Psychiatric Pain–Associated Co-morbidities

Affective Disorders (Depression)

About 8% of the U.S. population is depressed. Of those with chronic pain, 18–35% have been reported to be depressed and 6.3% to have major depressive disorder (MDD). Here, statistical association studies have demonstrated that pain is strongly associated with depressive disorders and may be a component of major depression. Painful physical symptoms have been reported to occur in 50–66.3% of patients with MDD and are an indicator of MDD. Finally, pain severity predicts the presence of depression and MDD ( ).

In tertiary treatment facilities for chronic pain patients (CPPs), depression has been reported in 10–100% of CPPs, MDD in 1.5–54%, dysthymia in 0–43.3%, and adjustment disorder with depressed mood in 28.3% ( ). Thus, the prevalence of some forms of depression is greater in tertiary pain facilities than in community CPPs.

There are diagnostic problems with arriving at a diagnosis of MDD by Diagnostic and Statistical Manual of Mental Disorders , Fourth Edition, Text Revision (DSM-IV-TR), criteria in CPPs. A number of criteria can be and are affected by pain (e.g., sleep, fatigue) and are pain dependent. For example, when all criteria are used regardless of whether they were pain dependent, 37.7% of CPPs had MDD. When pain-dependent symptoms were discounted, 19.4% had MDD. If non-somatic are substituted for somatic symptoms within the criteria, 30.3% of CPPs had MDD ( ). Accordingly, non-somatic symptoms may be a better gauge of depression in CPPs than somatic symptoms.

A major issue in the pain literature has been whether depression causes pain, a history of depression predisposes to or interacts in the development of pain, or depression is caused by the development of pain. An evidence-based structured review of 83 studies concluded that there was greater support for the pain causing depression and depression predisposing to the development of pain with injury hypotheses than for the depression causing pain hypothesis ( ). There has also been evidence presented for the pain causing depression hypothesis from a meta-analysis of 12 studies ( ). Recent evidence also indicates that formerly depressed individuals and never depressed individuals differ in how they cope with increased pain. Thus, it is likely that a history of depression predisposes to poorer handling of pain with injury and that pain predisposes to the development of depression in susceptible individuals.

Identification of depression in CPPs is clinically important. First, depression may affect the perception of pain. Recent studies in depressed patients using various forms of pain induction to measure pain threshold or tolerance have demonstrated the following: ischemic pain induction was perceived similarly in depressed patients and controls, but thermal pain induction and paresthesia were perceived as being more painful in depressed patients; depressed patients were hypoalgesic to heat and electrical stimulation but were hyperalgesic to ischemic muscle pain; and depressed patients demonstrated decreased pain thresholds with cold pain induction. It is also interesting to note that hyperalgesia to heat pain in depressed subjects appears to normalize with duloxetine treatment. Laboratory induction of sad mood has demonstrated the following: reduction of the pain threshold in healthy subjects; decreased pain tolerance in CPPs; an effect on perception of pain unpleasantness in healthy subjects; and using functional magnetic resonance imaging (fMRI) following induction of depressed mood, brain responses to noxious thermal stimuli were characterized by increased activity in a broad network with significantly less deactivation when compared with pain responses in a neutral mood. Thus, the depressive state may make the individual more sensitive to pain perception. Second, depression is associated with or correlates with perception of disability. In addition, depression predicts disability. Third, successful treatment of depression in CPPs appears to reduce the effect of pain interfering with work, whereas pain interferes with recovery from depression and the presence of pain predicts a poor outcome with treatment of depression. Non-responders to psychopharmacological treatment of depression report more severe baseline pain, and the presence of pain predicts a longer time to remission of the depression.

The issue of pain interfering with treatment of depression then leads to another question: “If one decreases pain, will this decrease depression?” Opioid treatment of pain in depressed CPPs will improve their depression. In addition, a reduction in pain mediates the relationship between chronicity and improvement in depressive symptoms. Similarly, effective treatment of the pain symptoms of MDD with duloxetine (shown to reduce the painful physical symptoms of depression) is associated with higher MDD remission rates, and improvement in overall pain severity is associated with a high probability of remission of MDD ( ).

Overall, this research indicates the following: depression is associated with chronic pain; chronic pain can have an impact on the difficulty in diagnosing depression; in predisposed individuals, pain may be etiologically related to the onset of depression; depression can affect pain perception in some modalities, such as ischemic muscle pain; depression is associated with greater disability; treatment of depression can improve disability; pain interferes with the treatment outcome of depression; and improving pain can improve depression.

Anxiety Disorders

In CPPs in tertiary pain facilities, the prevalence of anxiety disorders is as follows: panic disorder, 11%; panic disorder with agoraphobia, 2.1%; generalized anxiety disorder, 15–20%; obsessive–compulsive disorder, 1.1%; post-traumatic stress disorder (PTSD), 7–39%; phobic disorder, 9%; social phobia, 11%; and adjustment disorder with anxious mood, 42.8% ( ). Overall, the prevalence of any current anxiety (without adjustment disorder with anxious mood) disorder is 7–28.8%. Anxiety is therefore commonly found in CPPs, but these prevalence percentages are probably an underestimate. Evidence for this statement comes from studies of patients with anxiety syndromes who have chronic pain: in this case, 40–81% of patients with panic disorder have pain. Of the anxiety disorders, generalized anxiety disorder has the strongest association with pain. Finally, there is evidence that anxiety (panic disorder, generalized anxiety disorder) has an even stronger association with pain than depression does ( ).

Questionnaires are another approach to measuring anxiety in CPPs. Here, three aspects of anxiety have been emphasized: trait anxiety, which should be very closely related to generalized anxiety disorder; anxiety sensitivity; and health anxiety. CPPs with high trait anxiety scores report significantly more pain. Health anxiety is a dimensional construct characterized by extreme concern about one’s health. CPPs with high health anxiety scores have a poor prognosis, and the experience of chronic pain can increase health anxiety. Anxiety sensitivity (AS) refers to the extent to which an individual believes that anxiety or arousal can have harmful consequences. High AS scores are predictive of pain and anxiety during pain stimulation, and these CPPs are more negatively affected by their pain experience. A recent AS meta-analysis of 41 studies demonstrated that AS was strongly associated with fearful appraisal of pain ( ).

Overall, this research indicates that anxiety disorders are common in CPPs and that special attention should be paid to identifying CPPs with generalized anxiety disorder.

Post-Traumatic Stress Disorder

PTSD is an anxiety syndrome following exposure to an extreme traumatic stressor to one’s self or another person. In community studies the lifetime prevalence of PTSD is 8%, but in CPPs the prevalence has been reported to be 7–39%, and thus it appears to be greater than in the general population.

Pain following the aftermath of a traumatic event is a risk factor for PTSD. There is some evidence of mutual maintenance of pain and PTSD. This interaction has a negative impact on the course of either disorder and thus makes each disorder more difficult to treat ( ). PTSD symptoms have a direct influence on the severity of depressive symptoms in CPPs, with PTSD symptoms in CPPs being associated with increased pain, affective distress, and physical impairment/psychological distress. The relationship between pain and PTSD could be mediated by depression. PTSD significantly contributes to the sleep disorder found in CPPs. Finally, panic attacks are commonly associated with PTSD, but such attacks are more commonly associated with unexplained chronic pain.

Overall, this research indicates that CPPs involved in traumatic events should be evaluated for the presence of PTSD and asked whether they experienced pain after the event. In addition, further inquiry should be made into the relationship between some PTSD symptoms, such as flashbacks and pain. CPPs with PTSD should be evaluated for depression, panic, and sleep disorder.

Substance Dependence

A major clinical concern has been which CPPs fulfill this diagnosis—or have addiction—and what is the prevalence of this diagnosis in CPPs. Addiction is a chronic neurobiological disease with genetic, psychosocial, and environmental factors influencing its development and manifestations. It is characterized by impaired control over use, compulsive use, craving, and continued use despite physical, psychological, or social harm. Addiction may occur independent of tolerance or physical dependence (e.g., cocaine use), and physical dependence and/or tolerance may occur without addiction (e.g., most pain patients taking opioids). Addiction is incorporated within the DSM-IV-TR diagnosis of substance dependence. Unfortunately, a number of criteria within this diagnosis, such as “physical dependence,” “tolerance,” and “substance is often taken in larger amounts than intended,” are often seen in CPPs who do not have addiction. Consequently, when these criteria are applied to CPPs taking opioids, they are overly inclusive, thereby confounding an accurate diagnosis of addiction. The diagnosis of addiction can be made only prospectively over time while observing for a pattern of compulsive and impaired control over medication use, craving in the absence of out-of-control pain, and use despite harm. Most important, such behavior does not resolve easily when other possible causes of the behavior are appropriately addressed. However, so-called aberrant drug-related behavior (ADRB) is common in CPPs. ADRB consists of a group of problematic types of behavior surrounding opioid use (e.g., injecting oral formulations, which are red flags for addiction) ( ).

The prevalence of addiction in CPPs is subject to controversy. A much more relevant issue, however, is the incidence of de novo addiction on opioid exposure. In a review of 67 studies on de novo addiction on opioid exposure, the incidence of addiction was calculated for the total number of CPPs exposed ( ). The average incidence of addiction reported in this review was 3.27%; the average incidence of ADRB reported was 11.5% and 20.4% when urine toxicological studies were performed. For studies that preselected patients for no addiction history, the incidence of addiction decreased to 0.19% and that of ADRB to 0.59%. Thus, de novo opioid exposure may result in addiction in a small percentage of CPPs, but a much greater number of CPPs will demonstrate ADRB. The incidence of addiction and ADRB on de novo opioid exposure can be dramatically decreased by preselecting the exposed group for no history of addiction. Some recent studies ( ) not included in this review have indicated that the prevalence of current opioid dependence may be as high as 26%.

Overall, this literature indicates that this diagnosis, if applied to CPPs, will be overly inclusive and that the incidence of de novo opioid addiction can be decreased by careful selection of CPPs.

Smoking/Nicotine Dependence

Smoking is associated with a higher incidence and prevalence of non-specific low back pain (LBP). This may be related to more severe disc degeneration in smokers as a result of increased serum proteolytic activity caused by nicotine, which speeds up disc degeneration. Smokers have a higher prevalence of musculoskeletal disorders than do non-smokers reporting greater pain. Higher levels of smoking are also associated with greater levels of LBP disability and poor treatment outcome. The prevalence of smoking in CPPs has been reported to be 14.7–36.7% and may be greater than in the general population. Interestingly, nicotine has an antinociceptive (analgesic) effect, which could be the motivation for smoking behavior.

In psychiatric patients, smoking is associated with alcoholism, illicit drug use, depression, panic disorder, severe mental illness, cognitive decline, suicide attempts, suicidal thoughts, suicide completions, and decreased probability of recovery from suicidal ideation. The association of depression/suicide with smoking may be related to impairment of serotonin function by nicotine.

Smoking CPPs are more likely than non-smoking CPPs to use narcotics and to demonstrate a greater frequency of ADRB when taking opioids. Smoking also predicts long-term opioid use ( ) and is associated with greater opioid use in CPPs. Smoking chronic LBP patients are at greater risk for suicidal ideation than non-smoking chronic LBP patients are, and the risk is increased if the CPP is a heavy smoker and has a problem with alcohol ( ).

Overall, this literature indicates that smoking CPPs are a very distinct group that should be evaluated carefully for significant pain, psychiatric co-morbidity, addiction, and suicidality. These CPPs should be offered smoking cessation treatment.

Pain Disorder

In the DSM-III, pain disorder was called psychogenic pain. Here, the diagnosis revolved around two problematic criteria: the pain was inconsistent with its anatomic distribution, lacked organic pathology accounting for it, or was grossly in excess of that expected from the physical findings, and psychological factors had to be etiologically related to the development of pain. These criteria presented difficulty because few CPPs lack demonstrable organic pathology ( ) and it was unclear whether the pain demonstrated exceeded that expected from the physical findings. Therefore, the psychiatric examiner had to make a difficult judgment and then, in addition, had to judge the importance of psychological factors.

The DSM-III-R dropped the psychological criteria and renamed this disorder somatoform pain disorder, but the criteria remained similar. The DSM-IV-TR somatoform disorder work group found significant problems with the DSM-III-R criteria. For example, in one study, 91% of CPPs without objective physical findings responded to injections of intrathecal fentanyl as though the pain involved true nociception ( ). In another study, in a meta-analytic review of antidepressant treatment of pain in patients with somatoform pain disorder and psychogenic pain disorder reported that antidepressants were effective for the pain of somatoform pain disorder. The DSM-IV-TR work group then revised the criteria and named this condition pain disorder. This diagnosis now contains three criteria: (1) the pain must involve one or more anatomical sites and be the predominate focus of the clinical findings; (2) the pain must cause clinically significant distress and impairment in social, occupational, or other important areas of function; and (3) psychological factors must play an important role in the onset, severity, exacerbation, and maintenance of pain. There are, however, significant problems with these criteria: (1) the vast majority of CPPs have pain at more than one site; (2) most CPPs are significantly distressed and impaired in social, occupational, or other important areas of function; and (3) the last criterion resembles that used in the DSM-III and later eliminated in the DSM-III-R and requires a value judgment by the psychiatric examiner. These criteria should then also be of little diagnostic validity and should be overinclusive.

Recent evidence using fMRI also casts doubt on the validity of this diagnosis and its criteria. compared CPPs with a DSM-IV-TR diagnosis of pain disorder with non-pain controls in terms of cerebral activation on heat pain stimulation under fMRI observation. In comparison to controls, CPPs with pain disorder were found to have a hypoactive state of the ventromedial prefrontal/orbitofrontal cortex and a hyperactive state of the parahippocampal gyrus, amygdala, and anterior insula. This indicates that CPPs with pain disorder have a central pain-processing disorder. Such a finding casts doubts on the behavioral theory behind psychogenic pain/somatoform pain disorder/pain disorder.

Overall, this research indicates that the diagnosis of pain disorder has little validity in separating CPPs with explainable pain from those with unexplainable pain whose pain fits the conceptual paradigm behind this diagnosis. Accordingly, this diagnosis should be used with extreme caution until further research is available.

Conversion Disorder

Conversion disorder is characterized by the presence of a symptom or deficit affecting voluntary motor or sensory function that suggests a neurological or general medical condition. The two criteria for this disorder are the presence of the symptom and judgment by the psychiatric examiner that psychological factors are associated with initiation (preceded by conflict or other stressors) or exacerbation of the symptom. As noted later in the section on Waddell’s signs, non-dermatomal sensory abnormalities (NDSAs) and muscle weakness (paresis) are frequently found in CPPs and make the CPP a candidate for a diagnosis of conversion disorder. For example, reported that 37.8% of their CPPs had NDSAs. Motor deficits (paresis) are less frequent but are seen regularly.

There is now significant evidence that NDSAs are a neurophysiological central phenomenon. In rodent models of mononeuropathy, stocking or glove hypoesthesia is the result of extraterritorial sensory loss secondary to overlapping receptive fields of adjacent neurons in the spinal cord. Recent positron emission tomography scanning studies ( ) indicate that CPPs with NDSAs have decreases in cerebral blood flow in the thalamus and basal ganglia contralateral to the deficit and in cortical and subcortical areas. These cerebral blood flow problems resolve after recovery from the pain. Pain induction studies using fMRI have demonstrated that CPPs with NDSAs fail to activate brain areas normally responding to pain stimuli ( ). In addition, pain treatment studies have demonstrated that reduction of pain removes the NDSAs implicating pain in the phenomenon. Finally, NDSAs have been shown, in a statistically significant manner, to occur only on the painful site or limb ( ).

Pain has also been implicated in the paresis conversion problem. Motor conversion is almost always associated with the presence of pain in the affected extremity. If a painful limb is compared with a non-painful one, a 20–30% reduction in strength is observed. Muscle pain reduces maximal voluntary contraction and endurance time during submaximal contractions. In pain induction studies, motor unit firing rates are inversely related to reports of pain in rapidly activated painful and pain-free synergistic muscle.

Overall, this research indicates that the neurophysiological phenomena of pain are intimately involved in the generation of these two “conversion” symptoms and are not caused by behavioral issues as the diagnosis implies. Consequently and because the second criterion is essentially a value judgment, it may be inappropriate to apply this diagnosis to any CPP with these conversion symptoms, especially since there are medicolegal consequences in wrongfully applying this diagnosis ( ).

Sleep Disorders

Clinically, the vast majority (50–88.9%) of CPPs complain of sleep problems. Typically, pain does not allow them to go to sleep, and when they do go to sleep, the pain wakes them up. This raises the question of whether the sleep problem is etiologically related to the pain. A number of reviewers have concluded that there is an association between chronic pain and sleep problems. Polysomnography studies have demonstrated the following: (1) acute pain following various surgical procedures is associated with reduced and fragmented sleep and diminished amounts of slow wave and rapid eye movement (REM) sleep; (2) in CPPs, the specific nature of the sleep disorder is not clear but is characterized by reduced sleep efficiency and altered sleep architecture (increased wakefulness and stage 1 non-REM sleep); and (3) subjects subjected to noxious stimuli during all stages of sleep will demonstrate a sleep-awakening response and a moderate level of cortical arousal.

Within the DSM-IV-TR, sleep disorder related to a general medical condition involves “a prominent complaint of sleep disturbance that results from the direct physiological effect of a general medical condition on the sleep-wake system.” Pain from musculoskeletal disease (e.g., fibromyalgia) is one of the conditions that can cause this type of sleep disorder. recently addressed the CPP sleep disorder etiology question in a structured evidence-based review of 41 studies. Studies addressing primary sleep disorders, sleep disorder related to a mental disorder, and substance-induced sleep disorders were excluded. The remaining studies were subdivided by the type of analysis performed (prospective, path analysis, correlation, univariate, do non-sedating drugs that improve pain improve sleep). In all these groups, 100% of the studies supported the hypothesis of an etiological relationship between pain and sleep (pain causing the sleep disorder). However, a number of studies have demonstrated that sleep restriction in healthy pain-free normal sleepers will cause hyperalgesia on pain stimulation after sleep restriction. Thus, the results of this review did not preclude the pain–sleep relationship being bidirectional.

There are two other major issues in reference to chronic pain and sleep. Presently, a significant number of CPPs are treated with chronic opioid algesic therapy. CPPs taking opioids have a high prevalence of sleep apnea, but it is unclear whether this is etiologically related to opioids ( ). Second, nocturnal sleep disturbance is a predictor of suicide attempts in psychiatric outpatients. In addition, group membership in CPPs with suicidal ideation is predicted by sleep-onset insomnia and pain intensity.

Overall, this research indicates the following: CPPs should be routinely evaluated for sleep problems, those taking opioids should be evaluated for sleep apnea, CPPs with a sleep disorder and significant pain should be evaluated for suicidality, and any sleep disorders identified should be treated aggressively.