Sense of control buffers against stress
- Division of Psychology and Language Sciences, United Kingdom
- Department of Experimental Psychology, University of Oxford, United Kingdom
- Anna Freud, United Kingdom
- Yale Child Study Center, Yale School of Medicine, Yale University, United States
- Applied Computational Psychiatry Lab, Mental Health Neuroscience Department, Division of Psychiatry and Max-Planck UCL Centre for Computational Psychiatry and Ageing Research, Queen Square Institute of Neurology, United Kingdom
- North London NHS Foundation Trust, United Kingdom
Figures
Figure 1 with 1 supplement
Linear model predicting sense of control from WS task parameters.
(a) Parameter estimates across all participants who completed the WS task (n = 674). These were significantly different from 0 (one-sampled t-tests, FDR-corrected p values <0.001, see text for full statistical details). (b) Correlation coefficients between predicted and actual control ratings for both studies (Study 1 median = 0.54, Study 2 median = 0.69). Boxplots show the median and 25th and 75th percentiles. Whiskers extend to 1.5*IQR (inter-quartile range) from the quartiles. (c) Control rating responses over the time course of the experiment predicted from the model (red) plotted against the actual ratings (blue) for three randomly selected participants per study, with the correlation coefficient (r) per participant shown in the top right of each subplot.
Figure 1—figure supplement 1
Mean BIC (across all participants and both studies) for each of the 7 models with different regressors.
The bar shows the mean BIC (also added in text) for each model across all participants (n = 674). Error bars represent the standard deviation. The models are ordered in descending order of mean BIC left to right. The model furthest right was selected as the ‘winning’ model as it has the lowest BIC, although this was not significantly lower than the previous model. p values are from paired t-tests comparing BIC values between models in descending order.
Figure 2 with 2 supplements
Negative association between subjective control and subjective stress during the Wheel Stopping task for both studies.
Points represent raw data (Study 1: n = 473, 4 timepoints; Study 2: n = 201, 3 timepoints) and lines represent the estimated relationship from the linear mixed effects models in Table 1 (Study 1) and Supplementary file 3A (Study 2), showing the relationship between subjective control and subjective stress (Study 1: β=–0.13, p<0.001; Study 2: β=–0.33, p<0.001), after accounting for perceived task difficulty and random effects of participant and timepoint. Shaded regions represent 95% confidence intervals.
Figure 2—figure supplement 1
Associations between mean subjective control and questionnaire measures in (a) Study 1, and (b) Study 2.
Data points are the data per participant (mean values), and the line represents the estimated relationship from the linear mixed effects models in Supplementary files 2A and B.
Figure 2—figure supplement 2
Associations between mean subjective stress and questionnaire measures in (a) Study 1, and (b) Study 2.
Data points are the data per participant (mean values), and the line represents the estimated relationship from the linear mixed effects models in Supplementary files 2E and F.
Figure 3
Stress induction and stress relief in Study 2.
(a) Subjective stress ratings across the entire experiment (Study 2, n = 295) in the different experimental conditions. The point represents the mean per group, and the error bar represents standard error of the mean. Timepoints labelled WS 3, After Stressor, and After Stressor Debrief are the three timepoints isolating the stress induction and stress debrief (coded as timepoints 1, 2, 3). Jitter added to avoid overlap. (b) Stress Induction – the change in subjective stress from before to after the stressor. Data points show the difference between timepoints per participant. The black points show the mean estimate of the contrast between the two timepoints from the linear mixed effects models in Table 2 (with 95% confidence intervals as error bars). The comparison is the difference between these contrasts, showing that the stress induction was lower for the high control group than for the neutral control group (β=–7.78, SE = 3.07, t(291)=-2.54, p=0.012). (c) Stress Relief – the change in subjective stress from after the stressor to after the stressor debrief. Data points show the difference between timepoints per participant. The black points show the mean estimate of the contrast between the two timepoints from the linear mixed effects models in Table 2 (with 95% confidence intervals as error bars). The comparison is the difference between these contrasts, showing that the stress relief was greater for the high control group than for the neutral control group (β=–6.06, SE = 2.64, t(291)=-2.30, p=0.022).
Figure 4
Summary of task procedure for Studies 1 and 2.
The arrow shows the overall procedure with a simplified schematic of two Wheel Stopping task blocks. Participants also rated subjective stress levels on similar slider rating scales (not shown here) during the WS/Video tasks.
Tables
Table 1
The unique contributions of subjective control and perceived task difficulty on subjective stress during the Wheel Stopping task.
| Subjective Stress(Study 1) | ||
|---|---|---|
| Predictors | Estimates (95% CI) | p |
| (Intercept) | 31.13 (25.15–37.12) | <0.001 |
| Subjective Control | –0.13 (−0.20–-0.07) | <0.001 |
| Perceived Difficulty | 0.38 (0.31–0.44) | <0.001 |
| Random Effects | ||
| σ2 | 208.61 | |
| τ00 | 346.20 ppt | |
| 3.63 timepoint | ||
| ICC | 0.63 | |
| N | 473 ppt | |
| 4 timepoint | ||
| Observations | 1892 | |
| Marginal R2 /Conditional R2 | 0.154/0.684 | |
Table 2
Linear mixed effects models predicting subjective stress from two timepoints: before and after the stressor (stress induction, timepoints 1 and 2), and after the stressor and after the stressor debrief (stress relief, timepoints 2 and 3).
| Stress Induction | Stress Relief | |||
|---|---|---|---|---|
| Predictors | Estimates (95% CI) | p | Estimates (95% CI) | p |
| (Intercept) | 12.99 (3.57–22.41) | 0.007 | 114.30 (102.29–126.31) | <0.001 |
| Timepoint | 20.14 (15.30–24.99) | <0.001 | –29.91 (−34.07–-25.75) | <0.001 |
| Control [Neutral] | –12.35 (−28.19–3.49) | 0.126 | –11.02 (−31.60–9.56) | 0.293 |
| Stressor Intensity [Low] | 12.89 (0.57–25.20) | 0.040 | –56.19 (−72.35–-40.02) | <0.001 |
| Domain [Loss] | 10.64 (3.73–17.54) | 0.003 | 8.19 (1.28–15.10) | 0.020 |
| Timepoint ×Control [Neutral] | 10.92 (2.38–19.46) | 0.012 | 9.65 (2.32–16.99) | 0.010 |
| Timepoint ×Stressor Intensity [Low] | –14.66 (−21.46–-7.86) | <0.001 | 19.87 (14.03–25.71) | <0.001 |
| Control [Neutral]×Stressor Intensity [Low] | 11.56 (−10.26–33.38) | 0.298 | 13.40 (−15.23–42.03) | 0.358 |
| (Timepoint ×Control [Neutral])×Stressor Intensity [Low] | –6.28 (−18.32–5.77) | 0.307 | –7.19 (−17.54–3.16) | 0.173 |
| Random Effects | ||||
| σ2 | 301.09 | 222.45 | ||
| τ00 | 470.40 ppt | 510.27 ppt | ||
| ICC | 0.61 | 0.70 | ||
| N | 295 ppt | 295 ppt | ||
| Observations | 590 | 590 | ||
| Marginal R2 /Conditional R2 | 0.131/0.661 | 0.162/0.745 | ||
Table 3
Overview of the two studies.
| Study 1 | Study 2 | |
|---|---|---|
| N participants | 473 | 295 |
| N Conditions | 4 | 6 |
| WS Task Control Conditions | High, Low | High, Neutral (videos) |
| Stressor Intensity Conditions | High, Low | High, Low |
| Domain Conditions | Win | Win, Loss |
| Procedure Summary | Stressor then Wheel Stopping task | Mild stressor, Wheel Stopping task, then stressor |
| Questionnaire Measures Collected | STAI, PHQ, SPIN, LOC | STAI, PHQ, SPIN, LOC |
Table 4
Demographic information for the two study iterations.
| Study 1 | Study 2 | |
|---|---|---|
| Age – mean (SD) | 30.2 (8.18) | 28.6 (4.84) |
| Female – n (%) | 241 (51.2) | 148 (50.3) |
| Nationality UK – n (%) | 377 (80.0) | 239 (81.3) |
| First language English – n (%) | 399 (84.7) | 255 (86.7) |
| Ethnicity – n (%) Asian Black Mixed White Other | - | 30 (10.20) 24 (8.19) 12 (4.08) 219 (74.49) 6 (2.05) |
-
Notes: Contains missing demographic data for some participants. Missing: age from 5 participants (Study 1 n=3, Study 2 n=2), sex from 5 participants (Study 1 n=3, Study 2 n=2), nationality from 10 participants (Study 1 n=7, Study 2 n=3), first language from 10 participants (Study 1 n=8, Study 2 n=2), ethnicity from all Study 1 and from Study 2 n=4.
Additional files
-
Supplementary file 1
ICC results using a 1st/2nd half split of the data.
- https://cdn.elifesciences.org/articles/105025/elife-105025-supp1-v1.docx
-
Supplementary file 2
Associations between questionnaires and task based measures.
(A) Associations between questionnaire scores and mean task-level subjective control in Study 1, with WS control condition included as a covariate in the linear model. Adjusted p values (padj.) are FDR corrected p values given we ran five different models. (B) Associations between questionnaire scores and mean task-level subjective control in Study 2. The control condition was not included as a covariate in the linear model because the WS task was only presented in High Control. Adjusted p values (padj.) are FDR corrected p values given we ran five different models. (C) Associations between questionnaire scores and estimated intercept parameter from the computational model predicting control from WS task parameters in Study 1, with WS task control condition included as a covariate in the linear model. Adjusted p values (padj.) are FDR corrected p values given we ran five different models. (D) Associations between questionnaire scores and estimated intercept parameter from the computational model predicting control from WS task parameters in Study 2. Adjusted p values (padj.) are FDR corrected p values given we ran five different models. (E) Associations between questionnaire scores and mean task-level stress ratings, with external stressor intensity condition included as a covariate in the linear model for Study 1. Adjusted p values (padj.) are FDR corrected p values given we ran five different models. (F) Associations between questionnaire scores and mean task-level stress ratings, with external stressor intensity condition included as a covariate in the linear model for Study 2. Adjusted p values (padj.) are FDR corrected p values given we ran five models.
- https://cdn.elifesciences.org/articles/105025/elife-105025-supp2-v1.docx
-
Supplementary file 3
Additional analyses testing the association between subjective control and subjective stress for both studies.
(A) Relationship between subjective control, perceived difficulty and subjective stress during the WS Task in Study 2, also when removing the final WS timepoint and including Domain, or when including win rate. Predicted values from the leftmost column (Subjective Stress) model are presented in Figure 2. (B) Excluding the final timepoint to investigate the effects of control, difficulty and stress during the WS Task for Study 1 (left-hand model). Including all timepoints (as original model) and additionally including overall win rate as a covariate for Study 1 (right hand model).
- https://cdn.elifesciences.org/articles/105025/elife-105025-supp3-v1.docx
-
Supplementary file 4
Sensitivity and exploratory analyses for stress induction and stress relief.
(A) Linear mixed effects model for the stress relief when including initial stress level (after the WS/video task) as a covariate, predicting subjective stress from two timepoints: after the stressor and after the stressor debrief (timepoints 2 and 3). (B) Linear mixed effects models including total experiment time as a covariate, predicting subjective stress from two timepoints: before and after the stressor (stress induction, timepoints 1 and 2), and after the stressor and after the stressor debrief (stress relief, timepoints 2 and 3). (C) Linear mixed effects models including the interactions with Domain (rather than just as a covariate in the main analyses), predicting subjective stress from two timepoints: before and after the stressor (stress induction, timepoints 1 and 2), and after the stressor and after the stressor debrief (stress relief, timepoints 2 and 3). (D) Linear mixed effects models predicting subjective stress from two timepoints: before and after the stressor (stress induction, timepoints 1 and 2), and after the stressor and after the stressor debrief (stress relief, timepoints 2 and 3) in just the high control (WS task) condition, to test for the interactions with domain.
- https://cdn.elifesciences.org/articles/105025/elife-105025-supp4-v1.docx
-
Supplementary file 5
Additional methodological details.
(A) Descriptive statistics across the 4 conditions from Study 1. (B) Descriptive statistics across the 6 conditions in Study 2. Given that the Study 2 analyses compared group differences, we assessed group differences in demographic and questionnaire measures using a one-way ANOVA for continuous variables or a Chi-squared test for categorical variables. (C) Additional information about excluded participants. (D) Methodological details for both studies.
- https://cdn.elifesciences.org/articles/105025/elife-105025-supp5-v1.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/105025/elife-105025-mdarchecklist1-v1.docx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Sense of control buffers against stress
eLife 14:RP105025.
https://doi.org/10.7554/eLife.105025.3
