Edited by:
Reviewed by:
*Correspondence:
This article was submitted to Emotion Science, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Previous studies have shown changes in empathy in patients with depression, including an elevated level of trait personal distress. This study examined if low mood causes changes in self-reported empathic distress when seeing others in pain. To test this, we conducted an initial (
Empathy is an important aspect of social interactions as it promotes prosocial behaviors (
Mood induction is used in the current study to experimentally change healthy participants’ mood temporarily, in order to examine the effect of induced sadness on empathic personal distress. Mood induction has been used previously to study the effect of mood instead of examining clinically depressed participants, as it allows for better control of confounding variables, such as presence of comorbid disorders and use of medications (
Mood induction methods have been used in previous studies examining the effect of mood on first-hand pain. It appears that positive affect decreases pain perception, whereas negative affect may increase pain perception, although the later effect seems to be less consistent across studies (
In survey studies, trait empathy has been commonly examined using the self-report questionnaire Interpersonal Reactivity Index (IRI;
Behavioral studies have examined state empathy in depression, focusing on cognitive empathy, or cognitive evaluations of others’ emotional states, and they report conflicting results.
Empathy has also been an area of interest in neuroscience, particularly focusing on empathy for pain (e.g.,
In summary, the available questionnaire studies demonstrate that depression is accompanied by a heightened personal distress and altered cingulate activity at seeing others’ suffering. However, it is unclear if there is a direct and casual relationship between depression and exaggerated empathic distress. Also, there is relatively limited behavioral research on state emotional empathy in low mood and depression that involves an indicator of the observer’s own experience. As a first step in better understanding the relationship between mood state and empathic distress, the current study examines causal relationships between low mood and heightened empathic distress at seeing others’ pain in healthy participants. As explained above, mood induction in healthy participants could assist us in focusing specifically on the low mood component that is also prominent in depression, without the other potential confounding variables such as symptom complexity and medication effects. To the best of our knowledge, this is the first study to examine if there is a causal relationship between low mood and heightened personal distress.
To conclude, the current study aimed to establish a causal relationship between sad mood and increased personal distress at seeing another’s pain in healthy participants. It was hypothesized that there would be a larger difference between distress ratings for painful videos relative to non-painful videos under sad mood compared to neutral mood.
Caucasian volunteers without any self-reported current or previous psychiatric or neurological issues were recruited through online advertisements. Only Caucasian volunteers were recruited to limit any potential racial effect when viewing the video stimuli of Caucasian actors (
The study was conducted in two parts, comprising an initial study and a follow-up close replication study in independent groups of healthy participants. In the first study, 26 participants passed the screening and completed the experimental session (14 males; age
This study was reviewed and approved by the Behavioural and Social Sciences Ethical Review Committee at the University of Queensland and was conducted with written informed consent from all participants.
There were two sessions to the study. In the first, participants completed a screening process to ensure suitability, which involved completing the DASS (
In session 2, each participant received two mood inductions, a sad mood induction and a neutral mood induction (
During the 6-min mood induction, participants read statements on a computer display while listening to mood-coherent classical music through headphones. For the sad mood condition, participants listened to Prokofiev’s “Russia Under the Mongolian Yoke” at half speed while reading statements such as “There is no hope,” “I am tired of trying,” and “Life is such a heavy burden.” For the control condition, which was a neutral mood condition, Dvorak’s “Symphony from the New World” was played at its usual speed while participants saw statements such as “It snows in Scotland,” “An orange is a citrus fruit,” and “The Pacific Ocean has fish.” Interested readers could email the authors for a copy of the statements used in the mood inductions. Participants rated their mood prior to and after each mood induction on a scale from 0 = not at all to 10 = extremely for two items: “At this moment I feel SAD” and “At this moment I feel HAPPY.” To help maintain the induced mood, the mood-congruent music continued to play after the induction throughout the following empathy task. Participants rated their mood (on both the sadness and happiness questions) again on the 11-point scale after the empathy task, which was also the end of that mood condition.
For the empathy task, immediately after completing each of the mood inductions, participants watched videos of Caucasian actors receiving painful (with a syringe needle) and non-painful (with a cotton-tip) touches on the cheek. A total of 32 videos were shown, portraying all combinations of eight actors (four males and four females) receiving a painful or non-painful touch to the left or the right cheek of their face (3-s videos, 9 s inter-stimulus interval). To limit emotional contagion, the video clips were cut at the moment that the needle or the cotton-tip made contact with the cheek, so that participants did not observe emotional expressions of the actors in response to the touches. Participants were asked to rate how distressing they found watching each of the videos, by pressing one of the five keys on a keyboard (1 = not distressing at all to 5 = very distressing). There were a total of four blocks of this empathy task, two for each mood condition. A 6-min break time was scheduled after two blocks, during which the participant was told to relax and listen to a new piece of music. A positive piece of classical music, Delibes’ “Coppelia” (chosen based on
The analyses were carried out using SPSS 24 and the focus was on the participant’s responses to the mood induction paradigm, and ratings for the video stimuli on the empathy task.
A sad mood composite score was calculated using the formula {[(10-happy)+sad]/2} (
A mean distress rating was calculated from the participants’ scores on the five-point scale for each of the two types of videos (painful or non-painful) under each mood condition (sad or neutral). For participants with missing responses, the mean distress rating was calculated based on the available responses. Two participants had more than 10% responses missing due to technical issues during testing (14 and 27% responses missing, respectively), and their mean responses were included in the analysis.
Overall, the distress rating data were found to be highly skewed, due to low scores repeatedly given to the cotton-tip videos (more than 40% of the mean ratings given to the cotton-tip videos were ‘1’ or not distressing at all). Therefore, non-parametric methods were utilized to analyze the distress ratings. It should be noted that our initial analyses on the first sample of participants were conducted using repeated-measures ANOVA; however, on inspection of the skewness of the data (and following comments from an external referee), we decided on more conservative non-parametric analyses for the distress rating data across both the initial and replication studies.
For non-parametric analysis of the distress ratings, therefore, a Friedman Test was used first to examine whether the ratings were significantly different between the four conditions (painful in sad condition, non-painful in sad condition, and painful and non-painful in neutral condition). Follow up pairwise comparisons between conditions were then conducted using Wilcoxon Signed Rank Tests. Effect sizes were estimated with r using the formula of the absolute value of z divided by the square root of n (number of observations;
To further examine differences in distress ratings for painful relative to non-painful videos between the two mood conditions, a difference score was calculated between the distress rating given to the painful videos and that given to the non-painful videos, under each of the two mood conditions (i.e., pain minus non-pain for sad, compared with pain minus non-pain for neutral). These empathic distress difference scores were compared between mood conditions using a Wilcoxon Signed Rank Test. As consent for data sharing was not obtained from participants for this study, the individual data is not made publically available.
Participants’ responses to the mood induction, or their sad composite scores, were affected by the type of mood induction,
Participant sample one’s self-reported mood (composite score from 0 to 10, 0 = not sad at all and extremely happy; 10 = not happy at all and extremely sad) before and after two types of mood induction (sad or neutral), and at the end of the empathy task (bars shown are standard errors).
Importantly, the sad composite score was significantly greater both following the sad mood induction (
Most importantly, comparing neutral and sad mood induction conditions, participants’ sad composite scores were not significantly different before mood induction, but were significantly higher for the sad induction both immediately after mood induction and after the empathy task (both
Friedman test indicated that the rankings of the four distress ratings (given to the two types of videos under two mood conditions) varied significantly,
Participant sample one’s distress ratings on a scale from 1 to 5 (1 = not distressing at all, 5 = very distressing) for two types of stimuli (painful needles or non-painful cotton-tips) in the two mood conditions (bars shown are standard errors).
Follow-up pairwise comparisons with the Wilcoxon Signed Rank test and a Bonferroni adjusted α of 0.013 indicated that the needle videos were rated as more distressing under sad mood (
To further examine if the change in distress rating (painful vs. non-painful touch) was different across the two mood conditions, two difference scores were calculated for the distress rating given to the needle videos minus that given to the cotton-tip, under each mood condition. These difference scores represent the participant’s level of empathic distress, i.e., how much more distressing the participant experienced the needle videos relative to the cotton-tip videos. These empathic distress scores were marginally higher under sad mood compared with neutral mood, according to Wilcoxon Signed Rank test,
Participants’ responses to the mood induction, or their sad composite scores, were affected by the type of mood induction,
Participant sample two’s self-reported mood (composite score from 0 to 10, 0 = not sad at all and extremely happy; 10 = not happy at all and extremely sad) before and after two types of mood induction (sad or neutral), and at the end of the empathy task (bars shown are standard errors).
Importantly, the sad composite score was significantly greater both following the sad mood induction (
Most importantly, comparing neutral and sad mood induction conditions, participants’ sad composite scores were not significantly different before mood induction, but were significantly higher for the sad induction both immediately after mood induction and after the empathy task (both
Friedman test indicated that the rankings of the four distress ratings (given to the two types of videos under two mood conditions) varied significantly,
Participant sample two’s distress ratings on a scale from 1 to 5 (1 = not distressing at all, 5 = very distressing) for two types of stimuli (painful needles or non-painful cotton-tips) in the two mood conditions (bars shown are standard errors).
Follow-up pairwise comparisons with the Wilcoxon Signed Rank test and a Bonferroni adjusted α of 0.013 were conducted. As expected, the needle videos were rated as more distressing than the cotton-tip videos under both mood conditions (sad mood,
To further examine if the change in distress rating (painful vs. non-painful touch) was different across the two mood conditions, two difference scores were calculated for the distress rating given to the needle videos minus that given to the cotton-tip, under each mood condition. These difference scores represent the participant’s level of empathic distress, i.e., how much more distressing the participant experienced the needle videos relative to the cotton-tip videos. Crucially, these empathic distress scores were significantly higher under sad mood compared with neutral mood, according to Wilcoxon Signed Rank test,
A
Extending previous survey findings of a relationship between low mood and empathy (
It has been proposed that higher empathic distress contributes to guilt, which in turn makes depression more likely or contributes to its maintenance (
It may be noted that the difference between the distress ratings for painful videos relative to non-painful videos under sad mood compared to neutral mood was statistically significant in the second sample, but was not significant in the first sample. This is likely to be due to the limited power associated with the sample size of the first group. The first group did show a trend for the difference score to be larger in the sad condition and, importantly, the effect sizes for this difference were similar in both the first and second samples (
Also, it should be noted that the participants’ distress ratings given to the cotton-tip videos also increased following the sad mood induction, although the size of this effect was smaller than for the painful needle videos. One possibility is that this may reflect a heightened sensitivity toward these types of personal touch stimuli in general under low mood. While designed to appear neutral, and clearly rated by participants as less distressing than the needle-touch videos, the cotton-tip touch as depicted in the videos was not particularly pleasant or pleasurable and may therefore have been perceived as slightly more unpleasant or distressing for participants under low mood. Crucially, however, even when controlling for this difference in “baseline” distress ratings to cotton-tip videos with mood (i.e., by examining the relative difference between painful and non-painful touch), the increase in distress ratings seen for needle versus cotton-tip videos was significantly greater under sad mood than neutral mood.
A potential limitation of the current study was the playing of music during the empathy task, which is an adaptation of the original mood induction paradigm by
In summary, this study found a causal relationship between low mood and empathic distress when viewing other’s pain in healthy participants. This may also have important implications for people with depression, as low mood is one of the key features of depression, and heightened personal distress at seeing others’ pain may result in avoidance and social withdrawal. Further research on this topic, such as the underlying neural activity changes with low mood in response to seeing other’s pain, may help us better understand empathic and emotional processing under low mood, which have implications for better understanding of empathy in depression.
YC, RC, and GD had substantial contributions to the design of the work. YC completed the data acquisition and analysis. YC, GC, and RC contributed to the selection of statistical analysis and/or interpretation of the results. YC drafted the paper and GC, RC, and GD revised it critically. All authors approved this final version to be published.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.