Feature Article

Forgetfulness, stress or mild dementia? Cognitive assessment of older patients


A targeted neurological examination should be completed to assess for reversible causes. In particular, this should include assessment for:

  • any focal neurological deficits (suggesting stroke)
  • extraocular movements
  • abnormal gait or other movements (suggesting Parkinson’s disease, a Parkinson’s plus syndrome or dementia with Lewy bodies)
  • peripheral neuropathy (such as in vitamin B12 deficiency).

Neurological signs and their possible aetiology in patients with cognitive impairment are listed in Table 2.6

Cognitive screening

Several screening tools are available for the rapid assessment of cognitive function in the clinical setting. Probably the most widely known, well-validated tool is the Mini-Mental State Examination (MMSE).7 Although the MMSE does not predict who will develop dementia among those with MCI, a systematic review of more than 10,000 participants has shown that it has high sensitivity (0.81; 95% CI, 0.78 to 0.84) and specificity (0.89; 95% CI, 0.87 to 0.91) for the diagnosis of dementia.8,9 The limitations of the MMSE include the impact of practice effects, limited scope for short-term memory testing and failure to test frontal lobe function. The Montreal Cognitive Assessment (MoCA) has an additional component to test frontal lobe function and has comparable performance to the MMSE for the detection of MCI.9


The General Practitioner Assessment of Cognition (GPCOG) was designed specifically for use by GPs and consists of a short patient interview (less than four minutes) and an optional informant interview (two minutes).10 It has been translated into multiple languages, has a negative predictive validity similar to that of the MMSE and is freely available online from the NHMRC Dementia Collaborative Research Centre .

Several other screening tools are available but may be less useful to GPs because of cost and time prohibitions. A comprehensive summary and comparison of screening tools is available from the Dementia Knowledge Translation Hub.


Importantly, all cognitive screening tests need to be interpreted in the clinical context. For example, patients who are depressed may, in addition to having difficulty with short-term recall, appear tired and flat in affect and have difficulty sustaining effort during testing and a 10 generalised slow processing speed. These symptoms and signs are less likely in patients with MCI or dementia. Moreover, screening tools lack sensitivity in patients with MCI, especially those with high premorbid cognitive function. Consequently, persistent concerns of a patient or informant should not be dismissed, even if the patient scores above the cut-off for cognitive impairment on screening tests.

Although a cross-sectional cognitive assessment is of benefit, repeat testing after 12 months is often helpful to provide the longitudinal follow-up frequently needed to clarify the diagnosis.


Laboratory investigations

Laboratory testing should be completed to exclude reversible causes of cognitive impairment. Basic screening should include:

  • full blood count
  • biochemistry, including renal and liver function tests, measurement of electrolytes (including calcium, magnesium and phosphate), fasting blood glucose and lipid profile
  • thyroid function tests (TSH level)
  • serum vitamin B12 and folate levels.


Dr Takács is a Psychiatry Registrar at NorthWestern Mental Health, Melbourne, Vic. Dr Koncz is a Neuropsychiatry Fellow at the Neuropsychiatric Institute, Prince of Wales Hospital, Sydney; and a PhD Student at the Centre for Healthy Brain Ageing (CHeBA), UNSW Sydney. Dr Mohan is a Senior Lecturer and Research Fellow at CHeBA, UNSW Sydney; and a Neuropsychiatrist at the Neuropsychiatric Institute, Prince of Wales Hospital, Sydney; Professor Sachdev is Professor of Neuropsychiatry at UNSW Sydney; and Director of the Neuropsychiatric Institute at the Prince of Wales Hospital, Sydney, NSW.