Validation of two short dementia screening tests in Indonesia

Eef Hogervorst, Fidiansjah Mursjid, Raden Irawati Ismail, Sabarinah, Martina W Nasrun, Mochtar, Theresia Ninuk, Stephan Bandelow, Subarkah, Maria Francisca Lindawati Soetanto, Tri Budi Wahyuni Rahardjo

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

13 Citations (Scopus)

Abstract

Dementia is on the increase worldwide, and developing countries are expected to carry the burden of this. Relatively little is known about dementia prevalence in Indonesia. This chapter discusses two short screening tests to assess dementia in rural and urban Indonesian cohorts. At baseline in 2006/7, 719 elderly were included from rural and urban sites on Java. Large differences appeared in dementia prevalence in those over 60 years of age between urban (3%) and rural sites (7-16%) employing two dementia screening tests also used in Oxfordshire with the same cut-offs. An in depth study was performed on the rural sample from East Java to validate the cut-offs of the tests. For this study, Javanese Indonesian elderly from 4 villages around Borobudur were asked to participate. 113 agreed to participate and these were tested in a health center by medical experts and trained research assistants. The screening test cut-offs were validated against consensus based clinical dementia diagnoses by an expert psychiatrist, nurses and GP, which were based on a gold standard diagnostic instrument for dementia diagnoses from Cambridge. In addition, a sub sample of these participants was tested in depth by another psychiatrist using questions from our expert dementia diagnostic system developed at Oxford University. The adapted memory screening test was shown to have similar cut-offs for dementia (19.5 for controls and 14.5 for cases) as in Oxfordshire and the Mini Mental Status Examination (MMSE) had optimal sensitivity (100%) using a similar cut-off of 24. However, for optimal specificity, the MMSE was shown to require a lower cut-off of 21.5 and MMSE scores were also affected significantly by educational level. It was unclear how many of these cases had vascular dementia (VaD), as stroke, transient ischaemic attack and myocardial infract assessed by self-report were rare (n=1-2) and were only reported by controls, suggesting a recall bias. There was also no difference in diabetes mellitus or (high) blood pressure between cases and controls which could have increased risk. Physical examination suggested no other morbidity driving the dementia (e.g. infectious or lung disease). The cut-offs were also tested by another expert psychiatrist on a sub sample of these participants using her clinical assessment and aided by questions from our expert dementia diagnostic system. Agreement between psychiatrists was high (79%) on diagnoses of these 28 participants, with only 6 disagreed on. Of these, only diagnoses of 2 participants were disagreed on whether these had dementia or were controls. Of 59 elderly patients from the villages who were tested in depth by the second psychiatrist, 17 were thought to have dementia, with most (53%) having Alzheimer's disease (AD) and 6 (35%) suspected of having VaD, with only one mixed (with stroke) case. There was no clear indication of other types of dementia, but two cases with dementia (12%) were thought to be related to systemic disease. The 19 preclinical cases (possible dementia) all had memory complaints, but scored significantly higher on the adapted memory test and MMSE than those with dementia and scored lower than controls (but not significantly so on the memory test), independent of age and gender. Optimal cut-offs for dementia on the memory test were again 19.5 for the total immediate recall (100% sensitivity and 78% specificity) and 24 for the MMSE (88% sensitivity and 96% specificity). This study showed that two short cognitive tests can be used for dementia screening in rural Java. It has been hypothesized that VaD is more prevalent than AD in East Asian countries, but we could not substantiate this. Future studies should investigate in more detail the prevalence of vascular and secondary dementias (due to thyroid or infectious disease, nutritional deficiency etc.). Use of this screening instrument in other ethnic groups in other developing countries also needs to be explored.

Original languageEnglish
Title of host publicationVascular Dementia
Subtitle of host publicationRisk Factors, Diagnosis and Treatment
PublisherNova Science Publishers, Inc.
Pages235-256
Number of pages22
ISBN (Print)9781611223132
Publication statusPublished - 2011

Keywords

  • Alzheimer's disease
  • Dementia
  • Developing countries
  • Screening

Fingerprint

Dive into the research topics of 'Validation of two short dementia screening tests in Indonesia'. Together they form a unique fingerprint.

Cite this