Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 1.
Published in final edited form as: J Anxiety Disord. 2023 Jan 14;94:102674. doi: 10.1016/j.janxdis.2023.102674

Contrast avoidance predicts and mediates the effect of trait worry on problem-solving impairment

Sandra J Llera a, Michelle G Newman b
PMCID: PMC9987319  NIHMSID: NIHMS1867196  PMID: 36681059

Abstract

This study examined the relationship between the Contrast Avoidance Model (CAM; Newman & Llera, 2011) and impairment in the problem-solving process using an in-vivo laboratory-based problem-solving task. We also explored whether general emotional CA tendencies explained the relationship between trait worry and problem-solving outcomes. In this study, 185 participants (42 of whom met diagnostic criteria for generalized anxiety disorder) engaged in a problem-solving task, and reported outcomes related to ability to generate solutions, confidence in solutions, intention to implement solutions, and state anxiety levels. According to results, higher general emotional CA tendencies predicted significantly more difficulties on most problem-solving outcomes. Further, CA tendencies mediated between trait worry and some, but not all, problem solving outcomes. Overall, CA appears to be linked to problem-solving deficits, and may help to explain some of the association between trait worry and negative problem-solving outcomes.

Keywords: generalized anxiety disorder, contrast avoidance, problem solving, trait worry

1. Introduction

Chronic worry has been theorized to be associated with problem-solving difficulties. Impaired problem resolution could occur at any of several problem-solving stages. These include: a) attitudes toward problems, b) defining the problem, c) generating solutions, d) decision making, and e) solution implementation and verification (D’Zurilla & Goldfried, 1971). Studies support the association between trait worry and problem-solving difficulties. For example, trait worry was strongly associated with having a negative attitude toward problems in both clinical (Dugas et al., 1998; Fergus et al., 2015; Ladouceur et al., 1998) and non-clinical samples (Anderson et al., 2009; Robichaud & Dugas, 2005). Further, such negative attitude was more strongly associated with trait worry than other anxiety, mood, and obsessive symptoms (Fergus et al., 2015).

In behavioral observations of chronic worriers, problem-solving deficits were characterized by an over-attention to threat and a tendency to delay and/or overthink decision making. For example, Borkovec (1985) observed that chronic worriers were poor problem-solvers, yet they were “superb at identifying negative outcome possibilities and at pointing out problems with any self- or other-suggested solutions” (p. 481). Further, several studies showed that when performing difficult cognitive tasks, chronic worriers and GAD participants were significantly slower than non-anxious controls in arriving at an answer (LaFreniere & Newman, 2019; MacLeod & Donnellan, 1993; Stefanopoulou et al., 2014; Tallis et al., 1991), which could delay decision making and/or solution implementation in the face of problems. Indeed, in a study of real-life problem solving using diary data, a mixed depressed-anxious sample demonstrated fewer functional behaviors compared to controls (Anderson et al., 2009), and trait worry was negatively associated with problem-solving effectiveness in a community GAD sample (Pawluk et al., 2017).

Moreover, worrying during problem-solving can also interfere with the process. For example, in an experience sampling study intensity of worry was associated with more anticipation of negative outcomes, greater negative evaluation of solutions to problems, more self-blame, and lower rates of solution selection (Szabó & Lovibond, 2002, 2006). Similarly, during an experimental problem-solving task, a worry induction was associated with lower experimenter rated effectiveness in problem solutions compared to an objective thinking task, and state worry predicted participants endorsing lower likelihood of implementing solutions (Llera & Newman, 2020).

A wealth of research has explored factors that may explain why high trait worriers have problem-solving difficulties. For example, a review of evidence supports both an attentional (Goodwin et al., 2017) and interpretation bias (Hirsch et al., 2016) toward threat in chronic worriers. Across studies, high trait worriers and people with GAD were more likely than low trait worriers and those without GAD to interpret ambiguous information as threatening and have a harder time disengaging once threat was detected. This may influence the extent to which problems are coded as threats, rather than as opportunities to improve one’s circumstances. In addition, this could lead to a hyper-focus on potential negative outcomes at the expense of shifting into more solution-focused thinking. Those with GAD also demonstrated an anticipatory bias toward negative events in their own lives (LaFreniere & Newman, 2020), which could make failure appear to be the most likely outcome of problem-solving efforts. Another factor involved in problem-solving deficits may be intolerance of the uncertainty inherent to the problem-solving process (Dugas et al., 1997), resulting in a perseverative problem-solving style in which individuals must work through every possible outcome scenario in their minds before taking action. Indeed, chronic worriers have been described as having higher “evidence requirements” which delayed decision-making (Tallis et al., 1991), and trait worry was predicted by the belief that stressful problems required prolonged thinking (Sugiura, 2007). Working together, these factors may each serve to create or maintain difficulties in the problem-solving process for chronic worriers. However, we propose that an additional perspective centered on an emotion-focused coping style plays an important role as well.

The Contrast Avoidance Model (CAM; Newman & Llera, 2011) is a broad-based explanatory model of chronic worry, and may help to shed light on why these individuals hold such negative reactions toward life’s problems. The CAM posits that individuals with GAD adopt a self-protective stance of being emotionally prepared for the worst at all times. According to this theory, individuals with GAD fear and avoid sudden shifts into a negative emotional state, such as those that would accompany a negative outcome or event. To protect themselves from such shifts, they adopt the rather paradoxical safety behavior of maintaining a constant negative emotional state. In this way, they cannot feel much worse if something bad does happen. This provides a sense of safety by functioning to keep their “emotional shields” up at all times. It may also facilitate occasional shifts into positive emotion, such as relief if the feared outcome does not happen (Newman et al., 2022). These perceived benefits come at a high cost, however, such as living the majority of one’s life in a state of negativity, and associating positive feelings with danger (Kim & Newman, 2019). That is, in the case of a positive outcome, once they have savored the initial rush of relief, individuals with GAD may now feel as though they have let their guard down. This could cause them to shift back to a negative stance to reduce feelings of emotional vulnerability.

The CAM proposes chronic worry as a mechanism by which negative emotionality is maintained, based on abundant research showing that worry leads to negative arousal (for a review, see Newman & Llera, 2011; Newman et al., 2013). Research also shows that when individuals were in a negative emotional state, such as through state worry, they were less likely to experience an upward shift in negative emotions when exposed to an aversive stimulus (Jamil & Llera, 2021; Kim & Newman, 2022; Llera & Newman, 2010, 2014; Newman et al., 2019, 2022; Peasley-Miklus & Vrana, 2000). Individuals who were in a relaxed or neutral state prior to the aversive stimulus did respond with a surge in negative emotion. In the CAM literature, this emotional shift is labeled a negative emotional contrast (NEC), because the affective valence contrasts sharply with the previous state. Importantly, individuals who worried first still reached the same absolute levels of negative affect following the stressor as did those in other conditions (Llera & Newman, 2010, 2014); it is simply that they were already at that level prior to the stressor.

Research supports the applicability of the CAM, in that those with GAD report greater fear and avoidance of NECs. In a weekly diary study, GAD symptoms predicted rating NEC experiences as the worst event of the week (Crouch et al., 2017). In experimental studies, participants with GAD reported that it was easier to cope with stressors if they were able to avoid NEC, whereas non-anxious controls did not (Jamil & Llera, 2021; Kim & Newman, 2019, 2022; Llera & Newman, 2014). In a questionnaire designed to assess for CA tendencies, the extent to which individuals reported discomfort with NECs, and the tendency to create or sustain negative emotion to avoid them, successfully discriminated between participants with and without GAD (AUC = 0.96, 95% CI = .931–.997;Llera & Newman, 2017). This finding was replicated in a study conducted as part of this special series (AUC = 0.81, 95% CI = 0.78–0.84; Newman, Rackoff, Zhu, & Kim, 2023).

The CAM may also help to explain why chronic worriers experience difficulties with problem solving. For example, encountering a problem may be particularly triggering if an individual is also afraid of the emotional consequences of failure. Fear of such consequences could lead to reduced confidence in solutions, dismissing solutions as not good enough, and delaying action to avoid experiencing potential failure. Furthermore, it may feel emotionally protective to focus on potential negative outcomes, which could prevent NECs if problem-solving efforts fail (i.e., they will not be taken by surprise). In this way, hyper-focus on negative outcomes, pessimism about problem-solving efforts, and delayed decision-making/solution implementation could all be considered aspects of a coping strategy that avoids NECs. In support of this interpretation, a recent study tested the CAM using an in-vivo failure experience as the negative stressor (Jamil & Llera, 2021). Individuals engaged in a challenging cognitive task, and were then told they scored in the lower 16% relative to a national sample (i.e., false failure feedback). Those who were already in a negative emotional state (via prior worry or rumination inductions) experienced reduced NEC in terms of both subjective and physiological response to feedback, compared to those in prior relaxation or neutral inductions.

In sum, CA tendencies may fuel fear of negative problem outcomes and maintain negative behaviors and attitudes toward problems (e.g., focusing on threat, reduced confidence in solutions, delayed action) as a way to protect against the emotional repercussions of failure. Unfortunately, this coping style may only exacerbate problem-solving difficulties and perpetuate distress in the face of problems for high trait worriers. Despite these conceptual links, the role of CA tendencies has never been tested in relation to problem-solving, and in an in-vivo problem-solving exercise.

1.1. The Current Study

The purpose of this study, therefore, was to determine whether CA was linked to in-vivo problem-solving difficulties, and if CA tendencies could help to explain the longstanding relationship between high trait worry and impaired problem solving. We therefore examined whether CA tendencies could predict the success of solving real-world problems in a controlled laboratory setting. We chose to measure general emotional CA tendencies as this encapsulates both discomfort with emotional shifts and broad avoidance of emotional shifts (e.g., general negativity, pessimism, etc., to maintain a negative state). We also chose personally relevant problems over standardized hypothetical problem scenarios to allow for greater applicability of findings to problem-solving outside of the lab. We observed a range of outcomes related to the problem-solving experience, including number of solutions generated, participants’ confidence in solutions, reported intention to carry out solutions, and state anxiety levels in response to the problem-solving process. We believed that these outcomes best represented the typical problem-solving impairments experienced by chronic worriers, and those that would be most affected by CA tendencies. That is, fear of the emotional consequences of failure could interfere with generating solutions, reduce confidence in solutions, inhibit proactive behaviors, and increase stress.

We also examined whether CA would mediate the association between chronic worry and problem-solving outcomes. This was to determine if CA tendencies could explain the previously established relationship between high trait worry and impaired problem solving. Our study had two main hypotheses. Our first hypothesis was that higher CA tendencies would predict more difficulties with an in-vivo problem-solving task. Our second hypothesis was that CA tendencies would act as a mediator between high trait worry and negative problem-solving outcomes, thus serving as an explanatory link for the long-standing relationship between chronic worry and problem-solving impairment.

2. Method

2.1. Participants and Measures

The current study recruited 185 volunteers from psychology courses in a public university. Students received class credit as compensation. Participants were largely young adult (M = 19.61 years, SD = 2.45, range = 18–35), cisgender (100%) females (76.8%). Participants self-identified their race as White (58.4%), African American (24.3%), Asian American (7.6%), American Indian/Pacific Islander (1.1%), and other (e.g., “mixed race”, 7%). Of these, 5.9% identified their ethnicity as Hispanic/Latinx, 71.4% as non-Hispanic, and 17.8% as “Other”.

Participants were sampled to include a broad range of chronic worry and GAD symptoms based on their scores on the Penn State Worry Questionnaire (PSWQ; Meyer et al., 1990), a 16-item self-report measure designed to assess the frequency, intensity, and uncontrollability characteristics of trait worry, and the Generalized Anxiety Disorder Questionnaire (GAD-Q-IV; Newman et al., 2002), a 9-item self-report questionnaire based on diagnostic criteria for GAD. The PSWQ demonstrates strong internal consistency (Chronbach’s α = 0.91; Meyer et al., 1990) and retest reliability (.74 – .93; Molina & Borkovec, 1994). The GAD-Q-IV also demonstrates strong internal consistency (α = 0.94) and good retest reliability (Newman et al., 2002). Internal consistency for the current sample was high for both the PSWQ (α = .95) and the GAD-Q-IV (α = .91). For the PSWQ, a cut score of 45 has been recommended to obtain optimal sensitivity and specificity for GAD compared to nonanxious controls (Behar et al., 2003). For the GAD-Q-IV, a cut score of 5.7 has been used to identify those with GAD (Newman et al., 2002). In the current sample, scores on the PSWQ ranged from 19–80 (M = 54.42, SD = 16.73), whereas scores on the GAD-Q-IV ranged from 0–12.67 (M = 6.24, SD = 3.56), indicating that this sample included participants spanning a broad range of GAD symptoms. Based on answers to the GADQ-IV, 42 participants met self-reported diagnostic criteria for GAD.

Participants also completed the Contrast Avoidance Questionnaire-General Emotion (CAQ-GE; Llera & Newman, 2017). The CAQ-GE is a 25-item self-report measure that assesses general emotional contrast avoidance beliefs and behaviors, and includes two factors: 1) creating/sustaining negative emotion to avoid negative contrasts, and 2) discomfort with emotional shifts. Items are rated using a 5-point Likert scale (1 = “not at all true” to 5 = “absolutely true”). The CAQ-GE has demonstrated strong internal consistency, construct validity, and retest reliability (Llera & Newman, 2017; Rogers et al., 2022). Internal consistency for the current sample was high (α = .95).

2.2. Procedure

For the current study we utilized data collected as part of a larger IRB approved study investigating worry and problem solving (Llera & Newman, 2020). Here, we report procedures relevant to the current study. Participants were each tested alone in a private room equipped with a computer. All instructions and tasks were completed using the Qualtrics survey platform (Qualtrics, Provo, UT). Participants first provided informed consent, and then completed demographic questions along with the PSWQ, GAD-Q-IV, and CAQ-GE. They were next instructed to identify a current, real-life problem; specifically, one that was affecting them right now, and for which they had some control over the outcome. The latter requirement was to assist in identifying a problem for which there were possible solutions, as opposed to an uncontrollable issue (e.g., a loved one’s terminal illness). They were then asked to briefly describe their problem by typing it out on the computer.

Following this task, participants were asked to generate as many solutions to their problem as they could for 2 minutes. Solutions were also typed out on the computer. Next, they were instructed to reflect on these ideas and choose their “best, most effective” solution. Participants then reported how confident they felt that this solution would be effective (confidence), as well as how likely they were to actually carry it out (intention), on a scale of 0 (not at all confident/likely) to 100 (very confident/likely). To determine the presence of any lingering anxiety, participants also rated state anxiety levels after choosing their “best” solution.

2.3. Data Analytic Plan

Data analyses were conducted through Statistical Package for the Social Sciences (SPSS; IBM Corp., Armonk, NY, United States). One participant was removed as an outlier based on Mahalanobis, Cook’s distance, and leverage analyses, leaving a total sample of N = 184. We then ran zero-order correlations to observe the relationships between all variables (see Table 1).

Table 1.

Means, Standard Deviations, and Zero-Order Correlations between Measures

Measure 1 2 3 4 5 6 7
1. PSWQ 1
2. GAD-Q-IV .842** 1
3. CAQ-GE .612** .577** 1
4. # Solutions .030 .035 .048 1
5. Confidence −.189* −.205** −.185* 0.134 1
6. Intention −.158* −.215** −.229** −0.011 .435** 1
7. State Anxiety .577** .573** .463** −0.039 −.213** −.339** 1
Mean (SD) 54.42 (16.73) 6.24 (3.56) 55.41 (20.07) 5.86 (2.63) 71.26 (23.35) 66.31 (27.21) 32.37 (27.71)
Open in a new tab

Note. PSWQ = Penn State Worry Questionnaire; CAQ-GE = Contrast Avoidance Questionnaire, General Emotion Scale; # Solutions = number of solutions generated, Confidence = self-reported confidence in effectiveness of solutions; Intention = self-reported intention to implement solution; State Anxiety = self-reported anxiety after problem-solving task;

* =

p < .05;

** =

p < .01.

To test our first hypothesis, that CA could predict impairment in an in vivo problem-solving task, we ran a series of bootstrapped linear regressions (using 1000 samples). For all models, the CAQ-GE was tested for its ability to predict the following problem-solving outcomes: 1) number of solutions generated, 2) confidence in the effectiveness of their chosen solution, 3) intention to implement their chosen solution, and 4) state-anxiety levels after choosing their solution.

To test our second hypothesis, that the CAQ-GE would mediate between the PSWQ and each of the outcome measures, we ran a series of bootstrapped mediation analyses using the PROCESS extension for SPSS (model 4). Our analyses were based on 5,000 bootstrapped samples using 95% confidence intervals.

3. Results

Means, internal consistency reliability, and bootstrapped zero-order correlations are provided in Table 1. All variables were related in predicted directions, with the exception that number of solutions generated was not significantly associated with any other variables.

Our first hypothesis was that higher CA levels would predict participants providing 1) fewer solutions, and reporting 2) less confidence in the effectiveness of their solutions, 3) lower intention to implement solutions, and 4) greater state anxiety after choosing their “best” solution. This hypothesis was largely supported, with the CAQ-GE significantly predicting all problem-solving outcomes except for number of solutions generated (see Table 2).

Table 2.

CAQ-GE predicting problem-solving outcomes

Outcomes B Bootstrapped [95% CI] β t p R2 F p
 # Solutions .006 [−.012, .025] .048 0.645 .520 .002 0.416 .520
 Confidence −.215 [−.396, −.058] −.185 −2.535 .012 .034 6.426 .012
 Intention −.310 [−.494, −.141] −.229 −3.170 .002 .053 10.051 .002
 State Anxiety .640 [.488, .797] .463 7.040 <.001 .214 49.567 <.001
Open in a new tab

Note. Bootstrap results based on 1000 samples. # Solutions = number of solutions generated, Confidence = self-reported confidence in effectiveness of solution, Intention = self-reported intention to implement solution, State Anxiety = self-reported anxiety levels after problem-solving task.

Our second hypothesis was that CA tendencies would mediate the relationship between the PSWQ and each of our outcome measures. This hypothesis was partially supported. For number of solutions generated, the indirect effect of CA was not significant, b = .005, bootstrapped SE = .01, 95% BCa CI [−.01, .02], indicating that the CAQ-GE did not mediate between PSWQ and number of solutions. Although the PSWQ significantly predicted CAQ-GE scores, neither the PSWQ nor CAQ-GE were significant predictors of number of solutions (see Figure 1). In terms of predicting confidence levels, the indirect effect of CA was also not significant, b = −.10, bootstrapped SE = .08, 95% BCa CI [−.28, .05], indicating that the CAQ-GE did not mediate between PSWQ and reported confidence in solutions. Again, neither the PSWQ nor CAQ-GE were significant predictors of confidence levels. When predicting intention to implement solutions, however, there was a significant indirect effect of CA, b = −.21, bootstrapped SE = .09, 95% BCa CI [−.38, −.05]. In this case, CAQ-GE scores significantly mediated between PSWQ and lower reported intention to implement solutions. Further, the direct effect of the PSWQ on intention was no longer significant once the CAQ-GE was taken into account. For state anxiety levels after choosing a solution, there was a significant indirect effect of CA, b = .18, bootstrapped SE = .08, 95% BCa CI [.03, .34], as well as a significant direct effect of PSWQ, b = .78, SE = .14, 95% CI [.50, 1.06]. In this case, the CAQ-GE partially mediated the association between the PSWQ and state anxiety levels, given that PSWQ remained a significant predictor of state anxiety even when CA tendencies were taken into account.1

Figure 1.

Figure 1.

Open in a new tab

General emotional CA tendencies as a mediator of the relationship between trait worry and problem-solving outcomes.

Note. PSWQ = Penn State Worry Questionnaire; CAQ-GE = Contrast Avoidance Questionnaire, General Emotion Scale; # Solutions = number of solutions generated, Confidence = self-reported confidence in effectiveness of solutions; Intention = self-reported intention to implement solution; State Anxiety = self-reported anxiety after problem-solving task; * = p < .05; ** = p < .001.

4. Discussion

This was the first study to measure the ability for CA tendencies to predict impairment in the problem-solving process when tasked with solving a personally-relevant, real-world problem. Further, this study was the first to determine if CA tendencies could mediate the longstanding relationship between trait worry and various aspects of problem-solving difficulties, thus exploring the role that CA might play in fueling such impairment. This study expands our understanding of the CAM by observing how CA tendencies were associated with problem-solving outcomes in real time.

Results from this study partially supported our hypotheses about the role of CA in problem solving. First, we found that general emotional CA (i.e., fear of emotional shifts and maintaining a negative state to avoid NECs) was able to significantly predict most problem-solving outcome variables. That is, higher CA levels predicted lower participant-reported confidence in the solutions they generated, less intention to actually carry out these solutions in the real world, and greater state anxiety levels after having chosen a solution to their problem. This suggests that those who reported more fear and avoidance of emotional contrast via sustained negative emotionality were also more likely to experience impairment when attempting to solve their problems. Furthermore, engaging in the problem-solving process may have been experienced as threatening and led to continued anxiety about the problem for these individuals. On the other hand, CA did not predict generating fewer solutions to problems during the brainstorming process.

These results largely support the hypothesized link between CA tendencies and problem-solving difficulties. Findings suggest that those high in CA may experience impairment in problem solving in the real world. This may be due to their traditional coping patterns when faced with life’s stressors; that is, they may lower their expectations of their solutions leading to success (i.e., expect the worst), delay taking action, and maintain high levels of anxiety as a way to stay emotionally braced for negative outcomes. This process is likely to exacerbate rather than resolve life’s problems, and may reinforce beliefs that problems are insurmountable and dangerous. Although we cannot determine temporal precedence in the current study design, we can nonetheless conclude that CA and problem-solving difficulties are functionally related. However, number of solutions generated was not significantly correlated with either trait worry or CA levels. Together this suggests that whereas chronic worriers are not likely to be impacted in terms of how many solutions they generate while brainstorming solutions to problems, they may still feel more pessimistic about these solutions and less likely to put them into action.

We also examined the extent to which CA helped to elucidate the relationship between chronic worry and problem-solving difficulties. Our hypotheses for this series of analyses were partially supported. Results showed that this aspect of CA mediated the link between trait worry and lower intention to implement solutions, and between trait worry and higher state anxiety after engaging in the problem-solving task. However, CA did not mediate the link between trait worry and number of solutions generated or lower confidence in solutions.

Overall, these findings indicate that general emotional CA serves as a mediator between trait worry and some aspects of problem-solving difficulties. In terms of intention to implement the chosen solution, whereas higher trait worry initially predicted lower intention to implement solutions, this path was no longer significant when taking CA tendencies into account. This suggests that for those high in trait worry, fear of NEC and efforts to maintain negative emotion to avoid a negative contrast predict lower likelihood to put their solutions into action to resolve their problems. This could be because taking any action to resolve their problem could precipitate a NEC experience, whereas putting it off could avoid NEC, at least in the short term. This suggests that for those high on trait worry, attempting to prevent the possibility of NEC may lead to inaction, which could exacerbate problems. Furthermore, the link between trait worry and greater anxiety after choosing a solution to their problem was also mediated by greater fear and avoidance of NEC. This may suggest that the extent to which those higher in trait worry feel anxious in the face of problem-solving is in part because of their fear of the emotional repercussions given that problem-solving may put them at risk of NEC. Alternatively, this may reflect a desire for those high in trait worry to maintain higher negative affect as a way to feel braced for the worst outcome. Both of these findings help to elucidate the role of general emotional CA in the link between trait worry and problem-solving distress and impairment.

On the other hand, the fact that CA did not explain the link between trait worry and lower confidence in solutions suggests that this coping style does not influence all aspects of problem-solving difficulties for chronic worriers. Trait worry still predicted higher CA tendencies, which is consistent with the CAM, but CA levels did not mediate these specific negative outcomes. One possibility is that the link between chronic worry and poor confidence in solutions may be sufficiently strong regardless of the extent to which one fears and avoids emotional shifts. That is, lower confidence in solutions may not depend on a need to feel emotionally braced for the worst, but may have more to do with other characteristics of trait worriers. For example, self-confidence levels were shown to be substantially lower than the general public for those with GAD (Arshad et al., 2020) and recent studies found that increases in GAD symptoms predicted decreases in executive functioning capabilities over a 9-year period (Zainal & Newman, 2018, 2021). This latter finding is consistent with attentional control theory (Eysenck et al., 2007), which posits that worry itself can impair executive functioning. This may in turn reduce confidence in performance on cognitive tasks such as problem solving for chronic worriers.

There were some limitations to this study. First, although we tested a mediational model in which both trait worry and CA predicted in-vivo problem-solving outcomes, due to the cross-sectional nature of our data we cannot ascribe causality or temporal precedence to our variables. Both worry and CA were conceptualized and measured as trait variables and according to the CAM, may function to reciprocally maintain each other over time. Moreover, although we observed in vivo problem-solving impairments, we likewise cannot determine that these were caused by either of our predictors. Rather, our goal in conducting a mediation model was to test whether CA tendencies could explain a significant portion of the variance between trait worry and impaired problem solving. In support of our approach, Hayes (2013) states that it is possible to conduct mediation analyses even if “one cannot unequivocally establish causality” (pg. 89) as a way to run a mathematical test of a hypothesized relationship between variables. However, to determine if our hypothesized mediation path was the best fit to the data, we tested an alternative model in which trait worry and CA were switched in each mediational model. Results showed that CA was able to mediate more relationships than was trait worry, suggesting that our proposed mediational model provided a better explanation of problem-solving outcomes. Future research may wish to tease apart the development of these different variables to determine if one may lead to the other, or if they develop in tandem; however, that research question was beyond the scope of this study.

Second, though we deliberately recruited subjects across the spectrum on GAD symptomatology, this was a non-clinical sample and therefore results may not generalize to a clinical sample with GAD. However, our research question was not exclusive to those with GAD, as high trait worry itself is linked to problem-solving difficulties. Therefore, we chose to test this research question dimensionally. Future research may wish to replicate some of these connections within a treatment-seeking GAD sample. Finally, because participants came up with their own personally-relevant problems to solve, we could not ensure equivalence in terms of the difficulty or complexity of the problem-solving task. It is possible that some participants were attempting to solve more challenging problems than other participants. Further, we cannot be sure that chosen problems were reflective of participants’ typical worry topics, meaning that reflecting on problems could have led to more or less worry depending on the topic. (However, most problems listed by participants fell under the categories of academic, interpersonal, financial, and job-related, which tend to be among the most common topics about which people worry.) All of the above could have impacted our findings; however, our goal was to create a task with high external validity, and we believe this outweighed the concern over potential heterogeneity of the problem-solving task. Future studies may wish to replicate these findings using a more uniform problem-solving exposure.

In sum, this study expanded upon our knowledge of the relationship between CA tendencies and difficulties with problem solving, and shed some light on the CAM’s role in helping to explain some of the difficulties those with high trait worry may encounter when facing life’s problems. Given the severity of functional impairment for chronic worriers, identifying mechanisms that may influence problem-solving impairment can provide critical insights into how to address these issues. Treatments aimed to address problem-solving difficulties for high worriers may be enhanced by including a focus on the fear of the emotional repercussions of some aspects of the problem-solving process. If clients are uncomfortable facing problems in part because they believe they could not cope emotionally if something went wrong with their problem-solving efforts, this could be usefully targeted in treatment. Future research may wish to explore these relationships further to continue to expand our knowledge of the relationship between CA and problem-solving deficits in the context of chronic worry.

Supplementary Material

1

Highlights:

  • Chronic worriers have difficulties with problem solving.

  • Contrast avoidance (CA) may explain impaired problem-solving for chronic worriers.

  • Higher general emotional CA predicted more difficulties with an in vivo problem-solving task.

  • CA levels mediated between trait worry and some aspects of problem-solving impairment.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declarations of interest: none

1

Given the cross-sectional nature of our trait worry and CA variables, we wanted to confirm whether our theoretical pattern of relationships represented the best fit to the data. To do so, we tested a rival model in which trait worry and CA were switched in each mediational model, and re-ran all mediation analyses to test this approach. Whereas most results were comparable, trait worry did not mediate between CA and intention to implement solutions (see Figure 1 in Supplemental Material). This suggests that CA was a superior mediator to trait worry, significantly mediating more outcomes than did trait worry. As such, we feel more confident that our theoretical pattern of relationships was valid, and that trait worry did not better account for the link between CA and problem-solving difficulties.

References

  1. Anderson RJ, Goddard L, & Powell JH (2009). Social problem-solving processes and mood in college students: An examination of self-report and performance-based approaches. Cognitive Therapy and Research, 33(2), 175–186. 10.1007/s10608-007-9169-3 [DOI] [Google Scholar]
  2. Arshad S, Shakoor A, Sadaf B, & Ali S (2020). Emotion regulation and self-esteem among patients with generalized anxiety disorder. Psychosocial explorations of gender in society. In Kamal A, Masood S, Hanif R, Jami H, & Zubair A (Eds.), Psychosocial explorations of gender in society. (pp. 144–151). Cambridge Scholars Publishing. [Google Scholar]
  3. Borkovec TD (1985). Worry: A potentially valuable concept. Behaviour Research and Therapy, 23(4), 481–482. 10.1016/0005-7967(85)90178-0 [DOI] [PubMed] [Google Scholar]
  4. Crouch TA, Lewis JA, Erickson TM, & Newman MG (2017). Prospective investigation of the contrast avoidance model of generalized anxiety and worry. Behavior Therapy, 48(4), 544–556. 10.1016/j.beth.2016.10.001 [DOI] [PubMed] [Google Scholar]
  5. D’Zurilla TJ, & Goldfried MR (1971). Problem solving and behavior modification. Journal of Abnormal Psychology, 78(1), 107–126. 10.1037/h0031360 [DOI] [PubMed] [Google Scholar]
  6. Dugas MJ, Freeston MH, & Ladouceur R (1997). Intolerance of uncertainty and problem orientation in worry. Cognitive Therapy and Research, 21(6), 593–606. 10.1023/A:1021890322153 [DOI] [Google Scholar]
  7. Dugas MJ, Gagnon F, Ladouceur R, & Freeston MH (1998). Generalized anxiety disorder: A preliminary test of a conceptual model. Behaviour Research and Therapy, 36(2), 215–226. 10.1016/S0005-7967(97)00070-3 [DOI] [PubMed] [Google Scholar]
  8. Eysenck MW, Derakshan N, Santos R, & Calvo MG (2007). Anxiety and cognitive performance: Attentional control theory. Emotion, 7(2), 336–353. 10.1037/1528-3542.7.2.336 [DOI] [PubMed] [Google Scholar]
  9. Fergus TA, Valentiner DP, Wu KD, & McGrath PB (2015). Examining the symptom-level specificity of negative problem orientation in a clinical sample. Cognitive Behaviour Therapy, 44(2), 153–161. 10.1080/16506073.2014.987314 [DOI] [PubMed] [Google Scholar]
  10. Goodwin H, Eagleson C, Mathews A, Yiend J, & Hirsch C (2017). Automaticity of attentional bias to threat in high and low worriers. Cognitive Therapy and Research, 41(3), 479–488. 10.1007/s10608-016-9818-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hayes AF (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press. [Google Scholar]
  12. Hirsch CR, Meeten F, Krahe C, & Reeder C (2016). Resolving Ambiguity in Emotional Disorders: The Nature and Role of Interpretation Biases. Annual Review of Clinical Psychology, 12, 281–305. 10.1146/annurev-clinpsy-021815-093436 [DOI] [PubMed] [Google Scholar]
  13. Jamil N, & Llera SJ (2021). A transdiagnostic application of the contrast-avoidance model: The effects of worry and rumination in a personal-failure paradigm. Clinical Psychological Science, 9(5), 836–849. 10.1177/2167702621991797 [DOI] [Google Scholar]
  14. Kim H, & Newman MG (2019). The paradox of relaxation training: Relaxation induced anxiety and mediation effects of contrast avoidance in generalized anxiety disorder and major depressive disorder. Journal of Affective Disorders, 259, 271–278. 10.1016/j.jad.2019.08.045 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kim H, & Newman MG (2022). Avoidance of a negative emotional contrast from worry and rumination: An application of the contrast avoidance model. Journal of Behavioral and Cognitive Therapy, 32(1), 33–43. 10.1016/j.jbct.2021.12.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ladouceur R, Blais F, Freeston MH, & Dugas MJ (1998). Problem solving and problem orientation in generalized anxiety disorder. Journal of Anxiety Disorders, 12(2), 139–152. 10.1016/S0887-6185(98)00002-4 [DOI] [PubMed] [Google Scholar]
  17. LaFreniere LS, & Newman MG (2019). Probabilistic learning by positive and negative reinforcement in generalized anxiety disorder. Clinical Psychological Science, 7(3), 502–515. 10.1177/2167702618809366 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. LaFreniere LS, & Newman MG (2020). Exposing worry’s deceit: Percentage of untrue worries in generalized anxiety disorder treatment. Behavior Therapy, 51(3), 413–423. 10.1016/j.beth.2019.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Llera SJ, & Newman MG (2010). Effects of worry on physiological and subjective reactivity to emotional stimuli in generalized anxiety disorder and nonanxious control participants. Emotion, 10(5), 640–650. 10.1037/a0019351 [DOI] [PubMed] [Google Scholar]
  20. Llera SJ, & Newman MG (2014). Rethinking the role of worry in generalized anxiety disorder: Evidence supporting a model of Emotional Contrast Avoidance. Behavior Therapy, 45(3), 283–299. 10.1016/j.beth.2013.12.011 [DOI] [PubMed] [Google Scholar]
  21. Llera SJ, & Newman MG (2017). Development and validation of two measures of emotional contrast avoidance: The Contrast Avoidance Questionnaires. Journal of Anxiety Disorders, 49, 114–127. 10.1016/j.janxdis.2017.04.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Llera SJ, & Newman MG (2020). Worry impairs the problem-solving process: Results from an experimental study. Behaviour Research and Therapy, 135, 103759. 10.1016/j.brat.2020.103759 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. MacLeod C, & Donnellan AM (1993). Individual differences in anxiety and the restriction of working memory capacity. Personality and Individual Differences, 15(2), 163–173. 10.1016/0191-8869(93)90023-V [DOI] [Google Scholar]
  24. Meyer TJ, Miller ML, Metzger RL, & Borkovec TD (1990). Development and validation of the Penn State Worry Questionnaire. Behaviour Research and Therapy, 28(6), 487–495. 10.1016/0005-7967(90)90135-6 [DOI] [PubMed] [Google Scholar]
  25. Newman MG, Jacobson NC, Zainal NH, Shin KE, Szkodny LE, & Sliwinski MJ (2019). The effects of worry in daily life: An ecological momentary assessment study supporting the tenets of the contrast avoidance model. Clinical Psychological Science, 7(4), 794–810. 10.1177/2167702619827019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Newman MG, & Llera SJ (2011). A novel theory of experiential avoidance in generalized anxiety disorder: A review and synthesis of research supporting a Contrast Avoidance Model of worry. Clinical Psychology Review, 31(3), 371–382. 10.1016/j.cpr.2011.01.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Newman MG, Llera SJ, Erickson TM, Przeworski A, & Castonguay LG (2013). Worry and generalized anxiety disorder: A review and theoretical synthesis of research on nature, etiology, and treatment. Annual Review of Clinical Psychology, 9(1), 275–297. 10.1146/annurev-clinpsy-050212-185544 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Newman MG, Rackoff GN, Zhu Y, & Kim H (2023). A transdiagnostic evaluation of contrast avoidance across generalized anxiety disorder, major depressive disorder, and social anxiety disorder. Journal of Anxiety Disorders, 93, Article 102662. 10.1016/j.janxdis.2022.102662 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Newman MG, Schwob JT, Rackoff GN, Shin KE, Kim H, & Van Doren N (2022). The naturalistic reinforcement of worry from positive and negative emotional contrasts: Results from a momentary assessment study within social interactions. Journal of Anxiety Disorders, 92, Article 102634. 10.1016/j.janxdis.2022.102634 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Newman MG, Zuellig AR, Kachin KE, Constantino MJ, Przeworski A, Erickson T, & Cashman-McGrath L (2002). Preliminary reliability and validity of the Generalized Anxiety Disorder Questionnaire-IV: A revised self-report diagnositic measure of generalized anxiety disorder. Behavior Therapy, 33(2), 215–233. https://doi.org/doi: 10.1016/S0005-7894(02)80026-0 [DOI] [Google Scholar]
  31. Pawluk EJ, Koerner N, Tallon K, & Antony MM (2017). Unique correlates of problem solving effectiveness in individuals with generalized anxiety disorder. Cognitive Therapy and Research, 41(6), 881–890. 10.1007/s10608-017-9861-x [DOI] [Google Scholar]
  32. Peasley-Miklus C, & Vrana SR (2000). Effect of worrisome and relaxing thinking on fearful emotional processing. Behaviour Research and Therapy, 38(2), 129–144. 10.1016/S0005-7967(99)00025-X [DOI] [PubMed] [Google Scholar]
  33. Robichaud M, & Dugas MJ (2005). Negative problem orientation (Part II): construct validity and specificity to worry. Behaviour Research and Therapy, 43(3), 403–412. 10.1016/j.brat.2004.02.008 [DOI] [PubMed] [Google Scholar]
  34. Rogers TA, Gorday JY, Bardeen JR, & Benfer N (2022). Examining the factor structure and incremental utility of the Contrast Avoidance Questionnaires via bifactor analysis. Journal of Personality Assessment, 1–11. 10.1080/00223891.2022.2081921 [DOI] [PubMed] [Google Scholar]
  35. Stefanopoulou E, Hirsch CR, Hayes S, Adlam A, & Coker S (2014). Are attentional control resources reduced by worry in generalized anxiety disorder? Journal of Abnormal Psychology, 123(2), 330–335. 10.1037/a0036343 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sugiura Y (2007). Responsibility to continue thinking and worrying: Evidence of incremental validity. Behaviour Research and Therapy, 45(7), 1619–1628. 10.1016/j.brat.2006.08.001 [DOI] [PubMed] [Google Scholar]
  37. Szabó M, & Lovibond PF (2002). The cognitive content of naturally occurring worry episodes. Cognitive Therapy and Research, 26(2), 167–177. 10.1023/A:1014565602111 [DOI] [Google Scholar]
  38. Szabó M, & Lovibond PF (2006). Worry episodes and perceived problem solving: A diary-based approach. Anxiety, Stress & Coping, 19(2), 175–187. 10.1080/10615800600643562 [DOI] [Google Scholar]
  39. Tallis F, Eysenck M, & Mathews A (1991). Elevated evidence requirements and worry. Personality and Individual Differences, 12(1), 21–27. 10.1016/0191-8869(91)90128-X [DOI] [Google Scholar]
  40. Zainal NH, & Newman MG (2018). Executive function and other cognitive deficits are distal risk factors of generalized anxiety disorder 9 years later. Psychological Medicine, 48(12), 2045–2053. 10.1017/S0033291717003579 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zainal NH, & Newman MG (2021). Within-person increase in pathological worry predicts future depletion of unique executive functioning domains. Psychological Medicine, 51(10), 1676–1686. 10.1017/S0033291720000422 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1867196-supplement-1.docx (52.4KB, docx)

RESOURCES