Word recognition tracks amyloid where the audiogram does not

Design

cross-sectional study with amyloid positron emission tomography and multivariable linear regression adjusted for covariates

Population

223 late middle-aged adults, mean age 65.6 years (SD 3.8), mean better-ear pure-tone average 21.0 dB

Primary outcome

whole-brain and regional amyloid standardised uptake value ratio

Effect

per 10% fall in word recognition, global SUVR higher by β = 0.064 (P = 0.005) with consistent regional effects; pure-tone average significant only in left temporal region (0.024 per 10 dB, P = 0.048)

The association between hearing loss and dementia risk is well established and its mechanism is not. A study of 223 late middle-aged adults — mean age 65.6 years, mean pure-tone average 21.0 dB, so mostly mild loss — measured cerebral amyloid on positron emission tomography and asked which hearing measure it tracked.

The two measures behaved differently. Pure-tone average reached statistical significance in only one region, the left temporal lobe (amyloid standardised uptake ratio difference 0.024 per 10 dB, P = 0.048), though the direction and magnitude were similar across all regions. Word recognition score was associated with higher amyloid much more broadly: per 10% fall in word recognition, standardised uptake ratio was higher in left temporal (β = 0.053), right temporal (0.055), left parietal (0.068), right parietal (0.076), left cingulate (0.076), right cingulate (0.088) and right frontal (0.052) regions, and globally (0.064, P = 0.005). The authors note the associations appeared driven by participants at the extremes of both hearing loss and amyloid burden.

This is cross-sectional, the effect sizes are small, and nothing here says treating hearing loss changes amyloid. What it does suggest is that the speech discrimination score — which many audiology reports still treat as supplementary to the audiogram — may carry central information the pure-tone thresholds do not. In a patient whose word recognition is disproportionately poor for their thresholds, that discrepancy is worth noticing rather than attributing to test conditions.

• Report and act on word recognition score, not only the pure-tone average.
• Note a word recognition score that is out of proportion to the thresholds rather than dismissing it.
• Do not tell patients that hearing aids prevent amyloid deposition; nothing here supports that.
• Consider cognitive screening where speech discrimination is disproportionately impaired.
• Remember the cohort had mostly mild loss — this is not a finding about profound deafness.

Why it matters

It suggests the speech discrimination score, not the audiogram, is the hearing measure that tracks central change.

Don't overread it

Cross-sectional and small in effect — this cannot say whether hearing loss contributes to amyloid, results from it, or shares a cause with it.

The statistics, in plain English

These are small associations: a standardised uptake ratio difference of 0.064 globally per 10% fall in word recognition is a fraction of the range that separates amyloid-positive from amyloid-negative scans, so no individual's scan is being predicted here. The pure-tone result at P = 0.048 in one of several regions tested is exactly the kind of finding that multiple comparisons produce by chance, and the authors are careful to say the direction was consistent elsewhere rather than claiming regional specificity. That the association appeared concentrated in participants at the extremes means it may not hold across the ordinary range of hearing loss at all.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.