Share this article on:

Gas Exchange and Exercise Capacity Affect Neurocognitive Performance in Patients With Lung Disease

Parekh, Priti I. PhD; Blumenthal, James A. PhD; Babyak, Michael A. PhD; LaCaille, Rick PhD; Rowe, Sarah BS; Dancel, Liz BS; Carney, Robert M. PhD; Davis, R Duane MD; Palmer, Scott MD; for the INSPIRE Investigators

Original Articles

Objective: This study examined the relationship between cognitive functioning and the severity of underlying lung disease in patients awaiting lung transplantation.

Methods: Ninety-four patients with end-stage lung disease completed a test battery to assess cognitive performance in two domains: executive functioning/attention (Trails A and B, COWA, Animal Naming, Stroop Color–Word Test, Digit Symbol, and the 2 & 7 Test) and verbal memory (Digit Span–Backward and Forward, WMS-R Logical Memory and Paired Verbal Associates).

Results: Thirty-seven percent of the patients demonstrated moderate to severe cognitive impairment data on two or more tests. Adjusting for age and education, there were no statistically significant differences on executive functioning or verbal memory as a function of specific lung disease diagnosis. Lower PCO2 values were associated with better cognitive performance on latent measures of executive functioning and attention (p = .006) and verbal memory (p = .009), whereas higher PO2 values tended to be associated with better performance on the executive functioning/attention measure (p = .064). Distance walked in 6 minutes was positively related to verbal memory (p < .023).

Conclusions: Impaired neurocognitive functioning may be relatively common in patients awaiting lung transplantation and is associated with ineffective pulmonary gas exchange and reduced exercise tolerance.

CFI = Bentler’s Comparative Fit Index; COPD = chronic obstructive pulmonary disease; FEV1 = percent predicted forced expiratory volume in one second; FVC = forced vital capacity; INSPIRE = Investigational Study of Psychological Intervention in Recipients of Lung Transplant; MANOVA = multivariate analysis of variance; PCO2 = arterial pressure of carbon dioxide in the blood; PO2 = arterial pressure of oxygen in the blood; RMSEA = root mean square error of application; TLFI = Tucker-Lewis Index; WAIS-R = Wechsler Adult Intelligence Scale–Revised; WMS-R = Wechsler Memory Scale–Revised.

From the Departments of Psychiatry and Behavioral Sciences (P.I.P., J.A.B., M.A.B., R.L., S.R., L.D.), Surgery (R.D.D.), and Medicine (S.P.), Duke University Medical Center, Durham, North Carolina; and the Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri (R.M.C.).

INSPIRE Investigators: James A. Blumenthal, PhD (Director, Behavioral Medicine); Robert M. Carney, PhD (Codirector, Behavioral Medicine); R. Duane Davis, MD (Surgical Director); Scott M. Palmer, MD (Medical Director); Michael A. Babyak, PhD, Robyn Claar, PhD, Iris Csik, MSW, Liz Dancel, BS, Francis J. Keefe, PhD, Kenneth Freedland, PhD, Elizabeth Gullette, PhD, Joel Hughes, PhD, Rick LaCaille, PhD, Kari Merrill, PhD, Melissa Napolitano, PhD, Jennifer Norten, PhD, Priti I. Parekh, PhD, Sarah Rowe, BS, Victor Tapson, MD, and Elbert Trulock, MD.

Address correspondence and reprint requests to James A. Blumenthal, PhD, Box 3119, Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC 27710. E-mail:

Received for publication June 10, 2004; revision received December 27, 2004.

This study was performed at Duke University Medical Center and was supported by NHLBI grant HL 65503–01.

Back to Top | Article Outline


Morethan 25 million people in the United States are estimated to be currently living with chronic lung disease, whereas nearly 360,000 die each year from this condition (1). Lung transplantation is increasingly used as a treatment option for individuals with an array of end-stage pulmonary diseases, despite the myriad of associated risks (2,3). Currently, almost 4000 people await lung transplant in the United States and over 10,000 patients have benefited from surgery. Not surprisingly, advanced lung disease is associated with disturbances in psychologic adjustment and quality of life among candidates awaiting lung transplantation (4–8). Also, earlier studies found that cognitive impairments are often present in individuals with chronic obstructive pulmonary disease (COPD), with increased hypoxemia (decreased blood oxygen levels) associated with poorer neurocognitive test performance (9–11). Moreover, human and animal studies of chronic intermittent hypoxia resulting from sleep apnea also suggest that hypoxia leads to neurocognitive deficits and neurodegeneration (12). Stuss et al. (13) also have suggested that decreased pulmonary ventilation and hypercapnia (increased blood carbon dioxide levels) in COPD may result in more impaired cognitive deficits than those associated with hypoxemia alone.

The impact of end-stage lung disease on cognitive functioning may be relevant to patients’ medical and surgical management. Cognitive impairment in patients requiring organ transplantation may compromise their ability to comply with the complex posttransplant medical regimen, and noncompliance, particularly with immunosuppressive medications, significantly increases the risk for graft loss and death (14–17). Williams et al. (18) found mild to moderate deficits on tests of executive functioning, verbal memory, and naming in a sample of 39 lung transplant candidates. Although their sample was small, more than one third of the patients displayed cognitive deficits to the degree that impairment would likely affect their daily functioning. More recently, Crews and colleagues (19–21) examined 134 patients with end-stage pulmonary disease who were being evaluated for transplantation and found cognitive impairments primarily on tasks that involved verbal retrieval and immediate recall of a list of unrelated words. However, only a subset of patients completed the entire battery, and no measures of pulmonary function were obtained. In a small subset of 28 lung transplant candidates, Rodrigue and colleagues (22) found that the group mean scores on subtests from the Wechsler Adult Intelligence Scale–Revised (WAIS-R) assessing perceptual organization, attention, and short-term memory were in the average range; the percentage of patients showing impairment on the tests was not reported, however, and pulmonary function and exercise tolerance were not associated with cognitive functioning. In an effort to establish a more thorough normative database with 100 lung transplant candidates, Ruchinskas et al. (23) examined cognitive performance on several standardized neurocognitive measures such as the Trail Making Test and subtests of the Wechsler Memory Scale. Although the relationship between pulmonary function and neurocognitive performance was not assessed, the rates of neurocognitive impairment ranged from 5.3% to 48.9%, with measures of executive function revealing fewer deficits than did those assessing verbal, noncontextual memory.

Given the prior evidence linking hypoxia, hypercarbia, or both to neurocognitive impairments in patients with advanced lung disease, we hypothesized that cognitive impairments are likely to be highly prevalent among lung transplant candidates. Furthermore, we hypothesized that the severity of cognitive dysfunction would be related to the severity of objective pulmonary function. Prior studies in lung transplant candidates generally have relied on small samples and failed to consider the relationship between neurocognitive functioning and objective medical variables, including native disease, arterial oxygenation, and the results of the 6-minute walk test. Impaired functional capacity may be especially relevant for lung transplant candidates because lower levels of physical fitness have been shown to be associated with worse performance on neurocognitive tests, especially on tasks of executive functioning (24,25). It should be noted that PO2 and PCO2 reflect arterial blood measurements of oxygenation and carbon dioxide. Arterial blood generally has high levels of oxygen and low levels of carbon dioxide. Because the hemoglobin dissociation curve is exponential and not linear, humans can tolerate significant drops in the concentration of arterial blood oxygen before significant tissue hypoxia occurs. Once PO2 drops below 60 mm Hg, however, end-organ damage is possible. Organs such as the brain, the heart, and the kidneys are particularly sensitive to effects of tissue hypoxemia. The metabolic effects of elevated PCO2 are less well understood and perhaps less damaging because renal compensation mechanisms will retain bicarbonate to maintain a relatively normal tissue pH until very high levels of PCO2 occur. Because effects of PO2 and PCO2 should be consistent on the central nervous system and neurocognitive function regardless of the underlying disease, we focused on these parameters rather than forced expiratory volume in 1 second (FEV1), which varies considerably depending on the underlying pulmonary disease.

Therefore, the purpose of the current study was: 1) to investigate the neurocognitive functioning of patients with end-stage pulmonary disease awaiting lung transplantation by diagnostic category, and 2) to examine the relationship between cognitive functioning, exercise tolerance (6-minute walk test), and pulmonary gas exchange (PCO2 and PO2) in a relatively large sample of pulmonary patients awaiting lung transplantation.

Back to Top | Article Outline


Patient Population

A consecutive series of patients identified as candidates for lung transplantation at Duke University Medical Center were recruited for this study between November 2000 and April 2003. Patients were considered for lung transplantation if they were diagnosed with advanced lung disease and met general criteria for transplant eligibility as outlined by the International Society for Heart and Lung Transplantation (26). For example, patients with other significant medical comorbidity (e.g., prior cerebral vascular accident or dementia) were excluded from consideration of transplantation. Inclusion criteria for participation in this analysis were: age 18 years or older, current placement on the waiting list for an initial single or bilateral lung transplant, and the ability to complete verbal and written assessments (i.e., basic English proficiency; lack of significant hearing, speaking, or visual difficulties). This report is based on data obtained from 94 lung transplantation candidates who met these inclusion criteria and consented to participate in a larger behavioral intervention trial called INSPIRE (INvestigational Study of Psychological Intervention in REcipients of Lung Transplant).

Initially, a total of 434 patients were identified as potential candidates and approached to participate in this study. Of these patients, 247 (57%) were ineligible for the following reasons: not listed for lung transplantation (n = 135), listed for lung retransplantation (n = 1), required multiple organ transplantation (n = 19), underwent transplantation before study enrollment (n = 46), died before enrollment (n = 26), evidenced difficulty speaking secondary to the medical condition (n = 5), unable to schedule or complete the evaluation (n = 14), and already enrolled in a competing protocol (n = 1). Of the remaining 187 eligible patients, 93 (50%) refused to participate in the study. Patients who declined to participate in this study were younger (43 versus 50 years, p = .02) and had been on the waiting list longer (445 versus 251 days, p = .043) than persons who provided informed consent to participate in this study. However, there were no significant differences with respect to gender, pulmonary function, or other clinical or demographic characteristics between those who declined to participate and those who participated.

Back to Top | Article Outline


Patients were initially contacted either by mail or in person during a pulmonary clinic appointment. After patients provided Institutional Review Board-approved, written, informed consent to participate, they were individually administered a neurocognitive test battery by a trained psychometrician. Medical records were reviewed for pulmonary gas exchange (PCO2, PO2) and exercise tolerance (6-minute walk test) values closest to the date of the patients’ neurocognitive testing. These procedures have been shown to be reproducible (27–29) and are accepted measures oflung function.1 Because all patients have irreversible end-stage lung disease, results of these medical tests would be expected to remain unchanged or worsen over time. It should be noted that the timing of patents’ transplants relative to neurocognitive testing was highly variable, with transplants occurring from weeks to years (if at all) after the neurocognitive testing.

Back to Top | Article Outline

Neurocognitive Measures

The following instruments were administered individually to all participants in fixed order:

1. Digit Span Subtest–Forward (30). A subtest from the Wechsler Adult Intelligence Scale–Revised (WAIS-R), Digit Span–Forward requires patients to repeat progressively longer series of numbers, ranging from three to nine digits in length, immediately after they are read aloud by an examiner. The task is discontinued once two trials of the series of the same length are incorrectly repeated. Scores range from 0 to 14 with higher scores indicating better performance. Digit Span–Forward measures efficiency of attention and has satisfactory test–retest reliability ranging from 0.66 to 0.89 depending on the patient’s age and the interval length of test administration.

2. Digit Span Subtest–Backward (30). Digit Span–Backward requires patients to repeat a series of numbers, ranging in length from two to eight digits, in reverse sequence immediately after they are read aloud by the examiner. This subtest is discontinued and scored using the same procedures as the Digits Span–Forward subtest. Higher scores indicate better performance. Digit Span–Backward is thought to call on patients’ working memory and is estimated to have a test reliability of 0.83.

3. Trail Making Test—Parts A and B (31). The Trails A test requires patients to connect the numbers 1 to 25 in sequence. The score is determined by the amount of time required to complete the task, with higher scores indicating poorer performance. The Trails A is widely considered a measure of sustained attention and is thought to be sensitive to cognitive impairment, with satisfactory reliability (e.g., r = 0.79). Trails B requires patients to connect 25 numbers and letters in alternating sequence (i.e., 1-A–2-B, and so on). The score is the time in seconds required for completion, with higher scores reflecting poorer performance. The Trails B test is well established as a sensitive measure of cognitive flexibility, with studies typically reporting reliability coefficients in the 0.80 to 0.90 range.

4. Stroop Color–Word Test (32). The Stroop Color–Word Test is composed of three, 100-item trials using word, color, and interference lists. The first trial involves the patient reading aloud a list of color words as quickly as possible, and the second trial incorporates successive naming of the color of listed stimuli. The final trial uses the interference list, which consists of a series of color names that differ from the color of their print (e.g., the word “red” printed in blue ink). A higher score indicates better test performance. A measure of response inhibition and executive functioning, the Stroop Color–Word Test has been shown to have good reliability (33).

5. Controlled Oral Word Association (COWA) (34,35). The COWA requires patients to generate as many different words as possible that begin with a particular letter, excluding proper nouns and suffix variations. Three letters are used in the procedure with increasing level of difficulty, and there is a time limit of 60 seconds per letter. Scores are calculated by summing the acceptable words produced across the three letters. A higher score indicates better performance. The COWA has been shown to have high internal consistency (R = 0.83) and acceptable test–retest reliability (R = 0.74).

6. Animal Naming Task (34,35). Used to assess semantic verbal fluency, the Animal Naming Task requires patients to generate the names of as many animals as possible in 60 seconds. Scores are determined by summing correct responses, with higher scores reflecting better performance. Semantic verbal fluency tests such as the Animal Naming Task are considered indices of executive functioning and have been shown to be sensitive to detecting cognitive impairment in various patient groups (33).

7. Ruff 2 and 7 Selective Attention Test (36). This instrument requires patients to visually search for and identify the digits 2 and 7, which are randomly embedded within 20 alternating sets of letter and digit distractors. Completed within a 5-minute time limit, scores are calculated from the number of correctly identified targets with higher scores reflecting better performance. Test–retest reliability has been estimated between 0.84 and 0.97. The word and letter scores were standardized and combined to create a global score for selective attention.

8. Logical Memory I Subtest (37,38). The Logical Memory subtest from the Wechsler Memory Scale–Revised (WMS-R) was used in this study. This subtest requires patients to repeat two paragraph-length stories immediately after they are read aloud by the examiner. Scores are determined by summing the number of story details correctly recalled, with a higher score reflecting better performance. This subtest of the WMS-R measures the ability to accurately reproduce newly learned verbal information, with test–retest reliability for immediate recall estimated to be 0.70.

9. Digit Symbol Subtest (30). The Digit Symbol subtest from the WAIS-R involves patients drawing a series of symbols that match one of 10 digits using a key. Scores on this subtest are the number of correct symbols drawn in 90 seconds, with higher scores indicating better performance. The test–retest reliability of this measure for adults ranges from 0.82 to 0.86.

10. Verbal Paired Associates subtest (37). The Verbal Paired Associates subtest from the WMS-R requires patients to remember a set of word associations that are verbally presented. Immediately after presentation of a set of eight word pairs, half of which are semantically related (e.g., baby–cries) and half of which are unrelated (e.g., pen–grocery), patients were cued with one word from each pair and asked to produce the other. In a modification of the original WMS-R procedure, the test was discontinued after three trials regardless of whether the examinee was able to recall all the pairs correctly. Higher scores reflect better performance and are calculated by summing the total number of correctly recalled word associations across the three trials. The test–retest reliability for this measure has been shown to range from 0.53 to 0.72 depending on testing interval.

Neurocognitive tests were grouped a priori into two conceptual domains reflecting the underlying cognitive function that they are believed to assess: 1) Executive Functioning and Attention (Controlled Oral Word Association, Animal Naming, Stroop Color Word Test, Trail Making Test—Parts A and B, Digit Symbol, and Ruff 2 & 7 Selective Attention Test,); and 2) Verbal Memory (Digits Forward and Backward, Logical Memory, and Verbal Paired Associates).

Impairment ratings for each of the neurocognitive tests were determined by comparing patients’ test scores with published normative values as a function of gender, age, and education level (30–38).

Back to Top | Article Outline

Pulmonary Gas Exchange and Exercise Tolerance

Arterial blood analysis of carbon dioxide and oxygen (PCO2 and PO2) was routinely conducted during pulmonary clinic visits. The average interval between pulmonary gas exchange analysis and neurocognitive assessment was 38 days (standard deviation = 97 days; range = 0–600 days; median and mode = 0 days). It is well established that these measures are stable in patients with end-stage lung disease (27–29). The 6-minute walk test used a standard protocol to determine exercise tolerance by measuring the distance that patients were able to walk within a 6-minute time limit. Patients were asked to cover as much distance as possible at a self-selected pace and were provided with enough oxygen to maintain saturations of 90% or greater. No additional encouragement was provided. The test was performed by an experienced physical therapist at a dedicated pulmonary rehabilitation facility located on the Duke campus at the time of transplant evaluation.

Back to Top | Article Outline

Statistical Analysis

For the purposes of this study, we categorized patients into four groups based on their pulmonary diagnosis: obstructive disease (e.g., chronic obstructive pulmonary disease, emphysema, α1-antitrypsin deficiency), cystic fibrosis, noncystic fibrotic disease (e.g., idiopathic pulmonary fibrosis, sarcoidosis), and “other” (e.g., bronchiectasis, lymphangioleiomyomatosis, primary pulmonary hypertension).

We used structural equation modeling (SEM) as available in Muthén’s Mplus software (39) first to examine the tenability of the proposed grouping of the neurocognitive tests and then to simultaneously estimate the relation among neurocognitive function, demographic variables, and pulmonary function. This approach improves the reliability of the neurocognitive assessment and reduces the number of statistical tests on key parameters (namely, the structural relations between pulmonary function measures and neurocognitive performance). For the purpose of these analyses, we combined the letters and numbers subtests of the Ruff Selective Attention Test, and also log-transformed and reversed the scores from the Trail Making Test Parts A and B. We used these along with the remainder of the neuropsychologic test scores to create two a priori latent variables, or factors that measure two domains of neurocognitive functioning: verbal memory and executive function and attention. We then used confirmatory factor analysis to evaluate the extent to which the proposed factor structure was consistent with the observed data. After determining that this model fit the data adequately (Table 1), we proceeded to examine the relation between the factors and the following predictors: age, education, PCO2, PO2, and functional capacity as measured by the 6-minute walk test. Specifically, we regressed these two cognitive factors on these predictors in a simultaneous structural equation model, again using the Mplus software. We report the standardized form of the factor loadings and regression coefficients. We used maximum likelihood estimation for both the confirmatory factor analysis and structural parameters estimation. A bootstrap validation procedure was also conducted, yielding essentially identical results. We therefore report only the maximum likelihood results here.

Back to Top | Article Outline


Background Characteristics of the Sample

Table 2 shows the descriptive statistics for the 94 patients included in the study. The mean age of the sample was 50 years. Nearly half were women, most were white, and most had at least a high school education. The most frequent pulmonary diagnosis was obstructive disease, and 67% of patients reported a positive tobacco smoking history. Fifty patients (53%) met criteria for clinical hypercapnia (PCO2 >45 mm Hg) and/or clinical hypoxemia (PO2 <55 mg Hg).

Back to Top | Article Outline

Neurocognitive Test Scores and Impairment Ratings

The means and standard deviations, along with clinical impairment ratings for the individual neurocognitive measures, are provided in Table 3, with clinical ratings based on the established norms for each measure. Overall, 90 (95.7%) patients demonstrated at least mild impairment on one or more of the neurocognitive measures, whereas 57 (60.6%) patients evidenced moderate-to-severe impairment on at least one of the measures. Twenty-two (23%) patients showed moderate-to-severe impairment on just one test, 13 (14%) showed moderate-to-severe impairment on two tests, 11 patients (12%) showed impairment on three tests, eight patients (8.5%) showed impairment on four tests, two patients (2%) showed impairment on five tests, and one patient (<1%) showed impairment on six tests. Cognitive impairment was relatively uniform, although some neurocognitive measures (Animal Fluency, Ruff 2 & 7 Selective Attention Test, Verbal Paired Associates) appeared more sensitive than others (Digit Symbol, Digit Forward, Digit Backward) in detecting impairments.

Back to Top | Article Outline

Neurocognitive Test Scores and Diagnostic Category

Based on the multivariate test, there were no statistically significant differences among the disease groups on any of the tests (Wilk’s lambda = 0.71, df = 33, p = .710). Given the small cell sizes for some of the diagnostic categories, we also examined diagnostic group differences for each test individually and, despite the possible increased statistical power, also found no evidence of important differences (p values ranged from .23 to .70).

Back to Top | Article Outline

Neurocognitive Factors and Fit to the Observed Data

We examined whether the two factors we created were tenable given the observed data using confirmatory factor analysis using maximum likelihood estimation. We examined four standard indices of fit for confirmatory factor models: Bentler’s Comparative Fit Index (CFI), the Tucker-Lewis Index (TLFI), the root mean square error of approximation (RMSEA), and the chi-square. Each of these indices was within the acceptable range of a well-fitting model: CFI = 0.975, TLFI = 0.964, RMSEA = 0.044, and chi-square (df = 37) = 43.7, p = .207 (p values >.05 for the chi-squared test indicate that the model-implied variance–covariance matrix is not significantly different from the observed matrix), indicating that the a priori factor structure is reasonable given the observed data. The standardized loadings of the items, along with the variance explained by each factor, are shown in Table 1. The executive function and attention variable was predominated by the Digit Symbol and Stroop Test scores, whereas the verbal memory factor was most closely associated with Logical Memory and Paired Verbal Associates scores. Although Animal Naming and Digits Forward showed weaker loadings on their respective factors, these lower loadings do not diminish the reliability of the factor because the unshared variance is partitioned out into the error term. That is, the factors represent only the variance shared among their indicator variables.

Back to Top | Article Outline

Relation of Neurocognitive Domain Scores With Age, Education, PO2, PCO2, and 6-Minute Walk Test

Table 4 displays the results of regressing the verbal memory and executive function and attention factors on age, education, PO2, PCO2, and the 6-minute walk test. Age and PCO2 were inversely related to the executive function and attention factor, whereas education was positively related. PO2 was also positively related to the executive function factor, but it reached only marginal significance levels. The relation between the 6-minute walk test and the executive function and attention factor was weak and not statistically significant. For the verbal memory factor, PCO2 was inversely related to this factor, whereas education and the 6-minute walk test were positively related to the factor. Neither age nor PO2 were significantly related to the verbal memory factor. Figure 1 illustrates the relation between PCO2 and the executive function/attention and verbal memory factors.

Back to Top | Article Outline


Impaired neurocognitive functioning was prevalent in this series of patients awaiting lung transplantation. We found that 37% of patients with end-stage pulmonary disease demonstrated moderate to severe impairment on at least two measures of neurocognitive functioning. We found no significant differences in neurocognitive performance as a function of lung disease diagnosis, although as a result of the small numbers of patients in some categories, there may have been insufficient statistical power to detect group differences. Cognitive functioning was related to pulmonary gas exchange and exercise tolerance, however. Specifically, after adjusting for age and education, higher PCO2 values were associated with worse cognitive performance on the executive functioning/attention latent measure. Also, the distance walked in 6 minutes was positively related to the verbal memory construct after adjusting for age and education.

Overall, our finding of neurocognitive impairment among our sample of patients is consistent with other studies of lung transplant candidates, which also have reported deficits in executive functioning and verbal memory (18–23). We found the highest rate of moderate–severe impairment (28% of patients) on the WMS-R Verbal Paired Associates test, which assesses verbal memory for both related and unrelated word pairs. In contrast, we found the rate of moderate–severe impairment on the WMS-R Logical Memory test, a measure of contextual verbal memory, to be only 8%. These results are consistent with other studies that have found the most impairment to occur on tests of noncontextual verbal memory, with contextual verbal memory remaining relatively more intact (21,23) However, the present study also revealed some previously unreported deficits in executive function and attention in this population. We found that at least 35% of patients were impaired on verbal and category fluency tests, a result contrary to the only other study assessing fluency performance in lung transplant candidates, which found average performance on verbal fluency (18). Our results also suggest that attentional difficulties and deficits in executive functioning may be common in patients with end-stage lung disease, because we found that 15% of patients were impaired on three or more of the eight measures of executive function/attention assessed. Perhaps other neurocognitive studies in lung transplant candidates failed to identify such deficits because only simple tests of attention, such as Digit Span, were used. These results show that impairment in any one of a number of areas of cognitive functioning is common among patients awaiting lung transplant.

Our finding that pulmonary gas exchange was significantly associated with verbal memory in our sample is consistent with several studies in patients with COPD (9–11,40). Although only 50 (53%) of our patients met criteria for clinical hypercapnia (PCO2 >45 mm Hg) and/or clinical hypoxemia (PO2 <55 mg Hg), lower PCO2 levels were associated with better performance on both of the cognitive domains assessed, but higher PO2 values were only marginally associated with better performance on the executive functioning/attention construct. We found that PCO2 explained approximately the same amount of variance in the verbal memory latent variable and about half the amount of variance in the executive function/attention construct as did education, demonstrating that it is both clinically and statistically significant predictor of cognitive functioning in our sample. We hypothesized that both elevation in arterial carbon dioxide or decrements in oxygenation could lead to cellular injury in the brain and neurocognitive dysfunction. Although elevations in PCO2 and declines in PO2 were both associated with impairment of function, the effects with PCO2 were stronger. This effect would imply that the brain is more sensitive to changes in PCO2, which is not surprising because of the important effects of PCO2 on blood and tissue pH and metabolic cell function. These results suggest that monitoring of pulmonary gas indices, even if not in the clinically abnormal range, could help identify a subset of transplant candidates who may experience difficulties in cognitive functioning. Another implication of these findings is that oxygen therapy in appropriate candidates may improve attention by attenuating hypoxemia, although it is unclear to what extent, if any, changes are reversible.

Along with pulmonary gas exchange, exercise tolerance emerged as a significant predictor of neurocognitive performance such that greater distance walked on the 6-minute walk was related to better performance on verbal memory latent measures. The 6-minute walk test score was a relatively strong predictor, accounting for about as much variance in the verbal memory construct as education level. Although a previous, smaller study in transplant candidates failed to find a relationship between exercise capacity and subtests of the WAIS-R (22), our findings are consistent with a study of patients with COPD that found that higher exercise capacity was significantly related to less cognitive impairment on summary measures of the Halstead-Reitan battery (11). Moreover, a recent study by Etnier and colleagues (41) showed that 6-minute walk tolerance and age predicted fluid intelligence performance, whereas forced vital capacity (FVC), 6-minute walk, and age were associated with speed-of-processing. FVC was also related to working memory performance. Research also has shown that exercise training improves overall cognitive functioning in healthy older adults (42) and also improves verbal fluency (43) and fluid intelligence (44) in patients with COPD. It is possible that pulmonary rehabilitation in patients with end-stage lung disease may lead to benefits in cognitive functioning by improving exercise tolerance and physical conditioning. However, it should be noted that the effects of exercise on cognition are not consistent, with some evidence suggesting that tasks that assess executive functioning may be differentially improved by exercise training (24,25).

There are some limitations to interpreting the results of this study. First, in the absence of a healthy control group, we had to rely on published normative data for comparison with our sample of patients with end-stage lung disease. Ideally, it would have been preferable to include a sample of age-matched healthy controls for comparison. Also, although we report that one third of our sample showed impairment in the domains that we assessed, the extent to which neurocognitive test performance reflects deficits in everyday behaviors during routine activities of daily living is uncertain. Poor cognitive functioning could conceivably interfere with the patients’ ability to adhere to an often complex pretransplant medical regimen, but in the absence of data, we cannot determine whether the degree of cognitive impairment we found would affect medical compliance or posttransplant outcomes.

The measurement and statistical advantages of using factors, or latent variables, considerably enhances the ability to study behavioral and physiological relationships. However, the advantage of better reliability and improved power is gained at the expense of a more straightforward clinical application that would be available with observed individual test scores (for example, one cannot create a clinical nomogram based on latent variables). With respect to information about the specific test, we should point out that subsidiary analyses in the Mplus package showed that the individual tests were not related to the pulmonary measures over and above their corresponding latent variable. That is, the pulmonary and demographic variables predicted the shared variance among the neurocognitive tests shared, but did not predict the unshared, or unique, variance. We note that the generalizability of our study may also have been limited. Because our sample consists only of patients awaiting lung transplantation, the extent to which our results may generalize to all patients with end-stage lung disease or less severe pulmonary conditions cannot be determined. Patients must undergo medical and psychologic screening before listing for lung transplant, with dementia deemed a contraindication to the surgery. Therefore, listed patients may show less cognitive impairment than others with end-stage lung disease. On the other hand, because patients are less likely to be denied for listing as a result of cognitive deficits as long as a caregiver can assume responsibility for all medical care, and because patients awaiting lung transplant are likely to have the most severe lung disease, it is possible that patients awaiting transplant may actually exhibit greater cognitive impairment than the general population of patients with lung disease.

In summary, further studies are needed to characterize the cognitive functioning of patients awaiting lung transplant to confirm our findings of significant cognitive deficits related to pulmonary gas exchange and exercise tolerance in this population. Additional data, which the INSPIRE trial hopes to provide, are also needed to determine whether lung transplantation, by improving pulmonary status, could also improve cognitive functioning in patients with end-stage lung disease.

Back to Top | Article Outline


1.American Lung Association. Data and Statistics. Available at: (UNOS web site:
2.Arcasoy SM, Kotloff RM. Medical progress: lung transplantation. N Engl J Med 1999;340:1081–91.
3.Van Trigt P, Davis RD, Shaeffer GD, Gaynor JW, Landolfo KP, Higginbotham MB, Tapson VF, Ungerleider RM. Survival benefits of heart and lung transplantation. Ann Surg 1996;223:576–84.
4.Craven J, The Toronto Lung Group. Psychiatric aspects of lung transplant. Can J Psychiatry. 1990;35:759–64.
5.Parekh PI, Blumenthal JA, Babyak MA, Merrill K, Carney RM, Davis RD, Palmer SM. Psychiatric disorder and quality of life in patients awaiting lung transplantation. Chest 2003;124:1682–8.
6.Singer HK, Ruchinskas RA, Riley KC, Broshek DK, Barth JT. The psychological impact of end-stage lung disease. Chest 2001;120:1246–52.
7.Woodman CL, Geist LJ, Vance S, Laxson C, Jones K, Kline JN. Psychiatric disorders and survival after lung transplantation. Psychosomatics 1999;40:293–7.
8.Dew MA, Switzer GE, DiMartini AF. Psychiatric morbidity and organ transplantation. Curr Opin Psychiatry 1998;11:621.
9.Grant I, Heaton RK, McSweeny AJ, Adams KM, Timms RM. Neuropsychological findings in hypoxemic chronic obstructive pulmonary disease. Arch Intern Med 1982;142:1470–6.
10.Grant I, Prigatano GP, Heaton RK, McSweeny AJ, Wright EC, Adams KM. Progressive neuropsychologic impairment and hypoxemia. Arch Gen Psychiatry 1987;44:999–1006.
11.Prigatano GP, Parsons O, Wright E, Levin DC, Hawryluk G. Neuropsychological test performance in mildly hypoxemic patients with chronic obstructive pulmonary disease. J Consult Clin Psychol 1983;51:108–16.
12.Neubauer JA. Physiological and pathophysiological responses to intermittent hypoxia. J Appl Physiol 2001;90:1593–9.
13.Stuss DT, Peterkin I, Guzman DA, Guzman C, Troyer AK. Chronic obstructive pulmonary disease: effects of hypoxia on neurological and neuropsychological measures. J Clin Exp Neuropsychol 1997;19:515–24.
14.Arciniegas DB, Filley CM. Implication of impaired cognition for organ transplant candidacy. Curr Opin Org Transpl 1999;4:168–80.
15.Deshields TL, McDonough EM, Mannen RK, Miller LW. Psychological and cognitive status before and after heart transplantation. Gen Hosp Psychiatry 1996;18:62S–9S.
16.Nussbaum PD, Goldstein GG. Neuropsychological sequelae of heart transplantation: a preliminary review. Clin Psychol Rev 1992;12:475–83.
17.Temple RO, Putzke JD, Boll T. Neuropsychological performance as a function of cardiac status among heart transplant candidates: a replication. J Percept Mot Skills 2000;91:821–5.
18.Williams MA, LaMarche JA, Smith RL, Fielstein EM, Hardin JM, McGiffin DC, Zorn GL, Kirklin JK, Boll TJ. Neurocognitive and emotional functioning in lung transplant candidates: a preliminary study. J Clin Psychol Med Settings 1997;4:79–90.
19.Crews WD, Jefferson AL, Broshek DK, Barth JT, Robbins MK. Neuropsychological sequelae in a series of patients with end-stage cystic fibrosis: lung transplant evaluation. Arch Clin Neuropsychol 2000;15:59–70.
20.Crews WD, Jefferson AL, Bolduc T, Elliott JB, Ferro NM, Broshek DK, Barth JT, Robbins MR. Neuropsychological dysfunction in patients suffering from end-stage chronic obstructive pulmonary disease. Arch Clin Neuropsychol 2001;16:643–52.
21.Crews WD, Jefferson AL, Broshek DK, Rhodes RD, Williamson J, Brazil AM, Barth JT, Robbins MK. Neuropsychological dysfunction in patients with end-stage pulmonary disease: lung transplant evaluation. Arch Clin Neuropsychol 2003;18:353–62.
22.Rodrigue JR, Kanasky WF, Marhefka SL, Perri MG, Baz M. A psychometric normative database for pre-lung transplantation evaluations. J Clin Psychol Med Settings 2001;8:229–36.
23.Ruchinskas RA, Broshek DK, Crews WD, Barth JT, Francis JP, Robbins MK. A neuropsychological normative database for lung transplant candidates. J Clin Psychol Med Settings 2000;2:107–12.
24.Emery CF, Blumenthal JA. Effects of physical exercise on psychological and cognitive functioning of older adults. Ann Behav Med 1991;13:99–107.
25.Colombe S, Kramer AF. Fitness effects on the cognitive function of older adults: a meta-analytic study. Psychol Sci 2003;14:125–30.
26.Joint Statement of the American Society for Transplant Physicians/American Thoracic Society/European Respiratory Society/International Society for Heart and Lung Transplantation. International guidelines for the selection of lung transplant candidates. Am J Respir Crit Care Med 1998;158:335–9.
27.Mador MJ, Rodis A, Magalang UJ. Reproducibility of Borg scale measurements of dyspnea during exercise in patients with COPD. Chest 1995;107:1590–7.
28.Marcelo Velloso PT, Garcia Stella S, Cendon S, Silva AC, Jardim JR. Metabolic and ventilatory parameters of four activities of daily living accomplished with arms in COPD patients. Chest 2003;123:1047–53.
29.McKone EF, Barry SC, FitzGerald MX, Gallagher CG. Reproducibility of maximal exercise ergometer testing in patients with cystic fibrosis. Chest 1999;116:363–8.
30.Wechsler D. Wechsler Adult Intelligence Scale (WAIS-R) Manual. New York: The Psychological Corp; 1981.
31.Reitan RM. Validity of the trailmaking test as an indicator of organic brain damage. Percept Motor Skills 1958;8:271–6.
32.Golden CJ. Stroop Color Word Test. Chicago: Stoelting; 1978.
33.Lezak MD. Neuropsychological Assessment, 3rd ed. New York: Oxford University Press; 1995.
34.Ruff RM, Light, RH, Parker, SB, Levin, HS. Benton Controlled Oral Word Association test: reliability and updated norms. Arch Clin Neuropsychol 1996;11:329–38.
35.Gladsjo JA, Miller SW, Heaton RK. Norms for letter and category fluency: demographic corrections for age, education, and ethnicity. Lutz, FL: Psychological Assessment Resources, Inc; 1999.
36.Ruff RM, Allen C. Ruff 2 & 7 Selective Attention Test. FL: Psychological Assessment Resources, Inc; 1995.
37.Wechsler D. Wechsler Memory Scale–Revised Manual. San Antonio: The Psychological Corp; 1987.
38.Russell EW. A multiple scoring method for the assessment of complex memory functions. J Consult Clin Psychol 1975;43:800–9.
39.Muthén LK, Muthén BO. Mplus User’s Guide. 1998–2004, 3rd ed. Los Angeles: Muthén & Muthén.
40.Kozora E, Filley CM, Julian LJ, Collum CM. Cognitive functioning in patients with chronic obstructive pulmonary disease and mild hypoxemia compared with patients with mild Alzheimer disease and normal controls. Neuropsychiatry Neuropsychol Behav Neurol 1999;12:178–83.
41.Etnier J, Johnston R, Dagenbach D, Pollard RJ, Rejeski WJ, Berry M. The relationship among pulmonary function, aerobic fitness, and cognitive functioning in older COPD patients. Chest 1999;116:953–60.
42.Colcombe S, Kramer AF. Fitness effects on the cognitive function of older adults: a meta-analytic study. Psychol Sci 2003;14:125–30.
43.Emery CF, Schein RL, Hauck ER, MacIntyre NR. Psychological and cognitive outcomes of a randomized trial of exercise among patients with chronic obstructive pulmonary disease. Health Psychol 1998;17:232–40.
44.Etnier J, Berry M. Fluid intelligence in an older COPD sample after short- or long-term exercise. Med Sci Sport Exerc 2001;33:1620–8.

1The time interval between the neurocognitive testing and the pulmonary function tests was usually <30 days. We analyzed our data using only the subset of patients who underwent testing within 90 days of measurement of medical variables and results were consistent with our overall results. Cited Here...


lung transplantation; neurocognitive assessment; cognitive impairment; exercise capacity

Copyright © 2005 by American Psychosomatic Society