NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Boustani M, Peterson B, Harris R, et al. Screening for Dementia [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2003 Jun. (Systematic Evidence Reviews, No. 20.)

Cover of Screening for Dementia

Screening for Dementia [Internet].

Show details


Major Findings

Our systematic review found no randomized trial that evaluated the overall efficacy of dementia screening in primary care. Therefore, we attempted to evaluate screening by reviewing the literature using a step-wise analytic framework that focused on prevalence of undiagnosed dementia properties of screening tests, studies of treatment options for patients and caregivers, and adverse effects of screening and treatment.

The prevalence of the dementia syndrome increases rapidly in the seventh and eighth decades of life; this ailment affects more than 25% of people who are 85 years of age and older. The burden of this disease also extends to the caregivers. Among all primary care patients over age 65, 1.8% to 12% have undiagnosed dementia, and one-half to two-thirds of all cases of dementia in primary care populations are undiagnosed.

The Mini-Mental State Examination (MMSE) is the best-studied brief screening tool for dementia. A cut-point of 24 to 26 out of 30 points is usually accepted as a positive screen and should lead to diagnostic evaluation with further history, examination, and testing for dementia. Scores must be adjusted for educational attainment. The MMSE can identify cases of dementia with a sensitivity of 71% to 92% and specificity of 56% to 96%.

Dementia is often treatable but rarely curable or reversible. No more than 1.5% of all dementia cases are fully reversible. About 60% of people with dementia syndrome have Alzheimer's disease and 15% have vascular dementia. Nearly all cases of dementia are irreversible, so the clinical benefits of primary care screening will be heavily influenced by the benefits of early diagnosis and treatment of Alzheimer's disease, vascular dementia, and other irreversible etiologies.

Treatment of Alzheimer's disease with cholinesterase inhibitors for 6 to 12 months results in modest but consistent improvements in cognition and clinician global impression of change scores. Compared with patients receiving no such treatment, patients receiving cholinesterase inhibitors displayed clinically evident positive changes; their decline was delayed and they maintained independence equivalent to 3 to 5 months' delay in the natural history of the disease.

One randomized controlled trial (RCT) found that anti-oxidants (selegiline and vitamin E) delayed by 7 months a combined outcome of mortality, nursing home placement, and functional decline in patients with moderate stage Alzheimer's disease. RCTs of nimodipine and aspirin have shown no effect on vascular dementia.

Treatment of behavioral problems and depressive symptoms in early dementia is poorly studied. Two small RCTs demonstrated effective treatment of depression in early dementia. Neuroleptic medications and nonpharmacologic interventions reduce agitated behaviors in later stages of dementia, but these therapies are rarely studied in early dementia or in community or primary care settings. Intensive interventions to support caregivers delayed nursing home placement for Alzheimer's disease patients, but they had little or no demonstrated direct benefits on either patient or caregiver. These interventions are varied and complex; such multi-component interventional programs have not been evaluated in primary care setting and their effectiveness warrants further assessment.

In 2001, the American Academy of Neurology published an evidence-based review of the diagnosis, evaluation, and treatment of dementia.180 The authors found that, on average, cholinesterase inhibitors produce a small benefit in Alzheimer's disease patients. They concluded that insufficient data exist to make any recommendations regarding cognitive screening of asymptomatic individuals. The report recommended diagnostic evaluation and monitoring for persons with mild cognitive impairment because of their increased incidence of dementia.180

This conclusion is similar to that in the 1996 practice guideline issued by the Agency for Health Care Policy and Research75 and to the recommendations of the 2001 Canadian Consensus Conference on Dementia181 in that it recommends a complete evaluation and close follow-up of individuals with memory complaints or functional decline. These reviews did not include 2 recent 12-month RCTs, which showed that the cognitive and clinician global impression of change benefits of cholinesterase inhibitors extended to 1 full year and translated to up to a 5-month delay in clinically evident decline in patients' ability to maintain their independence in instrumental and basic activities of daily living (ADLs).93, 95 Although these trials may provide greater evidence for treatment of diagnosed disease than have been available heretofore, they do not directly address the issue of screening to promote earlier treatment.

The degree to which participants in treatment trials are representative of patients who would be identified by screening is not clear. Nonetheless, the mean MMSE score for individuals in treatment trials was similar to the scores for undiagnosed cases of dementia that were detected by screening.

Limitations of this Literature

Our review has some limitations. First, we limited our review to studies of mild to moderate dementia among community-dwelling patients. These criteria approximated the patient population likely to be identified by primary care screening, but they may exclude compelling studies of treatment outcomes. Second, we limited our search to English-language articles, and thus we may have excluded studies from similar non-English-speaking populations. We believe that our review successfully captured all studies that met inclusion criteria. A repeat search of 4 data sources—MEDLINE, PsycINFO, EMBASE, and the Cochrane Collaboration library—and extensive peer review of our draft systematic evidence review identified only 1 additional study that met our inclusion criteria.

Benefits and Harms

Data are insufficient to create an outcomes table of screening strategies for dementia. Harms of dementia screening have not been systematically studied. Potential harms include risk of depression and anxiety, the time and cost of screening, and possible labeling effects. Once a diagnosis of dementia is given, patients will be unlikely to qualify for long-term care insurance or acceptance into continuous care retirement communities.

In a survey of elderly and caregivers of Alzheimer's patients, most participants wanted to be told the diagnosis of dementia.182 In another study, Jha et al. investigated the reaction of elderly patients in outpatient clinics to the disclosure of their diagnosis of dementia compared with depression.183 The authors found no significant difference between patients with dementia or depression in their wish to know their diagnosis. Patients with dementia, even if they felt upset, preferred to be told their diagnosis.

Benefits of a comprehensive screening strategy have also not been studied directly, but they may be extrapolated from existing studies. Potential benefits would accrue to the 3% to 12% of primary care patients age 65 and older who have undiagnosed dementia. Based on the strong association between advancing age and the risk of dementia, a proposed screening strategy will have increased yield if it is begun at more advanced ages such as 75 years.

One unstudied potential benefit of dementia screening is the opportunity it affords for advance care planning. Individuals identified with early dementia by screening have the opportunity to discuss the nature of the syndrome, its prognosis, and future planning in regard to health care, safety, and financial planning. They may be able to formulate advance directives, choose a power of attorney for financial and personal care decision making, consent to participate in research, and contemplate issues such as motor vehicle driving, self-neglect, financial victimization, and housing relocation. Some may argue that earlier diagnosis of dementia syndrome, regardless of the cause, allows the patient, family, and physicians to plan more effectively for future events. It may also permit earlier and more effective administration of medication for other coexisting conditions by improving medication adherence and avoiding drug interactions. We found no study that verified or quantified these potential benefits.

Early detection of Alzheimer's disease can offer patients a chance to begin therapy with drugs such as cholinesterase inhibitors, vitamin E, or selegiline. Doing so may delay their functional decline and maintain their independence in performing both basic and instrumental ADLs for approximately 5 to 7 months. It may also lighten the burden on their caregivers and ultimately reduce total health care expenditure by delaying the time to nursing home admission and other costly outcomes.

Treatment of potentially reversible dementia is often proposed as a justification for screening, but we found few of these patients whose dementia was actually reversible with therapy. Thus, screening must benefit the large majority of patients with irreversible causes to demonstrate public health benefit. Patients with Alzheimer's disease account for 60% of cases of dementia, and both pharmacologic and nonpharmacologic interventions show some clinical benefits in this disease. The efficacy of treatments for other causes of dementia is unproven.

Future Research Needs

Our review highlights important limitations of the current research on screening and treatment of dementia. No study of the overall effectiveness or benefit of screening in primary care has been done. Given the high prevalence of undiagnosed dementia among primary care patients over age 65 and evidence of the efficacy of cholinesterase inhibitor treatment for mild to moderate Alzheimer's disease, a trial of screening would be very helpful. Such a trial should also monitor costs and harms. We are aware of one 4-year RCT at Indiana University that will evaluate the efficacy of an integrated program of screening, diagnosis, and management compared to usual care in a primary care setting (personal communication, Christopher C Callahan, MD; Associate Professor, Indiana University School of Medicine; Director IU Center for Aging, December 19, 2001).

The MMSE is the best studied screening test for dementia, but it has been criticized for limited specificity and the need to adjust scoring based on age and educational attainment. The Clock-Drawing Test is very easy to administer and has good screening test characteristics in enriched referral populations. Future research should examine other promising brief screening tools that may be less education dependent and test their positive and negative predictive value in primary care screening strategies.

Although caregiver burden, health care utilization, and complications in managing comorbid conditions are common in dementia, little work to date has dealt with these important aspects of the dementia syndrome. Future intervention trials, therefore, should examine outcomes not just for dementia syndrome patients but also for their caregivers.

Future treatment studies should consider expanding the usual outcome measures to include clinically important domains. In addition to standard measures of cognitive function and clinical global impression of change, we recommend inclusion of outcome measures for functional status and behavioral symptoms in all major clinical trials of dementia treatment. In addition, outcomes should be reported in temporal measures such as time to decline or survival analyses, to provide data on stabilization of disease course.

Despite the emergence of cerebrovascular disease as both a direct and an indirect cause for dementia syndrome, the literature offers very limited information about the potential benefit of modifying or treating cerebrovascular disease or atherosclerosis on dementia.47 One RCT found that therapy (nitrendipine with the possible addition of enalapril, hydrochlorothiazide, or a combination) for isolated systolic hypertension among elderly reduced the incidence of dementia from 7.7 to 3.8 cases per 1,000 person-years.184 Conversely, another trial did not detect any reduction in dementia incidence with the treatment of hypertension by low-dose diuretic, beta-blocker or both.185 Another one found no effect of antihypertensive treatment on the cognitive function of older people with isolated systolic hypertension.186 In addition to the previous primary prevention trials, our systematic review found 1 review that assessed the efficacy of aspirin in vascular dementia.117 A new RCT of aspirin with and without antihypertensive therapy for early vascular dementia is needed to improve the evidence base for treatment of this common disease.

Much of the current clinical approach to dementia is symptomatic treatment for psychological and behavioral problems related to dementia syndrome, yet study of these treatments is quite limited. Behavioral problems occur at all stages of dementia and they are the main reason for caregiver stress and institutional placement. Some RCTs have attempted to assess the effect of drug therapy on noncognitive symptoms, but they tended to use instruments that lacked the ability to detect clinically meaningful changes.140 Future research is needed to fill gaps in the data about the psychometric properties of such instruments, so that they can be applied with greater confidence in trials or effectiveness studies. Future RCTs of interventions to change behavioral symptoms should also include measures of cognition, functional status, and clinical global impression of change, to clarify the mechanism of behavioral change and possible adverse effects.

No RCT has yet evaluated the efficacy of nonpharmacologic interventions on behavioral symptoms in patients with mild to moderate dementia who were living at home. This may be another important direction for research. Depression occurs in about one-third of patients with mild or early dementia, but we found only 2 small RCTs that evaluated the efficacy of antidepressant therapy in this population.4, 111 Future research in dementia therapy should target these symptoms in mild to moderate stages. Such studies should expand outcome measures to include behavioral as well as cognitive and functional measures, and they should incorporate injury prevention, health care utilization, and effect on the clinical management of other comorbid conditions.

Mild cognitive impairment (MCI) has been considered a significant risk factor for the development of Alzheimer's disease because of its high annual conversion rate (close to 16%). Therefore, trials to evaluate the efficacy of the current treatment strategies for Alzheimer's disease, such as cholinesterase inhibitors, gingko biloba, vitamin E, aspirin, and hypertension control, in MCI are warranted.

PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...