Planungsleistungen bei Grundschülern mit ADHS und LRS

Background: The term executive functions (EF) depicts a set of higher-order cognitive abilities needed to manage and control individuals' thoughts and behaviours. Many daily life activities place high demands on EF. Converging evidence has shown that the ontogenetic development of EF follows a protracted pathway and that specific aspects of EF continue to mature into late adolescence and even young adulthood (e. g., Johnson, 2001, 2012). Thus, it is not surprising to learn that EF deficits may crucially hamper children's skill acquisition and learning progress. Furthermore, beyond exerting negative effects on individuals' school and professional careers, consistent experiences of (school-related) failure are frequently associated with psycho-emotional problems such as low self-esteem and school-related anxieties (e. g., Pixner & Kaufmann, 2013). Hence, the early detection of specific EF deficits in distinct neurodevelopmental disorders is crucial for (differential) diagnosis and treatment planning alike.

Attention deficit hyperactivity disorder (ADHD) is – among others – characterized by executive dysfunctions. EF is commonly reported to be underdeveloped in children diagnosed with ADHD and is thus considered to be one of the core deficits of ADHD. EF deficits observed in ADHD include inhibition of irrelevant stimuli, planning, cognitive flexibility, working memory and associative thinking (Barkley, 1997; Willcutt, Doyle, Nigg, Faraone & Pennington, 2005). Despite the fact that planning skills are considered to be a relevant aspect of EF in both adult and child models of EF (Miyake et al., 2000; Anderson, 2002; respectively), studies investigating planning skills in children with ADHD are scarce to date and largely restricted to English-speaking populations (Happé, Booth, Charlton & Hughes, 2006; Marzocchi et al., 2008; Gau & Shang 2010). Interestingly, while Marzocchi et al. (2008) reported planning deficits in children with ADHD but not in children with dyslexia, the results of other authors also disclosed deficient planning skills in children with dyslexia (Gioia, Isquith, Kenworthy & Barton, 2002). A plausible explanation for the latter contradictory findings are methodological differences between the two studies: Whereas the findings of Marzocchi et al. (2008) are based upon actual test performance (assessing EF in rather artificial testing environments), Gioia et al. (2002) used parent ratings that ought to estimate children's EF as they are relevant to daily living situations and thus, to orchestrating goal-directed behaviors. For elementary school children both the home and school environment may host various situations that impose high demands on planning skills. Importantly, proficient reading and writing alike requires mastery of many subskills that need to be integrated and automatized (Biederman et al., 2004). In particular, beyond working memory also planning skills are key to many academic endeavours such as story writing. However, studies targeted at examining planning skills in neurodevelopmental disorders are scarce. The present study aims to fill the aforementioned gaps by systematically investigating planning skills in elementary school children diagnosed with ADHD and other learning disabilities (i. e., dyslexia as well as a combined group diagnosed with ADHD+dyslexia). Furthermore, our findings are expected to contribute to more specific descriptions and better (differential) diagnoses of two highly prevalent childhood neurodevelopmental disorders known to have adverse effects on individuals' quality of life, self-esteem, and academic as well as occupational functioning.

Aims: The present study was targeted at assessing planning skills in elementary school children with ADHD and other forms of learning disabilities (i. e., dyslexia) upon utilizing both parent rating scales (tapping planning skills that are relevant for every-day-life situations and goal-directed behavior) and performance-based neuropsychological tests (i. e., a tower task thought to tap planning skills in a structured and somewhat artificial testing environment). Our working hypotheses were as follows: First, we hypothesized that EF deficits reported by parents should be more pronounced in children with ADHD and other learning disorders than in controls. Importantly, we were interested to investigate whether potential difficulties remain even after controlling for intelligence, age, and language skills. Second, we proposed that compared with children with dyslexia those with ADHD (with and without associated dyslexia) should display more severe planning deficits according to parent ratings, even after controlling for potentially confounding variables. Third, we assumed that compared with children diagnosed with ADHD solely those with an associated dyslexia should display worse planning skills according to actual test performance and parent ratings alike (see Gioia et al. (2002) who reported rating-based planning deficits to be associated with ADHD and – albeit to a lesser degree – with dyslexia in an English-speaking sample of school children). Finally, based on the Anglo-American literature correlations between parent ratings and neuropsychological test performance regarding ADHD were expected to be rather low (e. g., McCandless & O' Laughlin, 2007; Toplak, Bucciarelli, Jain & Tannock, 2009).

Methods: Participant characteristics.Overall, 85 children aged between eight and ten years participated in the present study (mean age 102.3 months, standard deviation/SD 8.4 months). Beyond being asked to complete the experimental tasks, all children were subjected to neuropsychological background tasks used for matching purposes (i. e., intelligence and language testing; see below for a more detailed description). Overall, our sample comprised four groups: children with ADHD (n = 28), children with dyslexia (n = 17), children with the comorbid condition (ADHD+dyslexia; n = 19) and healthy controls without reading and writing problems or clinically relevant symptoms of ADHD (n = 21). Importantly, and as described above, all children were matched for age, sex, IQ and language skills (speech comprehension, vocabulary). As regards ADHD, group membership was based upon clinical interviews and a frequently used questionnaire tapping ADHD symptoms (Döpfner, Görtz-Dorten & Lehmkuhl, 2008). With respect to dyslexia, group membership was based on the results of the standardized reading and writing test administered to all children (SLRT, Moll & Landerl, 2010; see the next paragraph for a more detailed description of the latter test).

Neuropsychological background tasks (tasks for sample description and group differentiation): (i) Nonverbal intelligence was estimated from the current German version of the Wechsler Intelligence Scales for Children (HAWIK-IV, Perceptual Reasoning Index; Petermann & Petermann, 2008). (ii) Language skills were examined by using a popular German-language vocabulary test (WWT6 – 10; Glück, 2007), as well as a language comprehension test (TROG-D; Fox, 2006). (iii) Reading and writing skills were assessed upon administering a widely used German-language standardized dyslexia test (SLRT II; Moll & Landerl, 2010). (iv) ADHD symptoms were reported by parents who were asked to complete an ICD-10 based structured questionnaire (FBB-ADHS; Döpfner et al., 2008).

Experimental tasks: First, all children were subjected to a computerized planning task (Stockings of 11-5Cambridge/SOC; subtest of the CANTAB, Cambridge Cognition, 2004 enabling us to record both response accuracy and latency. In particular, the following variables were used as dependent variables: (i) number of optimally solved tasks (i. e., solved within the minimum number of moves across all trials as well as across the most difficult trials requiring 5 moves at least), (ii) average number of moves used to solve the most difficult trials requiring 5 moves at least, and (iii) initial thinking time (i. e., latency to implement the first move). Second, parents were asked to complete a rating questionnaire targeted at assessing various aspects of EF as they are relevant for children's daily life activities (German-language version of the Behavior Rating Inventory of Executive Function (BRIEF; Drechsler & Steinhausen, 2013). Notably, the German-language version of the BRIEF is comparable to the original BRIEF (Gioia, Isquith, Guy & Kenworthy, 2000) with regard to its psychometric properties (Schöfl, Beitel, Kloo & Kaufmann, in press).

Statistical analyses. First, univariate analyses of variance (ANOVAs) and post-hoc analyses (Bonferroni corrections) were calculated to examine whether group differences on the eight BRIEF scales were significant. After correcting for multiple comparisons, the statistical significance level was set at p < .01 (p = .05/4 comparisons). Second, upon conducting further ANOVAS and post-hoc tests, potential group differences were examined with regard to planning skills (as estimated by parent ratings and derived from performance-based measures). Finally, correlation strengths between parent ratings (BRIEF results) and performance-based measures (results of the computerized SOC) were calculated to investigate whether the two assessment methods measure the same EF construct or not.

Results: In support of our first working hypothesis (i. e., EF deficits reported by parents should be more pronounced in children diagnosed with ADHD and dyslexia than in controls), parent reports revealed significant group differences and meaningful effect sizes in seven out of eight BRIEF-scales differentiating children with ADHD and controls: “inhibition”, “shifting”, “emotional control”, “initiation”, “working memory”, “planning and organization“ and “monitoring”. Importantly, the latter group differences reached significance at the Bonferroni-corrected significance level (i. e., p < .01) with at least moderate effect sizes (i. e., η_p²= .25, according to Ferguson, 2009). However, regarding the BRIEF scale “organization of materials“, main group effects were not significant at the adjusted p-level. Post-hoc pairwise comparisons revealed that with respect to the BRIEF scale “emotional regulation“, group differences became significant only between controls and children with ADHD (p < .01), whereas other group comparisons did not reach the adjusted significance level. Compared with healthy controls, children with dyslexia (without ADHD) were reported to exhibit significantly more deficits concerning „working memory“ (p < .001) as well as “planning and organizing“ (p < .01). Moreover, compared with controls, children with ADHD and the comorbid condition (ADHD+dyslexia) were estimated to exhibit significantly more deficits on six out of eight BRIEF scales (i. e., „inhibition“, „shift“, „initiate“, „working memory“, „planning and organizing“ (each at p < .001) and „emotional control“ (at p < .01). Overall, results of post-hoc analyses confirmed our first working hypothesis and revealed that according to parent ratings children with ADHD displayed the highest levels of executive dysfunctions regarding all BRIEF scales. Interestingly, according to parent ratings, children with dyslexia disclosed higher levels of deficiencies regarding „working memory“ (p < .001) and „planning and organization“ (p < .01) than healthy controls.

In our second working hypothesis we proposed that parent ratings of one specific executive dysfunction, namely planning deficits, should be more pronounced in ADHD (with and without associated dyslexia) than in dyslexia. Our findings are consistent with the latter hypothesis and reveal that, compared with controls, both children with ADHD and those with dyslexia exhibit significantly worse planning skills according to parent ratings, those with dyslexia were considered to display less severe planning deficits than children diagnosed with ADHD (p < .01).

Our findings do not support our third working hypothesis proposing that compared with children diagnosed with ADHD alone, those with the comorbid condition might exhibit worse planning skills. Rather, both groups (ADHD and ADHD+dyslexia) were found to display comparably poor planning skills. This holds true for both the computerized tower-task (SOC: number of problems solved in minimum moves, number of 5-move-problems solved, number of moves to solve 5-move-problems, initial thinking time: all p > .01) and the parent ratings of “planning and organizational” skills as tapped by the BRIEF questionnaire (p > .01). Nonetheless, both groups were described to show significantly more planning deficits than children with dyslexia solely (p < .01) or controls (p < .001).

Finally, our findings were fully consistent with our fourth and last working hypothesis proposing little concordance between neurocognitive test performance and parent ratings regarding planning skills (correlation strengths between the BRIEF-scale planning and i) SOC number of problems solved in minimum moves across all trials and across the most difficult trials requiring 5 moves at least (r = .13); ii) SOC number of problems solved in minimum moves across the most difficult trials requiring 5 moves at least (r = −.08); iii) SOC average number of moves used to solve the most difficult trials requiring 5 moves at least (r = .12); iv) SOC initial thinking time (r = −.13)). Nonetheless, bivariate correlations between parent ratings of ADHD symptoms (FBB-ADHS; Döpfner et al., 2008) and parent ratings of planning skills (BRIEF; Drechsler & Steinhausen, 2013) reached significance at p < .001, the highest correlations emerging between the inattention scale and planning (r = .70), followed by somewhat lower, but nevertheless moderate correlations regarding the links between the hyperactivity scale and planning (r = .54) as well as the impulsivity scale and planning (r = .42).

Discussion: Significant group differences regarding planning skills emerged upon utilizing parent ratings but not upon employing neuropsychological tests. Importantly, group differences remained even after controlling for intelligence, age and language skills. According to parent ratings and consistent with our respective working hypotheses, children with ADHD displayed the poorest functional levels regarding various aspects of EF.

Interestingly, children with dyslexia performed worse than healthy controls on everyday-life activities regarding working memory and planning (Gioia et al., 2002; see Drechsler & Steinhausen, 2013 for the German-language version of the BRIEF used in the present study). Thus, parent ratings – but not so neuropsychological test performance – revealed that compared with control children, those with dyslexia seemed to have worse planning skills (however, see Marzocchi et al., 2008).

Notably, parent ratings of EF did not differentiate between children with ADHD and a group with comorbid ADHD+dyslexia, thus failing to support our third working hypothesis assuming an additive effect (i. e., accumulating EF deficits) in the comorbid group. Rather, our findings suggest that the observed planning deficits in both groups (i. e., children diagnosed with ADHD solely and those with comorbid ADHD+dyslexia) may be primarily attributable to ADHD.

Finally, our findings disclosed only modest correlations between parent ratings of EF and performance-based measures and thus, corroborate similar results reported in English-speaking populations (Geurts, Verte, Oosterlaan, Roeyers & Sergeant, 2004; Happè et al., 2006) suggesting that tests (i. e., performance-based measures) and questionnaires (i. e., parent ratings) tap different aspects of EF. Hence, clinical assessments of EF may greatly benefit from the combined use of parent ratings and performance-based measures. Importantly, our results disclosing higher sensitivity of parent ratings compared with performance-based measures of EF nicely fit the findings of a recently published respective review based on the Anglo-American literature (Toplak, West & Stanovich, 2013). According to the latter authors, the differential sensitivity of actual test performance and ratings may be explained by the fact that parent ratings are more likely to index complex goal-directed behaviours (relevant for every-day-life situations) while performance-based measures assess more or less specific aspects of processing efficiency (under structured and somewhat artificial testing environments). Thus, performance-based measures are likely to overestimate children's EF because some of the characteristic features of EF (i. e., self-organisation, monitoring etc.) are difficult or even impossible to assess within a classical testing environment.

Taken together, our study is the first to systematically investigate planning skills in German-speaking elementary school children diagnosed with ADHD, dyslexia and those with comorbid ADHD+dyslexia. Our results are novel as they reveal planning deficiencies in the latter population that remain even after controlling for potentially confounding variables such as intelligence, age and language skills. Moreover, compared with neurocognitive performance-based measures parent ratings of planning skills seem to be more sensitive to deficiencies that are relevant to elementary school childrens' everyday-life activities. Future research is needed to replicate our findings with larger samples and a wider age range. Upon acknowledging that – among others – also planning skills are tightly linked to more or less complex academic skills (and their acquisition) our findings have important implications for educational practice and science. Future research endeavours are urgently needed to further elucidate the link between children's planning skills and their academic achievements.

We would like to thank Mr. Jason Fogler at Boston Childrens's Hospital for proof reading this extended abstract!