Treating reading disability without reading: evaluating alternative intervention approaches

1
Treating reading disability without reading
evaluating alternative intervention approaches

• Dorothy Bishop
• University of Oxford
• See Notes view (Edit menu, Edit slides) for text
  for this presentation

2
Conventional approaches to dyslexia

• Most children have problems with phonological
  awareness
• Interventions focus on training children to
  identify sounds in words and relate these to
  letters

3
Problems with conventional approach

• Intensive and prolonged phonological intervention
  can be effective in improving reading accuracy
• Reading fluency remains a problem
• Methods that are effective for most children,
  dont work for all A hard core of children
  remain very hard to treat

4
Neuroscience studies of developmental disorders

• The Holy Grail
• Develop a theory of the disorder that
• not only explains why it occurs, but also
  motivates effective intervention

5
Goals of this talk

• Identify some questions for parents considering a
  new treatments
• Illustrate with Dore method and fish oil

6
How to measure reading a brief digression
7
How to measure reading etc.
A 9-year-old reads 20 words
Reading Age is 8 years Sounds bad 12 months
behind age level
8
Reading age misleading does not take into
account spread of scores at a given age
non-linear relation with chronological age
A 9-year-old reads 20 words. Error bars show
middle 50 of children
Within normal range for children of this age
9
Better to measure reading in terms of statistical
abnormality at that age
Beware of studies that rely on reading age
Percentile of children of this age obtaining
this score or lower Also z-score or standard
score different way of expressing same idea
10
Some questions to ask of a new treatment

• Is the theory scientifically plausible?
• Does evidence for efficacy go beyond
  testimonials? Have studies been done with
  groups for whom treatment is recommended?
• Is there evidence that gains are due to treatment
  rather than maturation, practice, placebo, etc.?
• Are costs reasonable relative to benefits?

11
Dore method what is it?

• Method for curing cerebellar problems developed
  by Wynford Dore to help his dyslexic daughter
• Individualised program of exercises, done for
  around 10 mins, 2 x per day, assessed every 6
  weeks

12
Dore method the exercises

• Hundreds of exercises, e.g.
• standing on a cushion on one leg and throwing a
  beanbag from one hand to another for one minute
• hopping on one leg in large circle, clockwise
  then anticlockwise
• sitting upright in a chair, turning head from
  side to side, pausing to focus on chosen point
• balancing on a wobble board
• (Examples only full details confidential because
  commercially sensitive)

13
Dore method the theory

• Dyslexia and other learning difficulties arise
  when the cerebellum fails to develop normally
• Cerebellar impairments differ from person to
  person but can be diagnosed by specific tests of
  mental and physical co-ordination

14
Cerebellar theory of dyslexia

• Not proven, but some evidence for it
• Brain imaging and neuroanatomical studies offer
  some support
• Theory that dyslexia involves failure to
  automatise skills is plausible
• Associated deficits in motor co-ordination in a
  subset of people with dyslexia

15
Previous research on effectiveness of motor
training

• Training can improve performance on motor tasks,
  e.g. juggling
• In rats, exercise can reverse cerebellar deficits
  caused by prenatal alcohol or zero gravity
• But no evidence that motor training enhances
  development of non-motor skills

16
The Theory evaluation

• Notion that training motor skills will have
  effect on other skills
• This hypothesis required something of a leap of
  faith, in that it is generally believed that the
  cerebellum comprises a very large number of
  independent cerebro-cortical microzones, and so
  it is not clear why training on one sort of task
  should generalize to unrelated tasks
• (Reynolds et al, 2003, p 53)

17
The Theory evaluation

• If training focusing on one region of cerebellum
  had general effects on all cerebellar functions,
  then
• activities like juggling and skateboarding should
  protect against dyslexia
• sportsmen and women should have low risk of
  dyslexia

18
Questions

• 1. Is the theory scientifically plausible?
• Notion that cerebellum may be implicated in
  dyslexia is plausible though not proven
• Notion that motor exercises will have beneficial
  effect on regions of cerebellum concerned with
  learning is considerably less plausible

19
Does evidence for efficacy goes beyond
testimonials?

• One published study on Dore intervention
• Two papers in Dyslexia reporting different phases

20
Have studies been done with groups for whom
treatment is recommended?

• 2003 study 296 children from 3 school yrs
• Selected 35 at risk on basis of Dyslexia
  Screening Test strong risk in 34, mild in 21,
  remainder fall below at risk level
• Divided randomly into untreated and treated
  groups
• Previous diagnoses
• treated 4 dyslexic, 1 dyspraxic
• control 2 dyslexic, 1 dyspraxic, 1 ADHD

21
Results as reported by Dore organisation

• Dore (2006) results were stunning and
• reading age, increased 3 x
• comprehension age increased almost 5 x
• writing, increased by an extraordinary 17 x

22
Data from school-administered tests, treated
group only
improvement calculated by dividing orange line
by pink line, i.e. change from time 2 to 3,
divided by change from time 1 to 2 Conclude
reading age increased 3 times But misleading
depends on low score at time 2 Why use reading
age when test has scaled scores?
23
Data from SATS (treated children only)

• Designed for assessment of attainment rather
  than psychometric rigour (Reynolds Nicolson,
  2007)
• Level 2 average for typical 7 yr old
• Level 3 average for typical 9 yr old
• Level 4 average for typical 11 yr old
• One should not over-interpret these
  data(Reynolds Nicolson, 2007)

24
Q3. Is there evidence that gains are due to
treatment?
25
Uninteresting reasons why scores may improve - 1

• Maturation
• Children change with age
• Shoe size may go up after treatment, but does not
  mean that treatment made feet bigger
• Not an issue if age-adjusted scores used but
  problematic if reliant on reading age or tests
  with no age norms (e.g. balance tests)

26
Uninteresting reasons why scores may improve - 2

• Placebo effect / effect of other intervention
• Child may be having other help or may respond to
  increased attention

27
Uninteresting reasons why scores may improve - 3

• Practice effects
• Child does test better 2nd time around because
  they have done it before
• Numerous examples in research literature e.g.
  Dyslexia Screening Test manual recommends that
  semantic fluency subtest is not valid if given
  twice because children tend to practice once they
  have done the test

28
Uninteresting reasons why scores may improve - 4

• Regression to the mean
• Statistical artefact whereby someone selected for
  extreme score at time 1 will on average have less
  extreme score at time 2
• Regression to the mean is as inevitable as death
  and taxes
• Campbell Kenny (1999) A primer on regression
  artefacts

29
Regression to the mean
Correlation between time 1 and time 2 .06
30
Regression to the mean
Correlation between time 1 and time 2 .99
31
Regression to the mean
Correlation between time 1 and time 2 .76
Social scientists incorrectly estimate the
effects of ameliorative interventions.....and
snake-oil peddlers earn a healthy living all
because our intuition fails when trying to
comprehend regression toward the mean (Campbell
Kenny, 1999)
32
These unwanted sources of change can be
identified if we have a CONTROL GROUP

• Untreated matched group given same pre- and
  post-test will control for
• Maturation
• Effects of other intervention
• Practice effects
• Regression to the mean

33
Alternative treatment control group

• Crucial to see if improvement due to
• Placebo/expectation effects
• Child, parent, teachers all expect and want to
  see gains
• Child gets more attention, boosted confidence,
  etc.

34
Dore study did include untreated control group
35
Results total on dyslexia screening

• High score indicates more risk
• NB score include bead-threading/posture
• Treated mean fell from 0.74 to 0.34
• strong risk fell from 33 to 11
• Control mean fell from 0.72 to 0.44
• strong risk fell from 35 to 24
• Everyone improves, even if not treated

36
Significant group differences in gain on bead
threading, semantic fluency and reading
37
Control group subsequently given the treatment

• Results published in Dyslexia journal in 2007
• Control group now known as group D (delayed
  intervention), and compared with original
  intervention group (I)

38
Results on dyslexia screening test, time 1 and
time 4 N.B. No control data both groups now
treated
Note lack of stunning progress on literacy
tests
39
Is there evidence of gains due to treatment?

• Improvement looks best for measures where there
  is no control data
• On reading measures where control group
  available, initial gain in the treated group on
  reading was small and not sustained

40
Costs in relation to benefits

• Cost of treatment is around 1700-2000
• Surely it is a price worth paying in the
  attempt to transform the life of your child so
  that they are able to enjoy school, to develop
  social skills, to develop good sporting skills,
  to have good prospects in life?
• Dore (2006) p. 171
• Money-back guarantee
• But only if no physiological change i.e.
  child who improves on balance/eye tracking wont
  get refund, even if dyslexia/ADHD etc unchanged

41
Fish oil
42
The theory

• Certain highly unsaturated fatty acids (HUFAs)
  important in brain development and neuronal
  signal transduction
• Brain function may be affected by
• Dietary insufficiency
• Genetic abnormality in phospholipid metabolism
• Administration of HUFAs may improve synaptic
  transmission

43
Evidence of abnormal fatty acid levels in dyslexia

• Clinical signs of fatty acid deficiency found in
  adults with dyslexia (Taylor et al, 2000)
• Clinical signs of FAD correlate with severity of
  dyslexia in males only (Richardson et al, 2000)
• 7 items including dry skin/hair/nails, excess
  thirst, frequent urination

44
Evidence from treatment trials

• Significant reduction in ADHD symptoms in
  children with comorbid dyslexia/ADHD cf. placebo
  (Richardson Puri, 2002)
• reading not assessed (!!??)
• Cf. no improvement of ADHD symptoms vs. placebo
  in 2 studies of ADHD, though studies vary in
  fatty acid, sample, etc
• Hirayama et al. 2004
• Voigt et al. ,2001

45
Evidence from treatment trials

• Oxford-Durham study on children with
  developmental coordination disorder Treated show
  significantly more improvement in literacy
  (reading age!) and ADHD symptoms (Richardson
  Montgomery, 2005)
• Requests to see raw data to identify children
  with dyslexia from this sample get no response

46
Controlled trial of fish oil in dyslexic adults

• Cyhlarova et al, 2007 report baseline results -
  no differences in membrane fatty acid levels
  between dyslexic and control adults, though ratio
  of types of fatty acid differs
• Requests for information on progress of this
  treatment trial get no response

47
Q1. Is the theory scientifically plausible?

• Membrane phospholipid deficiency speculative
  theory developed to account for schizophrenia,
  extended to neurodevelopmental disorders
• Most plausible when applied to children who show
  physical symptoms suggestive of essential fatty
  acid deficiency

48
Q2. Does evidence for efficacy goes beyond
testimonials? have studies been done with
groups for whom treatment is recommended?

• Several clinical trials but only one specifically
  on children with dyslexia (and ADHD) - did not
  look at reading outcomes
• Study of developmental coordination disorder
  included measures of reading as part of outcome
  assessment

49
Q3. Is there evidence that gains are due to
treatment?

• Inclusion of control group makes it possible to
  take into account practice, maturation, etc.

50
Q4. Are costs reasonable?

• around 19.50 for 60 capsules (1 per day)
• Treatment may need to be long term
• 118 per year

51
Barriers to objective evaluation

• Failure to recognise important effects of
• expectations
• maturation
• practice
• statistical artefact

52
Human tendency to be impressed by testimonials

• N.B. Testimonials problematic because
• selective
• often at odds with objective evaluation

53
Human tendency to think something that has taken
time/effort/money was worthwhile
54
Trial of Sunflower therapy

• Includes applied kinesiology, physical
  manipulation, massage, homeopathy, herbal
  remedies and neuro-linguistic programming
• Similar gains in test scores for clinical and
  control children
• Higher academic self-esteem in those undergoing
  treatment
• 57 of parents thought Sunflower therapy was
  effective in treating learning difficulties

Bull, L. (2007). Sunflower therapy for children
with specific learning difficulties (dyslexia) A
randomised, controlled trial. Complement Ther
Clin Pract, 13, 15-24.
55
Human tendency to be impressed by neuroscientific
explanations
56
The seductive allure of neuroscience
explanations

• Weisberg et al. 2008. J. Cognitive Neuroscience
  20 470-7

People given explanations of psychological
phenomenon that were accepted or vacuous and
judged if satisfactory. With neuroscience just
added phrases such as brain scans indicated and
because of the frontal lobe circuitry involved
57
Conclusions

• Finding the neuroscientific basis of dyslexia is
  an important goal
• However, we are a long way from having reached
  that goal
• Even when we reach it, it may not be obvious how
  to translate knowledge into intervention
• We need to adopt as critical an approach
  neuroscientific explanations as we do to other
  aspects of dyslexia research claims that
  neuroscientific treatments are superior to
  conventional treatments are not, in our current
  state of knowledge, supported

58
Dorothy Bishop Oxford Study of Childrens Communic
ation Impairments, Department of Experimental
Psychology, South Parks Road, Oxford, OX1
3UD, England.
for reading list see http//www.psy.ox.ac.uk/osc
ci/
Photography Biljana Scott