As people with Down syndrome (DS) are living longer, it has become clear that dementia in adulthood is very common (Bittles & Glasson, 2004). Symptoms may begin at a younger age than in people without DS (Coppus et al., 2006) and the kind of symptoms seen first may also differ. Although we may think of dementia as covering a range of diseases that mainly affect people’s memory, in people with DS, carers may see personality changes and problems with planning and controlling behaviour before memory problems are noticeable (Ball et al., 2006).
As dementia may develop differently in people with DS, we do not know if the criteria that we use to diagnose dementia in the general population are suitable for this group. Sheehan and colleagues therefore looked at two of the most commonly used diagnostic manuals: the International Classification of Diseases (ICD) – 10 and Diagnostic and Statistical Manual of Mental Disorders-IV-Text Revision (DSM-IV-TR) and investigated whether they were reliable (i.e. gave consistent results) and valid (i.e. identified what they should identify) when diagnosing dementia in DS.
Clinical vignettes (narrative-style summaries) were written from the medical records of 85 adults with DS. These included information from memory assessments, but did not mention clinical diagnosis or treatment with drugs prescribed for dementia. For each assessment, the patient’s mood, cognitive ability, daily living skills, social behaviours, neurological function and mental state were examined. 64/85 (75.3%) had an existing diagnosis of dementia.
The vignettes were each reviewed by two independent raters, all of whom were clinicians (psychiatrists or psychologists) and part of a ‘dementia in intellectual disabilities’ special interest group.
The raters applied criteria from the DSM-IV TR and ICD-10, as summarised below. A further rating was completed using their clinical judgement. To assess whether dementia diagnoses remained stable over time, the outcomes of the rater’s assessments from the time of clinical diagnosis then two further assessments were compared. Three full assessments covering a maximum period of 18 months were available for 26/64 people with a recorded diagnosis.
Inter-rater reliability was assessed in a subset of 23 cases, to see how closely two different people applied the same criteria.
To receive a diagnosis of dementia, both ICD-10 and DSM-IV-TR require presence of memory impairment, in addition to at least one other impairment. For ICD-10 this may include decline in executive functions, behavioural or emotional function, emotional lability, irritability or a change in social behaviour, with symptoms being present for at least 6 months. For DSM-IV-TR the disturbance may be in executive function or may include aphasia, apraxia or agnosia. DSM-IV-TR criteria require a gradual onset and progression, but give no specific time frame. Both list delirium and any other CNS, systemic or substance-induced conditions as exclusion criteria, and for DSM-IV-TR there is an additional exclusion criterion for any other mental illness.
Age ranged from 35.5 – 70.9 years (mean ± SD = 55.4 ± 6.6 years). Level of intellectual disability (ID) was considered mild in 21 people, moderate in 40, severe or profound in 12 and unknown in the remaining 12.
• Raters agreed with an existing diagnosis of dementia in 84.4% (n=54) of cases, when using any of the available criteria. The same level of agreement was seen when using clinical judgement only.
• 45 (70.3%) cases met ICD-10 criteria for ‘dementia’ or ‘tentative dementia’. Of these, 28 (43.8) met ‘dementia’ and 17 (26.6%) met ‘tentative dementia’ criteria.
• 36 (56.3%) clinician-diagnosed cases of dementia met the DSM-IV-TR criteria for dementia.
• 32 (50%) of cases met criteria for dementia in all three measures.
• In 10 cases (15.6%) the raters did not diagnose dementia, although the patient had an existing diagnosis. When investigating further, these cases often had co-morbid mental and/or physical illness which could potentially explain the decline, the decline was very recent or there was insufficient evidence to confidently diagnose dementia.
Stability over time
Over the time course of the 3 assessments, a firm diagnosis was made in progressively more cases at each time point.
• ICD-10: Cohen’s kappa = 0.911
• DSM-IV-TR: Cohen’s kappa = 0.704
• Clinical judgement: Cohen’s kappa = 0.826
All significant at p ≤ .001. These results indicate that all criteria showed strong agreement between different raters.
The authors concluded that clinical diagnosis of dementia in DS is valid and reliable, and that experienced clinicians are able to diagnose dementia in this population earlier than would have been possible if relying on the criteria contained in either the ICD-10 or DSM-IV-TR diagnostic systems.
Although the ICD-10 and DSM-IV-TR criteria excluded several cases, their reliability between raters was strong. While clinical judgement may be considered more subjective, the inter-rater reliability of clinicians’ own judgements was actually equally strong. The authors therefore argue that clinical diagnosis can be used as the standard against which newer systems (e.g. the DSM-5) can be measured.
Strengths and Limitations
There is strength in the diversity of people included in this study: a wide age range, covering a range of abilities. Raters were kept blind to original diagnoses and medication that would have suggested an existing dementia diagnosis and they rated each case independently prior to reaching a consensus.
The raters all had substantial clinical experience in this field, however they were making their judgements based on data from existing medical records. While this would have enabled a potentially larger sample, there are inherent limitations in using data of this kind. The quality of data will be dependent on what was originally recorded. The records themselves were also transcribed into a vignette. This will have made the relevant information easier to read through and assess, but may also add another step in the process where data could be lost or changed.
There is also a question regarding the existing diagnosis of dementia. Although we are told that memory assessments covered a range of abilities and cognitive functions, we do not have access to the criteria the original diagnosis was made on.
The strong association between the raters’ clinical judgement and the original diagnoses shows that there are likely to be consistent ideas related to what dementia looks like in DS that clinicians are reliably using to identify the disease in this population. In those who had three separate assessments available, the proportion meeting ICD-10 and DSM-IV-TR criteria increased over time, suggesting that clinicians were accurately able to diagnose this disease early and that these individuals did go on to meet more stringent criteria.
However this suggests a need for further research with a focus on how dementia develops in DS. While these clinicians are able to draw on their experience of seeing dementia develop in their patients, for training new clinicians and developing future research projects we need clear criteria, which it would seem from these results are slightly different in DS to those used in the general population.
Sheehan, R., Sinai, A., Bass, N., Blatchford, P., Bohnen, I., Bonell, S., … Strydom, A. (2014). Dementia diagnostic criteria in Down syndrome. International Journal of Geriatric Psychiatry [abstract]
Ball, S. L., Holland, A. J., Hon, J., Huppert, F. A., Treppner, P., & Watson, P. C. (2006). Personality and behaviour changes mark the early stages of Alzheimer’s disease in adults with Down’s syndrome: findings from a prospective population-based study. International Journal of Geriatric Psychiatry, 21(7), 661–673. [abstract]
Bittles, A. H., & Glasson, E. J. (2004). Clinical, social, and ethical implications of changing life expectancy in Down syndrome. Developmental Medicine & Child Neurology, 46(4), 282–286. [abstract]
Coppus, A., Evenhuis, H., Verberne, G.-J., Visser, F., Van Gool, P., Eikelenboom, P., & Van Duijin, C. (2006). Dementia and mortality in persons with Down’s syndrome. Journal of Intellectual Disability Research, 50(10), 768–777. [abstract]