Updated profiles of everyday executive function in youth with Down syndrome using the BRIEF-2

Abstract

Background:

Executive function difficulties in youth with Down syndrome (DS) are well recognised using informant-report measures. However, the profile of relative challenges and strengths has not yet been evaluated using the Behavior Rating Inventory of Executive Function, Second Edition (BRIEF-2), which includes a new internal factor structure.

Method:

Using the BRIEF-2, profiles of everyday parent-reported executive function (EF) were evaluated in youth with DS (n=34) and compared to age- and sex-based norms. EF profiles were also compared across raters (parent vs. teacher, n=20) and relative to mental age-matched typically developing (MA-TD) controls (ns=19 in each group).

Results:

Although within group differences were not revealed on indexes, significant differences were found among BRIEF-2 scales. Across raters, teachers reported significantly more difficulties than parents. Compared to MA-TD controls, the DS group was rated more poorly on some but not all BRIEF-2 scales.

Conclusions:

At the scale, but not the index level, the BRIEF-2 identifies a variegated EF profile in children with DS. For several of the scales, significant differences were noted relative to both chronological age expectations (using norms) and mental-age expectations (using a developmentally-matched comparison group). At the scale level, the BRIEF-2 continues to be a sensitive tool for identifying executive function difficulties as well as profiles of relative strengths and weaknesses in children with DS.

Over the past several years, there has been increased interest among researchers studying the Down syndrome (DS) cognitive phenotype in characterising executive function (EF) skills. Broadly, EFs refer to aspects of cognition and behaviour required to successfully execute complex tasks (Denckla, 1996). These higher level abilities are thought to be related yet distinct (Gioia & Isquith, 2004; Miyake et al., 2000). They have been linked to a number of important outcomes. Specific to populations with DS, EFs have been associated with academic achievement and vocational status (Tomaszewski et al., 2018; Will et al., 2017). When EF is assessed with performance-based measures, children with DS generally perform below typically developing peers matched on mental age (MA-TD) on working memory (Carney et al., 2013; Lanfranchi et al., 2012), planning (Fidler et al., 2014; Kasari & Freeman, 2001), and cognitive flexibility/shifting (Costanzo et al., 2010; Edgin et al., 2010) tasks. Research examining inhibitory control with performance-based tasks is mixed, with some studies documenting group differences (Borella et al., 2013; Lanfranchi et al., 2010) and others failing to do so (Carney et al., 2013; Pennington et al., 2003).

Although EF skills have historically been assessed using performance-based measures administered in highly standardised conditions (Toplak et al., 2013), the use of rating scales of everyday EF difficulties has become increasingly common (Roth et al., 2005). Of the existing EF ratings, the Behavior Rating Inventory of Executive Function (BRIEF; Gioia et al., 2000) is the most commonly used (Toplak et al., 2013). The everyday EF profile in youth with DS has been examined using both the BRIEF and its preschool analogue, the BRIEF-P (Gioia et al., 2003), in numerous studies, with the profile of relative strengths and difficulties generally converging with studies employing performance-based measures described above. For example, in comparison to normative data from MA-TD peers on the BRIEF-P, Lee et al. (2011) found parent-reported elevations (i.e., greater difficulties) on the Global Executive Composite and Emergent Metacognition Index along with elevations on the Working Memory and Plan/Organize scales in a sample of young children with DS. Daunhauer and colleagues (2014) further characterised BRIEF-P parent and teacher ratings of children with DS relative to MA-TD peers. They reported challenges consistent with those found in Lee et al. (2011). In addition, parents, but not teachers of children with DS noted greater difficulties on the Inhibitory Self Control Index and the Inhibit scale than those noted for MA-TD peers. Relative strengths in Shift and Emotional Control were also noted by parents and teachers. A similar profile has been found on the school-age version of the BRIEF in samples with DS, with Working Memory indicated as a relative weakness and Emotional Control as a relative strength or an infrequent area of concern (Esbensen et al., 2019; Lee et al., 2015; Loveall et al., 2017).

The original BRIEF included eight scales which were stratified into two indexes: Behavior Regulation (BRI: Inhibit, Shift, Emotional Control) and Metacognition (MI: Initiate, Working Memory, Plan/Organize, Organization of Materials, and Monitor). The BRI and MI combined to create the Global Executive Composite (GEC). A revision of the BRIEF was published in 2015 (Behavior Rating Inventory of Executive Function, Second Edition; BRIEF-2; Gioia et al., 2015) as a result of a study by Gioia and colleagues (2002) which revealed a new internal factor structure. Rather than two indexes, the study by Gioia et al. (2002) found the BRIEF data to best fit a three-factor structure defined by three indexes consisting of nine subdomains: Behavior Regulation (Inhibit, Self-Monitor); Emotion Regulation (Shift, Emotional Control) and Cognitive Regulation (Initiate, Plan/Organize, Organization of Materials, Task-Monitor). These changes are reflected in the BRIEF-2 (Gioia et al., 2015). A breakdown of the structures of the BRIEF and the BRIEF-2 is provided in Figure 1.

Figure 1. Summary of composite, index, and scale score domains for the BRIEF and the BRIEF-2.

Abbreviations: GEC, Global Executive Composite; BRI, Behavior Regulation Index; MI, Metacognition Index; ERI, Emotion Regulation Index; CRI, Cognitive Regulation Index

To the best of our knowledge, no studies to date have examined the DS EF profile using the BRIEF-2. The current study sought to address this gap in the literature by asking three questions. First, what is the parent-rated profile of strengths and weaknesses of everyday EFs in youth with DS using the BRIEF-2 (Q1a), and how do EF difficulties experienced by youth with DS compare to the BRIEF-2's normative sample (Q1b)? Second, how do the profiles/severity of EF challenges in youth with DS vary by rater (parent vs. teacher; Q2); and finally, how do EF difficulties in school-age children with DS compare to those observed in younger children matched on developmental level (i.e., MA-TD controls; Q3)?

Method

All data were collected at Drexel University for a study of reading abilities in DS following procedures approved by the university’s Institutional Review Board.

Procedures

Informed consent was obtained consistent with guidelines in the Declaration of Helsinki. Assent was obtained from participants in a manner that was consistent with developmental level. Larger study procedures included the completion of cognitive testing for child participants and the completion of questionnaires, including the BRIEF-2, by parents and teachers of participants.

Participants

Fifty-three youth participated in the current investigation. Participants with DS (n=34) needed to meet the larger study’s inclusion criteria, which required participants to (a) have a medical diagnosis of DS, (b) be 6 to 17 years of age, and (c) have English as their first language. In addition, to be included in the current investigation, participants needed to have a parent-reported BRIEF-2 questionnaire. Although there was not an explicit inclusion criterion related to expressive language, nearly all participants had phrase speech (n=32; 94%) per parent response to the following question as a part of the initial enrollment phone call: “Does your child speak using phrases of 3 words?”

Of the larger DS sample, a subset (n=20) of participants also had teacher ratings and were the focus of our study question examining rater effects. Lastly, a subset of participants with DS (n=19) was matched to a younger, MA-TD group. MA-TD participants (n=19) were included if they had parent-reported BRIEF-2 ratings, were of a similar mental age to the DS group, and were free of developmental, neurological, and learning disabilities. See Table 1 for participant characteristics and the section that follows that describes mental age matching procedures.

TABLE 1.

Participant demographics

	Age			KBIT-2 IQ Composite Raw Score			KBIT-2 IQ Composite Standard Score			Maternal Ed. (Bachelor's degree)		Sex (Female)		Race1 (W, NH)
	M	SD	Range	M	SD	Range	M	SD	Range	N	%	N	%	N	%
DS group
Whole sample (n=34)	11.44	3.25	6-17	39.26	16.89	9-77	52.35	10.81	40-75	21	62	19	56	25	74
Teacher subsample (n=20)	11.82	3.12	6-17	42.45	17.30	12-77	51.55	9.86	40-71	10	50	11	55	17	85
MA-matched subsample (n=19)	12.31*	2.86	8-17	50.79	10.95	39-77	56.00*	8.64	42-75	13a	72	14	74	13a	72
MA-Matched TD group (n=19)	5.14*	0.95	3-7	50.89	9.97	35-72	113.17*a	10.72	97-130	17	89	9	47	11	58

^*p<.05; denotes statistically significant difference between DS MA-matched subsample and MA-matched TD group

^an=18; missing data for one participant in each of these cases. Maternal education was collected via free-form response. Responses were then stratified into either less than bachelor’s degree or bachelor’s degree or higher to simplify reporting and analyses.

¹Breakdown of race/ethnicity information for the DS (n=34) and MA-Matched TD (n=19) groups are as follows. DS: Asian (n=2); Multiracial (n=2); White, Hispanic (n=4); White, Non-Hispanic (n=25); missing (n=1). TD: Asian (n=1), Multiracial (n=5), White, Hispanic (n=2), White, Non-Hispanic (n=11).

Abbreviations: KBIT-2, Kaufman Brief Intelligence Test, Second Edition; Maternal ed = Maternal education; W, White; NH, Non-Hispanic

Mental-age matching procedures.

In order to complete analyses in which the DS group was compared to a group of younger participants with a similar developmental level – i.e., mental age – the following procedures were implemented. A subset (n=19) of the 34 participants with DS with BRIEF-2 data was matched groupwise to 19 typically developing participants from the larger study using Kaufman Brief Intelligence Test, Second Edition (KBIT-2; Kaufman & Kaufman, 2004) total raw scores. See description of the KBIT-II below in the Measures section.

Following guidelines from Kover and Atwood (2013), both the mean and variance of the KBIT-II raw scores for the DS and MA-TD groups were compared to create equivalently sized groups that had very similar mean KBIT-II raw scores (as ascertained via Cohen’s d) and a similar distribution of scores (as ascertained via a comparison of the variance). These groups each included 19 participants. The DS group had a mean KBIT-II raw score of 50.79 and range of 39-77; the MA-TD group had a mean KBIT-II raw score of 50.89 and range of 35-72. When additional participants with DS or participants in the MA-TD group were included, Cohen’s d became larger, and the variance moved further away from 1. As this was deemed undesirable from a matching/group equivalence perspective, 15 participants with DS and 4 TD-MA participants were excluded. The resulting sample of 38 participants was the largest matched sample of participants with DS and TD-MA controls in which Cohen’s d was closest to 0 (d < .02) and the variance was closest to 1 (per guidance form Kover & Atwood, 2013). Thus, this subsample was the focus of our analyses in which the DS group was compared to MA-TD peers.

Measures

Intellectual Ability.

To estimate overall cognitive ability and match the DS and MA-TD groups, participants completed the KBIT-2 which consists of two verbal (Verbal Knowledge, Riddles) and one nonverbal (Matrices) subtest. For the KBIT-2, no study-wide modifications were made to scoring or starting points. However, testers did use clinical judgment and dropped back to an earlier starting point for participants who had more profound challenges. This latter adjustment was implemented for seven of the 34 participants with DS. Note that none of these participants were included in the subgroup matched on KBIT-2 raw scores to the MA-TD group (described above). Scoring of the instrument followed standard guidelines of establishing basal and ceiling level items to calculate the raw score for each subtest.

Executive Functions.

Parents of youth in both groups and teachers of youth with DS completed the BRIEF-2 (Gioia et al., 2015) within the span of one year (average span = 2 months) following standard administration procedures. The BRIEF-2 assesses behavioural manifestations of EF impairments in youth ages 5-18 years (Gioia et al., 2015) using a 3-point Likert-scale measuring problem frequency: 1= Never; 2= Sometimes; 3=Often.

The BRIEF-2 yields sex- and age-adjusted normative T-scores (M=50, SD=10) for the nine scales, three indexes, and the Global Executive Composite (GEC). Higher T-scores denote greater executive function difficulties. See Table 2 for details about the BRIEF-2 scales and indexes, the psychometric characteristics of the instrument, and example items. Consistent with the strong psychometric characteristics of normative sample, high internal consistency was observed on the BRIEF-2 in the current study’s DS sample (α=0.93; n=34).

TABLE 2.

Summary of scales of the BRIEF-2 parent and teacher forms

Overall Composite 1	Indexes 1	Reliability P: Parent T: Teacher	Clinical Scales1	Description of Clinical Scale	Example Items
Global Executive Composite (GEC)2	Behavior Regulation Index (BRI)	Internal Consistency P: α = .90 T: α = .95	Inhibit	Inhibitory control	“Is fidgety” “Is impulsive”
		Test-Retest P: r = .83 T: r = .83	Self-Monitor	Awareness of effect of behaviour on others/outcomes	“Has poor understanding of own strengths and weaknesses” “Does not realize that certain actions bother others”
	Emotion Regulation Index (ERI)	Internal Consistency P: α = .92 T: α = .94	Shift	Flexibility in moving from one task/thought to another	“Thinks too much about the same topic” “Has trouble thinking of a different way to solve a problem…”
		Test-Retest P: r = .82 T: r = .88	Emotional Control	Ability to regulate emotional responses	“Has explosive, angry outbursts” “Small events trigger big reactions”
	Cognitive Regulation Index (CRI)	Internal Consistency P: α = .96 T: α = .98	Initiate	Ability to begin tasks or generate ideas/strategies	“Is not a self-starter” “Needs to be told to begin a task…”
		Test-Retest P: r = .89 T: r = .89	Working Memory	Ability to hold information in mind to complete tasks	“Has trouble remembering things, even for a few minutes” “Has trouble concentrating…”
			Plan/Organize	Ability to manage task demands	“Lacks follow-through” “Underestimates time needed to finish tasks”
			Task-Monitor	Ability to notice errors in tasks	“Work is sloppy” “Does not check work for mistakes”
			Organization of Materials	Ability to organise work/play/storage spaces	“Does not bring home homework, assignment sheets…” “Cannot find things in desk”

¹The BRIEF-2 provides age- and sex-adjusted T-scores with a M of 50 and SD of 10.

²Psychometric information for GEC: Internal Consistency (P: α = .97; T: α = .98); Test-Retest (P: r = .88; T: r = .90).

For Questions 1 and 2, BRIEF-2 T-scores for the GEC, indexes, and nine clinical scales were used to examine the EF profile in youth with DS relative to chronological age expectations (Q1) and to examine possible rater effects (i.e., parent vs. teacher BRIEF-2 ratings; Q2). Mean item ratings by scale (i.e., the average Likert-scale rating for the items belonging to each scale) were used in analyses for Question 3 in which the level of EF difficulties in participants with DS was compared to MA-TD peers. Mean item ratings by scale were used rather than age- and sex-adjusted normative T-scores for these analyses for the following reasons. First, this study question evaluated raw levels of EF difficulties for the two groups (not difficulty-levels adjusted for age and sex). Thus, un-normed raw scores were most appropriate to evaluate. Second, T-scores could not be calculated for the MA-TD group, given that most participants were too young for T-scores to be generated. Note for Question 3, six of the BRIEF-2 scales were included in analyses. These were the Inhibit, Emotional Control, Working Memory, Initiate, Shift, and Self-Monitor scales. These six scales were deemed developmentally appropriate for our chronologically younger TD group, many of whom were not yet in primary school. The following scales were excluded from analyses involving the younger, MA-TD group due to having items that referred to schoolwork: Plan/Organize, Task-Monitor, and Organization of Materials.

BRIEF-2 Quality Control Procedures:

Protocols were inspected for missing data. Six participants had one question missing from one or more BRIEF-2 scales (which is permitted per the BRIEF-2 manual). We followed procedures outlined in the measure’s manual to address this (i.e., missing items were assigned a score of 1). Additionally, one participant had two items missing on the Plan/Organize scale. As these data were being collected for research and not clinical purposes, we elected to include this participant and used the missing data rule described above. Finally, one participant with DS was excluded due to the number of missing items on the parent-reported BRIEF-2, and one participant’s teacher data were not included due to the number of missing items (this participant was also in the DS group; teacher analyses were not completed for the MA-TD group).

Data were inspected to evaluate the Inconsistency, Infrequency, and Negativity scales of the BRIEF-2. The Inconsistency scale measures the degree to which the respondent answered similar questions inconsistently as compared to the clinical sample. The Infrequency scale measures how often a respondent answers specific questions in an atypical manner. This scale includes three questions that are likely to be answered in a similar fashion by most respondents. Finally, the Negativity scale indicates the degree to which the respondent answered specific questions in a negative manner as compared to the clinical sample. Across the Inconsistency and Infrequency scales, no elevations were noted. Elevations on the Negativity scale occurred across both parent (n=12) and teacher (n=11) reports in the DS sample. We did not exclude these participants, as such elevations are to be expected for a portion of participants, given the well-documented profile of difficulties in EFs within DS.

Statistical Analyses

In addition to the quality control procedures completed for the BRIEF-2 described above, data were inspected for outliers and floor/ceiling effects. No outliers were noted. Floor/ceiling effects and instances of missing data were minimal. See Supporting Information for details. Data were visually inspected to evaluate normality and the presence outliers (i.e., scores > 3 SDs from the mean). Data were deemed to be normal based on visual inspection, and no outliers were noted. Descriptions of study questions, groups examined, scores used, and analyses implemented are provided in Table 3.

TABLE 3.

Summary of study questions, groups included, scores evaluated, and analytic technique employed

Study Question and Group(s)	Scores Used & Analyses Completed
Question 1: What is the parent-rated profile of everyday EF skills on the BRIEF-2 in school-age children with DS (a) and are scores elevated relative to normative means (b)? Group: Participants with DS; n=34	Scores Used: Age- and sex-adjusted normative T-scores from the BRIEF-2 (M = 50; SD =10) Analyses completed: 1a. Two repeated measures ANOVAs completed: One for index (BRI, ERI, CRI); One for scale (IH, SM, SH, EC, IN, WM, PO, TM, OM); Scale main effect followed up with 36 paired samples t-tests1 1b. Thirteen, one-sample t-tests comparing DS group mean to the normative mean for the GEC, BRI, ERI, CRI, and the 9 scales1
Question 2: Does the profile and severity of EF difficulties on the BRIEF-2 differ for parents and teachers? Group: DS subsample with data from parent and teacher; n=20	Scores Used: Age- and sex-adjusted normative T-scores from the BRIEF-2 (M = 50; SD =10) Analyses completed: Two repeated measures ANOVAs with rater (teacher, parent) and either index score (BRI, ERI, CRI) or scale (IH, SM; SH; EC; IN; WM; PO; TM; OM) as the within-subjects factors completed Main effect rater observed; Main effect of scale observed, followed-up with 36 paired samples t-tests averaged across rater1; additionally, 13 paired samples t-tests were completed to compare teacher and parent ratings on each index and scale1
Question 3: Does the severity of EF difficulties observed on the BRIEF-2 in youth with DS differ from typically developing controls matched on mental age (MA-TD)? Groups: DS subsample2 with similar mental age to MA-TD3 sample; ns=19	Scores Used: Mean item rating (raw score, unadjusted for age/sex) for six scales examined (item ratings range from 1 – 3) Analyses completed:  One mixed-model ANOVA with group (MA-matched DS and TD) as the between-subjects factor and scale (IH, EC IN, WM, SH, SM)4 as the within-subjects factor completed  Group x scale interaction observed; followed-up with 6 independent samples t-tests to compare the DS and TD groups on the individual BRIEF-2 scales1 and 15 paired samples t-tests within each group separately to compare each scale to every other scale1

Notes: ¹Multiple comparisons controlled for with Holm-Bonferroni adjustment; ²Thirteen participants with DS could not be matched to the TD sample due to KBIT-2 raw scores that were too low to be matched. ³Three TD participants were not matched due to having KBIT-2 raw scores that were too high to be matched to the DS group. ⁴Note for this analysis, six of the BRIEF-2 scales were included, as these scales were deemed developmentally appropriate for our chronologically younger TD group, many of whom were not yet in primary school. The following scales were excluded due to having items that referred to schoolwork: PO, TM, OM.

Key: Global Executive Composite [GEC]; Behavior Regulation Index [BRI]; Emotion Regulation Index [ERI]; Cognitive Regulation Index [CRI]; Inhibit [IH]; Self-Monitor [SM]; Shift [SH]; Emotional Control [EC]; Initiate [IN]; Working Memory [WM]; Plan/Organize [PO]; Task-Monitor [TO]; Organization of Materials [OM]

Results

Results are organised by study question which can be found in Table 3.

Parent-Reported BRIEF-2 T-Score Profile (Q1a)

The repeated measures ANOVA comparing index scores was not significant (F(2,66)=1.46, p>0.05, partial η² =0.04), indicating similar levels of impairment across indexes. In contrast, the results of the repeated measures ANOVA comparing the BRIEF-2 scales revealed a significant main effect of scale (F(8,264)=13.61, p<0.001, η²=0.29), indicating a variegated profile. To explore this effect, each scale was compared to all other scales, resulting in the completion of 36 paired samples t-tests. With Holm-Bonferroni correction, these t-tests revealed the following. Participants with DS were rated on Emotional Control as having fewer difficulties than: Inhibit, t(33)=−3.91, p<0.001, Cohen’s d=−0.67; Self-Monitor, t(33)=−4.63, p<0.001, Cohen’s d=−0.79; Shift, t(33)=−7.60, p<0.001, Cohen’s d=−1.30; Initiate, t(33)=−3.90, p<0.001, Cohen’s d=−0.67; Working Memory, t(33)=−6.13, p<0.001, Cohen’s d=−1.05; Plan/Organize, t(33)=−3.66, p=0.001, Cohen’s d=−0.63; Task-Monitor, t(33)=−6.61, p<0.001, Cohen’s d=−1.13. Additionally, participants with DS had fewer difficulties on Organization of Materials than: Inhibit, t(33)=−3.50, p=0.001 Cohen’s d=−0.60; Self-Monitor, t(33)=−5.23, p<0.001, Cohen’s d=−0.90; Shift, t(33)=−6.00, p<0.001, Cohen’s d=−1.03; Initiate, t(33)=−5.88, p<0.001, Cohen’s d=−1.01; Working Memory, t(33)=−7.84, p<0.001, Cohen’s d=−1.35; Plan/Organize, t(33)=−4.29, p<0.001, Cohen’s d=−0.74; Task-Monitor, t(33)=−6.38, p<0.001, Cohen’s d=−1.09. Finally, participants with DS were reported to have fewer difficulties on Inhibit than Shift, t(33)=−3.88, p<0.001, Cohen’s d=−0.67. See Figure 2.

Figure 2. DS BRIEF-2 T-Score Profile.

Panel a, top: Age- and sex-adjusted mean T-scores for parent report on the BRIEF-2 for the DS group (complete sample; n=34). Mean T-scores for the DS group are provided with demarcations for the instrument’s normative mean as well as the clinically elevated and possibly clinically elevated cut off scores. Note that when the DS profile on the BRIEF-2 was examined using repeated measures ANOVA, a main effect of scale emerged. To explore this, thirty-six paired samples t-tests comparing each scale to every other scale were completed. Holm-Bonferroni corrected statistically significant differences were as follows. EC fewer difficulties than: IH, SM, SH, IN, WM, PO, TM; OM fewer difficulties than: IH, SM, SH, IN, WM, PO, TM; IH fewer difficulties than: SH. See text for details.

Panel a, bottom: T-, p-, and Cohen’s d statistics for one-sample T-test results comparing the DS group mean to the normative sample mean of 50. In addition to the t, p, and Cohen’s d statistics, the percentages of the sample that received scores in the clinically elevated and possibly clinically elevated ranges are provided.

Panel b, top: Age- and sex-adjusted T-scores for subset of participants with both parent- and teacher-rated BRIEF-2 data (n=20). Mean T-scores per parent and teacher report are provided for the DS group with demarcations for the instrument’s normative mean as well as the clinically elevated and possibly clinically elevated cut off scores. Note that when the parent and teacher ratings on the BRIEF-2 were examined using repeated measures ANOVA, a main effect of rater emerged: teachers > parents (difficulties). A main effect of scale was also observed; thus, thirty-six paired sample t-tests comparing each scale to every other scale (averaged across rater: teachers + parents) were completed. Holm-Bonferroni corrected statistically significant differences were as follows. IH greater difficulties than: OM; SM greater difficulties than: OM; SH greater difficulties than: EC, PO, & OM; IN greater difficulties than: OM; WM greater difficulties than: EC, PO, & OM; PO greater difficulties than: OM; TM greater difficulties than PO & OM. See text for details.

Panel b, bottom: T-, p-, and Cohen’s d statistics for independent-samples T-test results comparing the parent and teacher raters on BRIEF-2 GEC, indexes, and scales. In addition, the percentages of the sample that received scores in the clinically elevated and possibly clinically elevated ranges for parent and teacher are provided.

Abbreviations: Global Executive Composite [GEC]; Behavior Regulation Index [BRI]; Emotion Regulation Index [ERI]; Cognitive Regulation Index [CRI]; Inhibit [IH]; Self-Monitor [SM]; Shift [SH]; Emotional Control [EC]; Initiate [IN]; Working Memory [WM]; Plan/Organize [PO]; Task-Monitor [TO]; Organization of Materials [OM] ; Note: Normative M=50; SD=10; Higher scores denote greater difficulties.​ C%: percentage of sample that received scores in the clinically elevated range; PC%: percentage of sample that received scores in the possibly clinically elevated range

Notes: Higher scores indicate greater difficulties; six of nine scales presented because three were deemed age-inappropriate for MA-TD group.

*Denotes that p-value survived Holm-Bonferroni correction.

Parent-Reported BRIEF-2 T-scores Relative to Norms (Q1b)

When comparing the GEC and index T-scores to the normative sample using Holm-Bonferroni corrected one-sample t-tests, the DS group was rated as having significantly more difficulties on the GEC and three indexes (ts > 5, ps < .001; see Figure 2). The same pattern was found for seven of the nine scales (ts > 4, ps < .001; see Figure 2), with the two exceptions being the Organization of Materials and Emotional Control scales (which did not differ from the normative mean). See Figure 2 for the t-, p-, and Cohen’s d statistics for these one-sample t-tests, and for the percentage of participants receiving scores at or above the BRIEF-2 cut-offs for ‘clinically elevated’ and ‘possibly clinically elevated’ scores. T-scores of 70 or more are considered to be ‘clinically elevated’, and T-scores of 65 to 69 are considered to be ‘possibly clinically elevated’.

Parent vs. Teacher BRIEF-2 T-scores (Q2)

The repeated measures ANOVA with rater (teacher, parent) and index as the within-subjects factors revealed a main effect of rater (Marginal Means: teacher=65.98; parent=60.03; higher scores=greater difficulties; F(1,19)=8.14, p=0.01, partial η²=0.30); however, there was not a main effect of index (F(2,38)=0.20, p>0.05, partial η²=0.01) nor an interaction between rater and index (F(2,38)=1.21, p>0.05, partial η²=0.06). A second repeated measures ANOVA was conducted with the nine scales and rater as the within-subjects factors. Again, a main effect of rater was found (Marginal Means: teacher=64.17; parent=59.30; higher scores=greater difficulties; F(1,19)=7.63, p=0.01, partial η²=0.29). There was also a significant effect of scale (F(8,152)=13.23, p<0.001, partial η²=0.41), but no rater x scale interaction (F(8,152)=1.56, p>0.05, partial η²=0.08).

The effect of scale was followed up with thirty-six paired samples t-tests comparing each scale to every other scale (averaged across rater: teachers and parents) with Holm-Bonferroni correction for multiple comparisons. Results were as follows. Inhibit greater difficulties than: Organization of Materials (t(19)=4.46, p<0.001, Cohen’s d=1.00). Self-Monitor greater difficulties than: Organization of Materials (t(19)=6.50, p<0.001, Cohen’s d=1.45). Shift greater difficulties than: Emotional Control (t(19)=4.37, p<0.001, Cohen’s d=0.98), Organization of Materials (t(19)=7.61, p<0.001, Cohen’s d=1.70), and Plan/Organize (t(19)=6.51, p<0.001, Cohen’s d=1.46). Initiate greater difficulties than: Organization of Materials (t(19)=8.21, p<0.001, Cohen’s d=1.84). Working Memory greater difficulties than: Plan/Organize (t(19)=4.89, p<0.001, Cohen’s d=1.09), Emotional Control (t(19)=4.26, p<0.001, Cohen’s d=0.95), and Organization of Materials (t(19)=10.23, p<0.001, Cohen’s d=2.29). Task-Monitor greater difficulties than: Organization of Materials (t(19)=6.30, p<0.001, Cohen’s d=1.41) and Plan/Organize (t(19)=3.97, p=0.001, Cohen’s d=0.89). Plan/Organize greater difficulties than: Organization of Materials (t(19)=4.58, p<0.001, Cohen’s d=1.02). See Figure 2.

In order to fully document differences in parent-teacher ratings, an additional 13 paired samples t-tests were completed to compare teacher and parent ratings on each index and scale (with Holm-Bonferroni for multiple comparisons). When unadjusted p-values were considered, teachers rated participants as more impaired on the GEC, ERI, BRI, and Self-Monitor, Emotional Control, and Working Memory scales. However, none of these survived correction for multiple comparisons. See Figure 2 for ts, ps, ds.

DS vs. MA-TD BRIEF-2 Raw Scores (Q3)

A two-way mixed-model ANOVA was conducted with group (DS, MA-TD) as the between-subjects factor and mean raw item ratings from the six age-appropriate BRIEF-2 scales as the within-subjects factor. There were significant effects of both group (Marginal Means: DS=1.79; TD=1.43; higher scores=greater difficulties; F(1,36)=14.48, p=0.001, partial η²=0.29) and scale (F(5,180)=4.35, p≤.001, partial η² =0.11); additionally, a significant group x scale interaction was observed F(5,180)=7.90, p<0.001, partial η²=0.18). To explore the group x scale interaction, six Holm-Bonferroni corrected independent samples t-tests were completed. They indicated that compared to the MA-TD group, the DS group had significantly greater difficulties on the Initiate, Working Memory, and Shift scales. The groups did not significantly differ on the Self-Monitor, Inhibit, or Emotional Control scales. See Figure 3. Next, within-group tests of simple effects (i.e., paired samples t-tests) comparing each scale with every other scale (in each group separately) were completed and yielded the following statistically significant differences after Holm-Bonferroni correction. Participants with DS had fewer difficulties on Emotional Control than: Shift, t(18)=−5.22, p<0.001, Cohen’s d=−1.20; Initiate, t(18)=−4.92, p<0.001, Cohen’s d=−1.13; Working Memory, t(18)=−4.65, p<0.001, Cohen’s d=−1.07. Paired samples t-tests did not reveal significant differences between scales for the MA-TD group.

Figure 3. DS vs. MA-TD BRIEF-2 Mean Item Ratings.

Top: Mean item rating for six of BRIEF-2 scales for DS and MA-TD control groups.

Bottom: T-, p-, and Cohen’s d statistics for independent samples t-tests comparing DS and MA-TD mean item ratings for each of the scales examined. A mixed model ANOVA comparing the profile of scores on the BRIEF-2 for DS and TD groups revealed a significant group x scale interaction. Between group tests of simple effects (i.e., independent samples t-tests) are summarised above. Within-group tests of simple effects (i.e., paired samples t-tests) comparing each scale with every other scale (in each group separately) yielded the following statistically significant differences after Holm-Bonferroni correction: participants with DS had fewer difficulties on Emotional Control than: SH, IN, WM. Paired samples t-tests did not reveal significant differences between scales for the MA-TD group. See text for details.

Abbreviations: Inhibit [IH]; Self-Monitor [SM]; Shift [SH]; Emotional Control [EC]; Initiate [IN]; Working Memory [WM]

Notes: Higher scores indicate greater difficulties; six of nine scales presented because three were deemed age-inappropriate for MA-TD group.

*Denotes that p-value survived Holm-Bonferroni correction for multiple comparisons.

Discussion

The current study is the first to use the BRIEF-2 with youth with DS to (a) evaluate everyday EF profiles, (b) contrast impairment across home and school settings, and (c) compare performance to MA-TD peers. Findings were as follows. With regard to BRIEF-2 profiles, index-level T-scores within the DS group were similar to one another but elevated (i.e., revealed greater difficulties) relative to normative expectations. In contrast, the BRIEF-2 clinical scale profile was characterised by both elevations relative to norms on most scales and variability between scales, in line with prior literature on the BRIEF and BRIEF-P. Consistent with Esbensen et al. (2019), Lee et al. (2015), and Loveall et al. (2017), youth with DS showed relative strengths on the Emotional Control and Organization of Materials scales. These scales did not differ from the normative mean. In contrast, T-scores revealed greater EF difficulties relative to norms on all other scales. Of these, consistently higher scores were noted on Shift; it was the only scale to reach subclinical and clinical elevations for parent- and teacher-report, respectively. Moreover, Shift (along with Working Memory and Initiate) was found to be elevated in excess of overall developmental level, as children with DS had more difficulties than MA-TD children.

Consistent with prior work by Daunhauer and colleagues (2014), teachers of youth with DS reported significantly more EF difficulties than parents. Also consistent with prior work, Emotional Control did not differ from MA-TD peers (Daunhauer et al., 2014; Lee et al., 2011). Inhibit was also found not to differ from MA-TD children in the current investigation, but this contrasted with findings from Daunhauer and colleagues (2014) in which parent-endorsed Inhibit weaknesses were reported. This discrepancy may be due to differences in BRIEF-2 versions (preschool vs school-age) or in the ages of the samples. From a developmental standpoint, expectations vary for inhibition skills. These varied developmental expectations may be reflected in the types of questions included on the two instruments. Although there is some overlap in questions, the BRIEF-P includes 16 Inhibit questions while the BRIEF-2 includes 8. Thus, these differences in the two instruments could contribute to the inconsistency with Daunhauer et al. (2014). Moreover, differences in the ages of the sample may be contributing to the inconsistency. The current sample (M age = 11.44 years) was chronologically older than the Daunhauer sample (M age = 8.05 years). Thus, inhibiting responses may become less problematic for youth with DS as they age. The nature of these two studies precludes evaluation of this possibility. Future research that employs a longitudinal study design is needed to evaluate this and describe the developmental trajectory of inhibition skills as youth with DS age.

The results of the current study, particularly the examination of everyday EF difficulties across settings/raters, have significant clinical implications for the identification of EF domains that could be the target of intervention in future research. Our results suggest future interventions could emphasise relative strengths in emotional regulation while supporting weaknesses in shifting/cognitive flexibility. Such focused interventions may have positive cascading effects on adult outcomes, as prior research has demonstrated EF skills are associated with employment outcomes in DS (Tomaszewksi et al., 2018). Promising interventions exist from the field of general education (e.g., Tools of the Mind; Bodrova & Leong, 2006) and neurodevelopmental disorders targeting EF difficulties. For example, “Unstuck and On Target” (Cannon et al., 2011) is demonstrated to improve flexibility in youth with autism spectrum disorder (Kenworthy et al., 2014). Given the cognitive flexibility difficulties highlighted here on both parent and teacher ratings of participants with DS, adaptations of such interventions may be worth exploring in future intervention research for this group.

With regard to study limitations, the investigation’s small sample limited power and the detection of small-to-medium effects. Additionally, due to the young age of the MA-TD participants, several of the BRIEF-2 scales were deemed age-inappropriate, limiting our ability to examine the complete BRIEF-2 profile relative to mental age expectations. An additional limitation is that parent and teacher ratings were not completed on the same day. Thus, situational factors across time may have influenced differences seen across raters. Lastly, the current study included only a typically developing comparison group, precluding evaluation of the syndromic-specificity of the DS BRIEF-2 profile.

Acknowledging these limitations, the current study extends knowledge about everyday EF profiles in youth with DS by comparing BRIEF-2 EF ratings to both chronological and mental age expectations and contrasting parent and teacher ratings. By comparing across both chronological and mental age, we were able to gain greater insight into which EF domains are impaired relative to age and developmental level. Further, by documenting that teachers observed greater EF challenges than parents during the school-age period, this study underscores the importance of EF skills to classroom behaviour and academic performance in DS. It is our hope that this may contribute to research and practice on educational strategies for youth with DS in the classroom.