Journal of Pediatric Psychology Advance Access published online on October 25, 2007
Journal of Pediatric Psychology, doi:10.1093/jpepsy/jsm080
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Factors Related to Changes in Cognitive, Educational and Visual Motor Integration in Children who Undergo Hematopoietic Stem Cell Transplant
1Department of Psychology, Division of Hematology/Oncology, The Hospital for Sick Children, 2University of Toronto, 3Child Health Evaluative Sciences, The Hospital for Sick Children, and 4Centre for Community Child Health Research
All correspondence concerning this article should be addressed to Maru Barrera, PhD, Department of Psychology, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada. E-mail: maru.barrera{at}sickkids.ca
 |
Abstract |
|---|
 Top Abstract MethodResultsDiscussionAcknowledgmentsReferences |
|---|
Objectives Investigate cognitive, educational, and perceptual motor skills up to 2 years posttransplant of pediatric hematopoietic progenitor cell transplantation (HPCT) survivors and their correlates. Methods Survivors were assessed at baseline, 12, and 24 months after transplant. Results Performance IQ improved over time and was negatively related to maternal depression. Full IQ and educational outcomes were positively related to child's age and mother's age. Low depression scores were associated with high Verbal IQ one and 2 years post-HPCT, and with high visual motor scores 2 years post-HPCT. Poor educational outcomes were related to increased time since diagnosis. Two years post-HPCT, Performance IQ and Processing Speed were above the norm values whereas arithmetic and motor scores were below. Conclusions Pediatric HPCT survivors do better cognitively than educationally. Maternal age and depression, child's age, and time since diagnosis are critical factors for these outcomes.
Key words: cancer; cognitive functioning; educational achievement; pediatric hematopoietic progenitor cell transplantation..
It is now well established that pediatric blood transplant protocols, including bone marrow, stem cell, or hematopoietic progenitor cell transplantation (HPCT), are aggressive yet life saving procedures for patients diagnosed with some cancers, blood, and metabolic disorders. Nevertheless, HPCT is associated with high mortality and morbidity. Evidence is growing regarding physical (Gassas, Sung, Saunders, & Doyle, 2006
) and psychological adverse effects of the procedure (Cool, 1996; Kramer, Crittenden, DeSantes, & Cowan, 1997; McGuire, Sanders, Hill, Buckner, & Sullivan, 1991; Phipps et al., 1995; Phipps, Dunavant, Srivastava, Bowman, & Mulhern, 2000; Smedler, Nilsson, & Bolme, 1995). To date, however, results regarding global and specific cognitive outcomes are inconclusive, and educational and perceptual motor outcomes are also not well understood. Additionally, research in this field has not been driven by theory with little being known about how child characteristics and family factors may contribute to these outcomes as they interact with clinical factors. This longitudinal study aims to address these gaps. Specifically, to better understand the cognitive, academic, and perceptual motor outcomes of pediatric HPCT, we conducted a comprehensive study driven by a conceptual framework that considers the contribution of clinical, child, and family factors to cognitive outcomes, prior to, 1 and 2 years post-HPCT. It also considers the contribution of these factors to educational and perceptual motor outcomes 2 years post-HPCT.
Previous studies have reported some cognitive deficits related to treatment, including cranial radiation therapy (CRT) or intrathecal methotrexate, in children treated for acute lymphoblastic leukemia (ALL) or for some brain tumors (Anderson, Godber, Smibert, Weiskop, & Ekert, 2000; Brown et al., 1992; Copeland, deMoor, Moore, & Ater, 1999; Espy et al., 2001; Mabbott et al., 2005; Mennes et al., 2005; Moore, 2005; Raymond-Speden, Tripp, Lawrence, & Holdwaway, 2000; Spiegler, Bouffet, Greenberg, Rutka, & Mabbott, 2004). Conditioning protocols for pediatric HPCT may also have detrimental effects on cognitive outcomes. These protocols vary with diagnoses, but can involve intrathecal chemotherapy and total body radiation (TBR) ranging from a single dose (300 cGy) on the day of transplant, to several doses (i.e., fractionated) totaling 1,200 cGy (or more). Prior to HPCT, children treated for ALL who received CRT, had lower IQ scores than children who were not treated with CRT (Cool, 1996).
After the HPCT procedure, some neurocognitive deficits have been reported in retrospective and prospective studies. Early retrospective studies examining cognitive and academic outcomes over a long posttransplant interval found cognitive deficits in children who had received CRT at a younger age (McGuire et al., 1991; Smedler et al., 1995) that appeared to increase with time (Smedler et al., 1995). The few prospective studies available report little if any general negative impact of HPCT treatment on cognition (Phipps et al., 1995), with the exception of CRT effects (Cool, 1996; Kramer et al., 1997). In the most recent prospective study examining cognitive, perceptual motor, and academic performance, no significant changes on global measures of IQ, perceptual motor, and academic ability were found from pre-transplant to 1 year post-HPCT or from 1 to 3 years post-HPCT. In addition, no relationship between TBR and IQ was found (Phipps et al., 2000). As previously reported (Cool, 1996; Kramer et al., 1997; McGuire et al., 1991), younger children at HPCT showed more IQ decline compared to older children. In these studies specific cognitive indices were not examined, likely because the complete set of IQ subtests were not administered, and familial factors were not investigated. Thus, the current literature suggests that cognitive, educational, and perceptual motor outcomes of survivors are within the normal range, with the possible exception of the outcomes of younger children. Less is understood about the mechanisms for these outcomes.
This study used the same conceptual model as in earlier studies of children who were diagnosed with cancer (Barrera et al., 2003) and their parents (Barrera, D’Agostino, Gibson, Weksberg, & Malkin, 2004) to examine the potential mechanism for cognitive, educational, and perceptual motor outcomes. Also, as in our previous work, we integrated Baron and Kenny's (1986) and Holmbeck's (1997) conceptualizations of mediators and moderators into the model of the relationship between the outcomes and their covariates. Briefly, this model postulates that the outcomes of childhood cancer, and in this case of HPCT, are a function of the child's characteristics and development and familial factors as they interact with illness and treatment factors. Examining some of the familial factors, in an earlier study, we found that at 6 months post-HPCT, children's quality of life scores were strongly correlated with pre-HPCT level of family cohesion (Barrera, Pringle, Sumbler, & Saunders, 2000). Evidence from the general population suggests that maternal age and depression may be associated with young children's cognitive development; with children of younger mothers or mothers with clinical depression obtaining lower IQ scores (Culp, Culp, Osofsky, & Osofsky, 1991; Fergusson & Woodward, 1999; Hay et al., 2001; Hubbs-Tait, Culp, Culp, & Miller, 2002; Keown, Woodward, & Field, 2001).
In this longitudinal study, we examined global and specific cognitive functioning of HPCT survivors at pre-HPCT, 1and 2 years post-HPCT, as well as their educational and perceptual motor outcomes at 2 years post-HPCT. Global IQ and specific cognitive outcomes, educational and perceptual motor outcomes were the focus of this investigation because the existing findings on these outcomes are still inconsistent across studies. We also examined the relationships between some clinical factors (history of CRT, diagnosis, and time since diagnosis), child's characteristics (age at HPCT, gender), and familial factors (mother's age, education, and distress, family cohesion) and these outcomes. We hypothesized that:
(1a) Cognitive, educational, and perceptual motor outcomes will be within the normal range and (1b) cognitive outcomes of HPCT survivors will not change significantly over time. These predictions are based on previous literature with this population. Based on the conceptual model described earlier and previous literature, we predicted variations in the outcomes depending on clinical variables, child and family factors and the relationship between these factors.
Clinical variables. (2a) Based on Cool's (1996) and Kramer et al.'s (1997) findings, children who received both CRT and TBR will obtain lower cognitive scores compared to children who had no history of cranial radiation; (2b) Children who had been treated for longer periods of time before the HPCT will show poorer cognitive, educational, and motor outcomes. This prediction is based on the well-established fact that experiencing long periods of medical treatment will cause disruption in their lives and in their development.
Child characteristics and family factors and their relationships. (3a) Younger children at HPCT will obtain the lowest scores across assessment times (for cognitive outcomes) or at 2 years post-HPCT (for educational and perceptual motor outcomes), given the fact that they had less opportunities for development and learning prior to the procedure; (3b) Children who had older mothers at HPCT will obtain higher scores across all assessment times for cognitive outcomes or at 2 years post-HPCT for educational and perceptual motor outcomes; also children whose mothers had high depression scores across time will obtain lower cognitive, educational, and motor scores; and family cohesion will have a positive effect on the cognitive, educational, and motor outcomes. These predictions are based on the evidence from the general population regarding the effect of maternal age (Culp et al., 1991) and depression (Hay et al., 2001) on children's IQ scores and our preliminary findings on family cohesion. (3c) If there is an effect of mother's age as well as an effect of child's age on the outcomes, we further predicted that maternal age at HPCT will mediate the effect of child's age at HPCT, so that children of older mothers at pre-HPCT will obtain better scores regardless of their own age at pre-HPCT. (3d) Finally, if there is an effect of time on the cognitive outcomes, we predicted that this effect will be moderated by maternal depression scores. This hypothesis is based on the assumption that when the child's health is significantly compromised, as is the case with the children in this study, particularly at pre-HPCT, parents are likely to experience high distress. However, parental distress is less likely to have an effect on the child's cognitive performance when the child's health is the most compromised (pre-HPCT) than when the child's health has improved, as we assume will be the case at 1 and 2 years post-HPCT.
|
Method |
|---|
|
TopAbstractMethodResultsDiscussionAcknowledgmentsReferences |
|---|
Participants and Recruitment
We approached 133 families who had a child between the ages of 3 and 17 years, eligible for a blood transplant; 120 (90.3%) of whom agreed to participate. Of these, 17 (15%) were unable to participate due to scheduling conflicts or because of a sudden change in health status of the ill child. Of the remaining 103 consenting families, 83 children and their parents were able to complete measures at pre-HPCT. Sample size was reduced significantly by 1 year post-HPCT due to mortality and poor health. Thus, at 1 and 2 years post-HPCT 49 and 48 survivors participated, respectively. This attrition rate is not uncommon in studies with HPCT survivors (Harder, Duivenvoorden, Van Gool, Cornelissen, & Van den Bent, 2006
). Table I presents clinical and demographic characteristics of all consenting families. Briefly, participating children's mean age was 8.9 years (SD = 4.2). The majority of children were diagnosed with ALL or other leukemia types (i.e., acute myelogenous leukemia, chronic myelogenous leukemia). Sixty-eight percent of the children were scheduled for an allergenic transplant. The mean time since diagnosis was 19.9 months (SD = 19.5). Most children came from a two-parent Caucasian family and a middle-class socio-economic background (Hollingshead, 1975).
|
Procedure
After approval from the Research Ethics Board was obtained, an invitation letter was sent to all potential participants describing the study and inviting them to participate. Informed consent and assent were obtained at the initial assessment, with the majority of families (57%) completing consent and the initial assessment within 1 week prior to the transplant. A research assistant explained instructions to parents (and the child when appropriate) for questionnaire completion, and administered the standardized tests to the child under the supervision of the first author, a registered psychologist. The primary caregiver for each child (the mother in every case) completed questionnaires about the child.
Outcome Measures
General Intelligence was measured using the Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R) for children 3–5 years of age (n = 27 at HPCT) (Wechsler, 1989), the Wechsler Intelligence Scale for Children Third Edition (WISC-III), for children 6–15 years of age (for the majority of the sample, n = 47 at HPCT) (Wechsler, 1991), and the Wechsler Adult Intelligence Scale-Revised (WAIS-R) for adolescents 16 years and older at diagnosis (n = 9 at HPCT) (Wechsler, 1981). Five children were tested with the WPPSI-R at pre-HPCT and with the WISC-III at 1 and 2 years post-HPCT. None of the children tested with the WISC-III at pre-HPCT were old enough to be tested subsequently with the WAIS-R. The Wechsler scales are well-normed, producing IQ estimates labeled Verbal (VIQ; which includes Information, Similarities, Arithmetic, Vocabulary, and Comprehension subtests), Performance (PIQ; which includes Picture Completion, Coding, Picture arrangement, Block Design, and Object Assembly), and Full Scale (FIQ). Using the WISC III subtests, including two supplementary subtests, Digit Span, and Symbol Search, we derived specific factor-based cognitive indices: Freedom from Distractibility (FD; which includes Arithmetic and Digit Span) and Processing Speed (PS; which includes Coding and Symbol Search). The two other indices, Verbal Comprehension and Perceptual Organization, were not used because they test basically the same constructs as VIQ and PIQ, respectively. Average reliability estimates for the IQ measures range from .91 to .96. The Wide Range Achievement Test (WRAT3) (Wilkinson, 1993) assesses educational achievement: reading, spelling, and arithmetic skills. Reliability estimates range from .92 to .95. Participants also completed the Developmental Test of Visual Motor Integration (VMI-4), a measure that assesses visual-motor skills in children 3–18 years of age (Beery, 1997). Overall reliabilities for this measure are .92 for the total VMI, .91 for the Visual test, and .89 for the Motor test. Although these scales have been revised since we started the first cohort of this study over 7 years ago, we continued to use the older versions to maintain consistency on the scores.
Measures of Related Factors
Family Adaptability and Cohesion Evaluation Scale (FACES-III, Olson, Portner, & Lavee, 1985) and the Beck Depression Inventory (BDI, Beck, 1978) were used as the measures of family cohesion and maternal distress, respectively. These measures were chosen because they are brief and widely used in families with childhood chronic illness or cancer and could be completed by a parent with minimal burden. FACES-III is a 20-item scale with two dimensions of family relations: adaptation and cohesion. Test–retest reliability over a 4 to 5-week period was r = .80 for adaptability, and r = .83 for cohesion, with high discriminative validity (Elliott, Ozolins, Olson, & Pruitt, 1985). In this study, the cohesion score was used as a measure of family functioning. The BDI is a self-administered 21-item questionnaire designed to measure depressive symptomatology. The BDI has been found to be highly sensitive in measuring change in depressive symptoms and severity (Katz, Katz, & Shaw, 1993). Concurrent validity of the BDI with most other self-report measures of depression has been consistently high (Katz et al., 1993). Both measures were completed by the mother.
Clinical and Demographic Factors
The child's disease and treatment information was obtained from the medical chart and confirmed by parental report. The clinical variables considered in this study were: diagnosis type (ALL, other leukemias, neuroblastoma, other tumors, and hematological disorders); time since diagnosis (less than 6 months, between 6 months to 1 year, and more than 1 year); history of radiation treatment (no radiation, TBR only and combined CRT and TBR) and radiation group (<300 cGy or >1,200 cGy). The personal factors included in this study were: child's age at transplant, gender, presence or absence of siblings, and birth order (oldest, youngest, only child). Mother's age and education were also considered, with the latter as an index of socio-economic status.
Statistic Analyses
The statistical software SAS version 9.1 (SAS Institute Inc. Cary, NC) was used for the statistical analysis. Descriptive statistics, frequency distributions, and percentages were calculated for the outcome variables and covariates of interest. T-tests of independent samples were calculated to compare the cognitive outcomes at pre-HPCT of the participants who continued in the study with those who did not. As well, each outcome measure was compared to the normative value at each assessment period (Hypothesis 1b) using one sample Z-tests, based on the fact that the population standard deviation was available from the normative values. Initial repeated measures ANOVAs were calculated using Global IQ scores (FIQ, VIQ, and PIQ) of participants who had complete data at every assessment. Given that almost half of the participants were too sick or not available for testing at some assessments or had died prior to the 1 year post-HPCT assessment, this analysis had a reduced sample size and did not represent the population of children who undergo the procedure. Thus, subsequent analyses were conducted using a mixed linear regression model, a multivariable regression model that allows for repeated measures, optimum use of all available data at each assessment time, and accounts for collinearity (Hypotheses 1a, 2a, b, and 3a, b). Bivariate analyses were conducted for the cognitive outcomes (FIQ, VIQ, PIQ and FD, PS) at each assessment time, and reading, spelling, arithmetic scores, and VMI scores at 2 years post-HPCT, with disease and treatment variables (history of cranial radiation and radiation group, diagnosis, type of transplant, time since diagnosis), personal and familial variables (child's age, gender, mother's age, depression scores and education, and family cohesion), in order to reduce the number of variables selected to be entered into subsequent mixed multivariable linear regression analyses.
In addition, to avoid collinearity, a pair-wise assessment of associations between predictors was examined. That is, chi-squared tests within categorical predictors, analysis of variance (ANOVA) for continuous and categorical, and correlation analyses for continuous variables were conducted. Variables found to have high potential collinearity were not included together in a multivariable model. In the cases of highly significant associations, a predictor was chosen based on its conceptual importance. Covariates significant at the 20% level in the bivariate analyses were included in the full multivariable linear mixed model. Because SES was highly correlated with maternal education, we used mother's education as a more reliable index of SES. Having siblings had no correlation with any of the outcome variables, nor did we find differences between the outcomes of surviving children who had siblings and those who had none. Thus, these variables were not considered further in the analyses. History of radiation was chosen over radiation group because preliminary analyses suggested potential association with some outcome variables.
Preliminary analyses set the stage for conducting linear mixed model analyses, with a compound symmetry covariance structure, for examining the longitudinal effect of time (Hypothesis 1a) along with the effects of clinical factors: history of Radiation (Hypothesis 2a), time since diagnosis (Hypothesis 2b); child's characteristics: age at HPCT (Hypothesis 3a); and family factors: maternal age, maternal depression, and family cohesion (Hypothesis 3b), on FIQ, VIQ, PIQ, FD, and PS. If there was a time effect, linear contrast analysis was conducted also for comparisons across time.
If maternal age had a significant effect on cognitive outcomes and if there was also a significant effect of child's age, we tested mother's age for a mediation effect (Hypothesis 3c). Thus, child's age was modeled along with maternal age for the specific outcome variable. Maternal depression scores were fit into a regression model and tested for a moderation effect after we found that maternal depression had a significant interaction with time (Hypothesis 3d). To examine the effects of potential predictors on reading, spelling, arithmetic, and VMI scores 2 years post-HPCT, we conducted generalized linear regression analyses for each outcome variable (Hypotheses 2a, b and 3a, b).
To determine clinical significance, effect size was calculated. An effect size of .5 corresponds to a minimal perceptible difference and hence could be considered of clinical significance (Norman, Sloan, & Wyrwich, 2003). As well, percentage of participants who scored one standard deviation below the mean of cognitive outcomes was also calculated at pre-HPCT, 1 and 2 years post-HPCT.
|
Results |
|---|
|
TopAbstractMethodResultsDiscussionAcknowledgmentsReferences |
|---|
Preliminary Analyses
Except for child's age (M = 9.09, SD = 4.16; M = 6.66, SD = 5.14, p < .01) and mother's age (M = 37.51, SD = 5.69; M = 33.40, SD = 5.74, p < .01), both older for the participants, there were no significant differences in demographic variables between consenting participants and nonparticipants. In terms of the clinical variables, there were no significant differences in the clinical variables between the two groups, except that there were more nonparticipants with a history of cranial radiation than were participants (
2 = 6.88, p = .03). Comparisons of cognitive scores at pre-HPCT between participants who remained in the study (survivors) and those who did not indicated that the mean FIQ and PIQ scores of the survivors were significantly higher than the mean scores of the nonsurvivors (M = 104.22, SD = 12.29; M = 95.32, SD = 15.93 for FIQ; M = 104.13, SD = 13.67, M = 96.21, SD = 15.30 for PIQ; p = .007 and p = .029 for FIQ and PIQ, respectively). The mean VIQ scores were also higher for survivors but did not reach significance (M = 101.5, SD = 13.61; M = 95.27, SD = 16.39).
Comparisons to Norms
Table II presents the means and standard deviations for the outcome measures at each assessment.
|
Cognitive Outcomes
At pre-HPCT, survivors’ mean FIQ, VIQ, and PIQ scores were not significantly different from the normative values. This remained the case for VIQ score at 1 and 2 years post-HPCT. At 2 years post-HPCT, survivors’ mean FIQ score was significantly above the norm, Z = 2.628, p < .01, ES = .36, but the effect size was not considered clinically significant. At 1 and 2 years post-HPCT, the mean PIQ scores were both statistically and clinically significantly higher than the norm, (Z = 3.341, p < .001, ES = .50 and Z = 5.044, p < .000, ES = .68, for 1 and 2 years post-HPCT, respectively) (Hypothesis 1a). To determine further the clinical significance of the scores we examined the frequency distribution of scores that were less than or equal to one standard deviation from the mean of a 100 at pre-HPCT, 1 and 2 years post-HPCT. At pre-HPCT the percentage of children whose scores were
85 was 13, 16, and 16%, for FIQ, VIQ, and PIQ scores, respectively. This percentage remained approximately the same for VIQ at 1 year and 2 years post-HPCT and for FIQ score at 1 year post-HPCT. For PIQ, however, the percentage of survivors with scores 85 decreased to 7% at 1 year and to 8% at 2 years post-HPCT. The cognitive indices (FD and PS) did not differ from the norm at pre-HPCT and 1 year post-HPCT. At 2 years post-HPCT, however, the mean score of FD was marginally lower than the norm (Z = –1.885, p = .059, ES = .26), but the mean score of processing speed was both statistically and clinically significantly higher than the norm (Z = 3.383, p < .001, ES = .57) (Hypothesis 1a).
Educational and Perceptual Motor Outcomes
Survivors’ mean standard arithmetic score was statistically and clinically significantly below the norm (Z = –3.128, p < .002, ES = .51), but their reading and spelling scores did not differ significantly from the norm. Survivors’ mean standard VMI score was also statistically below the norm, but not clinically significant (Z = –2.622, p < .01, ES = .43) (Hypothesis 1a).
Cognitive Outcomes over time and their Potential Predictors
For each cognitive variable we conducted multivariable linear mixed model analyses using time and the covariates (Hypotheses 1b, 2a and b, and 3a and b). These analyses were conducted both with and without the participants who did not complete the follow-up assessments. Because the results of the two analyses were remarkably similar and to ensure that this study represents the HPCT population, we decided to report the analyses that included all available data.
For the mean FIQ scores, we used time and the following covariates: child's age; time since diagnosis; history of radiation; mother's age; and maternal depression, in the mixed model analysis. There was no time effect, but, adjusting for the effect of time, there was a significant effect of mother's age [F(1,90) = 5.35, p = .023; 95% CI = 0.088–1.154] (Hypothesis 3b). In general, the mean FIQ score were higher for children who had older mothers at pre-HPCT. There was also an interaction between child's age and time [F(2,49) = 3.53, p = .037; 95% CI = 0.010–1.170 for year 1; and CI = 0.169–1.422 for year 2]. This interaction indicated that older children at transplant had a better score at 1 and at 2 years post-HPCT. Following Baron and Kenny's (1986) criteria for testing mediation, further analysis was conducted to examine the relationship between mother's age and child's age at HPCT on FIQ. Fitting regression of mother's age and child's age we found a significant association between these two variables [F(1,106) = 34.79, p < .0001; 95% CI = 0.443–.891]. In addition, mother's age pre-HPCT modeled with child's age and time were significantly associated with FIQ [F(1,88) = 4.88, p = .0297; CI = 0.067–1.255] (Hypothesis 3c). Thus, mother's age met the three conditions for mediation of the effect of child's age at pre-HPCT on the mean FIQ score over time (there were two causal paths feeding into the FIQ outcome, the direct path of the independent variable, child's age over time, the impact of the third variable, mother's age, as well as a path from child's age and mother's age).
For the mean VIQ scores, using time and the following covariates: history of radiation; diagnosis; child and mother's age; maternal depression; and family cohesion, again, we found no significant time effect. When adjusting for time, a strong direct effect of mother's age was found [F(1,85) = 9.90, p = .0023; 95% CI = 0.330–1.465]. As with the FIQ score, children with older mothers generally had a higher VIQ score. In addition, an interaction between time and maternal depression was found [F(2,71) = 3.93, p = .0240; 95% CI = –.897 to –.121 for year 1; and CI = –0.767 to –.0324 for year 2] (Hypothesis 3d), suggesting that maternal depression moderated the effect of time on the children's VIQ. At pre-HPCT children's VIQ scores were not related to mothers’ depression scores; at 1 and 2 years post-HPCT, however, children with higher VIQ had mothers with the lowest depression score.
For the mean PIQ scores, the linear mixed model procedure using time and the following covariates: time since diagnosis; history of radiation; child's age; mother's age; and depression scores, yielded a significant time effect [F(2,77) = 6.96, p = .0017; 95% CI = 1.884–6.489 for year 1; and CI = 3.178–11.627 for year 2] (Hypothesis 1b). Analysis of this effect indicated a statistically and clinically significant improvement from pre- to 1 year post-HPCT (7.7 point increment, Effect size = .53) and from pre- to 1 years post-HPCT (11.27 point increment, Effect size = .70). There was also a significant effect of maternal depression adjusting for time and the other covariates, [F(1,134) = 5.65, p = .0189; 95% CI = –.0793 to –.073], suggesting that in general children whose mothers had a high depression score had a lower PIQ score (Hypothesis 3b). Finally, a marginally significant interaction between radiation history and maternal depression was found [F(1,113) = 3.63, p = .059; 95% CI = –0.0279 to 1.367]. Analysis of this interaction trend indicated that children who received both CRT and TBR, and had mothers with a high depression score, regardless of time, obtained the lowest PIQ score.
Cognitive Indices
No significant time effect was observed with these indices. There was a marginal significant effect of maternal age for FD [F(1,90) = 3.52, p = .058; 95% CI = –0.010 to 0.352]. Adjusting for time, children with higher scores tended to have older mothers. For PS, after adjusting for time, family cohesion was a significant predictor, indicating that higher scores were obtained by children from cohesive families [F(1,62) = 5.16, p = .0266; 95% CI = –16.795 to –1.074) (Hypothesis 3b).
Potential Predictors of Perceptual Motor and Educational Outcomes
The generalized linear regression analysis of the VMI and educational outcomes at 2 years post-HPCT included the following variables: history of radiation; time since diagnosis; child's age; mother's age; and depression score. Adjusting for child's and mother's age, maternal depression was a significant predictor for the VMI outcome [F(1,34) = 4.37, p = .0440; 95% CI = –1.267 to –0.018] (Hypothesis 3b). Children with higher VMI had mothers with a lower depression score.
For educational outcomes, time since diagnosis was the critical factor for reading [F(1,38) = 8.81, p = .0052; 95% CI = –.384 to –.073] and for arithmetic scores [F(1,39) = 4.48, p = .0408; 95% CI = –.284 to –.006) (Hypothesis 2b). In both cases, the scores were lower the longer the time since diagnosis. Child's age was also related to reading [F(1,38) = 4.57, p = .0390; 95% CI = .061–2.251) and spelling scores [F(1,31) = 6.23, p = .0181; 95% CI = .246–2.441), with older children having better scores than younger children.
|
Discussion |
|---|
|
TopAbstractMethodResultsDiscussionAcknowledgmentsReferences |
|---|
The findings of this study present a comprehensive and positive picture of global and specific cognitive outcomes of HPCT longitudinally, perceptual motor skills and educational outcomes 2 years after transplant, and factors related to these outcomes.
Changes over time and Comparison to the Norms
As expected (Hypothesis 1b), no significant changes in full IQ, verbal IQ, or specific cognitive indices of Freedom from Distractibility and Processing Speed were found overtime. Contrary to prediction, however, survivors’ performance IQ significantly improved at one and 2 years post-HPCT, compared to pre-HPCT, with the percentage of survivors who obtained a score of 1SD below the norm dropping from 16% at pre-HPCT to only a 7 and 8%. The improvements in PIQ is consistent with Cool's (1996) results 1 year posttransplant, whereas the lack of absolute change in FIQ and VIQ scores is consistent with Phipps’ et al. (2000).
Moreover, contrary to expectations (Hypothesis 1a), the PIQ and the specific cognitive index of Processing Speed were markedly higher than normative values at the follow up assessments. The fact that by 2 years post-HPCT survivors’ health had improved, allowing them to focus better on the task at hand may account partially for these results. As well, since the performance scale of the Wechsler test is "more dependent on the child's immediate problem-solving ability" (Sattler, 1988; p. 172), the improvement in the PIQ scores, may also reflect the nature of the scale. It is possible, however, that this improvement reflects greater use of video games during the convalescence period.
In contrast, and contrary to expectations (Hypothesis 1a), arithmetic scores and perceptual motor scores at 2 years post-HPCT were significantly below the population norms. Numerical operations and perceptual motor skills are typically learned in the first years of school and form the basis for more complex learning of mathematics and writing. This finding is consistent with previously reported deficits in arithmetic in survivors of leukemia or CNS tumors (Barrera, Shaw, Speechley, Maunsell, & Pogany, 2005; Kaemingk, Carey, Moore, Herzer, & Hutter, 2004; Peckham, Meadows, Bartel, & Marrero, 1988), and therefore requires further investigation.
Relationships Among Clinical Factors, Child Characteristics, and Family Factors
Family factors and child characteristics appeared to have greater impact on the outcomes than clinical factors. Time since diagnosis was the only clinical factor that had a direct effect on educational (reading and arithmetic scores) and visual motor outcomes, but not on cognitive outcomes (Hypothesis 2b). This finding suggests that being ill for a prolonged period of time before transplant has a negative effect on learning but not on the overall cognitive skills. Although we found no major effect of combined CRT and TBR on any of the outcomes, a marginal but important relationship was found between combined CRT and TBR and maternal depression scores with the PIQ scores (Hypothesis 3d). Thus, it appears that having had CRT and TBR may impact negatively on survivors’ PIQ if maternal depression scores were high. The potential effect of cranial radiation on the outcomes of survivors of pediatric HPCT has been previously inconsistent (Cool, 1996; McGuire et al., 1991; Phipps et al., 2000). In depth analysis of pre-HPCT data indicates that 55% of the children who did not continue in the study had TBR and CRT, whereas of the participants who remained in the study, 94% did not have radiation, suggesting that high mortality but not morbidity, (measured by cognitive, educational, and perceptual motor outcomes) was directly related to history of radiation.
As hypothesized (Hypothesis 3a) and consistent with previous findings (Cool, 1996; Kramer et al., 1997; McGuire et al., 1991; Phipps et al., 2000), older children at transplant had a better FIQ score in general and better reading and spelling scores 2 years post-HPCT, highlighting the importance of normal cognitive development and learning during the early years of life. However, this age effect on FIQ was mediated by maternal age (Hypothesis 3c), a finding not previously reported with this population. Maternal age had a direct effect on VIQ and freedom from distractibility scores (Hypothesis 3b), regardless of the child's age, with children with high scores having older mothers. As with the general population (Culp, Osofsky, & O’Brien, 1996; Lacroix, Pomerleau, & Malcuit, 2002) the effect of mother's age stresses the importance of the caregiver environment and suggests that older mothers may have greater maturity and experiences, as well as financial stability, to provide a stimulating environment and learning strategies for their surviving children than younger mothers. Unlike in the general population, the relationships between children's cognitive development and maternal age did not seem to be confounded by socio-economic status, reflecting the socialized medicine of the health care system for the current sample.
Two other family factors seem to influence cognitive and perceptual motor outcomes: maternal depression and family cohesion. Maternal depression had a direct negative effect on PIQ and VMI scores at 2 years post-HPCT (Hypothesis 3b), moderated the time effect for VIQ scores (Hypothesis 3d), and possibly the effect of history of radiation on PIQ. The importance of maternal health on the cognitive outcomes of these survivors is consistent with cognitive outcomes of healthy children who had clinically depressed mothers (Hay et al., 2001). Finally, confirming the importance of the caretaking environment found in preliminary findings (Barrera et al., 2000), family cohesion contributed to better outcomes for HPCT survivors when the task required relative speed, short-term memory, and sustained attention.
Missing or having incomplete data is not unusual in longitudinal research with this medically vulnerable population faced with poor health, many stressors, and high mortality rates (Phipps et al., 2000). The lack of a comparison group is also a limitation in this study. On the other hand, the outcomes from this comprehensive longitudinal study seem to be conservative and valid. Lack of educational data at admission and at 1 year prevented us from comparing these outcomes over time, though a previous study found no differences in educational scores between baseline and 1 or 3 years posttransplant (Phipps et al., 2000). Sample self-selection might have biased the results towards those individuals who were healthier at pre-HPCT. Comparisons between patients who had only pre-HPCT and those who completed follow up indicated that those who had only pre-HPCT had lower IQ scores. By including all available participants at each assessment the potential for selection bias was likely reduced.
In conclusion, this study shows that in general, survivors of pediatric HPCT do well cognitively and improve over time in performance IQ, but do less well educationally and in terms of perceptual motor skills. Thus, it appears that the mechanisms involved in performing arithmetic operations, and to a lesser extent perceptual motor skills, appear somewhat impacted by the HPCT experience. Considering the conceptual model, this study illustrates the complex interaction of clinical, child and family factors for cognitive, educational, and perceptual motor development, after pediatric HPCT. Clinical variables seem to determine whether or not patients will survive 2 years after HPCT but have little effect on these outcomes. In contrast, family variables, maternal age and depression and family cohesion, seem to have the most impact on these outcomes. Finally, this study identifies critical factors for developing family focused psycho-educational interventions for at risk children who undergo aggressive treatment procedures such as HPCT.
|
Acknowledgments |
|---|
|
TopAbstractMethodResultsDiscussionAcknowledgmentsReferences |
|---|
The authors thank K. Sumbler and T. Smith for their help in collecting data for this study, F. Schulte for her comments on an early draft, and J. Pinto and T. Teall for helping to prepare the article. This research was supported by grants from the Hospital for Sick Children Foundation, New Initiatives; Elizabeth Lue Bone Marrow Foundation; and the National Cancer Institute of Canada.
Conflict of Interest: None declared.
Received January 16, 2007; revision received August 14, 2007; accepted August 16, 2007
|
References |
|---|
|
TopAbstractMethodResultsDiscussionAcknowledgmentsReferences |
|---|
Anderson VA, Godber T, Smibert E, Weiskop S, Ekert H. Cognitive and academic outcome following cranial irradiation and chemotherapy in children: A longitudinal study. British Journal of Cancer (2000) 82(2):255–262.[CrossRef][Web of Science][Medline]
Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology (1986) 51:1173–1182.[CrossRef][Web of Science][Medline]
Barrera M, D’Agostino N, Gibson J, Weksberg R, Malkin D. Predictors and mediators of psychological adjustment in mothers of children newly diagnosed with cancer. Psycho-Oncology (2004) 13:1–12.[Medline]
Barrera M, Pringle L.-AB, Sumbler K, Saunders F. Quality of life and behavioral adjustment after pediatric bone marrow transplantation. Bone Marrow Transplantation (2000) 26(4):427–435.[CrossRef][Web of Science][Medline]
Barrera M, Shaw AK, Speechley KN, Maunsell E, Pogany L. Educational and social late effects of childhood cancer and related clinical, personal, and familial characteristics. Cancer (2005) 104(8):1751–1760.[CrossRef][Web of Science][Medline]
Barrera M, Wayland LA, D’Agostino N, Gibson J, Weksberg R, Malkin D. Developmental differences in psychological adjustment and health related quality of life in pediatric cancer patients. Children's Health Care (2003) 12(3):215–232.
Beck AT. Beck depression inventory (1978) San Antonio, TX: Psychological Corporation.
Beery KE. The Beery-Buktenica developmental test of visual-motor integration. In: Administration, scoring and teaching manual (1997) 4th. Parsipanny, NJ: Modern Curriculum Press.
Brown RT, Madan-Swain A, Pais R, Lambert RG, Baldwin K, Casey R, et al. Cognitive status of children treated with central nervous system prophylactic chemotherapy for acute lymphocytic leukemia. Archives of Clinical Neuropsychology (1992) 7(6):481–497.[CrossRef][Web of Science][Medline]
Cool VA. Long-term neuropsychological risks in pediatric bone marrow transplant: What do we know? Bone Marrow Transplantation (1996) 18(3):45–49.
Copeland DR, deMoor C, Moore BD, Ater JL. Neurocognitive development of children after a cerebellar tumor in infancy: A longitudinal study. Journal of Clinical Oncology (1999) 17(11):3476–3486.
Culp AM, Osofsky JD, O’Brien M. Language patterns of adolescent and older mothers and their one-year-old children: A comparison study. First Language (1996) 16(46):61–75.
Culp RE, Culp AM, Osofsky JD, Osofsky HJ. Adolescent and older mothers' interaction patterns with their six-month-old infants. Journal of Adolescence (1991) 14(2):195–200.[CrossRef][Web of Science][Medline]
Elliott CH, Ozolins M, Olson RA, Pruitt SD. Behavioral management of children's distress during painful medical procedures. Behavior Research and Therapy (1985) 23(5):513–520.[CrossRef][Web of Science][Medline]
Espy KA, Moore IM, Kaufmann PM, Kramer JH, Matthay K, Hutter JJ. Chemotherapeutic CNS prophylaxis and neuropsychologic change in children with acute lymphoblastic leukemia: A prospective study. Journal of Pediatric Psychology (2001) 26(1):1–9.
Fergusson DM, Woodward LJ. Maternal age and educational and psychosocial outcomes in early adulthood. Journal of Child Psychology & Psychiatry & Allied Disciplines (1999) 40(3):479–489.[CrossRef][Web of Science][Medline]
Gassas A, Sung L, Saunders EF, Doyle JJ. Comparative outcome of hematopoietic stem cell transplantation for pediatric acute lymphoblastic leukemia following cyclophosphamide and total body irradiation or VP16 and total body irradiation conditioning regimens. Bone Marrow Transplantation (2006) 38(11):739–743.[CrossRef][Web of Science][Medline]
Harder H, Duivenvoorden HJ, Van Gool AR, Cornelissen JJ, Van den Bent MJ. Neurocognitive functions and quality of life in haematological patients receiving haematopoietic stem cell grafts: A one-year follow-up pilot study. Journal of Clinical & Experimental Neuropsychology: Official Journal of the International Neuropsychological Society (2006) 28(3):283–293.
Hay DF, Pawlby S, Sharp D, Asten P, Mills A, Kumar R. Intellectual problems shown by 11-year-old children whose mothers had postnatal depression. Journal of Child Psychology & Psychiatry & Allied Disciplines (2001) 42(7):871–889.[CrossRef][Web of Science][Medline]
Hollingshead A. Four factor index of social status (1975) New Haven, CT: Yale University Press.
Holmbeck GN. Toward terminological, conceptual, and statistical clarity in the study of mediators and moderators: Examples from the child-clinical and pediatric psychology literatures. Journal of Consulting and Clinical Psychology (1997) 65:599–610.[CrossRef][Web of Science][Medline]
Hubbs-Tait L, Culp AM, Culp RE, Miller CE. Relation of maternal cognitive stimulation, emotional support, and intrusive behavior during Head Start to children's kindergarten cognitive abilities. Child Development (2002) 73(1):110–131.[CrossRef][Web of Science][Medline]
Kaemingk KL, Carey ME, Moore IM, Herzer M, Hutter JJ. Math weakness in survivors of acute lymphoblastic leukemia compared to healthy children. Child Neuropsychology (2004) 10(1):14–23.[Medline]
Katz R, Katz J, Shaw BF. Beck depression inventory and hopelessness scale. In: Psychological testing: Treatment planning and outcome assessment—Maruish ME, ed. (1993) New York: Lawrence Erlbaum. 120–133.
Keown LJ, Woodward LJ, Field J. Language development of pre-school children born to teenage mothers. Infant and Child Development (2001) 10(3):129–145.[CrossRef][Web of Science]
Kramer JH, Crittenden MR, DeSantes K, Cowan MJ. Cognitive and adaptive behavior 1 and 3 years following bone marrow transplantation. Bone Marrow Transplantation (1997) 19(6):607–613.[CrossRef][Web of Science][Medline]
Lacroix V, Pomerleau A, Malcuit G. Properties of adult and adolescent mothers' speech, children's verbal performance and cognitive development in difference socioeconomic groups: A longitudinal study. First Language (2002) 22(65, Pt2):173–196.
Mabbott DJ, Speigler BJ, Greenberg ML, Rutka J, Hyder DJ, Bouffet E. Serial evaluation of academic and behavioral outcome after treatment with cranial radiation in childhood. Journal of Clinical Oncology (2005) 23(10):2256–2263.
McGuire T, Sanders JE, Hill D, Buckner CD, Sullivan K. Neuropsychological function in children given total-body irradiation for marrow transplantation. Experimental Hematology (1991) 19(6):578–578.
Mennes M, Stiers P, Vandenbussche E, Vercruysse G, Uyttebroeck A, De Meyer G, et al. Attention and information processing in survivors of childhood acute lymphoblastic leukemia treated with chemotherapy only. Pediatric Blood & Cancer (2005) 44(5):478–486.[Medline]
Moore BD. Neurocognitive outcomes in survivors of childhood cancer. Journal of Pediatric Psychology (2005) 30(1):51–63.
Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life: the remarkable universality of half a standard deviation. Medical Care (2003) 41(5):582–592.[CrossRef][Web of Science][Medline]
Olson DH, Portner J, Lavee Y. FACES III (1985) St. Paul, MN: University of Minnesota.
Peckham VC, Meadows AT, Bartel N, Marrero O. Educational late effects in long-term survivors of childhood acute lymphocytic leukemia. Pediatrics (1988) 81(1):127–133.
Phipps S, Brenner M, Heslop H, Krance R, Jayawardene D, Mulhern RK. Psychological effects of bone marrow transplantation on children and adolescents: Preliminary report of a longitudinal study. Bone Marrow Transplantation (1995) 15(6):829–835.[Web of Science][Medline]
Phipps S, Dunavant M, Srivastava DK, Bowman L, Mulhern RK. Cognitive and academic functioning in survivors of pediatric bone marrow transplantation. Journal of Clinical Oncology (2000) 18(5):1004–1011.
Raymond-Speden E, Tripp G, Lawrence B, Holdwaway D. Intellectual, neuropsychological, and academic functioning in long-term survivors of leukemia. Journal of Pediatric Psychology (2000) 25(2):59–68.
Sattler JM. Assessment of children (1988) 3rd. San Diego: Jerome M. Sattler.
Smedler AC, Nilsson C, Bolme P. Total body irradiation: a neuropsychological risk factor in pediatric bone marrow transplant recipients. Acta Paediatrica (1995) 84(3):325–330.[Web of Science][Medline]
Spiegler BJ, Bouffet E, Greenberg ML, Rutka JT, Mabbott DJ. Change in neurocognitive functioning after treatment with cranial radiation in childhood. Journal of Clinical Oncology (2004) 22(4):705–713.
Wechsler D. Wechsler adult intelligence scales revised (1981) New York, NY: The Psychological Corporation.
Wechsler D. Wechsler preschool and primary intelligence scales for children, revised (1989) New York, NY: The Psychological Corporation.
Wechsler D. Wechsler intelligence scale for children III (1991) New York, NY: The Psychological Corporation.
Wilkinson GS. WRAT3 wide range achievement test: Administration manual (1993) Wilmington: DE Wide Range Inc.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||