Which factors may explain why some children show unstable IQ scores over the course of their childhood?

Intelligence

The intelligence quotient (IQ) is often reported in studies of SDB although findings are not consistent. Lower IQ scores have been reported in children with SDB compared to controls, although these scores are typically still within the normal range.19,21,49,50,55,59,92–96 One study in children awaiting adenotonsillectomy found that compared to healthy non-snoring children, the snoring children had a 10-point reduction in IQ.95 Of course, the clinical significance of this remains to be shown for a high-functioning child, but a 10-point IQ difference could be rather significant in children performing at a lower level. Several studies fail to support findings of differences in full-scale IQ,44,91,97,98 although some have found lower scores for verbal IQ (language skills) in children with SDB.44,97

Lack of robust findings between SDB and IQ is perhaps not surprising given that measurement of IQ is complex and, in essence, measures performance across several tasks rather than a focus on a particular area of cognition. Standardized vocabulary tests, as a proxy measure of IQ and an excellent predictor of cognition and academic success,99 have demonstrated that the difference in scores between children with and without SDB may be equivalent to the impact of lead exposure.100 These findings clearly have great clinical significance for a child's future if indeed they are supported by additional studies. Nonetheless, there are many other factors which clearly impact a child's IQ and which require consideration in studies of SDB, including genetics, parental education level, as well as biological and environmental factors.23,101,102

The intelligence quotient

Intelligence quotient (IQ) refers to mental age (MA) expressed as a ratio of chronological age (CA) multiplied by 100. For IQ to remain stable, MA must increase with CA over time. This is true until around 18 years, when intellectual abilities are usually fully developed. As stated above, it was erroneously thought that intellectual ability started to decline after young adulthood. It now appears that although fluid intelligence may start to decrease, crystallized intelligence continues to develop throughout the lifespan. Intelligence is said to be normally distributed with a slight ‘bump’ at the lower end of the normal distribution curve, representing those with severe learning difficulties.

Epidemiology

Using an IQ of less than 70 (2 standard deviations below the mean) as the criterion for learning disability or mental retardation should result in prevalence rates of 2.3% based on the properties of the normal distribution. Studies of mild mental retardation have produced widely varying prevalence estimates, however, from less than 0.5% to over 8%. There are at least several reasons for this variation. First, identification, or ascertainment, of the ‘at risk’ population affects the rate, with studies that survey the entire population producing higher rates than those relying on cases formally identified because of health or educational problems, as many people with IQs between 50 and 70 may not be formally registered. Second, the test administered may affect the rates identified. Because performance on IQ tests has improved over the years (a phenomenon called the Flynn effect), the use of older and ‘easier’ tests may produce a lower rate of learning disability. There may also be real differences in the rate of mental retardation across different populations, with more advantaged populations having lower rates of learning disability. Some of this population difference may be due to bias in test content, so that children from developed countries who attend school from an early age may be more familiar with the type of items on conventional IQ tests. However, use of the supposedly culture-fair test does not eliminate the differences in rates of learning disabilities.

For severe mental retardation, however, the reported prevalence rates have been more consistent and average around 0.4–0.5%, which is about 10 times greater than expected where the normal distribution maintained. This extra ‘hump’ at the bottom of the normal IQ distribution is likely to represent the children whose learning disabilities have a clear ‘organic’ origin whether caused prenatally, perinatally, or postnatally.

These findings have led to a suggestion of a ‘two-group’ approach to learning disability. The first group represents the lower end of a normally distributed population, for which no organic cause can be ascertained and encompasses the majority of children with mild learning disability. Environmental deprivation has classically been cited as the cause of cognitive delay in this group, although this is now criticized for being too simplistic. The second group consists of those children with defined organic or biological cause for their cognitive impairment, which may be more severe. Again, this is likely to be an oversimplification of the heterogeneity of learning disability, and as advances are made in molecular genetics, increasing numbers of children with mild learning disability are likely to be found to have genetic disorders. However, the division of learning disability into these two broad etiological groups has been a useful starting point for researchers in this field.

There is also criticism of the use of IQ as sole discriminator of learning disability for both epidemiological studies and administrative purposes. In terms of judging prevalence rates based on IQ, problems arise due to the fact that IQ may not remain stable throughout development and IQ scores in infancy typically have only low correlations with scores in later childhood. Further, researchers found that although once formal schooling begins IQ scores tend to be more robust, individual cases can still show major gains or losses in cognitive ability. Up to now, however, there has been no agreement on the way in which adaptive behavior should be measured and impairment defined to meet this additional criterion for diagnosis.

The term ‘administrative prevalence’ has been used to mean the numbers for whom services would be required in a community which made provision for all who needed them. Here again, using IQ alone as discriminator is inadequate as IQ scores do not correlate with social adaptation in all instances. Indeed, researchers found that although 2.5% of children in the study could be classified as having learning disability based on their IQ scores, only half of these children were sufficiently impaired in their daily lives to require the additional provision of services. Functional impairment is dependent not only on the child’s developmental level, but on wider social issues of available family and community resources. Thus, it is unsurprising that administrative prevalence is somewhat higher in lower socioeconomic groups.

Gender differences are also apparent in the rates of learning disability with a male to female ratio of 1.5:1, which may reflect the male preponderance in certain genetically mediated disorders affecting cognitive ability. These differences are more clear-cut for those with severe learning disability.

Measures of Intellectual Ability

Currently, the IQ score is based on the standard statistical distribution of scores around a mean of 100 as described earlier in the section on types of scores. One implication of clinical importance is that at the statistical scoring extremes of a population, a few points of change can greatly affect school placement decisions, whereas a few points of change around the population average can be dismissed as random error. A new treatment that shifts the mean of the affected population from 94 to 100 substantially reduces the proportion of affected children with cognitive impairment.

One advance in the development of intelligence scales has been a greater attempt to organize test items into groupings that can remain consistent throughout the age span. This improvement, developed by Wechsler, may be responsible for the continued dominance of his series of tests including the Wechsler Adult Intelligence Scale (WAIS), which preceded the development of the comparable scales for children, now the Wechsler Intelligence Scale for Children–IV (WISC-IV) and the Wechsler Preschool and Primary Scale of Intelligence–Revised (WPPSI-R). Another important trend in intelligence test construction is the notion of subtypes of intelligence. In contrast to the original Stanford Binet, which yielded one global IQ score, measures that divide performance into at least two domains of verbal and nonverbal or performance abilities are now favored. These subtests of intelligence tests provide insight into the strengths and weaknesses of a child.

Sometimes intelligence tests are supplemented by additional measures of special ability, such as Visual Motor Integration, Wide Range Assessment of Memory and Learning, measures of information processing such as the Detroit Tests of Learning Aptitude–3, or language tests, to describe a child's skills or deficits further. A detailed review of various cognitive tests and approaches to administration and interpretation may be found in Sattler's authoritative text (Sattler, 2001) (see also Chapter 81).

Some limitations of attempts to divide intelligence into multiple domains may be related to measurement limitations, rather than the actual nature of individual abilities. When subscales contain few items, they are less likely to be reliable. They also are more likely to be in error. Some purported subtests do not prove to be statistically distinct. Using data from a variety of sources, however, Gardner (1983) has shown evidence for distinct “multiple intelligences.” The different abilities include logical-mathematical, spatial, bodily kinesthetic, musical, interpersonal, intrapersonal, and naturalist. Intelligence tests provide only a limited view of the abilities of individuals.

IQ test scores may be invalid if the child has a sensory or motor impairment that limits his or her ability to understand instructions and other such factors can adversely affect IQ test results. The diagnosis of intellectual disability or cognitive impairment requires not only low IQ test scores, but also comparable scores on a test of adaptive behavior (see Chapter 68). Some tests of adaptive behavior, such as the Vineland, generate standard scores, facilitating comparisons with IQ test scores.

Clinical features

Most patients have an IQ between 50 and 75, and often live at home. However, patients with severe learning impairment (IQ below 50) are often totally dependent on others. Brain damage may cause not only mental but also physical impairment and epilepsy; visual defects, hearing, speech or behavioural disorders, facial deformities or cardiac defects are often associated. The limitations cause children to learn and develop more slowly than normal children. They typically take longer to learn to speak, walk and take care of their personal needs, such as dressing or eating, and often have trouble learning. They may sit up, crawl, walk or talk late, or have difficulty speaking. They may find it hard to remember things, and have trouble solving problems or thinking logically. They may also find it difficult to understand social rules and how to pay for things and can have trouble seeing the consequences of their actions.

Other problems may include psychiatric disorders (symptoms are often modified by poor language development and other defects), hyperkinesis and stereotyped movements (body-rocking and self-mutilation are common; Fig. 28.7). Feeding difficulties may be present. Pica (the ingestion of inedible substances) is also fairly common.

6.1 Summary of the Results for the Client (Bryan, Teacher, School, and Parents)

Bryan has an average IQ, his perceptual organization is more than 1.5 SD below average, his attention and concentration are more than 1 SD below average, his relationship with the teacher is very tense, and Bryan avoids going to school. His self-esteem and achievement motivation are average.

Advice for treatment was concrete, and its implementation was feasible. It was considered that it would be better if Bryan could change class for the next year. The relationship with the teacher had to be improved. A daily planning of tasks was made and agreed on by both Bryan and the teacher, and the tasks were divided into small units. Instructions were short and clear, a time limit was given for each task, and he could ask questions at fixed times. The teacher was informed about the assessment results.

People with learning disabilities and the criminal justice system

Intellectual impairment (i.e. low IQ, but not necessarily learning disabilities) is a risk factor for offending in both adults and juveniles. There is little evidence to suggest that the presence of learning disability predisposes to criminal behaviour. There are particular difficulties looking at the numbers convicted; it is generally considered that the prevalence rates for offenders with learning disabilities may be higher than the general population, especially among those convicted of arson or sexual offences. There is also a predominance of males among offenders with learning disability.

People receiving specialist learning disability services are unlikely to offend, partly because the offending behaviour is re-labelled as ‘challenging behaviour’ and therefore decriminalized. However, in the UK, people with learning disabilities are over-represented among those arrested and taken to police stations, especially those with mild disability. The Police and Criminal Evidence Act 1984 (PACE) seeks to protect vulnerable people during the interview process, and requires the presence of an appropriate adult.

People with learning disabilities are also victims of crime. The fact that they have learning disabilities may prejudice their chances of being judged to be a reliable witness. Prosecutions are therefore not very common. The Bradley Report (DH 2009b) emphasizes that all parts of the process must be more aware and sensitive to the special needs of this vulnerable population.

Cognition and Executive Functioning

Children exposed prenatally to alcohol typically have intelligence quotient (IQ) scores in the low average to borderline range. Children with more dysmorphic features tend to have lower IQ scores, but cognitive problems are not limited to this group.27 Of those with FAS, around 25% to 50% will experience severe cognitive delay (IQ score <70), with the pooled prevalence of cognitive disability being 97 times higher in infants with FAS than the general population.28,29 In addition to global cognitive impairment, deficits in executive function and memory are also evident on neuropsychological testing. Executive deficits include problems with planning and organization, cognitive flexibility/set shifting, working memory, and behavioral inhibition, with parents reporting the greatest difficulty with inhibitory control and problem solving.30-33 In terms of memory deficits, problems encoding or memorizing information appear more prominent than problems with recall.34 Not surprisingly, given this constellation of cognitive impairments, learning problems are very common at school, even after controlling for IQ.35,36

The Case of Emotional Quotient Tests

Another question raised with respect to IQ tests relates to their inadequacy to measure success in real-life situations; assessing this has contributed to the creation of new knowledge. When psychologists began to write and think about intelligence, they focused on cognitive aspects, such as memory and problem solving. However, there were researchers who recognized early on that the noncognitive aspects were also important. As early as 1943, Wechsler (1896–1981) was proposing that the nonintellective abilities are essential for predicting a person's ability to succeed in life. Robert Thorndike (1943–), to take another example, was writing about “social intelligence” in the late 1930s. The work of these early pioneers was largely forgotten or overlooked until 1983, when Howard Gardner began to write about “multiple intelligence.” Gardner proposed that “intrapersonal” and “interpersonal” intelligences are as important as the type of intelligence typically measured by IQ and related tests. In 1990, Peter Salovey and John Mayer coined the term “emotional intelligence,” defining it as “a form of social intelligence that involves the ability to monitor one's own and others' feelings and emotions, to discriminate among them, and to use this information to guide one's thinking and action.” Salovey and Mayer also initiated a research program intended to develop valid measures of emotional intelligence and to explore its significance. In 1995, Daniel Goleman became aware of Salovey and Mayer's work, and this eventually led to his book, Emotional Intelligence. To be a scientist, probably an IQ of 120 or so is needed simply to get a doctorate and a job. But it is more important to be able to persist in the face of difficulty and to get along well with colleagues and subordinates than it is to have an extra 10 or 15 points of IQ. The same is true in many other occupations. In doing the research for his first book, Goleman became familiar with a wealth of research pointing to the importance of social and emotional abilities for personal success. Some of this research came from personality and social psychology, and some came from the burgeoning field of neuropsychology. Research now emerging suggests that emotional intelligence, and particularly the new measures that have been developed to assess it, is in fact a distinct entity.