Development of a measure to assess acceptance of headache: The Headache Acceptance Questionnaire (HAQ)

Abstract

Background

Disability resulting from headache disorders is attributable in part to avoidant coping. Acceptance of pain connotes a willingness to experience pain in the service of life values, such that meaningful activities and goals are pursued despite pain. Acceptance facilitates positive health outcomes but has rarely been investigated in headache. Because headache disorders manifest differently than other forms of chronic pain, the present study sought to develop and validate a measure of acceptance of headache.

Methods

Forty-five candidate items were developed and, with input from an expert panel, reduced to 24 items. Five items were eliminated following administration to a development sample and exploratory factory analysis. Nineteen items were administered to a validation sample for confirmatory factory analysis and assessment of psychometric properties.

Results

Factor analysis produced a unidimensional six-item measure, the Headache Acceptance Questionnaire (HAQ). The HAQ evidenced good internal consistency, convergent validity with headache disability and related psychological constructs, and divergent validity with social desirability. The measure also distinguished between headache diagnostic groups.

Conclusions

Pending further validation in clinical settings, the HAQ may have utility in assessing psychological responses to headache symptoms, identifying targets of treatment for interventions that focus on reducing avoidance, and studying mechanisms of change.

Keywords

Pain acceptance migraine Acceptance and Commitment Therapy disability

Introduction

A substantial portion of disability associated with headache disorders appears attributable to psychological factors, in particular common psychiatric comorbidities (1) and various psychological responses to pain (2,3). Foremost among the psychological factors that have received recent attention are those germane to avoidance and acceptance.

Individuals with headache disorders frequently utilize avoidant coping strategies intended to alter the form or frequency of headache (4). Often this avoidance manifests as attempts to evade or escape from environmental stimuli or private experiences that may trigger or worsen headache attacks, such as suppressing negative thoughts and emotions. These coping strategies are collectively referred to as experiential avoidance (5) and reduce pain-related fear in the short term but over time exacerbate disability and psychiatric symptomatology (4,6). Such avoidance becomes maladaptive when it generalizes to stimuli unrelated to headache, restricts involvement in valued daily activities, or begets social isolation and depression.

A growing body of literature suggests a need to revisit the role of avoidance in headache disorders. Multiple experimental studies show sensitivity to headache triggers decreases as a function of exposure (see Martin and MacLeod (2009) (7) for a review), and behavioral interventions that seek to minimize avoidance hold considerable promise in headache. Specifically, behavioral treatments involving prolonged exposure to triggers have shown preliminary utility in reducing headache symptoms and medication consumption (8), and studies utilizing mindfulness or Acceptance and Commitment Therapy (ACT) have shown favorable impact on headache-related disability and affective symptoms (9,10). Though the existing studies are relatively small, they parallel a larger trend in chronic pain broadly, in which advances in theory and intervention are guided by a focus on functional improvement and acceptance of pain (11).

In contrast to avoidance, acceptance involves a willingness to experience aversive internal events in the service of broader values. Acceptance of pain thus connotes a willingness to experience pain as evidenced by deliberate engagement in valued activities regardless of pain (12). Acceptance of pain is associated with reductions in psychiatric symptomatology and disability, increased daily uptime, and improved work status (11,13). Among headache samples, pain acceptance accounts for substantial unique variance in migraine-related disability (14) and is inversely related to catastrophizing and pain-related interference with daily functioning (15).

Despite its established importance as a psychological process related to headache disorders, measures to assess acceptance of headache are lacking. The 20-item Chronic Pain Acceptance Questionnaire Scale (CPAQ) (13) is the most widely used measure of pain acceptance but was developed and validated among individuals with chronic musculoskeletal pain. As such, it does not capture unique aspects of headache diseases, in particular responses to interictal periods when pain is absent or avoidance of potential attack triggers, and its length precludes routine use in medical settings. Therefore, it is important to develop and validate a brief measure that assesses acceptance of head pain, which was the aim of the present study. We hypothesized that the resulting measure would assess a unidimensional construct, exhibit adequate concurrent and discriminant validity with established measures, and be meaningfully associated with relevant headache variables.

Methods

This study was conducted in three stages during 2017 and 2018: a) Development of candidate items; b) administration to a development sample to determine factor structure and facilitate scale refinement; c) administration to a broader validation sample for confirmatory factor analysis (CFA) and assessment of scale validity. Following the recommendation of Meyers, Gamst, and Guarino (16) and assuming a moderate effect size (f = 0.25), a power level of 0.80, and an alpha level of 0.05, a priori power analysis indicated a total sample size of 500 (across stages 2 and 3) was required for primary factor analytic techniques. This study was approved by the Institutional Review Board at the University of Mississippi, and all participants provided informed consent before participating.

Stage 1: Item development

Scale development began with a large number of items to facilitate data reduction and to ensure a minimum of 10 participants per item for factor analysis (17–18). Items from existing measures of pain acceptance (CPAQ) and experiential avoidance (Acceptance and Action Questionnaire [AAQ-II]) (19) were adapted such that references to pain and negative emotions were replaced with “headache.” Novel items were generated to capture effortful behavioral engagement with precipitating and exacerbating factors of headache and common headache disorder symptoms besides pain. In total, 45 items reflecting willingness to experience headache, remain in the present moment while experiencing headache symptoms, and commitment to valued activities despite headache symptoms were generated. Items also probed individuals’ use of medications to control headache, activity engagement despite environmental sensitivities (e.g. phonophobia, photophobia) or premonitory symptoms (i.e. migraine aura), and responses to headache triggers and gastrointestinal symptoms. All candidate items were termed in the form of declarative statements with a Likert-type response format (1 = never true, 7 = always true).

The 45 candidate items were then distributed to an expert panel of five researchers internationally known in the fields of headache, chronic pain, and/or Acceptance and Commitment Therapy. These individuals were selected because they had worked on the development of existing pain acceptance/avoidance measures or had published psychometric papers relevant to headache acceptance/avoidance. They rated each item on a 0–10 scale in terms of how well it captured the construct of headache acceptance and provided detailed suggestions for improving content and readability of individual items.

Stage 2: Exploratory factor analysis and item refinement

Development sample, measures, and procedure

The development sample consisted of undergraduates at a southern US university who participated in exchange for course credit. As part of a larger battery of measures, they completed a computer-administered version of the Structured Diagnostic Interview for Headache (SDIH-3) (20), which assesses ICHD-3 diagnostic criteria (21) for various headache disorders. The participation announcement invited respondents who experienced any current pattern of headache attacks. Respondents meeting diagnostic criteria for migraine (allowing 2 hours’ minimum duration, as young adults often have otherwise prototypical attacks that do not last 4 hours) (22) or tension-type headache (TTH), those with “probable” migraine or TTH (e.g. meeting all diagnostic criteria except one), and those endorsing symptoms of medication overuse headache, headache subsequent to head injury, or a prior diagnosis of cluster headache were invited to complete the 24 candidate items remaining after stage 1 of development via Qualtrics (Provo, Utah). Five hundred and seventy-seven participants were screened initially, but 43.5% were removed for suspect effort (completing the items in >25% less time than the first author), not meeting inclusion criteria, or having incomplete data, for a final sample of 315 participants.

Statistical analyses

Initial psychometrics involved assessment of inter-item and item-total statistics. Exploratory factor analysis was conducted using principal component analysis (PCA), followed by principal axis factoring.

Stage 3: Scale validation

Validation sample

The validation sample was recruited via Amazon Turk Prime Marketplace, which affords quality data collection among diverse participant samples via crowdsourcing (23). Adults living in the United States who experienced headache were invited to participate in exchange for modest compensation ($0.25). As with the development sample, responses to the SDIH informed diagnostic classification and study inclusion. Three hundred and eighteen individuals completed validity measures described below via Qualtrics, and 33 (10.4%) were excluded for evidencing suspect effort, failing embedded attention checks within the survey, or withdrawing after providing informed consent. Notably, nine participants were excluded from further analyses due to missing data or no evidence of a clear headache diagnosis. The final sample of 276 thus adhered to Nunally’s recommendation (17) regarding the optimal ratio of participants to items.

Measures

In addition to the SDIH and 19 candidate items retained after Stage 2, the validation sample was administered measures to assess concurrent and discriminant validity.

Acceptance and Action Questionnaire-II

The AAQ-II (19) is a seven-item Likert-type measure that reflects respondents’ willingness to remain in contact with private experiences such as emotions, thoughts, and bodily sensations without attempting to alter or avoid them (i.e. experiential avoidance). Lower scores indicate greater acceptance of internal experiences, and high scores reflect experiential avoidance. The AAQ-II is internally consistent and has shown expected relations with measures of avoidant coping and emotional distress (19).

Pain Anxiety Symptoms Scale-Short Form 20

The 20-item Pain Anxiety Symptoms Scale- (PASS-20)²⁴ is a Likert-type measure that assesses fear of pain in relation to persistent pain behaviors. Higher scores reflect greater fear of pain. The PASS-20 has exhibited good convergent and divergent validity (24).

Depression Anxiety Stress Scale-21

The Depression Anxiety Stress Scale (DASS-21) (25) is a 21-item four-point Likert-type measure composed of subscales reflecting the titular negative affective states. The overall score and subscales have demonstrated good reliability and good convergent and divergent validity (26).

Headache Impact Test

The Headache Impact Test (HIT-6) (27) is a six-item self-report measure of headache-related disability as manifested by activity impairment and impact on cognitive and emotional functioning. Total scores range from 36 to 78, with higher scores indicating greater impairment. The HIT-6 has demonstrated high internal consistency, good test-retest reliability, and the ability to differentiate headache impact across diagnostic groups.

Marlowe-Crowne Social Desirability Scale

The Marlowe-Crowne Social Desirability Scale (MCSD) (28) assesses purposeful presentation of socially desirable responses. The MCSD contains 33 true-false items that query involvement in acceptable but improbable behaviors, as well as common improper behaviors. The MCSD was used to assess discriminant validity in the present study.

Statistical analyses

Structural equation modeling (SEM) was used to conduct confirmatory factor analysis based on factor structure derived from the development stage. Concurrent and discriminant validity was assessed via correlations with measures presumed to reflect related (AAQ-II, CPAQ, PASS-20, DASS-21, HIT-6) and unrelated (MCSD) measures, respectively. ANOVA with post-hoc tests was used to assess group differences in headache acceptance as a function of diagnostic status.

Results

Stage 1: Item development

Based on feedback from the expert panel, eliminations and revisions yielded a pool of 24 retained items. Items judged to be poor exemplars of headache acceptance (i.e. four focused on beliefs about headache medication, M = 4.8/10; five conflating fear of pain with acceptance, M = 6.2/10) were deleted. Qualitative feedback informed the removal of nine additional items referencing migraine-specific symptoms (e.g. photo/phonophobia, nausea), because the panel voiced concern that they would not apply to individuals with other headache disorders, and three others for interpretability concerns (e.g. referencing changing or controlling negative thoughts).

A recurrent theme from the panel’s qualitative feedback was to ensure scale items referenced acceptance using behaviorally active terms connoting activity engagement. Retained items without a clear behavioral referent or prone to subjective interpretation were thus revised in accordance with the panel’s suggested language (e.g. “… have a fulfilling life” was changed to “… doing things that I enjoy”). The 24 retained and revised items were administered to the development sample.

Stage 2: Exploratory factor analysis and item refinement

Participants

The 315 participants in the development sample ranged in age from 18 to 36 years (M = 19.2 [SD = 1.3]) and were primarily female (75.6%). Sixty-nine percent of participants self-identified as Caucasian, 21% as Black/African American, 4.1% as Asian, 3% as Hispanic, and 2% as another racial/ethnic background. The most common headache disorder diagnoses were TTH (43% episodic TTH, 2% chronic TTH, 12% probable TTH) and migraine (10% episodic migraine; 2% chronic migraine; 4% probable migraine). Participants reported experiencing headache attacks 8.6 (1.3) days per month with an average intensity of 5.2 (1.6) on a 0–10 scale.

Factor structure

Preliminary analysis

The response distributions of the 24 items met assumptions of normality. Internal consistency was excellent (α = 0.91). The inter-item correlation matrix indicated many items with correlation coefficients greater than r ≥ 0.30, and assessment of item-total statistics (see Table 1) revealed two items with alpha-if-item-deleted values greater than the final alpha (0.91) (18). These items (items 7 and 16) were removed, and the internal consistency of the remaining 22 items was α = 0.91. The Kaiser-Meyer-Olkin (KMO) test of sampling adequacy was significant (KMO = 0.91, p < 0.0001), as was the Bartlett’s Test of Sphericity (χ²(231) = 3086.087, p < 0.0001), both of which supported factorability of the correlation matrix.

Table 1.

Item-total statistics for 24 items.

Item	Corrected item-total correlation	Cronbach's alpha if item deleted
1. My headaches make it impossible to do important things in my life.	0.631	0.901
2. I must limit my activities to avoid anything that might trigger a headache	0.576	0.902
3. I avoid making plans if I think I might get a headache	0.606	0.902
4. Fear of headaches prevents me from living life to the fullest	0.590	0.902
5. I must control my headaches in order to do other things	0.544	0.903
6. I avoid putting myself in situations where I might get a headache.	0.547	0.903
7. I need to get rid of my headaches	0.310	0.908
8. My headaches keep me from trying to be productive	0.571	0.902
9. I would gladly sacrifice important things in my life to better control my headaches	0.566	0.902
10. I cancel my plans if I think I am about to have a headache	0.655	0.901
11. Controlling headache is more important than other goals in my life	0.607	0.902
12. I change my plans to avoid doing anything that might make a headache worse	0.585	0.902
13. I am able to pursue my goals, even though I have headaches	0.468	0.904
14. I can fully participate in my life between headache episodes	0.522	0.903
15. My headaches do not keep me from doing things that I enjoy	0.638	0.901
16. I am now sticking to a certain course in my life whether or not I have headaches	0.164	0.913
17. I try to put forth my best effort even when I have a bad headache	0.511	0.904
18. I can take care of my responsibilities when I have headache	0.551	0.903
19. I try to continue doing things that are important to me even if they might cause a headache	0.614	0.902
20. It’s OK to experience headaches	0.381	0.907
21. I am doing my best to live a normal life with my headaches	0.377	0.906
22. I still do many regular activities when I have a headache	0.646	0.901
23. I try to resume my normal activity as soon as a bad headache is over	0.313	0.907
24. I try to continue doing things that are important to me even if it means my headaches might get worse	0.502	0.904

Note: Values in bold denote items removed from further factor analytic evaluation due to alpha-if-item-deleted values greater than the final alpha value.

Principal component analysis and principal axis factoring

Cattell’s scree test (29) revealed three components above the elbow with eigenvalues greater than 1, which explained 53.5% of the variance. Component 1 yielded an eigenvalue of 8.05 (36.6% of variance), Component 2 of 2.31 (10.5% of variance), and Component 3 of 1.41 (6.4% of variance). Further investigation of these components suggested a two-factor solution, as Direct Oblimin rotation revealed that the first two components were highly correlated (r = 0.46), and the scree plot from parallel analysis also supported a two-component solution. Two items (items 13 and 20) exhibited low communality values (<0.4), indicating inadequate fit with other variables contained in the factor solution, and were thus removed.

An examination of the pattern matrix with the 20 remaining items, however, provided support against a two-component model, as multiple items loaded highly on both factors and both components were highly correlated (r = 0.44). Review of item content revealed a high likelihood that the indicator variability stemmed from method effects, such that all of the items loading on the first component were positively worded, and those comprising the second component were negatively worded. These observations suggested a need to explore a unidimensional factor solution, and principal axis factoring was executed to force a one-component factor solution.

Principal axis factoring

Factor extraction based on PAF evidenced satisfactory matrix factorability, as Kaiser-Meyer-Olkin (KMO) = 0.91, p < 0.001, and Bartlett’s Test of Sphericity was significant at χ² (171) = 2749.94, p < 0.0001. Internal consistency was excellent (0.91), and an additional item (item 23) was identified for exclusion based on an alpha-if-deleted value greater than the overall alpha. With the exception of this excluded item, all factor loadings exceeded the desirable cut-off of 0.32 (30). The retained 19 items were administered to the validation sample for confirmatory factor analysis.

Stage 3: Confirmatory factor analysis

Validation sample

The 276 participants in the validation sample were predominantly female (66%), with age ranging from 18 to 80 years (M = 37.9; SD = 11.4; see Table 2). Eighty percent identified as Caucasian, followed by Black/African American (6%), Asian (5%), Hispanic/Latino (4%), Multiracial (2%), and Other (2%). The overwhelming majority met diagnostic criteria for migraine (50%) or TTH (33%). They reported an average monthly headache attack frequency of 8.2 (6.1) days, moderate-to-severe pain intensity (M = 6.6 [1.9]), and substantial headache-related disability (HIT-6 M = 58.6 [7.8]).

Table 2.

Validation sample demographics.

	Migraine n = 142 (49.9%)				Tension-type headache n = 94 (33%)			Other n = 40 (14.1%)
	CM 13 (4.6%)	EM w/o A 35 (12.3%)	EM w/A 16 (5.6%)	PM 78 (27.4%)	CTTH 3 (1.1%)	ETTH 47 (16.5%)	PETTH 44 (15.4%)	CH 20 (7.0%)	PTHA 17 (6.0%)	MOH 3 (1.1%)
Age	35.69 (10.64)	38.94 (8.61)	39.19 (14.50)	38.10 (12.38)	41.00 (6.25)	40.62 (11.39)	31.61 (8.51)	44.25 (7.5)	39.24 (13.50)	41.33 (18.15)
AAQ-II	24.46 (11.27)	21.37 (10.68)	22.94 (12.77)	21.41 (10.62)	16.67 (2.08)	19.60 (9.32)	21.16 (10.07)	22.55 (9.77)	24.47 (8.16)	26.00 (4.00)
PASS-20	71.31 (17.55)	61.34 (17.65)	74.19 (17.18)	66.32 (19.87)	50.00 (3.61)	53.40 (14.43)	52.43 (14.70)	71.95 (20.97)	68.71 (18.08)	73.00 (20.07)
DASS-21 Depression	12.69 (6.45)	12.26 (5.46)	14.40 (6.85)	12.31 (5.95)	13.33 (8.39)	11.49 (5.44)	11.77 (5.19)	12.80 (5.39)	14.18 (4.88)	15.00 (7.00)
DASS-21 Anxiety	13.31 (5.20)	10.31 (4.06)	13.00 (6.31)	11.79 (5.20)	8.00 (1.73)	9.96 (3.85)	10.68 (3.42)	11.85 (4.53)	13.94 (4.97)	16.00 (6.56)
DASS-21 Stress	14.77 (6.00)	13.06 (4.92)	13.80 (5.31)	13.33 (5.26)	9.00 (1.00)	12.36 (4.06)	12.61 (4.34)	14.20 (4.62)	16.00 (5.15)	15.67 (6.35)
MCSD-13	53.46 (2.37)	53.34 (1.68)	52.75 (1.29)	52.70 (2.11)	51.67 (1.15)	53.19 (1.68)	53.59 (2.34)	53.15 (1.98)	52.18 (1.81)	52.33 (2.52)
Headache days/month	16.54 (1.94)	5.43 (2.80)	7.25 (5.77)	8.90 (6.21)	21.67 (5.77)	5.06 (3.05)	6.16 (4.41)	12.55 (6.03)	12.06 (8.40)	18.67 (2.31)
Hit-6	63.69 (5.7)	60.14 (5.65)	62.80 (5.89)	59.88 (6.51)	60.00 (8.19)	53.57 (6.80)	53.39 (7.86)	65.10 (7.69)	62.06 (8.39)	64.33 (4.62)

Note: Values are means (SD) or frequency counts (%).

CM: chronic migraine; EM w/o A: episodic migraine without aura; EM w/A: episodic migraine with aura; PM: probable migraine; CTTH: chronic tension-type headache; ETTH: episodic tension-type headache; PETTH: probable episodic tension-type headache; CH: cluster headache; PTHA: posttraumatic headache; MOH: medication overuse headache.

Model specification

Confirmatory factor analysis (CFA) via STATA SEM was utilized to distinguish the best fitting model for a unidimensional factor solution. Recommended conservative guidelines (31) were adopted for determining model fit: Non-significant model chi-square (χ²), Root Mean Square Error of Approximation (RMSEA) <0.05, Comparative Fit Index (CFI) >0.95, and Standardized Root Mean Square Residual (SRMR) <0.08. The initial model with 19 items exhibited poor model fit: χ²(170) =1051.00, p < 0.0001; RMSEA = 0.135; CFI = 0.681; SRMR = 0.116. As these indices suggested some degree of content saturation/redundancy, an iterative process of removing items with normalized covariance residuals ≥2 was undertaken (32). All indices reflected satisfactory fit after six additional iterations that removed 13 items (see Table 3): χ²(9) = 12.020, p > 0.05; RMSEA = 0.034; CFI = 0.995; and SRMR =0.02 (see Figure 1). The final six-item model accounted for a significant proportion of the variance explained (R²= 82%, p < 0.001) and evidenced good internal consistency (α = 0.82). The mean score was 28.37 (6.95) across all participants.

Table 3.

Summary of confirmatory factor analysis results.

Model iteration	Items removed based on normalized covariance residuals	χ ²	p	RMSEA	CFI	SRMR
1		170	<0.001	0.135	0.681	0.116
2	4, 5, 9, 12, 16	77	<0.001	0.120	0.836	0.116
3	11, 19	54	<0.001	0.107	0.893	0.051
4	8, 17	35	<0.001	0.116	0.900	0.049
5	1, 15	20	<0.001	0.080	0.958	0.037
6	20	14	<0.001	0.056	0.984	0.032
7	6	9	>0.05	0.034	0.995	0.022

Validity

As predicted, total scores on the measure exhibited a strong negative correlation with measures of headache-related disability (r = −0.63, p < 0.01) and fear of pain (r = −0.59, p < 0.01), such that individuals endorsing greater acceptance of headache reported lower disability and fear of pain. Additionally, the scale exhibited smaller but significant correlations with experiential avoidance (AAQ-II: r = −0.21, p < 0.01) and negative emotions (DASS-21: r = −0.29, p < 0.01). Discriminant validity was evident via a nonsignificant relationship with social desirability (MCSD: r = 0.08, p = 0.118).

To assess differences in headache acceptance as a function of diagnostic status, a one-way between-groups ANOVA was conducted across SDIH headache groups, revealing a significant omnibus difference, F (10, 268) =9.73, p < 0.001, η²= 0.27. Post-hoc tests indicated that participants with TTH variants evidenced significantly greater headache acceptance than other diagnostic groups (see Table 4).

Table 4.

Mean HAQ total scores by diagnostic group (SD).

	CM	EM w/o Aura	EM w/Aura	PM	CTTH	ETTH	PETTH	CH	PTHA	MOH
HAQ Total Score	24.23 (4.97)	27.89 (6.79)	26.44 (5.43)	26.96 (6.43)	36.67* (3.21)	33.06* (5.08)	32.30* (5.42)	22.20 (8.01)	23.53 (5.67)	23.33 (7.64)

Individuals with variants of tension-type headache (TTH) reported significantly higher (p < 0.05) headache acceptance than individuals with migraine (CM, EM with and without aura, PM) and those reporting symptoms of cluster headache (CH), post-traumatic headache (PTHA), or medication overuse headache (MOH).

Discussion

The present study endeavored to develop and provide initial validation data for the Headache Acceptance Questionnaire (HAQ; see Table 5). Items were generated then revised based on feedback from an expert review panel, and candidate items were then administered to two distinct samples for further reduction, factor analysis, and initial validation. This iterative effort culminated in a six-item measure with strong internal consistency and initial evidence of validity.

Table 5.

Headache Acceptance Questionnaire (HAQ).

Instructions: Below you will find a list of statements. Please rate how true each statement is for you using the scale below.
1	2	3	4	5	6	7
Never true	Very rarely true	Rarely true	Sometimes true	Frequently true	Almost always true	Always true
1. I must limit my activities to avoid anything that might trigger a headache. 2. I avoid making plans if I think I might get a headache. 3. I avoid putting myself in situations where I might get a headache. 4. My headaches keep me from trying to be productive. 5. I would gladly sacrifice important things in my life to better control my headaches. 6. I am doing my best to live a normal life with my headaches.

Scoring: Items 1–5 are reverse scored in order to obtain a total headache acceptance score. Total scores range from 6 to 42, with higher scores indicating greater acceptance of headache.

As hypothesized, final HAQ items reflect a single factor structure that endeavors to assess acceptance of headache as operationalized by willingness to experience pain via engagement in valued activities despite headache symptoms. Item content reflects behavioral manifestations of activity engagement both in anticipation of headache (three items) and when headache is present (three items). Attention to activity engagement in anticipation of pain is particularly germane to headache disorders, insofar as virtually all the major headache disorders manifest via episodic attacks and can be triggered by internal or external stimuli. The shared loading of these different items underscores the importance of trigger avoidance within (low) headache acceptance and provides further validation for the trigger-avoidance model of headache (7).

Evidence for initial validity of the HAQ was promising. Among a broad sample of adults with a variety of headache presentations, headache acceptance scores had strong inverse associations with headache-related disability and fear of pain, replicating and extending prior studies on general pain acceptance among individuals with migraine (14–15). Individuals with TTH reported greater headache acceptance than those with more severe and disabling forms of headache (e.g. migraine, cluster, post-traumatic), further supporting the construct validity of the scale. The relatively weaker relationship between headache acceptance and AAQ-II scores supports the notion that the HAQ assesses a construct related to, but somewhat distinct from, general experiential avoidance and psychological distress. The lack of a relationship between HAQ scores and social desirability not only provides evidence for discriminant validity but suggests that this new measure may not be particularly susceptible to attempts to present oneself in an overly positive light.

Future studies continuing validation efforts using treatment-seeking samples from headache clinics and further exploring differential utility of the HAQ versus existing measures of headache-related disability are needed. Although the correlation between HAQ and HIT-6 scores was moderate (r = −0.63), it was not at a level suggestive of multicollinearity. Pain acceptance is a psychological mechanism through which disability is affected, as shown in a recent mediation analysis of disability and psychiatric symptomatology (33). Nevertheless, additional studies comparing predictive utility of headache acceptance and disability would be valuable, given that both constructs involve activity engagement. Another promising area for future research is to explore relations between this measure and the Headache Triggers Sensitivity and Avoidance Questionnaire (HTSAQ) (34), which assesses potency, sensitivity, and avoidance of 24 common headache triggers. The observation of a significant inverse relationship between these measures would provide further evidence of convergent validity and validity of the underlying theoretical construct.

To the extent that the HAQ measures a change process fundamental to contemporary behavioral headache interventions, this new measure may prove valuable for various clinical purposes. Coupled with attention to unique aspects of headache, the brevity and readability (Flesch-Kincaid Grade Level (35) = 7.5) of the scale lend it well to administration within routine practice settings. Pending further investigation, treatment outcome research may reveal that HAQ scores may be utilized to identify individuals most likely to respond positively to ACT or exposure-based interventions and as a means for assessing both treatment outcome and underlying mechanisms of change.

Strengths of this study include utilization of three distinct phases for scale development and validation, assessment of headache profiles using a structured diagnostic interview, and inclusion of two different headache samples. Despite these strengths, limitations exist. The current study utilized convenience sampling of undergraduate students and an online sample obtained through Amazon Turk. Though not representative of treatment-seeking individuals, participants’ monthly headache frequency and disability reflected considerable impact of headache. Young adults were chosen because of their high prevalence of headache disorders and presumed low rates of confounding variables such as many years of medication overuse. Further, multiple studies have established the utility and reliability of samples gathered via MTurk (23). However, both samples were disproportionately female, Caucasian, and well-educated, and future work should attempt to validate the HAQ and its factor structure using more diverse treatment-seeking samples. A second limitation is that cluster and posttraumatic headache were each assessed via a single screener question (i.e. “Have you ever been diagnosed with cluster headaches” and “Did this headache develop shortly after a head injury or head trauma?”), which likely contributed to higher-than-expected rates of these conditions. Caution is thus warranted when applying these findings to these less common primary and secondary headache disorders. Pending further research, however, the present effort provides a novel instrument to assess acceptance of headache that may prove promising in both clinical and research settings, advancing both the science of headache medicine and the utility of behavioral interventions.

Article highlights

Headache acceptance is a distinct psychological construct that is meaningfully associated with variables of prognostic importance.

The six-item Headache Acceptance Questionnaire (HAQ) assesses acceptance of headache and displayed promising initial psychometric properties, including convergent validity with measures of disability and related constructs, divergent validity with a measure of social desirability, and the ability to distinguish between diagnostic groups.

Acceptance of headache warrants attention in clinical contexts as it may be a target for reducing headache disease-related disability and studying treatment outcomes and mechanisms of change.

Footnotes

Author note

This research was the doctoral dissertation of the first author.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: TAS has served as consultant for Alder. The other authors report no conflicts of interest.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Todd A Smitherman

References

Saunders

Merikangas

Low

, et al. Impact of comorbidity on headache-related disability. Neurology 2008; 70: 538–547.

Black

Fulwiler

Smitherman

. The role of fear of pain in headache. Headache 2015; 55: 669–679.

Seng

Buse

Klepper

, et al. Psychological factors associated with chronic migraine and severe migraine-related disability: An observational study in a tertiary headache center. Headache 2017; 57: 593–604.

Wieser

Walliser

Womastek

, et al. Dysfunctional coping in headache: Avoidance and endurance is not associated with chronic forms of headache. Eur J Pain 2011; 16: 268–277.

Hayes

Wilson

Gifford

, et al. Experiential avoidance and behavioral disorders: A functional dimensional approach to diagnosis and treatment. J Consult Clin Psychol 1996; 64: 1152–1168.

Vlaeyen

JWS

Linton

. Fear-avoidance and its consequences in chronic musculoskeletal pain: A state of the art. Pain 2000; 85: 317–332.

Martin

MacLeod

. Behavioral management of headache triggers: Avoidance of triggers is an inadequate strategy. Clin Psychol Rev 2009; 29: 483–495.

Martin

Reece

MacLeod

, et al. Behavioral management of the triggers of recurrent headache: A randomized controlled trial. Behav Res Ther 2014; 61: 1–11.

Andrasik

Grazzi

D’Amico

, et al. Mindfulness and headache: A “new” old treatment, with new findings. Cephalalgia 2016; 36: 1192–1205.

10.

Wells

Beuthin

Granetzke

. Complementary and integrative medicine for episodic migraine: An update of evidence from the last 3 years. Curr Pain Headache Rep 2019; 23: 10–10.

11.

McCracken

Vowles

. Acceptance and commitment therapy and mindfulness for chronic pain: Model, process, and progress. Am Psychol 2014; 69: 178–187.

12.

Vowles

Snowden

Ashworth

. A comprehensive examination of the model underlying acceptance and commitment therapy for chronic pain. Behav Ther 2014; 45: 390–401.

13.

McCracken

Vowles

Eccleston

. Acceptance of chronic pain: Component analysis and a revised assessment method. Pain 2004; 107: 159–166.

14.

Foote

Hamer

Roland

, et al. Psychological flexibility in migraine: A study of pain acceptance and values-based action. Cephalalgia 2016; 36: 317–324.

15.

Chiros

O’Brien

. Acceptance, appraisals, and coping in relation to migraine headache: An evaluation of interrelationships using daily diary methods. J Behav Med 2011; 43: 307–320.

16.

Meyers L, Gamst G and Guarino A. Principal components and factor analysis. In: Applied multivariate research: Design and interpretation. Thousand Oaks, CA: Sage Publications, 2006, pp. 465–514.

17.

Nunally

Bernstein

. Psychometric theory, 3rd edn. New York, NY: McGraw-Hill, 1994, pp. 209–337, 445–594.

18.

Tabachnick

Fidell

. Using multivariate statistics, 6th edn. Boston, MA: Allyn and Bacon, 2013, pp. 612–785.

19.

Bond

Hayes

Baer

, et al. Preliminary psychometric properties of the acceptance and action questionnaire–II: A revised measure of psychological inflexibility and experiential avoidance. Behav Ther 2011; 42: 676–688.

20.

Smitherman

Penzien

Rains

, et al. Headache, Boston, MA: Hogrefe Publishing, 2014, pp. 100–108.

21.

Headache Classification Committee of the International Headache Society. The International Classification of Headache Disorders, 3rd edition. Cephalalgia 2018; 38: 1–211.

22.

Rains

Penzien

Lipchik

, et al. Diagnosis of migraine: Empirical analysis of a large clinical sample of atypical migraine (IHS 1.7) patients and proposed revision of the IHS criteria. Cephalalgia 2001; 21: 584–595.

23.

Litman

Robinson

Abberbock

. TurkPrime.com: A versatile crowdsourcing data acquisition platform for the behavioral sciences. Behav Res 2017; 49: 433–442.

24.

McCracken

Dhingra

. A short version of the pain anxiety symptoms scale (PASS-20): Preliminary development and validity. Pain Res Manag 2002; 7: 45–50.

25.

Lovibond

. The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behav Res Ther 1995; 33: 335–343.

26.

Osman

Wong

Freedenthal

, et al. The Depression Anxiety Stress Scales—21 (DASS-21): Further examination of dimensions, scale reliability, and correlates. J Clin Psychol 2012; 68: 1322–1338.

27.

Kosinski

Baylis

Bjourner

, et al. A six-item short-form survey for measuring headache impact: The HIT-6. Qual Life Res 2003; 12: 963–974.

28.

Crowne

Marlowe

. A new scale of social desirability independent of psychopathology. J Consult Psychol 1960; 24: 349–354.

29.

Cattell

. The scree test for the number of factors. Multivariate Behav Res 1996; 1: 245–276.

30.

Yong

Pearce

. A beginner’s guide to factor analysis: Focusing on exploratory factory analysis. Tutor Quant Methods Psychol 2013; 9: 79–94.

31.

Kline

. Principles and practice of structural equation modeling, 4th edn. New York, NY: The Guilford Press, 2016.

32.

Bagozzi

. On the evaluation of structural equation models. JAMS 1988; 16: 74–79.

33.

Seng

Kuka

Mayson

, et al. Acceptance, psychiatric symptoms and migraine disability: An observational study in a headache center. Headache 2018; 58: 859–872.

34.

Kubik

Martin

. The Headache Triggers Sensitivity and Avoidance Questionnaire: Establishing the psychometric properties of the questionnaire. Headache 2017; 57: 236–254.

35.

Kincaid JP, Fishburne RP, Rogers RL, et al.; Institute for Simulation and Training. Derivation of new readability formulas (Automated Readability Index, Fog Count and Flesch Reading Ease Formula) for Navy enlisted personnel, https://stars.library.ucf.edu/istlibrary/56 (1975, accessed 27 January 2020).