Can someone hear that you’re depressed even if you say you’re fine?
You are sitting in the exam room at the doctor’s office when the doctor comes in.
“How are you doing?”
Your heart starts racing. How are you supposed to respond? What’s the protocol? Are you supposed to say “fine” now and then later share that you’ve been feeling down?
Or is now the time you are supposed to share that you are feeling down? Is that rude? You haven’t asked your doctor how they are doing yet. When do you move past the “fine” “fine” conversation and actually share how you are feeling?
Because you aren’t fine. You are struggling.
What if your voice could say more than your words?
You’ve probably heard that non-verbal communication is 80-93% of communication. While there are debates about the actual percentage, what is certain is that non-verbal communication is a valuable and impactful part of communication. Your voice, facial expressions, and body language all communicate.
Think back to the last time someone said, “I’m not angry” in a way that showed they were most definitely angry. Their increased rate of speech, change in pitch and volume, scrunched up face, and hands balled into fists all communicated something different than their words.
Our brains can pick up some of these cues, like the obvious ones I mentioned. But some changes to your voice are so subtle they aren’t picked up or consciously registered. How might people’s lives be different if we could identify more people with depression before they reach a breaking point? What if we got so good at identifying depression earlier that people stopped experiencing severe depression? How much struggle would be eliminated? How many lives could we save?
We’re able to generate data from speech, like we could analyze blood for information about someone’s health. With advances in technology and machine learning, this technology can be used outside the lab and without someone who has expertise in voice analysis.
Voice biometric technology turns our voices into a vital sign that can be used to detect signs of mental health, neurodegenerative, cardiovascular, and respiratory disorders through language (what you say) and acoustic (how your voice sounds) analysis.
This technology could listen to your “fine” “fine” and “the weather sure is awful/wonderful” conversation with your doctor and detect signs of disease, including depression from your voice.
How well does this work?
Studies that look at voice biometric technology often compare it to the nine-item Patient Health Questionnaire (PHQ-9), which is used to assess whether someone has depression, which has an accuracy of 85%. Voice biometric technology’s accuracy is lower than PHQ-9 according to a recent literature review (Donaghy, 2024), but there may be some models that have accuracy similar to or above the PHQ-9 according to another review (Briganti, 2025).
While most studies in the reviews looking at the accuracy of voice biometric technology recommend that the models continue to improve their accuracy, voice biometric assessment has some advantages that may make it worth considering now.
Because voice biometric assessment can be done without a specific task, it doesn’t take additional time to complete the screening. This also means that it can easily be done more and data can easily be compared to previous assessments, building out a profile so changes over time will become clearer.
Where could voice biometric technology listen to you?
Let’s go back to the exam room where you said you were “fine” and hoped for the courage and the opportunity later to say, “Actually, I’ve been struggling and I want some help.” With voice biometric technology, your doctor could see from the assessment data that you might be experiencing depression, which prompts a conversation that could lead to you getting help at a time when it can be hard to ask.
You might also experience voice biometric assessment in a virtual visit with a clinician, when you call to schedule an appointment with a therapist or when you call your insurance company. Those opportunities could also prompt the person you are speaking with to refer you to the support you need.
Your therapist or psychiatrist might use the technology to assess your mental health at each visit, which will help them in submitting a claim to insurance or adjusting your treatment plan.
Wellness apps might also include some voice assessment as well. Since wellness apps are not a medical device that is regulated for providing mental health treatment, you should not rely on wellness apps to provide mental health treatment.
Is your voice going to be assessed without your knowledge now?
No. It shouldn’t be a surprise. You should be aware when your voice is assessed.
Voice biometric technology can be integrated into just about any place where your voice is, but just like any medical treatment, before you participate, your informed consent should be given. You should be informed of and have an opportunity to ask questions about the purpose, duration, and methods of data collection and retention of the assessment.
In addition to needing informed consent, several states have passed laws or proposed legislation on biometric information privacy to secure your biometric data.
I’m curious, would you try it? Where do you think voice biometric technology will be in 5 years?
If you or someone you know is struggling with depression or thoughts of self-harm, help is available. In the U.S., you can call or text the Suicide & Crisis Lifeline at 988 for immediate support, or text HOME to 741741 to connect with a trained counselor. For resources outside the U.S., visit findahelpline.com to find support in your country.
Take aways
- Your voice is a vital sign that can be used to detect signs of mental health, neurodegenerative, cardiovascular, and respiratory disease.
- Voice biometric technology may not be as accurate as the PHQ-9 yet, but it provides conveniences that others don’t.
- Before your voice is assessed, you should provide informed consent to grant your permission.
References
Briganti, G., & Lechien, J. R. (2025). Speech and Voice Quality as Digital Biomarkers in Depression: A Systematic Review. Journal of Voice: Official Journal of the Voice Foundation, S0892-1997(25)00187-0. https://doi.org/10.1016/j.jvoice.2025.05.002
Donaghy, P., Ennis, E., Mulvenna, M., Bond, R., Kennedy, N., McTear, M., O’Connell, H., Blaylock, N., & Brueckner, R. (2024). A Review of Studies Using Machine Learning to Detect Voice Biomarkers for Depression. Journal of Technology in Behavioral Science. https://doi.org/10.1007/s41347-024-00454-2
Is Biometric Information Protected by Privacy Laws? (2024) In Bloomberg Law. https://pro.bloomberglaw.com/insights/privacy/biometric-data-privacy-laws/#state
Mazur, A., Costantino, H., Tom, P., Wilson, M. P., & Thompson, R. G. (2025). Evaluation of an AI-Based Voice Biomarker Tool to Detect Signals Consistent With Moderate to Severe Depression. The Annals of Family Medicine, 23(1), 60–65. https://doi.org/10.1370/afm.240091
PHQ-9. (2025). In Wikipedia. https://en.wikipedia.org/w/index.php?title=PHQ-9&oldid=1305790608
