As of our knowledge, our study is the first one using suicide completion as the outcome to investigate the mediation effect of and depressive symptoms, hopelessness, and perceived burden on the pain-suicide association and to quantify the individual mediational effect. We find that pain was an independent risk factor of suicide completion in older adults in rural China. Furthermore, our results indicated that depressive symptoms, hopelessness, and perceived burden significantly but partially mediated the pain-suicide association.
Previous studies found that patients with pain had increased risk of suicidal outcomes compared with the general population. A large cohort study demonstrated that veterans with self-reported severe pain were more likely to die by suicide than patients experiencing moderate or less pain (HR: 1.33, 95% CI: 1.15–1.54) (30). A case-control study among older adults in Canada also found that severe pain (OR: 4.07, 95% CI: 2.51–6.59) elevated the risk of suicide death (31). However, the study only considered patients with prescribed pain medication. Consistent with previous studies, the present study found that severe pain was more common in suicide cases than in controls (37.19% vs. 11.16%). We proved that older adults with severe pain predicted 3.46 times of suicide completion risk than those without pain. Older adults with severe pain may view suicide as a way to escape from suffering (32). The risk for suicide among the older population with severe pain merits particular attention.
One major finding was that 56.29% of the total effect of pain intensity on suicide was from pain directly, and perceived burden, hopelessness, and depressive symptoms significantly mediated 43.71% of the association between pain and suicide. These results were consistent with one recent study proving that depression and anxiety disorders had mediational effects on the relationships between pain-suicidal behaviors(17). Though this study only used suicide ideation and suicide attempt as outcomes, the study used a large, representative sample of the English adult population to prove the partially mediation effect of depression. However, several studies found different results that pain- suicide ideation relationship was fully mediated by depression(18, 33). A possible reason for this discrepancy might be due to the different outcomes (suicide completion vs. suicidal ideation/attempt) and different populations across these studies.
Moreover, we measured the individual mediational effect. Depressive symptoms, hopelessness, and perceived burden explained 17.39%, 17.63%, and 8.69% of the total effect of pain intensity on suicide completion, respectively. Many studies suggested that hopelessness significantly predicted suicidal behaviors. A meta-analysis of longitudinal studies found that hopelessness significantly predicted suicidal ideation, suicide attempt, and suicide death (34). Moreover, helplessness/ hopelessness in patients with pain has been reported as a significant predictor of suicidal ideation in previous studies (9, 15, 35). Our results suggest that further investigation should focus more on hopelessness in the pain-suicide association.
Consistent with prior evidences (36), we also found the mediating effect of the perceived burden on the pain-suicide association. The interpersonal theory of suicide proposed that thwarted belongingness and perceived burdensomeness were key factors contributing to a suicide desire (14). Perceived burden has been identified as a critical predictor of suicidality in patients with pain (15, 37). It is reasonable to expect that the older adults suffering from pain felt that they created a burden to their family. The current finding provided further evidence that the need to consider perceived burdens in older adults with pain in rural areas in suicide prevention.
Previous studies have reported depression as the most common mental illness among suicide decedents(20, 38, 39). Moreover, higher pain severity was a significant predictor of the onset of depression (40). Our results in line with previous evidence. We further demonstrated that depression partially mediated the impact of pain on suicide.
Interestingly, we found that only depressive symptoms could not mediate the association between pain and suicide completion. Nor did the perceived burden or hopelessness. The relationship between pain and suicide is complex and involves several psychological and psychosocial factors(32). Moreover, those factors might interact with each other, and the interactions may affect the pain-suicidality relationship. A study in back pain patients found that pain intensity showed indirect effects due to pain-related helplessness/ hopelessness on depression (41). Depressive symptoms were found to be a significant predictor of self-perceived burdensomeness in patients with chronic pain. Furthermore, a community-based study suggested that hopelessness and the interaction between perceived burdens and hopelessness were significantly associated with suicide plans and attempts (42).
Depression, hopelessness, and perceived burden might interact with and exacerbate each other, finally increasing the vulnerability to commit suicide. Though we found the mediators and their respective indirect effects, further researches on the interactions among mediators are needed for a comprehensive understanding of pain-suicide association.
The current study has several limitations. Though psychological autopsy is a widely used method to explore the risk factors of completed suicide, the validity of data provided by proxy informants is a concern. However, one previous study had proved that subject-proxy concordance for the GDS was fair (ICC = 0.590) in the living controls(26). Second, we measured pain with a single question concerning the intensity of pain without data on the location, type, and duration. More detailed questions on pain might be better for the comprehensive understanding between pain and suicide completion. However, most research generally examines only a limited number of types or locations of pain. Moreover, a review had suggested that pain itself, irrespective of which type or location, was a determinant of suicidality based on abundant evidence (43). Furthermore, it seems no strong link between pain duration and suicidality. As the duration of suffering pain increases, suicidal behaviors may be less related to pain symptoms (43). Also, prior studies had demonstrated that sleep disturbance was a mediator in the pain-suicidality link (19). Nevertheless, we did not include information on sleep quality. Further research is warranted to examine this possibility. Finally, we estimated the pain-suicide death association and mediational effects in older adults in rural China. The results should be interpreted with caution.