Sickness and sleep health predict frustration and affective responses to a frustrating trigger

Fluctuations in health and sleep are common, but we know surprisingly little about how these daily life stressors affect one's level of frustration and sensitivity to becoming frustrated. In this pre-registered study, 517 participants (Mage = 30.4, SD = 10.4) reported their current sickness symptoms, health status, sleepiness, and sleep duration and quality the previous night. They also rated their general frustration and mood before and after a mild frustration-eliciting task. In the task, participants were instructed to copy geometric shapes onto a piece of paper, without lifting the pen from the paper. Participants were given three minutes to copy the eight shapes, but in order to induce frustration half of them were unsolvable. The study was subsequently repeated in an independent sample (N = 113). Frustration increased in response to the task; however, those with the worst sickness symptoms or sleep health reduced or did not change their frustration levels. Instead, across both studies, frustration was already high at baseline for these individuals. These findings indicate that being sick or having poor sleep is related to high general frustration, but resilience to further frustration due to mild frustrating situations.

Factor loadings based on exploratory component analysis with promax rotation for the 25 sickness-related symptoms.The item "I want to keep still" was eliminated because the item failed to meet the criterium of having a factor loading of >0.3 and did not load onto any of the two factors. 1 = items used to calculate the composite Factor 1 score, 2 = items used to calculate the composite Factor 2; Factor 1 and Factor 2 explained 29.2% and 9.9% of the variance, respectively.
Factor  Table S4.Replication of the baseline and during-task responses in an independent sample (N = 113).Mean ratings (SEM) at baseline and during the Frustration Tolerance Task (the 'task' rating concerned their feeling during the task, rated immediately after the task) rated on 10-point Likert scales from 0 to 9. Higher scores indicate a greater level of the variable (e.g., higher anxiety, higher energy, greater effort used).Table S5.Mean ratings (SEM) at baseline and during the Frustration Tolerance Task (the 'task' rating concerned their feeling during the task, rated immediately after the task), shown separately for 'All participants' (n = 858), 'Max 1 unsolvable solved' (those who reported having solved an unsolvable shape none or one time; n = 517), and '0 unsolvable solved' (those who reported all four unsolvable shapes as unsolved; n = 382).Higher scores indicate a greater level of the variable (e.g., higher frustration, higher anger, higher energy).Results are mostly in the same direction as in the original study.However, due to the smaller sample size and thus lower variation in sickness and sleep health, we did not have enough statistical power to reach statistical significance for all interaction effects.

Sickness Questionnaire
On average, the sickness score of the adapted Sickness Questionnaire was 24.1 (SD = 14.9; range 0-75), calculated as the sum of all the 10 items rated on 10-point scales from 0 to 9.

bFigure S1 .
Figure S1.Replication of the task responses in independent samples.Mean (± SEM) ratings at baseline and during the Frustration Tolerance Task (the during-task ratings concerned their feeling during the task, rated immediately after the task) for (a) N = 517, with baseline ratings rescaled to the same during-task 0-100 VAS, and (b) N = 113, with baseline and during-task ratings measured on a 10-point (0-9) Likert scale.

Figure S3 .Figure S4 .
Figure S3.Negative (a, b, c) and positive (d, e, f) mood (means ± SEM) at baseline and during the Frustration Tolerance Task (the during-task ratings concerned their feeling during the task, rated immediately after the task) separated by sleep health groups (N = 517).Y-axes show the average rating of the mood items (0-100).Higher scores indicate greater negative (a, b, c) or greater positive (d, e, f) mood.Significant interaction effects are indicated with asterisks in the title.Asterisks in the legends indicate significance level for post-hoc task effects (baseline, task) in case a significant stressor x task interaction was evident.# p < .07,* p < .05,** p < .01,*** p < .001.

Figure
FigureS5shows the baseline and during-task frustration ratings separated by individuals

Table S2 .
Original (before rescaling) mean frustration ratings (SEM) at baseline rated on a 10point scale (0-9), separated by low/short, intermediate, and high/long stressor groups.Please note that lower depression-like sickness, infection-like sickness, and sleepiness indicate a better level of the stressor, while lower sleep quality and self-rated health indicate a worse level of the stressor.Short sleep duration refers to <7 hr, intermediate sleep duration to ≥7 to <9 hr, and long sleep duration to ≥9 hr.

Table S3 .
Descriptive characteristics of the study population of the replication study given as sample size (n) and percentage (%) and means (M) and standard deviation (SD).Percentages may not total 100 due to rounding.