Altered speech patterns in subjects with post-traumatic headache due to mild traumatic brain injury

Catherine D. Chong, Jianwei Zhang, Jing Li, Teresa Wu, Gina Dumkrieger, Simona Nikolova, Katherine Ross, Gabriela Stegmann, Julie Liss, Todd J. Schwedt, Suren Jayasuriya, Visar BerishaThe Journal of Headache and Pain 2021 · DOI

Abstract

Compared to healthy controls (HC), subjects with acute post-traumatic headache (PTH) demonstrate altered speech as measured by objective features of speech production. For individuals with PTH, speech production may have been more effortful, resulting in slower speaking rates and more precise vowel articulation during headache vs. when they were headache-free, suggesting that speech alterations were related to PTH and not solely due to the underlying mild traumatic brain injury (mTBI).

Introduction

Mild traumatic brain injury (mTBI), or concussion, frequently results in post-traumatic headaches (PTH) that closely resemble migraines. While patients often report "brain fog" or difficulty speaking during these episodes, clinical assessment usually relies on subjective self-reporting, which is prone to bias and recall errors.

There is a significant gap in objective, longitudinal tools to measure the "burden" of these headaches. This study investigates whether speech—a complex neuromuscular and cognitive task—changes predictably during a headache, potentially serving as a non-invasive digital biomarker for tracking recovery or treatment efficacy in mTBI patients.

Methods

The researchers enrolled 19 subjects with PTH (within 59 days of an mTBI) and 31 healthy controls. Over 12 weeks, participants used a mobile app every three days to complete sentence reading and spontaneous speech tasks. The team extracted several key acoustic features, including temporal features (speaking and pause rates), pitch, and articulation precision (measured via "Goodness of Pronunciation" scores).

Overview of the feature extraction and normalization pipeline, showing the processing of sentence reading and spontaneous speaking tasks.

To account for age and sex differences, the researchers normalized the features using a subset of the Mozilla Common Voice database. They then used a mixed-effects model to analyze both between-group differences and within-subject changes (headache vs. headache-free days).

Results

PTH vs. Healthy Controls

Regardless of whether a headache was present, PTH subjects showed significantly longer pause rates ($p=0.0098$) and reduced consonant ($p=0.0015$) and vowel ($p=0.0368$) precision compared to healthy controls. This suggests that the underlying mTBI creates a "baseline" deficit that persists even outside of acute headache episodes.

Example of a speech waveform used to calculate the spontaneous pause rate by identifying detectable pause periods within the total speaking time.

The Impact of Active Headaches

On days with active headaches, PTH subjects spoke significantly slower ($p < 0.0001$) compared to their own headache-free days. Interestingly, vowel articulation precision was significantly higher ($p = 0.0052$) during headaches. The authors hypothesize an "effortful speech" paradox: because speaking feels more difficult during a headache, patients exert more conscious cognitive and motor effort to enunciate, leading to slower but "sharper" vowels.

Discussion and Future Work

While vowel precision increased during headaches, consonant precision decreased. Consonants often require faster, more complex motor coordination, which may be more susceptible to the degradation caused by the headache compared to the more sustained sounds of vowels.

The study's primary limitation is the small sample size (19 PTH subjects), which prevented some complex statistical models from converging. Future research aims to use these speech markers to predict which patients are at high risk for developing chronic, persistent post-traumatic headaches, potentially allowing for earlier and more aggressive clinical intervention.

Summary updated

Abstract

Compared to HC, subjects with acute PTH demonstrate altered speech as measured by objective features of speech production. For individuals with PTH, speech production may have been more effortful resulting in slower speaking rates and more precise vowel articulation during headache vs. when they were headache-free, suggesting that speech alterations were related to PTH and not solely due to the underlying mTBI.
BibTeX
@article{chong2021altered,
  author = {Catherine D. Chong and Jianwei Zhang and Jing Li and Teresa Wu and Gina Dumkrieger and Simona Nikolova and Katherine Ross and Gabriela Stegmann and Julie Liss and Todd J. Schwedt and Suren Jayasuriya and Visar Berisha},
  title = {{Altered speech patterns in subjects with post-traumatic headache due to mild traumatic brain injury}},
  journal = {The Journal of Headache and Pain},
  year = {2021},
  doi = {10.1186/s10194-021-01296-6},
}
Figure