SEARCH

SEARCH BY CITATION

Keywords:

  • Childhood Cancer Survivor Study;
  • neurocognitive;
  • sleep;
  • fatigue;
  • vitality

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

BACKGROUND:

Long-term survivors of childhood cancer are at risk for fatigue, sleep problems, and neurocognitive impairment, although the association between these outcomes has not been previously examined.

METHODS:

Outcomes were evaluated in 1426 survivors from the Childhood Cancer Survivor Study using a validated Neurocognitive Questionnaire. Relative risks for neurocognitive impairment were calculated using demographic and treatment factors, and survivors' report on the Functional Assessment of Chronic Illness Therapy-Fatigue, the Short Form-36 Vitality Scale, the Pittsburgh Sleep Quality Index, and the Epworth Sleepiness Scale.

RESULTS:

Neurocognitive impairment was identified in >20% of survivors, using sibling-based norms for comparison. Multivariate logistic regression models revealed that fatigue (risk ratio [RR], 1.34; 95% confidence interval [CI], 1.13-1.59), daytime sleepiness (RR, 1.68; 95% CI, 1.55-1.83), poor sleep quality (RR, 1.23; 95% CI, 1.01-1.49), and decreased vitality (RR, 1.75; 95% CI 1.33-2.30) were all associated with impaired task efficiency. Likewise, fatigue (RR, 1.77; 95% CI, 1.23-2.55), sleepiness (RR, 1.38; 95% CI, 1.14-1.67), and decreased vitality (RR, 3.08; 95% CI, 1.98-4.79) were predictive of emotional regulation problems. Diminished organization was associated with increased sleepiness (RR, 1.80; 95% CI, 1.31-2.48) and decreased vitality (RR, 1.90; 95% CI, 1.37-2.63). Impaired memory was associated with poor sleep quality (RR, 1.45; 95% CI, 1.19-1.76), increased sleepiness (RR, 2.05; 95% CI, 1.63-2.58), and decreased vitality (RR, 2.01; 95% CI, 1.42-2.86). The impact of fatigue, sleepiness, sleep quality, and vitality on neurocognitive outcomes was independent of the effects of cranial radiation therapy, steroids and antimetabolite chemotherapy, sex, and current age.

CONCLUSIONS:

Neurocognitive function in long-term survivors of childhood cancer appears particularly vulnerable to the effects of fatigue and sleep disruption. These findings suggest sleep hygiene should be emphasized among survivors, as it may provide an additional mechanism for intervention to improve neurocognitive outcomes. Cancer 2011;. © 2011 American Cancer Society.

As enhanced medical treatments have contributed to an increase in the number of adult survivors of childhood cancer, research has focused on the medical and functional late effects of cancer therapy.1-4 Common functional late effects that significantly impact quality of life include neurocognitive impairment and fatigue.

Survivors of childhood cancer are at increased risk for neurocognitive impairment, as either a direct or indirect result of central nervous system (CNS) treatment. Cranial radiation treatment has long been associated with neurocognitive late effects,1-3 although antimetabolite chemotherapy and corticosteroids have also been implicated as potential contributors to poor outcomes.5 Neurocognitive impairment is among the most common late effects experienced by long-term survivors of pediatric acute lymphoblastic leukemia (ALL), with 20% to 40% of patients exhibiting deficits in 1 or more areas of function.3, 4 Adult survivors of pediatric CNS tumors also experience high rates of neurocognitive difficulties, which frequently include slow processing speed, inattention, and memory impairment.6

Reduced sleep quality and fatigue are also reported to impact neurocognitive functioning. Specifically, fatigue is associated with impairments in processing speed, attention, and memory functions.7 Among adults diagnosed with chronic fatigue syndrome, slow processing speed, impaired working memory, and poor memory and learning of new information have been reported.8, 9 Similarly, the presence of significant fatigue has been associated with poor neuropsychological functioning in adults with acute medical conditions.10-13

Fatigue and reduced sleep quality are common late effects experienced by long-term survivors of childhood cancer.14 It is estimated that up to 45% of the general population report symptoms of fatigue,15 whereas rates as high as 90% have been reported in cancer survivors.16-19 For example, ALL survivors typically report significant fatigue many years after completion of treatment,20, 21 whereas survivors of Hodgkin lymphoma report more fatigue than their siblings and population-based controls.14, 22, 23 Survivors of childhood brain tumor report more physical health difficulties than other pediatric cancer survivors, with poorer physical functioning associating with greater fatigue.24 In a comprehensive assessment of fatigue, sleepiness, and sleep quality, slightly higher rates of fatigue were reported among adult survivors of childhood cancer compared with a sibling cohort, although no significant differences were reported between siblings and survivors in regard to sleepiness or sleep quality.14

Although the differences between rates of fatigue and sleep problems among cancer survivors and siblings may be small, differential sensitivity of fatigue and poor sleep quality on neurocognitive functions within the different cohorts may exist. Sleep is important for neural recovery after brain injury.25 Furthermore, sleep deprivation among individuals with brain injury exacerbates the degree of neurocognitive impairment.26 Thus, although the rate of fatigue and poor sleep quality in survivors of childhood cancer may or may not be clinically significant, the impact of fatigue and sleep loss on neurocognitive performance may be more salient in the survivors who are at risk for brain injury after neurotoxic cancer therapy. Little is known of the association between fatigue or sleep quality and neurocognitive outcomes among adult survivors of childhood cancer. The primary objective of the current study is to elucidate this association. It was hypothesized that survivors with higher ratings of fatigue and sleepiness will report more problems with processing speed, attention, and memory functions, independent of the known effect of cranial radiation therapy on these neurocognitive outcomes.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Participants

Participants for these analyses included 1426 adult survivors from the Childhood Cancer Survivor Study and 384 sibling controls. The Childhood Cancer Survivor Study is a retrospective cohort study designed to investigate long-term medical, psychosocial, and functional outcomes of survivors of childhood cancer. Details of this study have been previously reported.27, 28 Briefly, eligible participants were treated for 1 of 8 childhood cancer diagnoses at 26 institutions between 1970 and 1986 when younger than 21 years of age. Cohort entry was limited to those individuals who survived for at least 5 years after their original diagnosis. The human subjects committee at each of the collaborating institutions approved the study protocol before participant enrollment. Participants provided informed consent for questionnaires and medical record abstraction. Study participants have completed multiple questionnaires since their original enrollment (the full survey questionnaires are available at http://ccss.stjude.org). The study population for the current analyses included all cancer survivors and a cohort of randomly selected sibling controls who completed both the 2003 Follow-up survey (conducted between 2002 and 2006) and the Fatigue and Sleep Survey (conducted between 2002 and 2004). The 2003 Follow-up survey was collected from 6739 survivors and 384 randomly select siblings between 2002 and 2006. The fatigue/sleep survey involved a separate mailing to 2645 randomly selected survivors and the siblings, and was collected between 2002 and 2004. A total of 1426 survivors and 384 siblings completed both surveys. Consistent with a prior published report on the prevalence of fatigue and sleep problems in the Childhood Cancer Survivor Study cohort,14 survivors diagnosed with and treated for Hodgkin lymphoma were over-represented in the current sample, given their apparently higher rates of reported fatigue and sleep problems.

In addition to the specific measures described below, the 2003 Follow-up survey included report of the following categories of information: medical care over the prior 2 years; medical screening tests; health behaviors; insurance coverage; family history of cancer, birth complication, and hereditary conditions; pregnancy and offspring; dental and bone health; current medication use; cancer relapse or recurrence; and current income.

Measures

Childhood Cancer Survivor Study Neurocognitive Questionnaire

Self-report of neurocognitive skills was assessed using the Childhood Cancer Survivor Study Neurocognitive Questionnaire, a measure that has been previously validated in adult survivors of childhood cancer.29 This 25-item questionnaire requires ratings of neurocognitive problems on a 3-point Likert scale (0 = “never a problem” to 2 = “often a problem”), and has been identified as comprising 4 primary factors of neurocognitive outcome: Task Efficiency, Emotional Regulation, Organization, and Memory. These factors provide measures of executive functioning (ie, Emotional Regulation and Organization factors), attention and processing speed (ie, Task Efficiency factor), and working and long-term memory (ie, Memory factor). Impaired performance was defined as a score falling below the 10th percentile based on sibling age and sex-adjusted norms.

Vitality

Survivors also completed the Medical Outcomes Short Form-36 (SF-36) to evaluate health-related quality of life.30 Within the SF-36, items comprising the Vitality Scale were used as a primary predictor. The 4 items comprising this measure are rated on a 6-point Likert scale (0 = “all of the time” to 5 = “none of the time”), and measure the degree of feeling energetic and full of life versus feeling tired and worn out. Poor vitality was defined as a T-score falling below 40 based on a national standardization sample.

Fatigue

The fatigue subscale of the Functional Assessment of Chronic Illness Therapy-Fatigue was also used as a primary predictor. This is a measure of the physical and functional consequences associated with fatigue.31 The 13 items comprising this measure are scored on a reverse 4-point Likert scale (4 = “not at all” to 0 = “very much”). Greater fatigue is represented by lower scores.

Sleep quality

The Pittsburgh Sleep Quality Index (PSQI) was administered to assess sleep quality over the month before survey completion.32 Several sleep quality components are measured by the PSQI, including subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbance, use of sleeping medications, and daytime dysfunction. For current analyses, the overall score comprised of all 19 items on the PSQI, which are scored on a 4-point Likert scale (0 = “not during the past month” to 3 = “three or more times a week”), was used as a measure of sleep quality, with higher scores indicating poorer quality of sleep.

Daytime sleepiness

Symptoms of daytime sleepiness were measured using the Epworth Sleepiness Scale (ESS).33 By using a 4-point Likert scale (0 = “would never doze” to 3 = “high chance of dozing”), participants rated the likelihood of falling asleep during routine situations, with higher scores indicative of greater daytime sleepiness.

The PSQI and Functional Assessment of Chronic Illness Therapy-Fatigue were originally standardized using a sample that was older on average than the age of the current survivor cohort (ie, 59.9 years for the PSQI and 45.7 for the Functional Assessment of Chronic Illness Therapy control samples, respectively, 35.9 years for the current cancer survivor cohort). In addition, the sample size for the control groups used in development of the PSQI (n = 52) and ESS (n = 30) were relatively small. Thus, to dichotomize the scales and identify levels of clinical problems, we compared survivors to an age- and sex-matched sibling cohort group (n = 384). Survivors scoring ≤10th percentile on the Functional Assessment of Chronic Illness Therapy-Fatigue or ≥10th percentile on the PSQI and ESS were classified as having significant fatigue, poor sleep quality, or daytime sleepiness, respectively. These cutoffs are consistent with the procedure used in a prior published report.14

Covariates

The Brief Symptom Inventory-18 was used as a measure of global distress.34 Given the association between fatigue and emotional distress, subscales for anxiety, depression, and somatization were used as covariates. Sex-specific scores were calculated based on standardized normative values, and scores <10th percentile were classified as demonstrating a clinical level of emotional distress. Only 48.2% of those survivors reporting significant depressive symptoms reported using antidepressant medication. Given the impact use of this medication has on fatigue and neurocognitive function,35 and its limited multicollinearity with depressive symptoms, antidepressant medication use was included as a covariate in modeling outcomes. Cancer treatment was also used as a covariate. Cranial radiation therapy was categorized by radiation dose intensity, resulting in 3 groups: None, Low (0 to <20 grays [Gy]), or High (≥20 Gy). The low-dose radiation exposure group included those with scatter exposure from radiation treatment of adjacent body areas. Treatment with CNS chemotherapy was included using dichotomous variables (yes/no) for antimetabolites and for corticosteroids. Additional covariates included age, sex, household income (<$20,000, ≥$20,000), and body mass index (BMI; nonobese, BMI <30; obese, BMI >30]).

For the primary predictors of fatigue, sleepiness, and sleep quality and for the neurocognitive outcomes, where higher ratings are suggestive of pathology, we used a cutoff score of great than or equal to that seen in the highest 10% of siblings. This is consistent with cutoffs used in the current literature.14 For the SF-36 Vitality score, where lower ratings are suggestive of pathology, we used a cutoff score ≤1 standard deviation from the population mean. This cutoff is also consistent with current reports in the literature and with the recommendation in the test manual and interpretation guide.30, 36

Statistical Analyses

Descriptive statistics were calculated to evaluate frequency, percentage, mean, standard deviation, and range of all outcomes, predictors, and covariates used in the analyses. Pearson correlations were conducted to evaluate associations between each variable included in the analyses. To address clinical implications for each of the 4 outcome variables on the Childhood Cancer Survivor Study Neurocognitive Questionnaire (Task Efficiency, Emotional Regulation, Organization, and Memory), performance was classified into yes/no impairment based on comparison to sibling norms, and a multivariate logistic regression model was then constructed. Fatigue, vitality, and sleep variables were also classified into yes/no problem level, as described above. Backward selection was performed for each of the models using PROC GENMOD with binomial distribution and log link. These analyses began with the full model, which included all of the predictors (vitality, fatigue, sleep quality, and daytime sleepiness) and covariates (anxiety, depression, somatization, cranial radiation therapy, antimetabolite chemotherapy, corticosteroid therapy, antidepressant medication, current age, age at diagnosis, sex, household income, and BMI). Then the least significant variables (largest P value) were excluded 1 at a time until all the variables left in the model were significant (P < .05). Factors having >2 levels were kept or removed based on the P value of the Likelihood Ratio Test. Given the association between current age and fatigue, current age was left in each model regardless of its significance. Finally, relative risks (RRs) and 95% confidence intervals (CIs) of predictors and covariates left in the final model for each outcome were calculated. Given the large sample size, this approach was perceived as appropriate for demonstrating increased risk of neurocognitive impairment associated with the primary predictors and covariates. Analyses were conducted separately for type of cancer diagnosis (eg, ALL, CNS brain tumor, Hodgkin disease, other) and type of treatment (eg, CNS radiation, CNS chemotherapy) because of multicolinearity between these variables. Because prior research has focused on treatment-related late effects associated with cancer, results from the models including treatment effects will be discussed in the following sections.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Table 1 presents descriptive statistics for survivor and sibling demographics and survivor treatment characteristics. As indicated above, survivors of Hodgkin lymphoma were oversampled to represent a majority of the cohort, given the increased rates of fatigue reported in this group. Roughly 78% of the survivor cohort received some level of cranial exposure to radiation, including 54.6% who received low-dose exposure such as scatter from radiation applied to adjacent body areas.

Table 1. Demographic and Treatment Characteristics of the Survivor and Sibling Cohorts
CharacteristicSurvivors, n = 1426Siblings, n = 384
Mean (SD)RangeMean (SD)Range
  • SD indicates standard deviation; CNS, central nervous system; IV, intravenous; IT, intrathecal; CRT, cranial radiation treatment; Gy, grays.

  • a

    Survivors diagnosed with and treated for Hodgkin lymphoma were over-represented in the current sample, given the higher rates of reported fatigue and sleep problems.

Age at diagnosis, y11.9 (5.6)0-21
Time since diagnosis, y24.0 (4.7)16.2-34.3
Current age, y35.9 (7.5)19.2-53.433.7 (8.4)17.8-58.4
 Frequency%Frequency%
Sex    
 Male68448.018147.9
 Female74252.019952.1
Income    
 <$20,000/y1359.5246.3
 $20,000-$39,999/y25017.55815.1
 ≥$40,000/y91964.527171.9
Cancer diagnosis    
 Leukemia20014.0
 CNS tumor21415.0
 Hodgkin lymphomaa76853.9
 Other cancer24417.1
Chemotherapy treatment    
 Alkylators72150.6
 Anthracycline40728.5
 Antimetabolite (IV)26518.6
 Antimetabolite (IT)79555.8
 Corticosteroids53838
 Epipodophyllotoxin453.2
CRT    
 No CRT30921.7
 CRT <20 Gy77954.6
 CRT ≥20 Gy20114.1

The frequency and rates of impairment on the 4 factors from the Childhood Cancer Survivor Study Neurocognitive Questionnaire are presented by cancer diagnosis in Table 2. Survivors of CNS tumors demonstrated the highest rates of neurocognitive problems, although rates of task efficiency problems were also elevated in survivors of leukemia and Hodgkin lymphoma.

Table 2. Neurocognitive Problems in Survivors by Cancer Diagnosis
DiagnosisNo.Task EfficiencyEmotion RegulationOrganizationMemory
No.ImpairedaNo.ImpairedaNo.ImpairedaNo.Impaireda
  • CNS indicates central nervous system.

  • a

    Impairment defined as a performance ≥ a symptom level reported by the highest 10% of sibling controls.

Leukemia2005025%2111%2714%2714%
CNS tumors2149243%2813%4421%5023%
Hodgkin lymphoma76812016%9212%9512%9312%
Other cancers2443514%208%229%167%
Total142629721%16111%18813%18613%

Descriptive statistics for the neurocognitive outcome measures stratified by primary predictors and covariates are provided in Table 3. As is evident in this table, problems in task efficiency, emotional regulation, organization, and memory were more common in those survivors reporting higher fatigue, poor sleep quality, daytime sleepiness, and low vitality. Survivors reporting use of antidepressants or displaying emotional distress (ie, depression, anxiety, somatization) were more likely to demonstrate neurocognitive problems.

Table 3. Descriptive Statistics for Primary Predictors and Covariates, N = 1426
 No.Task EfficiencyEmotional RegulationOrganizationMemory
Mean (SD)ImpairedaMean (SD)ImpairedaMean (SD)ImpairedaMean (SD)Impaireda
  • SD indicates standard deviation; BMI, body mass index.

  • a

    Impairment defined as a performance ≥ a symptom level reported by the highest 10% of sibling controls.

Siblings3842.9 (3.1)10.0%2.0 (1.6)10.0%1.6 (1.6)10.0%1.8 (1.8)10.0%
Survivors14263.8 (3.8)20.8%2.2 (1.7)11.3%1.7 (1.6)13.2%2.1 (2.0)13.0%
Primary predictors         
 Fatigued         
  Yes1977.3 (4.6)53.3%3.8 (1.7)39.6%2.4 (1.8)24.4%3.7 (2.4)37.1%
  No12293.2 (3.3)15.6%2.0 (1.6)6.8%1.5 (1.6)11.4%1.8 (1.9)9.2%
 Poor sleep quality         
  Yes2485.9 (4.4)40.7%3.3 (1.8)28.2%2.2 (1.8)23.0%3.2 (2.4)29.4%
  No11783.3 (3.5)16.6%2.0 (1.6)7.7%1.5 (1.5)11.1%1.9 (1.9)9.6%
 Daytime sleepiness         
  Yes1956.2 (4.5)43.1%3.2 (1.8)26.7%2.3 (1.8)25.1%3.4 (2.4)32.3%
  No12313.4 (3.5)17.3%2.1 (1.6)8.9%1.5 (1.6)11.3%1.9 (1.9)10.0%
 Decreased vitality         
  Yes5675.3 (4.3)33.2%3.0 (1.8)22.6%2.0 (1.8)19.8%2.8 (2.3)22.8%
  No8592.7 (3.1)12.7%1.7 (1.5)3.8%1.3 (1.5)8.8%1.6 (1.8)6.6%
Covariates         
 BMI ≥30 [obese]         
  Yes2813.9 (3.8)22.8%2.3 (1.7)12.8%1.7 (1.7)15.3%2.2 (2.1)14.9%
  No11093.7 (3.8)19.9%2.2 (1.7)11.1%1.6 (1.6)12.6%2.1 (2.1)12.4%
 Antidepressant use         
  Yes2276.2 (4.6)45.4%3.4 (1.7)28.6%2.3 (1.9)22.9%3.2 (2.4)28.2%
  No11843.3 (3.5)16.1%2.0 (1.6)8.0%1.5 (1.5)11.5%1.9 (1.9)10.1%
 Depression         
  Yes1397.8 (4.9)60.4%4.1 (1.6)45.3%2.6 (1.9)30.9%3.8 (2.4)38.1%
  No12873.3 (3.4)16.6%2.0 (1.6)7.6%1.5 (1.5)11.3%1.9 (1.9)10.3%
 Anxiety         
  Yes1007.0 (4.2)54.0%4.3 (1.4)50.0%2.4 (1.8)26.0%3.8 (2.6)39.0%
  No13263.5 (3.7)18.3%2.1 (1.6)8.4%1.5 (1.6)12.2%2.0 (2.0)11.1%
 Somatization         
  Yes2136.2 (4.3)41.3%3.3 (1.7)29.6%2.3 (1.7)24.4%3.4 (2.3)28.6%
  No12133.3 (3.6)17.2%2.0 (1.6)8.1%1.5 (1.6)11.2%1.9 (1.9)10.3%

Multivariate Prediction of Neurocognitive Impairment

Table 4 presents the RRs for impairment on the 4 neurocognitive outcomes. Multivariate logistic regression models revealed significant associations between all of the primary predictors and impaired task efficiency. Specifically, decreased vitality (RR, 1.75; 95% CI, 1.33-2.30), increased fatigue (RR, 1.34; 95% CI, 1.13-1.59), increased daytime sleepiness (RR, 1.68; 95% CI, 1.55-1.83), and poor sleep quality (RR, 1.23; 95% CI, 1.01-1.49) predicted impaired task efficiency. Among treatment variables remaining in the final model, survivors in the high-dose radiation group reported significantly greater impairments in task efficiency (RR, 1.46; 95% CI, 1.13-1.89). Decreased vitality (RR, 3.08; 95% CI, 1.98-4.79), increased fatigue (RR, 1.77; 95% CI, 1.23-2.55), and increased daytime sleepiness (RR, 1.38; 95% CI, 1.14-1.67) all significantly predicted impaired emotional regulation. Survivors in the high-dose radiation group reported significantly more impairment in emotion regulation (RR, 1.84; 95% CI, 1.19-2.84). Decreased vitality (RR, 1.90; 95% CI, 1.37-2.63) and increased daytime sleepiness (RR, 1.80; 95% CI, 1.31-2.48) both predicted impaired organization. No treatment variables remained in the final model predicting organization problems. Decreased vitality (RR, 2.01; 95% CI, 1.42-2.86), increased daytime sleepiness (RR, 2.05; 95% CI, 1.63-2.58), and poor sleep quality (RR, 1.45; 95% CI, 1.19-1.76) all significantly predicted impaired memory. Among treatment variables, survivors in the high-dose radiation group reported significantly greater memory impairments (RR, 2.14; 95% CI, 1.46-3.13).

Table 4. Relative Risk for the Prediction of Neurocognitive Outcome Comparing Adult Survivors of Childhood Cancer to National Norms, N = 1426
 Task EfficiencyEmotional RegulationOrganizationMemory
RR95% CIPRR95% CIPRR95% CIPRR95% CIP
  1. RR indicates relative risk; CI, confidence interval; Tx, treatment.

  2. Variables listed were included in the final model. — indicates variables dropped from the final model through the backward selection process.

Primary predictors            
 Vitality1.751.33-2.30<.00013.081.98-4.79<.00011.901.37-2.63.00012.011.42-2.86<.0001
 Fatigue1.341.13-1.59.00091.771.23-2.55.002
 Daytime sleepiness1.681.55-1.83<.00011.381.14-1.67.00081.801.31-2.48.00032.051.63-2.58<.0001
 Sleep quality1.231.01-1.49.041.451.19-1.76.0002
Behavioral covariates            
 Depression1.841.26-2.68.0021.561.09-2.23.015
 Anxiety1.661.20-2.30.002
 Somatization1.431.17-1.75.0005
 Antidepressant1.561.46-1.67<.00011.701.37-2.11<.0001
Cranial radiation Tx            
 None1.00  1.00  1.00  1.00  
  Low dose0.910.69-1.20.511.200.78-1.84.401.020.69-1.52.91
  High dose1.461.13-1.89.00381.841.19-2.84.0062.141.46-3.13<.0001
Sex        
 Male1.00  1.00  1.00  1.00  
 Female1.451.22-1.73<.0001
Current age0.990.98-1.00.0411.011.00-1.02.220.990.97-1.01.371.000.99-1.01.94
Household income        
 <$20,000/y1.501.39-1.62<.00011.761.26-2.47.001
 ≥$20,000/y1.00  1.00  1.00  1.00  

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Although the relationships between fatigue, sleep, and quality of life have been previously reported,14 this study is the first to extend the adverse impact of these factors to compromised neurocognitive functioning among adult survivors of childhood cancer. The RR for neurocognitive impairment associated with fatigue and sleep disturbance was roughly equivalent to that seen with high-dose radiation. When describing impairment observed among childhood cancer survivors, previous models of cognitive functioning have emphasized the direct insults associated with cancer treatment, whereas these findings suggest the additive contributions of less direct pathways (such as poor sleep and fatigue).

It is interesting to note that more predictors were significantly associated with task efficiency than any other outcome variable. Specifically, all 4 of the primary variables related to fatigue, vitality, sleep quality, and daytime sleepiness predicted impaired task efficiency, as did use of antidepressant medication, high-dose cranial radiation therapy, current age, and household income. Questions within this scale are generally related to attention and processing speed, which are often areas of vulnerability associated with late effects of cancer treatments, traumatic brain injury, pain, and psychological disorders such as depression. It is important to highlight that the impact of fatigue, vitality, daytime sleepiness, and poor sleep quality on task efficiency was independent of the significant effects of cranial radiation therapy, and use of antidepressant medication.

Although female sex was not a significant predictor of impaired task efficiency, organization problems or memory problems, it did contribute to the risk for problems with emotional regulation. Recent research demonstrates sex specificity to the attention problems detected in cancer survivors,37 although current measures may not be specific enough to demonstrate this differential pattern. Emotional regulation is associated with increased lability and tendencies for emotional explosiveness,38 and the current data suggest that female survivors are more prone to report these symptoms. Conversely, men may be less likely to acknowledge problems with emotional regulation or they may truly experience better emotional control. Of note, increased depression and anxiety were also identified risk factors for problems with emotional regulation. Problems with emotional regulation suggest increased emotional lability and/or explosiveness, and do not require the content of the lability to include either negative or positive emotions. However, the increased risk of poor emotional regulation found in those survivors reporting significant symptoms of depression and/or anxiety suggests that the increased lability or explosiveness is likely to involve negative emotions such as sadness, frustration, and nervousness.

Poor sleep quality was identified as being significantly associated with attention and processing speed problems (Task Efficiency) and memory problems. Rapid eye movement (REM) sleep is particularly important for healthy memory functions,39 and specific disruption of REM cycles has been associated with memory impairment in aging adult populations.40 Clearly, the PSQI does not permit identification of separate REM sleep stages. However, the association between the PSQI and memory problems is consistent with the literature on REM sleep and its role in memory consolidation. As such, REM deprivation may be part of the pathway underlying this association. In addition, sleep disorders in children often mimic symptoms of attention-deficit/hyperactivity disorder, including primary problems with inattention.41 Our data would suggest that although good sleep quality may not be necessary for emotional regulation and organization skills, it is related to attention and memory functions.

Decreased vitality and increased fatigue were also strongly associated with neurocognitive impairment, even when controlling for sleep quality. Although fatigue and low vitality may result from poor sleep quality, physical and mental fatigue can also be associated with metabolic or neuroendocrine dysfunction,42, 43 as well as cardiovascular disease.44 Because survivors of childhood cancer are at increased risk for symptoms related to these medical conditions,44, 45 the role they play in neurocognitive outcomes should be examined. We are in the process of conducting just such an investigation.

Because many quality of life outcomes are mediated by neurocognitive functions, the identification of modifiable risk factors that are associated with cognitive outcomes is a valuable and important target for intervention. In the case of childhood cancer survivors, results from this study highlight the importance of considering interventions to improve sleep hygiene and/or physical fitness as a nonpharmacological mechanism for improving neurocognitive functioning. The direct relationships between poor sleep and fatigue with self-reported neurocognitive outcomes were evidenced in our study. However, the more applied consequences of insomnia and fatigue extend to specific real world outcomes, including serious accidents, illness-related restricted activities, and psychiatric disorders, along with decreased work productivity.46 Interventions to improve sleep quality not only have the potential to enhance neurocognitive functioning, but may also have beneficial effects that extend to symptoms of depression, anxiety, and somatization, along with fatigue and vitality. Research suggests that the first-line interventions to improve sleep in this population should be behavioral in nature, as the use of hypnotics may be contraindicated because of increased risk for medical problems (including pulmonary, hepatic, and renal disease) secondary to cancer treatment.47, 48

Future research examining factors impacting neurocognitive functioning in survivors could be improved by the utilization of standardized direct performance measures to supplement self-report questionnaires like the Childhood Cancer Survivor Study-Neuroticism Scale Questionnaire. For example, a computerized continuous performance test would provide separate indices of processing speed and sustained attention, constructs that are combined in the Childhood Cancer Survivor Study Neurocognitive Questionnaire Task Efficiency scale. Although reliance on self-report measures does not represent the gold standard of a comprehensive neurocognitive direct performance assessment, self-report has been shown to be a valid and convenient method of measuring neurocognitive functioning.29 However, direct performance measures may provide additional details and identify more specific neurocognitive processes impacted by sleep disruption. For example, use of the California Verbal Learning Test would permit separation of poor memory encoding from retrieval deficits, which may suggest reduced functioning in hippocampal structures. The use of polysomnographic methods would also be advantageous and permit the characterization of sleep architecture, while at the same time allowing for the examination of affected sleep states and their impact on neurocognitive outcomes.

In the healthy population, disrupted cycles on the sleep-wake continuum, such as REM and slow wave sleep, adversely affect memory consolidation, perceptual and motor learning, and cognitive flexibility,49, 50 but the interaction between disrupted sleep cycles and childhood cancer treatment on neuropsychological outcomes has not yet been examined. Understanding the causal means by which factors such as reduced vitality, fatigue, daytime sleepiness, or sleep quality affect neurocognitive functioning will be important for improving the quality of life among survivors of childhood cancer. Clarity regarding the timing of neurocognitive compromise and the anatomical systems affected is also needed so that future therapies can reduce neurocognitive impairment or preserve further insult. Several of our significant findings related to covariate predictors also substantiate the need for future research to use comprehensive approaches that include the consideration of sociodemographic and psychological factors in predicting neurocognitive outcomes. Through this approach, we will maximize the likelihood of developing the most relevant models of cognitive functioning, which will have the greatest impact on quality of life outcomes among survivors of childhood cancer.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

This work was supported by the National Cancer Institute grant CA 55727 (L.L.R.). Support to St. Jude Children's Research Hospital was also provided by the National Cancer Institute Cancer Center Support grant (CA 21765) and by the American, Lebanese, Syrian Associated Charities.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES