Welkom bij THIM Hogeschool voor Fysiotherapie & Bohn Stafleu van Loghum
THIM Hogeschool voor Fysiotherapie heeft ervoor gezorgd dat je Mijn BSL eenvoudig en snel kunt raadplegen. Je kunt je links eenvoudig registreren. Met deze gegevens kun je thuis, of waar ook ter wereld toegang krijgen tot Mijn BSL. Heb je een vraag, neem dan contact op met helpdesk@thim.nl.
Om ook buiten de locaties van THIM, thuis bijvoorbeeld, van Mijn BSL gebruik te kunnen maken, moet je jezelf eenmalig registreren. Dit kan alleen vanaf een computer op een van de locaties van THIM.
Eenmaal geregistreerd kun je thuis of waar ook ter wereld onbeperkt toegang krijgen tot Mijn BSL.
Login
Als u al geregistreerd bent, hoeft u alleen maar in te loggen om onbeperkt toegang te krijgen tot Mijn BSL.
Being affected by a natural disaster or helping victims can impact mental health. This study aimed to examine whether a digital mindfulness intervention positively affects self-compassion, life satisfaction, and positive affect, and alleviates negative affect, perceived stress, and pathological symptoms in victims and volunteers of the flood disaster in Germany in July 2021.
Method
Participants were 146 people affected by the flood (via house, n = 88; friends, n = 100; workplace, n = 54) and/or volunteered after the disaster (n = 86). After the randomized assignment, 74 received a 6-week application-based mindfulness intervention, and 72 were allocated to the wait-control condition. The control group received the same intervention 6 weeks later. Before (T1) and after (T2) the intervention of the first group and after the intervention of the control group (T3), participants filled out questionnaires regarding different aspects of mental health.
Results
We found significant interaction effects of time and intervention for all variables (self-compassion, life satisfaction, positive and negative affect, perceived stress, and pathological symptoms), indicating effectiveness of the digital mindfulness intervention. No deterioration were found between T2 and T3 in the intervention group for any of the variables.
Conclusions
The 6-week digital mindfulness program positively affected the mental health of both disaster victims and volunteers. Therefore, a digital mindfulness intervention could be a good alternative in situations where group programs or face-to-face meetings are infeasible.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
A heavy rain event in western Germany caused devastating flooding and significant destruction in July 2021. In the Valley of the River Ahr, 134 people lost their lives to the flood, and 800 people were injured (Deutsche Presse-Agentur, 2021; Stern, 2022). In addition, four suicides have been linked to the flood disaster (Deutsche Presse-Agentur, 2021). The Supervisory and Service Directorate of Rhineland-Palatinate (Aufsichts- und Dienstleistungsdirektion, 2021) estimated that about 42,000 people were affected. At least 17,000 have lost their belongings or are facing significant damage (Stern, 2022). Investigations into the events of the night of the flood show that the warning and emergency system failed, and most people were on their own (Weidinger, 2021). However, a few days after the night of the flood, parallel to the governmental disaster relief, several tens of thousands of volunteers felt called to help in the disaster area to clean the streets and houses from debris and mud; organize accommodation, rideshares, food, and clothes; and give mental support (Aktuell, 2021). In some regions, 90% of the houses were still unoccupied 4 months after the flood (Muller & King, 2021).
The frequency of extreme natural events and the number of affected people has increased for decades. A significant increase can be observed for natural hazards of hydrological and meteorological origin, such as floods, mass movements, and storm events (McClean et al., 2020).
Flood survivors are at increased risk of post-traumatic stress disorder, anxiety, and depression (Azuma et al., 2014; Chen & Liu, 2015; Dai et al., 2017; Parel & Balamurugan, 2021; Stanke et al., 2012). Furthermore, the examination of secondary stressors by Tempest et al. (2017) and Paranjothy et al. (2011) illustrated that health concerns, relationship problems, and negative impacts on finances were also associated with depression and post-traumatic stress disorder as an indirect result of the floods. Disaster volunteers usually go unseen in such studies. A review by Thormar et al. (2010) pointed out that regardless of the type of disaster compared with professional workers, volunteers tend to have higher complaint levels similar to levels observed in direct survivors in the literature. The fact that such natural disasters increase worldwide and have a considerable impact on mental health demonstrates the relevance of examining an effective mental health intervention for flood survivors and disaster volunteers.
As already mentioned, high prevalence of post-traumatic stress disorder, other psychiatric disorders that affect the quality of life, and increased suicide rates have been found in survivors of natural disasters (Descilo et al., 2010; Golitaleb et al., 2022; Lee & First, 2022) Medical staff were also at risk of PTSD, depression, and anxiety following disasters (Naushad et al., 2019). Mindfulness interventions could reduce stress (Chiesa & Serretti, 2009; Vonderlin et al., 2020), anxiety (e.g., Blanck et al., 2018; Zhou et al., 2020), and depression symptoms (e.g., Blanck et al., 2018; Reangsing et al., 2022) and improve mental well-being (e.g., Lee et al., 2022; Querstret et al., 2020). In addition, meta-analyses (e.g., Hopwood & Schutte, 2017; Sun et al., 2021) found a reduction in post-traumatic stress, so mindfulness practice could help reduce these symptoms in survivors and helpers after natural disasters.
Mindfulness is typically defined as non-judgmental attention to experiences in the present moment (Kabat-Zinn, 1990). It is a naturally occurring cognitive process that increases with regular practice and can be further developed and improved (Bishop et al., 2004). Existing literature shows a variety of possible mechanisms of mindfulness meditation to enhance well-being (Baer et al., 2006; Brown et al., 2007; Hölzel et al., 2011; Shapiro et al., 2006). In the context of disaster survivors and aid workers, the following are discussed: Williston et al. (2021) offered three potential mechanisms: Improved regulation of interoception and arousal may increase the perception and control of internal bodily signals, acceptance could reduce avoidance and reactivity, and attentional flexibility can train the ability to shift the attention away from sad memories intentionally. Non-judgemental acceptance as a facet of mindfulness may also decrease shame in survivors (Panting et al., 2020). Mindfulness can aim at emotional under- and overmodulation (Boyd et al., 2018), and facets of mindfulness (acting with awareness and non-reactivity) were correlated with reduced PTSD symptoms and hyperarousal and emotional numbing (Harper et al., 2022; Stephenson et al., 2017). Breath awareness could reduce physiological arousal (Arch & Craske, 2006). Furthermore, Creswell et al. (2019) suggested that mindfulness interventions train central stress resilience pathways in the brain, which may buffer stress and boost positive emotions.
Cultivating self-compassion, primarily through metta mindfulness (Reilly & Stuyvenberg, 2023) that cultivates unconditional feelings of love, kindness, and acceptance (Salzberg, 1995), could also play a role. Winders et al. (2020) discussed five associations between trauma and self-compassion in a systematic review: High self-compassion was connected to a reduction of persistent internal threats such as shame and self-criticism; an increase in adaptive coping strategies with low levels of avoidance; improved emotion regulation; a strengthening of interpersonal connectedness, which may play a role in accessing social support and increased motivation to seek help; as well as integration of the trauma experience into the pre-existing worldview. Furthermore, there is evidence that higher levels of self-compassion are associated with higher levels of post-traumatic growth (Wong & Yeung, 2017; Yuhan et al., 2021).
Although all these findings suggest that mindfulness-based programs can positively impact those affected by and help in disaster areas, there is little research on post-disaster interventions (Brooks et al., 2018). Three studies examined the effects of mindfulness interventions on mental health workers who helped after a disaster (Waelde et al., 2018; Waelde et al., 2008) and on survivors of disasters (Descilo et al., 2010): A pilot study examining the impact of an 8-week meditation intervention with mental health workers after Hurricane Katrina in the USA found that symptoms of post-traumatic stress disorder and anxiety symptoms decreased significantly (Waelde et al., 2008). Another study analyzed the effects of a 4-hr workshop followed by an 8-week mindfulness intervention at home for counselors and psychologists affected by Typhoon Haiyan in the Philippines (Waelde et al., 2018). The results indicate lower depression severity in participants who spent more time with the mindfulness practice and overall high perceived usefulness and high expectancies that the training would help with survivors and self-care. Positive effects of an 8-hr yoga breath intervention on survivors of the Southeast Asia Tsunami were found for PTSD, depression symptoms, and general health (Descilo et al., 2010).
The present study aimed to investigate the effects of a meditation program for flood victims and volunteer helpers in the Ahr Valley. To this end, we randomized our sample to assign them to one of two groups, which received the intervention at two different periods. Furthermore, participants were asked about various aspects of their mental health (self-compassion, life satisfaction, positive and negative affect, perceived stress, pathological symptoms) at three different time points: T1 (before the intervention), T2 (after the first group completed the intervention and before the second group started it), and T3 (after the second group completed the intervention). Using this approach, our first hypothesis (H1) was to analyze the effect of the intervention (from T1 to T2 for the first and from T2 to T3 for the second group). Moreover, our second hypothesis (H2) was to analyze if this effect lasted even after the intervention was completed (from T2 to T3 for the first group).
Method
Participants
The sample size of cases included 146 subjects affected by the flood (first intervention group, n = 74; second intervention group, n = 72). Overall, 80.1% (n = 117) of participants were female, 19.2% (n = 28) were male, and one person was diverse (0.7%). The average age of the study sample was 43.4 years (SD = 15.3; range = 16–74). Most participants reported being employed (54.1%, n = 79), followed by self-employed (7.5%, n = 11) or retired (11.0%, n = 16). There were 9.6% (n = 14) who reported being in college, 4.1% (n = 6) were civil servants, 8.2% (n = 12) were in training, and 5.5% (n = 8) were unemployed. Fifty-seven participants (31 of the first and 26 of the second group) already had experience with meditation before the intervention, and 88 had no experience. The values of the six study variables did not significantly differ between people with and without meditation experience, with all p > 0.215.
Most participants were directly affected by the flood through damage to their apartment or house (60.3%, n = 88) and/or had family and friends who were affected (68.5%, n = 100); 16.4% (n = 24) were affected via their workplace, 37.0% (n = 54) helped as volunteers several times in the disaster area (multiple answers were possible), and 63.2% (n = 86) reported they had been on site on the night of the flood. To compare flood severity in both groups, a flood impact factor was determined by weighting and adding the severity types. The highest score was given to the apartment/house being affected, and the lowest score was given to regular helping. The flood impact factor resulted in a mean value of 4.64 (SD = 2.19; range = 1–10).
T-tests were performed to evaluate whether the experimental and control groups differed significantly in demographic characteristics. Neither groups differed significantly in age (t(144) = 0.89, p = 0.842, d = 0.03) and flood impact (t(144) = −0.50, p = 0.617, d = 0.08). A chi-square test revealed no significant deviation from an equal distribution between the intervention group and meditation experience (p > 0.05). We calculated a post hoc power analysis (G*Power, Faul et al., 2007) for the lowest significant effect (η2 = 0.04, f = 0.21), α = 0.05, n = 146, two groups, and three measurement points. It revealed a power of 1 − β = 0.999.
Procedure
Participants were recruited via social media (e.g., Instagram and Facebook), email lists from the local university, personal approaches, flyers in the care tents, and other points of contact for flood-affected people and helpers (e.g., trauma care center and shuttle bus stations), and via an appeal in regional newspapers. With the help of the online tool SoSciSurvey (Leiner, 2019), the registrations of the participants were recorded, and the issue and answering of the questionnaires were initiated. Study enrollment started in early December 2021 and ended in mid-January 2022. The study was accessible to people who (1) had their flat, house, or workplace affected by the flood disaster or had friends/relatives who had been affected in the flood areas or had regularly helped in the region; (2) were at least 16 years old and had their parents’ permission or were of legal age; (3) were willing to participate in a 6-week meditation program; (4) had a digital device to access the meditation program; and (5) spoke German.
A total of 183 individuals registered for the study. The 146 participants who answered all the T1 questionnaires were ultimately included in the analysis. For all participants who filled out the questionnaire at T1, we computed t-tests and chi-square tests to detect possible differences between people who completed the intervention and those who dropped out of the sample. Tests were calculated for all demographic variables and the study variables. A significant difference appeared for age and education. Participants who stayed in the sample were older (M = 45.86, SD = 14.30) and more educated (mean rank = 81.40) than those who dropped out (age, M = 39.03, SD = 16.30; education, mean rank = 66.14).
Before the first measurement point (T1), the subjects were randomly assigned to the first intervention group (n = 92) and the waiting control group (n = 91) who received the intervention later. The random assignment sequence was generated using a random number table in Excel. In addition to assessing perceived stress, pathological symptoms, life satisfaction, and self-compassion, information about the study setting was provided at the beginning of the baseline questionnaire, and informed consent was obtained. The baseline questionnaire also included questions about sociodemographics, individual flood impact, and meditation experience. Seventy-four people in the first group and 72 in the second group answered the first questionnaire thoroughly. After answering the first questionnaire, the subjects were told which group they were assigned to and received group-dependent information. In the following 6 weeks, the first group completed a meditation plan. At the second measurement point, 54 subjects from the first and 63 from the second groups participated.
Moreover, the control group was instructed to use the meditation app after the second measurement point. To ensure data protection and avoid bias, only pseudonymized data were analyzed. Adherence was assessed at the end of the 6 weeks. To answer the number of completed sessions, participants could check which courses had been completed in the app. Data from the backend of the meditation program, which could have provided information about the number of meditation sessions per participant, were not transmitted because encrypted extraction and transmission of the data was not possible. The order of administration of the questionnaires used was randomized. A participant flow diagram is shown in Fig. 1. Note that all participants who completed the questionnaires at T1 and were affected by the flood were included in the analyses. Analyses with participants who completed the intervention can be found in the Supplementary Information.
Fig. 1
Study flowchart
×
After completing the baseline questionnaire, the intervention group received all the information needed to set up and use the Balloon app. The 6-week meditation program was divided into the following four courses: “Mindful Living I,” “Strengthening Resilience,” “Being Happy,” and “Taking Care of Yourself.” In total, it comprised 42 sessions. The meditation sessions consist of audio recordings, and each includes a short intro in which an introduction to the topic is given, including psychoeducation or the approach to meditation. A 10–12-min meditation session is then guided. The participants could decide at which time of day they wanted to complete the sessions. In total, the participants of the first intervention group did not complete 8.60 of the 42 sessions, whereas those in the second group did not complete 7.44 sessions. This difference was not significant, t(90) = 0.71, p = 0.480, d = 0.15.
The sessions consisted of both mindfulness meditations and metta meditations. The “Being happy” course consisted of metta meditations, also called loving-kindness meditation, which focuses on people as the object of attention to increase feelings of empathy, love, and care for the self and others (Kok et al., 2013). Focused attention (FA) meditation elements were incorporated by guiding the meditators to focus on the breath, body, sounds, or a question. Later in the program, aspects of open monitoring (OM) meditation were used at the end of the meditations, where participants were asked to release conscious attention to an object and be open to any experience that might arise (Lutz et al., 2008).
A decision was made to deliver the intervention via mobile technology because infrastructure had yet to be fully restored in disaster areas. Face-to-face events would be another significant time commitment for participants, partly because the emergency shelters of those affected were spatially dispersed. In addition, people with limited mobility could participate this way, and the inhibition threshold and social desirability could be lowered, possibly due to greater anonymity. Furthermore, compliance with the applicable legal coronavirus regulations was always ensured. The corona rules during the study period (mid-January to early May 2022) included compliance with distance requirements, vaccination status checks, contact tracing measures, and wearing a medical or FFP2 mask, according to the Twenty-Ninth to Thirty-Third Corona Control Ordinance of Rhineland-Palatinate (Ministerium für Wissenschaft und Gesundheit Rheinland-Pfalz, 2023).
Measures
Self-compassion
Self-compassion was measured using the German short form of the Self-Compassion Scale (SCS-SF) by Hupfeld and Ruffieux (2011), which is based on the English version by Raes et al. (2011). The SCS-D short form measures several facets of self-compassion using 12 items (e.g., “I try to see my failings as part of the human condition.”), with a 5-point answering scale ranging from 1 = almost never to 5 = almost always. The mean values of all 12 items were calculated. Internal consistency of the scale was good (Cronbach’s α = 0.88–0.89, McDonald’s ω = 0.88–0.89).
Perceived Stress
The German version of the Perceived Stress Scale (PSS-10) by Schneider et al. (2020), based on the original English version by S. Cohen et al. (1983), was utilized to evaluate perceived stress. The PSS has 10 items assessing the degree to which situations in the past week were appraised as stressful (e.g., “How often have you been upset because something happened unexpectedly?”). Responses are given on a 5-point rating scale ranging from 1 = never to 5 = very often. A sum score of all 10 items was calculated. The results showed an excellent internal consistency (Cronbach’s α = 0.91–0.91, McDonald’s ω = 0.91–0.92).
Pathological Symptoms
Anxiety, depression, and somatization were measured using the German short version of the Brief Symptom Inventory (BSI-18) (Derogatis, 2001). The short version of the Brief Symptom Inventory includes 18 items, with 6 items assigned to each subscale (Spitzer et al., 2011). Rankings characterize the intensity of distress during the past 7 days. Symptom severity (e.g., pain in the chest, nervousness, and loneliness) was measured on a 5-point rating scale ranging from 0 = not at all to 4 = extremely. The total score was calculated as a sum of all 18 items. It had good internal consistency in the current sample (Cronbach’s α = 0.89–0.92, McDonald’s ω = 0.89–0.93).
Positive and Negative Affect
We measured positive and negative affect using the Positive and Negative Affect Schedule (PANAS). For this purpose, the German version of Breyer and Bluemke (2016) was used. It consists of 20 adjectives describing different sensations and feelings. Ten adjectives capture the independent dimensions of positive (PA) and negative affect (NA). The NA subscale contains words such as “distressed,” “afraid,” and “nervous,” and the PA subscale includes words such as “inspired,” “proud,” and “strong.” Respondents were asked to indicate the extent to which each item describes how they felt in the last week on a scale ranging from 1 = slightly or not at all to 5 = extremely. Mean scores of all 10 items for positive and negative affect, respectively, were calculated. In the present study, Cronbach’s α = 0.89–0.91 and McDonald’s ω = 0.89–0.91 could be determined for positive affect, and Cronbach’s α = 0.80–0.86 and McDonald’s ω = 0.81–0.87 for negative affect.
Life Satisfaction
The short-scale Life Satisfaction (L-1) by Beierlein et al. (2014) was used to assess general life satisfaction. The scale contains only one item with the following wording: “How satisfied are you currently, all things considered, with your life?” The 11 response categories ranged from 1 = not at all satisfied to 11 = completely satisfied. The examination of the quality criteria showed sufficient test-retest reliability (r = 0.67), good convergent, and sufficient discriminant construct validity (Beierlein et al., 2014).
Data Analyses
We used an intent-to-treat analysis with multiple imputations to ensure that participants who dropped out of the treatment and those who finished were equally considered in the study (Kaplan et al., 2022). Of the 149 participants who intended to participate in the study, three were eliminated because they were unaffected by the flood. The six study variables (self-compassion, perceived stress, life satisfaction, positive and negative affect, and pathological symptoms) at T1 showed no missing data, 19.9% of the values at T2 were missing, and an additional 15.7 to 17.2% at T3 (17.2% self-compassion, perceived stress, and pathological symptoms, 16.4% positive and negative affect, 15.7% life satisfaction). In total, 18.8% of the values were missing and imputed. Linear regression was used as an imputation method with ten imputations. The six variables at T1 were used as predictors, and those at T2 and T3 were used as predictors and variables to be imputed. Data were pooled for analyses using SPSS29.
To measure the effectiveness of the intervention on self-compassion, perceived stress, life satisfaction, positive and negative affect, and pathological symptoms (H1), two-way repeated-measures analyses of variance were used, including time (baseline, post-intervention, follow-up) as a within-subject factor and group (first intervention group, second intervention group) as a between-subjects factor. A significant interaction of time and group would indicate effectiveness. Simple effects for the measurement points and conditions were calculated in case of a significant interaction. The intervention can be considered adequate when the first intervention group differs between T1 and T2 and the second intervention group does not differ between T1 and T2 but between T2 and T3. Furthermore, baseline differences between both groups at T1 were analyzed, as well as group differences at T2 and T3. No group differences at T1 and T3, but significant differences at T2, would suggest an effective intervention. To analyze H2, no significant differences between T2 and T3 for the first intervention group would indicate that the effects last after the intervention was completed.
Partial eta squared (ηp2) is reported as an effect size indicator. Statistical significance was set at α < 0.05 for all tests performed. According to the common classifications (Cohen, 1988), f > 0.10 (ηp2 > 0.10) is considered a small effect, f > 0.25 (ηp2 > 0.06) a medium, and f > 0.40 (ηp2 > 0.14) a large effect.
Results
To describe the sample of our intervention study, Table 1 shows the scores of all six measurements for the first and second intervention groups at T1, T2, and T3 and the norm values for the scores. The mean values of self-compassion at T1 in our entire sample were more than one standard deviation lower than those of the validation sample and, therefore, below average. Values of life satisfaction, positive affect, and perceived stress significantly differed from the mean values of the validation sample (all p < 0.001) but were within the average range. Values of negative affect and pathological symptoms were above average. For pathological symptoms, the values of our sample did not differ from the clinical sample in the validation study (Spitzer et al., 2011), t(145) = −0.54, p = 0.590, d = 0.04.
Table 1
Means (M) and standard deviations (SD) for the study variables at T1, T2, and T3 for both groups and norm values
Norm values
First group
Second group
T1
T2
T3
T1
T2
T3
Self-compassion
M
SD
4.01
0.97
2.90
0.63
3.23
0.68
3.25
0.75
2.82
0.74
2.86
0.76
3.34
0.62
Life satisfaction
M
SD
8.18
2.07
6.27
2.37
7.44
2.17
7.85
1.73
6.17
2.51
6.24
2.59
7.83
2.18
Positive affect
M
SD
3.17
0.63
2.82
0.69
3.12
0.71
3.24
0.70
2.82
0.77
2.63
0.75
3.14
0.84
Negative affect
M
SD
1.72
0.57
2.41
0.70
2.23
0.61
1.97
0.63
2.47
0.66
2.62
0.82
1.97
0.57
Perceived stress
M
SD
28.33
6.97
31.21
6.71
27.55
6.94
26.63
6.99
31.02
7.03
32.38
7.37
26.48
6.75
Pathological symptoms
M
SD
3.87
4.64
20.44
12.91
18.35
13.75
14.25
10.47
24.16
14.84
22.69
13.95
14.40
9.69
Correlations of the study variables at T1 are presented in Table 2. Self-compassion was positively correlated with life satisfaction and positive affect and negatively with negative affect, perceived stress, and pathological symptoms. Life satisfaction was also negatively correlated with other negative dimensions of mental health. Similarly, positive affect was negatively correlated with negative affect, perceived stress, and pathological symptoms. The variables of negative affect, perceived stress, and pathological symptoms correlated strongly positively. All correlation coefficients were statistically significant at the 0.01 level.
Table 2
Correlations among the study variables at T1
1
2
3
4
5
6
1 Self-compassion
0.400**
0.490**
−0.488**
−0.646**
−0.426**
2 Life satisfaction
0.508**
−0.542**
−0.590**
−0.524**
3 Positive affect
−0.413**
−0.573**
−0.491**
4 Negative affect
0.671**
0.650**
5 Perceived stress
0.599**
6 Pathological symptoms
**p < 0.01. n = 146
Intervention Effects
For all six variables, ANOVAs were calculated to analyze the effects of time, the interaction of time and condition, and a quadratic effect for the interaction. Furthermore, simple effects were calculated for both groups to analyze the development between T1 and T2 and between T2 and T3, respectively, separately. All results are displayed in Table 3. For all six variables, both groups differed significantly at T2 (all p < 0.031, all η2 > 0.030) but not at T1 and T3.
Table 3
Statistics of the results
Time
Condition × time
First group
Second group
Interaction
Quadratic effect
T1, T2
T2, T3
T1, T2
T2, T3
Self-compassion
F(2, 288) = 58.29, p < 0.001, η2 = 0.29
F(2, 288) = 19.11, p < 0.001, η2 = 0.12
F(1, 144) = 28.50, p < 0.001, η2 = 0.17
pos
-
-
pos
Life satisfaction
F(2, 288) = 44.35, p < 0.001, η2 = 0.24
F(2, 288) = 6.17, p = 0.002, η2 = 0.04
F(1, 144) = 12.46, p < 0.001, η2 = 0.08
pos
pos
-
pos
Positive affect
F(2, 288) = 32.66, p < 0.001, η2 = 0.19
F(2, 288) = 12.50, p < 0.001, η2 = 0.08
F(1, 144) = 24.55, p < 0.001, η2 = 0.15
pos
-
neg
pos
Negative affect
F(2, 288) = 56.01, p < 0.001, η2 = 0.28
F(2, 288) = 10.22, p < 0.001, η2 = 0.07
F(1, 144) = 18.42, p < 0.001, η2 = 0.11
neg
neg
pos
neg
Perceived stress
F(2, 288) = 51.91, p < 0.001, η2 = 0.27
F(2, 288) = 19.19, p < 0.001, η2 = 0.12
F(1, 144) = 58.17, p < 0.001, η2 = 0.29
neg
-
pos
neg
Pathological symptoms
F(2, 288) = 51.51, p < 0.001, η2 = 0.26
F(2, 288) = 6.97, p = 0.001, η2 = 0.05
F(1, 144) = 9.09, p = 0.003, η2 = 0.06
neg
neg
neg
The last four columns describe the development of the values from T1 to T2 and from T2 to T3, respectively. pos, significant increase of the value; neg, significant decrease of the value; -, not significant
Self-compassion
For self-compassion, the ANOVA revealed a significant main effect of time and a significant interaction of time and condition (Fig. 2). Within-subjects contrasts showed a quadratic effect for the interaction. For the first group, simple effects of time were found between T1 and T2, p < 0.001, 95% CI [−0.416, −0.208], and between T1 and T3, p < 0.001, 95% CI [−0.421, −0.208], but not between T2 and T3, p = 0.819. For the second group, simple effects of time were found between T1 and T3, p < 0.001, 95% CI [−0.646, −0.430], and between T2 and T3, p < 0.001, 95% CI [−0.613, −0.372], but not between T1 and T2, p = 0.918. Comparing both groups, significant differences were found for T2, F(1, 144) = 8.38, p = 0.002, η2 = 0.06, 95% CI [0.132, 0.579], but not for T1 and T3.
Fig. 2
Self-compassion between conditions over time. Note. Solid lines, first group (intervention between T1 and T2); dashed lines, second group (intervention between T2 and T3). Asterisks between the solid and dashed lines indicate between-group differences at T2. Asterisks denote significant differences for the first intervention group on the top and the second intervention group on the bottom. Significant differences between T1 and T3 are not shown. **p < 0.01,***p < 0.001
×
Life Satisfaction
For life satisfaction, the main effect of time was significant, and the interaction of time and condition was also significant (Fig. 3). Within-subjects contrasts revealed a quadratic effect of the interaction. For the first group, simple effects of time were found between T1 and T2, p < 0.001, 95% CI [−1.512, −0.511] and between T1 and T3, p < 0.001, 95% CI [−2.003, −1.025], and between T2 and T3, p = 0.031, 95% CI [−0.956, −0.048]. For the second group, simple effects of time were found between T1 and T3, p < 0.001, 95% CI [−2.191, −1.199], and between T2 and T3, p < 0.001, 95% CI [−2.097, −1.176], but not between T1 and T2, p = 0.819. Comparing both groups, significant differences were found for T2, F(1, 144) = 5.24, p = 0.006, η2 = 0.05, 95% CI [0.315, 1.798], but not for T1 and T3.
Fig. 3
Life satisfaction between conditions over time. Note. Solid lines, first group (intervention between T1 and T2); dashed lines, second group (intervention between T2 and T3). Asterisks between the solid and dashed lines indicate between-group differences at T2. Asterisks denote significant differences for the first intervention group on the top and the second intervention group on the bottom. Significant differences between T1 and T3 are not shown. *p < 0.05, **p < 0.01,***p < 0.001
×
Positive Affect
For positive affect, both the main effect of time and the interaction of time and condition were significant (Fig. 4). Within-subjects contrasts showed a quadratic effect for the interaction. For the first group, simple effects of time were found between T1 and T2, p < 0.001, 95% CI [−0.409, −0.143], and between T1 and T3, p < 0.001, 95% CI [−0.547, −0.271], but not between T2 and T3, p = 0.065. For the second group, simple effects of time were found between T1 and T3, p < 0.001, 95% CI [−0.464, −0.184], between T2 and T3, p < 0.001, 95% CI [−0.656, −0.370], and also between T1 and T2, p = 0.006 95% CI [0.055, 0.324]. Comparing both groups, significant differences were found for T2, F(1, 144) = 15.65, p < 0.001, η2 = 0.10, 95% CI [0.231, 0.693], but not for T1and T3.
Fig. 4
Positive affect between conditions over time. Note. Solid lines, first group (intervention between T1 and T2); dashed lines, second group (intervention between T2 and T3). Asterisks between the solid and dashed lines indicate between-group differences at T2. Asterisks denote significant differences for the first intervention group on the top and the second intervention group on the bottom. Significant differences between T1 and T3 are not shown. **p < 0.01, ***p < 0.001.
×
Negative Affect
For negative affect, the ANOVA showed a significant main effect of time and a significant interaction of time and condition (Fig. 5). Within-subjects contrasts revealed a quadratic effect for the interaction. For the first group, simple effects of time were found between T1 and T2, p = 0.015, 95% CI [0.033, 0.302], between T2 and T3, p = 0.001, 95% CI [0.093, 0.365], and between T1 and T3, p < 0.001, 95% CI [0.267, 0.527]. For the second group, simple effects of time were found between T1 and T3, p < 0.001, 95% CI [0.362, 0.625], and between T2 and T3, p < 0.001, 95% CI [0.507, 0.783], and also between T1 and T2, p = 0.030, 95% CI [−0.288, −0.015]. Comparing both groups, significant differences were found for T2, F(1, 144) = 11.11, p = 0.001, η2 = 0.07, 95% CI [−0.604, −0.154], but not for T1 and T3.
Fig. 5
Negative affect between conditions over time. Note. Solid lines, first group (intervention between T1 and T2); dashed lines, second group (intervention between T2 and T3). Asterisks between the solid and dashed lines indicate between-group differences at T2. Asterisks denote significant differences for the first intervention group on the bottom and the second intervention group on the top. Significant differences between T1 and T3 are not shown. *p < 0.05, **p < 0.01, ***p < 0.001
×
Perceived Stress
For perceived stress, the main effect of time and the interaction of time and condition were significant (Fig. 6). Within-subjects contrasts revealed a quadratic effect for the interaction. For the first group, simple effects of time were significant between T1 and T2, p < 0.001, 95% CI [2.316, 4.459], and between T1 and T3, p < 0.001, 95% CI [2.747, 5.740], but not between T2 and T3, p = 0.184. For the second group, simple effects of time were significant between T1 and T3, p < 0.001, 95% CI [3.282, 6.316], and between T2 and T3, p < 0.001, 95% CI [4.812, 7.378], and also between T1 and T2, p = 0.020, 95% CI [−2.382, −0.210]. Comparing both groups, significant differences were found for T2, F(1, 144) =15.85, p < 0.001, η2 = 0.10, 95% CI [−6.727, −2.263] but not for T1 and T3.
Fig. 6
Perceived stress between conditions over time. Note. Solid lines, first group (intervention between T1 and T2); dashed lines, second group (intervention between T2 and T3). Asterisks between the solid and dashed lines indicate between-group differences at T2. Asterisks denote significant differences for the first intervention group on the bottom and the second intervention group on the top. Significant differences between T1 and T3 are not shown. *p < 0.05,***p < 0.001
×
Pathological Symptoms
For pathological symptoms, the ANOVA revealed a significant main effect of time and a significant interaction of time and condition (Fig. 7). Within-subjects contrasts showed a quadratic effect for the interaction. For the first group, simple effects of time were found between T1 and T2, p = 0.019, 95% CI [0.377, 4.082], and between T1 and T3, p < 0.001, 95% CI [2.892, 7.711], and also between T2 and T3, p = 0.004, 95% CI [0.986, 5.158]. For the second group, simple effects of time were found between T1 and T3, p < 0.001, 95% CI [7.031, 11.917], between T2 and T3, p < 0.001, 95% CI [6.436, 10.666], but not between T1 and T2, p = 0.333. Comparing both groups, significant differences were found for T2, F(1, 144) = 5.20, p = 0.024, η2 = 0.04, 95% CI [−9.383, −0.607], but not for T1 and T3.
Fig. 7
Pathological symptoms between conditions over time. Note. Solid lines, first group (intervention between T1 and T2); dashed lines, second group (intervention between T2 and T3). Asterisks between the solid and dashed lines indicate between-group differences at T2. Asterisks denote significant differences for the first intervention group on the bottom and the second intervention group on the top. Significant differences between T1 and T3 are not shown. *p < 0.05, **p < 0.01, ***p < 0.001
×
Discussion
First, our analyses showed that the mean values in our sample were above the average for negative affect and pathological symptoms and below the average for self-compassion. For life satisfaction, positive affect, and perceived stress, values were within the average range. This shows that the people in our sample were still affected by the natural disaster in some aspects of their mental health, even though the intervention started about 6 months after the flood.
The combination of significant interaction effects between time and condition and significant quadratic effects of those interactions for all six variables showed the success of the meditation intervention by increasing self-compassion, life satisfaction, and positive affect and decreasing negative affect, perceived stress, and pathological symptoms. Even considering the waitlist control group, we cannot be sure that the effects were only caused by the intervention.
An interaction effect of time and condition was also revealed for all variables. There was a moderate effect size for the interaction of condition and time on stress, self-compassion, positive and negative affect, and a small effect for life satisfaction and pathological symptoms. The interaction of time and condition hints toward different effects of time for both intervention groups. These interactions showed quadratic effects: large for perceived stress, positive affect, and self-compassion and moderate for life satisfaction, negative affect, and pathological symptoms. A quadratic effect of the interaction of time and condition illustrates that for self-compassion, life satisfaction, and positive affect, the values of the first intervention group from T1 to T2 increased more than the second intervention group and from T2 to T3 the values increased more for the second intervention group. This effect was reversed for pathological symptoms, negative affect, and perceived stress. The results indicated the positive effect of the meditation intervention, as the intervention group differed significantly from the waiting group at the second measurement point so that positive effects could be achieved within the intervention that could not be achieved by natural recovery of the waiting control group that did not yet receive the intervention at T2. Furthermore, simple effects for self-compassion, life satisfaction, positive and negative affect, perceived stress, and pathological symptoms appeared at the measurement points when the intervention took place (for the first group between T1 and T2, for the second between T2 and T3, but not between T1 and T2). No significant simple effects for all considered variables except pathological symptoms, life satisfaction, and negative affect between T2 and T3 in the first group showed that the interventional gains were maintained at the follow-up. Pathological symptoms and negative affect even decreased significantly from T2 to T3 in the first group, and life satisfaction increased furthermore between T2 and T3. Non-significant group differences in all variables at T1 demonstrated that there were no structural differences between the groups before the intervention. Non-significant group differences at T3 showed that the time of the intervention approx. 6 or 8 months after the disaster made no difference in its positive effects. Significant differences at T2 again illustrated the effectiveness of the intervention.
The main findings of our study suggest that mindfulness meditation could benefit flood victims and disaster volunteers. Therefore, government funding of such programs in natural disaster areas should be discussed to expand psychotherapy support. These programs could serve people waiting for a place in therapy, reluctant to undergo psychotherapy, or wanting anonymous help.
Our results align with other studies that show positive effects of online meditation programs and apps on self-compassion (Huberty et al., 2019), perceived stress (Schulte-Frankenfeld & Trautwein, 2021), anxiety (Duraimani, 2019; Huberty et al., 2019), emotional well-being (Athanas et al., 2019; Mahlo & Windsor, 2021), and life satisfaction (Mahlo & Windsor, 2021). Most recently, a pilot study by Sarsar et al. (2021) with a sample of firefighters responsible for transporting and treating COVID-19 patients showed that completing a 10-min daily meditation reduced anxiety symptoms and negative affect. A meta-analysis also supports these findings: Eisenstadt et al. (2021) found a reduction in mental health symptoms (i.e., anxiety, depression, stress) and an improvement in well-being (i.e., psychological, subjective, and emotional). Significant effects for stress, anxiety, depression, and psychological well-being, but not for general well-being, were found in the meta-analysis by Gál et al. (2021).
The current study did not explicitly address the mechanisms by which mindfulness meditation might improve mental health. As previously stated, reappraisal could be an essential factor. The large effect of self-compassion, closely related to cognitive reappraisal, supports this hypothesis. Further investigation of the mode of action of mindfulness mechanisms as a post-disaster intervention is needed.
The overall positive results achieved by the meditation program in the first intervention group compared to the control condition raises the question of whether the application effectiveness is also comparable to traditional forms such as stress reduction programs. Thus, comparing multiple groups implementing other intervention programs would be informative for future studies. This could include investigating the best time to begin a meditation intervention for those affected by natural disasters and answering the question of the extent to which natural disaster type influences effectiveness.
Limitations and Directions for Future Research
The sample consisted primarily of people open to a meditation intervention and had some prior interest and possibly positive expectations of mindfulness programs. Additionally, nearly 80% of the participants were female, which may limit generalizability. There were also few participants in the 25–30 age range. The reason for this could be the high level of employment at this age, which prevents posters and flyers from being read. To counteract this, in subsequent studies, study participation could be advertised in those facilities that this age group often frequents. Non-monetary or monetary incentives for the subjects would also increase motivation to participate in the study. It also remains unclear if participants’ mental health problems were related to the flood.
Furthermore, the subject of the study was not concealed, so participants could have looked through it and unconsciously answered the questionnaires according to the expected results or social desirability. Moreover, we did not control for multiple testing. This increases the risk of type I error. The dropout rates were in the normal range. However, it should not remain unmentioned that due to this also, some negative experiences with the meditation intervention could not be considered in the results. It should also be mentioned that the completers in this study differed from those who did not complete the intervention regarding age and education. Completers were, on average, 6 years older and had higher levels of education. This could affect the generalizability of the results to younger and less educated people. Finally, it can be seen as a limitation of our study that we used a single-item scale to measure life satisfaction. In further studies, a more detailed instrument should be used for a more valid and reliable measurement of life satisfaction. It should also be mentioned that the study was not blinded. This could have affected the interpretation of the results because participants could have given the answer they thought was wanted, especially when they liked the app.
Acknowledgements
We would especially like to thank all participants who took the time to participate in the study despite their difficult life situations. We wish them much strength for the future.
Declarations
Ethics Statement
The study protocol was approved by the Institute of Psychology at the University of Koblenz (21122021—LEK-379).
Informed Consent Statement
All participants gave written informed consent.
Conflict of Interest
The authors declare no competing interests.
Use of Artificial Intelligence
Grammarly has been used to correct grammatical issues.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Met BSL Psychologie Totaal blijf je als professional steeds op de hoogte van de nieuwste ontwikkelingen binnen jouw vak. Met het online abonnement heb je toegang tot een groot aantal boeken, protocollen, vaktijdschriften en e-learnings op het gebied van psychologie en psychiatrie. Zo kun je op je gemak en wanneer het jou het beste uitkomt verdiepen in jouw vakgebied.
