Cloninger’s TCI associations with adaptive and maladaptive emotion regulation strategies

Introduction

We deal with a variety of situations in everyday life, which leads to various emotional responses. Such emotional responses appear immediate and unchangeable at first. However, biological, social, and cognitive behavioral self-regulatory processes become activated internally. These processes are responsible for monitoring, evaluating and modifying emotional reactions in order to accomplish one’s goals in the context of human life, and such a process has been termed emotion regulation (Garnefski, Kraaij & Spinhoven, 2001; Thompson, 1994). Successful emotion regulation is associated with optimal physical and mental health outcomes (Aldao, Nolen-Hoeksema & Schweizer, 2010; Jo & Lee, 2013; Potthoff et al., 2016; Seol & Park, 2015), improved relationships (Cicchetti, Ackerman & Izard, 1995; Mihalca & Tarnavska, 2013) and better academic and work performance (Grandey & Melloy, 2017; Mega, Ronconi & De Beni, 2014). Difficulties in emotion regulation are associated with mental disorders including major depression, bipolar disorder, generalized anxiety disorder, social anxiety disorder, eating disorder, and alcohol and substance related problems (Aldao, Nolen-Hoeksema & Schweizer, 2010).

Two separate strategies can be broadly defined as (1) mature, adaptive and positive vs. (2) immature, maladaptive and negative, and they are the product of predisposed tendencies based on individual differences (Garnefski, Kraaij & Spinhoven, 2001).

These two types of strategies, adaptive (adaptive cognitive emotion regulation questionnaire: aCERQ) and maladaptive (maladaptive cognitive emotion regulation questionnaire: mCERQ), can be measured using the CERQ (Garnefski, Kraaij & Spinhoven, 2001). The CERQ consists of nine subscales corresponding to adaptive or maladaptive strategies. The aCERQ consists of five subscales: Putting into perspective (PIP), Refocus on planning (REP), Positive Refocusing (PRF), Positive Reappraisal (PRA) and Acceptance (ACC). The mCERQ is composed of four subscales: Rumination (RUM), Catastrophizing (CAT), Blaming Others (BLO) and Blaming Self (BLS).

In this context, personality has been suggested to predict the use of emotion regulation strategies of a given person (Kokkonen & Pulkkinen, 2001). Previous studies using temperamental models of personality, for example, have reported that those high in Neuroticism tend to use maladaptive emotion regulation strategies, while individuals high in Extraversion and Openness use adaptive emotion regulation strategies (Hassani et al., 2008; Kokkonen & Pulkkinen, 2001; Purnamaningsih, 2017; Sadr, 2016; Sohn & Yoo, 2011). However, personality also consists of psychosocial characteristics within an individual that determine behavioral, emotional and cognitive responses to internal and environmental stimuli that extend beyond temperament or stable automatic emotional responses (Cloninger, 2004). Thus, given the fact that successful emotion regulation in the face of stressful stimuli is an adaptive and mature coping strategy, we argue that what predisposes it is a stable temperament or the tendency to demonstrate stable emotional reactions (e.g., low scores in Neuroticism). Furthermore, mature character is reflected by changeable and mature goals and values in relation to the self, others, and the whole world (Eley et al., 2016; Kim, Lee & Lee, 2013; Moreira et al., 2015).

Cloninger’s biopsychosocial model of personality (Cloninger, 1987; Cloninger, Svrakic & Przybeck, 1993) consists of two domains, temperament and character, and explains the foundation and development of personality through their interaction. Temperament is an innate and biological foundation for character (Gray, 1982; Sjobring, 1973). In other words, it constrains character development but does not fully determine it because of the systematic effects of sociocultural learning and the stochastic effects of experience (Cloninger, 1987; Cloninger, Svrakic & Przybeck, 1993; Cloninger & Zwir, 2018; Garcia et al., 2014).

The temperament domain is composed of NS, HA, Reward Dependence (RD), and PS. NS is a personality trait associated with exploratory excitability in response to novel stimuli, and HA is associated with behavioral inhibition in reaction to dangerous or abhorrent stimuli. RD is a tendency to respond particularly to signals of social reward, and PS is a tendency to maintain behavior with intermittent reinforcement (Cloninger, Svrakic & Przybeck, 1993). The character domain is comprised of SD, CO, and Self-Transcendence (ST). SD indicates the tendency to identify the self as an autonomous individual and CO measures the degree to which one considers the self as integral to humanity. ST represents the degree to which one considers the self as an integral part of the universe as a whole (Cloninger, Svrakic & Przybeck, 1993). Generally speaking, there are advantages of both high and low temperament scores depending on the situations but from the perspective of character, only high scores are adaptive.

In addition, higher score in SD is a critical protective factor in psychopathology and also a factor in the promotion of mental health (Eley et al., 2013; Eley et al., 2016; Kim, Lee & Lee, 2013). Meanwhile, higher scores in HA are positively associated to maladaptive emotion regulation as measured by CERQ (Aldao, Nolen-Hoeksema & Schweizer, 2010; Ball, Smolin & Shekhar, 2002; Izadpanah et al., 2016; Manfredi et al., 2011; Peirson & Heuchert, 2001; Schreiber, Grant & Odlaug, 2012). The combination of higher scores in SD and lower scores in HA which reflect positive characteristics such as life satisfaction, optimism, and hopefulness (Cloninger, 2004) is important for mental health and longevity (DeNeve & Cooper, 1998; Lee et al., 2014).

However, previous studies were primarily dedicated to examining the strong association between maladaptive emotion regulation strategies, psychopathology and well-being (Aldao, Nolen-Hoeksema & Schweizer, 2010). Therefore, there is a need for research that considers positive and adaptive perspectives as well as negative and maladaptive perspectives because health is characterized by not only the absence of illness, but also a state of well-being (WHO, 2005). In addition, HA and SD traits along with other temperament and character traits should be examined together. For instance, resilience or a person’s ability to effectively adapt and flourish in face of adversity is associated not only with low scores in HA and high scores in SD, but also with the ability to be hard working and persistent (i.e., high scores in PS) (Eley et al., 2013).

For these reasons, the aim of this study was to explore the association between personality and adaptive and maladaptive cognitive emotion regulation by (1) assessing how much variance in emotion regulation scores can be explained by personality scores and by (2) identifying groups of emotion regulation style and assessing differences in personality scores between these groups (Cloninger et al., 2012; Eley et al., 2016). This will help us to understand how temperament and character interact to impact adaptive and maladaptive emotional responses to stress, hence providing foundations for promoting psychological well-being.

Materials & Methods

Results

LCA of CERQ subscale scores

LCA was performed (Table 4), and a four latent class model (Fig. 1A) was determined to be acceptable since all the criteria of model fit were satisfied: a four class model demonstrated smaller BIC (11212.21) and adjusted BIC (11053.71) compared to other models, more significant p value of VLMR (p = .009), LMR (p = .010), and BLRT ( p = .000), and a greater than .8 Entropy index (.82) (Table 4).

Table 4:

Information criterion for 2-5 latent classes of the nine subscales of the CERQ.

Model	BIC	adj. BIC	VLMR p	LMR p	BLRT p	Entropy
2 class	11353.27	11258.17	.025	.026	.000	.73
3 class	11281.05	11154.25	.694	.697	.000	.78
4 class	11212.21	11053.71	.009	.010	.000	.82
5 class	11216.99	11026.79	.468	.475	.000	.82

DOI: 10.7717/peerj.7958/table-4

Notes:

LCA

Latent Class Analysis

adj

adjusted

BIC

Bayesian Information Criterion

BLRT

Bootstrapped Likelihood Ratio Test

CERQ

Cognitive Emotion Regulation Questionnaire

LMR

Lo–Mendell–Rubin Likelihood Ratio Test

VLMR

Vuong-Lo-Mendell-Rubin Likelihood Ratio Test

The four latent CERQ classes were labeled as HALM or High Adaptive and Low Maladaptive (n = 78, 32.58%), MALM or Middle Adaptive and Low Maladaptive (n = 61, 24.70%), MAHM or Middle Adaptive and High Maladaptive (n = 91, 36.84%) and LAHM or Low Adaptive and High Maladaptive (n = 17, 6.88%) classes based on aCERQ and mCERQ scores. aCERQ and mCERQ total score of HALM (62.20 ±.74, 24.37 ± .65), MALM (43.17 ± .83, 19.14 ± .73), LAHM (29.17 ± 1.56, 37.59 ± 1.38), MAHM (49.08 ± .67, 35.51 ±.59) classes were significantly different from each other according to ANCOVA (F = 170.88, p < .001 in aCERQ; F = 128.84, p < .001 in mCERQ).

Significant differences in all nine CERQ subscales among the four latent classes were found when ANCOVA and Bonferroni post-hoc analysis were performed. In aCERQ subscales, the four latent classes were significantly different from each other in regard to PIP (F = 63.04, p < .001) and PRA (F = 108.69, p < .001) scores. REP score was significantly different (F = 22.58, p < .001) across the classes except between LAHM and MALM, and LAHM and MAHM. PRF score was significantly different (F = 38.73, p < .001) across the classes except between MALM and MAHM. ACC score was significantly different (F = 48.80, p < .001) across the classes except between MALM and LAHM. In aCERQ subscales, the four latent classes were significantly different from each other in PIP and PRA scores. REP score was significantly different across the classes except between LAHM and MALM, and LAHM and MAHM. PRF score was significantly different across the classes except between MALM and MAHM. ACC score was significantly different across the classes except between MALM and LAHM.

Discussion

This study investigated the effects of personality as individual traits as well as complex profiles on cognitive emotion regulation strategies using the TCI-RS and CERQ in a sample of Korean university students. Based on an exploration of latent classes depending on adaptive and maladaptive cognitive emotion regulation, we provide further support that character traits of higher scores in both SD and CO play a pivotal role in predicting psychological adaptation and adjustment.

First, as for the individual personality traits, the temperament trait HA was associated with maladaptive emotion regulation as measured by the CERQ (mCERQ) as in previous studies (Aldao, Nolen-Hoeksema & Schweizer, 2010; Ball, Smolin & Shekhar, 2002; Izadpanah et al., 2016; Manfredi et al., 2011; Peirson & Heuchert, 2001; Schreiber, Grant & Odlaug, 2012). That is, the present results found that temperament traits may guide the direction of cognitive emotion regulation strategy in adaptive or maladaptive ways, with high scores in HA being related to internalizing problems such as depression, anxiety, worry and catastrophization (Kim & Lee, 2018).

In addition, the temperament trait NS was found to be associated with adaptive strategies in this study. This may be expected given that an individual with a high score in NS typically seeks out and enjoys new or novel experiences, ideas and innovative activities. Extraversion and Openness of the five factor model (McCrae & John, 1992) which share a similar conceptual basis with NS are known to be correlated with the use of adaptive emotion regulation strategies (Aluja & Blanch, 2011; Hassani et al., 2008; Kokkonen & Pulkkinen, 2001; Purnamaningsih, 2017; Sadr, 2016; Sohn & Yoo, 2011). However, one of the four sub-dimensions of the NS temperament is impulsivity and impulsivity is typically not associated with adaptive emotion regulation (Schreiber, Grant & Odlaug, 2012). The association between impulsivity and adaptive emotion regulation should be further explored in future studies.

Second, the temperament trait PS showed a dualistic effect (Cloninger et al., 2012), concurrently strengthening both adaptive and maladaptive cognitive emotion regulation strategies. More specifically, PS demonstrated dualistic and opposite effects on cognitive emotion regulation depending on the particular adaptive or maladaptive strategy. This suggests that PS may sustain the employment of emotion regulation strategy in both adaptive as well as maladaptive ways.

PS is a tendency to maintain one’s behavior even in the face of punishment or compensation, and individuals with higher PS can be characterized as industrious, determined, ambitious and perfectionistic (Cloninger, 1987; Cloninger, Svrakic & Przybeck, 1993). An individual with high scores in HA and PS may tend to use maladaptive cognitive emotion regulation strategies since an individual with high scores in HA may focus on the dangers and worries related to the situation and employ pessimistic interpretation and avoidant responses. Therefore, they may be more sensitive to criticism and the possibility of punishment, which may further reinforce rigid maintenance of their cognitive style. However, an individual with high scores in NS and PS tended to use adaptive cognitive emotion regulation strategies. High scores in NS prevent loss of motivation resulting from non-reward or frustration despite the difficulty of the situation and this may make the person withstand and sustain their current cognitive style in a more adaptive way.

Third, it was also found that a high score in the character trait of SD promotes the use of adaptive emotion regulation and suppresses the use of maladaptive emotion regulation strategy in addition to the temperament traits such as higher HA and PS. According to Cloninger (Josefsson et al., 2013), SD is the first character trait that develops among humans and serves as the foundation of character development (Josefsson et al., 2013). Those who score high in the character trait of SD are considered as responsible, purposeful, resourceful, self-accepting, and disciplined, suggesting that such individuals regard themselves as autonomous (Cloninger, Svrakic & Przybeck, 1993). Therefore, high scores in SD are thought to serve as a protective factor in moderating unhealthy ways of thinking and encouraging the use of more well-functioning means of emotion regulation (Cloninger & Zohar, 2011; Josefsson et al., 2011).

When person-centered analyses were performed on the complex nature of temperament and character using LCA and Profile Analysis, there were four latent CERQ classes showing distinct personality classes. Individuals with a highly adaptive emotion regulation style/profile (i.e., HALM) exhibited the highest levels of responsibility, resourcefulness, and self-acceptance (i.e., high scores in SD) and also the highest levels of optimism and energy levels (i.e., low scores in HA). In contrast, individuals with a maladaptive emotion regulation style/profile (i.e., LAHM) were characterized by low self-esteem, feelings of victimization, and lack of goals (i.e., low scores in SD) and also by pessimism, low energy, and propensity for anxiety (i.e., high scores in HA) (Cloninger, 1987; Cloninger, Svrakic & Przybeck, 1993). Indeed, resilience has been proposed as the result of this specific personality configuration (Cloninger, 2004; Eley et al., 2013). That being said, resilience or adaptive functioning has also been associated with industriousness, perseverance, and hard-working style (i.e., high scores in PS). Thus, one remaining question is why individuals with a highly adaptive emotion regulation style/profile did not demonstrate higher scores in PS compared to individuals with a maladaptive emotion regulation style/profile. It may be that high levels of PS are only adaptive in combination with high scores in SD and low scores in HA. In other words, high levels of PS are also a feature of individuals who may exhibit high HA and low SD. This personality profile, however, is vulnerable for burn out. Hence, it is possible that the lack of difference in PS between individuals in the HALM and MALM groups is due to the fact that PS enables the maintenance of either adaptive or maladaptive emotion regulation strategies depending on how prone the individual is to worry (i.e., HA) and self-directedness (Cloninger et al., 2012). That is, the findings regarding the combination of high score in SD with low score in HA are consistent with previous findings that support the idea of psychological maturity or healthy functioning self (Josefsson et al., 2013), indicating that individuals may appropriately regulate their emotions depending on their circumstances.

Fourth, another important new finding in our study was that high scores in SD and high scores in CO and/or low scores in ST may play an important role in fast and adaptive emotional response or in the choice of cognitive emotion regulation strategy in stress situations which has not been previously reported. The importance of SD and CO in our stress response has been consistently reported (Eley et al., 2013; Eley et al., 2016; Kim, Lee & Lee, 2013; Lee, Choi & Chae, 2017; Sievert et al., 2016), and the combination of SD and CO scores representing psychological maturity was found to be an objective factor related to degree of social support, life satisfaction, subjective well-being, composite health index, and happiness (Eley et al., 2013). In addition, the combination of high scores in ST with low scores in SD is associated with maladaptive functioning and is an indicator of an immature character profile (Lee, Choi & Chae, 2017; Spittlehouse et al., 2014). That is, high scores in ST in combination with high scores in SD are advantageous in that this profile reflects creativity, holism, and contemplation but high scores in ST in combination with low scores in SD are disadvantageous as this may represent magical and unrealistic thinking. Once again, high scores in SD appear to be a prerequisite for advantages offered by high scores in ST (Cloninger & Zohar, 2011; Josefsson et al., 2011; Moreira et al., 2015).

While extant studies primarily focused on the role of temperament, the present study implicates the significant role of character in emotion regulation. The character domain of TCI-RS measures psychological maturity and is reported to be a protective factor in the development of psychopathology and is also related to resilience to stress and biopsychological well-being (Eley et al., 2013; Moreira et al., 2015). This study suggests that cognitive emotion regulation may be a mechanism that potentially explains the effects of character on psychological health. Furthermore, character may reflect an individual’s cognitive style which promotes reappraisal of a given environmental stimulus as a task to be solved, not as a dangerous threat, and thus prepare their behavioral response accordingly (Lee, Choi & Chae, 2017).

As such, temperament and character, which play a pivotal role in well-being and mental health, affect long-term development and continuation of emotion regulation style to everyday life stressors (Garnefski, Kraaij & Spinhoven, 2001), not only on an individual trait level but also through a complex interaction (Eley et al., 2016; Lee, Choi & Chae, 2017). In this way, person-centered analysis such as LCA or Profile Analysis on psychological characteristics is a great advantage of this study. The LCA and Profile Analysis (Eley et al., 2016) provide better clinical explanation compared to cluster analysis (Kim, 2008; Sievert et al., 2016) in which mature and adaptive strategy in relation to stress emerges through a complex and non-linear interaction (Cloninger et al., 2012). This study also analyzed the complex interaction of temperament and character traits in maladaptive and adaptive emotion regulation strategies which was not investigated before.

There are a number of limitations of this study. First, this study was conducted with a relatively small sample of Korean university students. The findings should be supported using a bigger sample balanced in sex and age. Second, the sex differences of Korean university students in personality and cognitive emotion regulation strategy should be further examined to determine whether cross-cultural differences (Ahn, Lee & Joo, 2013; Cho et al., 2007) exist in general.

This study examined maladaptive and adaptive cognitive emotion regulation strategies using the two perspectives of individual traits with two-step hierarchical regression analyses and the complex interaction across the classes simultaneously through LCA. The temperament domain guided the direction of emotion regulation such that high scores in HA and PS predicted maladaptive cognitive emotion regulation strategy and high scores in NS and PS predicted adaptive strategy. As for the character domain, SD and CO were correlated positively with adaptive and negatively with maladaptive cognitive emotion regulation strategy which may explain the significance of resilience to stress and promotion of well-being.

Conclusion

Specific combination of temperament and character traits appears to be particularly influential in predicting adaptive and maladaptive emotion regulation strategies and is also associated with both mental health and well-being. Mature personality as represented by character traits such as SD and CO rather than temperament traits was found again to account for temperamental vulnerability as well as specific cognitive emotion regulation. In particular, the effect of PS was once again found to be dualistic, supporting the concept of a non-linear model. In other words, the same high score in PS may exert an opposite tendency. It may be associated with adaptive emotion regulation when combined with high scores in NS but with maladaptive emotion regulation when combined with high scores in HA. Since adaptive emotion regulation is pivotal for psychological health, structured education and persistent interest on character development in Korean university students should be emphasized. This study provides a foundation for a potential psychological mechanism in stress response as well as existing psychopathological studies.

Supplemental Information