Introduction

Symptomatic, adult Chiari Malformation Type I (CMI) is a neurological condition characterized by high levels of chronic pain, disability, and depression, yet self-harming and suicidal behaviors have not yet been studied in this patient group [1–3]. Clinically, CMI involves herniation of the cerebellar tonsils through the foramen magnum at the bottom of the skull resulting in direct compression of neural tissue and disruption of the natural flow of cerebrospinal fluid across the cranio-vertebral junction [4,5]. Common symptoms include headaches (acute Valsava induced and chronic migraine type), neck pain, shoulder pain, balance issues, depression, anxiety, cognitive impairment, and trouble sleeping [6–9]. For reasons that have not yet been elucidated, close to 80% of adult CMI patients are female [10]. A subset of patients (~54%) [6] undergo neurosurgery to create more room around the compressed tissue and restore the natural flow of cerebrospinal fluid [5]. However, this procedure has limited long-term outcomes, and patients are often left with residual physical, psychological, and functional issues [1,2,11]. Although it has not been previously studied, both surgical and non-surgical CMI patients could be at elevated risk of self-harming and suicidal behaviors due to high levels of chronic pain, disability, and depression [1–3,12–14].

Specifically, it has been shown that 80% of adult CMI women have moderate-complete neck related disability as measured by the Neck Disability Index (NDI) [2]. In addition, pain levels reported by CMI patients on the Short Form McGill Pain Questionnaire 2 (SF-MPQ-2) are comparable to what is found in other chronic pain conditions [3]. Interestingly, both the disability and pain findings are independent of surgical status [2,3]. Meanwhile, Meeker et al. [15] found that more than 90% of adult CMI patients struggle with one or more aspects of daily living. Psychologically, Garcia et al. [1] found that 44% of adult CMI patients suffer from moderate-severe depression and 60% from moderate-severe anxiety (Depression, Anxiety and Stress Scale – 21, DASS-21), again independent of surgical status. Finally, a large survey found that 78% of patients still employed found it difficult to work due to CMI, and of those not employed 79% said it was due to CMI [6].

Both chronic pain and depression have been well demonstrated as major risk factors for self-harmful behaviors up to and including suicide [12,13]. In addition, a nationally representative survey of nearly 200,000 subjects found that people with functional disabilities were significantly more likely to report suicidal ideation, planning, and attempts than the general population [14].

The Columbia Suicide Severity Rating Scale – Screening Version (C-SSRS Screen) is a shortened version of the widely used Columbia-Suicide Severity Rating Scale [16]. Bjureberg et al. [16] evaluated the utility of the C-SSRS Screen in predicting 7-day, one month, and one year completed suicide in more than 18,000 consecutive psychiatric emergency room patients and found the C-SSRS Screen score was significantly associated with higher odds of completed suicide at all three time points.

The purpose of this cross-sectional study was to quantify the suicide risk among a relatively large sample (N = 372) of CMI adults (65% post-surgical) using the C-SSRS Screen administered through a web-based anonymous survey, and further to examine the relationship between the level of risk with pain, disability, and psychological distress.

We hypothesized that CMI adults would exhibit high levels of suicide risk and that pain, disability, and psychological distress would be significant, independent predictors of the C-SSRS Screen scores.

Materials and methods

The research was approved by the Institutional Review Board at The University of Akron (#2022065), and all participants provided written, electronic informed consent. A web-based, anonymous survey was constructed using Qualtrics (Provo, UT: Qualtrics; 2022). The survey consisted of demographic and Chiari history questions, plus eight validated scales to assess suicide risk, pain, disability, and mental health.

Given the nature of the study, the participant debrief included instructions on how to contact the National Suicide Prevention Lifeline as well as a web link for additional resources. While the study of suicidal ideation and behaviors raises valid ethical concerns, Gould et al. found in a large, randomized control trial of high school students that screening did not increase distress levels, ideation, or attempts [17]. In a review of the literature, Dazzi et al. also found no evidence that screening increased ideation [18].

Results

The average age of the study sample was 43 years and 92.7% were female. Sixty-five percent had undergone Chiari surgery. Overall, 162 (43.5%) participants were in the High-Risk group based on the CSSRS-Screen (Table 1). The High-Risk group was 5 years younger on average than the Low-Risk group (p = .0005) but the differences in gender and Chiari surgical history were not significant. The High-Risk group scored significantly higher on every scale measure, with all but Loneliness (p = .0046) maintaining significance even after correcting for multiple observations (p < .0045).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic, functional, and psychological distress measures of adult CMI patients scoring at high (N = 162) and low (N = 210) risk on the C-SSRS screen.

https://doi.org/10.1371/journal.pone.0334599.t001

Categorizing the variables (Table 2) demonstrated that the odds of scoring at High-Risk for suicide decreased significantly with older age, with participants aged 30 years and older having less than half the odds of being High-Risk as those aged 18–29. Participants with moderate Depression were more than three times as likely to be in the High-Risk group as those who scored in the normal range (OR = 3.50, 95% CI = 1.79–6.86), while participants who scored as extremely severe in Depression were more than six times as likely to be High-Risk (OR = 6.56, 95% CI = 3.60–11.94). Severe Neck Disability was significantly associated with High-Risk status (OR = 4.71, 95% CI = 1.62–13.73), as was Pain at every quartile above the first. Participants with IES-R scores suggestive of PTSD were twice as likely to be in the High-Risk group (OR = 2.38, 95% CI = 1.56–3.64), while those meeting the cut-off for Pain Catastrophizing were three times as likely to be in the High-Risk group (OR = 3.2, 95% CI = 2.08–4.93).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Univariate odds ratio analysis for scoring at high-risk on C-SSRS screen among adults with CMI (N = 372).

https://doi.org/10.1371/journal.pone.0334599.t002

The collinearity check showed all VIF < 3, indicating collinearity among the input variables would not influence the regression models. The first regression model (Continuous) demonstrated good concordance (c = .76), with Age, Depression, and Neck Disability as the only significant predictors (p < .05) when controlling for all other variables (Table 3). The second logistic regression model (Binary) also demonstrated good concordance (c = .73), with Depression at the moderate level or above having the highest adjusted odds ratio of being in the High-Risk group (AOR = 4.27, 95% CI = 2.58–7.05). In addition, those under the age of 30 years had three times the odds of being in the High-Risk group (AOR = 3.10, 95% CI = 1.67–5.78), and those with severe or complete Neck Related Disability had twice the odds of being in the High-Risk group (AOR = 2.02, 95% CI = 1.22–3.33).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Logistic regression models with continuous and selected binary factors for C-SSRS screen high-risk score among adults with CMI (N = 371).

https://doi.org/10.1371/journal.pone.0334599.t003

Discussion

This cross-sectional study found that 44% of adults with CMI were at an elevated risk of suicide as defined by the CSSRS-Screen. In addition, being under the age of 30, having moderate or higher depression (DASS-21), or severe to complete neck related disability (NDI) significantly increased the odds of being classified as High-Risk for suicide. However, gender and history of Chiari surgery were not related to suicide risk.

Suicide risk has not been previously studied in the CMI population. However adult CMI patients exhibit several factors linked to increased risk of suicidal behavior, including high levels of depression, chronic pain, and disability [1–3,12–14]. In a review of the epidemiology of suicide, Bachmann [12] noted, “Depression is the leading cause of death by suicide worldwide…half of all completed suicides are related to depressive and other mood disorders.” In a separate review, Lepine and Briley [31] found that the mortality risk for suicide among depressed patients is more than twenty times greater than the general population. Our study found that depression (at the moderate or higher level on the DASS-21) was a strong, independent predictor of being classified in the High-Risk group.

Chronic pain is also a major risk factor for suicidal behaviors [13,32]. In a review, Tang and Crane [13] found that relative to controls the risk of death is double for chronic pain patients and the lifetime prevalence of suicidal ideation is around 20%. Calati et al. [32] found in a meta-analysis that individuals with any type of pain were significantly more likely to experience death wish, suicidal ideation, planning, attempts, and completions. They concluded that physical pain is a consistent risk factor for suicidal thoughts and behaviors. While our study showed that on average participants in the High-Risk group reported pain scores that were 23.5% higher than the Low-Risk group (p < .0001), pain was not an independent, significant predictor in the regression models.

In a secondary analysis of data collected through the 2020 Nations Survey on Drug Use and Health, Marlow et al. [14] found that people with disabilities were more than twice as likely to report suicidal ideation, planning, and attempts. Our study found similar results, with participants who reported severe to complete neck related disability being twice as likely to be in the High-Risk group. While these categories represent high levels of disability, they are common among adult CMI patients. Specifically, 31% of the study sample were classified as such, and in a separate study of 474 adult females with CMI, Labuda et al. [2] found that 44% reported severe to complete neck related disability.

The present study also found that older participants had significantly lower odds of scoring High-Risk compared to younger participants. This is counter to global, general population suicide rates where older adults (age 65+) are the most at risk age group, especially men [33]. This could be due in part to the composition of the study group, with only 4% of participants being 65 years or older and with no men in this category. A post-hoc analysis demonstrated a weak but significant negative correlation between age and both depression (r = −.13, p = .0149) and anxiety (r = −.18, p = .0007). However, this seems unlikely to provide a full explanation of this result, especially since all three variables were included in the regression modeling. Further study may be warranted to elucidate the nature of the association.

CMI patients often seek care from a number of providers during the course of their diagnosis and clinical care, including neurosurgeons, neurologists, and pain specialists. However, based on our experience, screening for depression and/or suicide risk is not routinely performed for CMI patients. The results of our survey suggest that suicidal ideation rates are unusually high in both surgically treated and untreated Chiari patients, and that the CSSRS-Screen may be a useful tool to identify those at risk. Our data further supports the importance of a study of suicide risk and its clinical and health implications in the CM1 patient population.

The present study has two main limitations. First, the cross-sectional design precludes any conclusions regarding causality. While several predictor variables showed strong statistical associations with an elevated risk score on the C-SSRS Screen, the precise nature of the associations cannot be definitively determined and requires additional research.

Second, because participants were recruited through a patient organization’s website and social media, the study sample may include selection bias which could affect the accuracy of the point estimates and potential generalizability. Specifically, it is possible that the study sample may include subjects with more severe symptoms and/or worse outcomes than the general Chiari population. The surgical rate – which can be used as a crude surrogate for disease severity in the group – was 65% in the study sample compared to 54% in a much larger sample [6]. However, previous research and this study’s results indicate that suicide risk scores, neck related disability, and depression are all comparable among surgical and non-surgical patients [1,2], suggesting that the impact of a surgically skewed bias would be minimal. While clinic-based studies of suicide risk would be valuable, the logistics of performing a comparable study to the present one would be challenging. First, not many clinical centers have the patient flow for a study this size, but more importantly it is our experience that most surgical CMI patients are only followed clinically for 12 months after surgery. Therefore, a clinic-based sample would only be looking at a small window of time in the patient experience and would not be representative of the general CMI population.

It is worth noting that the potential effects of delivering the C-SSRS screen anonymously through the web are not known. However, it should be emphasized that the Screen was designed using plain language and intended for use by anyone without any training. It seems reasonable that the anonymous, remote format may actually result in more accurate responses than in-person assessments, since the respondents are not concerned with the potential consequences of their answers, and thus more willing to disclose any ideation or previous suicide attempts. Finally, it is important to note that the study data was collected leading up to the holiday season which could have biased the C-SSRS Screen scores. However, 92.2% of the responses were collected prior to 10/31, so the magnitude of any such bias is likely minimal.

Conclusion

Forty-four percent of adults with CM1 surveyed were at an elevated risk of suicidal behavior as measured by the C-SSRS Screen, independent of surgical treatment. Age under 30 years, depression at the moderate level or worse as scored on the DASS-21, and severe to complete neck related disability significantly increased the odds of scoring at higher risk. This study is the first to examine suicidal ideation and risk in the adult CMI population.

Acknowledgments

The authors would like to acknowledge and thank Dorothy Loth, MS, of Northeastern University for performing the Qualtrics programming.