Normative voice data for younger and older adults

Objectives: We constructed age- and gender-specific norms for healthy adults on a large number of speech and voice measures using standardized recording procedures.

Study design: Participants were 159 younger (mean = 19.1 years, standard deviation [SD] = 1.4) and 133 older (mean = 72.0 years, SD = 4.8) healthy native English male and female speakers who did not currently smoke and had typical hearing for their age group.

Methods: Participants phonated the vowel [a] under various instructions and read an abbreviated version of the Rainbow Passage. Voice measures based on the productions of [a] included fundamental frequency (F(0)), jitter, shimmer, harmonics-to-noise ratio, noise-to-harmonics ratio, maximum phonation time, minimum phonation intensity, maximum pitch, and the Dysphonia Severity Index (DSI). Mean speaking fundamental frequency (SF(0)), SF(0) standard deviation (SF(0)SD), and signal intensity were measured from the reading passage.

Results: Noteworthy age-related differences were found for males and females. Older females had a lower F0 and SF(0) and smaller SF(0)SD than younger females, but younger and older males did not differ. Shimmer increased with age for males, but neither jitter nor shimmer increased with age for females, whereas noise measures were similar for both ages. Younger and older males had a similar DSI, whereas older females had a higher DSI than younger females.

Conclusions: This study provides a unique database containing a wide variety of voice measures collected from a large sample of adults in good health using standardized recording procedures. Males and females differed on the type and extent of age-related vocal changes, but overall age-related differences were limited.