Considering the emergence and maturation of biological psychiatry over the last decade, it seems unusual from the perspective of a neuroscientist to be considering neurological disease as separate from psychiatric disease, especially from the standpoint of risk factors. Indeed, psychiatric symptoms are more often than not present in neurologic patients as well as neurologic symptoms being present in psychiatric patients. However, if one considers that primary neuropathology in neurology is defined with greater certainty than is often the case in psychiatry, perhaps the distinction remains useful.

Neurological diseases encompass a wide range of disorders with symptoms spanning from the relatively subtle and stable as with essential tremor to those that are progressive, degenerative, and devastating to quality of life of both patients and their caregivers. While the relationship between primary psychiatric diseases with suicidal behavior is often relatively clear, anxiety, depression, or loss of impulse control secondary to or comorbid with neurological disease may not be as clearly understood. For example, symptoms of depression in an Alzheimer’s dementia patient may be masked by the symptoms of the primary disease, or a Parkinson’s patient may not even recognize changes in their impulse control or mood to report these to their physician (Brand et al. 2007). Even if psychiatric symptoms are recognized in a patient with neurological disease, there may be an assumption on the part of the physician that treatment of the primary disease will positively affect psychiatric symptoms which are secondary. That of course may or may not be true, depending upon the treatment and the patient.

Some special consideration within, but not limited to neurology patient populations are the degree of verbal fluency and fidelity, degree (or ability) of agreement between self-report, caregiver report, physician’s assessment, disease severity and progression, instruments used to garner information, and age. Additionally, there is a growing awareness of the bi-directional impact of disease comorbidity, that is, patients with neurological disease are at greater risk of developing major depressive disorder (MDD), (Rickards 2005, 2006) which can then lead to suicidal behavior, and likewise, patients with lifelong MDD are likely more prone to develop other diseases including neurological diseases over the course of their lives. Untreated psychiatric symptoms in patients with neurological disorders are associated with poorer prognosis and obviously poorer quality of life (Agrawal and Rickards 2011, Rickards 2005). Therefore, ensuring that symptoms are recognized and effectively treated will result in overall better outcomes for the patient.

While the scope of the present chapter cannot encompass all neurological diseases, we intended to focus on some of the more common, as well as several which not have received much prior attention. Additionally, Traumatic Brain Injury is extremely well-covered in another chapter (Nazem et al. 2008, This Volume). Given the diversity of neurological diseases, the risk factors may be weighted very differently among different sets of neurology patients.

Autism Spectrum Disorder (ASD) is characterized fundamentally as a disorder in social interaction which can range from the social awkwardness inherent in Asperger’s to the near-total isolation of a severely affected autistic patient. Prevalence estimates have been increasing over the last few decades, and the CDC now estimate that more than 1 % of the pediatric population have ASD (CDC 2012). Contributing factors to social dysfunction include impaired cognition, attention, verbal, and nonverbal communication, and a restricted behavioral repertoire (DSM 5). The neurobiological underpinnings of autism have been the focus of a great deal of research over the last 100 years, and while etiology has remained elusive, there do appear to be structural abnormalities in the brains of patients (Brambilla et al. 2003). Among the brain regions which have been most consistently described as abnormal are increased total brain volume, increased cerebellar, amygdala and caudate volume, and reduced corpus callosum volume (Verhoeven et al. 2010; Stanfield et al. 2008). Structural differences imply functional differences, and indeed application of functional imaging has demonstrated that autistic patients appear to in some cases recruit different brain regions during the processing of cognitive tasks than normal (Brambilla et al. 2003). Overall, despite what can be very marked consequences of the disorder, the neuropathology is relatively subtle and variable, which suggests in turn that connectivity among brain regions may be of primary importance rather than malfunction of any collection of brain regions. Further, it is clear that neuropathic changes do evolve over time, and how that can be reflected in behavioral changes if of particular interest. Standard of care is highly tailored to the degree of impairment and can range from custodial to no treatment.

In a sample of 350 children with autism 6–16 years of age, mothers reported depressed mood in 54 % of children with high-functioning autism and 42 % with low-functioning autism (Mayes et al. 2011; Murray et al. 2011). Anxiety disorders are also well-known to co-occur in ASD in a much higher rate than in the general population, and is considered to be a primary contributing factor to behavioral disturbances in the population (Kim et al. 2000; Muris et al. 1998).

Given that social isolation is a strong risk factor in suicidal behavior and is a primary characteristic of ADS, it would seem that ASD patients would be at an especially high risk, assuming a certain level of social awareness in this population. Social awareness will vary with disease severity, age, and level of intervention. Diagnosis is most commonly done in early childhood, and it may be useful to consider the youth population separately from those in adolescence and then in later adulthood. In a study of 102 youths aged 7–12 years of age (Storch et al. 2013), identified an 11 % incidence of suicidal ideation and behaviors, which is comparable or even lower that to age-matched peers (Carter et al. 2008). Depression and posttraumatic stress disorder were associated with suicidal thoughts and behaviors (Storch et al. 2013). In a larger (N = 791) study (Mays and Gillon 1993), evaluated a primarily high-functioning (IQ >= 80) population of autistic youths and found a 14 % overall incidence of ideation and behavior, with greatest risk factors being black or Hispanic ethnicity, lower socioeconomic status, male gender, and comorbid depression. In patients with all four risk factors, the incidence increased to as high as 70 %.

It is well known that the incidence of suicidal behavior and of depression can spike in early adulthood, but overall, there is a strong trend for increased probability of appearance with age. This is reflected in observations that the proportion of the population over age 50 have a much higher incidence of MDD and of suicidal behavior than that under age 50 (See Szanto, this volume). Whether this is true in ASD has been addressed by a number of investigators, despite special challenges inherent in diagnosing comorbidities in this population. In a longitudinal study spanning 30 years (Mouridsen et al. 2008), examined cause of death among 341 Danish ASD patients (median age 43 years) and found that their rate of mortality was twice that which would be expected versus the general population. Several subsequent studies have confirmed the mortality rate in ASD at more than 2X the overall population in the US (Pickett et al. 2006; Shavelle et al. 2001). In the Danish study, epilepsy was associated with one-third of the deaths (8 of 26), whereas clearly identified suicide accounted for two of the 26 deaths. More prominent were accidental drowning and as well as motor vehicle accidents which were seen as more related to patients’ difficulties in communications, and these causes of death were also noted in the Shavelle and Pickett studies. While difficult to conclude completed suicide rate from these studies, it remains clear that more work is needed to better understand the subpopulations that may be at highest risk for suicide.

ADHD is characterized as a persistent pattern of inattention and/or hyperactivity-impulsivity that is more frequently displayed and is more severe than is typically observed in individuals at a comparable level of development. There are three subtypes of ADHD, differentiated by (1) inattention as a primary symptom, (2) hyperactivity, or (3) mixed, although systematic study of each as a risk factor has not been conducted. While usually diagnosed in childhood, it is now recognized as a potentially lifelong condition (Wilens et al. 2002; Faraone 2005), and affects roughly 5 % of the population overall (Faraone and Biederman 2005). Response to both stimulant (e.g., d-amphetamine, methylphenidate) and non-stimulant (atomoxetine, guanfacine) medications implicates both dopaminergic and adrenergic systems in the pathophysiology of the disorder. Imaging studies have demonstrated dysregulation in processing of the prefrontal cortex, its inputs, and target areas, although the precise etiology remains elusive (Arnsten 2009).

Several studies have attempted to address whether the incidence of suicide completion or ideation is higher in the ADHD population than in the general population. Most recently (Impey and Heun 2012), have extensively reviewed the literature surrounding the association of ADHD with suicidality. Data were analyzed for completed suicides, attempters, and ideation without attempt, and concluded that despite the variability among populations addressed by various studies, there did appear to be a positive association in the vast majority of studies, but only in males. Attention deficit hyperactivity disorder (ADHD) symptoms occur more frequently in suicidal populations and subjects with ADHD are more likely to have suicidal ideation and to attempt suicide. As in previous discussions, addition of other risk factors, such as substance abuse, anxiety, and depression heighten the risk in this population.

The chronic pain population is extremely diverse, encompassing neuropathies, MS, migraine, cluster headache, osteoarthritis, fibromyalgia, and low-back pain, to name a few. A number of studies have previously reported higher rates of suicidal behavior in chronic pain populations (Breslau 1992; Magni et al. 1998; Hitchcock et al. 1994; Fishbain 1996; Smith et al. 2004). Depression is highly comorbid in some of these populations, creating ambiguity whether suicidal behavior is secondary to the psychiatric condition or a combination of both the chronic pain and depressive symptoms. Braden and Sullivan (2008) using the National Comorbidity Survey Replication sample found that any increased risk for suicidal behavior in reported chronic pain conditions (arthritic, neck, back, or ‘other chronic pain’) was largely accounted for by presence of depression or anxiety disorder, with the exception of headache and ‘other’ pain. However, in a larger sample (Scott et al. 2010), found that many chronic medical conditions, such as hypertension, heart disease, in addition to pain-related conditions such as back and neck pain, in general were associated with higher lifetime rates of suicidal ideation, even after controlling for psychiatric illnesses. Sample and methodological differences likely account for the differing conclusions of these studies, but clearly, chronic pain may present risk, which will be heightened by concomitant depression. Assessment of the relationship between management of chronic pain, pain severity, and suicidal behavior will be crucial in better understanding the risks in these patients.

Besides epilepsy, the neurological condition most clearly associated with suicidal behavior and independent of other factors such as depression is chronic, recurrent headache, such as migraine and cluster headache. Migraine is characterized by moderate to severe headache, usually accompanied by sensory sensitization such as photophobia and phonophobia. There are two main subtypes, delineated by absence or presence of preceding neurologic symptoms referred to as aura. It is clear that irregularity in cerebrovascular and cranial nerve (trigeminal and cervical) function is involved in the disorder, but there is no agreement on initiators or even the exact nature of the irregularity (Cutrer 2010). There are effective acute-phase treatments, however, including the tryptan class of serotonergic 1b/d agonists, and a variety of anticonvulsants and antidepressants have shown some efficacy in phrophylaxis (Pringsheim et al. 2010). The prevalence of migraine in the general population has been recently estimated to be about 11 % (Buse et al. 2012), and migraine is much more prevalent in in and more likely to cause disability in females (Buse et al. 2013a, b). Comorbidity of migraine with risk factors associated with suicidal behavior, such as depression and anxiety disorders, is roughly twice the overall lifetime population averages (Breslau and Davis 1992; Buse et al. 2013a, b). Overall, migraine patients do appear to have a higher risk of suicide attempts than the general population, and the risk may be amplified by comorbid psychiatric illness as well as with increased pain intensity (Breslau et al. 2012). Interestingly, from the standpoint of suicidal ideation, there were few differences among migraine patients with aura and those without. There is no data at present that indicates completed suicide attempts are higher in this population, although previous attempt is a powerful predictor of successive attempts.

Cluster headache is characterized by severe unilateral orbital, supraorbital, or temporal headache lasting between 15 min and up to several hours. Its incidence lower than migraine, recently estimated to be about 0.1–0.2 % of the general population (Fischera et al. 2008). The pathophysiology of cluster headache appears to involve activation of the trigeminal pain pathways, vasodilation of intracranial arteries during attacks, and inflammation of the cavernous sinus, although which of these correlates are primary or secondary is still poorly understood (Leone and Bussone 2009). Involvement of hypothalamic regions is suggested by activation of hypothalamic areas during active headache. Treatments include NSAIDS, calcium channel blockers, such as verapamil, mixed-mechanism anticonvulsant compounds, such as topiramate, lithium and opioids.

While migraine’s psychiatric comorbidities have been fairly well studied, perhaps due to the lower population incidence of cluster headache, this linkage appears to have received less attention (Robbins 2013). Rozen and Fishman (2012) surveyed more than 1,000 cluster headache patients and found that depression was observed to be the most frequently comorbid medical condition (nearly 25 %) and the majority (55 %) of patients exhibited suicidal ideation. Among these, 2 % have made suicide attempts. Additionally, half of cluster headache patients have exhibited some form of self-injurious behavior during an episode.

Parkinson’s disease is the second most common neurodegenerative disease, with a 1 % incidence in the adult population over 60 years of age (de Lau and Breteler 2006). The disease is characterized primarily by motor dysfunction due to loss of dopamine-containing neurons in the midbrain’s substantia nigra and hence, loss of their projections to forebrain motor regions. The loss of synaptic input of dopamine can be reflected by a loss of forebrain dopamine transporters or postsynaptic dopamine receptors. Hallmark symptoms are tremor while awake that progresses distally to proximally over the disease course, a shuffling gait, and ultimately progresses to relative immobility. While the motor symptoms are iconic in Parkinson’s, it is clear that as importantly the disease results in cognitive, behavioral, and affective symptoms which can be positively or negatively altered by the standard of care-dopamine replacement therapy, as well as by emerging therapies, such as deep brain stimulation. Etiology is multivariate and heterogeneous, and includes factors such as exposure to environmental toxins, coupled to genetic vulnerability and to nutritional history and status. Recently, impulse control disorders which are likely part of the disease and/or treatment (Dagher and Robbins 2009) have been the subject of a number of studies and are characterized by a failure to resist impulses that can lead to harmful outcomes for both themselves and those around them. Examples of these are gambling, which is the most common, but also include hypersexuality, compulsive shopping, and compulsive eating. Impulse control disorders can affect as many as 15 % of patients with Parkinson’s disease. It is interesting to consider whether impulsivity, related to the disorder or to its treatment, plays a role in suicidality in the disease. Additionally, given the pivotal role of the dopaminergic system in reward, pleasure, and in cognition, it would certainly be predicted that Parkinson’s patients would suffer from depression in much higher proportions than in the general population. This has been demonstrated by Nazem et al. (2008), who estimated that the incidence of depression may be as high as 40 % in the Parkinson’s population, accompanied by an alarmingly high level of suicidal ideation of >30 %.

Degeneration of other monoamine neurons could also play a role in depressive symptoms and suicidal behavior in light of evidence that restoration of dopaminergic neural networks by transplantation of dopamine-rich fetal mesencephalic tissue failed to alleviate depressive symptoms despite improving motor symptoms (Politis et al. 2012). Serotonergic neurons and serotonergic terminal fields did not demonstrate recovery with the transplantation and are a reasonable candidate to explain the failure to improve depressive symptoms. This notwithstanding, Parkinson’s patients have been shown to have a lower prevalence of completed suicides than age-matched controls (Myslobodsky et al. 2001; Stenager et al. 1994). Depression associated with Parkinson’s, however, appears to increase the risk of an attempt, and treatment of the depression therefore becomes quite important (Nazem et al. 2008). Recently, if Parkinson’s patients who have undergone DBS treatment are considered, a number of case studies (Burkhard et al. 2004; Soulas et al. 2008), followed by controlled trials (Voon et al. 2008) indicated that DBS patients were indeed at higher risk for suicidal behaviors including completed suicide attempt, especially in the first years following surgery. Interestingly, the increase in suicidal behavior in this population occurs despite otherwise excellent symptomatic outcome (Wolters et al. 2008). There is some retrospective evidence of both affective (Berney et al. 2002) and cognitive (Aybek et al. 2007) changes following DBS surgery, which could potentially play a role in outcome in some patients, although this remains unclear to what degree (if any) the surgery may unmask, induce, or accelerate these symptoms. There does appear to be a clear association with depression in this population, and therefore careful presurgical screening and postoperative monitoring can be useful tools in mitigation.