despair, it is likely that many more individuals with dementia are exposed to similar stressors and that their mental well-being is affected.
In adjusted analyses, individuals with dementia who had a history of a mental disorder or previous suicide attempt were found to have elevated rates of suicidal behavior when compared to those with no such history. Individuals with physical frailties, measured through hospital diagnoses or by being a recipient of home care, were also found to have excess risk of suicidal behavior. These findings are in line with general findings for older adults, which have shown that vulnerable groups have elevated risks of suicidal behavior (Conwell et al., Reference Conwell, Duberstein, Hirsch, Conner, Eberly and Caine2010; Fassberg et al., Reference Fässberg, Cheung, Canetto, Erlangsen, Lapierre, Lindner, Draper, Gallo, Wong, Wu, Duberstein and Wærn2016). Concerningly, individuals with dementia who were born outside the country were also found to have elevated risks of suicide when compared to those born in the country. It could, however, not be established whether these individuals were foreign nationals and, if so, what was their country of origin. Depending on this, information on dementia and psychoeducation regarding mental well-being for individuals of non-Swedish origin might be indicated.
Interestingly, individuals with milder stages of dementia seemingly have higher risks of suicidal behavior when compared to those in more advanced stages. Adjusted analyses of Danish linkage data during 1980–2016 and presented as supplementary material showed that rates of suicide were 3-fold (aIRR, 3.0, CI 95%, 1.9–4.6) higher among those who had been first-time diagnosed with dementia within the last month when compared to individuals who had not been diagnosed with dementia (Erlangsen et al., Reference Erlangsen, Stenager, Conwell, Andersen, Hawton, Benros, Nordentoft and Stenager2020). Using the same comparison group, significantly lower rates were found for individuals who had been first-time diagnosed 1–2 years (aIRR, 0.7, CI 95%, 0.5–0.9) and 3–5 years (aIRR, 0.5, CI 95%, 0.4–0.7) ago. Although using different measures, these findings might be comparable and relate to the same phenomena. There are different possible explanations for these observations. Firstly, changes in cognition, which occur due to the progression of the disorder, may reduce physical abilities for carrying out a suicidal act. Thus, a lower the risk of suicide among individuals in an advanced state of dementia is likely to be the case, i.e. some years after first diagnosis (Erlangsen et al., Reference Erlangsen, Stenager, Conwell, Andersen, Hawton, Benros, Nordentoft and Stenager2020). Secondly, receiving the news that one has a severe, chronic disorder can be very distressing, and elevated risks of suicide have been demonstrated during the first months after being diagnosed with a range of physical disorders (Erlangsen et al., Reference Erlangsen, Stenager, Conwell, Andersen, Hawton, Benros, Nordentoft and Stenager2020; Petersen et al., Reference Petersen, Stenager, Mogensen and Erlangsen2020). Both findings point to the time of the diagnosis being a crucial point for intervention.
Although most people have some knowledge about dementia and the course of the disorder, they may lack insights regarding treatment and support options, which can help improve quality of life also in advanced stages of the disorder. Specific personality traits have been suggested as prevalent among those older adults who died by suicide, for instance, a wish for “being in control” of one’s life (Kjolseth et al., Reference Kjolseth, Ekeberg and Steihaug2010). Consequently, fear of losing this autonomy when becoming dependent on help from others may lead to suicidal thoughts. The anticipation of these problems, and a wish to act while one is still apt to do so, might explain the higher risk of suicidal behavior during early stages of the disorder. Also, it underscores some opportunities for intervention. Previous efforts of psychoeducation in combination with assignment of a case manager provided in primary care settings have shown promising results in terms of reducing levels of suicidal thoughts among older adults with mood disorders (Gallo et al., Reference Gallo, Morales, Bogner, Raue, Zee, Bruce and Reynolds2013; Unutzer et al., Reference Unützer, Tang, Oishi, Katon, Williams, Hunkeler, Hendrie, Lin, Levine, Grypma, Steffens, Fields and Langston2006). Community-based efforts to support individuals with dementia and their informal caregivers exist in several countries and might be intensified around the time of diagnosis.
In the study by Hedna et al. (Reference Hedna, Sigström, Johnell and Waern2024), being of older age or recipient of home care was also linked to an excess risk of suicidal behavior. These markers may, in principle, be related to the above-mentioned features of a more advanced stage of dementia, while higher risks were found for those of younger age and those who did not require home care. Previous findings have suggested that individuals who experience an onset of dementia at an unexpected age, i.e. 50–69 years versus 70 years and over, were associated with higher rates of suicide (Erlangsen et al., Reference Erlangsen, Zarit and Conwell2008).
A limitation of studies based on secondary data is that information on relevant factors, which might provide an understanding of causal links, may not be available. Based on the existing evidence, it seems that efforts to prevent adverse mental health and suicidal thoughts should best be offered to individuals in milder stages of dementia, ideally at the time of first diagnosis.
You have
Access
Recent trends in longevity promise longer life spans for each new generation in many countries of the world (Oeppen & Vaupel, Reference Oeppen and Vaupel2002). However, reaching higher ages also implies an elevated risk of developing dementia (Corrada et al., Reference Corrada, Brookmeyer, Paganini-Hill, Berlau and Kawas2010). Also, suicide rates have consistently been found to be high among older adults, especially at very high ages (Shah et al., Reference Shah, Bhat, Zarate-Escudero, DeLeo and Erlangsen2016). It is, thus, of relevance to consider whether dementia may be linked to elevated risks of suicide. Given that suicide is a relatively rare event, the research findings, which have sought to answer this question, have been of a lower level of evidence, and data on large samples have only been available in few countries (Erlangsen et al., Reference Erlangsen, Zarit and Conwell2008; Haw et al., Reference Haw, Harwood and Hawton2009).
The rate of suicide deaths is highest among older adults in many high-income countries when compared to younger age groups (Shah et al., Reference Shah, Bhat, Zarate-Escudero, DeLeo and Erlangsen2016). It has, however, been more complicated to identify the true rate of non-fatal events, i.e. suicide attempts, as these might escape detection among older adults. For instance, it might be challenging to determine whether an overdose of a prescribed medication originated from an intentional or accidental act (Qin & Mehlum, Reference Qin and Mehlum2020). As with suicide deaths, older adults who have a suicide attempt seem to be more determined when compared to younger adults. This is also supported by the finding that risk factors associated with serious suicide attempts have been shown to resemble those, which are linked to suicide deaths (Beautrais, Reference Beautrais2001). In other words, older adults who have a suicide attempt might be comparable to those who die by suicide in terms of what sets them at elevated risks of suicidal behaviors and it can be meaningful to examine the two outcomes jointly, especially when studying rare events.
Obtaining representative data for individuals diagnosed with dementia is difficult. In countries where linkage or administrative data are available, hospital registers may provide a helpful resource for identifying individuals diagnosed with dementia, albeit only being available for those individuals who were given diagnosis in a hospital setting (Erlangsen et al., Reference Erlangsen, Qin and Mittendorfer-Rutz2018). A substantial proportion of individuals with dementia are likely to only be diagnosed in the primary care sector. Clinical databases, such as the one used by Hedna et al. (Reference Hedna, Sigström, Johnell and Waern2024), provide therefore a unique resource in terms of having data on virtually all individuals who had been diagnosed with dementia either in a primary care office or a memory clinic and were residing in Sweden. Further, the database was based on validated diagnoses and included other relevant and detailed information.
By analyzing data on a large and community-based population of individuals with dementia, valuable insights were secured regarding the possible association between dementia and suicide for individuals aged 75 years and over (Hedna et al., Reference Hedna, Sigström, Johnell and Waern2024). Having data on newly recorded cases in the Swedish Dementia Registry over a 10-year period, the authors had access to detailed information regarding type of dementia and cognitive status at the time of diagnosis, as measured by the Mini-Mental State Examination (MMSE). Using a unique personal identifier, data from the dementia database were linked with national Swedish register data (Erlangsen et al., Reference Erlangsen, Qin and Mittendorfer-Rutz2018). This provided individual-level information regarding in-home care, prescribed medication for mental disorders, and in- and outpatient hospital care for mental disorders, such as depression. Among these data, were several covariates, such as MMSE scores and in-home care, which not previously have been examined for national samples and in relation to dementia. Likewise, information on suicide attempts and suicides was available via national register data and examined jointly as suicidal behavior.
Dementia was found to be associated with suicidal behavior (Hedna et al., Reference Hedna, Sigström, Johnell and Waern2024). While suicidal behavior may be considered an extreme act of despair, it is likely that many more individuals with dementia are exposed to similar stressors and that their mental well-being is affected.
In adjusted analyses, individuals with dementia who had a history of a mental disorder or previous suicide attempt were found to have elevated rates of suicidal behavior when compared to those with no such history. Individuals with physical frailties, measured through hospital diagnoses or by being a recipient of home care, were also found to have excess risk of suicidal behavior. These findings are in line with general findings for older adults, which have shown that vulnerable groups have elevated risks of suicidal behavior (Conwell et al., Reference Conwell, Duberstein, Hirsch, Conner, Eberly and Caine2010; Fassberg et al., Reference Fässberg, Cheung, Canetto, Erlangsen, Lapierre, Lindner, Draper, Gallo, Wong, Wu, Duberstein and Wærn2016). Concerningly, individuals with dementia who were born outside the country were also found to have elevated risks of suicide when compared to those born in the country. It could, however, not be established whether these individuals were foreign nationals and, if so, what was their country of origin. Depending on this, information on dementia and psychoeducation regarding mental well-being for individuals of non-Swedish origin might be indicated.
Interestingly, individuals with milder stages of dementia seemingly have higher risks of suicidal behavior when compared to those in more advanced stages. Adjusted analyses of Danish linkage data during 1980–2016 and presented as supplementary material showed that rates of suicide were 3-fold (aIRR, 3.0, CI 95%, 1.9–4.6) higher among those who had been first-time diagnosed with dementia within the last month when compared to individuals who had not been diagnosed with dementia (Erlangsen et al., Reference Erlangsen, Stenager, Conwell, Andersen, Hawton, Benros, Nordentoft and Stenager2020). Using the same comparison group, significantly lower rates were found for individuals who had been first-time diagnosed 1–2 years (aIRR, 0.7, CI 95%, 0.5–0.9) and 3–5 years (aIRR, 0.5, CI 95%, 0.4–0.7) ago. Although using different measures, these findings might be comparable and relate to the same phenomena. There are different possible explanations for these observations. Firstly, changes in cognition, which occur due to the progression of the disorder, may reduce physical abilities for carrying out a suicidal act. Thus, a lower the risk of suicide among individuals in an advanced state of dementia is likely to be the case, i.e. some years after first diagnosis (Erlangsen et al., Reference Erlangsen, Stenager, Conwell, Andersen, Hawton, Benros, Nordentoft and Stenager2020). Secondly, receiving the news that one has a severe, chronic disorder can be very distressing, and elevated risks of suicide have been demonstrated during the first months after being diagnosed with a range of physical disorders (Erlangsen et al., Reference Erlangsen, Stenager, Conwell, Andersen, Hawton, Benros, Nordentoft and Stenager2020; Petersen et al., Reference Petersen, Stenager, Mogensen and Erlangsen2020). Both findings point to the time of the diagnosis being a crucial point for intervention.
Although most people have some knowledge about dementia and the course of the disorder, they may lack insights regarding treatment and support options, which can help improve quality of life also in advanced stages of the disorder. Specific personality traits have been suggested as prevalent among those older adults who died by suicide, for instance, a wish for “being in control” of one’s life (Kjolseth et al., Reference Kjolseth, Ekeberg and Steihaug2010). Consequently, fear of losing this autonomy when becoming dependent on help from others may lead to suicidal thoughts. The anticipation of these problems, and a wish to act while one is still apt to do so, might explain the higher risk of suicidal behavior during early stages of the disorder. Also, it underscores some opportunities for intervention. Previous efforts of psychoeducation in combination with assignment of a case manager provided in primary care settings have shown promising results in terms of reducing levels of suicidal thoughts among older adults with mood disorders (Gallo et al., Reference Gallo, Morales, Bogner, Raue, Zee, Bruce and Reynolds2013; Unutzer et al., Reference Unützer, Tang, Oishi, Katon, Williams, Hunkeler, Hendrie, Lin, Levine, Grypma, Steffens, Fields and Langston2006). Community-based efforts to support individuals with dementia and their informal caregivers exist in several countries and might be intensified around the time of diagnosis.
In the study by Hedna et al. (Reference Hedna, Sigström, Johnell and Waern2024), being of older age or recipient of home care was also linked to an excess risk of suicidal behavior. These markers may, in principle, be related to the above-mentioned features of a more advanced stage of dementia, while higher risks were found for those of younger age and those who did not require home care. Previous findings have suggested that individuals who experience an onset of dementia at an unexpected age, i.e. 50–69 years versus 70 years and over, were associated with higher rates of suicide (Erlangsen et al., Reference Erlangsen, Zarit and Conwell2008).
A limitation of studies based on secondary data is that information on relevant factors, which might provide an understanding of causal links, may not be available. Based on the existing evidence, it seems that efforts to prevent adverse mental health and suicidal thoughts should best be offered to individuals in milder stages of dementia, ideally at the time of first diagnosis.
Conflict of interest
None.
Description of author(s)’ roles
The authors, Lauren L. Brown and Catherine García, equally contributed to the manuscript, revised, read, and approved the submitted version.