Article Text


Mild impairments in cognition in patients with type 2 diabetes mellitus: the use of the concepts MCI and CIND
  1. E van den Berg,
  2. R P C Kessels,
  3. E H F de Haan,
  4. L J Kappelle,
  5. G J Biessels
  1. Department of Neurology, Rudolf Magnus Institute of Neuroscience, University Medical Centre, Utrecht, Netherlands
  1. Correspondence to:
 Dr Geert Jan Biessels
 Department of Neurology, G03.228, University Medical Center Utrecht, PO Box 85500, 3508 GA Utrecht, Netherlands;

Statistics from

Type 2 diabetes mellitus (DM2) is associated with moderate cognitive impairment in verbal memory, mental flexibility, and information processing speed, while other cognitive functions remain relatively unaffected.1 Moreover, epidemiological studies have shown that DM2 patients have a twofold increased risk of developing either vascular dementia or Alzheimer’s disease.1,2 In the present study we examined whether mild cognitive impairment (MCI) and “cognitive impairment, no dementia” (CIND)—two concepts that are used to describe cognitive impairment in the transitional state between normal aging and early dementia—can be applied to the cognitive impairments encountered in a population based sample of DM2 patients. Recently, these concepts have attracted considerable attention, as individuals who meet the criteria for either MCI or CIND are known to have a substantially increased risk of developing dementia.3,4 MCI is defined as a memory deficit without impairments in other cognitive domains.3 Patients with MCI develop Alzheimer’s disease at an annual incidence of between 6% and 25%, compared with 0.2–3.9% in the general population of the same age. The broader concept of CIND is used to describe more general cognitive impairments, often encountered in relation to vascular risk factors. The diagnosis requires impairment in one or more cognitive domains and no dementia.5 A fivefold increased risk of developing Alzheimer’s disease, vascular dementia, or mixed Alzheimer/vascular dementia has been reported in patients with CIND.4 If either of the concepts MCI or CIND is found to be useful in describing the cognitive problems in DM2, it may help identify DM2 patients who are at increased risk of developing dementia.

Participants were recruited for the Utrecht diabetic encephalopathy study, which assesses the impact of macrovascular and microvascular disease on cognition in DM2. This research was approved by the medical ethics committee of the University Medical Centre Utrecht. Patients (n = 90) were aged between 60 and 75 years, were known to have had DM2 for at least one year, and were recruited through their general practitioner. Age and education matched control participants (n = 40) were recruited through the patient (mostly spouses) (table 1). Exclusion criteria were a psychiatric or neurological disorder (unrelated to DM2) that could influence cognitive functioning, a history of alcohol or substance abuse, and dementia. All participants were functioning independently at home and had intact comprehension of the Dutch language. The participants had an extensive neuropsychological examination (11 tasks) addressing the following cognitive domains in both a verbal and a non-verbal form: abstract reasoning, memory, working memory, information processing speed, visuo-construction, attention, and mental flexibility.

Table 1

 Characteristics of the participants and the number and relative frequencies of participants classified as having CIND or MCI

Previous studies used variable case definitions of CIND and MCI, often based on the diagnostic opinion of experienced clinicians. We preferred a numerical approach, comparing test scores with available age and education adjusted normative data. This procedure results in an objective classification and facilitates comparison of different studies. Performance of the participants on each test was rated as either within the normal range (0), below average (1), or impaired (2). “Normal performance” was defined as performance between −1 SD and +1 SD from the normative mean, “below average” as between −1 SD and −1.65 SD from the normative mean (the lowest 16% of the normal population), and “impaired” as below −1.65 SD from the normative mean (the lowest 5% of the normal population). Performance on a cognitive domain as a whole was classified as impaired when the average rating of tests in that domain was >1.

Participants were classified as having CIND if they were impaired in one or more of the cognitive domains.5 When memory was the only affected domain, the participant was also classified as having MCI, applying the Petersen criteria.3 As a decrease in information processing speed is common in an elderly population, impairment in this domain had to be accompanied by a rating of 1 or 2 in more than half the tasks in another domain (average rating of tests in that domain >0.5) for the individual to be classified as CIND.

Overall, the number of tasks on which performance was impaired was higher in DM2 patients than in controls (patients: five tasks, interquartile range two to eight; controls: two tasks, interquartile range one to five; Mann–Whitney U test, p<0.05). The domains that were affected most often in both groups were mental flexibility (patients 11%; controls 10%) and information processing speed (patients 17%; controls 11%). Memory was relatively spared (patients 4%; controls 3%). Significantly more patients than controls met the criteria for CIND (χ2 test, p<0.05; table 1). The proportion of participants classified as having MCI did not differ between the two groups. In addition, hypertension, major vascular events (excluding non-invalidating stroke), retinopathy, and neuropathy were more common in the DM2 patients (table 1). Within the DM2 group no significant differences were found between the patients with or without CIND for age, diabetes duration, HbA1c, or any of the factors described above. The same applied for the controls with or without CIND.


The results show that DM2 patients overall had more cognitive impairments than control participants, predominantly affecting mental flexibility and information processing speed; these cognitive domains are known to be the most sensitive to cognitive decline associated with aging. The prevalence of MCI and CIND in the control group was comparable to previous population based studies.3,5 CIND, but not MCI, was significantly more common in DM2 patients than in controls. Memory impairment, which is the main feature of MCI, was not the most prominent impairment in the DM2 patients. Rather, a more general pattern of cognitive impairment, affecting multiple domains, was observed. This pattern fits better within the broader concept of CIND. These results illustrate that, because of its focus on memory impairments, the concept of MCI should be used with caution outside the field of Alzheimer’s disease research. The concept of CIND appears to be applicable more widely, but its current broad and non-specific definition remains a limitation. Thus the nature and magnitude of the cognitive impairments should also be taken into account when the classification of CIND is used. Future prospective studies, using clear criteria, should resolve whether the concept of CIND could serve to identify DM2 patients who are at increased risk of developing dementia.


This research was supported by grant 2001.00.023 from the Dutch Diabetes Foundation. The Utrecht Diabetic Encephalopathy Study Group consists of (departments in alphabetical order, all part of the University Medical Centre, Utrecht, Netherlands): Department of Clinical Neurophysiology, A C van Huffelen; Department of Epidemiology, Julius Centre of Health Sciences and Primary Care, and Department of Neurology, A Algra; Department of Family Practice, G E H M Rutten; Department of Internal Medicine, H W de Valk; Department of Medical Pharmacology, W H Gispen; Department of Neurology, G J Biessels, L J Kappelle, S M Manschot, J van Gijn; Department of Neuropsychology and Helmholtz Instituut, A M A Brands, E H F de Haan, R P C Kessels, E van den Berg; Department of Radiology, J van der Grond.

View Abstract


  • Competing interests: none declared

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.