Negative affect reliably predicts paranoid thoughts. Previous studies point to the importance of emotion regulation for paranoid thoughts but have not yet focused on effects of acute increases and variability in strategy use.
Methods
We conducted an experience-sampling study for one week in a subclinical sample. Acute increases in the intensity of strategy use from one measurement point until the next, between-strategy variability (i.e., standard deviation between all strategies at one measurement point), and within-strategy variability (i.e., standard deviation of each strategy over one day) were analyzed for effects on negative affect and paranoid thoughts.
Results
Multi-level-models indicated that acute increases in acceptance and reappraisal are associated with less negative affect. Acute increases in acceptance, but not in reappraisal, were related to less paranoid thoughts. In contrast, acute increases in rumination and suppression were associated with more negative affect and paranoid thoughts. Between- and within-strategy variability were no significant predictors.
Conclusions
Acute increases in the intensity of but not variability in emotion regulation strategies are related to negative affect and paranoid thoughts in daily life. Future studies are needed to examine whether improving emotion regulation leads to sustainable reductions in symptoms.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
Paranoid thoughts are defined as distressing beliefs that a persecutor is intentionally harming or going to harm the individual in the future (Freeman & Garety, 2000). They are linked to a reduced quality of life (Watson et al., 2018) and lower psychological well-being (Freeman et al., 2014). In accordance with the continuum of psychotic experiences (Linscott & Van Os, 2013), paranoid thoughts are also common in non-clinical groups (Freeman et al., 2019) though attenuated positive psychotic symptoms predict the transition to psychosis in individuals at clinical high risk (for a meta-analysis see Oliver et al., 2020). To prevent manifest paranoid symptoms, it is thus important to understand the development of paranoid thoughts. One factor, which has been consistently found to predict paranoid thoughts, is negative affect. This is evident in clinical samples (e.g., Ben-Zeev et al., 2011; Krkovic et al., 2020; Ludwig, et al., 2019a) but also in samples across the paranoia continuum (e.g., Kramer et al., 2014; Krkovic et al., 2020; Lincoln et al., 2009; Thewissen et al., 2011). Therefore, a better understanding of how negative affect can be effectively regulated could be crucial to preventing manifest disorders.
To determine the relevance of emotion regulation strategies for psychotic symptoms such as paranoia, researchers have mainly been looking at how frequently different types of strategies are used and how this is related to negative affect and paranoid thoughts. Cross-sectional studies in general population samples found that suppression and rumination, commonly classified as dysfunctional strategies (Aldao et al., 2010), were associated with more paranoid thoughts (Grezellschak et al., 2017; Simpson et al., 2012), whereas reappraisal and acceptance, commonly classified as functional strategies, were associated with less paranoid thoughts (Osborne et al., 2017; Perchtold et al., 2019). Next to questionnaire studies, other studies have made use of experimental designs and experience sampling methods (ESM).
A recent meta-analysis of questionnaire studies examining the frequency of strategies showed that individuals with psychosis typically report to use more dysfunctional and less functional strategies than healthy controls (Ludwig et al., 2019b). In contrast to the consistent picture presented by these studies, experimental studies that commonly induce affect along with the instruction to use a certain emotion regulation strategy tend to find that individuals with psychosis are as successful as healthy controls in effectively down-regulating their negative affect (Grezellschak et al., 2015; Opoka et al., 2020, 2021; Perry et al., 2012; Van Meer et al., 2014). Putative reasons for the discrepancy between the findings in questionnaire and experimental studies are a retrospective recall bias in the questionnaire studies and the artificialness of the instructions in the experimental design and the induced emotion, which questions their ecological validity. Studies using an ESM are advantageous in this regard because they can assess both emotion regulation and psychotic symptoms at the point of their occurrence, which minimizes recall biases and enables to analyze predictive relationships. Moreover, ESM was already shown to be highly accepted and feasible in clinical samples (for reviews see Menon et al., 2017; Myin-Germeys et al., 2016, 2018; Naslund et al., 2015). Existing ESM studies observed that individuals with psychotic disorders tend to use emotion regulation strategies more frequently than healthy controls, but yielded mixed results for the effectiveness of these specific strategies in terms of reducing negative affect (Ludwig et al., 2020; Strauss et al., 2019; Visser et al., 2018). Beyond that, ESM studies focusing on paranoid thoughts as the dependent variable found that the intensity of the dysfunctional strategies suppression (Nittel et al., 2018) and rumination (Hartley et al., 2014) predicted the subsequent occurrence of paranoid symptoms in clinical samples. However, only one study with a time-lagged design included functional strategies and did not find functional strategies to predict a change in paranoid thoughts (Nittel et al., 2018). To sum up, although the ESM is ideal to study the questions of interest, results are still inconsistent concerning the question of which strategies are effective in regulating negative affect as well as which directly impact on paranoid thoughts. This may be due to the fact that previous studies focused on the absolute intensity of emotion regulation strategies at a given point in time (Hartley et al., 2014; Ludwig et al., 2020; Strauss et al., 2019; Visser et al., 2018) or the deviation in intensity from the personal average (Nittel et al., 2018) without evaluating temporal changes or the onset of strategy use. However, for an effective regulation, being able to change the intensity of used strategies from one moment to another, to prioritize between different strategies, and to vary strategies across time is also crucial.
It is increasingly discussed that not only the overreliance on dysfunctional and the neglect of functional strategies, but also a rigid and inflexible pattern of regulation could have a negative impact on behavioral and emotional functioning (Coleman & Oliveros, 2020; Kobylińska & Kusev, 2019). Because the functionality of emotion regulation strategies also depends on whether they are implemented in response to the context (Aldao & Nolen-Hoeksema, 2012; Bonanno & Burton, 2013; Gross, 2015; Ma et al., 2018), purely considering the average amount of the use of single strategies is likely to be overly simplistic. To better capture moments in which the intensity of a strategy is acutely increased, it may be advantageous to focus on time points at which the employment of a strategy is changed compared to the moment before. For an adaptive response to a changing environment, it is also necessary to prioritize single strategies over others at one emotional episode (i.e., between-strategy variability), and to vary the intensity of a particular strategy across time (i.e., within-strategy variability; Aldao et al., 2015; Blanke et al., 2020). By using data from various ESM studies in non-clinical samples, Blanke et al. (2020) recently found that higher between-strategy variability was linked to lower levels of negative affect suggesting that prioritizing strategies may be effective for regulating affect. Similarly, higher within-strategy variability was associated with lower levels of negative affect. Until now, no studies have addressed the role of acute increases in the intensity of emotion regulation strategies and their variability in relation to paranoia. Examining these characteristics of emotion regulation could help to gain a consistent picture of the role of specific strategies for the regulation of negative affect and paranoid thoughts and thus to optimize therapy and prevention programs.
The aim of the study is therefore to investigate whether acute increases in the intensity of emotion regulation strategy use (I), and the variable use of strategies (II + III) predict negative affect and paranoid thoughts in daily life. We investigate this question in a subclinical sample using a prospective design with the ESM and assessing emotion regulation, negative affect, and paranoid thoughts nine times a day over the course of one week. We hypothesize that (Ia) acute increases in the intensity of acceptance and reappraisal from one measurement point to the next, are predictive of lower levels of negative affect and paranoid thoughts, whereas (Ib) acute increases in the intensity of rumination and suppression are predictive of higher levels of negative affect and paranoid thoughts. Furthermore, we expect that (II) more between-strategy variability at one measurement point is predictive of less negative affect and paranoid thoughts at the subsequent measurement point and (III) more within-strategy variability of a single strategy during one day predict lower levels of negative affect and paranoid thoughts during the same day.
Method
Participants
Participants were recruited in the city of Hamburg, Germany, via postings at several public places, at the university, and in different mental health supportive facilities to increase the probability to recruit a sample with subclinical symptoms.
Inclusion criteria were a minimum age of 18, sufficient knowledge of German to participate and the experience of subclinical psychotic experiences during the previous four weeks assessed with the Community Assessment of Psychic Experiences (Stefanis et al., 2002). Only participants with scores above nine were included as these correspond to a score above the 50th percentile (Med = 8.0) within a large community sample (Lincoln et al., 2017). Although we assessed the diagnosis of a psychotic disorder or another mental disorder by self-report, these were not defined as exclusion criteria.
Our final sample (N = 34) for the ESM study consisted of 70.4% female participants with a mean age of 29.9 years (SD = 10.2). For current employment status, 17.2% reported working full time and 26.6% worked part time, 47.0% were students, 5.9% were retired and 9.3% were currently unfit for work. The average years of education were relatively high (M = 15.7, SD = 3.3). More than half of the participants (52.8%) reported a mental health problem in the course of their lifetime. Most frequently, they mentioned the diagnoses of depression (41.2%) and anxiety disorders (20.9%). No one reported the diagnosis of a psychotic disorder. Comparable with samples fulfilling criteria for a psychotic disorder (Schlier et al., 2015), the CAPE positive symptoms score was M = 14.52 (SD = 4.73), the CAPE negative symptoms score was M = 18.64 (SD = 6.76) and the CAPE depressive symptoms score was M = 8.92 (SD = 4.15). Further analyses and descriptions of the sample can be found in Krkovic et al. (2020).
Procedure
The participants first filled out an online prescreening to assure that they met the inclusion criteria and were likely to experience paranoid thoughts in daily life. Only participants who fulfilled the criteria received an invitation for the main study.
The main study started with a baseline assessment in the laboratory. After giving informed consent, data from various questionnaires and a baseline recording of an electrocardiogram was collected, which is published elsewhere (Bahlinger et al., 2020). The participants received a study smartphone on which the application Movisens XS was preinstalled. They were instructed how to use the application and the meaning of the ESM items (particularly the different emotion regulation strategies) was explained. For the following week, participants rated their current negative affect, symptoms of paranoia, and usage of emotion regulation strategies nine times a day. Furthermore, current experience of stress and symptoms of depression were assessed, which are not reported in detail here. The 63 measurement points were distributed semi-randomly between 9 a.m. and 10 p.m. and were at least 45 min apart. The measurement point could be postponed for six min before it was classified as ignored. During the whole week, participants were able to call or contact the experimenter via email if they struggled with the data collecting devise or experienced any other problems regarding the study. After seven days, participants returned the study smartphones to the experimenter and in a short interview, they were asked to report any technical complications, problems with the ESM procedure or items. They had the opportunity to report any particular distressing experience during the past week. Afterwards they were debriefed, possible concerns were discussed, and they received monetary compensation for participation.
ESM Parameters
All items of the ESM assessment are provided in Table S1 of the supplementary material.
Emotion Regulation
We assessed four strategies of emotion regulation: acceptance and reappraisal as putatively functional strategies and rumination and suppression as putatively dysfunctional strategies (see Aldao et al., 2010). The participants were able to report the usage of several strategies at once, reflecting the fact that emotions are often polyregulated (Ford et al., 2019). Each strategy was represented by three items rated on a seven-point Likert scale ranging from “not at all” to “very much”. For reappraisal and rumination, we used the German version of the well-validated Cognitive Emotion Regulation Questionnaire (CERQ; Loch et al., 2011). For acceptance, we assessed the corresponding subscale of the Emotion Regulation Skills Questionnaire (SEK-27; Berking & Znoj, 2008), since acceptance in the CERQ reflects acceptance of the situation while our focus was on dealing with the emotional experience. For suppression, which is not included in the CERQ, we used the German version of the Emotion Regulation Questionnaire (ERQ; Abler & Kessler, 2009). Because the corresponding subscales of the CERQ and ERQ consist of more than three items, we extracted the items with the highest factor loading to ensure that all strategies were measured by the same number of items. We adapted the framing for the state assessment and asked the participants: “Please focus on the negative emotions that you experience most right now. We would like to know how you deal with these current, most negative feelings.”.
Negative Affect
Negative affect was defined as current experiences of fear, anger, sadness, and shame. Each of these emotions were assessed by one item consisting of four describing adjectives (e.g., sad/depressed/miserable/dejected; Stemmler et al., 2001). The participants were asked to rate the intensity of their current experience of the given emotion on a seven-point Likert scale ranging from “not at all” to “very much”.
Paranoid Thoughts
To assess paranoid thoughts we used the Brief State Paranoia Checklist (Schlier et al., 2016), which was derived from a state-adapted version of the Paranoia Checklist (Freeman et al., 2005). The instrument consists of five statements like “I need to be on my guard against others.” or “My actions and thoughts might be controlled by others.” and the participants had to rate how much these apply to them “at the moment”. In accordance with the other ESM items, we also used a seven-point Likert scale ranging from “not at all” to “very much”.
Acute Increases in the Intensity of Emotion Regulation Strategy Use
To determine acute increases in the intensity of strategy use, we adapted an approach for measuring affective instability in ESM data (Jahng et al., 2008). Therefore, we calculated a dichotomous variable of acute increase that indicated whether the intensity of an emotion regulation strategy increased acutely from one measurement point to the next. Acute increases were defined statistically as there was no theoretical or empirical basis to define a cut-off. First, successive differences for each emotion regulation strategy were calculated. Second, successive differences, which were more than + 1 SD above the total sample’s mean of successive differences, were defined as “acute increase” (similar to Schneider et al., 2012). The dichotomous predictors of acute increases are 1, when the intensity of the emotion regulation strategy increased between the two measurement points (i.e., t − 1 and t) and 0, when no acute increase occurred.
Calculation of Between- and Within-Strategy Variability
As suggested by Aldao et al. (2015) and recently implemented by Blanke et al. (2020), we used standard deviations to indicate a variability of emotion regulation strategies. To measure between-strategy variability, we calculated the SD of the ratings of all four emotion regulation strategies at one measurement point. Thus, a high between-strategy variability indicates that some strategies are used to a high and others to a low extent at the same time. For within-strategy variability, we calculated the SD of all measurements of one strategy within one day. Thus, a high within-strategy variability indicates a variation in the extent that one specific strategy is used across time.
Statistical Analysis
Due to the hierarchical structure of our data, we used multilevel modeling (MLM). Level 1 consisted of the repeated momentary measurements, which were clustered in participants (i.e., level 2). We assumed medium effect sizes based on previous studies investigating the associations of the intensity of emotion regulation with negative affect and paranoid thoughts in daily life (Ludwig et al., 2020; Nittel et al., 2018; Visser et al., 2018). We analyzed whether we had a sufficient sample size on level 2 to detect medium effect sizes on level 1 with a targeted power of ≥ 80%. We estimated power based on 1000 Monte Carlo simulations using the R package SIMR, which allows to calculate power for hierarchically structured data (Green & Macleod, 2016). For hypothesis III, we determined a required sample size of n = 31 on level 2 to achieve a power of 89.4%. Due to aggregating measurement points for within-strategy variability, the number of observations on level 1 was smaller for examining within-strategy variability than for the other two hypotheses. Therefore, the required sample sizes to detect medium effect sizes were below n = 31 on level 2 for hypothesis I and II. Multilevel reliabilities for two-level α were estimated applying multilevel confirmatory factor analyses (Geldhof et al., 2014). The reliability analyses were run with Mplus (version 7; Muthén & Muthén, Los Angeles, CA). All remaining statistical analyses were conducted with IBM SPSS (version 25; IBM Corp., Armonk, NY).
All predictors were on level 1 and therefore person mean centered. All models were calculated with a random intercept and a random slope and we used the heterogeneous first-order autoregressive structure as covariance type (ARH1). ARH1 assumes that the correlation between measurements of successive points is highest. To calculate the intraclass correlations (ICCs) of the dependent variables, we conducted null models without any predictors with a random intercept and a fixed slope and used variance components as covariance structure. The ICCs were calculated as the ratio of between-person to total variance in the null model.
To test the hypothesis (I)—that an acute increase in the intensity of a certain emotion regulation strategy between t − 1 and t predicts a change in negative affect or paranoid thoughts from t − 1 until t—we calculated separate MLMs for the two outcome variables. In the MLMs, we entered acute increases in the intensity of emotion regulation strategies as predictors of negative affect and paranoid thoughts at measurement point t, while controlling for the outcome variable at measurement point t − 1. For a graphical depiction of the operationalization of hypothesis (I), see Fig. 1. For hypothesis (II)—that between-strategy variability at one measurement point is predictive of negative affect and paranoid thoughts at the subsequent measurement point—we calculated separate MLMs with between-strategy variability at t − 1 predicting negative affect and paranoid thoughts at t while controlling for the outcome variable at t − 1. To test hypothesis (III)—that within-strategy variability is predictive of negative affect and paranoid thoughts on that day—we entered within-strategy variabilities of all four strategies in the MLMs predicting the average level of negative affect and paranoid thoughts on the same day. As a high variability is not possible at extremely low or high levels of mean emotion regulation, the variability effect can be confounded with mean emotion regulation use. Thus, we controlled for mean emotion regulation use in the analyses for hypotheses (II) and (III). As an effect size measure we provide standardized regression coefficients (β) of all fixed effects (Lorah, 2018).
Fig. 1
Operationalization of hypothesis I. t = time point for which negative affect and paranoid thoughts are predicted. t − 1 = time point before t. Acute increases in the intensity of an emotion regulation strategy are defined as present when successive differences between t − 1 and t were more than + 1 SD apart from the total sample’s mean of successive differences
×
Results
Descriptive Statistics and Compliance Rate
The mean scores, standard deviations and ranges of the ESM variables are presented in Table 1. The ICCs for the dependent variables were 0.405 for state negative affect, 0.422 for state paranoid thoughts, 0.663 for paranoid thoughts per day, and 0.658 for negative affect per day.
Table 1
Mean scores, standard deviations and range of the ESM variables
M
SD
Range
Acceptance
4.47
1.12
1.90–6.89
Reappraisal
4.30
1.35
1.08–6.84
Rumination
3.92
1.11
1.66–6.20
Suppression
3.71
1.16
1.04–6.30
Negative affect
2.16
0.65
1.04–3.28
Paranoia
2.44
0.92
1.03–4.95
ESM experience sampling method
For state emotion regulation, Cronbach’s Alpha for the strategies was excellent at the between subject level, which describes the scale as a whole, and acceptable at the within subject level, reflecting the variance of the state measurements (acceptance: αbetween-subject-level = .968; αwithin-subject-level = .608; reappraisal: αbetween-subject-level = .977; αwithin-subject-level = .695; rumination: αbetween-subject-level = .927; αwithin-subject-level = .588; suppression: αbetween-subject-level = .986; αwithin-subject-level = .681). For state negative affect and paranoid thoughts, Cronbach’s Alpha was good at between subject level (αnegativeaffect = .883; αparanoia = .846) and acceptable at within subject level (αnegativeaffect = .723; αparanoia = .757).
Of the 2142 planned measurement points (i.e., 63 measurement points in 34 participants) 2041 were delivered. Out of these 329 were ignored (15.8%) and 13 were incomplete (0.6%). Thus, the final data set consisted of 1699 measurement points. The average compliance rate was 83.34% with a range between 41.27 and 100%, which is a typical rate in mental health research (Vachon et al., 2019).
Acute Increases in Emotion Regulation as a Predictor for Negative Affect and Paranoia (I)
Results of MLMs for hypothesis (I) are depicted in Fig. 2. The detailed results can be found in Table S2 in the supplementary material available online.
Ia) An acute increase in acceptance and reappraisal between two measurement points t − 1 and t was predictive of lower levels of negative affect at t when controlling for negative affect at t − 1. An acute increase in acceptance between t − 1 and t predicted less pronounced paranoid thoughts at t when controlling for paranoid thoughts at t − 1. An acute increase in reappraisal was not a significant predictor of paranoid thoughts.
Ib) An acute increase in rumination and suppression between t − 1 and t predicted higher levels of negative affect at t when controlling for negative affect at t − 1 and more pronounced paranoid thoughts at t when controlling for paranoid thoughts at t − 1.
Fig. 2
Acute increases in the intensity of emotion regulation strategies predicting negative affect and paranoid thoughts (I). t = time point for which negative affect and paranoid thoughts are predicted. t − 1 = time point before t. Acute increases in the intensity of an emotion regulation strategy are defined as present when successive differences between t − 1 and t were more than + 1 SD apart from the total sample’s mean of successive differences. n.s. = not significant. Fixed effects of the multi-level models when controlling for negative affect/paranoid thoughts at t − 1. β = standardized coefficient
×
Between- and Within-Strategy Variability as a Predictor of Negative Affect and Paranoia (II + III)
Between-strategy variability at t − 1 was not a significant predictor of negative affect or paranoid thoughts at t when controlling for the outcome variable at t − 1 and mean emotion regulation use at t − 1 (see Table 2).
Table 2
Negative affect and paranoid thoughts predicted by between-strategy variability (II)
Negative affect
Paranoid thoughts
Between-strategy variability
b
− 0.058
− 0.024
SE
0.188
0.207
95% CI
[− 0.430, 0.313]
[− 0.431, 0.383]
p
.758
.909
β
− 0.006
− 0.002
Outcome variable at t − 1
b
0.300
0.325
SE
0.045
0.039
95% CI
[0.209, 0.391]
[0.246, 0.404]
p
< .001
< .001
β
0.243
0.214
Mean ER use
b
0.006
− 0.003
SE
0.149
0.184
95% CI
[− 0.286, 0.298]
[− 0.377, 0.370]
p
.968
.986
β
0.001
0.000
ER emotion regulation, β standardized coefficient
Within-strategy variability did not predict mean levels of negative affect and paranoid thoughts across that day when controlling for mean emotion regulation use that day (see Table 3).
Table 3
Negative affect and paranoid thoughts predicted by within-strategy variability (III)
Negative affect
Paranoid thoughts
Variability of
b
0.495
1.281
acceptance
SE
0.565
0.660
95% CI
[− 0.622, 1.611]
[− 0.021, 2.583]
p
.383
.054
β
0.040
0.067
Variability of reappraisal
b
0.290
0.552
SE
0.438
0.638
95% CI
[− 0.575, 1.154]
[− 0.842, 1.945]
p
.509
.404
β
0.030
0.036
Variability of rumination
b
0.567
0.081
SE
0.506
0.698
95% CI
[− 0.434, 1.567]
[− 1.387, 1.548]
p
.265
.909
β
0.053
0.005
Variability of suppression
b
0.830
0.453
SE
0.503
0.613
95% CI
[− 0.164, 1.824]
[− 0.773, 1.679]
p
.101
0.463
β
0.081
0.028
Mean ER use
b
− 0.343
1.153
SE
0.432
0.722
95% CI
[− 1.195, 0.509]
[− 0.349, 2.656]
p
.428
.125
β
− 0.036
0.078
ER emotion regulation, β standardized coefficient
Discussion
In this study, we investigated whether acute increases in the intensity of emotion regulation strategies and their variable use predict negative affect and paranoid thoughts in daily life. As expected, acute increases in the intensity of emotion regulation strategy use were associated with negative affect and paranoid thoughts. However, our results neither confirmed the assumption that between-strategy variability would predict subsequent negative affect and paranoid thoughts, nor that within-strategy variability would be associated with negative affect and paranoid thoughts at the same day.
By examining acute increases in the intensity of strategy use, we showed that the acute onset of emotion regulation strategy use is related to negative affect. We found that negative affect was low following an increase in the intensity of acceptance and reappraisal. Moreover, levels of negative affect were high when rumination and suppression were increased since the previous time point. Thus, when investigating possible deficits in emotion regulation in individuals with psychotic symptoms in future studies, it could prove useful to include the aspect of temporal changes in strategy use and not only focus on the intensity and frequency of single strategies. To explain the puzzling finding that individuals with psychosis showed increased levels of negative affect although reporting to use emotion regulation strategies more frequently than healthy controls (Ludwig et al., 2020; Visser et al., 2018), it may be helpful to focus on deficits in the targeted use of strategies. Therefore, future studies should investigate not only whether functional strategies are employed but whether their intensity is increased acutely in response to environmental demands.
Additionally, acute increases in the intensity of emotion regulation strategies were directly linked to the occurrence of paranoid thoughts. Rumination and suppression were already found to predict more paranoid thoughts in clinical samples (Hartley et al., 2014; Nittel et al., 2018), which we could now show in individuals with subclinical psychotic experiences. Furthermore, we found that acute increases in acceptance predicted fewer paranoid thoughts. Existent cross-sectional studies had already indicated an association of acceptance with positive outcomes, such as quality of life and low distress, in individuals experiencing psychotic symptoms (Osborne et al., 2017; Vilardaga et al., 2014). Until now, it had remained unclear whether acceptance is related to paranoid thoughts in daily life. Therefore, our results extend findings from previous studies by not only showing that rumination and suppression predict paranoia in individuals with subclinical psychotic experiences but also that intensifying acceptance towards the emotional experience is associated with less paranoid thoughts.
Against our expectation, increases in the intensity of reappraisal were not linked to paranoid thoughts. Reappraisal is commonly considered a functional emotion regulation strategy (Webb et al., 2012), but recently it was discussed that its use is not always beneficial (Ford & Troy, 2019). Some studies have even found reappraisal to be associated with more pronounced paranoid thoughts (Kimhy et al., 2020; Westermann et al., 2012) and to increase negative affect during psychotic experiences (Strauss et al., 2019). For determining the functionality of reappraisal, the reappraised thought contents and the context, in which it is used, may need to be considered. If the alternative thoughts involved blaming others or self-blaming, this could even trigger delusional interpretations. This is supported by a finding by Westermann et al. (2013) showing that the emotion regulation strategy of self-blame predicted paranoid thoughts one month later while controlling for paranoia at baseline. Furthermore, evidence shows that in some contexts, for example when a situation is perceived as controllable (Haines et al., 2016) or has a high emotional intensity (Hay et al., 2015; Scheibe et al., 2015; Shafir et al., 2016), reappraisal is not the best choice. Assessing the content of thoughts and the context could contribute to answering how and when a targeted use of reappraisal can help to deal with paranoid thoughts.
In contrast to our hypothesis, when individuals prioritized strategies (i.e., high between-strategy variability) instead of simultaneously using all strategies to the same extent, they did not have lower levels of subsequent negative affect and paranoid thoughts. This was unexpected given that previous studies have found that using multiple strategies at once is associated with more negative affect (Aldao & Nolen-Hoeksema, 2013) and that clinical samples report to use more strategies simultaneously than healthy controls (Gruber et al., 2013; Ludwig et al., 2020), which was not only explained by a higher emotional reactivity (Gruber et al., 2012). However, previous work also found that employing several strategies at once that were experienced as helpful was found to benefit mood (Heiy & Cheavens, 2014) and combining the strategies of accepting and tolerating emotions with comprehending the emotional experience was related to less subsequent paranoid thoughts (Wittkamp et al., 2021). This indicates that a pure assumption of “less is more” is too simplified and the effectiveness of combinations of strategies may depend crucially on which specific strategies are combined.
Furthermore, we did not find that individuals who applied the same emotion regulation strategy at differing intensities during one day (i.e., high within-strategy variability) had lower levels of negative affect and less paranoid thoughts. Thus, considering a high global variability irrespectively of the need and context as functional may be too short-sighted. Rather, in line with the extended process model of emotion regulation (Gross, 2015), it needs to be taken into account whether stopping, switching, and maintaining of strategies fits the context. When regulation is stopped before the emotion is effectively downregulated or when switching between strategies does not follow a temporal order in search for the best fitting strategy but proceeds chaotically, high variability may reflect a highly unstable and rather ineffective regulation pattern. Therefore, neither extremely high variability nor high inertia in the use of strategies may be advantageous to regulate emotions effectively but an optimal level of moderate variability (Blanke et al., 2020). In line with that, it is supposed that in a stable context, maintaining one strategy can be even more adaptive than a switch of strategy (Pruessner et al., 2020). A direct examination of whether changes in the intensity of the strategies covary with environmental changes (see Aldao et al., 2015) and investigating the temporal sequences of switches between strategies could help to further determine the effectiveness of the variable use of emotion regulation strategies.
Some limitations of our study need to be considered. First, since we did not assess the regulatory context, we cannot answer the question whether the intensity of strategies was increased in order to encounter environmental demands. In future studies, an assessment of context variables and regulatory goals would be beneficial to further investigate the effectiveness of the intensity of and variability within emotion regulation strategies. As the choice of emotion regulation strategies depends on regulatory goals (e.g., Eldesouky & English, 2019; English et al., 2017; Tamir & Ford, 2012; Wilms et al., 2020), it should be examined whether the effectiveness of specific strategies for reducing negative affect and paranoid thoughts depends on which goal is currently activated. Second, we focused on the self-report of emotion regulation, negative affect, and symptoms. Deficits in emotion awareness were reported even in subclinical samples (Kimhy et al., 2016), which renders the additional assessment of psychophysiological variables worthwhile. Third, we only investigated the four most commonly investigated emotion regulation strategies, although other strategies, such as distraction (Martinelli et al., 2013) or social support (Ludwig et al., 2020; Nittel et al., 2018), may also be relevant to the questions of interest. Forth, as we based our power calculations on the assumption of medium effect sizes, we acknowledge that we could have missed smaller effects. Finally, to address which deviations of emotion regulation are ultimately predictive of eventually developing a clinical disorder, examining samples at clinical high risk in a longitudinal design would be informative.
Notwithstanding these limitations, our study strengthens the view that emotion regulation is linked to the occurrence of paranoid thoughts in individuals with subclinical paranoia. More precisely, acute increases in the intensity of the usage of single strategies were found to relate to the experience of negative affect and paranoid thoughts. In light of our study’s results, it seems recommended to further investigate whether emotion regulation plays a causal role in the development of paranoia. For this purpose, it can be helpful to apply ecological momentary interventions (EMI), which deliver psychological interventions in the daily life of participants (Myin-Germeys et al., 2016) and could thereby change emotion regulation when it takes place. By testing whether these changes in the regulation of negative affect directly lead to a sustainable reduction of psychotic symptoms, EMI can evaluate causal relationships between regulation, negative affect, and paranoid thoughts (i.e., ecological interventionist causal models, Reininghaus et al., 2016). This could shed light on the question whether emotion regulation interventions may help to prevent paranoia in the future.
Acknowledgements
We would like to thank Philipp Kraatz for his support in data collection.
Declarations
Conflict of Interest
Katrin Bahlinger, Tania M. Lincoln and Annika Clamor declare that they have no conflict of interest.
Ethical Approval
All procedures performed in the study were in accordance with the provisions of the World Medical Association Declaration of Helsinki (as revised in 2013). The study received approval by the local ethics committee of the Universität Hamburg.
Informed Consent
Informed consent was obtained from all individual participants included in the study.
Animal rights
No animal studies were carried out by the authors for this article.
Open AccessThis 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.
