Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

Diseases and/or disorders of the pulmonary system may affect brain behavior relationships due to impaired oxygen delivery (i.e., anoxia, hypoxia). Neurons are dependent on oxygen and without oxygen cellular function is disrupted and damage to cell structure leading to neuronal death may follow. A variety of respiratory disorders such as cardiac or respiratory arrest, carbon monoxide poisoning, obstructive sleep apnea, chronic obstructive pulmonary disease, and acute respiratory distress syndrome result in anoxia or hypoxia which can result in anoxic brain injury. The neuronal injury is manifest structurally by lesions and neuronal atrophy and functionally as neurocognitive and neuropsychiatric impairments.

The incidence of cardiac arrest with anoxia and cerebral ischemia occurs in more than 400,000 cases per year, of which more than 80% of these patients are likely to have poor neurological outcomes [1, 2]. Improvements in emergency and critical care medicine have resulted in approximately 200,000 cardiac resuscitations per year of which over 70,000 patients survive but constitute only 1% of those admitted to brain injury rehabilitation centers [3]. Other respiratory disorders associated with anoxia or hypoxia may also cause anoxic brain injury. The severity of anoxia/hypoxia does not appear to be related to development of neuropsychological impairments. However, the degree of neuropsychological impairment appears parallel to the degree of morphologic abnormalities as demonstrated by quantitative MRI image analysis [4, 5]. Neuropsychological deficits are common in respiratory disorders with concomitant hypoxia including impaired memory [68], executive function [9, 10], apperceptive agnosia [11], visual–spatial deficits [12], and generalized neurocognitive decline [6, 13, 14].

Effects of Hypoxia

The human brain constitutes approximately 2% of the total body mass but utilizes 20% of the total oxygen consumption [15]. The brain requires oxygen to produce energy and uses aerobic glucose oxidation to produce 95% of the brain’s adenosine triphosphate (ATP). ATP serves as a source of energy for many metabolic processes including neural function. ATP releases energy when it is broken down into ADP by hydrolysis during cell metabolism. Neocortical and subcortical functions depend upon continuous supply of oxygen, as neurons are not able to store oxygen and glucose for later use [16]. Hypoxia or anoxia damages multiple organ systems especially those with high oxygen utilization such as the central nervous system. Oxygen and glucose are required to maintain the function of the central nervous system (CNS) and reduction or depletion of oxygen and glucose results in neuronal injury. Slight decreases in oxygen delivery may cause permanent biochemical and morphological changes. Anoxia is defined as absence of oxygen in arterial blood or tissues, hypoxia as tissue oxygen deprivation, and hypoxemia as reduced oxygenation of the blood [17, 18]. Hypoxia and anoxia are often used interchangeably as anoxia is severe hypoxia.

Regional brain oxygen utilization is not homogeneous with some brain regions more vulnerable to the effects of anoxia/hypoxia, particularly structures at the end of the vascular supply or with high metabolic rates [19]. Selective vulnerability of some brain regions has been attributed to vascular or hemodynamic specificity [19], increased regional metabolism of glucose [20], and/or proximity to structures with high levels of excitatory amino acids such as glutamate [21, 22]. Vulnerable brain regions include the neocortex, hippocampus, basal ganglia, cerebellar Purkinje cells, primary visual cortex, frontal regions, and thalamus [2325].

Mechanisms of Brain Injury

Anoxia or ischemia causes a pathophysiological cascade that leads to neuronal damage and death (for reviews of the mechanisms see [17, 26]). Mechanisms of anoxic-induced neuronal injury include the following: (1) Decreased ATP production without decreasing ATP utilization, resulting in energy depletion, ionic pump failure, K+ outflow, and inflow of Ca2+ [27]; (2) lactic acidosis due to anaerobic metabolism [28]; (3) excitotoxic damage due to excessive glutamate release leading to increased neuronal firing, calcium influx, and neuronal death [26]; (4) increased calcium influx and intracellular accumulation of calcium due to ionic pump failure [29]; (5) the formation of oxygen radicals during reperfusion or reoxygenation [17]; (6) nitric oxide synthase leading to impaired neurotransmission, protein synthesis, and membrane peroxidation [17]; and (7) anoxia or ischemia resulting in neuronal necrosis and/or apoptosis or programmed cell death [30, 31]. Controversy exists in the literature regarding whether hypoxia in the absence of ischemia can result in brain injury [32]. Neuropsychological sequelae following hypoxia without ischemia occurred in 22 patients with hypoxia without hypotension, all were comatose and recovery to the premorbid level of function occurred in only 50% of the patients [33]. In fact three patients with hypoxia (PO2 less than 45 mmHg) without hypotension died of cardiac failure, indicating that factors other than ischemia contributed to poor outcome [34]. Further, neuropsychological impairments are common in patients with pulmonary disorders in which continuous or intermittent hypoxia or hypoxemia occur without ischemia. For example, patients with pulmonary disorders including chronic obstructive pulmonary disease (COPD) and obstructive sleep apnea (OSA) with concomitant hypoxia have neuropsychological deficits similar to patients with anoxia due to cardiac or respiratory arrest.

Neuroimaging Findings

As stated previously, some brain regions are more vulnerable to the effects of anoxia/ischemia, particularly structures at the end of the vascular supply, with high metabolic rates [19], and/or proximity to structures that contain excitatory amino acids such as glutamate [21, 22]. Anoxic brain injury results in focal and diffuse neuropathologic lesions and atrophy [7, 3537] including lesions in the hippocampus [38, 39], basal ganglia, cerebellum [40], subcortical and periventricular white matter lesions [41], and atrophy of the corpus callosum [42]. Generalized brain volume loss leading to ventricular enlargement and sulcal widening [36] and hippocampal atrophy are also common [7, 43]. A review of anoxic brain injury (N = 90) found that 44% of individuals had cortical edema or atrophy, 33% had cerebellar lesions, 22% had basal ganglia lesions, 21% had hippocampal atrophy, and 3% had thalamic lesions [36]. Hippocampal damage, including lesions and atrophy [38, 39], has long been established as a common consequence of anoxia. Hippocampal atrophy can be identified on magnetic resonance scans as volume reduction. Previous research has suggested that the hippocampus may be more vulnerable to hypoxic injury than adjacent medial temporal lobe structures such as the parahippocampal gyrus or temporal lobes [44].

Neurological and Neuropsychological Sequelae

Poor neurological outcomes after brain injury include death, coma, vegetative state, severe neurologic disability [45], neurocognitive sequelae, and development of new psychiatric disorders [35, 36]. Neuropsychological deficits following anoxia brain injury are heterogeneous and include agnosia [11], impaired memory [8, 39, 46], executive dysfunction [9, 10], impaired visual–spatial skills [12], generalized neurocognitive impairments [14], and motor disturbances [47]. Psychological and behavioral changes include euphoria, irritability, emotional volatility, depression, and anxiety [48, 49]. This chapter will review some common respiratory disorders and associated neurocognitive and neuropsychiatric sequelae.

Chronic Obstructive Pulmonary Disease

Chronic obstructive pulmonary disease (COPD) refers to a group of pulmonary diseases with airflow limitation that is not fully reversible including chronic bronchitis and emphysema, which occur without an asthmatic component. The airflow limitation is usually progressive and associated with an abnormal inflammatory response of the lungs [50]. Both chronic bronchitis and emphysema are characterized by airway obstruction that may be partially reversible. COPD is the fourth leading cause of death in the United States and leads to serious, long-term disability [51]. The estimated prevalence of COPD in the United States in the adult population is 5–10% [52]. Chronic obstructive pulmonary disease (COPD) is a growing cause of morbidity and mortality worldwide and the prevalence of stage II (forced expiratory volume of 30 to <50% predicted) or higher COPD was 10.1% overall, 11.8% for men, and 8.5% for women [53]. More than 14 million people are currently diagnosed with COPD and an additional 12 million likely have COPD that has not been diagnosed [54]. COPD is a progressive, degenerative disease process that results in airflow obstruction, air trapping, hyperinflation of the lungs, and impaired gas exchange. The most important risk factor for development of COPD is cigarette smoking [55]. Other risk factors include family history of pulmonary disease, exposure to allergies and/or irritants, and pulmonary infection [56]. Symptoms include shortness of breath, dyspnea, cough, increased sputum production, and wheezing. As the disease process advances, the pulmonary changes lead to abnormal sleep structure, sleeplessness, poor physical and neurocognitive function, poor exercise tolerance, lack of appetite, weight change, fatigue, and dyspnea [57]. Dyspnea or air hunger, manifest as difficult or labored breathing [58], can lead to hypoxia/hypoxemia, which, as noted above, is linked to brain injury and development of neurocognitive impairments [59], depression, and anxiety [56].

Neurocognitive Morbidity

Neurocognitive impairments are common in patients with chronic obstructive pulmonary disease (COPD) [60]. The pattern, extent, and severity of neurocognitive impairments in COPD patients are variable, but are associated with hypoxemia [61, 62]. Patients with COPD have impaired memory, executive function [63], flexible thinking [61], attention, and slow mental processing speed [64]. In general, neurocognitive impairments correlate with the duration and severity of the hypoxia [62, 65]. However, even patients with mild hypoxemia have neurocognitive impairments in a variety of cognitive domains [61, 62]. The severity of the neurocognitive impairments is also associated with older age and duration of COPD [66]. Further, older age, poorer aerobic fitness, and reduced pulmonary function predict worse neurocognitive performance [67]. Frequently reported neurocognitive impairments include executive dysfunction; reduced perceptual motor speed, impaired memory, and attention; and reduced intellectual function [68]. Investigations to date find great inter-individual variability in neurocognitive impairments in COPD patients. Memory is one of the most commonly affected neurocognitive domains in COPD. For example, impaired verbal memory is associated with poor adherence to the patients’ medication regimen [69].

As noted above, neurocognitive impairments in COPD patients correlate with hypoxia/hypoxemia [61, 62]. Moderate to severe hypoxemia deficits lead to poor motor skills, abstract reasoning, attention learning, and memory and language skills [70]. Prigatano et al. [60] found that COPD patients with mild hypoxemia had mild impairments in “higher cerebral problem-solving skills.” Other studies find impaired memory, problems forming new concepts, problems with flexible thinking [61], impaired attention, and slow mental processing speed [64] correlate with the severity of hypoxia [62, 65]. In fact COPD patients with mild hypoxia, the majority of who were treated with supplemental oxygen, had neurocognitive impairments [63]. Several studies suggest that long-term oxygen treatment improves cognitive functioning [62, 65]. Alternatively, acute oxygen treatment may not improve cognitive functioning in some COPD patients [71]. One study found that COPD patients had significantly worse intellectual function and attention than controls and the neurocognitive impairments did not correlate with disease severity (i.e., lung function, blood gas analysis, nocturnal oxygen saturation) [72]. Current data suggest the ability to predict neurocognitive impairments on the basis of the severity of the disease or hypoxemia is poor and the effectiveness of oxygen therapy for improving cognitive functioning of COPD patients is unclear.

Potential reasons for the lack of relationship between hypoxemia and neurocognitive impairments may due to non-compliance with oxygen therapy or other factors such as variable duration of illness and illness severity (such as variable duration and severity of hypoxemia), comorbid disorders, poor medication adherence, and reduced functional capacity and physiologic reserve to name a few [73]. For example, COPD patients that have impaired global cognitive performance and memory had poor adherence to medication regimes [69]. Physical exercise may improve cognitive function in healthy individuals through improved cerebral metabolism and oxygenation [74]. Pulmonary rehabilitation consisting of exercise, education, and psychosocial counseling improved psychomotor speed and mental flexibility [75]. However, memory and concentration did not improve [75], neurocognitive impairments that are associated with hypoxia/hypoxemia induced hippocampal damage rather than those associated with fatigue. Risk factors and mechanisms of neurocognitive impairments, as well as interventions that may lead to improved cognitive functioning in COPD patients, remain to be determined.

In addition to neurocognitive impairments COPD patients have impaired physical function, health status, and reduced quality of life. These patients often are impaired in their ability to perform activities of daily living (i.e., bathing, dressing, toileting, transfer, continence, and feeding) which are related to their pulmonary function, cough, wheezing, and dyspnea [76]. COPD patients report poor physical function that contributed to reduced quality of life compared to healthy control subjects [60]. Further, the degree of physical limitations paralleled the severity of the COPD. Decreased exercise tolerance and dyspnea are related to poor health status and functional abilities [77]. Even COPD patients with mild hypoxemia have significantly lower quality of life (measured using the Medical Outcome Survey short form-36; SF-36) compared to healthy controls [78]. Higher levels of depression and anxiety symptoms are also associated with reduced quality of life in COPD patients [79].

Neuropsychiatric Morbidity

Neuropsychiatric disorders, especially depression and anxiety, are common in patients with COPD. The prevalence of neuropsychiatric disorders is as high as 30–58% of COPD patients [80]. Other studies note the prevalence of depression is as high as 42% [81] and 37% of patients have anxiety disorder [82]. Depression increases with increased hypoxemia, carbon dioxide levels, or dyspnea [83]. Hypoxia may be a cause factor in the development of depression and anxiety in COPD due to brain injury (see section “Mechanism of Brain Injury” ) [84]; however, treatment with oxygen results in little or no improvement in the severity of the depression [85]. Even patients who undergo long-term oxygen therapy have severe depression [86]. It is unclear why oxygen treatment improves neurocognitive function but not neuropsychiatric function, but may be due to physical or psychological responses. Another possible cause of neuropsychiatric morbidity may be negative self-perception and restrictions in behavioral functioning due to reduced physical capacity [75]. Exercise rehabilitation in COPD patients improves depression and anxiety [75].

Anxiety is also common in patients with pulmonary disorders such as COPD. Vögele et al. found 55% of COPD had a diagnosis of an anxiety disorder compared to 30% of controls [87]. Vögele and colleagues found no associations between anxiety levels and respiratory function in the COPD group, but there was positive association between anxiety levels and physical symptoms, as well as negative cognition in COPD patients with anxiety disorders. The above findings confirm the high prevalence rate of neurocognitive and neuropsychiatric disorders in patients with COPD [87].

COPD Summary

COPD is a chronic obstructive pulmonary disorder associated with poor physical health, reduced physical function, neurocognitive impairments, and neuropsychiatric manifestations including depression and anxiety. Oxygen treatment may improve neurocognitive function in some patients. Studies using exercise rehabilitation show improvements in physical functioning, endurance, neurocognitive functioning, and improved neuropsychiatric well-being [70, 75].

Acute Respiratory Distress Syndrome

The acute respiratory distress syndrome (ARDS) is a common cause of mortality and morbidity and affects 150,000 people per year [88] or more [89] in the United States. Patients are critically ill and survival has improved from less than 50% to approximately 70% [9092] resulting in approximately 100,000 ARDS survivors per year. Acute respiratory distress syndrome is characterized by acute lung injury, hypoxemia, reduced total thoracic compliance, and diffuse bilateral infiltrates [93, 94]. Although the pathophysiology of ARDS is unclear it occurs in response to a variety of insults including sepsis, shock, trauma, pneumonia, massive transfusion, and other medical/surgical conditions. ARDS is frequently accompanied by organ dysfunction, which includes central nervous system dysfunction. Treatment of ARDS requires aggressive supportive care including positive pressure ventilation [92] and increased oxygen concentration with risks of barotrauma, oxygen toxicity, and nosocomial infection. Survivors of ARDS are often left with pulmonary function abnormalities, neuromuscular weakness, diminished health-related quality of life, neuropsychiatric and neurocognitive deficits [9597].

Investigations of the neurologic dysfunction in ARDS survivors have been relatively neglected compared to other organ systems. A study that assessed CNS dysfunction using the Glasgow Coma Scale score found that greater severity of the initial neurologic dysfunction (lower Glasgow Coma Scale scores) and no change or worsening of the neurologic dysfunction were associated with higher 30-day mortality [98]. Neurologic complications of ARDS involve the central and peripheral nervous systems and contribute to significant mortality and morbidity. Brain imaging data in 15 ARDS patients suggest that neurologic morbidity includes neuropathologic changes including generalized brain atrophy and structural lesions [99]. The patients with ARDS and brain CT scans had a longer hospital and ICU length of stays, duration of mechanical ventilation, and lower FiO2 compared to ARDS patients without brain imaging. The patients who underwent brain CT imaging had significantly larger ventricular volumes for the lateral ventricles, III ventricle, temporal horns of the lateral ventricles, total ventricular volume, and ventricle-to-brain ratio (a measure of diffuse atrophic change and a general index of white matter integrity; p values 0.02–0.008) compared to the normal control subjects [99]. Clinical radiological brain CT reports identified seven patients with atrophy that support the finding of ventricular enlargement and brain atrophy. Six of the 15 ARDS patients had mild to moderate cerebral atrophy or ventricular enlargement including hippocampal atrophy in one patient and one patient had increased temporal horn size [99]. The observed brain atrophy may be due to hypoxia during critical illness [96, 100]. The nonspecific brain injury manifested by reduced gyral volume, increased sulcal space, passive increase in ventricular volume, and increased cerebrospinal fluid is common following hypoxic brain injury [101]. While the sample size is small, the data suggest that longer ICU and hospital length of stays, longer duration of mechanical ventilation and lower FiO2 may be risk factors for brain injury [99]. Both the patients with ARDS with and without CT scans had significant cognitive impairments including impaired memory, attention, mental processing speed, and executive function at hospital discharge and 1 year post-hospital discharge. These findings suggest that ARDS may result in neuropathologic injury and concomitant neuropsychological impairments.

Mechanisms of Injury

The mechanisms of ARDS-induced brain injury are just beginning to be elucidated, but hypoxemia is undoubtedly implicated [96, 100]. Hopkins et al. measured pulse oximetry in a prospective cohort of mechanically ventilated ARDS survivors and assessed the relationship between the duration and severity of mean oxygen saturation below 90, with neurocognitive outcome [96] (pulse oximetry: SpO2 level <90 is approximately a PaO2 of 50% or severe hypoxia). The pulse oximetry was measured for a total of 31,665 h, excluding data without a good pulse waveform. Patients’ mean saturations were below 90% for 122 ± 144 h per patient. The degree of hypoxemia correlated significantly with neurocognitive sequelae [96]. Supportive evidence for the role of hypoxia in brain injury includes CA1 neuronal in the hippocampus due to hypoxia and increased S-100B protein serum levels in pigs with acute lung injury [102]. Other mechanisms of brain injury following ARDS include hyperglycemia [103], delirium [104], hypotension [105], and the use of sedatives or analgesics [106]. The mechanisms of neurologic dysfunction are likely multi-factorial in nature. Thus, ARDS may result in significant long-term brain-related morbidity manifest by neurocognitive impairments, neuropsychiatric morbidity, and decreased quality of life.

Neurocognitive Morbidity

Neurocognitive impairments in ARDS patients are long lasting and are reported at 6 months [107], 1 year [96, 105, 108], 2 years [100], and 6 years following hospital discharge [109, 110]. Hopkins and colleagues found ARDS survivors had global neurocognitive decline and impaired memory, attention, concentration, mental processing speed, and global neurocognitive decline [96]. At 1-year follow-up, 30% of the 55 patients had lower intellectual function and 78% had impaired memory, attention, concentration, and/or mental processing speed. In other studies of ARDS survivors neurocognitive sequelae occurred in 73% (54 of 74) of survivors at hospital discharge, 46% (30 of 66) at 1 year, and 47% (29 of 62) at 2 years [100, 105]. Regarding global intellectual function, Hopkins et al. showed that ARDS patients’ estimated premorbid IQ was significantly higher than their measured IQ at hospital discharge but improved to their premorbid level by 1-year follow-up [100]. The finding that patients recovered over time with regard to IQ does not necessarily suggest a comparable recovery in all cognitive domains, as data from traumatic and anoxic brain injury literature suggest that some cognitive abilities are more likely to improve.

The neurocognitive impairments appear to improve during the first 12 months post-hospital discharge [100]. For example, at hospital discharge 70% of ARDS survivors had neurocognitive impairments whereas only 45% had neurocognitive impairments at 1 and 2 years following hospital discharge [100]. While highly prevalent, cognitive impairment demonstrated by ICU survivors is also often quite severe. The aforementioned ARDS patients with cognitive sequelae all fell below the sixth percentile of the normal distribution of cognitive functioning, displaying marked neuropsychological deficits in memory, executive functioning, attention, and mental processing abilities. Impairment does not impact all domains equally – and deficits in some areas rebound relatively more completely than others. The duration of the neurocognitive impairments in at-risk ARDS survivors lasts years and may be permanent. Two studies found that 25–33% of ARDS survivors have neurocognitive impairments 6 years after ARDS [109, 110]. The observed neurocognitive impairments are similar to those reported in other ARDS survivors [108, 109, 111], medical ICU survivors [112], following carbon monoxide poisoning [113], and several years after elective coronary artery bypass surgery [114]. Risk factors for acute and chronic neurocognitive impairments following ARDS are unknown and should be the subject of future studies.

Neuropsychiatric Morbidity

Psychiatric or neurobehavioral morbidity following ARDS is common and includes depression, anxiety, and posttraumatic stress disorder (PTSD). It is unclear whether psychiatric disorders are a psychological reaction to extraordinary emotional and physiologic stress, sequelae of brain injury sustained due to ARDS and its treatment, or all of the above. The combination of medications (e.g., sedatives, narcotics, atypical antipsychotic medications, physiological changes, pain, altered sensory inputs, and an unfamiliar environment may contribute to emotional changes) [115117]. The prevalence and severity of depression, anxiety, and PTSD in survivors of critical illness are variable [109, 117119]. Depression occurs in a quarter [100] to over half of ARDS survivors [118]. For example, one study found 43% of ARDS patients had depression [120] and another study reported that over 50% of ARDS survivors had depression 1 year after intensive care unit treatment [118]. A longitudinal study found that ARDS survivors have moderate to severe depression (16 and 23%) and anxiety (24 and 23%) at 1 and 2 years, respectively [100, 105]. While data are accumulating regarding depression following ARDS, less is known regarding anxiety. Anxiety occurs in as many as 41% of ARDS survivors [121, 122]. A longitudinal ARDS study found that anxiety occurred in 24% at 1 and 2 years [100, 105].

The most commonly reported anxiety disorder in ARDS populations is posttraumatic stress disorder (PTSD). Posttraumatic stress disorder is the development of characteristic symptoms that occur following a traumatic event(s) where triggers include a serious personal threat experienced with helplessness and intense fear [123]. Schelling and colleagues were the first to report PTSD following ARDS and intensive care unit treatment [124]. Almost a third of the ARDS survivors reported impaired memory, bad dreams, anxiety, and sleeping difficulties after ICU discharge, with a prevalence rate of PTSD of 28%. Others have reported high rates of PTSD in ARDS survivors [125]. The prevalence of PTSD is as high as 38% [126] and persists years after intensive care unit discharge [109]. For example, PTSD has been reported at hospital discharge and 8 years after discharge [124]. One treatment study found that ARDS patients treated with hydrocortisone had a significant reduction in the development of PTSD compared to patients without treatment (19 vs. 59%).

ARDS Summary

The significant and sometimes permanent effects of ARDS on neurocognitive and neuropsychiatric functioning are increasingly recognized in the intensive care community regarding the importance of this issue; however, it is less recognized in the psychological or neuropsychological communities. Since the presence of cognitive impairment among ARDS survivors was first systematically identified a decade or so ago, progress has been made to study and better characterize this phenomenon. Neurocognitive impairments following ARDS are prevalent, occur in wide-ranging cognitive domains, and are functionally disruptive. Key questions remain unanswered with regard to determining mechanisms, risk factors, and the degree to which brain injuries associated with ARDS are amenable to rehabilitation.

Carbon Monoxide Poisoning

Carbon monoxide is a colorless, odorless gas produced as a by-product of combustion. Common sources of CO are internal combustion engines and faulty furnaces [127]. Carbon monoxide is the leading cause of poisoning injury and death worldwide [128] and the most common cause of accidental and intentional poisoning in the United States. Carbon monoxide results in approximately 40,000 emergency department visits [129] and 800 deaths per year in the United States [130]. The acute symptoms of CO poisoning are nonspecific and are similar to those associated with flu-like illness, which can make diagnosis of CO poisoning difficult. The brain and heart are particularly vulnerable to the pathological effects of CO [128].

Carbon monoxide poisoning results in focal and generalized neuroanatomical abnormalities observed on magnetic resonance (MR) and computed tomography (CT) imaging. Brain lesions following CO poisoning occur in the cortex [131], cerebellum [132], thalamus [133], and substantia nigra [134]. Lesions also occur in subcortical structures including white matter [135] and basal ganglia including the globus pallidus [136], caudate, and putamen [137, 138]. White matter hyperintensities occur in the periventricular and centrum semiovale or deep white matter regions [41]. In addition to neural lesions, carbon monoxide poisoning may cause neuronal cell loss and concomitant structural atrophy. Atrophy occurs in the fornix [139], hippocampus [37], and corpus callosum [42]. Generalized atrophy is also reported with brain volume reduction manifested by reduced gyral volume, increased sulcal space, and passive ventricular enlargement [37]. One study found CO-poisoned patients had atrophy in the putamen, caudate, and globus pallidus [140].

Neurocognitive Morbidity

Neurocognitive impairments commonly occur following CO poisoning in previously healthy individuals [113]. It is estimated that between 15 and 49% of individuals diagnosed with acute CO poisoning will develop neurocognitive sequelae [141]. A recent review of 18 group studies (N = 979) and 16 case studies (N = 35) found that 94% of the case studies and 33.9% of patients in the group studies had cognitive impairments [142]. Neurocognitive sequelae of CO poisoning are heterogeneous regarding onset, severity, neurocognitive domain affected, and the pattern of neurocognitive deficits is variable [37]. Carbon monoxide-related neurocognitive impairments include impaired memory [143], executive function [144], slow mental processing speed, decreased intellectual function [37], apraxia, aphasia, and agnosia [145]. Neurocognitive sequelae lasting 1 month [146, 147], or more [113, 148], occur in 25–50% of patients with loss of consciousness or COHb levels greater than 25% [147, 149].

Weaver and colleagues studied individuals with acute carbon monoxide poisoning who were compared for neurocognitive outcome following either hyperbaric oxygen or normobaric oxygen treatment in a randomized double blind clinical trial. The neurocognitive impairments were significantly more frequent in the normobaric oxygen group (14.5%) as compared with the hyperbaric oxygen group (3.9%; p = 0.03) [113]. Hyperbaric oxygen therapy reduced neurocognitive impairments by 46% at 6-week outcome. Both groups improved with time, but the difference in neurocognitive impairments between the groups was maintained at 12 months [113]. Thus, treatments such as hyperbaric oxygen may potentially prevent or reduce the neurocognitive impairments that occur following CO poisoning. Risk factors for development of neurocognitive sequelae following acute CO poisoning were assessed using multivariable logistic regression in 163 CO-poisoned patients not treated with hyperbaric oxygen [150]. Of the 163 patients, 68 (42%) manifested neurocognitive sequelae [150]. The risk factors for development of neurocognitive sequelae were aged ≥36 years (odds ratio 2.6; p = 0.005) and CO exposure intervals ≥24 h (odds ratio 2.4; p = 0.019).

Neuropsychiatric Morbidity

Neuropsychiatric morbidity following CO poisoning are common and include depression, anxiety [151], obsessive and compulsive behavior, hallucinations [152], violent outbursts [145], elated mood [146], irritability, and decreased frustration tolerance [153]. The prevalence of CO-related neuropsychiatric morbidity ranges from 33 to 100% [37, 146, 154]. For example, a study by Jasper et al. found significant depression and anxiety in CO-poisoned patients: 45% at 6 weeks, 44% at 6 months, and 43% at 12 months [151]. Accidentally CO-poisoned patients are as likely as individuals with intentional CO poisoning to have depression and anxiety at 6 and 12 months. Patients with neurocognitive sequelae had a higher rate of depression and anxiety at 6 weeks compared to those with no neurocognitive sequelae, but not at 12 months. Patients with intentional CO poisoning have a higher rate of depression and anxiety at 6 weeks compared to accidental CO-poisoned patients, but not at 6 and 12 months. Although there was some subgroup improvement in depression and anxiety over time, the overall prevalence did not change. Hyperbaric oxygen therapy did not reduce the rate of depression and anxiety, but did reduce neurocognitive sequelae [151]. Consistent findings across CO studies to date are the high rate of depression and anxiety following CO poisoning. Similar prevalence rates of depression and anxiety occur in patients with traumatic brain injury and stroke [155], chronic obstructive pulmonary disease [156], acute respiratory distress syndrome [118], and acute myocardial infarction [157].

Accurately predicting outcomes in CO poisoning is difficult as markers of poisoning severity do not appear to predict outcomes. The severity of CO poisoning (measured by COHb level <15% without loss of consciousness) did not result in lower rates of neurocognitive sequelae in less severe CO-poisoned patients compared to patients with more severe poisoning (COHb >15% or loss of consciousness) [158]. Other studies have similarly found COHb levels are not related to neurocognitive deficits [139, 148, 159]. Furthermore, COHb levels and loss of consciousness are neither associated with nor predict clinical outcome [148, 160]. Neither symptoms of poisoning, neurocognitive impairment [37, 148], white matter hyperintensities [41], fornix atrophy [139] nor corpus callosum atrophy [42] is related to CO poisoning severity (e.g., loss of consciousness or COHb level).

CO Summary

Carbon monoxide poisoning may result in significant neurocognitive and neuropsychiatric sequelae which persist 12 months or more post-CO poisoning. Patients with neurocognitive sequelae have a higher rate of depression and anxiety at 6 weeks compared to those with no neurocognitive sequelae, but not at 12 months. Patients with intentional CO poisoning have a higher rate of depression and anxiety at 6 weeks compared to accidental CO-poisoned patients, but not at 6 and 12 months for those with intentional vs. accidental CO poisoning. Hyperbaric oxygen therapy did not reduce the rate of depression and anxiety, but does reduce neurocognitive sequelae. Clinicians need to be aware of neurocognitive and neuropsychiatric morbidity following CO poisoning and remain vigilant about CO prevention.

Obstructive Sleep Apnea

As many as 18 million Americans suffer from obstructive sleep apnea (OSA). Obstructive sleep apnea is more common among men and individuals who snore, are overweight, have high blood pressure, or have physical abnormalities in their upper airways [161, 162]. The incidence of obstructive sleep apnea in this patient population is greater than 70% and increases in incidence as the body mass index increases [163]. Obstructive sleep apnea is a sleep disorder that results in the absence (apnea) or reduction (hypopnea) of airflow lasting at least 10 s despite normal respiratory efforts [164, 165]. Apnea and hypopnea result in hypoxemia and disrupt or fragment the sleep cycle. OSA affects an estimated 2–4% of the middle-aged population and the prevalence increases with age [166, 167]. Common symptoms include excessive daytime sleepiness (EDS), snoring, gasping or choking during sleep, headaches (especially upon waking), irritability, mood disturbance, personality change, motor restlessness, and neurocognitive complaints [168, 169]. OSA is associated with development of pulmonary hypertension, cardiovascular and cerebrovascular disease, hypertension, arrhythmias, and hormonal abnormalities in adults [170172]. Although OSA is associated with medical morbidity such as cardiovascular disease, its most functionally disruptive effects in adults appear to be neurocognitive and neuropsychiatric in nature [173].

Neurocognitive Morbidity

Patients with OSA may exhibit impairments in vigilance, attention, memory, general intellectual functioning, problem solving [174], executive dysfunction, visuospatial abilities [175], and psychomotor speed [174, 176]. Impairments in memory and attention are the most commonly reported [174]. The neurocognitive and neuropsychiatric morbidities associated with OSA are associated with both intermittent hypoxia and sleep fragmentation [176178]. Further, neurocognitive impairments appear to be exacerbated by the severity and duration of hypoxemia [176, 179], higher apnea hypopnea index scores [180], and sleep arousals [180]. There is continued discussion regarding which of the above aspects of OSA are associated with neurocognitive impairments. Previous research has reported both improvement [181] and no change in cognitive function following nCPAP [182].

A study comparing neuroimaging and neuropsychological findings in 14 patients with OSA and 20 CO-poisoned patients found hippocampal and generalized atrophy and neuropsychological impairments in CO poisoning and OSA [144]. Hippocampal atrophy occurred in both groups; however, increased VBR due to generalized cerebral atrophy (i.e., whole brain volume loss) was greater in the CO group [144]. The groups may differ due to moderation of tissue damage from intermittent hypoxia observed on OSA instead of a single episode of longer duration as occurs in CO poisoning. Therefore, the duration of hypoxia may account for the more severe generalized brain atrophy observed in the CO patients. Alternatively, the long-term effects of chronic intermittent hypoxia, such as occurs in OSA, may result in cerebral vascular problems, neurodegeneration, and neurocognitive deficits due to the cumulative effects of the hypoxia [183]. The CO group consistently performed worse on most neurocognitive measures while the OSA group had more selective neurocognitive impairments (predominately executive dysfunction and impaired memory). Improvement in neurocognitive function in OSA patients following 6 months of nasal continuous positive airway pressure (nCPAP) treatment was limited to executive function. Differences in test performance between the CO and OSA groups were more pronounced after the OSA group had received 6 months of nCPAP treatment [144]. Previous research has suggested that the hippocampus is more vulnerable to hypoxic injury than adjacent structures such as the parahippocampal gyrus or temporal lobes [44]. Alternatively, injury to the prefrontal cortex, rather than medial temporal lobe structures may be responsible for the neurocognitive and neuropsychiatric morbidity associated with OSA [173].

Neuropsychiatric Morbidity

Neuropsychiatric morbidity including depression and anxiety are common in patients with OSA [177, 164]. The prevalence of depression is 24% [184] to as high as 45% of individual with OSA [185]. Depression is associated with fatigue, feeling tired, sleepiness, and reduced motivation in OSA patients [186]. A study that assessed the prevalence of neuropsychiatric disorders in 171 patients with sleep disorders (83% with OSA) found that 11% of patients had major depression, 7% minor depression, 3% panic disorder, and 12% anxiety [187]. The rate of depression and anxiety in OSA is higher than population norms [188]. Gender differences in neuropsychiatric disorders appear to be more common in women than men with OSA, similar to that observed in the general population. While the prevalence of OSA is higher in men, McCall et al. [189] found women with OSA had more common and more severe symptoms of depression and milder hypoxemia was associated with worse depression [189]. Nasal CPAP does not appear to improve depression or anxiety [190]. However, one study noted some improvement after long-term nCPAP therapy (approximately 1 year) [191]. In addition uvulopalatal flap surgery improved depression and anxiety in OSA patients [188].

OSA Summary

The neurocognitive and neuropsychiatric impairments are common in individuals with OSA with intermittent hypoxia. The neurocognitive and neuropsychiatric sequelae are similar to those observed in CO poisoning and other pulmonary disorders and are associated with acute hypoxia. Research is needed to determine mechanisms, risk factors, and treatment for OSA-associated neuropsychiatric morbidity.

Rehabilitation Outcomes Following Anoxia

Outcome following severe anoxia is variable, however, the majority of patients have poor outcome [3]. Information regarding the effects of rehabilitation on neurocognitive outcome following anoxic brain injury is limited. Survival rates following post-anoxic coma range from 9 to 40% [192, 193]. Patients who survived anoxic coma regain mobility and ability to perform activities of daily living but not neurocognitive [194]. Outcome following anoxic coma was not predicted by age, sex, site of resuscitation, cause of anoxia, nor presence of post-anoxic seizures [193]. A single case suggested that “relatively” good neurocognitive function 1 month post-anoxic coma suggesting some recovery and benefit of rehabilitation [194]; however, this finding is not generally reported.

Groswasser et al. [195] followed a group of 31 comatose patients following anoxic brain injury, 13 were independent in activities of daily living, 2 regained premorbid neurocognitive functioning, and 4 returned to work, but only 1 to the same job. Patients who were younger with shorter coma had “relatively better outcomes”. The differences in recovery may be due to the interaction of the diffuse damage and delayed cell death, but not the etiology of the anoxic brain injury. Armengol [195] reported eight individuals with severe anoxia who were treated in a long-term neurobehavioral rehabilitation program. Six of the eight individuals had poor outcome with significant impairments in attention, executive function, memory, reasoning, language, visuospatial, and motor skills, while two patients exhibited mild neurocognitive impairments. In-patient rehabilitation appears to improve functional status, with individuals who had higher Functional Independence Measure scores on admission had the best outcome; however, few resumed their previous jobs and level of function [196].

Little is known regarding rehabilitation outcomes in many hypoxic disorders. Further, it is unknown if the severity of hypoxia/hypoxemia is related to rehabilitation outcomes. For example, there are no studies that assess rehabilitation after ARDS. A few studies assess rehabilitation following COPD and OSA. As noted above, pulmonary rehabilitation consisting of exercise, education, and psychosocial counseling improved psychomotor speed and mental flexibility but not memory and concentration in COPD patients [75]. One study found lung volume reduction surgery plus pulmonary rehabilitation (exercise and education) compared to pulmonary rehabilitation alone and found improved neurocognitive and neuropsychiatric function in the lung volume reduction group [197]. It remains to be determined if rehabilitation on neurocognitive and neuropsychiatric morbidity in patients with hypoxia/hypoxemic disorders is effective.

Conclusions

Patients with respiratory disorders and concomitant brain injury exhibit both diffuse and focal brain injury and concomitant neurocognitive and neuropsychiatric sequelae. Respiratory disorders are heterogeneous and include cardiac or respiratory arrest, COPD, CO poisoning, OSA, and ARDS. The associated hypoxic or anoxic brain injury results in focal and diffuse neuropathologic lesions and atrophy including hippocampal, basal ganglia, cerebellar, and white matter abnormalities. Neuropsychological impairments include generalized intellectual decline, memory deficits, decreased attention, visuo-perceptual, problem solving, executive dysfunction, and decreased mental processing speed. Further, these individual may experience a high rate of neurobehavioral disorders including euphoria, irritability, hostility, depression, and anxiety and personality changes. Thus, respiratory disorders and their associated hypoxia and ischemia result in significant neurological structural and functional abnormalities, and neuropsychological impairments.

Questions remain regarding risk factors for development of neurocognitive and neuropsychiatric sequelae, precise mechanisms of brain injury, and whether there are treatments that will prevent or ameliorate these sequelae. Further, physicians and other health-care providers often are unaware and do not assess for the presence of neurocognitive impairments. A recent study found 42% of ARDS survivors underwent rehabilitation therapy, but most were not evaluated for neurocognitive impairments, with only 12% identified as having neurocognitive impairments by the clinical rehabilitation team [100]. Neurocognitive impairments appear to be under recognized by both intensivists and rehabilitation providers. Studies suggest that in non-critical care clinical settings many physicians fail to recognize (or assess) neurocognitive impairment in 35–90% of patients [198]. Increased identification of neurocognitive impairments patients with respiratory disorders may benefit patients by raising physician awareness potentially leading to increased referrals to rehabilitation specialists, neuropsychologists, speech and language therapists, and other health-care providers who can provide interventions such as cognitive remediation. It should be noted that there is a paucity of data regarding interventions for neurocognitive impairments or the potential benefit of such interventions in critically ill patients.

Today, it is recognized that neurocognitive sequelae are common in patients with respiratory disorders, especially those with concomitant hypoxia. The neurocognitive impairments are long lasting, and may be permanent, although substantial research needs to be done to fully understand the prevalence, nature, risk factors, etiology, and nuances of the neurocognitive and neuropsychiatric impairments in this population. Referrals to colleagues in rehabilitation medicine, psychiatric, neurology, or psychology would facilitate evaluation of potential areas of concern. Attention to proximal determinants and possible interventions to prevent neurocognitive morbidity are warranted and should be an emphasis in outcomes research. Such research will likely yield valuable insights into identification, the natural history, prognosis, and potential mechanisms of the neurocognitive deficits and guide the development, implementation, and fine-tuning of intervention programs.