**11. Diagnostic substitution and the influence of society and culture**

When symptoms of two or more conditions are shared, whatever the psychological mecha‐ nisms (whether or not there are shared underlying cognitive deficits, and /or genetic and neurological differences) then the area of functioning that is highlighted as a problem may depend on which tests are administered. In our recent research we followed a six year old child who was assessed by three educational psychologists and one multidisciplinary team, each blind to the findings of the others. One concluded that the child had dyspraxia, two that the child had dyslexic difficulties, and a third that borderline AS was likely. We inter‐ preted these differences in the use of diagnostic labels as dependent on settings that varied during assessments, and assessment methods that exposed different types of behaviour (Russell, Norwich, & Gwernan-Jones, 2012). This work suggests that which diagnosis is as‐ signed depends to some extent on social and cultural factors as well as actual symptoms. If a child has symptoms of several disorders, then one context or test may draw out symptoms associated with one disorder, whereas another setting may expose symptoms of another. Thus for co-occurring symptoms it is difficult to differentiate between disorders and the likelihood that a co-morbid disorder will be missed is increased. This emphasizes the need for assessment in multiple settings and reassessment over time.

**1980 Diagnosis (ICD 9 codes) N of children Percentage of total examined %**

Total 259 1.961

**Table 1.** Named conditions using ICD-9 categories for 10 year old children in 1980 (n=13201).

25 (1 ADHD co-morbid)

Co-Occurrence of Developmental Disorders: Children Who Share Symptoms of Autism, Dyslexia and Attention Deficit

Among the 14,043 children in the 2007 cohort, 209 (1.49%) were reported to have ASD, and 180 (1.28%) were reported having been given an ADHD diagnosis by a clinician (unweight‐ ed figures). There was disproportional stratification in the Millennium Cohort, meaning that all analyses were weighted to account for the clustering and over-inclusion of participants from disadvantaged areas. After weighting, 1.7 % of children were reported as having an ASD (95% CI, 1.4-1.99). 1.3% of these were boys, and 0.25% girls, giving boy girl ratio of ap‐ prox 5:1 for ASD. Surprisingly, the figure for ADHD was lower. After weighting, 1.4% of the population were reported as having ADHD (95% CI, 1.2-1.7). Of these, 2.3% were boys and 0.25% girls, giving a gender ratio of approximately of 1 girl to every 4 boys with ADHD.

One interpretation of the historical shift is that diagnostic substitution has occurred: children with similar symptoms in 1980 may have been more likely to receive generalised labels of 'delays in learning & development' than ASD or ADHD. So changing diagnostic practice, cultural factors and context may do much to explain both co-morbidity and rising preva‐ lence. The steep rise in children assigned these diagnoses cannot be totally explained by the substitution mechanism- twice as many children were given either ASD or ADHD diagno‐ ses in 2009 as the total number diagnosed with any type of developmental disorder in 1980.

Context also has a big part to play in the identification of difficulties, in terms of what is con‐ sidered to be 'disordered'. Social constructionists have also pointed out that the conceptuali‐ zation of difficulties associated with both dyslexia and ASD as 'disorders' is itself a product of social and cultural standards, and of course the definition of each disorder has changed over time. This has prompted calls for the term autism spectrum 'conditions' to replace au‐ tism spectrum 'disorders' (2009). Our own analysis of the Millennium Cohort has shown a strong association between ADHD and poverty, reflecting findings from US studies which have also found differing levels of ADHD amongst various ethnic groups- Hispanic children were more likely to be identified with ADHD in a study by Akinbami et al. (2011). It is un‐ clear whether this is entirely due to greater awareness and access to health care in some groups, differential reporting about the same level of difficulties between ethnic groups or whether children in different groups have truly varying symptom levels (Boyle et al., 2011). A study by Cuccaro et al. (1996) showed the nature of diagnosis of developmental disorders varied according to the socio-economic status of the child's family; autism was more likely to be identified in children of higher income families, although no biases of SES were found for identification with ADHD. Cooper (2001) points out that the behaviour symptomatic of ADHD becomes problematic where high value is placed on ability to remain sedentary and sustain attention on tasks, in other words, in schools. Hulme and Snowling (2009) describe how differences of this nature must therefore be thought of as *both* biological *and* as a prod‐

0.379

Hyperactivity Disorder

377

http://dx.doi.org/10.5772/54159

Unspecified delays in development

uct of the social and environmental world.

(319)

One of the most compelling cross cultural descriptions of how autism is regarded across various cultures was the book *Unstrange Minds.* Written by the anthropologist Roy Grinker (2008), Grinker explains how the category of ASD is contingent on the culture through which it is expressed- the condition is associated with differing levels of stigma in different cultures. In the US, several studies have also shown that clinicians may diagnose ASD when resources are targeted at the diagnosis, whereas previously, under other circumstances, they may have diagnosed another category of childhood disorder. Paul Shattuck has written about the extent to which increases in the administrative prevalence of autism have been as‐ sociated with corresponding decreases in the use of other diagnostic categories, mental re‐ tardation and learning disabilities (2006). This process of 'diagnostic substitution' he argues, may partially explain the rise in prevalence in autism in the US.

Our own work suggests that since the 1980s, the recorded prevalence of both ASD and ADHD in the UK has increased dramatically. We examined data from both the Millennium Cohort Study, (the large cohort of around 19,000 children who have been followed from their birth through to seven years old and beyond), and another cohort, called the British Cohort Study, where children were born thirty years previously. Both cohorts were repre‐ sentative of the UK as a whole, and medical reports of both ASD and ADHD were given when children were age seven for in 2007-9 and ten in 1980. The results from 2007 contrast‐ ed with the 1980 sample at age 10. Only 11 children in the 1970 British Cohort Study were reported as having ADHD in their medical exam, giving an estimated prevalence of 0.083%. The autism diagnosis was rarely used with just 3 children assigned the label; 0.023% of chil‐ dren. A number of other child psychiatric diagnoses were available and many of these were diagnosed during the medical exams. Details of these alternative labels are given in Table 1.



**Table 1.** Named conditions using ICD-9 categories for 10 year old children in 1980 (n=13201).

Thus for co-occurring symptoms it is difficult to differentiate between disorders and the likelihood that a co-morbid disorder will be missed is increased. This emphasizes the need

One of the most compelling cross cultural descriptions of how autism is regarded across various cultures was the book *Unstrange Minds.* Written by the anthropologist Roy Grinker (2008), Grinker explains how the category of ASD is contingent on the culture through which it is expressed- the condition is associated with differing levels of stigma in different cultures. In the US, several studies have also shown that clinicians may diagnose ASD when resources are targeted at the diagnosis, whereas previously, under other circumstances, they may have diagnosed another category of childhood disorder. Paul Shattuck has written about the extent to which increases in the administrative prevalence of autism have been as‐ sociated with corresponding decreases in the use of other diagnostic categories, mental re‐ tardation and learning disabilities (2006). This process of 'diagnostic substitution' he argues,

Our own work suggests that since the 1980s, the recorded prevalence of both ASD and ADHD in the UK has increased dramatically. We examined data from both the Millennium Cohort Study, (the large cohort of around 19,000 children who have been followed from their birth through to seven years old and beyond), and another cohort, called the British Cohort Study, where children were born thirty years previously. Both cohorts were repre‐ sentative of the UK as a whole, and medical reports of both ASD and ADHD were given when children were age seven for in 2007-9 and ten in 1980. The results from 2007 contrast‐ ed with the 1980 sample at age 10. Only 11 children in the 1970 British Cohort Study were reported as having ADHD in their medical exam, giving an estimated prevalence of 0.083%. The autism diagnosis was rarely used with just 3 children assigned the label; 0.023% of chil‐ dren. A number of other child psychiatric diagnoses were available and many of these were diagnosed during the medical exams. Details of these alternative labels are given in Table 1.

**1980 Diagnosis (ICD 9 codes) N of children Percentage of total examined %**

62 (1 autism co-morbid)

34 (1 ADHD co-morbid)

22 (1 ADHD co-morbid)

13 0.098

81 0.614

0.462

0.258

0.166

Autism (299.0/1/8/9) 3 0.023 ADHD (314.00/01, 314.9) 11 0.083 Disturbance in emotions (313) 7 0.053

Impulse control (312.3/9) 1 0.007

Delays in development: Reading

Delays in learning & development

(315.0)

(315.3)

(317)

(315.2/8/9/5)

Delays in language

Mild mental retardation

Other specified delays in development (318)

for assessment in multiple settings and reassessment over time.

376 Recent Advances in Autism Spectrum Disorders - Volume I

may partially explain the rise in prevalence in autism in the US.

Among the 14,043 children in the 2007 cohort, 209 (1.49%) were reported to have ASD, and 180 (1.28%) were reported having been given an ADHD diagnosis by a clinician (unweight‐ ed figures). There was disproportional stratification in the Millennium Cohort, meaning that all analyses were weighted to account for the clustering and over-inclusion of participants from disadvantaged areas. After weighting, 1.7 % of children were reported as having an ASD (95% CI, 1.4-1.99). 1.3% of these were boys, and 0.25% girls, giving boy girl ratio of ap‐ prox 5:1 for ASD. Surprisingly, the figure for ADHD was lower. After weighting, 1.4% of the population were reported as having ADHD (95% CI, 1.2-1.7). Of these, 2.3% were boys and 0.25% girls, giving a gender ratio of approximately of 1 girl to every 4 boys with ADHD.

One interpretation of the historical shift is that diagnostic substitution has occurred: children with similar symptoms in 1980 may have been more likely to receive generalised labels of 'delays in learning & development' than ASD or ADHD. So changing diagnostic practice, cultural factors and context may do much to explain both co-morbidity and rising preva‐ lence. The steep rise in children assigned these diagnoses cannot be totally explained by the substitution mechanism- twice as many children were given either ASD or ADHD diagno‐ ses in 2009 as the total number diagnosed with any type of developmental disorder in 1980.

Context also has a big part to play in the identification of difficulties, in terms of what is con‐ sidered to be 'disordered'. Social constructionists have also pointed out that the conceptuali‐ zation of difficulties associated with both dyslexia and ASD as 'disorders' is itself a product of social and cultural standards, and of course the definition of each disorder has changed over time. This has prompted calls for the term autism spectrum 'conditions' to replace au‐ tism spectrum 'disorders' (2009). Our own analysis of the Millennium Cohort has shown a strong association between ADHD and poverty, reflecting findings from US studies which have also found differing levels of ADHD amongst various ethnic groups- Hispanic children were more likely to be identified with ADHD in a study by Akinbami et al. (2011). It is un‐ clear whether this is entirely due to greater awareness and access to health care in some groups, differential reporting about the same level of difficulties between ethnic groups or whether children in different groups have truly varying symptom levels (Boyle et al., 2011). A study by Cuccaro et al. (1996) showed the nature of diagnosis of developmental disorders varied according to the socio-economic status of the child's family; autism was more likely to be identified in children of higher income families, although no biases of SES were found for identification with ADHD. Cooper (2001) points out that the behaviour symptomatic of ADHD becomes problematic where high value is placed on ability to remain sedentary and sustain attention on tasks, in other words, in schools. Hulme and Snowling (2009) describe how differences of this nature must therefore be thought of as *both* biological *and* as a prod‐ uct of the social and environmental world.

### **12. Conclusion**

Two conclusions can be drawn. First, co-morbidities between developmental disorders are common, and second, the causes of these overlapping difficulties are likely to be complex, multifactorial and interacting. Firstly, the high overlap between symptoms of different de‐ velopmental disorders has been identified in a number of studies and there is an interna‐ tional consensus on this overlap. Studies from Canada, the UK, USA and Scandinavia all show how hard it is provide an unequivocal diagnosis, leading to the quote from Kaplan and her colleagues (2001) *in developmental disorders co-morbidity is the rule, not the exception.* This was informed by the group's work studying a population-based sample of 179 children receiving special support in Calgary: If the children met the dyslexia criteria, there was a 51.6% chance of having another disorder. If the children met the ADHD criteria there was an 80.4% chance of having another disorder. They criticize the term 'comorbidity', as it implies unsubstantiated presumption of independent aetiologies. The authors argue that discrete categories do not exist in real life.

tion, rather than classing children into dichotomous 'disorder' categories. Taylor (2011) notes that for many children, it is better to think of changes in cognitive style, learning and motivation rather than symptoms. Both conclude that it is important to examine children for evidence of co-occurring disorders, and not simply continue to examine the areas which we expect to be impaired according to categorization. The practical application of assessing chil‐ dren for a range of difficulties is that children will be best helped not by any all encompass‐ ing diagnosis, but by individual analysis of their strengths and weaknesses. Future research may be wise to focus on the individual profiles of children across a broad range of areas, looking at the unique strengths, as well as the weaknesses of the individual children, so that parents and educators may adapt their support accordingly, regardless of the diagnostic la‐

Co-Occurrence of Developmental Disorders: Children Who Share Symptoms of Autism, Dyslexia and Attention Deficit

Hyperactivity Disorder

379

http://dx.doi.org/10.5772/54159

bel a child receives.

**Author details**

2 University of Milan, Italy

Ginny Russell1

**Figure 3.** Schematic of interacting causal mechanisms for co-morbidity.

and Zsuzsa Pavelka2

1 University of Exeter Medical School, ESRC Centre for Genomics in Society, UK

Secondly, in considering the reasons for co-morbidities, a complex bio-psycho-social model is required that leads to symptoms that may result in diagnosis. The nature of the diagnosis itself may depend on social context as well as an individual child's behaviour. A hint of this complexity is achieved in Figure 3, which is a schematic diagram of various potential causal pathways. It is plausible that the same underlying genetic or neurological mechanisms may underlie co-occurrence of dyslexia, ADHD and ASD. The reverse pathways are not at first so obvious. But recent advances in systems biology have shown that the environment of the cell affects gene expression and protein synthesis at molecular levels. Thus environmental influences can alter 'core' biology: for example Mack and Mack (1992) describe how tweak‐ ing rats' whiskers changes gene expression in the sensory cortex. In systems theory, genetic influences are conceptualised more like a set of piano keys on which notes may be played or not played, played slowly or quickly, and there is enormous variation in the music pro‐ duced even with the same basic set of keys. So the cellular environment can affect genetic expression. A simplified model underlying much behaviour genetics research envisages a direct linear relationship between individual genes and behaviours. The reality is likely to be far more complex with gene networks and multiple environmental factors impacting brain development and function, which in turn will influence behaviour (Hamer, 2002). Kar‐ miloff-Smith (2007) emphasizes how learning and experience effects gene expression in hu‐ mans. Such scholars demonstrate that the social can affect the biological as well as the more intuitive path of genetic origin leading to neurological development leading to aberrant be‐ haviour. Diagnosis itself may influence behaviour too, through differential treatment and in‐ terventions. Thus the pervasive developmental environment is composed of many related factors, environmental stresses, and genetic predispositions, and the social contexts all of which may interact to produce developmental outcomes that themselves may contribute to predicting ongoing child development.

Snowling (2012) suggests a new dimensional classification of disorder, where deficits in dif‐ ferent components of learning are seen as additive, impacting on the potential for remedia‐ tion, rather than classing children into dichotomous 'disorder' categories. Taylor (2011) notes that for many children, it is better to think of changes in cognitive style, learning and motivation rather than symptoms. Both conclude that it is important to examine children for evidence of co-occurring disorders, and not simply continue to examine the areas which we expect to be impaired according to categorization. The practical application of assessing chil‐ dren for a range of difficulties is that children will be best helped not by any all encompass‐ ing diagnosis, but by individual analysis of their strengths and weaknesses. Future research may be wise to focus on the individual profiles of children across a broad range of areas, looking at the unique strengths, as well as the weaknesses of the individual children, so that parents and educators may adapt their support accordingly, regardless of the diagnostic la‐ bel a child receives.

**Figure 3.** Schematic of interacting causal mechanisms for co-morbidity.
