Speech changes in old age: Methodological considerations for speech-based discrimination of healthy ageing and Alzheimer's disease.
Olga Ivanova, Israel Martínez-Nicolás, Juan José García Meilán
PMID 37140204WHAT IT FOUND
Speech timing and pausing distinguish Alzheimer's from healthy ageing better than other speech features.
Tasks with higher cognitive load, like recalling memories, reveal these differences more clearly. Automated analysis shows promise for screening, but it cannot diagnose Alzheimer's on its own.
Key findings
01Temporal speech parameters, such as pause duration and speech rate, are the strongest predictors of cognitive impairment and distinguish Alzheimer's from healthy ageing more effectively than acoustic or prosodic features.
02Elicitation tasks with higher cognitive load, particularly those involving memory recall, are more sensitive for classifying clinical groups and detecting pathological ageing.
03Automated speech analysis can serve as a non-invasive screening tool to discriminate between healthy and pathological ageing, but it is not intended to diagnose Alzheimer's disease.
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The review included only 24 studies, and many had small sample sizes. There is a significant lack of data on how speech changes vary across different languages, limiting generalizability. Most automated analysis tools were trained on small, specific datasets and have not been widely validated in clinical settings. The studies primarily focused on discrimination between groups, not on diagnosing individual patients.
Declared interests
The authors declare no conflicts of interest.
The easy way to misread this
Do not use speech analysis alone to diagnose Alzheimer's disease. The review emphasizes that speech markers are useful for screening and discriminating groups, but diagnosis requires confirmation through neuropsychological testing and clinical evaluation. Additionally, the accuracy of automated tools may not generalize to patients who speak languages or dialects not represented in the training data.