Isometric handgrip exercise speeds working memory responses in younger and older adults

Analysis and manuscript code for: Bachman, S.L., Attanti, S., Mather, M. (2022). Isometric handgrip exercise speeds working memory responses in younger and older adults.

Data for this project are available on Open Science Framework.

For questions, contact sbachman@alumni.usc.edu.

contents

• functions/: MATLAB functions used for data preprocessing
• handgrip-WM_analysis.Rmd: main statistical analysis document that calls relevant scripts
• scripts/: scripts used for data preprocessing and statistical analysis

dependencies

• MATLAB, including the Statistics and Machine Learning Toolbox
• R
• R packages needed for preprocessing (run scripts/0_install-scripts.R to install)
• (R packages needed for statistical analysis are installed when knitting handgrip-WM_analysis.Rmd)
• PhysioNet Cardiovascular Signal Toolbox (link; for compatibility with the scripts, install into a subdirectory toolboxes)

Preprocessing and analysis for Bachman et al. (2022) were performed using R Version 4.0.4, MATLAB Version R2021b, using a linux kernel with version 5.13.0-7614. See preprint for the specific package versions that were used.

instructions for reproducing analyses in Bachman et al., 2022

prepare project directory

• clone this directory
• initialize an RStudio project in the project directory
• create a subdirectory named data in the project directory

download data

• download the dataset from Open Science Framework, using the "Download as zip" option to download all files in rawdata/ as a zipped directory
• unzip rawdata.zip so that rawdata is a subdirectory of data

compile behavioral data

• in R, run scripts/1_compile-behavior/A_compile-nback.R to compile n-back task data

preprocess ECG & EMG data

• in MATLAB, run scripts/2_process-physio/A_find_beats_ECG.m to identify r-peaks in ECG data

  • beforehand: update dir_root to location of the project directory on your machine

• in R, run scripts/2_process-physio/B_analyze-niHR-HRV.R to generate heart rate time series

  • beforehand: update line 22 with the name of the directory in data/derivatives containing physionet output (you should just need to update the date in the directory name)

• in MATLAB, run scripts/2_process-physio/C_analyze_neuECG.m to apply the neuECG method to ECG data

  • beforehand: update dir_root to location of the project directory on your machine

• in R, run scripts/2_process-physio/D_summarize-neuECG.R to aggregate neuECG results

  • beforehand: update line 18 with the name of the directory in data/derivatives containing neuECG output (you should just need to update the date in the directory name)

• in R, run scripts/2_process-physio/E_summarize-EMG.R to aggregate EMG metrics

preprocess eyetracking data

• in R, run scripts/3_process-et/A_preprocess-et-data_baseline.R to preprocess baseline eyetracking data
• in R, run scripts/3_process-et/B_preprocess-et-data_handgrip.R to preprocess handgrip eyetracking data
• in R, run scripts/3_process-et/C_preprocess-et-data_nback.R to preprocess n-back eyetracking data

statistical analysis for manuscript

• update the file scripts/4_analysis/01_load-data.R with correct names of results files you created in the steps above
• knit the document handgrip-WM_analysis.Rmd to html