Suicide is an important public health problem [1–3] and a priority for the World Health Organization (WHO) [1]. Approximately 703,000 deaths by suicide occur each year, and in 2019 they represented 1.3% of global deaths [1]. Moreover, suicide has great importance in young populations because, in people aged 15–29 years, it constitutes the fourth leading cause of death globally and the second in Spain [1, 2]. Additionally, suicide is preventable [1] and the WHO recommends that suicide prevention strategies should be adapted to each country, by identifying vulnerable groups in each specific context [4]. In this regard, it has been suggested that having a disability could be a potential risk factor for suicide [5].

The concept of disability includes “impairments, activity limitations, and participation restrictions” [6] arising from the interaction between a health condition and the contextual factors of the individual [6], and it is considered to have a prevalence of 11.8% in higher income countries and 18.0% in lower income countries [6]. People with disabilities have an increased risk of depression [6, 7], anxiety [8], sleep problems [8], worse socioeconomic status [6], discrimination [6], isolation/solitude [8], and feelings of burdensomeness [9]. Some of these conditions may be related to an increased risk of suicide [4] and differ by sociodemographic groups [6].

A number of publications have specifically studied the relation between experiencing a disability and suicide [7, 9–16]. However, some of these were not representative of the general population and/or did not control for important confounders [13, 14, 16]. A limited number of population-based follow-up studies, have analyzed the likelihood of suicide in people with global assessments of disability, by sex and age groups [7, 10].

We previously studied the relation between disability and all-cause mortality in a representative sample of the adult population in Spain, finding a positive association [17]. In that work, we also showed a positive effect of disability on suicide mortality, among a group of 89 other causes of death. However, we consider it important to carry out a specific analysis that goes deeper into this relation, especially focusing on possible differences by sex and age groups, since the rates appear to be very different. In the present work, carried out in the same sample, the analyses will include effects measures that take into account the competing risks and additional co-variables. In addition, standardizations will be computed using the distribution of co-variables of the disabled population, which will provide a more approximate causal interpretation.

Thus, the objective of this research is to assess, in a representative sample of the adult population in Spain, the association of disability with suicide mortality, separately for women and men, globally and for the different groups of age and other sociodemographic factors.

In the community-dwelling adult population of Spain, the prevalence of disability (95% CI) was 12.9% (12.6% to 13.1%) in women and 9.0% (8.7% to 9.2%) in men. Women with disability were older, had lower educational level and household income, and were more likely to live alone in small municipalities in the central and southern regions of Spain than women without disability. Similar but less pronounced sociodemographic differences were observed between men with and without disability (Table 1).

During 800,754 person-years of follow-up, 29 women died from suicide and 7,932 from other causes, corresponding to mortality rates of 3.6 and 952.0 deaths per 100,000 person-years. Among men, there were 97 deaths from suicide and 9,020 from other causes during 735,709 person-years of follow-up, with higher mortality rates of 11.6 and 1118.0 deaths per 100,000 person-years (Table 2 ). The unstandardized cumulative suicide mortality risks per 100,000, comparing disabled with non-disabled women and men, are provided in Table 2. After standardizing to the weighted distribution of baseline sociodemographic characteristics in the community-dwelling disabled population and accounting for competing deaths from other causes, suicide mortality was consistently higher at any follow-up time among disabled women and men in comparison with non-disabled (Figure 1). Compared with adults without a baseline disability, the standardized cumulative suicide mortality (95% CI) at 5 and 10 years of follow-up was 54.4 deaths (−17.2 to 126.1) and 54.4 deaths (−22.7 to 131.6) per 100,000 disabled women; and 122.2 deaths (4.1 to 240.3) and 83.1 deaths (−40.7 to 206.8) per 100,000 disabled men. Similarly, the standardized suicide mortality risk ratios (95% CIs) at 5 and 10 years of follow-up were 3.39 (0.61 to 18.66) and 2.53 (0.72–8.93) comparing disabled with non-disabled women, and 2.54 (1.22 to 5.27) and 1.57 (0.86 to 2.87) comparing disabled with non-disabled men (Table 2 ).

In subgroup analyses, the excess risk of suicide death associated with disability was larger among women younger than 65 years at baseline (P for homogeneity = 0.11) and those living in more populated areas (P = 0.12). The standardized 5-year suicide mortality risk (95% CI) was 100.0 deaths (−27.4 to 227.4) per 100,000 disabled women aged 18–64 years, and 73.5 deaths (−13.2 to 160.2) per 100,000 disabled women residing in municipalities larger than 20,000 inhabitants compared to women with no disability (Figure 2). In addition, disability was associated with more marked outcomes in suicide risk among less educated men (P for homogeneity = 0.10) and those living in the East and South regions (P = 0.06). The standardized 5-year suicide risk (95% CI) was 85.9 deaths (−28.4 to 200.2) per 100,000 disabled men with primary or lower educational level and 125.0 deaths (−10.4 to 260.4) per 100,000 disabled men residing in Eastern, Southern Spain, or the Canary Islands (Figure 2).

In this large, population-based follow-up study, the cumulative suicide mortality for both sexes was higher at any follow-up time among disabled people, compared to non-disabled people. Nevertheless, it remained relatively stable after the fifth year of follow-up. These results are consistent with studies reporting that, in persons with disability, time to death by suicide is skewed towards early points [7]. Additionally, the impact of disability appeared to differ by sex and age group. In women, the increased risk of suicide mortality was only noteworthy in those younger than 65. In men, the estimates were similar for both age groups. However, these results should be viewed with caution due to the imprecision of estimates. Further studies are needed to confirm these patterns.

A limited number of studies have assessed the association between disability and suicide mortality in nationally representative samples and, in general, our results are consistent with their findings. To the best of our knowledge, the works of Park et al. [10], and Lee et al. [7] are the only studies that analyzed the hazard ratio (HR) of suicide in people with any of the different types of disability, by sex and age groups in population-based follow-ups. Park et al. in an 11-year population-based follow-up study, found that men with disabilities had an adjusted risk of suicide 1.60 times higher than men without disabilities [10], which is roughly similar to our results. It is interesting to highlight that, in their study, this association was smaller for women (adjusted risk of suicide 1.26) [10]. Additionally, they found that the association between disability and risk of suicide was substantial only for those younger than 60 years old and increased when decreasing the age group [10] which is comparable to the results for severe disability found in other studies [11]. On the other hand, Lee et al., in a 10-year population-based follow-up study, reported that the adjusted HR for people with disability was 1.9 times higher compared to those without disability [7]. Moreover, in contrast to Park et al., the HR were very similar for both sexes [7]. They also found a decreased risk when increasing the age group, and suggested that the age at which the disability occurs is a more important risk factor for suicide than the current age [7]. Also, Onyeka et al. [11] in a 4-year population-based follow-up study, showed that adults with a disability were over 1.5 times more likely to die by suicide compared to those with no limitations, after adjustment for poor mental health. Moreover, and consistent with Park et al. [10] the relationship between severe disability and death by suicide was notable only for those under 60 years old [10, 11]. Along the same lines, Turvey et al., in a longitudinal cohort study of people 65 years and older, did not find increased odds ratios of suicide in those with functional impairment; however, due to small numbers, they did not provide estimates by sex [15]. It is relevant to highlight that neither of these studies analyzed this association for the different age groups by sex [7, 10, 11, 15], and our results suggest that the lower suicide risk for the older age group [7, 10, 15] could be attributable, at least partially, to the weight of estimates for women. Cao et al. [13] found that older people from rural China with a severe disability were 1.4 more likely to die by suicide, compared to those without disability. Kaplan et al. also found that people reporting a functional limitation (through an ad hoc question that did not included disability duration) had roughly 3 times higher suicide mortality risk [12] and, finally, Kim et al. [14] found that people with disability had 2.5 times higher mortality rates by suicide than the general population, which is consistent with our results for men.

Numerous potential underlying pathophysiological mechanisms for the relation between disability and suicide have been described. Most are related to psychiatric and sociodemographic factors that are more likely in people with disability (depression [5, 7, 8], worse socioeconomic conditions [6], discrimination [6], isolation/solitude [8], and feelings of burdensomeness [9]) and could be associated with an increased risk of suicide [4]. It is interesting to highlight, that the Interpersonal Theory of Suicide, hypothesizes that suicide ideation is fomented by the integration of perceived burdensomeness and thwarted belongingness (feeling not to be part of a social group) [27, 28]. Moreover, it is considered that the relationship between perceived burdensomeness, thwarted belongingness, and suicidal ideation is mediated by lack of meaning in life that underlies the concept of demoralization [28].

Thus, it is interesting to consider the information about the subjective experiences of these patients in relation to their disability. Constanza et al. in a quantitative-qualitative observational mixed method study analyzed information from validated quantitative questionnaires and qualitative open-ended questions of seventy participants at the Multidisciplinary Pain Center of the Geneva University Hospitals. They assessed that in those patients, social interactions and activities were related to meaning in life, and pain could impact on meaningful relationships and activities, the ability to enjoy and feeling pleasure, lack or loss of vital objectives and meaning, fear of the future, mood disturbances, a sense of loss and despair that induced deep moral suffering. More specifically, the patients refer that the disability puts them in the situation of reviewing life goals and expectations, that has an important impact in their meaning in life [29]. Thus, the disability undermines meaning in life, directly or by impairing its two domains (social interactions and activities) in a vicious circle related to suicidal ideation [29].

In any case, the analysis of disability associated with suicide mortality is complex, and, as already suggested, suicide mortality risk in people with disability may be disproportionately distributed in relation to the different types of disability [7, 10], severities of disability [10, 11], and sociodemographic groups [7, 10, 11]. Furthermore, some potentially disabling conditions may lead to psychiatric symptoms [30] and there appears to be a bidirectional relation between psychiatric symptoms and function [31]. On the other hand, it is very unlikely that diseases that may lead to disability (other than depression and psychiatric disorders) would increase suicide mortality risk directly, and not through mechanisms related to disability and/or psychiatric disorders and, thus, it is not very probable that they would act as confounding factors.

The analyses performed in the present study suggest some interesting patterns. Our estimates are relatively invariant across age groups for men; however, for women, they are only remarkable for those <65 years. Thus, it appears that disability-related suicide for older women is different than for younger women and/or men. As a possible explanation, it has been suggested that older people may anticipate activity limitations related to age and better manage their role expectations [6, 11] while younger people may experience such situations as a heavier burden not being able to perform the activities expected of a person their age [11]. According to our results, this possible explanation would apply mainly to women. Qualitative studies could be of interest to assess the subjective experience of these persons that may provide additional information to explain this pattern.