In this national population-based Swedish study, being married was a protective factor for non-fatal SH in men aged 75 years and above but not in women. Being foreign-born was associated with an increased risk of non-fatal SH, in particular among women. In AD users, an elevated risk of non-fatal SH was found in both men and women with higher education.
The foreign-born older adults seem to constitute a particular risk group for non-fatal SH. This expands on findings from our previous study that focused only on new users of ADs [10]. Taken together, our results point to a need to improve the monitoring of AD treatment and to reach the mental health needs of older adults with foreign background, and this pertains especially to women. Previous research conducted in younger age groups has shown that immigrant women do not seek or receive adequate mental healthcare [12], because of language limitations and cultural barriers such as the fear of stigma related to mental illness [13, 14]. The higher risk found in the older adults from other Nordic countries advocates the need for more research to better elucidate this association. Differences in drinking behaviours might in part explain the disparity.
In the total cohort, risk of non-fatal SH was elevated in those with higher education. This was somewhat unexpected, as we have previously reported an association between lower educational level and non-fatal SH [15]. That study employed a different methodology, involving a clinical cohort and a population-based comparison group. It is probable that depressed older individuals would be less likely to participate in the latter group, which may have affected results. In the current study, the association between higher education and non-fatal self-harm was particularly pronounced in AD users. Loss of social status might help to explain the risk increase, but it is unclear why we did not see an association in the non-user group.
The lower risk of non-fatal SH in the oldest AD users (80+) may be explained by the extensive use of ADs in in nursing homes and in those with dementia, which are both associated with lower non-fatal SH risk in our cohort. Frailty and severe illnesses in some nursing home residents may make them physically or mentally incapable of planning and carrying out a suicidal act.
Our finding of a protective effect of being married on non-fatal SH in older men is in accordance with previous research [16]. The increased risk observed for single and widowed men, in both AD users and non-users, suggests the need to not only focus on men’s use and adherence to prescribed AD therapies but also to recognise and meet the need of social support in older men without partners who are potentially more prone to social isolation.
Methodological considerations
The use of total population registers provides a more accurate risk estimation as selection bias is eliminated. Further, the large size of the population allowed the inclusion of a wide range of covariates in the adjusted analyses. However, we lacked information on numerous pertinent behavioural risk factors including social isolation, degree of hopelessness, problematic alcohol use, and suicidal behaviour earlier in life.
Our study design detected persons who received care in hospital and at specialised outpatient services. The public health significance of studying this group is emphasized by a finding from a British multicentre study estimating that older adult self-harmers presenting to hospital have a risk for suicide that is 67 times that of the general population [17]. However, a limitation of our study is that older adults who self-harmed but did not seek help at hospital or specialised outpatient clinics may have erroneously been considered as controls. This may have caused an underestimation of the risk ratios. The Swedish Prescribed Drug Register does not include medications prescribed in inpatient settings. As such, there may be a risk that some people classified in this study as non-users of ADs had in fact received this therapy during a hospital stay. We believe this is unlikely to have any noticeable impact on the observed outcome as in most cases, AD therapy initiated in hospital will be continued after discharge, and thus recorded in the register. Finally, we are not able to know whether patients who filled prescriptions actually consumed their medication. We note, however, that there is some evidence that older adults taking AD report that the positive aspects of treatment outweigh the negatives [18]. The purchase of AD was used as a proxy for depression in our population but we acknowledge a possibility of an indication bias or a residual confounding as AD may be prescribed in late-life for other indications than depression [19].
We lacked data on date of immigration which is a limitation since time spent in the host country may influence the degree of acculturation [20]. Our results may not be extrapolated to other regions or cultures due to differing patterns of suicidal behaviours and availability of healthcare.
Implications and conclusion
The finding of elevated non-fatal SH risk in persons born outside of Sweden both with and without AD therapies indicates a need for a better understanding of potential barriers to adequate support by mental healthcare service providers, primary care services and social services alike. Our study supports the potential value of mental health promotion strategies that encourage men to deconstruct the hegemonic ideal of the man who is emotionally and socially self-sufficient, and to extend their social emotionally supportive relationships other than their partners to achieve this [21]. Our findings are of importance to identify and monitor individuals at higher risk for suicidal behaviours, and highlight the need for more research on tailored multifaceted gender-specific preventive strategies to reduce suicidal behaviour in older adults.