Cognitive Disengagement Syndrome in Pediatric Spina Bifida

Abstract

Objective:

The primary aim of this study was to characterize Cognitive Disengagement Syndrome (CDS) symptomatology in youth with spina bifida (SB.)

Methods:

169 patients aged 5 to 19 years old were drawn from clinical cases seen through a multidisciplinary outpatient SB clinic at a children’s hospital between 2017 and 2019. Parent-reported CDS and inattention were measured using Penny’s Sluggish Cognitive Tempo Scale and the Vanderbilt ADHD Rating Scale. Self-reported internalizing symptoms were measured with the 25-item Revised Children’s Anxiety and Depression Scale (RCADS-25).

Results:

We replicated Penny’s proposed 3-factor structure of CDS with slow, sleepy, and daydreamer components. The slow component of CDS overlapped heavily with inattention, while the sleepy and daydreamer components were distinct from inattention and internalizing symptoms. 18% (22 of 122) of the full sample met criteria for elevated CDS, and 39% (9 of 22) of those patients did not meet criteria for elevated inattention. Diagnosis of myelomeningocele and presence of a shunt were associated with greater CDS symptoms.

Conclusions:

CDS can be measured reliably in youth with SB and can be discriminated from inattention and internalizing symptoms in this population. ADHD rating scales measures fail to identify a substantial portion of the SB population with attention-related challenges. Standard screening for CDS symptoms in SB clinics may be important to help identify clinically impairing symptoms and design targeted treatment plans.

Spina bifida (SB) is a chronic condition that has a multisystemic impact on the physical, neurocognitive, social, and psychological functioning of children and adolescents (Alriksson-Schmidt, Wallander, & Biasini, 2007; Holmbeck & Devine, 2010). Youth with SB are at risk for orthopedic, bowel/bladder, neurological, and psychological challenges (Copp et al., 2015). However, there is a wide range of outcomes, and several markers of disease severity have been linked to morbidity, including type of spina bifida, lesion level, and hydrocephalus, Hydrocephalus occurs in approximately 80% of children with the most common and severe type of SB, myelomeningocele (SBM), with the large majority requiring shunting (Sqouros, 2005). SBM with hydrocephalus (SBM-H) has been associated with a range of cognitive weaknesses, and attention has been identified as a core impacted domain (Dennis & Barnes, 2010). Individuals with SBM-H have elevated rates of attention deficit-hyperactivity disorder (ADHD), particularly the inattentive subtype (Burmeister et al., 2005). However, there is evidence that the nature of attention problems differs between SBM-H and idiopathic ADHD (Dennis & Barnes, 2010, Burmeister et al, 2005). In the broader field of attention disorders, there has been mounting interest in a construct called sluggish cognitive tempo (SCT), a term that was recently revised to cognitive disengagement syndrome (CDS; Becker et al., 2022), which may help to describe some of the attention impairments in SBM-H, but to date, only one published manuscript has characterized CDS in this population (Smith, Castillo, Clark, & Holmbeck, 2022). The purpose of this study was to add to this literature by characterizing CDS symptomatology in youth ages 5 to 19 followed in a multidisciplinary spina bifida clinic using a measure developed specifically to assess CDS.

Neurocognitive functioning in SB

An influential neuropsychological model proposed by Dennis and Barnes (2010) argued that SBM-H is characterized by both core neurocognitive deficits as well as cross-domain processing biases leaving some aspects of thinking intact and others impaired. The proposed core deficits are in attention, timing, and movement; although evidence supports these as part of the modal profile of SBM-H, there is considerable inter-individual variability and not all individuals with SBM-H show weaknesses in all core areas (Dennis et al., 2006; Dennis & Barnes, 2010).

Within the core domain of attention, skills are not uniformly impaired. Individuals with SBM-H tend to have greater impairment with engaging and shifting of attention than with sustaining attention (Burmeister et al., 2005; Dennis et al., 2006). Even though a significant portion of individuals with SBM-H meet criteria for Attention Deficit/Hyperactivity Disorder – Inattentive Presentation (ADHD-I), idiopathic ADHD is associated with the opposite pattern of attentional difficulties (i.e., more difficulties sustaining attention than with initial engagement) (Burnmeister et al., 2005; Dennis & Barnes, 2010).

Compared to individuals with SBM-H, individuals with SBM without shunted hydrocephalus and those with other forms of SB (e.g., meningocele and lipomyelomeningocele) are at relatively lower risk for neurocognitive weaknesses, though they remain at risk for disruptions to orthopedic, bowel/bladder, and psychosocial functioning (Hampton et al., 2011; Iddon et al., 2004). Higher lesion level is a marker of disease severity that has consistently been linked to greater morbidity within the SB population as a whole (Copp et al., 2015) as well as to higher risk for neurocognitive challenges in SBM-H (Fletcher et al., 2005.)

Sluggish Cognitive Tempo

CDS is a construct related to, yet partly distinct from, inattention and internalizing symptoms and includes difficulties with initiation/motivation, sluggishness/sleepiness, and daydreaming (Becker et al., 2013). Although CDS was initially thought to fall under the umbrella of attention problems seen in ADHD, studies over the past two decades both in adults (Barkley, 2012) and children (Penny et al., 2009; Jacobson et al., 2012; Barkley 2013) have found that CDS symptoms comprise a partly separable set of dimensions from the traditional two ADHD symptom dimensions (i.e., inattention and hyperactivity/impulsivity). In a large nationally representative sample of children and adolescents in the United States, Barkley (2013) found that 59% of children meeting criteria for clinical CDS also had clinically elevated ADHD while 39% of children meeting criteria for clinical ADHD also had clinically elevated CDS. Similarly, Severa, Saiz, Burns, & Becker, (2018) found that, across mother, father, and teacher informants, 44–45% of children with elevated CDS met criteria for elevated ADHD and 27–35% of children with elevated ADHD met criteria for elevated CDS. CDS has also been associated with internalizing psychopathology (Becker & Willcutt, 2019) including social isolation and withdrawal (Becker et al., 2019; Marshall et al., 2014, Wilcutt et al., 2014), depressive symptoms (Becker, Webb, & Dvorsky, 2019), and conflicted shyness (Becker et al., 2018; Saez et al., 2019).

Previous factor analyses (Jacobson et al., 2012; Barkley 2013) have identified some domains of CDS that are relatively more independent from ADHD symptoms than others. The CDS dimensions labeled “sluggish “and “daydreaming” were more related to each other than each was to inattention and hyperactivity (Barkley, 2013). Though results have been mixed, elevated CDS has been associated with demographic factors such as older age, lower household income and parent education (Becker et al., 2016, Smith et al., 2022), and – in contrast to a male predominance among individuals with ADHD – an approximately equal gender prevalence (Becker et al., 2016).

There has been growing interest in investigating CDS in several medical populations including children with prenatal alcohol exposure (Graham et al., 2013), acute lymphoblastic leukemia (Reeves et al., 2007), and epilepsy (Loutifi, Carvalho, Lamouneir, & Nascimento, 2011). In all these populations, elevated rates of CDS were found compared to healthy controls (Graham et al., 2013; Reeves et al., 2007; Loutifi, Carvalho, Lamouneir, & Nascimento, 2011). Of particular relevance to SBM-H, hydrocephalus was associated with CDS symptoms among pediatric brain tumor survivors (Williard et al., 2013). Characteristics of CDS, including slow initiation and engagement, are similar to the type of attentional challenges described in individuals with SBM-H. CDS may therefore be an important construct for understanding cognitive and psychosocial impairment among children with SBM, but only one published study to date has investigated CDS in this population. Smith and colleagues (2022) recently reported that among youth with SB, greater neurocognitive difficulties were associated with higher levels of four CDS symptoms on the broadband Child Behavior Checklist. Furthermore, worse working memory and cognitive flexibility at baseline predicted an increase in CDS symptoms as youth with SB age (Smith et al., 2022).

The current study

The primary aim of this study was to characterize CDS symptomatology in children ages 5 to 19 followed in a multidisciplinary spina bifida clinic and compare the level of CDS symptomatology in this population to multiple unselected and clinical samples in the extant literature that reported descriptive statistics on the Penny et al. (2009) scale or a similar scale with overlapping CDS items. We had three hypotheses. First, we predicted that symptoms of CDS would be associated with, but still distinct from, inattention and internalizing symptoms. Second, we hypothesized that the level of CDS would be elevated in this population compared to community and clinically referred samples. Lastly, we hypothesized that higher CDS symptoms would be associated with markers of disease severity. Based on previous research into neurocognitive challenges in SB, we predicted that diagnosis of myelomeningocele, higher lesion level, and presence of shunt would all be linked to higher CDS symptoms. We also tested for a relationship between ambulation difficulties and CDS symptoms since the broader SB literature often includes ambulation status as a measure of disease severity (e.g., Holmbeck, Willis, & Coers, 1999, Bellin et al., 2012).

Methods

Participants

The current project was a retrospective chart review of clinical data collected as part of a larger study (the National Spina Bifida Patient Registry, sponsored by the Centers for Disease Control and Prevention) approved by the university-affiliated institutional review board. Patients were drawn from clinical cases seen through a multidisciplinary outpatient spina bifida clinic at a children’s hospital between 2017 and 2019. Patients were considered eligible if they were between ages 5 and 19 years old at the time of their clinic appointment and had a diagnosis of SB (types included myelomeningocele, lipomyelomeningocele, and meningocele). This yielded a total of 169 patients. Of these patients, 66% had SBM-H, 50% were female, and 43% identified as Hispanic. Further information about the sample is included in Table 1.

Table 1.

Sample demographics and characteristics

Characteristic	Range (M ± SD) or % (n)
Age, range (M ± SD)	5–19 (11.0 ± 3.6)
Gender, % (n)
Female	50.3 (n =85)
Male	49.7 (n =84)
Race, % (n)
White	71% (n =120)
Black	1% (n =1)
Asian	8% (n =13)
Multiracial	4% (n =7)
Other	15% (n =25)
Not reported	2% (n =3)
Ethnicity, % (n)
Hispanic	43% (n =72)
Non-Hispanic	56% (n =95)
Not reported	1% (n =2)
Spoken Language, % (n )
English	80% (n =135)
Spanish	20% (n =34)
Insurance, % (n )
Medicaid or similar	63% (n =104)
Private	37% (n =62)
Shunt status, % (n)
No	34% (n =58)
Yes	66% (n =111)
Spina bifida diagnosis, % (n)
Myelomeningocele	83% (n =135)
Lipomyelomeningocele	14% (n =23)
Meningocele	3% (n =4)
Lesion level, % (n)
Sacral	28% (n =46)
Low lumbar	9% (n =15)
Mid lumbar	49% (n =80)
High lumbar	6% (n =9)
Thoracic	7% (n =12)
Parent-rated ambulatory status, % (n)
Non-ambulator	28% (n =44)
Therapeutic/household	13% (n =20)
Community	60% (n =94)
Symptom ratings, range (M ± SD)
CDS (total)	0–19 (4.5 ± 4.3)
Inattention (total)	0–27 (9.7 ± 6.6)
Anxiety (T-score)	27–72 (43.7 ± 9.4)
Depression (T-score)	29–80 (41.9 ± 9.0)

Measures

Sluggish Cognitive Tempo (SCT) Scale (Penny et al., 2009).

The SCT scale is a 14-item teacher- or parent-report rating scale of symptoms that correspond to the CDS construct. For the present study, only parent ratings were used, and the items are ordered as they appear in Table 2 of Penny et al. (2009). Ratings are made on a four-point Likert scale (ranging from 0 = Never to 3 = Very Often). Total CDS composite score is the sum of the ratings on all 14 items. Internal consistency of the total scale in this clinical sample was excellent (α = .92).

Table 2.

Confirmatory and Exploratory Factor Analyses of Cognitive Disengagement Syndrome and Inattention

Model	df	χ2	p	CFI	TLI	RMSEA [90% CI]	MC	df	χ2	p
14-item CDS CFA							EFA Model Comparisons
3 Factors	74	142.9	<.001	.985	.981	.087 [.066, .109]	–	–	–	–
14-item CDS & IN EFAs
2 Factors	208	507.8	<.001	.955	.946	.097 [.086, .108]	–	–	–	–
3 Factors	187	291.3	<.001	.984	.979	.060 [.047, .074]	3v2 Factor	21	125.5	.000
8-item CDS & IN EFAs
2 Factors	103	372.9	<.001	.947	.930	.131 [.117, .145]	–	–	–	–
3 Factors	88	140.8	<.001	.990	.984	.063 [.043, .081]	3v2 Factor	15	129.5	.000

Note. CDS = cognitive disengagement syndrome; IN = inattention; CFI = comparative fit index; TLI = Tucker-Lewis Index; RMSEA = root-mean-square error of approximation; CI = confidence interval; MC = model comparison; C/EFA = confirmatory/exploratory factor analysis. WLSMV estimator used, items modeled as categorical.

Vanderbilt ADHD Diagnostic Parent Rating Scale (VADPRS).

The VADPRS (Wolraich et al., 2003) asks parents to rate the frequency of occurrence for each of the 18 DSM-V symptoms of ADHD on a four-point Likert scale (ranging from 0 = Never to 3 = Very Often). The scale is broken down into two subscales measuring inattentive and hyperactive/impulsive behaviors. For the present study, only the inattentive (IN) subscale (items 1–9; N = 151) was completed by caregivers, since problems with IN are common in SB, while problems with hyperactivity/impulsivity are not (De la Torre, Martin, Cervantes, Guil, & Mestre, 2017).

Revised Child Anxiety and Depression Scale (RCADS).

Self-reported internalizing symptoms were measured with the short form (25-item version) of the RCADS (Chorpita, Yim, Moffitt, Umemoto, & Francis, 2000). The RCADS 25-item version contains 15 anxiety items and 10 depression items that are totaled and converted to standard T-scores based on child gender and grade level (3 to 12). The shortened RCADS-25 has demonstrated acceptable reliability in clinic-referred and school-based samples (Ebesutani et al., 2012). In the present study, 119 youth aged 3 to 12 completed the RCADS-25.

Analytic Strategy

Analyses were conducted using the Mplus statistical software package (Version 8.1.5; Muthen & Muthen, 2012).

Measurement of CDS in SB.

Penny’s et al. (2009) 14-item SCT scale was first subjected to confirmatory factor analysis (CFA) to replicate the originally proposed three-factor structure. Given prior studies (Jacobson et al., 2012; Barkley 2013) that have found inadequate discriminant validity for this model in relation to IN, including Penny’s et al. (2009) original study, we anticipated this initial CFA would provide adequate fit but insufficient discrimnant validity when IN was also included in the model. As such, Penny’s SCT scale and the 9 IN items were next subjected to exploratory factor analysis (EFA), in which all items were pemitted to load on all latent factors, in order to test the generalizability of Penny’s et al. two and three factor CDS and ADHD principal component analyses. We required indicators to demonstrate loadings of at least .60 to be assigned to a primary factor (Becker et al., 2016; Willcutt et al., 2014). Items were considered to be cross-loaded if loadings on another factor were greater than .30 and within .20 of the item’s primary factor loading (Willcutt et al., 2014).

For item-level analyses, items were treated as ordered categorical manifest variables using the robust weighted least estimator (WLSMV). For structural equation model analyses using parcels (see below), parcels were treated as approximately continuous manifest variables using the robust maximum likelihood estimator (MLR). Robust estimation was used for both types of analyses in order to adjust for the non-normality that is characteristic of symptom ratings data. Model fit was assessed with the robust comparative fit index (CFI; study criteria of at least .90, with ≥ .95 being ideal), the Tucker-Lewis Index (TLI; study criteria ≥ .90), and the robust root-mean-square error of appoximation (RMSEA; study criteria ≤ .08).

Some analyses used parcels, whereby individual items are combined (e.g., taking the mean of multiple items) and the new composites are used as manifest variables. Parceling reduces the amount of unreliable (i.e., error) and item-specific variance and decreases the likelihood of Type II errors (Little, 2013; Little, Rhemtulla, Gibson, & Schoemann, 2013). Unstandardized loadings from the categorical confirmatory factor analyses were used to assign items to a parcel that would maximize the likelihood of homogenous parcels (Burns, Servera, Bernad, Carrillo, & Geiser, 2014; Little, 2013). The items with the highest and lowest unstandardized loadings were assigned to parcel 1, followed by the next highest and next lowest items to parcel 2, and so on until all items were assigned to a parcel. For example, CDS Parcel 1 consisted of the appears to be sluggish, sleepy-eyed appearance, and gets lost in thoughts symptoms; CDS Parcel 2 the seems drowsy, underactive, and in a world of his/her own symptoms; and CDS Parcel 3 the appears tired or lethargic and daydreams symptoms. Each parcel was then used as a manifest variable. Thus, the latent CDS, IN, depression, and anxiety constructs were represented by three manifest variables.

Relationships among CDS, inattention, and internalizing symptomatology.

A measurement model including these CDS, IN, depression, and anxiety factors was estimated in order to assess the unique relationships among these dimensions of psychopathology. We then used a Multiple Indicators Multiple Causes (MIMIC) model to examine whether demographic factors (e.g., age and gender) and markers of disease severity (i.e., presence of shunted hydrocephalus, ambulatory status, diagnosis, lesion level) incrementally predicted these clinical factors. Models from the full sample were re-estimated with the subset of patients with SBM-H.

Documenting levels of CDS in SB.

Finally, to evaluate levels of parent reported CDS in youth with SB, current participants were compared to multiple unselected and clinical samples in the extant literature that reported descriptive statistics on the Penny et al. (2009) scale or a similar scale with overlapping CDS items. We included group comparisons of IN symptomatology as well. Comparisons were made using unequal variances t-tests (i.e., Welch’s t-test), Hedges’ g as an effect size estimate due to sample size discrepancies, and a family-wise Bonferroni correction for multiple comparisons was employed.

The full SB sample, as well as the subgroup of patients with SBM-H, was compared to the following samples: 1) the initial scale development sample from Penny et al. (2009); 2) an unselected sample of twins from Leopold et al. (2016); 3) a sample of adolescents diagnosed with ADHD from Smith et al. (2018); a population-based sample)Barkley, 2013) as well as subgroups within that sample that had been characterized as having elevated IN, CDS, or neither; and a diverse sample of patients referred for clinical neuropsychological evaluations (Jacobson et al., 2018).

We also documented the percentage of our sample with clinically elevated IN and CDS. A traditional symptom count procedure (i.e., six or more ADHD-IN symptoms endorsed at the often or very often level) defined the IN group. Barkley’s (2013) three symptom cut-off definition was used to define elevated CDS.

Missing data.

Of the 169 SB patients in this retrospective sample, 47 were missing parent CDS ratings and 24 were missing IN ratings. Among the subgroup of 111 patients with SBM-H, 34 were missing parent CDS ratings and 17 were missing IN ratings. Patients with and without missing data did not differ in age, sex, ethnicity, presence of shunt, or number of shunt revisions for those with a shunt (all p>.27). Patients without CDS ratings also did not differ from those with complete ratings on the level of parents-endorsed IN (p=.37). For SEM analyses, covariance coverage for patients’ and parents’ ratings ranged from 0.54 to 0.99. Analyses of individual items used the WLSMV estimator with a pairwise approach to missing data. Analyses using item parcels used the MLR estimator with a full information maximum likelihood approach to missing data.

Results

Factor structure and reliability of CDS and IN

The initial CFA of Penny’s et al. (2009) 14-item SCT rating scale replicated the 3-factor (i.e., slow, sleepy, and daydreamer) structure (Table 1) with very good global fit (CFI and TLI > .98) and mediocre local fit (RMSEA=.087). When the 14 CDS and 9 IN items were subjected to EFA, however, the first 6 items of Penny’s scale – the slow factor – heavily cross-loaded with ADHD IN (Table 2) and were therefore removed from the set of items used to measure CDS in the present study. Using only the 8 remaining CDS items that demonstrated discriminant validity from the 9-item IN scale, exploratory factor analyses of the two scales demonstrated that a 3-factor solution (Table 1) – composed of the 9-item IN, 5-item sleepy CDS, and 3-item daydreamer CDS factors – fit the data best (CFI=.990; TLI=.984; RMSEA=.063 [.043, .081]).

Analyses using manifest items

Individual CDS and IN items loaded strongly on their corresponding latent trait (mean standardized loading $\overline{\lambda }$ = .82 for CDS and .81 for IN). Mean standardized loadings were adequate and consistent with prior factor analyses of the RCADS-25 (e.g., Klaufus et al., 2020) (depression $\overline{\lambda }$ = .60 and anxiety $\overline{\lambda }$ = .58). Similarly, estimates of internal consistency were good to excellent for composite measures of CDS (14-item α = .92; 8-item α = .88) and IN (α = .92), but acceptable to good for composite measures of internalizing psychopathology (depression α = .75 and anxiety α = .80).

Analyses using parceled items

Mean standardized loadings of the parcels on the CDS, IN, depression, and anxiety factors were high (CDS $\overline{\lambda }$ = 0.89; IN $\overline{\lambda }$ = 0.91; depression $\overline{\lambda }$ = 0.68; anxiety $\overline{\lambda }$ = 0.74), and composite reliability coefficients (true-score variance) for each factor’s three parcels were excellent for CDS (α = .92) and IN (α = .94) and acceptable for depression (α = .72) and anxiety (α = .78). These results confirmed that the models with parcels as manifest variables displayed acceptable psychometric properties.

Latent measurement model

A 4-factor confirmatory factor analysis with latent traits for CDS, IN, depression, and anxiety demonstrated a very good fit to the data in both the full sample (X²(48)=64.9; p=.052; CFI=.98; RMSEA=.046) and in the subgroup of patients with SBM-H (X²(48)=63.2; p=.070; CFI=.97; RMSEA=.055). The CDS, IN, depression, and anxiety traits demonstrated moderate to large correlations with one another (Figure 1).

Fig. 1.

Multiple indicators, multiple causses model of the latent relationships between cognitive disengagement syndrome (CDS), inattention (IN), depression, and anxiety using parceled indicators. N = 165. Lines with single arrowheads represent directed regression paths, whereas lines with double arrowheads represent bivariate correlations. Standardized path weights are depicted. RCADS = Revised Children’s Anxiety and Depression Scale.

We next tested adding covariates to the 4-factor model. As expected, patient diagnosis and the need for shunting were associated with greater CDS (Diagnosis: λ = .21; Shunt: λ = .18) and IN (Diagnosis: λ = .23; Shunt: λ = .29) symptoms in the full sample. Lesion level and ambulatory status, however, were not significantly associated with CDS or IN ratings (all p > .30) in the full and SBM-H subsample models. Age also showed a significant positive association with CDS symptom burden (standardized path loading λ = .15) such that older patients were rated as exhibiting higher levels of CDS. When this MIMIC model was estimated among the SBM-H subsample, age was no longer significantly associated with CDS. The final MIMIC model in the full sample is shown in Figure 1. (The final MIMIC model in the SBM-H subsample was similar but did not include a covariate for shunt because all participants in that group had a shunt.) The final models provided a good fit to the data in both the full sample (X²(69)=91.1; p=.039; CFI=.98; RMSEA=.044) and the SBM-H subgroup (X²(59)=74.1; p=.089; CFI=.97; RMSEA=.050).

Level of CDS symptomatology

Rates of clinically elevated CDS and inattention.

For the full sample, the study criteria for clinical elevations resulted in 18% (22 of 122) and 21% (30 of 145) of the full sample meeting elevated CDS and IN criteria, respectively. Many (39%) of the patients with elevated CDS didn’t meet criteria for IN, and almost two-thirds (65%) of patients with IN didn’t meet criteria for elevated CDS. For the subgroup of patients with SBM-H, 22.1% had elevated CDS and 28.8% had elevated IN. Half (50%) of SBM-H patients with elevated CDS did not meet IN criteria, and two-thirds (66.7%) of those with elevated IN did not meet CDS criteria.

Comparisons to extant samples.

Table 3 compares CDS and IN ratings for current participants (both the full SB sample and the subset of those with SBM-H) to extant published results. For the full sample, patients with SB were rated higher on CDS and IN items than unselected and non-clinical samples, including Penny’s et al. (2009) original sample of school children, a community sample of twins described in Leopold et al. (2016), and the group from Barkley (2013) without an ADHD or CDS diagnosis. Of note, our sample was on average older than the Penny (2009) sample, younger than the Leopold et al. (2016) sample, and of similar age to the Barkley (2013) sample, indicating that age differences across studies are unlikely to account for these findings. For CDS items, effect sizes were larger for the sleepy subscale than the daydream subscale. The SB patients were also rated modestly higher (g=.49) on the sleepy subscale than the participants from Barkley (2013) who met study criteria for ADHD without CDS (and who were of similar mean age to the current participants). However, patients with SB demonstrated lower parent-rated levels of CDS and IN than adolescents diagnosed with ADHD (Smith et al., 2018), a heterogeneous sample of patients referred for neuropsychological evaluation (Jacobson et al., 2018), and the CDS and ADHD+CDS groups from Barkley (2013). Again, these findings appear not to be accounted for by age differences across samples in that the current sample is, on average, older than the Jacobson et al (2018) sample but younger than the Smith et al. (2018) sample and Barkley’s (2013) CDS group.

Table 3.

Exploratory Factor Analyses of CDS and IN

		2-Factor EFA		3-Factor EFA
Construct	Item	F1	F2	F1	F2	F3
Slow	CDS 1: Is apathetic; shows little interest in things or activities	.52	.20	.02	.63	−.05
	CDS 2: Is slow or delayed in completing tasks	.28	.71	.47	.58	−.04
	CDS 3: Is unmotivated	.32	.58	.49	.45	−.03
	CDS 4: Lacks initiative to complete work	.31	.70	.45	.56	.04
	CDS 5: Effort on tasks fades quickly	.32	.66	.34	.55	.15
	CDS 6: Needs extra time for assignments	.17	.72	.11	.69	.14
Sleepy	CDS 7: Appears to be sluggish	.60	.32	.73	.16	.01
	CDS 8: Seems drowsy	.72	.23	.88	.03	.05
	CDS 9: Appears tired of lethargic	.79	.16	.93	.01	−.03
	CDS 10: Has a yawning, stretching, or sleepy-eyed appearance	.76	.08	.89	−.13	.09
	CDS 11: Is underactive, slow-moving, or lacks energy	.69	.19	.86	−.01	.01
Daydream	CDS 12: Daydreams	.84	−.06	.01	.02	.86
	CDS 13: Gets lost in his or her own thoughts	1.02	−.09	.11	−.02	.96
	CDS 14: Seems to be in a world of his or her own	.92	.00	.16	.03	.86
Inattention	IN 1: Does not pay attention to details or makes careless mistakes	−.09	.93	.05	.96	−.22
	IN 2: Has difficulty keeping attention to what needs to be done	.02	.86	.01	.87	.01
	IN 3: Does not seem to listen when spoken to directly	.03	.62	.07	.64	−.07
	IN 4: Does not follow through when given directions and fails to finish activities	−.12	.89	−.02	.93	.17
	IN 5: Has difficulty organizing tasks and activities	−.03	.91	.00	.93	−.06
	IN 6: Avoids, dislikes, or does not want to start tasks that require ongoing mental effort	.07	.72	.04	.72	.04
	IN 7: Loses things necessary for tasks or activities	−.01	.81	−.08	.84	.04
	IN 8: Is easily distracted by noises or other stimuli	.16	.64	.15	.70	.28
	IN 9: Is forgetful in daily activities	.01	.85	.04	.86	.03

Note. CDS = cognitive disengagement syndrome; IN = inattention; EFA = exploratory factor analysis.

A similar pattern emerged for the subgroup of patients with SBM-H. This group also demonstrated higher IN and CDS symptoms than unselected/non-clinical samples and higher sleepy symptoms than adolescents with ADHD but not CDS. Effect sizes were higher than for the full SB group, consistent with greater neurobehavioral risk among the SBM-H subgroup. Furthermore, these patients were similar to the group of clinically referred patients in Jacobson et al. (2018) on the sleepy subscale, again indicating relatively higher risk for symptoms in the SBM-H subgroup compared to the full SB sample. As was the case for the full SB sample, the SBM-H subgroup demonstrated lower parent-rated levels of CDS and IN than an adolescent ADHD sample (Smith et al., 2018) and the CDS and ADHD+CDS groups from Barkley (2013).

Discussion

Difficulties with attention are common among individuals with SB and can contribute to functional impairment at home, at school, and in the community. Many people with SB are diagnosed clinically with ADHD, but the causes and profile of attentional difficulties can differ in this population from those with idiopathic ADHD. A better characterization of the attentional challenges in SB could support more accurate identification and, ultimately, more effective and personalized interventions. The construct of CDS has been the focus of increased scientific inquiry in recent years and may help better characterize some of the cognitive and behavioral challenges in SB. To date, however, only one published study has explored CDS among people with SB, and that study did not use a measure designed to assess this construct.

We first explored the psychometric properties of a full CDS measure (Penny et al., 2009) in our sample of 169 youth ages 5 to 19 who are followed in a multidisciplinary SB clinic. We found that the measure had good to excellent reliability, and we replicated the factor structure proposed by Penny, which produced three CDS subscales (called slow, sleepy, and daydreamer). Consistent with previous research in other populations (Jacobson et al., 2012; Barkley 2013), the slow subscale did not show adequate discriminant validity from IN. Our final model, which fit the data well, supports measuring the sleepy and daydreamer aspects of CDS in this population using items from the Penny scale.

The primary aim of this study was to better characterize CDS symptomatology in children and adolescents with SB. We had three hypotheses. First, we predicted that symptoms of CDS would be associated with, but still distinct from, IN and internalizing symptoms. This prediction was supported, at least for the sleepy and daydreamer subscales of the Penny measure. Consistent with previous literature demonstrating that CDS and IN are separable constructs, the correlation between the latent traits of CDS and IN was .58 in our final model, indicating that about two thirds of the variance in CDS was independent from IN. Furthermore, the scales identified largely different individuals as having clinical problems. Up to half of the patients with elevated CDS didn’t meet criteria for elevated IN, and about two-thirds of patients with elevated IN didn’t have elevated CDS.

Our results are consistent with a growing body of literature demonstrating that CDS and IN are partly separate constructs in other populations (Becker et al., 2016; Hartman et al, 2004; Jacobson et al, 2012). CDS and IN also show somewhat different developmental trajectories, further supporting their discriminant validity. Symptoms of IN tend to be stable or decline slightly from school age through adolescence (Leopold et al., 2016), while previous work has found modest increases in parent reported CDS during this same age period (Jacobson et al., 2018; Leopold et al., 2016, Smith et al., 2022). Consistent with this prior literature, we found a statistically significant relationship between older age and higher CDS in our full sample, but no relationship between age and IN.

The association between CDS and internalizing symptoms is also well-documented in pediatric samples (Becker et al., 2016; Becker & Langberg, 2013; Smith and Langberg, 2017) and remains significant while accounting for symptoms of ADHD (Bauermeister et al., 2012). In our sample of youth with SB, self-reported internalizing symptoms were significantly associated with CDS symptoms as well as with IN. Effect sizes were generally moderate (i.e., latent factor correlations in the 0.3–0.5 range), meaning that most of the variance in CDS was not accounted for by internalizing symptoms.

Our second hypothesis was that the level of CDS in this sample would be elevated compared to community and clinically referred samples. This prediction was partly supported. Parent-reported rates of CDS were higher in the present sample of youth with SB than community samples of typically developing children and adolescents (Penny et al., 2009; Leopold et al., 2016), particularly for the sleepy subscale. However, the full sample of youth with SB demonstrated lower parent reported CDS than a large sample of adolescents diagnosed with ADHD (Smith et al., 2018) and a large, heterogenous sample of children and adolescents with psychiatric and medical diagnoses who had been clinically referred for neuropsychological evaluation (Jacobson et al., 2018). Because of the documented relationship between older age and higher CDS symptomatology, we investigated whether these findings could reflect age differences across samples, but that did not appear to be the case. The current sample demonstrated higher CDS symptoms than even an unselected sample that was, on average, older (Leopold et al., 2016) as well as lower CDS symptoms than a clinically referred sample that was, on average, younger (Jacobson et al., 2018). A largely similar pattern emerged for the subsample of current participants with SBM-H. Although we were not able to make statistical comparisons with the Smith et al., (2022) study due to different questionnaires used to measure CDS, parent-reported rates of CDS in the present sample are nearly twice as elevated as parent-reported rates of CDS in the sample of youth with SB in the Smith et al., (2022) study. Overall, our results indicate that the level of CDS in youth with SB is higher than in unselected samples, but that clinically elevated symptoms of CDS are evident only in a substantial minority of the SB population.

In retrospect, the finding that our sample had lower rates of CDS than some referred samples is not surprising. Our multidisciplinary clinic cares for a heterogenous group of youth with SB, many of whom do not meet criteria for ADHD or other psychiatric diagnoses. For youth aged 5 and up, yearly multidisciplinary clinic visits are recommended. Providers from the following disciplines are scheduled to see most patients: rehabilitation medicine, neurosurgery, orthopedics, urology, genetics, physical therapy, nutrition, rehabilitation psychology, and social work. Many patients are also seen by the hospital’s adaptive sports program coordinator, who provides information regarding local adaptive sport opportunities and assists with registration. The standard questionnaires gathered during clinic are handed out by a medical assistant to parents and patients upon check-in. Parents and patients are provided a brief description of the measures and instruction on how to complete them. The rehabilitation psychologist utilizes the questionnaire data to assess child psychosocial functioning and to screen for attentional difficulties during the clinic visit. The psychologist uses this information in conjunction with clinical interview to decide whether to place an internal referral for comprehensive clinical outpatient neuropsychological evaluation. These referrals are typically placed to help answer clinical questions related to neurocognitive functioning (e.g., learning, attention, intellectual functioning), adaptive functioning, and/or educational/community impairment. If concerns with patient or parent psychosocial functioning are identified, brief intervention, in addition to internal or community referrals for follow-up behavioral health treatment when indicated, is provided by the psychologist during the clinic appointment.

Finally, we predicted that higher levels of CDS would be associated with markers of disease severity, including diagnosis of myelomeningocele, higher lesion level, ambulatory difficulties, and presence of a shunt. This prediction was also partly supported. Myelomeningocele was associated with higher CDS symptomatology than other SB diagnoses, and participants with shunted hydrocephalus had more CDS symptoms than participants without a shunt. These findings are consistent with research demonstrating that patients with SBM-H are at high risk for a range of neurocognitive challenges, while other forms of SB may primarily impact mobility and bowel/bladder functioning (Fletcher et al., 2005). Lesion level and ambulation status, on the other hand, were not associated with CDS in our sample. Taken together, these results indicate that the difficulties captured by the CDS measure likely reflect atypical brain functioning at least in large part, rather than an exclusively orthopedic issue. This confirmation is important since some but not all CDS items (e.g., “Is underactive, slow-moving, or lacks energy”) could be related to mobility challenges.

The current findings must be interpreted in the context of several limitations. First, we utilized a clinically referred sample of convenience. Second, we did not include a control group of children without SB and comparisons were thus made statistically with previously published results. This means that some of the obtained differences could be due to differences in the demographic characteristics between our sample and the other samples, rather than an effect of SB per se. Age differences across samples do not appear to explain the findings, and given the lack of association between gender and CDS, gender differences are also unlikely to contribute, but it is certainly possible that some other demographic characteristic partly explains some of the observed differences. Third, we did not have ratings of IN or psychopathology for all the comparison samples. Fourth, the results may be impacted by method variance, given that internalizing symptoms were measured by the patient and CDS and inattention symptoms were measured by the caregiver. Lastly, we only had a single reporter (parent report) of CDS symptoms. Recent literature highlights the value of including multiple reporters of CDS, as CDS can manifest differently in home and school contexts (Saez et al., 2019; Smith et al., 2022).

Despite these limitations, the current study represents an important contribution to the CDS and SB literatures. We demonstrated that CDS can be measured reliably in youth with SB, and that the sleepy and daydreamer aspects of this construct can be discriminated from IN and internalizing symptoms in this population. We are not aware of a previously published study that has characterized CDS in SB using a measure designed to assess this construct. Overall, psychometric properties of the CDS scale we used (Penny et al., 2009) were similar in this sample as in other unselected and clinical samples. We found that CDS symptomatology was significantly higher in youth with SB than unselected populations, with most effect sizes falling in the moderate to large range. About 20% participants with SB had clinically elevated CDS, and many of those individuals did not meet symptom criteria for ADHD. Thus, while screening individuals with SB for ADHD is important, standard ADHD measures fail to identify a substantial portion of the SB population with attention-related challenges. Given the growing body of evidence indicating an association of higher levels of CDS with academic, social, and functional impairment (Barkley, 2012; Becker and Langberg, 2013; Creque & Willcutt, 2021; Fassbender et al., 2015; Jacobson et al., 2018; Kim & Kim, 2021; Wåhlstedt & Bohlin, 2010; Willcutt et al., 2014), standard screening for CDS symptoms in SB clinics for all patients age 5 and older may help identify clinically impairing symptoms and design targeted treatment plans. Based on findings from the current study and a recent systematic review (Becker et al., 2021), we recommend that clinicians and researchers assess CDS using measures that demonstrate discriminant validity from ADHD-IN items. Examples of such measures include the Child and Adolescent Behavior Inventory (CABI) and/or the Child Concentration Inventory, Second Edition (CCI-2).

Table 4.

Composite Effect Size Comparison of CDS and IN Between Full Spina Bifida, SBM-H Subset, Elevated CDS Subset, and Extant Samples

Demographics	Penny et al. (2009)	Leopold et al. (2016)	Smith et al. (2018)	Barkley (2013)					Jacobson et al. (2018)
Sample size	n=325	n=478	n=255	n=1922	n=1714	n=102	n=43	n=63	n=566
Sample type	Unselected	Unselected	ADHD	Total	No Dx	ADHD Only	CDS Only	ADHD+CDS	Hospital NP
Sex (% female)	55.5%	47.0%	26.3%	50.0%	51.3%	35.9%	53.7%	34.4%	55.0%
Age (mean (SD)) [range]	8.5 (2.2) [4–13]	12.9 (2.5) [9–17]	12.0 (1.1) [10–15]	11.4 (3.5) [6–17]	11.4 (3.5) [6–17]	11.4 (3.4) [6–17]	13.2 (3.0) [6–17]	12.0 (3.0) [6–17]	10.4 (2.8) [6–16]
Composites	Full Spina Bifida Sample (n=122) versus Extant Samples
Sleepy Scale	.67	.62 *	−.11	.72	1.42	.49	−1.07	−1.24	−1.75
Daydream Scale	.18		−.77	.65	1.24	.06	−.99	−1.87	−.67
IN Total	–	.79	–	–	–	–	–	–	−.83
	SBM-H (n=77) versus Extant Samples
Sleepy Scale	.86	.85 *	.02	.91	1.78	.68	−.92	−1.06	−.17
Daydream Scale	.37		−.57	.89	1.63	.27	−.81	−1.69	−.51
IN Total	–	1.06	–	–	–	–	–	–	−.63
	Elevated CDS (n=22) versus Extant Samples
Sleepy Scale	2.38	2.93 *	1.01	2.44	4.44	2.15	.15	−.20	.84
Daydream Scale	1.56		.85	2.37	3.99	1.66	.26	−.59	.41
IN Total	–	2.12	–	–	–	–	–	–	.09

Note. Bold effect sizes (Hedges’ g) indicate significance at p ≤ .004 (family-wise Bonferonni correction applied for 3 composite and 3 subgroup comparisons (.05/9 ≈ .006). Positive and negative values indicate higher and lower symptomatology, respectively, among current vs. comparison samples. All group comparison t-statistics and Hedges’ g are from unequal variances t-tests.

^*= total of items 7, 8, and 12 to match identical items from Leopold et al. (2016). Jacobson et al. (2018) neuropsychological evaluation-referred sample includes epilepsy, oncology, brain injury, and neurofibromatosis patients; CDS = cognitive disengagement syndrome; SBM-H = spina bifida myelomeningocele with hydrocephalus; ADHD = attention-deficit/hyperactivity disorder.