Alzheimer's Disease Assessment Scale-Cognitive
- Introduced by
- William Rosen
- Year
- 1984
- Type
- Clinician-administered cognitive scale for Alzheimer's disease
- Family
- cognitive assessment
What does it do?
The Alzheimer's Disease Assessment Scale-Cognitive (ADAS-Cog) is a clinician-administered cognitive assessment instrument designed specifically to measure cognitive decline in Alzheimer's disease.
Developed by Rosen, Mohs, and Davis in 1984 and published in the American Journal of Psychiatry, the ADAS-Cog has become the gold standard outcome measure in pharmaceutical trials of antidementia drugs. It is sensitive to disease progression and capable of detecting cognitive change over periods as brief as 6–12 months.
The intuition behind the method
The ADAS-Cog assesses the cognitive domains most commonly affected by Alzheimer pathology: memory, language, praxis, and visuospatial function. The instrument samples each domain with multiple items and weights them to reflect clinical relevance. Higher scores indicate greater cognitive impairment. The original 11-item version (ADAS-Cog11) and the extended 14-item version (ADAS-Cog14) are both widely used; the 14-item variant adds items assessing delayed recall, digit span, and letter fluency, enhancing sensitivity to treatment effects.
History
The ADAS-Cog was developed by Rosen, Mohs, and Davis at Mount Sinai School of Medicine in 1984 as part of the larger Alzheimer's Disease Assessment Scale (ADAS), which includes a non-cognitive behavioral subscale. The cognitive subscale rapidly became the standard outcome measure in pharmaceutical development for Alzheimer's disease. In 1997, Mohs and collaborators published an extended version (ADAS-Cog14) adding items for delayed recall, digit span, and letter fluency to enhance sensitivity to cognitive change in treatment trials. Both versions remain widely used; regulatory agencies (FDA, EMA) accept ADAS-Cog as a primary cognitive efficacy measure.
When to use it
The ADAS-Cog is indicated for measuring cognitive decline in patients with mild-to-moderate Alzheimer's disease, tracking disease progression in clinical practice, and as the primary cognitive outcome in pharmaceutical trials. It is less suitable for very mild cognitive impairment (ceiling effect) or severe dementia (floor effect). The ADAS-Cog is not appropriate for screening or diagnosis alone and should always be integrated with clinical assessment, functional evaluation, neuroimaging, and biomarkers. It is particularly valuable in research settings and specialized memory clinics.
Limitations and considerations
- Long administration time — 30–40 minutes is burdensome in busy clinical settings and unsuitable for frequent monitoring or cognitively impaired patients with poor cooperation.
- Limited utility at extremes — ceiling effect in cognitively normal individuals and those with minimal impairment; floor effect in severe dementia where almost all patients score near maximum.
- Requires clinical training — administration and scoring are more complex than the MMSE and demand clinician expertise to ensure standardization and reliability.
- Not diagnostic — ADAS-Cog measures cognitive change but does not diagnose Alzheimer's disease or differentiate Alzheimer's from other dementias.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Gold standard in Alzheimer drug trials — sensitivity to treatment effects and disease change has been established across thousands of patients in clinical trials, making it the benchmark cognitive measure in neurodegenerative disease research.
- Sensitivity to change — detects clinically meaningful cognitive decline over 6–12 months in mild-to-moderate Alzheimer's disease, more reliably than the MMSE in this range.
- Comprehensive cognitive sampling — assesses memory, language, praxis, attention, and visuospatial function, providing a multidimensional profile rather than a single global score.
- Extensive normative data — available for diverse populations, ages, and disease stages, with published conversion tables linking ADAS-Cog to MMSE scores.
Applications
- Clinical trial efficacy endpoint: A pharmaceutical company enrolls 300 patients with mild-to-moderate Alzheimer's disease in a 12-month, placebo-controlled trial of a novel γ-secretase inhibitor. ADAS-Cog14 is administered at baseline, 6, and 12 months; the drug-treated group shows 2-point improvement (indicating slowing of decline) versus 6-point decline in placebo, meeting the primary efficacy hypothesis.
- Memory clinic disease monitoring: A neurologist uses ADAS-Cog scores every 12 months to track cognitive decline in a patient with diagnosed Alzheimer's disease on cholinesterase inhibitor therapy. Baseline score is 18; after 2 years, score rises to 26, indicating worsening cognition despite treatment, prompting discussion of disease stage and care planning.
- Research cohort characterization: A longitudinal cognitive aging study enrolls 500 patients across the cognitive spectrum (normal aging, mild cognitive impairment, dementia). ADAS-Cog scores at baseline stratify participants by disease severity and predict conversion rates from MCI to dementia over 3 years.
- International Alzheimer collaboration: An international consortium harmonizes ADAS-Cog data across 10 memory clinics in 8 countries using standardized administration training and video-recorded item scoring to minimize examiner variance and combine data for cross-national phenotyping.
How it works
How it works
The ADAS-Cog is a clinician-administered instrument requiring 30–40 minutes to complete. The examiner directly administers tasks to the patient and scores responses according to detailed criteria provided in the manual. The ADAS-Cog11 comprises 11 items; the ADAS-Cog14 adds 3 additional items (delayed word recall, digit span, letter fluency). No special equipment is required beyond paper, pencil, and a stopwatch.
Each item on the ADAS-Cog is scored on a scale indicating severity of impairment; the precise scoring depends on the specific item. The raw score is obtained by summing individual item scores. The original ADAS-Cog11 total score ranges from 0 to 70, where 0 indicates no cognitive impairment and 70 indicates maximum impairment. The ADAS-Cog14 total score ranges from 0 to 100. Lower scores reflect better cognitive performance. No subscales are routinely reported, although memory, language, and praxis items can be examined separately.
The ADAS-Cog is a continuous measure without categorical cutoffs. Average MMSE-equivalent scores at baseline are typically 18–24 in mild Alzheimer's disease, 10–17 in moderate disease, and <10 in severe disease. The instrument is most sensitive to change in the mild-to-moderate range. A change of ≥4 points on ADAS-Cog11 over 6–12 months is considered clinically meaningful decline. Improvement (negative change) of 2–3 points may reflect drug effect or practice effects. Interpretation requires comparison to baseline and reference to population-specific norms.
Common pitfalls
- Using ADAS-Cog11 and ADAS-Cog14 scores interchangeably — the scoring scales differ (0–70 vs. 0–100), and mixing versions across longitudinal assessments introduces error and misinterpretation.
- Over-interpreting small score changes (<4 points over 6 months) as clinically meaningful — practice effects, mood, fatigue, and measurement noise may account for modest fluctuations.
- Administering to patients with severe dementia or those unable to cooperate — scoring becomes unreliable when the patient cannot engage with test items; the instrument is optimized for mild-to-moderate disease.
- Failing to control for examiner effects — variation in task administration, cueing, and scoring across different clinicians can inflate apparent decline. Use the same clinician longitudinally when possible.
Frequently asked questions
Should I use the 11-item or 14-item version of the ADAS-Cog?
The ADAS-Cog11 is the historical standard and most commonly used in longitudinal clinical practice. The ADAS-Cog14 is preferred in pharmaceutical trials because the additional items increase sensitivity to treatment-related cognitive change. If you are comparing to published norms or prior assessments, maintain consistency with the version used previously. Never mix scores across versions in the same patient's longitudinal record.
What is a clinically meaningful change on the ADAS-Cog?
A change of 4 or more points on the ADAS-Cog11 over 6–12 months is considered clinically meaningful. Improvement (negative change) of 2–3 points may reflect drug effect in trials. In routine practice, a change of 3–5 points over 12 months often reflects true disease progression. Always interpret change in context of the baseline score, disease stage, and interval between assessments.
Is ADAS-Cog sensitive enough to detect drug effects in mild cognitive impairment?
The ADAS-Cog has limited sensitivity in pure MCI because scores are near-normal; a ceiling effect limits the range of possible change. The ADAS-Cog is optimized for mild-to-moderate Alzheimer's disease (MMSE equivalent 10–24). For MCI trials, supplementary measures of memory and executive function (e.g., neuropsychological batteries, RBANS) are typically needed alongside ADAS-Cog.
Can the ADAS-Cog be administered remotely by videoconference?
Remote administration is feasible but reduces standardization and scoring reliability, particularly for items requiring precise visual or motor response assessment (e.g., object naming, construction). If remote administration is essential, use video conferencing with high video quality, ensure a quiet environment, and document the method. In-person administration is strongly preferred for research and regulatory trials.
Related methods
Sources
Rosen, W. G., Mohs, R. C., & Davis, K. L. (1984). A new rating scale for Alzheimer's disease. American Journal of Psychiatry, 141(11), 1356-1364.
Open DOI referenceCopied.Mohs, R. C., Knopman, D., Petersen, R. C., et al. (1997). Development of cognitive instruments for use in clinical trials of antidementia drugs: Additions to the ADAS and MMSE. Alzheimer Disease and Associated Disorders, 11(Suppl 2), 13-21.
Open publication referenceCopied.Pfeffer, R. I., Inoue, S. K., & Chance, G. R. (2000). Diagnostic criteria for dementia: Revision of the DSM-IV and ICD-10. Journal of the American Geriatrics Society, 48(12), 1572-1578.
Open publication referenceCopied.