Diagnostic Tool for Alzheimer’s Shows High Accuracy

Alzheimer's disease
Source: istock

Researchers in Sweden have developed a new diagnostic tool for Alzheimer’s that can predict the disease with 90% accuracy.

Alzheimer’s is an irreversible neurological disorder that accounts for 60-70% of dementia cases worldwide. The progressive brain disease results in lapses in memory, cognitive impairment, problem performing routine tasks such as bathing, and changes in behaviour. People are mostly affected after the ages of 65 years. Diagnosis is based on a complete history, tests for memory and cognition, and brain scans such as MRI or CT. Often doctors may measure levels of disease-specific proteins in spinal fluid samples or using PET scans. However, these are expensive and not readily available. Moreover, 20-30% of patients receive the wrong diagnosis resulting in a delay in treatment and increased risk of death. Therefore, there is an urgent need for a better diagnostic tool for Alzheimer’s disease.

Led by Professor Oskar Hansson, researchers at Lund University have developed an algorithm that combines results of blood tests and memory tests to predict whether one will develop Alzheimer’s. They tested their algorithm in a group of over 800 patients from across Sweden and North America. The results are published in the journal Nature Communications.

Our algorithm is based on a blood analysis of phosphylated tau and a risk gene for Alzheimer’s, combined with testing of memory and executive function. We have now developed a prototype online tool to estimate the individual risk of a person with mild memory complaints developing Alzheimer’s dementia within four years

Professor Sebastian Palmqvist, study author

The researchers used a simple blood test that measures level of tau protein and a risk gene for Alzheimer’s. An abnormal buildup of tau protein within the brain is believed to be responsible for causing the disease. Additionally, the researchers also conducted three brief memory tests.

Early Diagnosis Equals Better Survival

First, the team examined 340 patients in Sweden with mild memory impairment. They compared findings of the new algorithm with older diagnostic tools: brain scans and medical history. Aim of the diagnostic tool was to predict who would develop Alzheimer’s disease within four years.

According to the results, older methods accurately identified patients who would develop Alzheimer’s 72% of the time. However, the new algorithm accurately predicted the disease 83% of the time when using only the data from blood tests. But when the algorithm used a combination of both blood tests and memory tests, the accuracy went over 90%.

They then further confirmed their results by conducting it in 543 patients in North America.

The algorithm will enable us to recruit people with Alzheimer’s at an early stage, which is when new drugs have a better chance of slowing the course of the disease

Professor Oskar Hansson, lead researcher

The team has developed the algorithm for use in primary care settings and low-income countries with limited resources. Moreover, their research will likely lead to early diagnosis; thus, allowing patients to get appropriate care before the progression of the disease.


Palmqvist, S., Tideman, P., Cullen, N. et al. Prediction of future Alzheimer’s disease dementia using plasma phospho-tau combined with other accessible measures. Nat Med (2021). https://doi.org/10.1038/s41591-021-01348-z


Please enter your comment!
Please enter your name here