Pediatric Quality of Life Inventory (PedsQL) in Fragile X Syndrome

Abstract

To date, health related quality of life (QoL) has not been systematically evaluated in youth with fragile X syndrome (FXS), the most common single gene cause of autism and the most common inherited form of developmental disability. We describe QoL data gathered using the Pediatric Quality of Life Inventory (PedsQL) completed online by 364 parents of youth with FXS. Parents consistently reported across all gender and age groups that their children experienced the highest QoL in Physical functioning and the lowest QoL in Cognitive functioning. Overall, older children with FXS had increase QoL ratings in the domains of School and Cognitive function.

Fragile X syndrome (FXS) is a neurodevelopmental disorder caused by a CGG trinucleotide repeat expansion (> 200 repeats) within the 5’ untranslated region of the fragile X mental retardation gene (FMR1) on the X chromosome (Pieretti et al. 1991). This expansion leads to methylation and subsequent reduction or absence of fragile X mental retardation protein (FMRP), which is important for synaptic plasticity and experience-dependent learning (Bagni et al. 2012). Despite being a single gene disorder, FXS is marked by a heterogeneous presentation of significant cognitive, language, sensory, behavioral, and emotional impairments, and autism-like symptoms or a comorbid diagnosis of autism spectrum disorder (ASD).

Given the complex phenotype, individuals with FXS often present to clinical settings for numerous concerns, including intellectual disability (ID), hyperarousal to sensory stimuli, anxiety, aggression, sleep issues, hyperactivity, and social and communication difficulties among other concerns (Berry-Kravis and Potanos 2004). These common clinical issues associated with FXS can have a lifelong impact on the affected child, parent(s), and the overall well-being of the family of the child (Wheeler et al. 2008a, b; Bailey et al. 2010). Thus, the severity of these clinical issues and, in turn, the quality of life (QoL) of the child and family may be important outcome measures following intervention.

Health-related quality of life (HRQoL) is a multidimensional concept involving physical, psychological, social, and cognitive aspects of life (Cummins 2005; Schalock et al. 2002). In children, examining HRQoL and global functioning is particularly important for the evaluation and optimization of their health and development (Varni et al. 1999a, b, 2001a, b, 2002). The United States Food and Drug Administration (FDA) has stated that QoL as a general concept implying evaluation of the effect of all aspects of life on general well-bring is too general to be considered appropriate when evaluating the efficacy or effectiveness of treatment. In contrast, the FDA has affirmed that HRQoL as defined above can be evaluated as an outcome measure of change with treatment.

To date, no studies have systematically evaluated HRQoL in FXS. Some data is available describing HRQoL in persons with idiopathic autism and with non-specific developmental delay. For example, cognitive and gross-motor delays generally in children have been demonstrated to negatively impact HRQoL in youth (Alonso and Sorensen 2009; Hsieh et al. 2013). Additionally, HRQoL was found to be lower in children with autism and intellectual disability compared to typically developing peers (Kuhlthau et al. 2010; Limbers et al. 2009; Kamp-Becker et al. 2010).

Utilization of health-related QoL (HRQoL) instruments, which allow for comparisons across chronic and healthy conditions, has significantly increased in the past two decades (Varni et al. 2005). One such measure, the Pediatric Quality of Life Inventory (PedsQL) 4.0 Generic Core Scales (Varni et al. 1999a, b) is an instrument for measuring HRQoL in children and adolescents ages 2 to 18 using a child self-report version and/or a parent-proxy report version. Originally developed to focus on HRQoL, the PedsQL does largely focus on functional status which may not be synonymous with well-being. The PedsQL 4.0 Generic Core Scales consist of four scales measuring HRQoL in four domains: Physical, Emotional, Social, and School functioning. The PedsQL Cognitive Functioning Scale is a symptom-specific instrument to measure the impact of cognitive function on HRQoL. The PedsQL Family Impact Module (Varni et al. 2004) measures the impact of pediatric chronic health conditions on parents’ own health related quality of life and the family overall functioning. In the Parent Report versions of the measures, a parent/caregiver proxy-reporter answers questions about their child’s functioning, their own functioning, and their family’s functioning to capture a comprehensive view of HRQoL. PedsQL demonstrated high reliability and construct validity in both health and acute- and chronic-disease pediatric populations, responsiveness to clinical changes, and may be applicable in clinical trials (Desai et al. 2014; Varni et al. 2001a, b, 2007a). This study aims to evaluate the HRQoL of individuals with FXS, their parents, and their families using the PedsQL 4.0 Generic Core Scales, Cognitive Functioning Scale, and Family Impact Module. We first use the scale to describe the impact of having a child with FXS on HRQoL, and the role that sex and age of the child may play on HRQoL. We then conduct a pilot examination of the scale characteristics when used in this population.

Methods

Participants

Three-hundred fifty-five parents completed the PedsQL online via SurveyMonkey. These reports were on 364 youth with full mutation FXS (7 parents reported on two children; 1 parent reported on 3 children). Parents completed the PedsQL as proxy reporters for their children. The National Fragile X Foundation facilitated survey completion with a nationwide email that directed families impacted by FXS to the survey link. Only deidentified data was analyzed in this report and the project was determined to be exempt from IRB review by our local IRB. In addition to online completion of the PedsQL survey, parents self-reported the age of their child with full mutation FXS in years and the sex of their child. For the purposes of this report, we included the surveys of parents of children with self-reported full mutation FXS ages 5 to 17 years old (mean age = 10.41, SD = 3.7). Consistent with the ratio of males to females with full mutation FXS, more surveys were completed by parents of males with FXS (n = 294, 80.8%) compared to females with FXS (n = 70, 19.2%).

Measures

The measures included the Parent Report for Children versions of the PedsQL 4.0 Generic Core Scales and Cognitive Functioning Scale, as well as the PedsQL Family Impact Module. Combined, the PedsQL 4.0 Generic Core Scales and Cognitive Functioning Scale include twenty-nine items measuring five core dimensions of health: Physical Functioning (8 items; example, “problems with…walking more than one block”), Emotional Functioning (5 items; example, “problems with…feeling afraid or scared”), Social Functioning (5 items; example, “problems with…getting along with other children”), School Functioning (5 items; example, “problems with…paying attention in class”), and Cognitive Functioning (6 items; example, “problems with…difficulty remembering what people tell him/her”). The Parent Report versions used in this study assess parents’ perceptions of their child’s HRQoL by asking how much of a problem their child has had with specific activities within each dimension of health in the past month.

The PedsQL Family Impact Module (PedsQL Family) includes twenty-eight items measuring HRQoL of the caregiver across six dimensions of health: Physical Functioning (6 items), Emotional Functioning (5 items), Social Functioning (4 items), Cognitive Functioning (5 items), Communication (3 items), and Worry (5 items; referred to as “Caregiver QoL”). In addition, eight items measured the QoL of the caregiver’s family as a result of the affected child’s health across two dimensions: Daily Activities (3 items) and Family Relationships (5 items; referred to as “Family QoL”).

All items across the three measures used are scored using a five-point Likert scale: 0 = never a problem, 1 = almost never a problem, 2 = sometimes a problem, 3 = often a problem, 4 = almost always a problem. Based on standard scoring practice, all responses were reverse-scored and then linearly transformed to a 0–100 scale (0 = 100,1 = 75, 2 = 50, 3 = 25, 4 = 0). Subscale scores were determined as the sum of the item scores divided by the number of items answered (to account for missing data). Higher scores indicated higher QoL/functioning.

Statistical Analyses

For all analyses, subjects were divided into two groups based on a median age split (median = 10), which resulted in a younger group (ages 5–10, n = 189, 51.9%) and an older group (ages 11–17, n = 175, 48.1%). All analyses were conducted in SPSS Version 23 (IBM Analytics). Preliminary analyses found very high correlations between the subscales within the Caregiver QoL and within the Family QoL (r’s > .60). Thus, a principal components analysis was conducted separately for the items for each of these measures with inter-item reliabilities tested using Cronbach’s alphas. Then, correlations and stepwise linear regressions were conducted to test associations between the study measures resulting from the principal components analyses. Specifically, two stepwise linear regressions were conducted. For each regression, sex and age group were entered on the first step as control variables and all five dimensions of health were entered for the second step. The Caregiver and Family QoL variables were outcome variables. Next, a within-subjects ANOVA tested mean differences across the five dimensions of QOL testing the within-subject main effect (i.e. the five dimensions), a sex X within-subjects interaction, and a sex X within-subject interaction. Preliminary analyses found no significant three-way interaction between age, sex, and dimension. Finally, independent t tests tested sex (Fig. 1) and age differences (Fig. 2) for all the study measures. Preliminary analyses found no age X sex interactions for any study measures.

Fig. 1.

Bar graphs depicting the means for males and females for each of the five dimensions of HRQoL

Fig. 2.

Bar graphs depicting the means for 5 to 10 years old participants and 11 to 17 years old participants for each of the five dimensions of HRQoL

Principal components analyses conducted separately for the Caregiver QoL items and the Family QoL items found that the Caregiver QoL items and Family QoL items were each reduced to one factor (Eigenvalues 3.85 and 1.63; Percentage of variance explained 64.18% and 81.52%, respectively) with high inter-item reliabilities for each factor as indicated by the Cronbach’s alpha (see Table 1). Thus, all Caregiver QoL items and all Family QoL items from the Family Impact Module were summed and averaged to create a summary Caregiver QoL score and a summary Family QoL score.

Table 1.

Correlations and inter-item reliabilities across all study measures and subscales

Dimension	1	2	3	4	5	6	Reliability
1. Physical	–						.84
2. Emotional	.34***	–					.72
3. Social	.48***	.44***	–				.82
4. School	49***	.47***	.54***	–			.77
5. Cognitive	.33***	.37***	.52***	.62***	–		.90
6. Caregiver	.44***	.46***	.46***	.53***	.50***	–	.96
7. Family	.35***	.48***	.45***	.52***	.44***	.76***	.94

Reliabilities are Cronbach’s alpha coefficients

^*p < 0.05;

^**p < 0.01;

^***p < 0.001

Results

Associations Between Child QoL, Caregiver QoL, and Family QoL

Correlations between all study measures in Table 1 show significant associations. Overall, the strongest correlation was between the Caregiver QoL and Family QoL. For correlations between the five dimensions of health for the child, the strongest correlation was between Cognitive functioning and School functioning, and the weakest between Cognitive functioning and Physical functioning. Both the Caregiver QoL and Family QoL summary scores showed the strongest correlation with the child’s School functioning, and the weakest correlation with the child’s Physical functioning out of the five dimensions of health.

Table 2 provides the findings for each regression. Higher scores (better QoL) in Physical, Emotional, and Cognitive functioning for the child were positively associated with Caregiver QoL. Higher scores (better QoL) in Emotional, Social, and School functioning for the child were positively associated with Family QoL. Sex and age group did not moderate any of the main effects of the dimensions of health in either regression.

Table 2.

Findings of stepwise regressions examining the associations between each of the five health dimensions and parent reporter and family variables

	b	t-value	β	ΔR2
Parent reporter
Step 1				.04***
Age	1.04	3.51***	.19
Sex	2.52	0.10***	.05
Step 2				.38***
Physical	0.16	3.37***	.18
Emotional	0.24	4.70***	.24
Social	0.09	1.64	.09
School	0.12	1.65	.11
Cognitive	0.22	3.96***	.30
PedsQL family impact model
Step 1				.04***
Age	0.84	2.35*	−.13
Sex	8.67	2.58*	−.14
Step 2				.37***
Physical	0.05	0.84	.04
Emotional	0.36	5.67***	.29
Social	0.15	2.26*	.13
School	0.30	3.42***	.23
Cognitive	0.12	1.75	.10

^*p < 0.05;

^***p < 0.001

Mean Differences Among Dimensions of QOL

Based on the parent/caregiver report for each child on the PedsQL 4.0 Generic Core Scales and Cognitive Functioning Scale, the within-subject measures ANOVA found a significant main effect for QoL within each dimension of health, as well as two interaction effects: sex X dimension and age group X dimension (see Table 3). For the total sample, the dimensions of health went from highest mean score (higher QoL/functioning) to lowest mean score (lower HRQoL/functioning) in the following order: Physical functioning, Emotional functioning, School functioning, Social functioning, and Cognitive functioning; i.e. Physical functioning was the least affected and Cognitive functioning was the most affected. The order of dimensions—i.e. the dimensions of health ranked from highest HRQoL/functioning to lowest HRQoL/functioning by mean score—only slightly differed depending on sex. For males, the order was the same as the total group with a significant difference between each dimension. Females with FXS also showed the highest HRQoL in Physical functioning and lowest HRQoL in Cognitive functioning like males and the total group, but there were no significant differences found between HRQoL scores for Emotional, Social, and School functioning.

Table 3.

Analyses testing differences across the means of the five health dimensions for the total sample and across sex and age group

			Sex					Age
	Total		Male		Female		t test	5–10		11–17		t test
	Mean	SD	Mean	SD	Mean	SD		Mean	SD	Mean	SD
Physical	63.48a	21.5	63.46a	21.0	63.56a	23.6	0.32	61.58a	21.0	65.59a	21.8	1.65
Emotional	57.55b	19.3	58.45b	18.9	53.77b	20.6	1.65	55.71b	18.6	59.58b	20.0	1.92
Social	47.19d	21.1	46.07d	20.6	51.86b	22.5	2.04*	45.91d	20.6	48.60c	21.7	1.04
School	53.96c	18.8	53.64c	18.5	55.30b	20.0	0.46	49.23c	18.5	59.18b	17.7	2.96**
Cognitive	38.57e	21.4	38.02e	20.5	40.86c	24.9	1.14	35.02e	21.8	42.49d	20.4	3.44**
With-in subjects F-values	Within-subjects F (4,1376) = 22.16***		Sex X within-subjects F (4,1376) = 4.41**					Age X within-subjects F (4,1376) = 3.64**

Superscript lower case letters designate significant (all ps < .05 adjusted for number of comparisons) differences between means across the five dimensions of health found in follow-up analyses as a result of significant main and interaction effects from within-subjects ANOVA. All t-tests are independent sample t-tests comparing across groups (i.e. male vs. female persons with FXS and 5–10 year olds vs. 11–17 year olds) for each dimension of health

^*p < 0.05;

^**p < .0.01;

^***p < 0.001

There were also differences in the order of dimensions from highest to lowest QoL between age groups. For the younger group (ages 5–10), the order of the dimensions from highest to lowest HRQoL/functioning was the same as the total group (i.e. Physical, Emotional, School, Social, Cognitive functioning), but there were slight differences in the order for the older group (ages 11–17). For the older group, Physical functioning still had the highest scores, followed by Emotional and School functioning with no differences between these two dimensions, then Social functioning having the next lowest scores, and Cognitive functioning with the lowest scores.

Next, we explored sex and age group differences for each dimension (see Tables 3, 4). There was only one sex difference found for the five dimensions. Females with FXS had higher scores in Social functioning than males. Between age groups, which included both males and females, there were two significant differences for the five dimensions. The older group had significantly higher scores for both School and Cognitive functioning than the younger group.

Table 4.

Mean differences across the five dimensions of HRQoL

Dimensions	Quality of life
	M	SD
Physical	63.48a	21.5
Emotional	57.55b	19.3
Social	47.19d	21.1
School	53.96c	18.8
Cognitive	38.57e	21.4

Superscript lower case letters designate significant (all ps < .05 adjusted for number of comparisons) differences between means across the five dimensions of health found in follow-up analyses

Finally, we examined sex and age group differences for both the Caregiver HRQoL variable and the Family HRQoL variable. Females with FXS (M = 52.96, SD = 23.4) had higher Family HRQoL scores than males (M = 44.47, SD = 23.7; t (326) = 2.54, p < .05). There were no sex differences for the Caregiver HRQoL measure (t (329) = 0.65, p ns). There were significant group differences for both Caregiver HRQoL and Family HRQoL variables (t (326) = 3.45, p < .01; t (326) = 2.44, p < .05, respectively). The older group had higher Caregiver HRQoL (M = 55.13, SD = 20.2) and Family HRQoL (M = 49.40, SD = 25.1) scores than the younger group (Caregiver HRQoL M = 47.65, SD = 19.3; Family HRQoL M = 43.03, SD = 22.3).

Discussion

This study describes HRQoL and its associations among the dimensions of health in youth with FXS, their parents/caregivers, and their families, as reported by caregivers on the PedsQL. Notably, caregivers consistently reported across all sex and age groups that their children experienced the highest HRQoL in Physical functioning and the lowest HRQoL in Cognitive functioning. We also found that families of females with FXS had higher reported Family HRQoL scores than families of males with FXS. The generally less impaired phenotype in females with FXS may have translated to higher Family HRQoL. Interestingly, Caregiver and Family HRQoL scores were higher for the older group (11–17) than the younger group (5–10). This is consistent with most studies in ASD on the impact of a child’s age on parental stress (Barker et al. 2011; Fitzgerald et al. 2002; Gray 2002; Lounds et al. 2007), suggesting that families and parents adapt to the behavior and capabilities of their child over time and/or youth with FXS improve in functioning over time thus enhancing Family HRQoL. It also may be that over time treatment may improve HRQoL including ongoing educational, behavioral, occupational, speech therapies, and potentially pharmacotherapy alone or in combination thus resulting in our finding of enhanced HRQoL in the older youth cohort.

Overall, our PedsQL results provide a window into where the greatest perceived impairments are with FXS and how those impairments affect the HRQoL of patients and their caregivers/families. Previous studies robustly have demonstrated the measures’ responsiveness to clinical changes (Desai et al. 2014), validity, and reliability (Varni et al. 2007b). Likewise, our project eliciting a large sample of HRQoL reports may indicate the feasibility of using the measures in this population. Using a variety of measures, several previous studies have examined the HRQoL of mothers of children with FXS (Abbeduto et al. 2004; Lewis et al. 2006; Wheeler et al. 2008a, b) and of patients with FXS (Chevreul et al. 2015; DaWalt et al. 2017). Although the PedsQL has been widely used to assess HRQoL in patients with ASD and ID (Ikeda et al. 2014), this is the first report to date that has used the Peds QL to systematically evaluate HRQoL of patients with FXS and their families.

Despite numerous clinical trials of potential targeted treatments, there is no FDA-approved medication for FXS to date. This is due, in part, to the difficulty the FXS field faces in identifying appropriate standardized outcome measures (Erickson et al. 2017; Berry-Kravis et al. 2017) and lack of validated measures of QoL in FXS that may be sensitive to intervention. Previous trials in FXS have focused on several caregiver proxy report measures of behavior and psychiatric symptoms including the Aberrant Behavior Checklist-Community Edition (ABC-C) (Aman et al. 1995), Anxiety Depression and Mood Scale (ADAMS) (Esbensen et al. 2003), Vineland Adaptive Behavior Scale (VABS) (Sparrow and Cicchetti 1985) and ADHD Rating Scale-IV (Pappas 2006), among others, as well as direct assessments of patient functioning, including the Mullen Scales of Early Learning (MSEL) (Mullen 1995), as primary study outcome measures (Jacquemont et al. 2011; Berry-Kravis 2015; Hess et al. 2016). Although more recent trials have begun to utilize physiological measures such as eye tracking, electroencephalography (EEG), and changes in various blood protein levels for more objective measures of treatment efficacy and target engagement (Erickson et al. 2017), the association between changes in these physiological measures and improvements in functioning level or quality of life is less clear. In the future the establishment of QoL or HRQoL measures with strong psychometric properties measuring personal and family well-being in the FXS field hold promise as potential outcome measures in treatment development.

Waters et al. (2009) have provided a thorough evaluation of instrument use focused on HRQoL evaluation in the neurodisabilities field in youth including specific evaluation of the PedsQL. In choosing an instrument for use, in this case the PedsQL, it is appropriate to understand the strengths and potential weaknesses of the instrument to aid in results interpretation in our pilot work in FXS. While the original purpose of the PedsQL was the evaluation of HRQoL, the focus of the measure on functioning is not directly related in all cases to QoL. The focus on functioning and impairment does elicit some concern that negative wording involved with reporting using the PedsQL could threaten self-esteem. It will be important in future use of the PedsQL in FXS to evaluate parental and when possible individual patient reactions to use of the measure.

Although the primary aim of this study was to evaluate initial feasibility and descriptive characteristics of use of the PedsQL in the FXS population, we note several limitations. First, due to the nature of the online survey, we were unable to confirm diagnosis. We also did not include additional parent-report measures of functioning or other clinical relevant variables that may impact HRQoL. Also, we did not include a direct comparison control groups matched to our FXS sample. We were additionally unable to evaluate the specific potential impact of pharmacotherapy on QoL in youth with FXS. Although the normative PedsQL healthy, acutely ill, and chronically ill child samples had consistently numerically higher mean QoL scores across all domains compared to our FXS youth cohort (Varni et al. 2001a, b), future studies still are needed to study the PedsQL and potential HRQoL measures in families impacted by FXS directly compared to other groups of individuals. For example, a comparison of age and IQ-matched youth with idiopathic ASD or developmental delay without ASD could be appropriate comparison groups for the FXS sample in additional to a more traditional typically developing match control group. We only provided the Parent Report version for the online survey. While a parent likely can provide the best outside report of their child’s experience, we remain unable to capture the individual perspective of the vast majority of youth with FXS who lack the ability to self-report using a measure such as the PedsQL. Our report also does not annotate the presence or absence of comorbid medical conditions/diagnoses that could impact HRQoL outside of FXS. Finally, further work with significant additional data collection and substantial additional analyses will be required to thoroughly establish the psychometric properties of the PedsQL in FXS including work to establish test-retest reliability, validity, and factor structure in this population.

Examining QoL among individuals with FXS is particularly important given the lifelong impact of the disorder, and its influence across contexts and various forms of functioning. Many of the current treatment targets for patients with FXS involve and influence the dimensions of health that comprise HRQoL, including physical, social, and adaptive functioning. Yet, HRQoL and overall well-being are rarely systematically evaluated in clinical or research settings. In clinical trials for children with FXS, assessing the HRQoL of subjects could help identify and evaluate the efficacy of the interventions beyond treating a specific symptom(s) that may present in differing intensity across youth. Likewise, HRQoL measures like the PedsQL may provide a more complete view of the health of children and their family and could be important components of natural history study to assess HRQoL over time. Thus, our study was able to demonstrate the potential of using the PedsQL in this population, but future studies establishing the test-retest reliability over brief time periods, the developmental trajectory of HRQoL in individual youth with FXS, and evaluation of HRQoL in adults with FXS are needed.