Can listeners assess men's self-reported health from their voice?

Men's voices may provide cues to overall condition; however, little research has assessed whether health status is reliably associated with perceivable voice parameters. In Study 1, we investigated whether listeners could classify voices belonging to men with either relatively lower or higher self-reported health. Participants rated voices for speaker health, disease likelihood, illness frequency, and symptom severity, as well as attractiveness (women only) and dominance (men only). Listeners' were mostly unable to judge the health of male speakers from their voices; however, men rated the voices of men with better self-reported health as sounding more dominant. In Study 2, we tested whether men's vocal parameters (fundamental frequency mean and variation, apparent vocal tract length, and harmonics-to-noise ratio) and aspects of their self-reported health predicted listeners' health and disease resistance ratings of those voices. Speakers' fundamental frequency (f_o) negatively predicted ratings of health. However, speakers' self-reported health did not predict ratings of health made by listeners. In Study 3, we investigated whether separately manipulating two sexually dimorphic vocal parameters—f_o and apparent vocal tract length (VTL)—affected listeners' health ratings. Listeners rated men's voices with lower f_o (but not VTL) as healthier, supporting findings from Study 2. Women rated voices with lower f_o and VTL as more attractive, and men rated them as more dominant. Thus, while both VTL and f_o affect dominance and attractiveness judgments, only f_o appears to affect health judgments. Results of the above studies suggest that, although listeners assign higher health ratings to speakers with more masculine f_o, these ratings may not be accurate at tracking speakers' self-rated health.