*3.2. Participants*

Among the 15 selected articles, 7 used the Stroop task for patients with general chronic pain [9,69,74–78], while 3 used it for FMS patients [2,25,26], 2 for CLBP patients [1,28], 1 for migraine patients [7], 1 for TMD patients [27], and 1 for CNP patients [73].

The total study sample (*n* = 677) was divided into two main groups: a clinical group (*n* = 386) and a control group (with slightly fewer participants; *n* = 291). The clinical group included 11 women with CNP (age range: 48–57 years) [73] and 17 with TMDs (age range: 34–36 years) [27], 17 participants with migraine (85% women) [7], 19 women, 11 men and 25 participants of unspecified gender with CLBP (age range: 35–49 years) [69,74], 77 women with FMS (age range: 47–54 years) [2,25,26], 33 participants of unspecified gender with chronic back and/or neck pain (age range: 30–36 years), and 94 women, 50 men and 33 participants of unspecified gender with general chronic pain (age range: 38–83 years) [9,69,74–78].

It should be noted that two of the studies did not have a control group [1,28]. Regarding subjects' sex, there were more female than male participants [2,7,9,25–28,73,74,76]. Nevertheless, five studies included both men and women [7,9,28,74,76]. Notably, none of the reviewed studies included a sample composed entirely of men. Further, five studies did not provide information about the sex of the participants [1,69,75,77,78].

In the selected articles, the majority of the chronic pain participants did not have comorbid psychiatric illnesses (e.g., depression or anxiety) [1,7,9,25–28,74,75,77]. Furthermore, some of them used these conditions as exclusion criteria [7,25–27,69,74,76,77]. Other studies considered these conditions as a symptom of chronic pain, especially in FMS [2,25]. However, these conditions could negatively influence emotional Stroop task responses [76,82].

#### *3.3. Quality of Selected Studies*

The quality assessment was conducted independently by two researchers (L.A.-D. and C.I.M.-A.), and the initial agreement was 93%. To achieve a consensus, the interpretation and monitoring of the criteria was discussed with a third reviewer (C.M.S.-G.). This quality assessment was focused on the analysis of the limitations of the selected studies.

The authors of the reviewed studies indicated various limitations of their research, such as small sample and effect sizes [2,7,9,26,73–75,77], an absence of a neutral category of words [28], issues with the methods and/or criteria used for the diagnosis of chronic pain (e.g., ACR criteria for FMS and IHS criteria for migraine) [2,75], and with pharmacological treatments [2,25,69,75], non-control of the medication status (pharmacological) of the patients and healthy controls [75], low statistical power and non-inclusion of an additional experimental condition for the Stroop task [26], non-inclusion of a masked version of the test and failure to record each participant's pain level at the time of the test [74], use of non-specific stimuli [1,7,9,28,73,74], non-inclusion of additional measures of psychological distress to improve construct validity [77], a non-pragmatic approach to the study patients [69], failure to assess sensitivity to anxiety or fear of pain, failure to screen for psychiatric disturbances through screening interviews, and non-inclusion of a pain comparison group [9], and failure to control for the effect of some variables (e.g., time since surgery, dose of chemotherapy received, type of chemotherapy drugs, current medications and menopausal) [73].

Additional limitations were identified during this review, including non-randomization of participants to different groups in most studies [1,2,7,9,25–28,69,74–78], the absence of a control group to compare the results [1,28], non-blinding of participants, personnel and outcome assessments [1,2,7,9,25–28,69,74–78], non-specification of the criteria used to diagnose the disease [69,73,74,77], failure to report effect size measures [1,7,25,27,28,69,73,75–78], failure to indicate the sex ratio in some studies [1,69,75,77,78], and failure to report analyses by sex [69,76]. Moreover, some studies did not report the mean age [7,69,76] or standard deviation of their sample [69,76], and provided incomplete data on the task performed (e.g., failure to disclose the total duration of the task) [1,7,9,27,28,69,77].

#### **4. Discussion**

The present systematic review aimed to analyze studies that used an emotional Stroop task in patients with chronic pain, and assessed associated alterations of specific brain regions and behavioral performance (e.g., attentional biases). In general, and as reported in the literature, the emotional Stroop task proved to be a valid tool to assess emotional and pain processing in patients with chronic pain [2,9,69]. Most studies reported the activation of certain brain regions (the somatosensory region, and cingulate and prefrontal cortices, among other regions) during the emotional Stroop task; these regions are related to pain and emotional regulation in patients with chronic pain [27,69,73].

First, patients' performance in the emotional Stroop task, as well as the presence of attentional biases, will be discussed, followed by a brief overview of the brain areas showing neural activation in relation to the performance of the emotional Stroop task. Finally, the benefits and effects of psychological therapies that can reduce the neural activation observed in patients with chronic pain will be discussed.

Regarding performance on the emotional Stroop task, greater processing of negative and/or positive words was observed in patients with FMS, suggesting the existence of an underlying interference process, triggered by events capable of immediately capturing attention (i.e., those conveying affective meaning) [78,81]. Studies such as that of Algom et al. [83] indicate that this interference effect in the emotional Stroop task is mediated by pre-attentive inhibition, associated with the threat of negative emotional stimuli presented during the task. However, this inhibition mechanism is considered to be independent from that of selective attention [83]. In FMS patients, delayed responses to pain words were associated with pain-specific anxiety and cognitive interference, as well as low sensitivity to anxiety [75]. Some studies indicated that the slowness in color naming during the emotional Stroop task seen in FMS patients is associated with the presence of a generalized hypervigilance response [12,26]. This response is associated with a tendency for FMS subjects to be slower with respect to the color naming of symptomatic (pain-related words) and arousing negative words, depending on the degree of perceived unpleasantness [9,26]. However, it is suggested that, in larger samples, more significant interactions between patient and control groups would be seen, and that it is necessary to compare these findings with those for other diseases [9,12,26,69]. In patients with CLBP and FMS, attentional bias to sensory pain words was associated with the emotional load of the words presented in the emotional Stroop task [1,2,9,74]. This provides clear evidence of the presence of emotion-driven selective attention in FMS and CLBP [1,2,30,84]. In fact, the existence of attentional bias towards negative information seems to play an important mediating role in the relationship between a negative affective state and heightened pain [2,30,84]. In the study by Duschek et al. [2], such attentional bias was also observed in patients with FMS; they showed a specific bias towards negative information, which led to an increase in pain intensity. This further supports the findings of the literature reviewed herein. In CLBP, attentional bias was even greater in the context of words related to back pathology, and in association with increased pain intensity [1,76]. However, the causal nature of the relationship between attentional bias and pain could not be established, as most of the included studies used a cross-sectional design. On the other hand, there are data showing that individuals with greater attentional bias towards negative affective stimuli (i.e., words associated with pain) may be more prone to chronic pain symptoms [85]. In fact, attentional bias in these individuals may be a risk factor for the development of chronic pain and could also serve as a prognostic factor [71]. Attentional bias has been consistently linked to individuals' anticipation and/or experience of pain across different chronic pain conditions [70,85].

In terms of neuronal activation, in patients with chronic pain in general, greater activation was observed when performing the emotional Stroop task [69]. Compared to the healthy group, greater activation in the anterior cingulate cortex, insula, and the primary and secondary somatosensory cortex was seen [69]. More specifically, pain-related words in the Stroop task were associated with significant differences between chronic pain patients and healthy controls, in terms of activation of the pain-processing centers of the brain (i.e., the anterior cingulate cortex, insula, parietal operculum, and the primary and secondary somatosensory cortices) [11,69]. Greater activation of brain areas related to attention, cognition, and motor planning in patients with TMDs compared to controls was also found [27]. TMD subjects showed increased task-evoked responses in prefrontal, lateral, and inferior parietal areas, as well as in the amygdala, pregenual anterior cingulate, primary motor areas, and the medial prefrontal and posterior cingulate areas [27,86]. In addition, patients also showed dissociations with respect to the activity of the prefrontal cortex and cingulate, and of the amygdala and cingulate, which are normally correlated [27,86–88]. Hence, the prominence of chronic pain (which requires attention) and slow behavioral responses may be explained by attenuated, or slow and/or desynchronized, recruitment of attentional processing areas [27,86–88].

Some of the studies reviewed herein focused on specific psychological therapies, such as the mindfulness-based stress reduction technique, which yielded a significant reduction in brain activity in regions related to pain, emotional regulation, and cognitive processing (i.e., regions in the left somatosensory cortex, left precuneus, and left dorsolateral prefrontal cortex) in patients with CNP, using the emotional Stroop task as a measure of emotional reactivity [73]. This demonstrates the impact of psychological therapy on the neural correlates of pain processing and attention [73]. Mindfulness-based psychological therapies seem to be a viable complementary treatment for people suffering from CNP [73,89]. Indeed, the reduction in cerebral activity observed after mindfulness treatment suggested that the emotionally charged words presented during the task had a diminished capacity to capture attention after the therapy compared to before the therapy [90,91]. Thus, the application of this technique reduces brain activation and pain perception, where trait mindfulness is a major component of the therapy [90,91]. However, to draw firm conclusions, longitudinal studies regarding the effect of this type of psychological therapy on patients with chronic pain are needed. Moreover, other therapies, such as cognitive behavioral therapy, have been used in patients with TMDs to effectively reduce the abnormal neuronal and brain activation seen in patients after performing the emotional Stroop task [27]. Likewise, in patients with FMS, cognitive therapy for chronic pain has focused on reducing the negative attentional bias exhibited by these patients [2]. The self-control strategies involved in this therapy promote conscious withdrawal of attention from dysfunctional cognitions and possible stressors, such as emotionally charged negative words, after the application of the Emotional Stroop task [2]. Similarly, techniques such as attention training, focusing, and exposure (cognitive behavioral therapy) have proven useful in patients with FMS, to reduce the activation of emotional and pain processing areas after the application of the emotional Stroop task [26]. This therapy also reduces the hypervigilance exhibited by patients with FMS, and attentional bias to negative emotions [26].

An important limitation of the present review is that the majority of the sample was female [2,7,9,25–28,73,74,76], where the overall gender ratio of the studies was not equal. However, as previously noted, chronic pain is more prevalent in females, in whom it also tends to be overdiagnosed [34], and so studies frequently include a larger female sample. Another limitation is the lack of information on effect sizes [2,7,9,26,73–75,77]; this lack of information on the magnitude of the differences found limited the interpretability of the results. In addition, some of the studies did not specify the clinical criteria used to diagnose the different types of chronic pain [2,75], which calls into question whether the diagnoses were made on the basis of valid criteria. Furthermore, only one of the selected studies did not obtain statistically significant results [77]. A possible explanation for this may be insufficient sample sizes, which tend to preclude large variability in the results. In addition, a more accurate study quality assessment tool will be necessary for future studies. To further elaborate on the results obtained by each study, future reviews could perform a meta-analysis and also compare the findings with those of other emotional tasks, such as the dot-probe or spatial cueing task. Finally, to overcome the absence of effect sizes in the studies [1,7,25,27,28,69,73–78], calculation (and pooling) of Cohen's d or standardized measures of means would be useful (although this is more crucial and typical for meta-analyses).

The main strength of the present review was that it strictly followed a systematic methodological approach in accordance with the study protocol, which was previously registered in PROSPERO, and was prepared in accordance with the updated PRISMA guidelines [72]. Further, in terms of the thematic focus of the systematic review, this is the first review to relate the emotional Stroop task to chronic pain.

In light of the findings of the present review and the analyzed literature, continuing to examine the efficacy of the emotional Stroop task in patients with chronic pain is of high clinical relevance. Future research on chronic pain and the emotional Stroop task should aim to uncover neurobiological correlates in chronic pain patients during performance of the task. Once the precise neuroanatomical correlates underlying the disease are known, specific and integrated research and/or intervention protocols can be established to improve health-related quality of life. Given the negative attentional biases that chronic pain patients exhibit during the performance of the emotional Stroop task, a treatment aimed at the conscious redirection of attention against negative aspects could be implemented, along with relaxation techniques, modification of beliefs about pain, enhancement of coping skills, and targeted treatment of anxiety and/or depression [66]. Future research aiming to establish a relationship between attentional bias and the anticipation and/or experience of pain would be also useful to identify individuals at risk of developing chronic pain, as well as prognostic factors. Psychological treatments can be as effective as surgery for alleviating chronic pain symptoms, by altering the central processing of pain sensation [66]. In this sense, another therapy suitable for chronic pain patients is Acceptance and Commitment Therapy (ACT), the mindfulness component of which is the basis of mindfulness therapy. This therapy has proven effective for people with chronic pain [92–95]. In addition, chronic pain patients with a history of psychosocial trauma may benefit from exposure and emotional processing techniques, which have proven effective [96]. Following the application of the emotional Stroop task, studies suggest that the use of other psychological therapies may be beneficial in reducing brain activation (e.g., the cingulate, amygdala, and medial prefrontal cortex) [2,26,27]. Specifically, Weissman-Fogel et al. [27] suggested that cognitive behavioral therapy could be effective for reducing brain activation after the application of the emotional Stroop task in patients with TMDs. Unfortunately, there are currently no studies that have evaluated the effectiveness of the aforementioned techniques in reducing activation in these areas (i.e., the cingulate, amygdala, and medial prefrontal cortex). Nonetheless, such therapy (e.g., self-control, focusing, exposure, and attentional training) and cognitive therapy in patients with FMS can reduce hypervigilance towards negative stimuli with emotional content [2,26].

Numerous models of the development and/or maintenance of chronic pain suggest that attentional biases are important therein [97,98]. These models support the findings of the present systematic review. Furthermore, each model attributes different roles to attentional processes [97,98]. However, they all make the same assumption; people pay excessive attention to painful stimuli when they experience pain and feel fearful of, or threatened by, pain [97,99]. Indeed, as it was mentioned above, there is an attentional bias towards negative stimuli in people with chronic pain [2]. Furthermore, increased attention to threat/negative cues has been observed in patients with other conditions apart from chronic pain (e.g., anxiety disorders and post-traumatic stress disorder) [9,73]. This bias might be explained by the Threat Interpretation Model [100], which suggests a relationship between threat, threat interpretation, and stimuli, through the vigilance– avoidance hypothesis [100]. This hypothesis states that individuals usually pay more attention to threat stimuli and this attentional bias is usually accompanied by an avoidance of negative/threat stimuli [100]. This pattern of vigilance–avoidance may be variable across individuals, as the interpretation of stimuli may differ according to the degree of importance assigned to the perceived stimuli [100]. Furthermore, this model may generate verifiable predictions about the role of attentional processes, and how they are influenced by interpretations [100], which would have a positive impact on clinical practice (i.e., the development and improvement of chronic pain treatments).

In conclusion, after performing the emotional Stroop task, specific brain areas (e.g., the prefrontal cortex, somatosensory cortex, cingulum, and amygdala) related to emotional and pain processing are activated in patients with chronic pain (FMS, migraine, CNP, TMDS, and CLBP). During the task, chronic pain patients showed longer reaction times and delayed responses to words with negative emotional content. They also showed attentional biases towards pain sensory words. Therefore, the use of psychological therapies (e.g., mindfulness, cognitive, and cognitive behavioral therapies) will help reduce the brain activation and attentional bias produced by the emotional Stroop task in these patients.

**Author Contributions:** Conceptualization, L.A.-D., C.I.M. and C.M.G.-S.; methodology, L.A.-D., C.I.M. and C.M.G.-S.; formal analysis, L.A.-D., C.I.M. and C.M.G.-S.; resources, L.A.-D., C.I.M., L.R.F.-J. and C.M.G.-S.; data curation, L.A.-D., C.I.M. and C.M.G.-S.; writing—original draft preparation, L.A.-D., C.I.M. and C.M.G.-S.; writing—review and editing, L.A.-D., C.I.M., L.R.F.-J. and C.M.G.-S.; visualization, L.A.-D., C.I.M., L.R.F.-J. and C.M.G.-S.; supervision, C.I.M. and C.M.G.-S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

