Parameters of anger as related to sensory-affective components of pain

Pain definition

Pain definitions have undergone successive refinements over the last century [1], and across these variations there has been a consistent reference to sensation and emotion in pain. Thus, the International Association for the Study of Pain (IASP) recently proposed that pain is: “an unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage” [2]. The emotional or affective component of pain has been further elaborated in the latest updated International Classification of Diseases (ICD-11), which states that chronic primary pain is “characterized by significant emotional distress (anxiety, anger/frustration or depressed mood) or functional disability (interference in daily life activities and reduced participation in social roles)” [3]. In other words, chronic pain is not just aversive or unpleasant in a general sense but is associated with specific types of distress such as anxiety, depression, and anger or, for that matter, any type of negative affect (e.g., guilt, shame, regret).

Anger in the affective component of pain

Of the many types of negative affect possibly associated with pain, anxiety and depression have attracted much systematic research; anger has been the subject of clinical observations and anecdotes but remains relatively under-researched in this population [4, 5]. One recent striking illustration of anger in pain patients is that of a 45-year-old man in the US. He complained of persistent back pain following surgery and upon failing to get additional care, he blamed and fatally shot the physician who had treated him plus three others at the hospital before committing suicide [6]. While such rage by no means typifies chronic pain patients, it saliently points to the comorbidity of chronic pain and anger as a priority for research and an imperative for treatment.

Association between anger and pain

In a classic study involving the World Health Organization, 2,400 schoolchildren ages 11–12 and 15–16 were randomly sampled in Iceland [7]. They completed a 256-item questionnaire about pain, distress, attitudes, and behaviors. The most common concern turned out to be anger, affecting 76.5% of those with weekly back pain, 76.9% of those with weekly headaches, and 78.3% of those with weekly stomach pain. Moreover, anger increased with pain frequency and the number of pains, and it outstripped anxiety and other aspects of negative affect.

In an adult population of chronic pain patients, 69% of chronic pain patients were found to be angry at someone or another [8]. The intensity of this anger was, on average, in the moderate range of a 0–10 scale.

Numerous studies have gone beyond frequency data to standardized measures of anger in relation to various chronic pain syndromes. A narrative review of these studies has led to the conclusion that those reporting higher levels of anger report higher levels of pain [9].

Aggregation of quantitative data from at least two decades of research on anger and pain has now been accomplished in a meta-analysis [10]. Most of the 20 studies analyzed utilized the State-Trait Anger Expression Inventory [11]. As reported, pain was significantly related to state anger, r=0.44, and much less so to trait anger, r=0.18. The correlation coefficient between non-specific anger measures and pain/disability was +0.35.

The anger-pain association may depend on the particular measures chosen. The STAXI essentially operationalizes state anger as what is felt “momentarily” and trait anger as how often certain behaviors occur in general [11]. It is one of 16 anger assessment tools that have been critically reviewed from a psychometric standpoint [12]. Some of these alternative assessment tools offer additional perspectives for assessing anger in pain patients [13].

Rationale

The Anger Parameters Scale (APS) is premised on the view that anger is measurable not only according to how often it occurs. Like any emotion, anger is quantifiable in terms of five parameters: frequency (number of occurrences), duration (length of time), intensity (magnitude), latency (immediacy), and threshold (sensitivity) [14]. As further detailed in the Method section, the APS is a self-report questionnaire for measuring each of these parameters. Additionally, as defined earlier, pain can be dichotomized into sensory vs. affective components [2]. As detailed in the Method section, these can be measured using unidimensional rating scales. Regarding how anger parameters might be related to pain components, our approach was to look beyond co-occurrence to the actual strength of association. Although the pain components were expected to be closely coupled, it was hypothesized that anger scores (individually and collectively) would be correlated more strongly with the affective than the sensory component of pain.

Participants

Participants were volunteers who met criteria for a study approved by the Institutional Review Board of the authors’ institution. The criteria for inclusion were (a) having chronic pain (minimum 6 months), (b) recent/current use of analgesic medication (minimum 3 months) in the past 12 months, (c) having anger related to pain or other circumstances as indicated by mid-range scores on STAXI, (d) speaking English, (e) having reliable access to phone and internet, and (f) being between the ages of 18–65. Furthermore, exclusion criteria were: (a) serious psychiatric illness (e.g., autism, schizophrenia, or dementia), (b) current or recent suicidal thoughts, (c) currently taking anti-psychotic medication, and d) currently breastfeeding (as stipulated by the Institutional Review Board because of the possible pain and non-use of analgesics during the weeks postpartum).

A total of 1,061 participants responded to our advertisements for volunteers, of which 184 met criteria for inclusion while not meeting criteria for exclusion. Of those passing the screening process, 82 participants expressed interest in further participation in a subsequent study on the treatment outcome evaluation of anger and they were sent the study questionnaire with items relating to pain, anger, and demographic variables (gender, race/ethnicity, and educational level). As a precursor, this study was designed to address the question of the strength of relationship between pain components and anger parameters, whereas the subsequent treatment outcome study was for the specific purpose of evaluating a new psychological treatment program that had been developed for regulating anger.

Subsequently, 28 participants were dropped due to non-communication or non-completion, resulting in 54 participants who completed the study requirements. Of these, 3 participants were excluded (with consensus from two members of the research team) because of missing or contradictory responses to questions about pain and anger. This yielded a final sample of n=51 participants with analyzable data for this precursor study on the relationship between anger and pain. This sample size was deemed sufficient for 70% power in detecting a medium-sized correlation. The demographic breakdown of the sample is shown in Table 1.

Measures

Anger was measured using the Anger Parameters Scale (APS) as described earlier [14]. This self-report questionnaire is designed to assess five parameters of anger: frequency, duration, intensity, latency, and threshold. Frequency refers to how often an individual gets angry, duration refers to how long an individual stays angry, intensity refers to how strong the individual’s anger is, latency refers to how quickly the individual’s anger occurs in response to provocation, and finally, threshold refers to how sensitive to provocation the individual is.

The APS has 30-items organized into six items per parameter. Sample items are as follows: Frequency: “Anger is my least common emotion”; Duration: “My anger is prolonged”; Intensity: “I explode with anger”; Latency: “I am slow to anger”; Threshold: “A single provocation is enough to anger me.” Each statement is meant to be rated on a 0 to 4 scale (fractional ratings permitted), 0 meaning the statement is absolutely false or inapplicable to the participant and 4 meaning that it is absolutely true or totally applicable to the participant. Half of the 30 items of the APS are negatively keyed to control for response sets and therefore they are reverse-scored. For each parameter, scores range from 0 to 24, the total score for the instrument having a maximum value of 120. For individual parameters and for the total scores, higher values indicate greater maladaptiveness [14].

Evidence has confirmed the factorial structure of the APS [15]. Internal consistency as indexed by Cronbach’s alpha=0.85 for frequency, 0.90 for duration, 0.62 for intensity, 0.88 for latency, and 0.74 for threshold [16].

To assess pain, two separate unidimensional scales were employed, one pertaining to the physical sensation of pain and the other pertaining to the degree of distress from pain. The scales were anchored at 0 for absence of the phenomenon being rated, 1 for minimally detectable level of the phenomenon being rated, and 10 for maximum level of the phenomenon being rated. This method of twin unidimensional scales for sensory vs. affective pain has been successfully used in a long line of research [17, 18].

Procedures

Participants were recruited largely through posts on the following social media platforms: Facebook, Reddit, Twitter, and Instagram. Electronic announcements were also emailed to various organizations that expressed willingness to disseminate flyers regarding the study. These included chronic pain support groups, police interest groups, warehouse worker groups/unions, construction/manual labor groups/unions, and veteran support groups including the American Legion Posts.

Those interested were directed to an online Qualtrics survey so that informed consent was electronically obtained before participants proceeded to answer specific questions pertaining to eligibility. To ascertain anger, participants were asked to complete the 10 items making up the trait section of the STAXI-2. To be eligible for inclusion, the participant had to score at least 19 out of a maximum of 40 on this subscale. This would put their score in the mid-range of trait anger.

Research assistants screened participants for inclusion and exclusion criteria. Participants who passed the screening process were then granted access to the main study questionnaire containing the APS and scales for measuring pain. The questionnaires were hosted on Qualtrics and disseminated by email as participants became eligible.

Limitations

The present study is limited by the size and composition of the sample. Of n=51, 78% were female, and 57% identified as “White”. Although these are representative of the gender and ethnicity/race of the population from which participants were drawn, our findings would need to be replicated in a much larger and demographically heterogenous sample before they could be deemed generalizable.

The methodology of the study is essentially correlational. Causal inferences are not justifiable at this stage. At least five major interactions between negative affect and pain might be entertained [4, 24]:- firstly, anger may precede pain as a predisposing factor that exerts a long-term influence on the latter; secondly, anger may be a precipitating factor that suddenly triggers pain; thirdly, anger may not be an antecedent but a consequence of pain; fourthly anger may be an exacerbating factor that augments pain intensity; and lastly, anger may be a perpetuating factor that prolongs the duration of pain. In order to tease apart which of these interactions holds, other methods including experimental manipulation would be applicable. As it stands, the present findings do build on the existing literature on co-prevalence of anger and pain by measuring the strength of association between the two, but the findings do not explicate the more intriguing question of how each dynamically interacts with the other. Tackling that would entail other methodological designs such as prospective studies, longitudinal observations, and time series analyses.