Message Framing and the Willingness to Pursue Behavioral Therapy: A Study of People With Migraine
Publication: The Journal of Neuropsychiatry and Clinical Neurosciences
Abstract
Objective:
Behavioral treatments for migraine prevention are safe and effective but underutilized in migraine management. Health message framing may be helpful in guiding patients with treatment decision making. The authors assessed associations between message framing and the willingness to seek migraine behavioral treatment among persons with a diagnosis of migraine headache.
Methods:
A total of 401 individuals (median age=34 years [interquartile range, 12 years]) who screened positive for migraine, as determined by the American Migraine Prevalence and Prevention questionnaire, were assessed. Participants were randomly assigned to receive one of four message frames using TurkPrime: specific loss framing (N=101), specific gain framing (N=98), nonspecific loss framing (N=102), and nonspecific gain framing (N=100). The message frames were initially piloted for 56 participants and then revised by a headache specialist, with input from a communications specialist, and randomly distributed to the larger sample.
Results:
More than two-thirds of participants (70.3%) were women. The median number of headache days per month was 5 (interquartile range, 5.3). Some of the participants (12.5%) had previously used evidence-based behavioral therapy for migraine. No significant differences in the willingness to pursue behavioral treatment for migraine between the four message framing groups were found. The median for all four types of message frames was 4 (interquartile range, 1; Kruskal-Wallis H, p=0.41).
Conclusions:
Findings revealed that message framing was not associated with willingness to seek behavioral therapy for migraine.
The Centers for Disease Control and Prevention and the National Institutes of Health describe health communication as “the study and use of communication strategies to inform and influence individual decisions that enhance health” (1). One main purpose of health communication is to disseminate health information to influence personal health choices and to improve literacy. In the field of neurology and psychiatry, health communication is essential (e.g., much of a neurologist’s time is spent taking a detailed history and counseling patients about diagnoses and treatment decision making). There has been some research on health communication in neurology. When eliciting a patient’s history, neurologists tend to use closed-ended questions, not the ask-tell-ask approach (2).Although recent research has begun to investigate various methods for delivering information and for decision making, to date, such research has examined recommending, versus providing, options (3, 4), and no data support one method over the other. This is especially important in the fields of neurology and psychiatry, in which adherence to treatment has been proven to be a challenge in a wide variety of neurologic and psychiatric conditions (5–7). However, little research has been conducted on how to frame neurologists’ and psychiatrists’ recommendations to improve adherence to therapies.
The term “health message frames” refers to how health messages can be written or framed to highlight the benefits of engaging in a behavior (a gain frame) or the costs of not engaging in a behavior (a loss frame) (8, 9). Although findings are mixed based on the behavioral goal, in general, gain-framed messages have been shown to be more effective than loss-framed messages in promoting disease prevention and encouraging health-promoting behaviors (8, 9). Yet, to our knowledge, this method of health communication has not been examined in the field of neurology.
In the present study, we sought to examine how message framing might be useful to patients in adhering to recommendations to initiate behavioral therapy. Behavioral therapy has many applications within the field of neurology. Unfortunately, despite the level-A evidence supporting several treatments for neurologic conditions, patients often do not adhere to recommendations to pursue these treatments. This was observed in a recent study that found that only about one-half of patients with migraine initiated scheduling for a behavioral therapy appointment for migraine prevention as recommended by a health care provider (10). In another study of psychogenic nonepileptic seizures, 80% of the study subjects attended their first psychiatric appointment, but only 42% attended the second, and 24% attended the third, with only 14% remaining adherent at the fourth visit (11). Thus, using migraine, the second most disabling condition worldwide in terms of disability-adjusted years lost (12), we sought to examine whether positive or negative gain frames, as well as whether specific frames (more detailed description), were associated with patients’ likelihood to pursue level-A evidence-based therapies for migraine prevention. In addition, we used TurkPrime, a platform for participant recruitment in the behavioral sciences (13, 14), to explore the impact of message framing on the willingness of patients with migraine to participate in behavioral therapy, as well as to examine whether this accounted for patient demographic variables, such as age, gender, ethnicity, education level, and income (13, 14).
We hypothesized that the gain-framed messages would be more effective than the loss-framed messages and that the more specific messages would be more effective than the less specific messages.
Methods
Study Design and Participants
This was a cross-sectional study assessing the intention of patients with migraine to pursue behavioral treatment on the basis of the way in which behavioral therapy for migraine was framed to them in an online survey. Participants were recruited online via TurkPrime, and TurkPrime administered the survey to eligible individuals. Inclusion criteria were age 18–89 years, proficiency in English, and migraine as a diagnostic condition. These inclusion criteria were established with the prequalifier panel, which was set up by TurkPrime. Individuals with fewer than four headache days per month were excluded. The exclusion criterion enabled us to examine only those individuals who had migraine and would likely qualify for migraine preventive treatment. TurkPrime contacted the participant pool with information about the study (study description, monetary compensation, and time required to complete the study). Individuals who were interested in participating were asked to complete a migraine screening questionnaire to ensure that they met criteria for a migraine diagnosis. Predictor variables were demographic characteristics, headache characteristics, and types of message frames. The primary outcome was likelihood to pursue behavioral therapy for migraine prevention. The objective of this study was to assess whether different message frames would affect the likelihood to pursue behavioral therapy for migraine prevention.
Study Questionnaire
The initial screening questionnaire to identify participants with migraine used the American Migraine Prevalence and Prevention (15) questions for migraine diagnosis. By using branching logic, participants who screened positive for migraine were randomly assigned by TurkPrime to receive one of four hypothetical messages (below) to assess the impact of health message framing on their willingness to engage in behavioral therapy for migraine. The messages were gain-framed or loss-framed and either specific or nonspecific. The survey and the initial messages were developed by a headache specialist and a health psychologist. Eight message frames about behavioral treatment for migraine were initially developed (six specific messages and two nonspecific messages), with one-half being gain-framed and the other half being loss-framed. These initial message frames were randomly distributed to 56 participants who met the study criteria (seven participants per frame). On the basis of the initial feedback from participants and feedback from two health communications experts (AF and AL), four final message frames were used in the full study, with the goal being a sample size of around 400 participants (100 participants per frame). The message provided to the participants was as follows:
Behavioral therapy for migraine helps people with migraine change their lifestyle and day-to-day approach to migraine attacks to better manage migraine. Examples of behavioral therapy include cognitive behavioral therapy, biofeedback and/or relaxation therapy. Cognitive behavioral therapy helps people with migraine change how they think about and respond to migraine attacks. It also teaches them lifestyle changes to help reduce migraine frequency. Relaxation training teaches people with migraine how to control their bodies’ reaction to stress, which can reduce migraine frequency. Biofeedback gives people with migraine insight into how their bodies react to stress, and how they can use relaxation training, by giving them literal feedback (e.g., computer printout, audible tone, etc.) into their bodies’ stress responses.
We then provided the participants with one of the four message frames listed below.
1. Nonspecific loss frame: “Not participating in behavioral therapy for migraine is a lost opportunity to improve both 1) the number of headache days per month and 2) the severity of your pain. This could lead to more days with headache in the long-term.”2. Nonspecific gain frame: “Participating in behavioral therapy for migraine may reduce both 1) the number of headache days per month and 2) the severity of your pain. This could lead to fewer days with headache in the long-term.”3. Specific loss frame: “Not participating in behavioral therapy for migraine is a lost opportunity to improve both 1) the number of headache days per month by at least half and 2) the severity of your pain. This could lead to more days with headache in the long-term.”
Study participants were then asked, “Given what you know about behavioral treatment for migraine, how likely are you to do behavioral treatment?” Responses were captured by using a 5-point Likert scale (1=not at all likely to 5=strongly likely).
Statistical Analyses
The distribution of the data was assessed for normality with the Kolmogorov-Smirnov test. Nonparametric data are reported as medians and interquartile ranges and were analyzed with the Kruskal-Wallis H test. Chi-square and linear regressions were also performed. The statistical analyses were performed with SPSS Statistics, version 25.0 (IBM, Armonk, N.Y.).
The study was approved by the institutional review board at New York University Langone Health. Participants were paid $0.75 for each completed survey.
Results
Demographic Variables
A total of 401 individuals participated in the study. The median age was 34 years (interquartile range, 12). The majority of participants were women (N=282 [70.3%]) and Caucasian (N=302 [75.3%]). Most were employed full-time (N=252 [62.8%]). A complete summary of the participants’ demographic characteristics is presented in Table 1.
| Characteristic | Total sample (N=401) | Specific gain framing (N=98) | Nonspecific gain framing (N=100) | Specific loss framing (N=101) | Nonspecific loss framing (N=102) | p | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Median | Interquartile range | Median | Interquartile range | Median | Interquartile range | Median | Interquartile range | Median | Interquartile range | ||
| Age (years) | 34 | 12 | 34 | 12.8 | 32 | 10 | 32 | 13 | 34 | 13 | 0.094a |
| N | % | N | % | N | % | N | % | N | % | ||
| Female | 282 | 70.3 | 68 | 69.4 | 67 | 68.4 | 72 | 73.5 | 75 | 76.5 | 0.267b |
| Race/ethnicity | 0.573b | ||||||||||
| Caucasian | 302 | 75.3 | 70 | 71.4 | 73 | 73 | 80 | 79.2 | 79 | 77.5 | |
| African American | 50 | 12.5 | 13 | 13.3 | 13 | 13 | 11 | 10.9 | 13 | 12.7 | |
| Hispanic | 26 | 6.5 | 5 | 5.1 | 9 | 9 | 7 | 6.9 | 5 | 4.9 | |
| Asian, Native American, or other | 23 | 5.7 | 10 | 10.2 | 5 | 5 | 3 | 3.0 | 5 | 4.9 | |
| Education (years) | 0.33b | ||||||||||
| High school | 55 | 13.7 | 10 | 10.2 | 15 | 15 | 11 | 10.9 | 19 | 18.6 | |
| Some college and vocational training | 112 | 27.9 | 26 | 26.5 | 31 | 31 | 30 | 29.7 | 25 | 24.5 | |
| Bachelor’s or associate degree | 185 | 46.1 | 47 | 48.0 | 42 | 42 | 50 | 49.5 | 46 | 45.1 | |
| Master’s or doctorate degree | 49 | 12.2 | 15 | 15.3 | 12 | 12 | 10 | 9.9 | 12 | 11.8 | |
| Employment | 0.48b | ||||||||||
| Full-time | 252 | 62.8 | 69 | 70.4 | 57 | 57 | 56 | 55.4 | 70 | 68.6 | |
| Other | 149 | 37.2 | 29 | 29.6 | 43 | 43 | 45 | 44.6 | 32 | 31.4 | |
a
Data were determined with Kruskal-Wallis H test.
b
Data were determined with chi-square test.
Headache characteristics.
The median number of headache days per month was 5 (interquartile range, 5.3). Nearly one-half of participants (N=181 [45.1%]) reported at least two headaches per week that were of moderate to severe intensity. Approximately 11% (N=45) of participants reported having more headache days than days without a headache. The median headache intensity rating was 7 out of 10 (interquartile range, 2). The median score on the Migraine Disability Assessment test was 28 (interquartile range, 29). Nearly one-third of participants (N=117 [29.2%]) reported taking at least one migraine preventive medication. One in eight individuals (N=50 [12.5%]) had previously tried at least one evidence-based cognitive-behavioral treatment for migraine prevention.
There was no difference in age, gender, ethnicity, education, or employment status by message frame (Table 2). In addition, there was no difference in the message frame-perceived clarity between the four study groups (nonspecific loss, nonspecific gain, specific loss, and specific gain).
| Variable | Total sample (N=401) | Specific gain framing (N=98) | Nonspecific gain framing (N=100) | Specific loss framing (N=101) | Nonspecific loss framing (N=102) | p | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Median | Interquartile range | Median | Interquartile range | Median | Interquartile range | Median | Interquartile range | Median | Interquartile range | ||
| Number of headache days per month | 5 | 5.3 | 4.2 | 6.25 | 4.8 | 7 | 5 | 4 | 5 | 4.7 | 0.56a |
| Headache intensityb | 7 | 2 | 7 | 2.8 | 7 | 2 | 7 | 2 | 8 | 3 | 0.46a |
| Migraine Disability Assessment score | 28 | 29 | 26.5 | 32 | 29 | 29.8 | 28 | 24 | 31.5 | 28.7 | 0.94a |
| N | % | N | % | N | % | N | % | N | % | ||
| Prior preventive medications | 117 | 29.2 | 30 | 30.6 | 24 | 24 | 30 | 29.7 | 33 | 2.4 | 0.58a |
| Prior evidence-based behavioral treatment | 50 | 12.5 | 11 | 11.2 | 15 | 15 | 11 | 10.9 | 13 | 12.7 | 0.98c |
| ≥2 Moderate to severe headaches per week | 181 | 45.1 | 38 | 38.8 | 40 | 40 | 53 | 52.5 | 50 | 49.0 | 0.053c |
| ≥15 Headache days per month | 45 | 11.2 | 7 | 7.14 | 17 | 17 | 14 | 13.9 | 7 | 6.9 | 0.74c |
a
Data were determined with Kruskal-Wallis H test.
b
The headache intensity rating was measured on a scale of 1–10.
c
Data were determined with chi-square test.
There was no difference between the four groups with regard to headache frequency, headache intensity, score on the Migraine Disability Assessment, the number of participants taking preventive medications, or previous use of evidence-based cognitive-behavioral treatments for migraine prevention (Table 2).
Message frame type and likelihood to pursue behavioral treatment.
There was no association between the type of message frame and likelihood to seek in-person behavioral treatment, with a median of 4 (interquartile range, 1) for all four types of message frames (Kruskal-Wallis H, p=0.41, η2H of 0.00030). The effect size was not significant.
Discussion
We did not find any significant change in the use of positive versus negative message framing or specific versus nonspecific message framing among participants’ willingness to initiate behavioral therapy for migraine. Evidence-based behavioral treatments for migraine are underutilized despite the fact that they are effective in migraine prevention and are fairly safe and free of many of the adverse side effects that may occur with migraine preventive medications (10, 16). Previous research has found that a lack of belief in efficacy or in the importance of behavioral treatment can be a barrier for patients pursuing behavioral treatment for migraine prevention. However, our results do not support that message frames with dissemination of knowledge about behavioral treatment efficacy is superior to message frames on behavioral treatment without efficacy. A better understanding of what facilitates the willingness of patients with migraine to participate in behavioral therapy is needed.
Research has shown that the impact of a health message also depends on the patient’s involvement (17), the efficacy of the proposed intervention, the ease of uptake of the intervention, and other individual differences (e.g., risk perception and tendency toward behavioral activation) (9). For example, experiences with prior interventions are particularly salient for message framing, because previous research suggests that patients who have had negative experiences with prior interventions are more likely to respond to loss-frame messages (17). In addition, there is a large disconnect between people’s intentions to alter their behavior and the actual likelihood that they will do it (9). Our results suggest that message framing is not adequate, at least not the examples we analyzed, in increasing people’s intention of the likelihood to pursue behavioral therapy. Perhaps a more tailored health message frame approach, such as SEABIT [Self-Administered Behavioral Intervention using Tailored messages] for migraine, rather than generic one-size-fits-all messages, would be more useful (18).
There are several limitations of this study. First, the quantitative design may have limited our findings. Future work might include qualitative research in order to understand what additional information people may want before deciding whether to pursue behavioral therapy. Second, there may have been selection bias, because the participants were paid volunteers. Third, the study did not account for any influence a health care provider may have had in encouraging a patient to pursue behavioral therapy. Fourth, we did not examine whether neural mechanisms involved in decision making might affect a patient’s decision to seek behavioral therapy or whether migraine as a chronic pain disorder may bias brain-based decision making. Lastly, we did not assess whether the use of migraine preventive medications (and which class of medications) might affect neural mechanisms and thus may be associated with the decision making involved in pursuing behavioral therapy.
Conclusions
To our knowledge, no research to date has evaluated message framing to promote the uptake of behavioral treatment. Our study did not reveal an association between health message framing and willingness to initiate behavioral therapy for migraine prevention. More research on the most impactful ways to convey information to patients in neurology clinics is needed.
References
1.
Gateway to Health Communication and Social Marketing Practice: https://www.cdc.gov/healthcommunication/healthbasics/WhatIsHC.html
2.
Hahn SR, Lipton RB, Sheftell FD, et al: Healthcare provider-patient communication and migraine assessment: results of the American Migraine Communication Study, phase II. Curr Med Res Opin 2008; 24:1711–1718
3.
Toerien M, Shaw R, Reuber M, et al: Initiating decision-making in neurology consultations: recommending versus option-listing and the implications for medical authority. Sociol Health Illn 2013; 35:873–890
4.
Chappella P, Toeriena M, Jacksona C, et al: Following the patient’s orders? recommending vs offering choice in neurology outpatient consultations. Soc Sci Med 2018; 205:8–16
5.
Gewirtz A, Minen M: Adherence to behavioral therapy for migraine: knowledge to date, mechanisms for assessing adherence, and methods for improving adherence. Curr Pain Headache Rep 2019; 23:3
6.
Matsuzawa Y, Lee YSC, Fraser F, et al: Barriers to behavioral treatment adherence for headache: an examination of attitudes, beliefs, and psychiatric factors. Headache 2019; 59:19–31
7.
Farooq S, Naeem F: Tackling nonadherence in psychiatric disorders: current opinion. Neuropsychiatr Dis Treat 2014; 10:1069–1077
8.
Rothman AJ, Bartels RD, Wlaschin J, et al: The strategic use of gain- and loss-framed messages to promote healthy behavior: how theory can inform practice. J Commun 2006; 56:S202–S220
9.
Gallagher KM, Updegraff JA: Health message framing effects on attitudes, intentions, and behavior: a meta-analytic review. Ann Behav Med 2012; 43:101–116
10.
Minen MT, Azarchi S, Sobolev R, et al: Factors related to migraine patients’ decisions to initiate behavioral migraine treatment following a headache specialist’s recommendation: a prospective observational study. Pain Med 2018; 19:2274–2282
11.
Tolchin B, Dworetzky BA, Baslet G: Long-term adherence with psychiatric treatment among patients with psychogenic nonepileptic seizures. Epilepsia 2018; 59:e18–e22
12.
GBD 2015 Neurological Disorders Collaborator Group: Global, regional, and national burden of neurological disorders during 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Neurol 2017; 16:877–897
13.
Litman L, Robinson J, Abberbock T: TurkPrime.com: a versatile crowdsourcing data acquisition platform for the behavioral sciences. Behav Res Methods 2017; 49:433–442
14.
Buhrmester M, Kwang T, Gosling SD: Amazon’s Mechanical Turk: a new source of inexpensive, yet high-quality, data? Perspect Psychol Sci 2011; 6:3–5
15.
Stewart WF, Lipton RB, Celentano DD, et al: Prevalence of migraine headache in the United States: relation to age, income, race, and other sociodemographic factors. JAMA 1992; 267:64–69
16.
Ernst MM, O’Brien HL, Powers SW: Cognitive-behavioral therapy: how medical providers can increase patient and family openness and access to evidence-based multimodal therapy for pediatric migraine. Headache 2015; 55:1382–1396
17.
Rothman AJ, Updegraff JA: Specifying when and how gain- and loss- framed messages motivate healthy behavior: an integrated approach, in Perspectives on Framing, 1st ed. Edited by Keren G. New York, Psychology Press, 2010, pp 257–277
18.
Nicholson R, Nash J, Andrasik F: A self-administered behavioral intervention using tailored messages for migraine. Headache 2005; 45:1124–1139
Information & Authors
Information
Published In
History
Received: 12 March 2019
Revision received: 29 May 2019
Accepted: 7 June 2019
Published online: 9 August 2019
Published in print: Spring 2020
Keywords
Authors
Competing Interests
The authors report no financial relationships with commercial interests.
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).