Stroop vs. Rest

Goal: Assess blood pressure differences between rest and the Stroop task and whether this difference is different in OSA than healthy controls (HCs).

The Stroop task is a color-word naming task consists of 3 separate parts:

Naming colors of non-word stimuli (rectangles)
Reading color words (congruent)
Naming the color of a word that describes a color but is printed in a different color. This condition is an incongruent set of stimuli and requires inhibition of our automatic mental reading abilities. Therefore, it requires more time (i.e. response time) to name the colors.

Each block had a maximum duration of 270 seconds or 4.5 minutes.

A Stroop effect or score can be computed from the accuracy (number of mistakes) and time to read through the list. Thereby, timing and accuracy of all 3 conditions are considered.

In this figure, the first 2 conditions are reversed and the control list is X’s instead of rectangles.

Recording of the exact timing and accuracy of each word, however, is difficult in the laboratory. The instructions are to name the colors as fast as possible without making mistakes. Different people will use different strategies and trade off one vs. the other. They also self-correct a lot when they say the wrong color. In our case, there was a set time for each block. If participants finished the list before the block ended, they could manually discontinue by pressing a button. The block will then be shown as ended in Acqknowledge. However, if the participant took longer than the provided time, they were cut off in the middle of the naming list. We are not interested in the exact cognitive performance (i.e. time and accuracy), but the physiological reactions to this task, as it is cognitively demanding and therefore more “stressful”.

Data

OSA 2019: OSA vs. CONTROL

OSA2013: OSA vs. CONTROL

OSA2010: OSA vs. CONTROL

Signals

Blood pressure
Respiration as a control to verify the timing of signals 1; i.e. while we don’t have exact timing of the onset of the spoken words, the respiration signal tells us roughly when the person inhaled before speaking.
The cardiac signal makes less sense to investigate, since it’s highly affected by the word speaking.

Analysis steps Acqknowledge

Copy raw data into shared folder → copy to local folder
Separate the Stroop task and save as a separate file (keep about 90 seconds before and after Stroop block wherever possible)
Create focus areas for each task/file:
1. Color naming
2. Word reading
3. Incongruent word-color naming
Run blood pressure signal detection and manually correct markers for each focus area
1. Export excel spreadsheet with blood pressure data: systolic and diastolic
Run respiration peak detection and manually correct markers
1. Find cycle inspiration-to-inspiration to get breath-to-breath breathing rate
2. Find cycle inspiration-to-expiration for inspiration time
3. Find cycle expiration-to-inspiration for expiration time

Excel calculations

Create excel spreadsheet for each task and subject that contains:

Systolic blood pressure
Diastolic blood pressure
Breathing rate
Inspiration time
Expiration time

Matlab analysis

Import excel spreadsheets into Matlab

Research questions

Does blood pressure differ between rest, congruent and incongruent conditions of the Stroop task?
Do OSA patients respond differently to the Stroop than healthy controls?
Can the differences be explained by the breathing/speaking rate of the Stroop task?

Analyses

Repeated measures ANOVA:

Within-subject factor: condition
1. Word reading
2. Color naming
3. Incongruent
4. Rest period
Between-subject factor:
1. OSA vs. healthy control (HC)
Dependent variables:
1. Breathing rates
2. Inspiratory duty cycle (inhale/(inhale+exhale))
3. Systolic blood pressure
4. Diastolic blood pressure
5. Mean blood pressure

Correlations:

Breathing rate per condition X blood pressure per condition
Breathing rate difference (congruent - incongruent) X blood pressure (congruent - incongruent)
Are the differences in correlations between OSA and healthy controls significant?