Down syndrome (DS) is characterized by difficulties in both intellectual functioning and adaptive behavior. These sets of abilities are considered as separate but related domains with small to moderate correlations. The main objective of this study was to explore the relationship of intellectual functioning and adaptive behavior in adolescents with DS because previous studies have shown different relationship patterns between these constructs across other syndromes. Fifty-three adolescents with DS were assessed regarding their intellectual functioning whereas adaptive behavior was reported by parents and teachers. Participants showed a better performance on verbal than nonverbal tasks when assessing intellectual functioning, contrary to previous findings. Regarding adaptive behavior, higher social skills were reported than conceptual and practical skills. Intellectual functioning and adaptive behavior showed a medium correlation, consistent with observations in typical population. These results support the exploration of the variability across the DS phenotype.
Down syndrome (DS) is the most common chromosomal disorder caused by a complete or partial extra copy of chromosome 21 (Steingass, Chicoine, McGuire, & Roizen, 2011). Globally, there has been an increase in pregnancies resulting in DS (from 13.1 to 18.2 in 10,000), that is largely explained by the rise in maternal age (Carothers, Hecht, & Hook, 1999). However, due to the development of better prenatal diagnostic techniques and the increased use of procedures to terminate a pregnancy, the incidence of DS has remained stable at about 0.8 in 10,000 (Cocchi et al., 2010). In Chile, where the deliberate termination of pregnancies is less common, the rise in incidence of DS increased from 1.96 per 1,000 between 1982 and 2001 to 2.47 per 1,000 between 1995 and 2008 (Nazer & Cifuentes, 2011). The high incidence of DS has prompted the need to better understand how their developmental pathways may differ from children and adolescents without DS.
DS is usually associated with an atypical development and difficulties in both intellectual functioning and adaptive behavior (Chapman & Hesketh, 2000; Grieco, Pulsifer, Seligsohn, Skotko, & Schwartz, 2015; Steingass et al., 2011). Intellectual functioning is usually conceptualized as a general factor of intelligence that includes reasoning, planning, problem solving, thinking, comprehending complex ideas, learning quickly, and learning from experience (Gottfredson, 1997; Tassé, Luckasson, & Schalock, 2016). Adaptive behavior, on the other hand, corresponds with learned conceptual, social, and practical skills that are performed by people in their everyday lives (Tassé et al., 2012). These two areas are key in the diagnosis of intellectual disability (ID), which is characterized by significant limitation in both intellectual functioning and adaptive behavior, originating before age 18 (Schalock et al., 2010). Thus, DS has usually been associated with ID (Caban-Holt, Head, & Schmitt, 2014), although great variability exists (Karmiloff-Smith et al., 2016).
Regarding intellectual functioning in DS, studies generally describe gradual declines in IQ from infancy—rather than a deterioration of cognition—due to slowed development relative to their typically developing peers (Tsao & Kindelberger, 2009). Hence, although longitudinal studies have reported decreasing IQ scores in tests of intelligence, most raw scores and mental age scores tend to rise with chronological age (Grieco et al., 2015). On average, IQ scores in people with DS range from 20 to 70, with the mean being in the 40s (Steingass et al., 2011).
The cognitive phenotype reported is often characterized by a relative strength in visual problem-solving skills and relative weakness in expressive language, short-term verbal memory, long-term memory, and working memory (Chapman & Hesketh, 2000; Grieco et al., 2015; Steingass et al., 2011). Specifically, language is usually characterized by delays in syntax and phonological processing that can affect both expression and higher order comprehension. There is also a trend to analyze visuospatial information using a global approach. The observed deficits in tasks of explicit long-term memory seem to be associated with problems in encoding and retrieval (Grieco et al., 2015).
Adaptive behavior is commonly mentioned as a strength in people with DS, not only within themselves as compared to their cognitive abilities, but also when compared with other genetic syndromes associated with ID (Chapman & Hesketh, 2000; Di Nuovo & Buono, 2011; Rodrigue, Morgan, & Geffken, 1991; Steingass et al., 2011). The common adaptive behavior profile described in young children with DS is characterized by relative strengths in the social domain and daily living skills, and deficits in communication and motor skills (Fidler, Hepburn, & Rogers, 2006). Several studies have pointed out that the relative weakness in communication observed throughout development is associated with deficits in expressive language and is not a generalized limitation in all aspects of communication (e.g., Chapman & Hesketh, 2000; Di Nuovo & Buono, 2011; Dykens, Hodapp, & Evans, 1994; Fidler et al., 2006).
Although several studies have reported on the behavioral phenotype that characterizes DS, researchers have also pointed out the great variability within these individuals. As with the typical population, individual differences in genetic, brain, cognition, and behavior have been described in DS. Authors specify the need to not only consider them as a group, but also to study their variability (Karmiloff-Smith et al., 2016). For example, in a study with children with DS, Tsao and Kindelberger (2009) identified four cognitive profiles: (a) one cluster with relatively equivalent scores in verbal and nonverbal tasks, close to the means; (b) a second cluster with difficulties across tasks, but particularly in verbal tests; (c) a third cluster with better performance on verbal tests; and (d) a fourth cluster with strengths in nonverbal tasks.
As far as we know, there are no studies that have specifically explored the relationship between intellectual functioning and adaptive behavior in DS. The precise relationship between intellectual functioning and adaptive behavior has been debated in the diagnosis of ID in general. A study that modeled different forms of relationship between these two constructs (unrelated, related but different, and identical) concluded that intellectual abilities and adaptive behavior are better explained as separate but related domains, with small to moderate correlations. This suggested that inclusion of both measures in an assessment is appropriate and provides complementary information (Keith, Fehrmann, Harrison, & Pottebaum, 1987). Tassé and colleagues (2016) stressed the importance of considering both constructs as equally important in the diagnosis of ID and to understand that their relationship is correlational, not causal.
In a study comparing children with autism spectrum disorder (ASD) and children with ID (without ASD), matched on IQ and chronological age, it was observed that the correlations between intellectual and adaptive functioning were higher for the ASD group (Carpentieri & Morgan, 1996); however, in the ID group, the only significant correlations were between the cognitive measures and the communication domain of adaptive behavior (Carpentieri & Morgan, 1996). These differences show the possibility of specific patterns of correlations across syndromes.
In DS other variables have been studied. For example, one study found a positive correlation between adaptive behavior and chronological age in children between 1 and 11.5 years old, with stronger correlations in younger children and increased intragroup variability beginning at age 6 or 7 (Dykens et al., 1994). Libb, Myers, Graham, and Bell (1983) reported negative correlations between age and IQ scores and age and adaptive behavior scores, respectively, in their longitudinal study with 3-month-olds to 22-year-olds. The authors hypothesized that these results might reflect a decline in these abilities and/or generational changes in parenting behaviors towards children with DS. Authors also reported positive correlations between intelligence and adaptive behavior with the level of parental education.
Differences have also been reported along varied demographic characteristics. For example, when comparing institutionalized and home-based groups, researchers reported that both groups had similar declines in IQ with aging but found significant differences in the rate of social/adaptive skills decline, with more pronounced slopes observed in the home-based population (Brown, Greer, Aylward, & Hunt, 1990). They hypothesized that these differences are related to an emphasis on structured daily routines and semi-independent skills in the institution, whereas in a home environment other members of the house can assume more responsibilities (Brown et al., 1990). Another study found that children of mothers with higher educational levels had significantly higher IQ scores when compared to groups with mothers with lower educational levels (Sharav, Collins, & Shlomo, 1985).
As an exploratory study, our objective was to explore the relationship between performance on measures of intellectual functioning and adaptive behavior and how chronological age, education, and socioeconomic status (SES) impact intellectual functioning and adaptive behavior in adolescents with DS. As stated before, correlations between intellectual functioning and adaptive behavior have not been explored in DS. Furthermore, the effects of sociodemographic variables on development have been studied mostly in younger children, whereas the effects of age have been studied either across the different stages of childhood or between wide age ranges up until adulthood, but not specifically in adolescence. In this regard, adolescence is a time period of great changes in intellectual and adaptive development (Steinberg & Morris, 2001).
For this study 56 adolescents were included between the ages of 12 to 17 years old at the initial assessment, all with a medical diagnosis of DS confirmed with a karyotype. Three participants were excluded due to the presence of a comorbid ASD diagnosis. The final sample consisted of 53 adolescents with DS (16 females and 37 males) with a mean age of 14.57 years (SD = 1.39). Participants were recruited with the assistance of DS organizations in Santiago, Chile. More descriptive information can be found in Table 1.
Wechsler Intelligence Scale for Children-3rd edition, (WISC-III), Chilean adaptation and standardization (
Ramirez & Rosas, 2010; based on Wechsler, 1991)
The WISC-III is a comprehensive intelligence test for children ages 6 years, 0 months to 16 years, 11 months. The test comprises six verbal subtests and seven performance subtests. The Verbal Intelligence Quotient (VIQ) and the Performance Intelligence Quotient (PIQ) are obtained by adding the five main subtests of each area, and the Full-Scale Intelligence Quotient (FSIQ) is based on the total raw score. The WISC-III also includes four index scores: (1) Verbal Comprehension; (2) Perceptual Organization; (3) Freedom from Distractibility; and (4) Processing Speed. Each of these scales and indexes are expressed in standardized IQ scores (M = 100, DS = 15). The Chilean version was adapted from the Argentinian Spanish translation and standardized on 1,924 Chilean children. Reliability and validity were established (Ramírez & Rosas, 2007).
Wechsler Adult Intelligence Scale-4th edition, (WAIS-IV), Chilean adaptation and standardization (
Rosas, Tenorio, & Pizarro, 2013; based on Wechsler, 2008)
The WAIS-IV is the adult version of the intelligence test for individuals between 16 years, 0 months to 90 years, 11 months. The test yields four index scores: (1) Verbal Comprehension (VCI); (2) Perceptual Reasoning (PRI); (3) Working Memory (WMI); and (4) Processing Speed (PSI). The first two index scores include three of the main subtests and two complementary ones, whereas the last two index scores include two other main subtests and a complementary one. The FSIQ is obtained by adding the scores of each main subtests. All index scores and the FSIQ are transformed onto a standardized IQ metric. The Chilean version was standardized on 887 adults based on the original U.S. version. Reliability and validity were established (Rosas et al., 2014).
Adaptive Behavior Assessment System-II Spanish adaptation and standardization (
Montero-Zenteno & Fernández-Pinto, 2013; based on Harrison & Oakland, 2004)
The ABAS-II is a scale used to assess a wide range of skills necessary for personal and social sufficiency in daily activities in children and adults from birth to 89 years. The ABAS-II is completed directly by a respondent, typically the assessed person's parents and/or teacher(s) in the case of infants, children, and adolescents. The ABAS-II yields standard scores in three main domains: (1) Conceptual (2) Social, and (3) Practical, and a General Adaptive Composite (GAC) score. The three domain scores and the GAC are transformed onto a standard metric (M = 100, SD = 15). Reliability and validity have previously been established (Montero-Zenteno & Fernández-Pinto, 2013). Because there are no standardized Chilean tests for the assessment of adaptive behavior, we opted for the use of this Spanish adaptation normed on a comparable Spanish population.
The data for this study were obtained from a larger study investigating the effects of an intervention program in the development of executive functions in adolescents with DS. (The corresponding author can be contacted for more information about the larger study.) Parents of the participants were invited to an interview where the objectives and procedures of the study and intervention were explained to them. If they agreed to participate, consent and assent forms (for the participant adolescent) were reviewed and signed. Participants were then assessed following a standard procedure including either the WISC-III (Ramirez & Rosas, 2010) or the WAIS-IV (Rosas et al., 2013), depending on their chronological age; the Wisconsin Card Sorting Test (Heaton, Chelune, Talley, Kay, & Curtiss, 1993); and the Corsi Block-Tapping Task (Corsi, 1973; the results of the later two are not considered for this study). The parents and teachers of the participants were asked to independently complete the ABAS-II (Montero-Zenteno & Fernández-Pinto, 2013).
This study was conducted in accordance with the ethical standards presented in the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board for Ethical Research at the Universidad de los Andes (Chile).
For this study, intellectual functioning, adaptive behavior, age, education, and SES were measured. Intellectual functioning was measured with either the WISC-III or the WAIS-IV, depending on the age of the participant (12- to 15-year-olds were assessed with the WISC-III, and 16- to 17-year-olds with the WAIS-IV). From these tests, we obtained the FSIQ as a measure of general intellectual functioning. For verbal functioning, we analyzed the VIQ from the WISC-III and the VCI from the WAIS-IV. For nonverbal functioning, we analyzed the PIQ from the WISC-III and the PRI from the WAIS-IV.
Adaptive behavior was measured with the ABAS-II with separate reports from parents and teachers. For the study, the nine ability scales (i.e., Communication, Functional academics, Self-direction, Leisure, Social, Community use, Home/school living, Health and safety, and Self-care), the three main domains (i.e., Conceptual, Social, and Practical), and the GAC were used. An explanation of each scale and how are they distributed between domains can be seen in Table 2.
Education was measured in two ways: First we divided the adolescents between the ones enrolled in a school (in school) and the ones not attending school (not in school). We called that variable schooling. We created a second related variable called type of education, in which we divided the participants that attended school between those integrated into general education classrooms (general) and those enrolled in special schools for students with disabilities (special). It is important to note that all students that did not attend school, and four students (11.1%) who did attend school, were also enrolled in institutions that offered nonacademic programs for people with DS (e.g., independent living skills, vocational training, arts).
SES (low, middle, high) was assessed using the type of funding received by the schools the participants attend. In Chile, schools are classified into three categories, depending on their source of funding: (1) public schools, which are fully funded by the state; (2) combined-funded schools that receive funding from both the state and parents (i.e., private pay); and (3) private schools, which are fully supported by fees paid by the families. Research has shown that the type of school attended by children and adolescents in Chile is directly related to their family's SES level, particularly their income: low SES families typically enroll their children in public schools, middle SES families typically enroll their children in combined-funded schools, and high SES families typically enroll their children in private schools (García-Huidobro & Belleï, 2003). Hence, type of school attended by the child and its source of funding has been used as a proxy for family SES in several other studies with Chilean population (e.g., Catalán & Santelices, 2014; del Río & Balladares, 2010; Hernández, Unanue, Gaete, Cassorla, & Codner, 2007).
For some of the analyses, the participants were classified in three age groups: (a) early adolescence (12–13 years old); (b) middle adolescence (14–15 years old); and (c) late adolescence (16–17 years old).
Data were analyzed using SPSS Statistics 24 (IBM Corporation, 2016). Fisher's exact tests showed no statistically significant differences in the distribution of gender (male, female) across age groups (early adolescence, middle adolescence, late adolescence), p = .081; schooling (in school, not in school), p = .530; type of education (general, special, not in school), p = .681; and SES (low, middle, high), p = .664. Age groups were also equally distributed across schooling, p = .140; and type of education, p = .076; but not across SES, p = .021, were a large part of participants (35.8%) belong to the middle adolescence, high SES group.
A t test for correlated samples revealed that the adolescents had significantly better scores in VIQ/VCI than PIQ/PRI, t(52) = 3.83, p < .001 (two-tailed). IQ means and standard deviations can be seen in Table 3.
We then examined the three main domains of the ABAS-II reported by parents (i.e., Conceptual, Social, and Practical). A Mauchly's test showed that sphericity cannot be assumed, χ2(2) = 11.86, p = .003; thus, we used a Huynh-Feldt correction to interpret our results. A repeated-measures ANOVA determined that the means were significantly different, F(1.70, 88.61) = 69.20, p < .001. A follow-up test for correlated samples with Bonferroni adjusted alpha (α = .016) indicated that the Social domain scores were significantly higher than both the Conceptual domain scores, t(52) = −9.84, p < .001 (two-tailed) and the Practical domain scores, t(52) = 8.05, p < .001 (two-tailed). The Practical domain scores were significantly higher than the Conceptual domain scores, t(52) = −3.59, p = .001 (two-tailed). The same pattern of results was seen on the ABAS-II reported by teachers: the Mauchly's test showed that sphericity cannot be assumed, χ2(2) = 8.27, p = .016. Thus, we used a Huynh-Feldt correction for the repeated-measures ANOVA that revealed significantly different means, F(1.78, 87.41) = 61.46, p < .001. Using a correlated samples t test with Bonferroni adjusted alpha (α = .016) we observed that the Social domain scores were significantly higher than both the Conceptual domain scores, t(50) = −9.68, p < .001 (two-tailed) and the Practical domain scores, t(49) = 5.89, p < .001 (two-tailed). The Practical domain scores were significantly higher than the Conceptual domain scores, t(49) = −6.49, p < .001 (two-tailed). Mean domain scores obtained from parent's and teacher's ABAS-II ratings are presented in Table 4.
Comparisons using t tests for correlated samples indicated no significant differences between the ABAS-II Conceptual, Social, and Practical domain scores, and the GAC, when comparing the results reported by parents and teachers, all p > .050. Considering each specific scale (raw scores), t tests for correlated samples revealed significant differences between reports by parents and teachers in communication, t(50) = 2.90, p = .006 (two-tailed); leisure, t(50) = 4.95, p < .001 (two-tailed); social, t(50) = 3.72, p = .001 (two-tailed); home/school living, t(49) = −2.13, p = .038 (two-tailed); health and safety, t(50) = 5.20, p < .001 (two-tailed); and self-care, t(50) = 6.04, p < .001 (two-tailed). Refer again to Table 4.
When analyzing data separately by age group (i.e., early adolescence, middle adolescence, and late adolescence), we found differences in FSIQ, VIQ/VCI, and PIQ/PRI. A one-way ANOVA indicated that these differences were statistically significant, FSIQ: F(2,50) = 8.87, MSE = 32.17, p = .001; VIQ/VCI: F(2,50) = 9.42, MSE = 72.09, p < .001; PIQ/PRI: F(2,50) = 9.11, MSE = 51.20, p < .001. Post hoc analyses using Fisher's LSD showed the same pattern of results across IQ scores: For FSIQ, there were significant differences between early (M = 40.42, SD = 1.17) and late adolescence (M = 47.93, SD = 10.99), p = .001; and middle (M = 40.52, SD = .98) and late adolescence, p < .001. For VIQ/VCI, there were significant differences between early (M = 46.42, SD = 2.78) and late adolescence (M = 57.79, SD = 15.12), p = .001; and middle (M = 46.26, SD = 4.60) and late adolescence, p < .001. For PIQ/PRI, there were significant differences between early (M = 45.83, SD = 2.04) and late adolescence (M = 54.21, SD = 13.29), p = .004; and middle (M = 44.33, SD = 2.90) and late adolescence, p < .001 (see Figure 1). In adaptive behavior, parents reported statistically significant differences in self-direction between age groups, F(2,50) = 4.40, MSE = 10.66, p = .017. Post hoc analyses determined that the significant difference was between early (M = 2.33, SD = 1.67) and late adolescence (M = 6.14, SD = 4.11), p = .005. Teachers did not report significant differences in adaptive behavior scores across age groups.
Regarding schooling (i.e., in school, not in school), there were no significant differences in IQ or adaptive behavior scores (reported by parents or teachers). When grouping the sample by type of education (i.e., general, special, not in school), a one-way ANOVA showed significant differences in communication, F(2,50) = 5.29, MSE = 14.68, p = .008; functional academics, F(2,50) = 3.69, MSE = 4.20, p = .032; and the Conceptual domain, F(2,50) = 4.14, MSE = 150.51, p = .022; as reported by parents. As can be seen in Figures 2 and 3, post hoc analyses showed a significant difference in communication between adolescents in general schools (M = 6.76, SD = 4.36) and adolescents in special schools (M = 2.80, SD = 2.49), p = .008; and between adolescents in general education schools and adolescents not in school (M = 3.72, SD = 3.61), p = .013. In functional academics, the same pattern of differences was observed: There was a difference between adolescents in general education schools (M = 2.60, SD = 2.93) and adolescents in special schools (M = 1.00, SD = .00), p = .042; and between adolescents in general education schools and adolescents not in school (M = 1.11, SD = .47), p = .023. And again, the same was observed for the Conceptual domain: There was a difference between adolescents in general education schools (M = 70.80, SD = 15.36) and adolescents in special schools (M = 59.80, SD = 6.37), p = .020; and between adolescents in general education schools and adolescents not in school (M = 62.00, SD = 9.39), p = .024. No significant differences by type of education were observed in IQ scores and adaptive behavior scores reported by teachers.
Analyses conducted when grouping by SES (i.e., low, middle, high) only showed significant differences in community use as reported by teachers, F(2,48) = 3.69, MSE = 7.48, p = .032. Post hoc analyses determined that the significant difference was between low (M = 1.38, SD = .89) and high SES (M = 3.49, SD = 3.67), p = .009.
Finally, as seen in Table 5, linear correlation analyses using Pearson r showed significant moderate to strong positive correlations between FSIQ and communication, functional academics, self-direction, self-care, and Conceptual domain, as reported by parents and teachers; plus, the Practical domain, and GAC as reported by parents. VIQ/VCI also showed moderate to strong positive correlations with communication, functional academics, self-direction, self-care, and Conceptual domain, as reported by parents and teachers; plus health and safety and GAC, as reported by parents. PIQ/PRI maintained moderate to strong positive correlation with communication, functional academics, self-direction, self-care, and Conceptual domain, as reported by parents and teachers; plus health and safety and GAC, as reported by parents.
The objective of this study was to explore intellectual functioning and adaptive behavior in adolescents with DS and to analyze how different sociodemographic variables impact these constructs. Specifically, we compared VIQ/VCI and PIQ/PRI within participants, as well as the three main domains of adaptive behavior, and the GAC reported separately by parents and teachers. We then compared the nine ABAS-II scales, along with the three main domains, and the GAC, between parents and teachers, to look for possible differences on how they assess adaptive functioning in the adolescents with DS. To evaluate possible sociodemographic differences in intellectual functioning and/or adaptive behavior, we grouped the participants by age group, schooling, type of education, and SES to conduct between-group analyses. Finally, we explored the correlation between intellectual functioning and adaptive behavior in our sample.
We found that participants obtained higher scores on verbal tasks (VIQ/VCI) as compared to performance tasks (PIQ/PRI) on tests of intelligence. This is a rather interesting finding since language has been consistently reported as an area of relative weakness for people with DS whereas visual problem-solving skills are considered a relative strength (Chapman & Hesketh, 2000; Grieco et al., 2015; Steingass et al., 2011). At this time, we do not have an accurate explanation for these findings, although hypotheses arise. One possibility is that these results are related to successful intervention programs. In our sample, 96.2% of the parents reported that their children received early intervention beginning before age 1, with an average start at 4.26 months of age (SD = 7.58) and a mean duration of 87.28 months (SD = 69.65). Unfortunately, we do not have precise information on the specific areas intervened or current treatments or intervention programs the adolescents might be receiving (e.g., speech-language therapy, occupational therapy). Hence, we cannot conclude that these interventions are responsible for better performance on the verbal tasks, although a possibility worth analyzing in the future. Another factor that should be taken into account is the Tsao and Kindelberger (2009) study that identified a cluster of children with DS that had better performance on verbal tasks. This finding opens the possibility of an alternative cognitive profile in DS that is different from what has previously been reported. Further research would be necessary to understand why this cluster seems to be overrepresented in our sample.
We found significant differences in intellectual functioning when separating by age group, with late adolescents having consistently higher FSIQ, VIQ/VCI, and PIQ/PRI compared to early and middle adolescents. Nagle and Lazarus (1979) previously reported that the WAIS yielded significantly higher IQ scores than the WISC-R on 16-year old adolescents with ID. Gordon, Duff, Davidson, and Whitaker (2010), had the same findings when comparing the WAIS-III and WISC-IV in 16-year old special education students. Thus, a possibility exists that in general, the WAIS produces higher IQ scores than the WISC in ID populations. This could explain why our 16- and 17-years old adolescents that were assessed with the WAIS-IV have higher IQs than the 12- to 15-years old participants that were tested with the WISC-III. This will mean that these results could be related to the test differences, rather than actual differences in their intellectual abilities. It is important to take into account that the WAIS and WISC versions from previous studies differ from the ones used in our study; thus, we cannot be certain that our results are completely explained by them. Moreover, as Gordon and colleagues (2010) explain, we currently do not know why this difference is observed in ID and not in the typically developing population.
The fact that we did not find IQ differences between schooling or type of education might mean that intellectual functioning was not a factor considered when school placement was decided for these adolescents. On the other hand, another possibility is that the educational experiences received by the adolescents are not significantly different in areas that could affect IQ. More research is needed to directly assess these hypotheses.
Another interesting result from our study is that we did not find a significant difference between IQ and SES. Previous research has consistently shown that poverty, SES, and SES related factors, such as maternal level of education, are related to the child's intellectual functioning and development (e.g., Bates, Lewis, & Weiss, 2013; Bradley & Corwyn, 2002; Ronfani et al., 2015). Previous research in Chile have found significant differences between IQ and SES in typical development population (Rosas & Santa Cruz, 2013) and differences in developmental trajectories and SES in DS (Arango, Aparicio, & Tenorio, 2018; in both studies, SES was defined using the type of funding received by the school, as in our study). We did not have an equal distribution of age groups by SES in our sample. If we consider our previously discussed results that the late adolescents have higher IQs associated with being tested with the WAIS-IV instead of the WISC-III, it is possible that this alters the mean IQ scores, inflating the results from the low SES group, and lowering the scores from the middle and high SES. Thus, we cannot draw any definitive conclusions regarding the impact of SES might have on IQ scores in adolescents with DS.
Our general findings in adaptive behavior are concordant with previous research (Chapman & Hesketh, 2000; Fidler et al., 2006): both parents and teachers reported stronger social skills than practical skills and communication.
Across age groups, parents reported higher self-direction skills in late adolescence compared to early adolescence. Teachers, on the other hand, did not report significant differences in adaptive skills across age groups. This discrepancy between parents and teachers could be related to the different environments and contextual expectations on these adolescents. Parents might be more prone to assist younger adolescents in their chores whereas teachers might expect their students to be able to start and finish assignments and follow instructions. This, in turn, will promote the use of different adaptive skills between the home and the school environment.
Parents of children and adolescents with special educational needs in Chile can consider either special education or integration into general education classrooms as school choices for their students. Special education schools follow a modified version of the general curriculum, adapted for the special educational needs of their students. Some special education schools are focused exclusively on specific conditions or types of disabilities (e.g., ASD, ID, deafness) while others receive children with different conditions. General education schools with integration programs offer diverse forms of accommodations in order to integrate the student with special educational needs into the general education classroom and curriculum. As per guidelines, special education schools are designed to educate students that require more supports than those enrolled in integration programs; however, there is great variability (Godoy, Meza, & Salazar, 2004). As stated previously, we found no differences between intellectual functioning and placement in different types of education, but we did find significant differences in adaptive behavior. Parents reported better communication, functional academics, and Conceptual skills in adolescents with DS that were enrolled in general education schools compared to the ones enrolled in special education or not in school. Because we believe intellectual functioning was not a factor, or the main factor, when considering placement of these students, these results might show that adaptive skills could have played a more significant role in the parent's decision regarding what type of school in which to enroll their child. This would be consistent with the notion that adaptive behavior is a stronger correlate to intensity of support needs than intellectual functioning (American Psychiatric Association, 2013; Schalock et al., 2010). Another possibility is that the type of school and its specific model of intervention has helped students further develop these abilities. If we consider the type of skills that showed significant differences in favor of students in general education schools, it is possible that the greater amount of contact with typically developing peers, along with integration into the general curriculum, promoted greater development. Further research is needed to evaluate the relationship between type of school and adaptive behavior, particularly to evaluate possible causal relationships.
Our results regarding correlations between intellectual functioning and adaptive behavior replicate previous findings: Most of the significant correlations found are in the moderate range, showing that these constructs are related but different, as stated by Keith and colleagues (1987). Furthermore, most of the correlations found were between the three IQ scores (FSIQ, VIQ/VCI, and PIQ/PRI) and the Conceptual domain and its scales, which replicated previous findings in similar populations with idiopathic ID (Carpentieri & Morgan, 1996). Due to the characteristics of those adaptive behavior skills, which are more centered on language, academic skills, and self-control, it seems reasonable for them to be related to intellectual functioning.
There are several limitations to our study. First, as a cross-sectional study, we cannot attribute any differences found to development without considering possible cohort differences. Second, the reported differences in IQ scores between the WISC-III and WAIS-IV imply that all of our results in intellectual functioning should be interpreted with caution. As reported in previous research (Gordon et al., 2010), we do not know what causes these differences and we do not know which score is more representative of the actual level of intellectual functioning. Thus, there is the possibility of confounding variables altering our results. Third, all of our adaptive behavior scores measured with the ABAS-II should be interpreted with caution since the norms used are Spanish and not Chilean. Fourth, we have uneven distributions across the groups tested, particularly when grouping by age and SES. As stated previously, and in relation to the observed differences in scores between early and middle adolescents (WISC-III) and late adolescents (WAIS-IV), we cannot make certain conclusions about the IQ distribution across SES.
Positive contributions of our study include the presentation of in-depth assessments of an adolescent population with DS, including both intellectual functioning and adaptive behavior. We also explored differences across groups and correlations between variables that have not been previously reported in this type of sample.
Future directions would include further explorations into the differences in IQ scores between assessments with the WISC and WAIS and to explore what causes the observed differences in adaptive behavior between adolescents with DS integrated into general education schools versus the ones enrolled in special education schools or not in school. Do better adaptive behavior abilities influence the decision of parents regarding where to enroll their children, or does the type of school assist or hinder adaptive behavior? Further research will also help in the understanding of the DS behavioral phenotype along with within-group variability.
Our study shows that intellectual functioning and adaptive behavior in DS follow the same pattern of relationships that was observed previously in the typically developing (TD) population (Keith et al., 1987) and ID population (Carpentieri & Morgan, 1996), with correlations in the medium range and most of them between IQ scores and the Conceptual area of adaptive behavior.
Interestingly, we did not observe the common cognitive phenotype described for DS because our sample had greater verbal abilities, extending the need for future research to explain this finding. Adaptive behavior skills did show the expected pattern, with stronger social skills than practical skills and communication.
These findings provide support for the exploration of the variability across the DS phenotype. Following Karmiloff-Smith and colleagues (2016) and Tsao and Kindelberger (2009), it seems clear that the commonly reported DS cognitive phenotype does not always match the profile of skills in this population. It is crucial to take this into account for future research and particularly when planning interventions with people with DS. Along with this, our results support the need to also consider both intellectual and adaptive skills at the moment of designing and implementing interventions for these individuals. Just as with TD population, these two constructs are related in DS and might influence each other. Further research is necessary to explore causal relationships that can be applied in practice.
This work was supported by the Fondo Nacional de Desarrollo Científico y Tecnológico (FONDECYT), grant number: 11150800.
The authors would like to express their appreciation for the families and institutions that participated in this research. We would also like to give a special thanks to Javiera Donoso, Consuelo Reyes, Diego Urzúa, Sandra Canales, and the students that worked during the interventions.