Selective Difficulties in Lexical Retrieval and Nonverbal Executive Functioning in Children With HbSS Sickle Cell Disease

Selective Difficulties in Lexical Retrieval and Nonverbal Executive Functioning in Children With... Abstract Language deficits in multilingual children with sickle cell disease (SCD) are poorly understood. We tested the hypothesis that selective language deficits in this population could relate to an impaired frontal lobe functioning often associated with high-risk homozygous HbS disease (HbSS). In all, 32 children from immigrant communities with HbSS SCD aged 6 to 12 years (mean age = 9.03, n = 9 with silent infarcts) and 35 demographically matched healthy controls (mean age = 9.14) were tested on their naming skills, phonological and semantic fluency, attention, and selected executive functions (response inhibition and planning skills). Analyses of variance showed significant differences between patients and controls in inhibition and planning (p = .001 and .001), and phonological fluency (p = .004). The poorer performance in phonological fluency of the children with SCD was not associated with any visible brain damage to language areas. Hierarchical regression analyses showed that, whereas the control children’s vocabulary knowledge explained their performance in the phonological fluency tasks, only inhibition skills accounted for variance in the performance of the children with SCD. These results suggest a selective impairment of verbal and nonverbal executive functioning (i.e., planning, inhibition, and phonological fluency) in children with SCD, with deficits possibly owing to frontal area hypoxia. executive functioning, language skills, lexical retrieval, sickle cell disease, verbal fluency Sickle cell disease (SCD) is an inherited disorder of hemoglobin, endemic in some regions of Africa, which has also spread in the European Union and the United States owing to migration flows. The disease is characterized by chronic hemolytic anemia, recurrent vaso-occlusive events, and a progressive vasculopathy, which may have negative effects on children’s health and neurocognitive functioning (DeBaun & Kirkham, 2016; Hijmans et al., 2011; Kral, Brown, & Hynd, 2001; Schatz & McClellan, 2006). The severity of the disease and the likelihood of neurocognitive deficits depends on the patient’s SCD genotype. Homozygous HbS disease (HbSS) and HbS0 thalassemia are characterized by lower hemoglobin levels, an earlier onset of cerebral vasculopathy, and a more severe phenotype, with a higher risk of cerebrovascular events than in the double heterozygous variants (HbSC and HbSß+) (Manara et al 2017; Schatz & McClellan, 2006). It is estimated that approximately 11% of children with HbSS SCD suffer from overt cerebral strokes. Occult neurological complications may develop too, however, and lead to selective cognitive impairments. These include silent infarcts, or subclinical cerebral strokes visible on brain imaging, affecting approximately 28% of children with SCD (Burkhardt, Lobitz, Koustenis, Rueckriege, & Hernáiz Driever, 2017), and small vessel vasculopathy, or insufficient oxygen and glucose delivery to brain tissues, which may result in brain function impairments in the absence of visible cerebral lesions (Baldeweg et al., 2006; Brousse, Kossorotoff, & de Montalembert, 2015; Colombatti et al., 2015, 2016; DeBaun & Kirkham, 2016). Children with SCD may consequently have a variety of neurocognitive deficits affecting their attention (Hijmans et al., 2011), executive functions such as inhibitory control, planning, and working memory (Brandling-Bennett, White, Armstrong, Christ, & DeBaun, 2003; Hijmans et al., 2011), visuospatial skills (Armstrong et al, 1996; Schatz & McClellan, 2006), and language skills (Berkelhammer et al., 2007; Kral et al., 2001; Sanchez, Schatz, & Roberts, 2010; Schatz, Puffer, Sanchez, Stancil, & Roberts, 2009), as well as a more general cognitive impairment (Berkelhammer et al., 2007). Although the neurocognitive profiles vary considerably, deficits in executive functioning (EF) seem to be characteristic of these children. Vaso-occlusion and hypoxia often affect the frontal brain regions, that is, areas implicated in response inhibition and higher-order EF, such as planning (Berkelhammer et al., 2007; Brown, Davis, Lambert, Hsu, Hopkins, & Eckman, 2000; Burkhardt et al., 2017; Hijmans et al., 2011; Kral et al., 2001; Schatz & McClellan, 2006). Deficits in verbal abilities are also fairly common among children with SCD (Brandling-Bennett et al., 2003; Sanchez et al., 2010; Schatz et al., 2009; Tarazi, Grant, Ely, & Barakat, 2007), but our understanding of these problems is still limited. Previous studies have shown that these impairments primarily affect children with the HbSS genotype, and relate more to the neurological effects of the disease than to its general medical complications (e.g., absences from school owing to hospitalizations; Schatz et al., 2009). Moreover, they are associated with cerebrovascular disruptions and high cerebral blood flow velocities on transcranial Doppler (TCD) ultrasound (Sanchez et al., 2010). Language deficits in SCD are not always associated with brain damage, however, as they sometimes emerge in children with normal neuroimaging findings too (Bernaudin et al., 2000; Schatz, Finke, Kellett, & Kramer, 2002). Some studies suggest a general weakness in the verbal abilities of children with high-risk SCD across language domains (syntax, semantics, and phonology), but their performance within these domains is not impaired to the same degree (Schatz et al., 2009). Research findings to date are also frequently inconsistent regarding the type of language skills that can be affected in children with SCD. Armstrong et al. (1996) reported deficits in vocabulary skills, whereas Brown et al. (2000) and Schatz et al. (2009) did not find any significant difficulties in expressive or receptive vocabulary, but they did find impairments in rapid naming (Brown et al., 2000) and syntactic skills (Schatz et al., 2009). Some researchers have suggested an association between deficits in verbal abilities and lesions in frontal areas (Sanchez et al., 2010; Schatz et al., 1999), but the nature of this association is still not clear. Verbal deficits could be the consequence of neurological insult to frontal lobe areas directly related to language processing, such as Broadmann’s areas 44 and 45 (Sanchez et al., 2010; Schatz et al., 2009). On the other hand, they could be a side effect of other neurocognitive impairments caused by SCD, such us deficits in working memory or EF (Brandling-Bennett et al., 2003). This article aims to further explore the nature of language problems in children with HbSS, examining the association between their performance in lexical retrieval tasks (picture naming, phonological fluency, and semantic fluency), and in tasks for testing attention and EF (planning and response inhibition) that are significantly affected by frontal lobe dysfunctions (Della Sala, Gray, Spinnler, & Trivelli, 1998; Friesen, Luo, Luk, & Bialystok, 2015). Lexical retrieval demands not only language proficiency, but also search strategies that require executive control (Friesen et al., 2015; Luo et al., 2010). In particular, in phonological and semantic fluency tasks, children must employ attentional resources to access their lexical knowledge and inhibit responses that do not fit the (semantic or phonological) criterion (Friesen et al., 2015). By examining children’s lexical retrieval across three types of tasks requiring different degrees of executive control (the Boston Naming Test, a phonological fluency task, and a semantic fluency task), and correlating performance in these tasks with measures of attention, planning, and inhibition, we tested the hypothesis of a selective impairment of lexical skills related to EF deficits. Verbal (phonological and semantic) fluency tasks are used to assess verbal EF, and are particularly effective for distinguishing between the role of executive control and language (vocabulary) knowledge (Friesen et al., 2015; Luo et al., 2010). Although both rely on vocabulary knowledge (i.e., on linguistic proficiency), they place different executive demands on lexical retrieval. Retrieving words based on semantic categories, as in semantic fluency tasks, is fairly automatic in children because words are organized in our memory in a semantic network, with items linked by semantic associations (Luo et al., 2010; Mulatti, Peressotti, Job, Saunders, & Coltheart, 2012). By contrast, phonological fluency requires greater executive control and inhibition skills because, to retrieve words from the same phonemic category (with the same initial sound), children must suppress the automatic retrieval of the lexical items (i.e., words) activated by semantic associations (Friesen et al., 2015). Past research has shown that environmental factors such as socioeconomic disadvantage or bilingualism also contribute to the language problems that may affect children with SCD (Drazen, Abel, Gabir, Farmer, & King, 2016; Montanaro et al., 2013; Tarazi et al., 2007). In the United States, most children with SCD come from African American communities, from families that are not first-generation immigrants or bilingual (Hassell, 2010). In the European Union, where immigration is a more recent phenomenon, most children with SCD come from immigrant and multilingual communities (Hijmans et al., 2011; Montanaro et al., 2013). Children of immigrant families typically show a slower growth in the vocabulary of their second language because they are bilingual (Calvo & Bialystok, 2014), or owing to the low socioeconomic status of their families (Blair & Raver, 2016; Calvo & Bialystok, 2014). Both these factors can affect their language performance: for example, socioeconomic status accounts for up to 47% of the variance in the language skills of preschoolers with SCD (Tarazi et al., 2007). In the present study, we compared the performance of children with HbSS SCD and healthy, demographically matched controls in naming and verbal fluency tasks (semantic and phonological fluency). The study had three main goals: (1) to investigate the influence of neurological (HbSS SCD) and environmental (socioeconomic status and bilingualism) factors on the lexical abilities of children with SCD; (2) to explore whether the children with HbSS SCD showed a profile of selective impairment in nonverbal and verbal tasks consistent with the hypothesis of a selective deficit in EF; and (3) to test the association between an impaired inhibitory control and the performance of these children in phonological and semantic fluency tasks. If, as expected, the cognitive and language problems of children with HbSS SCD are associated with the neurological risk characteristic of their disease, their performance in tasks of verbal fluency and nonverbal EF (inhibition and planning) should be impaired, even by comparison with that of healthy controls matched on socioeconomic level and bilingualism. If the children with HbSS SCD reveal selective problems with verbal language owing to their EF profile, then significant differences in the two groups’ language performance should emerge especially in word retrieval tasks that demand a greater executive control (i.e., phonological fluency). Methods Participants Thirty-two children with HbSS SCD aged 6 to 12 years (mean age = 9.03, SD = 2.04) and 35 healthy demographically matched controls (mean age = 9.14, SD = 1.40) were recruited and agreed to participate in the study. Table I shows the participants’ characteristics. Table I. Characteristics of Participants With HbSS SCD (n = 32) and Demographically Matched Controls (n = 35) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) Table I. Characteristics of Participants With HbSS SCD (n = 32) and Demographically Matched Controls (n = 35) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) Participants with SCD were selected for the study if they had the high-risk HbSS phenotype, were from 6 to 12 years old, had no history of overt cerebral strokes, and were in a steady state defined as at least 4 weeks since any vaso-occlusive crisis (VOC) or hospital admission. In addition to these inclusion criteria, the clinical psychologist at the hospital (the second author) who conducted the routine assessments of the children with SCD and the class teacher (for the controls) were consulted to ensure that participants had sufficient verbal skills to understand the instructions and complete the tasks involved. Thirty-four children with HbSS SCD who met these criteria were initially identified, but two did not complete all the tasks and were excluded post hoc. The final sample thus included 32 patients with HbSS SCD. Patients with HbSC SCD were also initially recruited for the study to enable a comparison between children at high (HbSS) and low (HbSC) neurological risk, but this proved impossible because only four patients with HbSC met the inclusion criteria. TCD velocities were available for all patients: four children had a history of abnormal TCD findings (all had silent infarcts too), while one had a history of conditional TCD. Magnetic resonance imaging (MRI) was available for most of the patients (n = 28). Nine children had silent infarcts in the white matter in the border zones of the middle cerebral artery (MCA; four of them had a history of abnormal TCD). Six patients had lesions involving both the left and the right MCA (these lesions were larger on the left in five-sixth cases), two had left MCA lesions, and one had a right MCA lesion. Seventeen patients (17 of 32, 53%) were receiving a disease-modifying treatment: five were on chronic transfusions owing to abnormal/conditional TCD; and 12 were taking hydroxyurea (HU) for previous recurrent VOCs or acute chest syndromes, or previous anemia <8 g/dl (these are standard indications for HU treatment in Italy) (Colombatti et al., 2018) The healthy controls enrolled for the study came from the same (mainly African and Eastern European) immigrant communities as the children with SCD; they were 6 to 12 years old and had no known cognitive, motor, or sensory disabilities. All participants (SCD group and control group) were from immigrant families. Their ethnicity was predominantly African (30 participants in the SCD group, and 29 controls). Most of the participants were born in Italy (26 SCD patients, 28 controls). Three children (one with SCD, and two controls) had migrated to Italy when they were ≤1 year old. All participants reportedly spoke more than one language and used languages other than Italian at home. Five controls and five children in the SCD group were sequential bilinguals (i.e., they were exposed to Italian as an additional language after the age of 2). The mean time of exposure to the second language (Italian) was 3.4 years for controls and 4.7 years for the SCD group. A Mann–Whitney test revealed no statistically significant differences between the two groups, p = .55. Two children in the SCD group were trilingual: Italian was their third language in both cases, and they had been exposed to Italian for 6 or 7 years. The controls were matched with the SCD patients by age, gender, ethnicity, years spent in Italy, and socioeconomic status. The children with SCD had a mean Intelligence quotient (IQ) of 89.6 (SD = 15.3). Procedure The study was approved by the ethical committee for psychological research at the University of Padova. Written informed consent to use the child’s data for research purposes was obtained from parents, and verbal assent was obtained from the children. Participants with SCD were recruited during routine visits to the Veneto Region’s reference center for SCD. The third author (a psychologist) introduced the study to the parents and informed them that participation was on a voluntary basis. Children were tested individually, before or after their scheduled visits. Healthy controls were recruited and tested at schools in the same geographical area as those attended by the children with SCD. Parents were informed about the study by the class teachers, who also collected their informed consent forms. Children whose families consented to their participation were assessed individually during school hours in a quiet room by a trained psychologist (the third author of this study). A questionnaire was used to collect relevant sociodemographic information from parents, that is, the family’s country of origin, the number of years the child had spent in the adopted country, the language(s) spoken at home, and the parents’ formal education and occupation. As mentioned above, candidates for the study were identified in advance by the second author (clinical psychologist at the hospital) and their class teacher (for the controls), based on their prior knowledge of the child, to ensure that participants had sufficient verbal skills to complete the tasks. For children not born in Italy, language skills were also briefly assessed by the third author of the study by means of a short conversation about their school experience, hobbies, and favorite games. All children had sufficient language skills to describe in simple terms their everyday experiences and talk about their personal interests (Council of Europe, 2001). Before performing each task, the children were administered two or three practice items, and their comprehension of the tasks was ascertained from their performance. Measures Socioeconomic Index The children’s socioeconomic status was judged from their parents’ formal education, scored from 0 (less than elementary school) to 4 (college), and their occupation, scored from 1 (unemployed) to 4 (professional roles). A composite score was calculated as the sum of the highest education score and the highest occupation score obtained by either parent (see also Calvo & Bialystok, 2014). Naming The Boston Naming Test (Kaplan, Goodglass & Weintraub, 1983) was administered to assess the children’s naming skills. The task involves naming aloud 60 objects pictured in black and white. Normative data exist for Italian school-age children (aged 5–11 years, from 5 years and 11 months to 11 years and 4 months; Riva, Nichelli & Devoti, 2000). The test has also been used to assess bilingual naming skills (Kohnert, Hernandez, & Bates, 1998). The procedure used by Riva et al. (2000) was also adopted in our study to enable comparisons. No cues were given to help the children name the pictures, and the test was interrupted after six consecutive wrong or missed responses. The total number of correct responses was scored, and z-scores were calculated based on age-related normative data (Riva et al., 2000). Riva et al.’s normative data for 11-year-olds were used to compute the z-scores of the 12-year-old participants in this study (the difference between the two age groups ranged from 6 to 8 months). Two standardized tasks were used to assess fluency in word retrieval, one for phonological fluency and the other for semantic fluency (Riva et al., 2000). Both tasks have been used in prior research also with bilingual school-age children (age 7 and 10 years, Friesen et al., 2015). Phonological Fluency In this task, the child is allowed 1 min to produce as many different words as possible that begin with a given letter or sound (B and S). The mean number of correct items retrieved in the two conditions (B/S) is computed. Repeated words, proper names, and invented words are disregarded. Z-scores are calculated, based on available age-related norms (Riva et al., 2000). Semantic Fluency The semantic fluency task involves producing in 1 min as many different words as possible in a given category (Animals or Food). Here again, the mean number of correct items retrieved in the two conditions is computed, and z-scores are calculated, based on available normative data (Riva et al., 2000). Selected nonverbal EFs were assessed with Elithorn’s task and the Flanker task. Elithorn’s Perceptual Maze Test This is a traditional “frontal” neuropsychological test, considered particularly sensitive to frontal lobe deficits (Della Sala et al., 1998). It assesses nonverbal planning (Spinnler & Tognoni, 1987) and has been used in previous research to examine the strategic planning skills of children with executive dysfunctions (Leuzzi et al., 2004; Locascio, Mahone, Eason, & Cutting, 2010), and patients with intellectual impairment (Bertella et al., 2005). A short version of the task, standardized for Italian adolescents aged 12–18 years (Batteria per la Valutazione Neuropsicologica; Gugliotta et al., 2009), has also been used successfully in studies with younger children, from the age of 8 years (Leuzzi et al., 2004), and by the first author of this study in another recent study on a sample (N = 57) of typically developing 6-year-olds (Arfé, Vardanega, & Bertin, Unpublished). The same short version was used in the present study too. The test consists of eight items (mazes), and the child is asked to connect a number of black dots, arranged randomly on grids, while obeying three rules: tracing lines from the bottom up; not crossing over the grid; and not backtracking. A maximum time of 2 min is allowed to complete each maze. Three scores are computed: (1) response latency times (in seconds), corresponding to the time elapsing between completing the instructions and when the child starts tracing the path on the grid (which provides a measure of the time spent planning); (2) response times (in seconds), or the speed with which the child completes the maze (which gives a measure of executive ability); and (3) accuracy, corresponding to the overall number of mazes correctly completed within 2 min. After receiving the instructions, the children could practice with three examples. Even the youngest children in the study understood the rules and how to perform the practice items. The test started when the children demonstrated that they fully understood the task and the rules. Flanker Task Experimental Flanker tasks have been used to assess bilingual children’s attention and inhibitory control (Calvo & Bialystok, 2014). In our implementation of the task, the stimuli consisted of a horizontal target arrow displayed in the middle of the computer screen, pointing to the left or right, and flanked by two distractors on both sides. The distractors could be straight lines (neutral condition), or arrows pointing either in the same direction as the target (congruent condition), or in the opposite direction (incongruent condition). The three experimental conditions (neutral, congruent, and incongruent) were equally represented in the experiment (with 96 trials for each one). The stimuli were displayed in black against a white background. Each trial started with a fixation cross displayed for 400 ms, immediately followed by the appearance of the stimulus, which remained on the screen until the child gave a response, or for up to 2 s. The next trial began immediately afterward. Participants were instructed to indicate whether the central arrow pointed to left or right, ignoring the flankers. Participants sat in front of the computer screen, and gave their answer by pressing keys on the keyboard (A for a target pointing to the left, L for a target pointing to the right). The experimental procedure and data acquisition were controlled with E Prime 2 software (Psychology Software Tool, Inc.), recording accuracy and reaction times for each response. Accuracy was scored for each condition as the proportion of correct responses out of the overall number of items presented. Practice trials were used at the beginning of the testing session to enable the child to become familiar with the task. Spatial Relations Subtest, Primary Mental Abilities Scale As deficits in visuospatial skills are sometimes reported in the literature on SCD (Armstrong et al, 1996; Schatz & McClellan, 2006), and may affect performance in the Flanker and Elithorn tasks, this subtest of the Primary Mental Abilities was used to estimate participants’ visuospatial abilities (Thurstone & Thurstone, 1963). Using the Spatial Relations subtest also enabled us to estimate nonverbal (visuospatial) intelligence in the control group, for whom no IQ scores were available (Carretti, Motta, & Re, 2016). The test is traditionally used with school-age children (Carretti, Motta, & Re, 2016). Children are asked to identify one of four shapes that complete a geometrical figure. They are given 6 min to complete all the task items. Response accuracy and total time are recorded. Z-scores are computed based on the test age-related norms. Data Analysis Preliminary between-group comparisons were carried out to check that the HbSS SCD and control groups did not differ on relevant sociodemographic variables. Analyses of variance (ANOVAs) were performed to test between-group differences in age, years in the country, and socioeconomic index. Differences in ethnicity were tested with a chi-square analysis. Then between-group ANOVAs were run to test for differences in visuospatial skills, attention, inhibitory control, planning, and verbal language skills (naming and fluency) between the children with HbSS SCD and their healthy demographically matched controls. Because multiple comparisons were performed, the level of significance was set at .005 to control for Type I error (Bonferroni formulas). Age-adjusted scores were available for the visuospatial, naming, and verbal fluency tasks, so variance owing to age was not further controlled in these analyses. For the Flanker and Elithorn tasks, age was covaried in the analyses because age-equivalent scores were unavailable. The association between children’s executive control and their verbal fluency was explored using Pearson’s correlations and hierarchical regression analyses. The correlations were run on the whole sample and correlations with the dummy factor group (SCD group. 1, and control group. 0) were observed. Hierarchical regression analyses were only performed for phonological fluency because the ANOVAs only revealed significant differences between the two groups in this area (see Results section). Phonological fluency can be influenced by both language proficiency (i.e., breadth of expressive vocabulary) and executive control (i.e., inhibition processes) (Friesen et al., 2015), and these skills could play a different role in children with HbSS SCD and controls. The regression analysis was consequently conducted on the overall sample of participants. The unique contribution of group, breadth of expressive vocabulary, and inhibition skills, and the interaction between group and the other two factors, were thus explored. The Group factor was entered at Step 1, the Boston naming test z-scores at Step 2, the inhibition z-scores (for the mean proportions of accurate answers in the incongruent Flanker task condition) at Step 3, the interaction between Group and Boston naming test z-scores at Step 4, and the interaction between Group and inhibition z-scores at the fifth and final step. Finally, all analyses were replicated after excluding the nine participants in the SCD group who revealed silent infarcts to ensure that the results were not owing to their inclusion in the study sample. Results Preliminary Analyses As shown in Table I, the two groups did not differ on any of the sociodemographic factors considered (age, gender, ethnicity, years in the country, socioeconomic index). The difference in socioeconomic index approached significance, however, with children with SCD having relatively higher socioeconomic scores than controls. Visuospatial Skills, Verbal Language Skills (Naming and Fluency), Attention, Inhibitory Control, and Planning All between-group differences went in the expected direction. Visuospatial skills. No differences emerged between the two groups’ visuospatial skills, F(1, 65) = 0.52, p = .82, ηp2 = .001, and remarkably, both groups’ performance was age-appropriate, as shown in Figure 1. Figure 1. View largeDownload slide Children’s performance (in z-scores) in visuospatial, Boston naming, phonological fluency, and semantic fluency tests. Figure 1. View largeDownload slide Children’s performance (in z-scores) in visuospatial, Boston naming, phonological fluency, and semantic fluency tests. Naming and verbal fluency. No significant differences emerged between the two groups for the Boston naming or semantic fluency tests [F(1, 65) = 2.28, p = .14, ηp2= .03 and F(1, 65) = 0.15, p = .69, ηp2 = .002], whereas the two groups differed significantly in the phonological fluency task, F(1, 65) = 9.09, p = .004, ηp2= .12 (see also Figure 1). Attention, inhibitory control, and planning. The differences between the two groups (HbSS SCD vs. controls) are summarized in Table II. Table II. Differences Between Children With SCD and Controls in the Flanker and Elithorn Tasks (Accuracy and Response Times) HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 Table II. Differences Between Children With SCD and Controls in the Flanker and Elithorn Tasks (Accuracy and Response Times) HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 Flanker task: The two groups differed in performance, in both the congruent (attention) and incongruent (inhibition) conditions, F(1, 64) = 4.30, p < .05, ηp2 = .06, and F(1, 64) = 13.92, p < .001, ηp2= .18, respectively. Only the difference in the incongruent condition (inhibition) remained significant after applying Bonferroni corrections, however. Elithorn task: The analyses revealed significant differences between the two groups in response latency (i.e., planning), F(1, 64) = 15.42, p < .001, ηp2 = .19, and in accuracy, F(1, 64) =11.02, p = .001, ηp2= .14, but not in response times (execution speed), F(1, 64) = 0.01, p = .91, ηp2 < .001. Association Between Children’s Lexical Retrieval Skills and Nonverbal Executive Functioning Correlational Analyses Table III shows the correlations. Group correlated negatively with phonological fluency, r = −.35, p < .005, latency time in the Elithorn task (planning), r = −.44, p < .001, accuracy in the Elithorn task, r = −.38, p = .001, and accuracy in the incongruent (inhibition) and congruent (attention) Flanker task conditions, r = −39, p = .001 and r = −25, p < .05, respectively. The analyses also showed that phonological fluency was significantly associated with accuracy in the Elithorn task, r = .43, p < .001, accuracy in the incongruent Flanker task condition (inhibition), r = .41, p = .001, and accuracy in the congruent Flanker task condition (attention), r = .26, p < .05. When the significance level was adjusted using Bonferroni’s correction to .006, only the first four correlations involving the Group factor (with phonological fluency, planning times, planning accuracy, and inhibition), and the first two correlations between phonological fluency and accuracy in the Elithorn task, and inhibition in the Flanker task, reached statistical significance. The correlation between accuracy in the Elithorn task (i.e., planning) and in the incongruent Flanker task condition (i.e., response inhibition) was also significant and strong, r = .62, p < .001, indicating that the Elithorn task probably demands significant inhibitory control skills. Table III. Correlations Between Factors: Group, Boston Naming, Semantic Fluency, Phonological Fluency, Elithorn Latency Times, Elithorn Accuracy, Flanker Accuracy (in Congruent and Incongruent Conditions) Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Note. *p < .05; **p < .005. Table III. Correlations Between Factors: Group, Boston Naming, Semantic Fluency, Phonological Fluency, Elithorn Latency Times, Elithorn Accuracy, Flanker Accuracy (in Congruent and Incongruent Conditions) Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Note. *p < .05; **p < .005. Hierarchical Regressions Variables were entered in five steps: Group at Step 1, Boston naming test scores at Step 2, inhibition scores at Step 3, the interaction between Group and Boston naming test scores at Step 4, and the interaction between Group and inhibition scores at Step 5, resulting in five regression models. The parameters for the five regression models are listed in Table IV. The first three steps led to significant increases in terms of the variance explained (with differences in R2 of 0.15, 0.09, and 0.14, respectively, and ps equal to .001, .006, and <.001). When the inhibition scores were added at Step 3, the Group factor was no longer significant, suggesting that differences in inhibition scores could partly account for the differences between the two groups. Adding the Group by Boston naming test z-scores interaction prompted an increase in the amount of variance explained, which approached conventional levels of significance (ΔR2 = .04, p = .05). Follow-up models run separately for each group revealed that performance in the Boston naming test significantly predicted phonological fluency in the control group (β = 1.76, t = 3.24, p = .003), but not for children with HbSS SCD (β = .30, t = 0.44, p = .66). Flanker task inhibition z-scores were also entered in these analyses because the two groups differed in terms of inhibition skills. Only the inhibition z-scores explained the clinical group’s performance in phonological fluency. However, including the interaction between Group and Flanker task inhibition z-scores in Step 5 prompted no significant increase in the amount of variance explained (ΔR2 = 0.01, p = .22). Table IV. Hierarchical Multiple Regression Analyses Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Note. Dependent variable: Phonological fluency. Table IV. Hierarchical Multiple Regression Analyses Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Note. Dependent variable: Phonological fluency. To sum up, language proficiency (i.e., Boston naming test scores) only contributed significantly to performance in terms of phonological fluency in the control group, whereas the clinical group’s performance in phonological fluency was only significantly associated with inhibitory control skills. Further, although the Flanker task inhibition z-scores by Group interaction was not significant, including these scores in our regression model at Step 4 cancelled the effect of Group, suggesting that a part of the difference between the groups in terms of phonological fluency might be captured by differences in their inhibitory skills. Nine children in the SCD group revealed silent infarcts in the white matter in the MCA border zones. To ensure that our findings were not explained by their inclusion in the SCD group, we replicated the analyses after excluding these subjects. Even after this further selection, the two groups remained comparable for age, socioeconomic index, years in the country, and ethnicity. These repeat analyses confirmed the original results of the study. Discussion Children with SCD may show language deficits that are not always associated with evident neurological events or brain damage (Armstrong et al., 1996; Bernaudin et al., 2000; Brown et al., 2000; Schatz et al., 2009; Steen, Fineberg-Buchner, Hankins, Weiss, Prifitera, & Mulhern, 2005; Tarazi et al., 2007). The nature and origin of these problems are still unclear. In this study, we explored whether selective language deficits in children with high-risk SCD (HbSS genotype) could be associated with selective impairments in EF (e.g., attentional control and inhibition). To explore this hypothesis, we assessed children’s lexical retrieval across tasks requiring different degrees of executive control (picture naming, phonological fluency, and semantic fluency), and examined the association between their performance in these tasks and in nonverbal tasks assessing EF (attention, planning, and response inhibition). The influence of environmental factors (i.e., linguistic or socioeconomic disadvantage) on children’s performance was controlled by comparing children with HbSS SCD with healthy controls matched for socioeconomic level and bilingualism. In line with our initial hypothesis, our results showed that children with HbSS SCD only had a selective impairment in verbal and nonverbal tasks that demanded greater executive control, that is, planning, response inhibition, and phonological fluency. These deficits were apparently unrelated to any brain damage: 9 of the 32 participants with SCD had silent infarcts in the white matter in the MCA border zones, but our findings remained the same regardless of whether these cases were included or not. Socioeconomic disadvantage and bilingualism probably influenced the performance of the clinical group and the controls in the tasks involving lexical retrieval. On average, the performance of the children with SCD and the corresponding controls was >1 SD below the norm in the Boston naming test (Figure 1). It is clear from our findings, however, that the neurological condition associated with SCD contributed to the lexical retrieval difficulties of the children with HbSS over and above any environmental factors (see Schatz et al., 2009 for similar findings). In fact, the HbSS SCD group fared significantly worse in EF (inhibitory skills and planning skills) and phonological fluency than their demographically matched controls. These findings are consistent with previous research showing that environmental factors can contribute to, but not fully explain, the cognitive and language problems of children with SCD (Schatz et al., 2009; Steen et al., 2005; Tarazi et al., 2007). The most significant finding of this study, however, relates to the performance of the children with HbSS SCD in the lexical retrieval tasks. It was only in terms of phonological fluency that they performed significantly worse than the controls. Phonological fluency tasks are typically more demanding than naming or semantic fluency tasks for typically developing children too (Friesen et al., 2015; Riva et al. 2000). This result is also consistent with the hypothesis that the phonological fluency task demands more strategic skills for word searching than the other two tasks, and that phonological fluency thus relies more on executive control abilities (Friesen et al., 2015). Unlike findings in studies on monolingual children (Riva et al., 2000), the bilingual controls in our study found object naming and semantic fluency tasks considerably more difficult than the phonological fluency tasks. At first glance, this might seem surprising, but this result is in line with an extensive body of literature indicating that, while bilingualism delays language development (and vocabulary growth in the second language), it also enhances children’s executive control (Calvo & Bialystok, 2014; Luo et al., 2010). The advantage that our bilingual controls showed in the phonological fluency task may reflect this enhanced executive control effect (Luo et al., 2010). It is therefore likely that the children with SCD were unable to benefit from the effects of their bilingualism in the same way because of their disease (which significantly affected their response inhibition skills). The regression analyses revealed an interesting difference in the performance of the two groups (SCD vs. controls) in the phonological fluency task: whereas the control group’s language proficiency (i.e., breadth of expressive vocabulary) was significantly associated with performance in phonological fluency, this association was not significant for the group with HbSS SCD, and it was only their inhibition skills that accounted for their performance. The two groups did not differ significantly in terms of object naming skills, but the control group had a broader expressive vocabulary, which probably supported their retrieval of lexical items in the phonological fluency task. Their more efficient inhibition skills also helped them to better suppress irrelevant items. Other studies have shown that some degree of language proficiency is needed before bilingualism can significantly enhance a child’s EF (Bosma, Hoekstra, Versloot, & Blom, 2017). There may be thus reciprocal effects between the lower EF of the children with HbSS SCD and their slower second language acquisition. On the other hand, bilingualism could have a protective effect on children with SCD, and this would explain why their phonological fluency skills, though weaker than in the controls, were not significantly impaired vis-à-vis age-related norms (they were about 0.70 SD below the norms, see Figure 1). In future studies, comparisons between monolingual and bilingual children with SCD would be useful to explore the protective role of bilingualism in the development of the neurocognitive symptoms of SCD. Further research should also control for time of exposure to the additional language, as this variable could mediate the effects of bilingualism. An alternative interpretation of our findings could be that our children with SCD had silent focal brain strokes in areas serving phonological fluency processes. Broadmann’s area 44 in the frontal lobe seems to specialize in phonological processes (it is more active during phonological fluency tasks; Heim, Eickhoff, & Amunts, 2008, 2009). A focal cerebrovascular accident to this area 44 might therefore be consistent with our findings, but none of our nine patients with MRI abnormalities had any visible lesions involving Broadmann’s area 44. Based on these findings, it seems more likely that selective deficits occur in children with HbSS SCD owing to impairments in executive control and inhibition. As suggested by other studies (Brousse et al., 2015; Colombatti et al., 2015; DeBaun & Kirkham, 2016), these selective cognitive impairments in SCD could be due to chronic cerebral hypoxia, even in the absence of visible brain damage. This would explain the presence of significant selective EF deficits in our patients too, even though the majority (n = 23, 72%) of them showed no signs of brain damage. An important limitation of our study lies in the lack of any comparison between children with more and less severe forms of SCD, and different neurological complications (i.e., HbSS and HbSC; see Schatz et al., 2009). Such a comparison would have provided stronger evidence of how the related neurological impairments influence the children’s EF and language abilities. Unfortunately, the sample originally enrolled for the present study included only four children with low-risk (HbSC) SCD, and so we preferred to limit our analysis to the children with HbSS SCD. This also means that our findings cannot be generalized to children with SCD in general. Another issue limiting their generalizability concerns the specific sociodemographic profile of our study participants. Our sample only included children from immigrant communities who are bilingual. In other countries, such as the United States, such a situation is no longer common among children with SCD. Despite these limitations, the results of our study shed further light on the nature of the language problems experienced by children with high-risk SCD. A close relationship seems to emerge between their language difficulties and impairments in the EF domain. If these findings are supported by further studies, then recommendations should be developed on how to integrate the scaffolding of EF in language learning and rehabilitation interventions for children with SCD (see Hussey & Novick, 2012). Providing more opportunities for language learning, or speech therapy to support their language development, may not be enough to ensure significant language improvements in these children. Another noteworthy suggestion coming from this study is that bilingualism could have a protective role in SCD. If this hypothesis is supported by further research, there would be a clear indication for bilingual education to combat the negative sequelae of this disease. As mentioned above, studies comparing monolingual and bilingual children with SCD across cognitive and language tasks will be needed to confirm this hypothesis. Although such research is challenging, it is extremely important to address this issue to clarify the effects of bilingualism on cognitive development, and to make informed decisions in educational and rehabilitation settings. More work on specific sub-domains or skills (such as lexical retrieval) is in general necessary to map the neurocognitive and language deficits of children with SCD in greater detail, and to identify the most effective remedial interventions for this population. Funding The research was partially supported by grants from the Fondazione Città della Speranza 11/02, 14/02, and 16/04. Aknowledgments The authors thank Patrizia Montanaro for her invaluable help in participants recruitment and all children and families who took part in this study. Conflicts of interest: None declared. References Arfé B. , Vardanega T. , Bertin M. ( Unpublished ). The effects of computational thinking on first graders’ planning and response inhibition. Armstrong F. D. , Thompson R. J. Jr. , Wang W. , Zimmerman R. , Pegelow C. H. , Miller S. , Moser F. , Bello J. , Hurtig A. , Vass K. ( 1996 ). Cognitive functioning and brain magnetic resonance imaging in children with sickle cell disease . Pediatrics , 97 , 864 – 870 . Google Scholar PubMed Baldeweg T. , Hogan A. M. , Saunders D. E. , Telfer P. , Gadian D. G. , Vargha-Khadem F. , Kirkham F. J. ( 2006 ). Detecting white matter injury in sickle cell disease using voxel-based morphometry . Annals of Neurology , 59 , 662 – 672 . doi: 10.1002/ana.20790 Google Scholar CrossRef Search ADS PubMed Berkelhammer L. D. , Williamson A. L. , Sanford S. D. , Dirksen C. L. , Sharp W. G. , Margulies A. S. , Prengler R. A. ( 2007 ). Neurocognitive sequelae of pediatric sickle cell disease: A review of the literature . Child Neuropsychology , 13 , 120 – 131 . doi: 10.1080/09297040600800956 Google Scholar CrossRef Search ADS PubMed Bernaudin F. , Verlhac S. , Fréard F. , Roudot-Thoraval F. , Benkerrou M. , Thuret I. , Mardini R. , Vannier J. P. , Ploix E. , Romero M. , Cassé-Perrot C. , Helly M. , Gillard E. , Sebag G. , Kchouk H. , Pracros J. P. , Finck B. , Dacher J. N. , Ickowicz V. , Raybaud C. , Poncet M. , Lesprit E. , Reinert P. H. , Brugières P. ( 2000 ). Multicenter prospective study of children with sickle cell disease: Radiographic and psychometric correlation . Journal of Child Neurology , 15 , 333 – 343 . doi: 10.1177/088307380001500510 Google Scholar CrossRef Search ADS PubMed Bertella L. , Girelli L. , Grugni G. , Marchi S. , Molinari E. , Semenza C. ( 2005 ). Mathematical skills in Prader-Willi syndrome . Journal of Intellectual Disability Research , 49 , 159 – 169 . Google Scholar CrossRef Search ADS PubMed Blair C. , Raver C. C. ( 2016 ). Poverty, stress, and brain development: New directions for prevention and intervention . Academic Pediatrics , 16 , S30 – S36 . Google Scholar CrossRef Search ADS PubMed Bosma E. , Hoekstra E. , Versloot A. , Blom E. ( 2017 ). The minimal and short-lived effects of minority language exposure on the executive functions of Frisian-Dutch bilingual children . Frontiers in Psychology , 8 , 1453 . http://doi.org/10.3389/fpsyg.2017.01453 http://dx.doi.org/10.3389/fpsyg.2017.01453 Google Scholar CrossRef Search ADS PubMed Brandling-Bennett E. M. , White D. A. , Armstrong M. M. , Christ S. E. , DeBaun M. R. ( 2003 ). Patterns of verbal long-term and working memory performance reveal deficits in strategic processing in children with frontal infarcts related to sickle cell disease . Developmental Neuropsychology , 24 , 423 – 434 . Google Scholar CrossRef Search ADS PubMed Brousse V. , Kossorotoff M. , de Montalembert M. ( 2015 ). How I manage cerebral vasculopathy in children with sickle cell disease . British Journal of Haematology , 170 , 615 – 625 . http://dx.doi.org/10.1111/bjh.13477 Google Scholar CrossRef Search ADS PubMed Burkhardt L. , Lobitz S. , Koustenis E. , Rueckriege S. M. , Hernáiz Driever P. ( 2017 ). Cognitive and fine motor deficits in a pediatric sickle cell disease cohort of mixed ethnic origin . Annals of Hematology , 96 , 199 – 213 . Doi: 10.1007/s00277-016-2861-1 Google Scholar CrossRef Search ADS PubMed Brown R. T. , Davis P. C. , Lambert R. , Hsu L. , Hopkins K. , Eckman J. ( 2000 ). Neurocognitive functioning and magnetic resonance imaging in children with sickle gels disease . Journal of Pediatric Psychology , 25 , 503 – 513 . Google Scholar CrossRef Search ADS PubMed Carretti B. , Motta E. , Re A. M. ( 2016 ). Oral and written expression in children with reading comprehension difficulties . Journal of Learning Disabilities , 49 , 65 – 76 . doi: 10.1177/0022219414528539 Google Scholar CrossRef Search ADS PubMed Calvo A. , Bialystok E. ( 2014 ). Independent effects of bilingualism and socioeconomic status on language ability and executive functioning . Cognition , 130 , 278 – 288 . doi: 10.1016/j.cognition.2013.11.015 Google Scholar CrossRef Search ADS PubMed Colombatti R. , Ermani M. , Rampazzo P. , Manara R. , Montanaro M. , Basso G. , Meneghetti G. , Sainati L. ( 2015 ). Cognitive evoked potentials and neural networks are abnormal in children with sickle cell disease and not related to the degree of anaemia, pain and silent infarcts . British Journal of Haematology , 169 , 597 – 600 . Google Scholar CrossRef Search ADS PubMed Colombatti R. , Lucchetta M. , Montanaro M. , Rampazzo P. , Ermani M. , Talenti G. , Baracchini C. , Favero A. , Basso G. , Manara R. , Sainati L. ( 2016 ). Cognition and the default mode network in children with sickle cell disease: A resting state functional MRI study . PLoS One , 11 , e0157090. Google Scholar CrossRef Search ADS PubMed Colombatti R. , Palazzi G. , Masera N. , Notarangelo L. D. , Bonetti E. , Samperi P. , Barone A. , Perrotta S. , Facchini E. , Miano M. , Del Vecchio G. C. , Guerzoni M. E. , Corti P. , Menzato F. , Cesaro S. , Casale M. , Rigano P. , Forni G. L. , Russo G. , Sainati L. ; Italian Multicenter Study of Hydroxyurea in Sickle Cell Anemia Investigators . ( 2018 ). Hydroxyurea prescription, availability and use for children with sickle cell disease in Italy: Results of a national multicenter survey . Pediatric Blood and Cancer , 65 . Online first: doi: 10.1002/pbc.26774 Council of Europe . ( 2001 ). Common European framework of reference for languages: Learning, teaching, assessment . Cambridge University press . Retrieved from https://rm.coe.int/1680459f97 DeBaun M. R. , Kirkham F. J. ( 2016 ). Central nervous system complications and management in sickle cell disease . Blood , 127 , 829 – 838 . doi: 10.1182/blood-2015-09-618579 Google Scholar CrossRef Search ADS PubMed Della Sala S. , Gray C. , Spinnler H. , Trivelli C. ( 1998 ). Frontal lobe functioning in man: The riddle revisited . Archives of Clinical Neuropsychology , 13 , 663 – 682 . Google Scholar PubMed Drazen C. H. , Abel R. , Gabir M. , Farmer G. , King A. A. ( 2016 ). Prevalence of developmental delay and contributing factors among children with sickle cell disease . Pediatric Blood and Cancer , 63 , 504 – 510 . doi: 10.1002/pbc.25838 Google Scholar CrossRef Search ADS PubMed Friesen D. C. , Luo L. , Luk G. , Bialystok E. ( 2015 ). Proficiency and control in verbal fluency performance across the lifespan for monolinguals and bilinguals . Language, Cognition and Neuroscience , 30 , 238 – 250 . doi: 10.1080/23273798.2014.918630 Google Scholar CrossRef Search ADS PubMed Gugliotta M. , Bisiacchi P. S. , Cendron M. , Tressoldi P. E. , Vio C. ( 2009 ). BVN 12-18 Batteria di Valutazione Neuropsicologica per l’adolescenza [it. tr. Battery for the neuropsychological assessment in adolescence] . Trento : Erickson . Hassell K. L. ( 2010 ). Population estimates of sickle cell disease in the U.S . American Journal of Preventive Medicine , 38 , S512 – S521 . doi: 10.1016/j.amepre.2009.12.022. Google Scholar CrossRef Search ADS PubMed Heim S. , Eickhoff S. B. , Amunts K. ( 2008 ). Specialisation in Broca's region for semantic, phonological, and syntactic fluency? Neuroimage , 40 , 1362 – 1368 . doi: 10.1016/j.neuroimage.2008.01.009 Google Scholar CrossRef Search ADS PubMed Heim S. , Eickhoff S. B. , Amunts K. ( 2009 ). Different roles of cytoarchitectonic BA 44 and BA 45 in phonological and semantic verbal fluency as revealed by dynamic causal modelling . Neuroimage , 48 , 616 – 624 . http://doi.org/10.1016/j.neuroimage.2009.06.044 http://dx.doi.org/10.1016/j.neuroimage.2009.06.044 Google Scholar CrossRef Search ADS PubMed Hijmans C. T. , Fijnvandraat K. , Grootenhuis M. A. , van Geloven N. , Heijboer H. , Peters M. , Oosterlaan J. ( 2011 ). Neurocognitive deficits in children with sickle cell disease: A comprehensive profile . Pediatric Blood and Cancer , 56 , 783 – 788 . doi: 10.1002/pbc.22879 Google Scholar CrossRef Search ADS PubMed Hussey E. K. , Novick J. M. ( 2012 ). The benefits of executive control training and the implications for language processing . Frontiers in Psychology , 3 , 158 . doi: 10.3389/fpsyg.2012.00158 Google Scholar CrossRef Search ADS PubMed Kaplan E. , Goodglass H. , Weintraub S. ( 1983 ). The Boston naming test . Philadelphia, PA : Lea & Fibiger . Kohnert K. , Hernandez A. E. , Bates E. ( 1998 ). Bilingual performance on the Boston naming test: Preliminary norms in Spanish and English . Brain and Language , 65 , 422 – 440 . http://dx.doi.org/10.1006/brln.1998.2001 Google Scholar CrossRef Search ADS PubMed Kral M. C. , Brown R. T. , Hynd G. W. ( 2001 ). Neuropsychological aspects of pediatric sickle cell disease . Neuropsychology Review , 11 , 179 – 196 . doi: 10.1023/a: 1012901124088 Google Scholar CrossRef Search ADS PubMed Leuzzi V. , Pansini M. , Sechi E. , Chiarotti F. , Carducci C. , Levi G. , Antonozzi I. ( 2004 ). Executive function impairment in early-treated PKU subjects with normal mental development . Journal of Inherited Metabolic Disease , 27 , 115 – 125 . doi: 10.1023/B: BOLI.0000028781.94251.1f Google Scholar CrossRef Search ADS PubMed Locascio G. , Mahone E. M. , Eason S. H. , Cutting L. E. ( 2010 ). Executive dysfunction among children with reading comprehension deficits . Journal of Learning Disabilities , 43 , 441 – 454 . http://doi.org/10.1177/0022219409355476 http://dx.doi.org/10.1177/0022219409355476 Google Scholar CrossRef Search ADS PubMed Luo L. , Luk G. , Bialystok E. ( 2010 ). Effect of language proficiency and executive control on verbal fluency performance in bilinguals . Cognition , 114 , 29 – 41 . doi: 10.1016/j.cognition.2009.08.014 Google Scholar CrossRef Search ADS PubMed Manara R. , Talenti G. , Rampazzo P. , Ermani M. , Montanaro M. , Baracchini C. , Teso S. , Basso G. , Sainati L. , Colombatti R. ( 2017 ). Longitudinal evaluation of cerebral white matter hyperintensities lesion volume in children with sickle cell disease . British Journal of Haematology , 176 , 485 – 487 . Google Scholar CrossRef Search ADS PubMed Montanaro M. , Colombatti R. , Pugliese M. , Migliozzi C. , Zani F. , Guerzoni M. , Manoli S. , Manara R. , Meneghetti G. , Rampazzo P. , Cavalleri F. , Giordan M. , Paolucci P. , Basso G. , Palazzi G. , Sainati L. ( 2013 ). Intellectual function evaluation of first generation immigrant children with sickle cell disease: The role of language and sociodemographic factors . Italian Journal of Pediatrics , 39 , 36 . doi: 10.1186/1824-7288-39-36. Google Scholar CrossRef Search ADS PubMed Mulatti C. , Peressotti F. , Job R. , Saunders S. , Coltheart M. ( 2012 ). Reading aloud: The cumulative lexical interference effect . Psychonomic Bulletin and Review , 19 , 662 – 667 . 10.3758/s13423-012-0269-z Google Scholar CrossRef Search ADS PubMed Riva D. , Nichelli F. , Devoti M. ( 2000 ). Developmental aspects of verbal fluency and confrontation naming in children . Brain and Language , 71 , 267 – 284 . doi: 10.1006/brln.1999.2166. Google Scholar CrossRef Search ADS PubMed Sanchez C. E. , Schatz J. , Roberts C. W. ( 2010 ). Cerebral blood flow velocity and language functioning in pediatric sickle cell disease . Journal of the International Neuropsychological Society , 16 , 326 – 334 . doi: 10.1017/s1355617709991366 Google Scholar CrossRef Search ADS PubMed Schatz J. , Craft S. , Koby M. , Siegel M. J. , Resar L. , Lee R. R. , Chu J.-Y. , Launius G. , Dadash-Zadehm M. , DeBaun M. R. ( 1999 ). Neuropsychologic deficits in children with sickle cell disease and cerebral infarction: Role of lesion site and volume . Child Neuropsychology , 5 , 92 – 103 . doi: 10.1076/chin.5.2.92.3170 Google Scholar CrossRef Search ADS Schatz J. , Finke R. L. , Kellett J. M. , Kramer J. H. ( 2002 ). Cognitive functioning in children with sickle cell disease: A meta-analysis . Journal of Pediatric Psychology , 27 , 739 – 748 . Google Scholar CrossRef Search ADS PubMed Schatz J. , McClellan C. B. ( 2006 ). Sickle cell disease as a neurodevelopmental disorder . Mental Retardation and Developmental Disabilities Research Reviews , 12 , 200 – 207 . doi: 10.1002/mrdd.20115 Google Scholar CrossRef Search ADS PubMed Schatz J. , Puffer E. S. , Sanchez C. , Stancil M. , Roberts C. W. ( 2009 ). Language processing deficits in sickle cell disease in young school-age children . Developmental Neuropsychology , 34 , 122 – 136 . doi: 10.1080/87565640802499191 Google Scholar CrossRef Search ADS PubMed Spinnler H. , Tognoni G. ( 1987 ). Standardizzazione e taratura di test neuropsicologici . The Italian Journal of Neurological Sciences , 8 (Suppl 6) , 1 – 120 . Steen R. G. , Fineberg-Buchner C. , Hankins G. , Weiss L. , Prifitera A. , Mulhern R. K. ( 2005 ). Cognitive deficits in children with sickle cell disease . Journal of Child Neurology , 20 , 102 – 107 . doi: 10.1177/08830738050200020301 Google Scholar CrossRef Search ADS PubMed Tarazi R. A. , Grant M. L. , Ely E. , Barakat L. P. ( 2007 ). Neuropsychological functioning in preschool-age children with sickle cell disease: The role of illness-related and psychosocial factors . Child Neuropsychology , 13 , 155 – 172 . doi: 10.1080/09297040600611312 Google Scholar CrossRef Search ADS PubMed Thurstone T. G. , Thurstone L. L. ( 1963 ). Primary mental ability . Chicago, IL : Science Research . © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Pediatric Psychology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Pediatric Psychology Oxford University Press

Selective Difficulties in Lexical Retrieval and Nonverbal Executive Functioning in Children With HbSS Sickle Cell Disease

Loading next page...
 
/lp/ou_press/selective-difficulties-in-lexical-retrieval-and-nonverbal-executive-9YbQKKbqo6
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Pediatric Psychology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0146-8693
eISSN
1465-735X
D.O.I.
10.1093/jpepsy/jsy005
Publisher site
See Article on Publisher Site

Abstract

Abstract Language deficits in multilingual children with sickle cell disease (SCD) are poorly understood. We tested the hypothesis that selective language deficits in this population could relate to an impaired frontal lobe functioning often associated with high-risk homozygous HbS disease (HbSS). In all, 32 children from immigrant communities with HbSS SCD aged 6 to 12 years (mean age = 9.03, n = 9 with silent infarcts) and 35 demographically matched healthy controls (mean age = 9.14) were tested on their naming skills, phonological and semantic fluency, attention, and selected executive functions (response inhibition and planning skills). Analyses of variance showed significant differences between patients and controls in inhibition and planning (p = .001 and .001), and phonological fluency (p = .004). The poorer performance in phonological fluency of the children with SCD was not associated with any visible brain damage to language areas. Hierarchical regression analyses showed that, whereas the control children’s vocabulary knowledge explained their performance in the phonological fluency tasks, only inhibition skills accounted for variance in the performance of the children with SCD. These results suggest a selective impairment of verbal and nonverbal executive functioning (i.e., planning, inhibition, and phonological fluency) in children with SCD, with deficits possibly owing to frontal area hypoxia. executive functioning, language skills, lexical retrieval, sickle cell disease, verbal fluency Sickle cell disease (SCD) is an inherited disorder of hemoglobin, endemic in some regions of Africa, which has also spread in the European Union and the United States owing to migration flows. The disease is characterized by chronic hemolytic anemia, recurrent vaso-occlusive events, and a progressive vasculopathy, which may have negative effects on children’s health and neurocognitive functioning (DeBaun & Kirkham, 2016; Hijmans et al., 2011; Kral, Brown, & Hynd, 2001; Schatz & McClellan, 2006). The severity of the disease and the likelihood of neurocognitive deficits depends on the patient’s SCD genotype. Homozygous HbS disease (HbSS) and HbS0 thalassemia are characterized by lower hemoglobin levels, an earlier onset of cerebral vasculopathy, and a more severe phenotype, with a higher risk of cerebrovascular events than in the double heterozygous variants (HbSC and HbSß+) (Manara et al 2017; Schatz & McClellan, 2006). It is estimated that approximately 11% of children with HbSS SCD suffer from overt cerebral strokes. Occult neurological complications may develop too, however, and lead to selective cognitive impairments. These include silent infarcts, or subclinical cerebral strokes visible on brain imaging, affecting approximately 28% of children with SCD (Burkhardt, Lobitz, Koustenis, Rueckriege, & Hernáiz Driever, 2017), and small vessel vasculopathy, or insufficient oxygen and glucose delivery to brain tissues, which may result in brain function impairments in the absence of visible cerebral lesions (Baldeweg et al., 2006; Brousse, Kossorotoff, & de Montalembert, 2015; Colombatti et al., 2015, 2016; DeBaun & Kirkham, 2016). Children with SCD may consequently have a variety of neurocognitive deficits affecting their attention (Hijmans et al., 2011), executive functions such as inhibitory control, planning, and working memory (Brandling-Bennett, White, Armstrong, Christ, & DeBaun, 2003; Hijmans et al., 2011), visuospatial skills (Armstrong et al, 1996; Schatz & McClellan, 2006), and language skills (Berkelhammer et al., 2007; Kral et al., 2001; Sanchez, Schatz, & Roberts, 2010; Schatz, Puffer, Sanchez, Stancil, & Roberts, 2009), as well as a more general cognitive impairment (Berkelhammer et al., 2007). Although the neurocognitive profiles vary considerably, deficits in executive functioning (EF) seem to be characteristic of these children. Vaso-occlusion and hypoxia often affect the frontal brain regions, that is, areas implicated in response inhibition and higher-order EF, such as planning (Berkelhammer et al., 2007; Brown, Davis, Lambert, Hsu, Hopkins, & Eckman, 2000; Burkhardt et al., 2017; Hijmans et al., 2011; Kral et al., 2001; Schatz & McClellan, 2006). Deficits in verbal abilities are also fairly common among children with SCD (Brandling-Bennett et al., 2003; Sanchez et al., 2010; Schatz et al., 2009; Tarazi, Grant, Ely, & Barakat, 2007), but our understanding of these problems is still limited. Previous studies have shown that these impairments primarily affect children with the HbSS genotype, and relate more to the neurological effects of the disease than to its general medical complications (e.g., absences from school owing to hospitalizations; Schatz et al., 2009). Moreover, they are associated with cerebrovascular disruptions and high cerebral blood flow velocities on transcranial Doppler (TCD) ultrasound (Sanchez et al., 2010). Language deficits in SCD are not always associated with brain damage, however, as they sometimes emerge in children with normal neuroimaging findings too (Bernaudin et al., 2000; Schatz, Finke, Kellett, & Kramer, 2002). Some studies suggest a general weakness in the verbal abilities of children with high-risk SCD across language domains (syntax, semantics, and phonology), but their performance within these domains is not impaired to the same degree (Schatz et al., 2009). Research findings to date are also frequently inconsistent regarding the type of language skills that can be affected in children with SCD. Armstrong et al. (1996) reported deficits in vocabulary skills, whereas Brown et al. (2000) and Schatz et al. (2009) did not find any significant difficulties in expressive or receptive vocabulary, but they did find impairments in rapid naming (Brown et al., 2000) and syntactic skills (Schatz et al., 2009). Some researchers have suggested an association between deficits in verbal abilities and lesions in frontal areas (Sanchez et al., 2010; Schatz et al., 1999), but the nature of this association is still not clear. Verbal deficits could be the consequence of neurological insult to frontal lobe areas directly related to language processing, such as Broadmann’s areas 44 and 45 (Sanchez et al., 2010; Schatz et al., 2009). On the other hand, they could be a side effect of other neurocognitive impairments caused by SCD, such us deficits in working memory or EF (Brandling-Bennett et al., 2003). This article aims to further explore the nature of language problems in children with HbSS, examining the association between their performance in lexical retrieval tasks (picture naming, phonological fluency, and semantic fluency), and in tasks for testing attention and EF (planning and response inhibition) that are significantly affected by frontal lobe dysfunctions (Della Sala, Gray, Spinnler, & Trivelli, 1998; Friesen, Luo, Luk, & Bialystok, 2015). Lexical retrieval demands not only language proficiency, but also search strategies that require executive control (Friesen et al., 2015; Luo et al., 2010). In particular, in phonological and semantic fluency tasks, children must employ attentional resources to access their lexical knowledge and inhibit responses that do not fit the (semantic or phonological) criterion (Friesen et al., 2015). By examining children’s lexical retrieval across three types of tasks requiring different degrees of executive control (the Boston Naming Test, a phonological fluency task, and a semantic fluency task), and correlating performance in these tasks with measures of attention, planning, and inhibition, we tested the hypothesis of a selective impairment of lexical skills related to EF deficits. Verbal (phonological and semantic) fluency tasks are used to assess verbal EF, and are particularly effective for distinguishing between the role of executive control and language (vocabulary) knowledge (Friesen et al., 2015; Luo et al., 2010). Although both rely on vocabulary knowledge (i.e., on linguistic proficiency), they place different executive demands on lexical retrieval. Retrieving words based on semantic categories, as in semantic fluency tasks, is fairly automatic in children because words are organized in our memory in a semantic network, with items linked by semantic associations (Luo et al., 2010; Mulatti, Peressotti, Job, Saunders, & Coltheart, 2012). By contrast, phonological fluency requires greater executive control and inhibition skills because, to retrieve words from the same phonemic category (with the same initial sound), children must suppress the automatic retrieval of the lexical items (i.e., words) activated by semantic associations (Friesen et al., 2015). Past research has shown that environmental factors such as socioeconomic disadvantage or bilingualism also contribute to the language problems that may affect children with SCD (Drazen, Abel, Gabir, Farmer, & King, 2016; Montanaro et al., 2013; Tarazi et al., 2007). In the United States, most children with SCD come from African American communities, from families that are not first-generation immigrants or bilingual (Hassell, 2010). In the European Union, where immigration is a more recent phenomenon, most children with SCD come from immigrant and multilingual communities (Hijmans et al., 2011; Montanaro et al., 2013). Children of immigrant families typically show a slower growth in the vocabulary of their second language because they are bilingual (Calvo & Bialystok, 2014), or owing to the low socioeconomic status of their families (Blair & Raver, 2016; Calvo & Bialystok, 2014). Both these factors can affect their language performance: for example, socioeconomic status accounts for up to 47% of the variance in the language skills of preschoolers with SCD (Tarazi et al., 2007). In the present study, we compared the performance of children with HbSS SCD and healthy, demographically matched controls in naming and verbal fluency tasks (semantic and phonological fluency). The study had three main goals: (1) to investigate the influence of neurological (HbSS SCD) and environmental (socioeconomic status and bilingualism) factors on the lexical abilities of children with SCD; (2) to explore whether the children with HbSS SCD showed a profile of selective impairment in nonverbal and verbal tasks consistent with the hypothesis of a selective deficit in EF; and (3) to test the association between an impaired inhibitory control and the performance of these children in phonological and semantic fluency tasks. If, as expected, the cognitive and language problems of children with HbSS SCD are associated with the neurological risk characteristic of their disease, their performance in tasks of verbal fluency and nonverbal EF (inhibition and planning) should be impaired, even by comparison with that of healthy controls matched on socioeconomic level and bilingualism. If the children with HbSS SCD reveal selective problems with verbal language owing to their EF profile, then significant differences in the two groups’ language performance should emerge especially in word retrieval tasks that demand a greater executive control (i.e., phonological fluency). Methods Participants Thirty-two children with HbSS SCD aged 6 to 12 years (mean age = 9.03, SD = 2.04) and 35 healthy demographically matched controls (mean age = 9.14, SD = 1.40) were recruited and agreed to participate in the study. Table I shows the participants’ characteristics. Table I. Characteristics of Participants With HbSS SCD (n = 32) and Demographically Matched Controls (n = 35) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) Table I. Characteristics of Participants With HbSS SCD (n = 32) and Demographically Matched Controls (n = 35) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) HbSS SCD Controls Statistics p Age, M (SD) 9.03 (2.04) 9.14 (1.40) F = 0.069 .79 Gender, n girls (%) 17 (53.13%) 21 (60%) χ2 = 0.32 .57 Born in Italy (%) 81.25% 80% χ2 = 0.17 .897 Years in Italy, M (SD) 8.09 (2.63) 8.31 (2.61) F = 0.12 .73 Ethnicity χ2 = 1.89 .17 African, n 30 29 Other, n 2 6 Socioeconomic index (range 1–8) M (SD) 5.15 (0.95) 4.59 (1.09) F = 3.63 .063 Silent strokes, n (%) 9 (28%) – – – TCD velocities Abnormal, n (%) 4 (12%) Conditional, n (%) 1 (3%) Participants with SCD were selected for the study if they had the high-risk HbSS phenotype, were from 6 to 12 years old, had no history of overt cerebral strokes, and were in a steady state defined as at least 4 weeks since any vaso-occlusive crisis (VOC) or hospital admission. In addition to these inclusion criteria, the clinical psychologist at the hospital (the second author) who conducted the routine assessments of the children with SCD and the class teacher (for the controls) were consulted to ensure that participants had sufficient verbal skills to understand the instructions and complete the tasks involved. Thirty-four children with HbSS SCD who met these criteria were initially identified, but two did not complete all the tasks and were excluded post hoc. The final sample thus included 32 patients with HbSS SCD. Patients with HbSC SCD were also initially recruited for the study to enable a comparison between children at high (HbSS) and low (HbSC) neurological risk, but this proved impossible because only four patients with HbSC met the inclusion criteria. TCD velocities were available for all patients: four children had a history of abnormal TCD findings (all had silent infarcts too), while one had a history of conditional TCD. Magnetic resonance imaging (MRI) was available for most of the patients (n = 28). Nine children had silent infarcts in the white matter in the border zones of the middle cerebral artery (MCA; four of them had a history of abnormal TCD). Six patients had lesions involving both the left and the right MCA (these lesions were larger on the left in five-sixth cases), two had left MCA lesions, and one had a right MCA lesion. Seventeen patients (17 of 32, 53%) were receiving a disease-modifying treatment: five were on chronic transfusions owing to abnormal/conditional TCD; and 12 were taking hydroxyurea (HU) for previous recurrent VOCs or acute chest syndromes, or previous anemia <8 g/dl (these are standard indications for HU treatment in Italy) (Colombatti et al., 2018) The healthy controls enrolled for the study came from the same (mainly African and Eastern European) immigrant communities as the children with SCD; they were 6 to 12 years old and had no known cognitive, motor, or sensory disabilities. All participants (SCD group and control group) were from immigrant families. Their ethnicity was predominantly African (30 participants in the SCD group, and 29 controls). Most of the participants were born in Italy (26 SCD patients, 28 controls). Three children (one with SCD, and two controls) had migrated to Italy when they were ≤1 year old. All participants reportedly spoke more than one language and used languages other than Italian at home. Five controls and five children in the SCD group were sequential bilinguals (i.e., they were exposed to Italian as an additional language after the age of 2). The mean time of exposure to the second language (Italian) was 3.4 years for controls and 4.7 years for the SCD group. A Mann–Whitney test revealed no statistically significant differences between the two groups, p = .55. Two children in the SCD group were trilingual: Italian was their third language in both cases, and they had been exposed to Italian for 6 or 7 years. The controls were matched with the SCD patients by age, gender, ethnicity, years spent in Italy, and socioeconomic status. The children with SCD had a mean Intelligence quotient (IQ) of 89.6 (SD = 15.3). Procedure The study was approved by the ethical committee for psychological research at the University of Padova. Written informed consent to use the child’s data for research purposes was obtained from parents, and verbal assent was obtained from the children. Participants with SCD were recruited during routine visits to the Veneto Region’s reference center for SCD. The third author (a psychologist) introduced the study to the parents and informed them that participation was on a voluntary basis. Children were tested individually, before or after their scheduled visits. Healthy controls were recruited and tested at schools in the same geographical area as those attended by the children with SCD. Parents were informed about the study by the class teachers, who also collected their informed consent forms. Children whose families consented to their participation were assessed individually during school hours in a quiet room by a trained psychologist (the third author of this study). A questionnaire was used to collect relevant sociodemographic information from parents, that is, the family’s country of origin, the number of years the child had spent in the adopted country, the language(s) spoken at home, and the parents’ formal education and occupation. As mentioned above, candidates for the study were identified in advance by the second author (clinical psychologist at the hospital) and their class teacher (for the controls), based on their prior knowledge of the child, to ensure that participants had sufficient verbal skills to complete the tasks. For children not born in Italy, language skills were also briefly assessed by the third author of the study by means of a short conversation about their school experience, hobbies, and favorite games. All children had sufficient language skills to describe in simple terms their everyday experiences and talk about their personal interests (Council of Europe, 2001). Before performing each task, the children were administered two or three practice items, and their comprehension of the tasks was ascertained from their performance. Measures Socioeconomic Index The children’s socioeconomic status was judged from their parents’ formal education, scored from 0 (less than elementary school) to 4 (college), and their occupation, scored from 1 (unemployed) to 4 (professional roles). A composite score was calculated as the sum of the highest education score and the highest occupation score obtained by either parent (see also Calvo & Bialystok, 2014). Naming The Boston Naming Test (Kaplan, Goodglass & Weintraub, 1983) was administered to assess the children’s naming skills. The task involves naming aloud 60 objects pictured in black and white. Normative data exist for Italian school-age children (aged 5–11 years, from 5 years and 11 months to 11 years and 4 months; Riva, Nichelli & Devoti, 2000). The test has also been used to assess bilingual naming skills (Kohnert, Hernandez, & Bates, 1998). The procedure used by Riva et al. (2000) was also adopted in our study to enable comparisons. No cues were given to help the children name the pictures, and the test was interrupted after six consecutive wrong or missed responses. The total number of correct responses was scored, and z-scores were calculated based on age-related normative data (Riva et al., 2000). Riva et al.’s normative data for 11-year-olds were used to compute the z-scores of the 12-year-old participants in this study (the difference between the two age groups ranged from 6 to 8 months). Two standardized tasks were used to assess fluency in word retrieval, one for phonological fluency and the other for semantic fluency (Riva et al., 2000). Both tasks have been used in prior research also with bilingual school-age children (age 7 and 10 years, Friesen et al., 2015). Phonological Fluency In this task, the child is allowed 1 min to produce as many different words as possible that begin with a given letter or sound (B and S). The mean number of correct items retrieved in the two conditions (B/S) is computed. Repeated words, proper names, and invented words are disregarded. Z-scores are calculated, based on available age-related norms (Riva et al., 2000). Semantic Fluency The semantic fluency task involves producing in 1 min as many different words as possible in a given category (Animals or Food). Here again, the mean number of correct items retrieved in the two conditions is computed, and z-scores are calculated, based on available normative data (Riva et al., 2000). Selected nonverbal EFs were assessed with Elithorn’s task and the Flanker task. Elithorn’s Perceptual Maze Test This is a traditional “frontal” neuropsychological test, considered particularly sensitive to frontal lobe deficits (Della Sala et al., 1998). It assesses nonverbal planning (Spinnler & Tognoni, 1987) and has been used in previous research to examine the strategic planning skills of children with executive dysfunctions (Leuzzi et al., 2004; Locascio, Mahone, Eason, & Cutting, 2010), and patients with intellectual impairment (Bertella et al., 2005). A short version of the task, standardized for Italian adolescents aged 12–18 years (Batteria per la Valutazione Neuropsicologica; Gugliotta et al., 2009), has also been used successfully in studies with younger children, from the age of 8 years (Leuzzi et al., 2004), and by the first author of this study in another recent study on a sample (N = 57) of typically developing 6-year-olds (Arfé, Vardanega, & Bertin, Unpublished). The same short version was used in the present study too. The test consists of eight items (mazes), and the child is asked to connect a number of black dots, arranged randomly on grids, while obeying three rules: tracing lines from the bottom up; not crossing over the grid; and not backtracking. A maximum time of 2 min is allowed to complete each maze. Three scores are computed: (1) response latency times (in seconds), corresponding to the time elapsing between completing the instructions and when the child starts tracing the path on the grid (which provides a measure of the time spent planning); (2) response times (in seconds), or the speed with which the child completes the maze (which gives a measure of executive ability); and (3) accuracy, corresponding to the overall number of mazes correctly completed within 2 min. After receiving the instructions, the children could practice with three examples. Even the youngest children in the study understood the rules and how to perform the practice items. The test started when the children demonstrated that they fully understood the task and the rules. Flanker Task Experimental Flanker tasks have been used to assess bilingual children’s attention and inhibitory control (Calvo & Bialystok, 2014). In our implementation of the task, the stimuli consisted of a horizontal target arrow displayed in the middle of the computer screen, pointing to the left or right, and flanked by two distractors on both sides. The distractors could be straight lines (neutral condition), or arrows pointing either in the same direction as the target (congruent condition), or in the opposite direction (incongruent condition). The three experimental conditions (neutral, congruent, and incongruent) were equally represented in the experiment (with 96 trials for each one). The stimuli were displayed in black against a white background. Each trial started with a fixation cross displayed for 400 ms, immediately followed by the appearance of the stimulus, which remained on the screen until the child gave a response, or for up to 2 s. The next trial began immediately afterward. Participants were instructed to indicate whether the central arrow pointed to left or right, ignoring the flankers. Participants sat in front of the computer screen, and gave their answer by pressing keys on the keyboard (A for a target pointing to the left, L for a target pointing to the right). The experimental procedure and data acquisition were controlled with E Prime 2 software (Psychology Software Tool, Inc.), recording accuracy and reaction times for each response. Accuracy was scored for each condition as the proportion of correct responses out of the overall number of items presented. Practice trials were used at the beginning of the testing session to enable the child to become familiar with the task. Spatial Relations Subtest, Primary Mental Abilities Scale As deficits in visuospatial skills are sometimes reported in the literature on SCD (Armstrong et al, 1996; Schatz & McClellan, 2006), and may affect performance in the Flanker and Elithorn tasks, this subtest of the Primary Mental Abilities was used to estimate participants’ visuospatial abilities (Thurstone & Thurstone, 1963). Using the Spatial Relations subtest also enabled us to estimate nonverbal (visuospatial) intelligence in the control group, for whom no IQ scores were available (Carretti, Motta, & Re, 2016). The test is traditionally used with school-age children (Carretti, Motta, & Re, 2016). Children are asked to identify one of four shapes that complete a geometrical figure. They are given 6 min to complete all the task items. Response accuracy and total time are recorded. Z-scores are computed based on the test age-related norms. Data Analysis Preliminary between-group comparisons were carried out to check that the HbSS SCD and control groups did not differ on relevant sociodemographic variables. Analyses of variance (ANOVAs) were performed to test between-group differences in age, years in the country, and socioeconomic index. Differences in ethnicity were tested with a chi-square analysis. Then between-group ANOVAs were run to test for differences in visuospatial skills, attention, inhibitory control, planning, and verbal language skills (naming and fluency) between the children with HbSS SCD and their healthy demographically matched controls. Because multiple comparisons were performed, the level of significance was set at .005 to control for Type I error (Bonferroni formulas). Age-adjusted scores were available for the visuospatial, naming, and verbal fluency tasks, so variance owing to age was not further controlled in these analyses. For the Flanker and Elithorn tasks, age was covaried in the analyses because age-equivalent scores were unavailable. The association between children’s executive control and their verbal fluency was explored using Pearson’s correlations and hierarchical regression analyses. The correlations were run on the whole sample and correlations with the dummy factor group (SCD group. 1, and control group. 0) were observed. Hierarchical regression analyses were only performed for phonological fluency because the ANOVAs only revealed significant differences between the two groups in this area (see Results section). Phonological fluency can be influenced by both language proficiency (i.e., breadth of expressive vocabulary) and executive control (i.e., inhibition processes) (Friesen et al., 2015), and these skills could play a different role in children with HbSS SCD and controls. The regression analysis was consequently conducted on the overall sample of participants. The unique contribution of group, breadth of expressive vocabulary, and inhibition skills, and the interaction between group and the other two factors, were thus explored. The Group factor was entered at Step 1, the Boston naming test z-scores at Step 2, the inhibition z-scores (for the mean proportions of accurate answers in the incongruent Flanker task condition) at Step 3, the interaction between Group and Boston naming test z-scores at Step 4, and the interaction between Group and inhibition z-scores at the fifth and final step. Finally, all analyses were replicated after excluding the nine participants in the SCD group who revealed silent infarcts to ensure that the results were not owing to their inclusion in the study sample. Results Preliminary Analyses As shown in Table I, the two groups did not differ on any of the sociodemographic factors considered (age, gender, ethnicity, years in the country, socioeconomic index). The difference in socioeconomic index approached significance, however, with children with SCD having relatively higher socioeconomic scores than controls. Visuospatial Skills, Verbal Language Skills (Naming and Fluency), Attention, Inhibitory Control, and Planning All between-group differences went in the expected direction. Visuospatial skills. No differences emerged between the two groups’ visuospatial skills, F(1, 65) = 0.52, p = .82, ηp2 = .001, and remarkably, both groups’ performance was age-appropriate, as shown in Figure 1. Figure 1. View largeDownload slide Children’s performance (in z-scores) in visuospatial, Boston naming, phonological fluency, and semantic fluency tests. Figure 1. View largeDownload slide Children’s performance (in z-scores) in visuospatial, Boston naming, phonological fluency, and semantic fluency tests. Naming and verbal fluency. No significant differences emerged between the two groups for the Boston naming or semantic fluency tests [F(1, 65) = 2.28, p = .14, ηp2= .03 and F(1, 65) = 0.15, p = .69, ηp2 = .002], whereas the two groups differed significantly in the phonological fluency task, F(1, 65) = 9.09, p = .004, ηp2= .12 (see also Figure 1). Attention, inhibitory control, and planning. The differences between the two groups (HbSS SCD vs. controls) are summarized in Table II. Table II. Differences Between Children With SCD and Controls in the Flanker and Elithorn Tasks (Accuracy and Response Times) HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 Table II. Differences Between Children With SCD and Controls in the Flanker and Elithorn Tasks (Accuracy and Response Times) HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 HbSS SCD Controls Statistics p M (SD) M (SD) Elithorn accuracy n correct 4.19 (2.4) 5.89 (1.60) F = 11.02 .001 Elithorn Perceptual Maze—latency (seconds) 13.88 (11.36) 26.53 (14.40) F = 15.42 .001 Elithorn Perceptual Maze—response time (seconds) 5.54 (3.45) 5.76 (2.28) F = 0.012 .913 Flanker in congruent condition—accuracy (%) 0.85 (0.15) 0.92 (0.105) F = 4.30 .042 Flanker in incongruent condition—accuracy (%) 0.62 (0.23) 0.80 (0.20) F = 13.92 .001 Flanker task: The two groups differed in performance, in both the congruent (attention) and incongruent (inhibition) conditions, F(1, 64) = 4.30, p < .05, ηp2 = .06, and F(1, 64) = 13.92, p < .001, ηp2= .18, respectively. Only the difference in the incongruent condition (inhibition) remained significant after applying Bonferroni corrections, however. Elithorn task: The analyses revealed significant differences between the two groups in response latency (i.e., planning), F(1, 64) = 15.42, p < .001, ηp2 = .19, and in accuracy, F(1, 64) =11.02, p = .001, ηp2= .14, but not in response times (execution speed), F(1, 64) = 0.01, p = .91, ηp2 < .001. Association Between Children’s Lexical Retrieval Skills and Nonverbal Executive Functioning Correlational Analyses Table III shows the correlations. Group correlated negatively with phonological fluency, r = −.35, p < .005, latency time in the Elithorn task (planning), r = −.44, p < .001, accuracy in the Elithorn task, r = −.38, p = .001, and accuracy in the incongruent (inhibition) and congruent (attention) Flanker task conditions, r = −39, p = .001 and r = −25, p < .05, respectively. The analyses also showed that phonological fluency was significantly associated with accuracy in the Elithorn task, r = .43, p < .001, accuracy in the incongruent Flanker task condition (inhibition), r = .41, p = .001, and accuracy in the congruent Flanker task condition (attention), r = .26, p < .05. When the significance level was adjusted using Bonferroni’s correction to .006, only the first four correlations involving the Group factor (with phonological fluency, planning times, planning accuracy, and inhibition), and the first two correlations between phonological fluency and accuracy in the Elithorn task, and inhibition in the Flanker task, reached statistical significance. The correlation between accuracy in the Elithorn task (i.e., planning) and in the incongruent Flanker task condition (i.e., response inhibition) was also significant and strong, r = .62, p < .001, indicating that the Elithorn task probably demands significant inhibitory control skills. Table III. Correlations Between Factors: Group, Boston Naming, Semantic Fluency, Phonological Fluency, Elithorn Latency Times, Elithorn Accuracy, Flanker Accuracy (in Congruent and Incongruent Conditions) Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Note. *p < .05; **p < .005. Table III. Correlations Between Factors: Group, Boston Naming, Semantic Fluency, Phonological Fluency, Elithorn Latency Times, Elithorn Accuracy, Flanker Accuracy (in Congruent and Incongruent Conditions) Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Variable 1 2 3 4 5 6 7 8 1. Group 1 2. Boston naming −.18 1 3. Semantic_fluency .05 .02 1 4. Phonol_fluency −.35** .40** .10 1 5. Elithorn_latency −.44** .02 .05 .19 1 6. Elithorn_accuracy −.38** .09 .06 .43** .42** 1 7. Flanker_accuracy_congruent −.25* −.08 .04 .26* .03 .42** 1 8. Flanker_accuracy incongruent −.39** .06 −.05 .41** .18 .62** .66** 1 Note. *p < .05; **p < .005. Hierarchical Regressions Variables were entered in five steps: Group at Step 1, Boston naming test scores at Step 2, inhibition scores at Step 3, the interaction between Group and Boston naming test scores at Step 4, and the interaction between Group and inhibition scores at Step 5, resulting in five regression models. The parameters for the five regression models are listed in Table IV. The first three steps led to significant increases in terms of the variance explained (with differences in R2 of 0.15, 0.09, and 0.14, respectively, and ps equal to .001, .006, and <.001). When the inhibition scores were added at Step 3, the Group factor was no longer significant, suggesting that differences in inhibition scores could partly account for the differences between the two groups. Adding the Group by Boston naming test z-scores interaction prompted an increase in the amount of variance explained, which approached conventional levels of significance (ΔR2 = .04, p = .05). Follow-up models run separately for each group revealed that performance in the Boston naming test significantly predicted phonological fluency in the control group (β = 1.76, t = 3.24, p = .003), but not for children with HbSS SCD (β = .30, t = 0.44, p = .66). Flanker task inhibition z-scores were also entered in these analyses because the two groups differed in terms of inhibition skills. Only the inhibition z-scores explained the clinical group’s performance in phonological fluency. However, including the interaction between Group and Flanker task inhibition z-scores in Step 5 prompted no significant increase in the amount of variance explained (ΔR2 = 0.01, p = .22). Table IV. Hierarchical Multiple Regression Analyses Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Note. Dependent variable: Phonological fluency. Table IV. Hierarchical Multiple Regression Analyses Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Steps Predictors β t p Step 1 Group −.39 −3.43 .001 Step 2 Group −.33 −3.03 .004 Boston z .31 2.83 .006 Step 3 Group −.17 −1.60 .11 Boston z .32 3.15 .003 Flanker z .40 3.77 .000 Step 4 Group −.38 −2.55 .013 Boston z .47 3.76 <.001 Flanker z .41 3.93 <.001 Boston z × Group −.34 −1.98 .05 Step 5 Group −.39 −2.58 .01 Boston z .46 3.73 <.001 Flanker z .28 1.78 .08 Boston z × Group −.34 −1.98 .05 Flanker z × Group .17 1.16 .25 Note. Dependent variable: Phonological fluency. To sum up, language proficiency (i.e., Boston naming test scores) only contributed significantly to performance in terms of phonological fluency in the control group, whereas the clinical group’s performance in phonological fluency was only significantly associated with inhibitory control skills. Further, although the Flanker task inhibition z-scores by Group interaction was not significant, including these scores in our regression model at Step 4 cancelled the effect of Group, suggesting that a part of the difference between the groups in terms of phonological fluency might be captured by differences in their inhibitory skills. Nine children in the SCD group revealed silent infarcts in the white matter in the MCA border zones. To ensure that our findings were not explained by their inclusion in the SCD group, we replicated the analyses after excluding these subjects. Even after this further selection, the two groups remained comparable for age, socioeconomic index, years in the country, and ethnicity. These repeat analyses confirmed the original results of the study. Discussion Children with SCD may show language deficits that are not always associated with evident neurological events or brain damage (Armstrong et al., 1996; Bernaudin et al., 2000; Brown et al., 2000; Schatz et al., 2009; Steen, Fineberg-Buchner, Hankins, Weiss, Prifitera, & Mulhern, 2005; Tarazi et al., 2007). The nature and origin of these problems are still unclear. In this study, we explored whether selective language deficits in children with high-risk SCD (HbSS genotype) could be associated with selective impairments in EF (e.g., attentional control and inhibition). To explore this hypothesis, we assessed children’s lexical retrieval across tasks requiring different degrees of executive control (picture naming, phonological fluency, and semantic fluency), and examined the association between their performance in these tasks and in nonverbal tasks assessing EF (attention, planning, and response inhibition). The influence of environmental factors (i.e., linguistic or socioeconomic disadvantage) on children’s performance was controlled by comparing children with HbSS SCD with healthy controls matched for socioeconomic level and bilingualism. In line with our initial hypothesis, our results showed that children with HbSS SCD only had a selective impairment in verbal and nonverbal tasks that demanded greater executive control, that is, planning, response inhibition, and phonological fluency. These deficits were apparently unrelated to any brain damage: 9 of the 32 participants with SCD had silent infarcts in the white matter in the MCA border zones, but our findings remained the same regardless of whether these cases were included or not. Socioeconomic disadvantage and bilingualism probably influenced the performance of the clinical group and the controls in the tasks involving lexical retrieval. On average, the performance of the children with SCD and the corresponding controls was >1 SD below the norm in the Boston naming test (Figure 1). It is clear from our findings, however, that the neurological condition associated with SCD contributed to the lexical retrieval difficulties of the children with HbSS over and above any environmental factors (see Schatz et al., 2009 for similar findings). In fact, the HbSS SCD group fared significantly worse in EF (inhibitory skills and planning skills) and phonological fluency than their demographically matched controls. These findings are consistent with previous research showing that environmental factors can contribute to, but not fully explain, the cognitive and language problems of children with SCD (Schatz et al., 2009; Steen et al., 2005; Tarazi et al., 2007). The most significant finding of this study, however, relates to the performance of the children with HbSS SCD in the lexical retrieval tasks. It was only in terms of phonological fluency that they performed significantly worse than the controls. Phonological fluency tasks are typically more demanding than naming or semantic fluency tasks for typically developing children too (Friesen et al., 2015; Riva et al. 2000). This result is also consistent with the hypothesis that the phonological fluency task demands more strategic skills for word searching than the other two tasks, and that phonological fluency thus relies more on executive control abilities (Friesen et al., 2015). Unlike findings in studies on monolingual children (Riva et al., 2000), the bilingual controls in our study found object naming and semantic fluency tasks considerably more difficult than the phonological fluency tasks. At first glance, this might seem surprising, but this result is in line with an extensive body of literature indicating that, while bilingualism delays language development (and vocabulary growth in the second language), it also enhances children’s executive control (Calvo & Bialystok, 2014; Luo et al., 2010). The advantage that our bilingual controls showed in the phonological fluency task may reflect this enhanced executive control effect (Luo et al., 2010). It is therefore likely that the children with SCD were unable to benefit from the effects of their bilingualism in the same way because of their disease (which significantly affected their response inhibition skills). The regression analyses revealed an interesting difference in the performance of the two groups (SCD vs. controls) in the phonological fluency task: whereas the control group’s language proficiency (i.e., breadth of expressive vocabulary) was significantly associated with performance in phonological fluency, this association was not significant for the group with HbSS SCD, and it was only their inhibition skills that accounted for their performance. The two groups did not differ significantly in terms of object naming skills, but the control group had a broader expressive vocabulary, which probably supported their retrieval of lexical items in the phonological fluency task. Their more efficient inhibition skills also helped them to better suppress irrelevant items. Other studies have shown that some degree of language proficiency is needed before bilingualism can significantly enhance a child’s EF (Bosma, Hoekstra, Versloot, & Blom, 2017). There may be thus reciprocal effects between the lower EF of the children with HbSS SCD and their slower second language acquisition. On the other hand, bilingualism could have a protective effect on children with SCD, and this would explain why their phonological fluency skills, though weaker than in the controls, were not significantly impaired vis-à-vis age-related norms (they were about 0.70 SD below the norms, see Figure 1). In future studies, comparisons between monolingual and bilingual children with SCD would be useful to explore the protective role of bilingualism in the development of the neurocognitive symptoms of SCD. Further research should also control for time of exposure to the additional language, as this variable could mediate the effects of bilingualism. An alternative interpretation of our findings could be that our children with SCD had silent focal brain strokes in areas serving phonological fluency processes. Broadmann’s area 44 in the frontal lobe seems to specialize in phonological processes (it is more active during phonological fluency tasks; Heim, Eickhoff, & Amunts, 2008, 2009). A focal cerebrovascular accident to this area 44 might therefore be consistent with our findings, but none of our nine patients with MRI abnormalities had any visible lesions involving Broadmann’s area 44. Based on these findings, it seems more likely that selective deficits occur in children with HbSS SCD owing to impairments in executive control and inhibition. As suggested by other studies (Brousse et al., 2015; Colombatti et al., 2015; DeBaun & Kirkham, 2016), these selective cognitive impairments in SCD could be due to chronic cerebral hypoxia, even in the absence of visible brain damage. This would explain the presence of significant selective EF deficits in our patients too, even though the majority (n = 23, 72%) of them showed no signs of brain damage. An important limitation of our study lies in the lack of any comparison between children with more and less severe forms of SCD, and different neurological complications (i.e., HbSS and HbSC; see Schatz et al., 2009). Such a comparison would have provided stronger evidence of how the related neurological impairments influence the children’s EF and language abilities. Unfortunately, the sample originally enrolled for the present study included only four children with low-risk (HbSC) SCD, and so we preferred to limit our analysis to the children with HbSS SCD. This also means that our findings cannot be generalized to children with SCD in general. Another issue limiting their generalizability concerns the specific sociodemographic profile of our study participants. Our sample only included children from immigrant communities who are bilingual. In other countries, such as the United States, such a situation is no longer common among children with SCD. Despite these limitations, the results of our study shed further light on the nature of the language problems experienced by children with high-risk SCD. A close relationship seems to emerge between their language difficulties and impairments in the EF domain. If these findings are supported by further studies, then recommendations should be developed on how to integrate the scaffolding of EF in language learning and rehabilitation interventions for children with SCD (see Hussey & Novick, 2012). Providing more opportunities for language learning, or speech therapy to support their language development, may not be enough to ensure significant language improvements in these children. Another noteworthy suggestion coming from this study is that bilingualism could have a protective role in SCD. If this hypothesis is supported by further research, there would be a clear indication for bilingual education to combat the negative sequelae of this disease. As mentioned above, studies comparing monolingual and bilingual children with SCD across cognitive and language tasks will be needed to confirm this hypothesis. Although such research is challenging, it is extremely important to address this issue to clarify the effects of bilingualism on cognitive development, and to make informed decisions in educational and rehabilitation settings. More work on specific sub-domains or skills (such as lexical retrieval) is in general necessary to map the neurocognitive and language deficits of children with SCD in greater detail, and to identify the most effective remedial interventions for this population. Funding The research was partially supported by grants from the Fondazione Città della Speranza 11/02, 14/02, and 16/04. Aknowledgments The authors thank Patrizia Montanaro for her invaluable help in participants recruitment and all children and families who took part in this study. Conflicts of interest: None declared. References Arfé B. , Vardanega T. , Bertin M. ( Unpublished ). The effects of computational thinking on first graders’ planning and response inhibition. Armstrong F. D. , Thompson R. J. Jr. , Wang W. , Zimmerman R. , Pegelow C. H. , Miller S. , Moser F. , Bello J. , Hurtig A. , Vass K. ( 1996 ). Cognitive functioning and brain magnetic resonance imaging in children with sickle cell disease . Pediatrics , 97 , 864 – 870 . Google Scholar PubMed Baldeweg T. , Hogan A. M. , Saunders D. E. , Telfer P. , Gadian D. G. , Vargha-Khadem F. , Kirkham F. J. ( 2006 ). Detecting white matter injury in sickle cell disease using voxel-based morphometry . Annals of Neurology , 59 , 662 – 672 . doi: 10.1002/ana.20790 Google Scholar CrossRef Search ADS PubMed Berkelhammer L. D. , Williamson A. L. , Sanford S. D. , Dirksen C. L. , Sharp W. G. , Margulies A. S. , Prengler R. A. ( 2007 ). Neurocognitive sequelae of pediatric sickle cell disease: A review of the literature . Child Neuropsychology , 13 , 120 – 131 . doi: 10.1080/09297040600800956 Google Scholar CrossRef Search ADS PubMed Bernaudin F. , Verlhac S. , Fréard F. , Roudot-Thoraval F. , Benkerrou M. , Thuret I. , Mardini R. , Vannier J. P. , Ploix E. , Romero M. , Cassé-Perrot C. , Helly M. , Gillard E. , Sebag G. , Kchouk H. , Pracros J. P. , Finck B. , Dacher J. N. , Ickowicz V. , Raybaud C. , Poncet M. , Lesprit E. , Reinert P. H. , Brugières P. ( 2000 ). Multicenter prospective study of children with sickle cell disease: Radiographic and psychometric correlation . Journal of Child Neurology , 15 , 333 – 343 . doi: 10.1177/088307380001500510 Google Scholar CrossRef Search ADS PubMed Bertella L. , Girelli L. , Grugni G. , Marchi S. , Molinari E. , Semenza C. ( 2005 ). Mathematical skills in Prader-Willi syndrome . Journal of Intellectual Disability Research , 49 , 159 – 169 . Google Scholar CrossRef Search ADS PubMed Blair C. , Raver C. C. ( 2016 ). Poverty, stress, and brain development: New directions for prevention and intervention . Academic Pediatrics , 16 , S30 – S36 . Google Scholar CrossRef Search ADS PubMed Bosma E. , Hoekstra E. , Versloot A. , Blom E. ( 2017 ). The minimal and short-lived effects of minority language exposure on the executive functions of Frisian-Dutch bilingual children . Frontiers in Psychology , 8 , 1453 . http://doi.org/10.3389/fpsyg.2017.01453 http://dx.doi.org/10.3389/fpsyg.2017.01453 Google Scholar CrossRef Search ADS PubMed Brandling-Bennett E. M. , White D. A. , Armstrong M. M. , Christ S. E. , DeBaun M. R. ( 2003 ). Patterns of verbal long-term and working memory performance reveal deficits in strategic processing in children with frontal infarcts related to sickle cell disease . Developmental Neuropsychology , 24 , 423 – 434 . Google Scholar CrossRef Search ADS PubMed Brousse V. , Kossorotoff M. , de Montalembert M. ( 2015 ). How I manage cerebral vasculopathy in children with sickle cell disease . British Journal of Haematology , 170 , 615 – 625 . http://dx.doi.org/10.1111/bjh.13477 Google Scholar CrossRef Search ADS PubMed Burkhardt L. , Lobitz S. , Koustenis E. , Rueckriege S. M. , Hernáiz Driever P. ( 2017 ). Cognitive and fine motor deficits in a pediatric sickle cell disease cohort of mixed ethnic origin . Annals of Hematology , 96 , 199 – 213 . Doi: 10.1007/s00277-016-2861-1 Google Scholar CrossRef Search ADS PubMed Brown R. T. , Davis P. C. , Lambert R. , Hsu L. , Hopkins K. , Eckman J. ( 2000 ). Neurocognitive functioning and magnetic resonance imaging in children with sickle gels disease . Journal of Pediatric Psychology , 25 , 503 – 513 . Google Scholar CrossRef Search ADS PubMed Carretti B. , Motta E. , Re A. M. ( 2016 ). Oral and written expression in children with reading comprehension difficulties . Journal of Learning Disabilities , 49 , 65 – 76 . doi: 10.1177/0022219414528539 Google Scholar CrossRef Search ADS PubMed Calvo A. , Bialystok E. ( 2014 ). Independent effects of bilingualism and socioeconomic status on language ability and executive functioning . Cognition , 130 , 278 – 288 . doi: 10.1016/j.cognition.2013.11.015 Google Scholar CrossRef Search ADS PubMed Colombatti R. , Ermani M. , Rampazzo P. , Manara R. , Montanaro M. , Basso G. , Meneghetti G. , Sainati L. ( 2015 ). Cognitive evoked potentials and neural networks are abnormal in children with sickle cell disease and not related to the degree of anaemia, pain and silent infarcts . British Journal of Haematology , 169 , 597 – 600 . Google Scholar CrossRef Search ADS PubMed Colombatti R. , Lucchetta M. , Montanaro M. , Rampazzo P. , Ermani M. , Talenti G. , Baracchini C. , Favero A. , Basso G. , Manara R. , Sainati L. ( 2016 ). Cognition and the default mode network in children with sickle cell disease: A resting state functional MRI study . PLoS One , 11 , e0157090. Google Scholar CrossRef Search ADS PubMed Colombatti R. , Palazzi G. , Masera N. , Notarangelo L. D. , Bonetti E. , Samperi P. , Barone A. , Perrotta S. , Facchini E. , Miano M. , Del Vecchio G. C. , Guerzoni M. E. , Corti P. , Menzato F. , Cesaro S. , Casale M. , Rigano P. , Forni G. L. , Russo G. , Sainati L. ; Italian Multicenter Study of Hydroxyurea in Sickle Cell Anemia Investigators . ( 2018 ). Hydroxyurea prescription, availability and use for children with sickle cell disease in Italy: Results of a national multicenter survey . Pediatric Blood and Cancer , 65 . Online first: doi: 10.1002/pbc.26774 Council of Europe . ( 2001 ). Common European framework of reference for languages: Learning, teaching, assessment . Cambridge University press . Retrieved from https://rm.coe.int/1680459f97 DeBaun M. R. , Kirkham F. J. ( 2016 ). Central nervous system complications and management in sickle cell disease . Blood , 127 , 829 – 838 . doi: 10.1182/blood-2015-09-618579 Google Scholar CrossRef Search ADS PubMed Della Sala S. , Gray C. , Spinnler H. , Trivelli C. ( 1998 ). Frontal lobe functioning in man: The riddle revisited . Archives of Clinical Neuropsychology , 13 , 663 – 682 . Google Scholar PubMed Drazen C. H. , Abel R. , Gabir M. , Farmer G. , King A. A. ( 2016 ). Prevalence of developmental delay and contributing factors among children with sickle cell disease . Pediatric Blood and Cancer , 63 , 504 – 510 . doi: 10.1002/pbc.25838 Google Scholar CrossRef Search ADS PubMed Friesen D. C. , Luo L. , Luk G. , Bialystok E. ( 2015 ). Proficiency and control in verbal fluency performance across the lifespan for monolinguals and bilinguals . Language, Cognition and Neuroscience , 30 , 238 – 250 . doi: 10.1080/23273798.2014.918630 Google Scholar CrossRef Search ADS PubMed Gugliotta M. , Bisiacchi P. S. , Cendron M. , Tressoldi P. E. , Vio C. ( 2009 ). BVN 12-18 Batteria di Valutazione Neuropsicologica per l’adolescenza [it. tr. Battery for the neuropsychological assessment in adolescence] . Trento : Erickson . Hassell K. L. ( 2010 ). Population estimates of sickle cell disease in the U.S . American Journal of Preventive Medicine , 38 , S512 – S521 . doi: 10.1016/j.amepre.2009.12.022. Google Scholar CrossRef Search ADS PubMed Heim S. , Eickhoff S. B. , Amunts K. ( 2008 ). Specialisation in Broca's region for semantic, phonological, and syntactic fluency? Neuroimage , 40 , 1362 – 1368 . doi: 10.1016/j.neuroimage.2008.01.009 Google Scholar CrossRef Search ADS PubMed Heim S. , Eickhoff S. B. , Amunts K. ( 2009 ). Different roles of cytoarchitectonic BA 44 and BA 45 in phonological and semantic verbal fluency as revealed by dynamic causal modelling . Neuroimage , 48 , 616 – 624 . http://doi.org/10.1016/j.neuroimage.2009.06.044 http://dx.doi.org/10.1016/j.neuroimage.2009.06.044 Google Scholar CrossRef Search ADS PubMed Hijmans C. T. , Fijnvandraat K. , Grootenhuis M. A. , van Geloven N. , Heijboer H. , Peters M. , Oosterlaan J. ( 2011 ). Neurocognitive deficits in children with sickle cell disease: A comprehensive profile . Pediatric Blood and Cancer , 56 , 783 – 788 . doi: 10.1002/pbc.22879 Google Scholar CrossRef Search ADS PubMed Hussey E. K. , Novick J. M. ( 2012 ). The benefits of executive control training and the implications for language processing . Frontiers in Psychology , 3 , 158 . doi: 10.3389/fpsyg.2012.00158 Google Scholar CrossRef Search ADS PubMed Kaplan E. , Goodglass H. , Weintraub S. ( 1983 ). The Boston naming test . Philadelphia, PA : Lea & Fibiger . Kohnert K. , Hernandez A. E. , Bates E. ( 1998 ). Bilingual performance on the Boston naming test: Preliminary norms in Spanish and English . Brain and Language , 65 , 422 – 440 . http://dx.doi.org/10.1006/brln.1998.2001 Google Scholar CrossRef Search ADS PubMed Kral M. C. , Brown R. T. , Hynd G. W. ( 2001 ). Neuropsychological aspects of pediatric sickle cell disease . Neuropsychology Review , 11 , 179 – 196 . doi: 10.1023/a: 1012901124088 Google Scholar CrossRef Search ADS PubMed Leuzzi V. , Pansini M. , Sechi E. , Chiarotti F. , Carducci C. , Levi G. , Antonozzi I. ( 2004 ). Executive function impairment in early-treated PKU subjects with normal mental development . Journal of Inherited Metabolic Disease , 27 , 115 – 125 . doi: 10.1023/B: BOLI.0000028781.94251.1f Google Scholar CrossRef Search ADS PubMed Locascio G. , Mahone E. M. , Eason S. H. , Cutting L. E. ( 2010 ). Executive dysfunction among children with reading comprehension deficits . Journal of Learning Disabilities , 43 , 441 – 454 . http://doi.org/10.1177/0022219409355476 http://dx.doi.org/10.1177/0022219409355476 Google Scholar CrossRef Search ADS PubMed Luo L. , Luk G. , Bialystok E. ( 2010 ). Effect of language proficiency and executive control on verbal fluency performance in bilinguals . Cognition , 114 , 29 – 41 . doi: 10.1016/j.cognition.2009.08.014 Google Scholar CrossRef Search ADS PubMed Manara R. , Talenti G. , Rampazzo P. , Ermani M. , Montanaro M. , Baracchini C. , Teso S. , Basso G. , Sainati L. , Colombatti R. ( 2017 ). Longitudinal evaluation of cerebral white matter hyperintensities lesion volume in children with sickle cell disease . British Journal of Haematology , 176 , 485 – 487 . Google Scholar CrossRef Search ADS PubMed Montanaro M. , Colombatti R. , Pugliese M. , Migliozzi C. , Zani F. , Guerzoni M. , Manoli S. , Manara R. , Meneghetti G. , Rampazzo P. , Cavalleri F. , Giordan M. , Paolucci P. , Basso G. , Palazzi G. , Sainati L. ( 2013 ). Intellectual function evaluation of first generation immigrant children with sickle cell disease: The role of language and sociodemographic factors . Italian Journal of Pediatrics , 39 , 36 . doi: 10.1186/1824-7288-39-36. Google Scholar CrossRef Search ADS PubMed Mulatti C. , Peressotti F. , Job R. , Saunders S. , Coltheart M. ( 2012 ). Reading aloud: The cumulative lexical interference effect . Psychonomic Bulletin and Review , 19 , 662 – 667 . 10.3758/s13423-012-0269-z Google Scholar CrossRef Search ADS PubMed Riva D. , Nichelli F. , Devoti M. ( 2000 ). Developmental aspects of verbal fluency and confrontation naming in children . Brain and Language , 71 , 267 – 284 . doi: 10.1006/brln.1999.2166. Google Scholar CrossRef Search ADS PubMed Sanchez C. E. , Schatz J. , Roberts C. W. ( 2010 ). Cerebral blood flow velocity and language functioning in pediatric sickle cell disease . Journal of the International Neuropsychological Society , 16 , 326 – 334 . doi: 10.1017/s1355617709991366 Google Scholar CrossRef Search ADS PubMed Schatz J. , Craft S. , Koby M. , Siegel M. J. , Resar L. , Lee R. R. , Chu J.-Y. , Launius G. , Dadash-Zadehm M. , DeBaun M. R. ( 1999 ). Neuropsychologic deficits in children with sickle cell disease and cerebral infarction: Role of lesion site and volume . Child Neuropsychology , 5 , 92 – 103 . doi: 10.1076/chin.5.2.92.3170 Google Scholar CrossRef Search ADS Schatz J. , Finke R. L. , Kellett J. M. , Kramer J. H. ( 2002 ). Cognitive functioning in children with sickle cell disease: A meta-analysis . Journal of Pediatric Psychology , 27 , 739 – 748 . Google Scholar CrossRef Search ADS PubMed Schatz J. , McClellan C. B. ( 2006 ). Sickle cell disease as a neurodevelopmental disorder . Mental Retardation and Developmental Disabilities Research Reviews , 12 , 200 – 207 . doi: 10.1002/mrdd.20115 Google Scholar CrossRef Search ADS PubMed Schatz J. , Puffer E. S. , Sanchez C. , Stancil M. , Roberts C. W. ( 2009 ). Language processing deficits in sickle cell disease in young school-age children . Developmental Neuropsychology , 34 , 122 – 136 . doi: 10.1080/87565640802499191 Google Scholar CrossRef Search ADS PubMed Spinnler H. , Tognoni G. ( 1987 ). Standardizzazione e taratura di test neuropsicologici . The Italian Journal of Neurological Sciences , 8 (Suppl 6) , 1 – 120 . Steen R. G. , Fineberg-Buchner C. , Hankins G. , Weiss L. , Prifitera A. , Mulhern R. K. ( 2005 ). Cognitive deficits in children with sickle cell disease . Journal of Child Neurology , 20 , 102 – 107 . doi: 10.1177/08830738050200020301 Google Scholar CrossRef Search ADS PubMed Tarazi R. A. , Grant M. L. , Ely E. , Barakat L. P. ( 2007 ). Neuropsychological functioning in preschool-age children with sickle cell disease: The role of illness-related and psychosocial factors . Child Neuropsychology , 13 , 155 – 172 . doi: 10.1080/09297040600611312 Google Scholar CrossRef Search ADS PubMed Thurstone T. G. , Thurstone L. L. ( 1963 ). Primary mental ability . Chicago, IL : Science Research . © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Pediatric Psychology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Pediatric PsychologyOxford University Press

Published: Feb 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off