Hi Clarissa,

Have you considered the Zoo Map test - admittedly for planning, but has a problem solving element to it. Has been found useful in MCI/dementia see below -

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2744431/

Perhaps the 20 questions from the D-KEFS? I haven't used it myself, but I have been considering it in a clinical setting, and would like to know what others think of its utility with older adults.

Card Sorting tasks are traditionally good measures I suppose, but the WCST is too long for older people, & I haven't used other versions myself. Might be worth considering.

The Cognitive Estimates Test is interesting, but might be a little ambiguous for someone with cognitive impairment already. Plus might need to be adapted for different units of measurement (pounds vs kilograms etc).

Interested in hearing your thoughts.

Thank you Himani, really interesting answer. I have had a look at the D-KEFS subtests, and I think that that will be the best option for assessing early dementia for my study! I will have to take a look at those other tests as well though.

Thanks again!

Don't forget the classic Trails Making Test..

Free and quick to use, robust and very well researched.

From my own literature review:

AD Diagnosis

TMT performance was shown to be significantly different between a number of groups, based on diagnosis, suggesting discriminative capability. The performance of those with AD was shown to be significantly worse than the performance of controls for TMT-B error rates (Ashendorf et al, 2011; Stockholm et al, 2005), TMT-B time to complete (Baudic et al, 2006; Brown et al, 2011; Chen et al, 2000; Jungwirth et al, 2009; Stockholm et al, 2005; Yuspeh et al, 2002), TMT-A time to complete (Chen et al, 2000; Gleichgerrcht et al, 2011; Nathan et al, 2001), TMT-B – TMT-A (Yuspeh et al, 2002) and completion status (Schmitt et al, 2010). None of the studies failed to support these findings.

MCI conversion to AD

The performance of those with MCI was shown to be significantly better than AD patients and worse than controls for TMT-B error rates (Ashendorf et al, 2008), TMT- B time to complete (Brown et al, 2011, Marshall et al, 2011) and TMT-A time to complete (Brown et al, 2011). This includes all of the sourced papers that explored these group differences.

There were a number of studies that retrospectively or prospectively explored the value of TMT performance in predicting later conversion from MCI to AD. Ewers et al’s (2010) analysis of the ADNI (The Alzheimer’s Disease Neuroimaging Initiative study) 3-year longitudinal cohort database indicated that TMT-B (time to complete) baseline performance alone was as accurate as a multimarker model (64.1%) (including memory performance, TMT performance and CSF biomarkers) in predicting conversion from AD to MCI in either a 2 or 3 year prediagnosis period (64.6% accuracy). This is supported by Saxton et al’s (2004) 8 year longitudinal study in which TMT-B (time to complete) baseline performance correlated with AD onset at 3.5-5 years prior. However, in this case, verbal and delayed memory were more sensitive, predicting AD at 5.1-8.1 years prior. Grober et al’s (2008) 15 year longitudinal study indicated that TMT-B (time to complete) performance decline predicts conversion to AD 3-5 years prior to diagnosis, less than memory performance decline which predicted conversion 5-7 years prior. Sinai et al (2010) proposed, from their 4-year longitudinal study, that MCI with poor TMT-B/A (time to complete) and impaired episodic memory indicated a high probability of conversion to AD.

A number of studies have sought to identify the optimal assessment battery for the prediction of conversion to AD.  The most sensitive, in terms of duration prior to diagnosis is provided by Dickerson et al (2007). They suggest that baseline performance, 5 years prior to diagnosis, can predict conversion from MCI to AD if TMT-B (time to complete), clinical impairment (Clinical Dementia Rating) and verbal memory performance comprise the assessment model.

Understanding the conversion pathway is complex, however, TMT performance and decline is implicated in all models at least in the 1-5 year period prior to a diagnosis of AD.

Good point, thanks for this!  Very helpful

Typically problem solving is not affected in aMCI...which would be the MCI likely to lead to AD.  You could use the Ravens Progressive Matrices...the short form. Another possibility would be the Tower of London (Hanoi)...that is another classic problem solving task. 

Executive function tasks like Trails B are frequently good for predicting future decline in aMCI to AD (Albert et al) but most of the tasks described previously are not really problem solving tasks.  

We reviewed and compared all evidence based single tests for dementia in this publication: http://www.ncbi.nlm.nih.gov/pubmed/20808094

I use the Montreal Cognitive Assessment (MoCA).

Hi,

Maybe this article will interest you:

Article The Mindstreams MCI score: a practical clinical tool for ear...

Best regards,

Ya'akov