Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
ISSN 0017-8748 doi: 10.1111/j.1526-4610.2011.01866.x Published by Wiley Periodicals, Inc. Headache © 2011 American Headache Society Research Submission Preventive Medication Adherence in African American and Caucasian Headache Patients head_1866 520..532 Bernadette D. Heckman, PhD; Gary Ellis, MS Study Objectives.—To examine race-related differences in adherence to preventive medication agents in headache patients and identify factors predictive of medication adherence in Caucasian and African American headache patients. Methods.—Using a longitudinal naturalistic study design, data from 77 Caucasian and 32 African American headache patients were collected through (1) 30-day daily diaries that assessed medication adherence, headache frequency, and headache episode severity; (2) self-administered surveys that assessed headache management self-efficacy; and (3) telephoneadministered psychiatric interviews that yielded psychiatric diagnoses. Using daily diary adherence data, patients’ adherence to preventive agents was dichotomized as “Inconsistent” (ie, adhered fewer than 80% of days) or “Consistent” (ie, adhered ⱖ80% of days during the past month). Results.—The proportion of adherent African American patients (69%) did not differ significantly from the proportion of adherent Caucasian patients (82%). Exploratory univariate logistic regression analyses found that preventive medication adherence levels of 80% or less were associated with being diagnosed with major depressive disorder and lower levels of headache management self-efficacy. Conclusions.—Future research should test if interventions that reduce depressive symptoms and increase patients’ levels of headache management self-efficacy can produce concomitant increases in adherence to preventive headache agents. Key words: headache, medication adherence, race-related health disparities Abbreviations: AIDS acquired immune deficiency syndrome, ANOVA analysis of variance, HIV human immunodeficiency virus, HMSE headache management self-efficacy, ICHD-II International Classification of Headache Disorders, 2nd Edition, MEMS medication event monitoring system, PRIME-MD Primary Care Evaluation of Mental Disorders (Headache 2011;51:520-532) schedules is associated with greater mortality, more rapid disease progression, more frequent hospitalizations, and poorer clinical health outcomes in patients with a variety of chronic health conditions.4-7 Moreover, in the USA alone, the economic impact of appointment and medication nonadherence exceeds $100 billion dollars annually.8 While medication adherence has been studied extensively across a variety of health conditions, such as human immunodeficiency virus/acquired immune deficiency Approximately 30% to 60% of patients living with chronic health conditions report inconsistent adherence to medication regimens (ie, typically ⱕ80%).1-3 Inconsistent adherence to medication From the Department of Psychology, Ohio University, Athens, OH, USA. Address all correspondence to B.D. Heckman, Department of Psychology, Ohio University, Athens, OH 45701, USA. To download a podcast featuring further discussion of this article, please visit www.headachejournal.org Conflict of Interest: There are no conflicts of interests associated with this study. Accepted for publication January 23, 2011. 520 Headache syndrome (HIV/AIDS),9,10 cardiovascular disease,11 and asthma,12 far less research has examined medication adherence in patients with headache disorders. Pharmacological treatments for headache episodes include abortive agents, such as triptans, oral nonsteroidal anti-inflammatory drugs, or combination analgesics taken during the course of the headache attack to stop or limit headache symptoms.13 However, the overutilization of abortive headache medications can lead to medication overuse headaches, also known as “rebound headaches.”13 For individuals who experience frequent headaches, preventive (ie, prophylactic) pharmacotherapies, such as b-blockers, anticonvulsants, and tricyclic antidepressants/selective serotonin reuptake inhibitors (SSRIs), are indicated treatments and significantly reduce the number of headache attack days, improve daily physical functioning, and enhance headache patients’ overall quality of life.14-18 The benefits of preventive headache agents, however, are largely dependent upon consistent adherence (although minimum levels of adherence that must be achieved to obtain clinical benefits in this population are unknown). Packard et al19 found a low level of adherence (52%) to acute and preventive headache agents in patients with headache disorders, suggesting that patients found it difficult to adhere consistently to headache medications or that adherence-related benefits did not outweigh adherence-related challenges in many patients (eg, economic costs, negative side effects, and perceived inefficacy of treatments). An additional explanation for low rates of medication adherence in headache patients is that, compared with many other chronic illnesses, such as diabetes, osteoarthritis, and HIV/ AIDS, headaches can be relatively intermittent in nature (eg, some patients experience only a few headaches per month), which may predispose some headache patients to poorer adherence. Only 2 studies have specifically examined adherence to preventive agents in headache patients. Steiner et al20 investigated adherence to pizotifen in migraine patients in a headache specialty clinic. Patients took the correct number of pills on 58% of days over an 8-week period; however, only 32% of doses were taken on schedule during this period. 521 Mullerners et al21 examined adherence to preventive medications in migraine patients in a London headache specialty clinic. When assessed via self-report pill-count, the average 2-month adherence rate was 91%. However, during the same time period, medication event monitoring systems (MEMS) recorded a mean adherence rate of only 66% (suggesting that self-report may inflate adherence rates in headache patients). When regimen complexity was considered, patients taking only 1 medication dosage per day selfreported an adherence rate of 79%, patients prescribed twice-daily preventive medications reported a 60% adherence rate, and patients prescribed a 3-time daily preventive medication reported an adherence rate of 54%. It is important to note, however, that these studies were based on small samples (eg, n = 4 and n = 29, respectively). The low rates of adherence reported in these studies underscore the need to better understand, and potentially intervene upon, medication adherence in headache patients. No research has examined if race-related differences in medication adherence observed in persons with other chronic health conditions (eg, AIDS, diabetes)22-25 generalize to individuals with headache disorders or if predictors of medication adherence differ by race. Racial differences in adherence to headache medications have not been studied even though: (1) 3.3% of African Americans in the USA are diagnosed with a headache disorder;26 (2) African Americans living with a variety of chronic health conditions (eg, HIV/AIDS, hypertension, diabetes) often report poorer adherence to medication regimens than Caucasians;27-30 and (3) African American and Caucasian headache patients respond equally well to contemporary preventive pharmacotherapies, highlighting the importance of maximizing medication adherence in both racial groups.31 This exploratory study examined rates and predictors of adherence to preventive medication agents in patients in headache subspecialty treatment clinics and if predictors of adherence differed by race. The study tested the following 5 hypotheses (the rationale for each is described in detail below): (1) African Americans would report lower rates of medication adherence than Caucasians; (2) patients diagnosed with major depressive disorder (MDD) would report 522 poorer medication adherence than patients with no diagnosis of MDD; (3) patients with greater headache management self-efficacy (HMSE) would report more consistent medication adherence; (4) headache frequency would predict medication adherence more strongly in African Americans than in Caucasians; and (5) headache episode severity would predict medication adherence more strongly in African Americans than Caucasians. African Americans report poorer adherence to medication regimens than Caucasians across a number of chronic health conditions, including HIV/ AIDS, hypertension, and diabetes.25-28 Poorer rates of medication adherence in African Americans compared with Caucasians have been observed across different adherence assessment methodologies, including self-report,32 pharmacy refills,33 and MEMS.34 Reasons for poorer medication adherence in African Americans in the USA include cost concerns,35 beliefs that medications are symbols of mental illness,36 fear of becoming addicted to medications,36 and conspiracy beliefs regarding the healthcare system and the use of pharmacotherapies with African Americans in the USA.37 The hypothesis that being diagnosed with MDD would lead to poorer medication adherence in headache patients is based on several studies linking depression with poor adherence in other illnesses.22-24 Patients with MDD may experience cognitive symptoms of depression that result in one forgetting to take medications and/or confusion about one’s medication regimen. Moreover, depressed headache patients may experience increased apathy, low selfesteem, and overly pessimistic and irrational thoughts about medication efficacy, resulting in poorer medication adherence. The hypothesis that greater HMSE would promote greater medication adherence is based on numerous studies demonstrating a positive relationship between self-efficacy and medication adherence.38-46 Self-efficacy is a person’s beliefs about his/her abilities to influence events and behaviors that affect one’s life.47 Headache patients who are confident that they can effectively prevent and manage their headaches are likely to adhere more consistently to headache medication regimens. Conversely, April 2011 headache patients who lack a sense of mastery or who have difficulty managing their headaches (based either on current headache activity or past headache management failures) may question their headache management skills and experience greater difficulty adhering to medication regimens. Finally, this study tested the hypothesis that patients who reported more severe and frequent headaches at treatment initiation would report greater headache medication adherence at 6-month follow-up but that these relationships would be stronger in African Americans than in Caucasians. These predictions are based on clinical and experimental research showing that African Americans report less pain tolerance and greater pain unpleasantness than their Caucasian counterparts.48,49 Because of their heightened sensitivity to pain,African Americans with more severe and frequent headaches at treatment initiation are expected to perceive a greater need to adhere more consistently to preventive headache agents to better manage their headache disorders. Findings from the current study can identify individuals at elevated risk for poor adherence to preventive headache agents and inform the development of culturally contextualized interventions to increase preventive medication adherence in headache patients. METHODS Participants and Inclusion Criteria.—This study was a secondary analysis of a larger dataset of 311 patients in a longitudinal study that examined racerelated differences in headache treatment outcomes. Of the 311 patients enrolled in the original study, 109 (77 Caucasian and 32 African American headache patients) provided adherence data through 30-day daily diaries. Between July 2004 and June 2008, all patients were recruited into the study through 4 outpatient headache specialty treatment clinics located in Cincinnati, Cleveland, Columbus, and Toledo, Ohio. Study inclusion criteria were: (1) 18 years of age or older; (2) self-identifying as African American or Caucasian; (3) meeting International Classification of Headache Disorders, 2nd Edition (ICHD-II) criteria50 for episodic migraine, chronic migraine, episodic tension-type headache, or chronic tension-type headache; (4) the physician believed that the patient Headache 523 Pre treatment Visit Patient enters treatment 0 1-Mo FU Visit Patient begins new preventive therapy 1 2-Mo FU Visit 2 Patients complete the first 30-day daily diary 6-Mo FU Visit 3 Months 4 5 6 Patients complete the second 30-day daily diary Figure.—Study design. FU = follow-up; Mo = month. would benefit from a new preventive headache agent; and (5) the patient was willing to delay the initiation of new preventive treatment for 1 month (so that pretreatment headache activity could be recorded). The project’s protocol was approved by the university’s institutional review board and no adverse events were reported during the study. Previous publications from this dataset have examined how treatment appointment attendance51 and treatment outcomes differ by patient race,31 how treatment outcomes relate to the presence of a psychiatric condition,52 and the extent to which commonly used psychosocial headache measures are appropriate for use with Caucasian and African American headache patients.53 Procedure.—Recruitment brochures that described the study were distributed to potential patients in reception areas and waiting rooms of participating clinics as they awaited treatment. Because patients of color were relatively underrepresented in all 4 treatment centers, all racial minority patients who presented for treatment were approached for study enrollment. The larger pool of Caucasian patients enabled the project to use a recruitment strategy in which only every fifth Caucasian patient was approached for study enrollment. Eligible patients who decided to enroll into the study provided written informed consent in the treatment clinic. As shown in the Figure, the study used a prospective, longitudinal design and assessed patients at pre- treatment and 1-, 2-, and 6-month follow-up. These follow-up time periods were used because they represented patients’ typically scheduled visits to participating clinics. Acute therapies were either initiated or adjusted by study physicians at the patients’ pretreatment visit. Patients were reevaluated during the second visit that occurred 1 month after the pretreatment visit. During this visit, study physicians assessed the need for preventive medication(s) using the 30-day headache diary completed by the patient. Patients determined to be in need of preventive a gents were prescribed medications as clinically indicated. Patients were then scheduled to participate in 2- and 6-month follow-up visits. The study’s experimental design and procedures are described in greater detail elsewhere.31 Assessment Methodologies.—Headache Diagnosis.—During the initial patient–physician interaction, the physician diagnosed the patient’s current headache disorder(s) using ICHD-II.50 The physician also documented the patient’s past and current pharmacological headache treatments. Psychiatric Diagnosis.—Psychiatric disorders were diagnosed using the Primary Care Evaluation of Mental Disorders (PRIME-MD).54 The PRIME-MD was administered by trained research staff during a telephone interview conducted within 2 days after the pretreatment visit. The PRIME-MD yields a subset of diagnoses included in the Diagnostic and Statistical 524 Manual of Mental Disorders, including mood and anxiety disorders, alcohol and substance abuse/ dependence disorders, eating disorders, and somatoform disorders. In this study, the PRIME-MD was used to identify patients with MDD (0 = no, 1 = yes). 30-Day Daily Headache Diary.—At the conclusion of the pretreatment visit, each patient received a self-administered, paper-and-pencil daily diary to record the frequency and severity of headaches experienced during the assigned 30-day period. The first 30-day period was the interval between the patient’s initial assessment and his or her second visit (when new preventive therapy was prescribed). This headache diary was also completed during the 30 days that followed the patient’s 6-month follow-up visit. Headache frequency was operationally defined as the number of days over the 30-day period during which patients experienced a “mild” “moderate,” or “severe” headache.” Headache episode severity was assessed for each headache using a 4-point scale (1 = “No pain,” 2 = “Mild,” 3 = “Moderate,” or 4 = “Severe”). On days that patients did not experience a headache, a value of “1” was recorded. On days that patients did experience a headache, the severity of the headache episode was rated using response options “2” through “4.” Mean headache episode severity was calculated by summing headache severity ratings and dividing this sum by the number of headaches the patient experienced over the 30-day reporting period. Preventive Medication Adherence.—Adherence to preventive medication was assessed only in the 30-day headache diary that followed the 6-month follow-up visit. Based on patients’ daily diary responses, adherence to preventive medication agents was dichotomized as 0 = “Inconsistent” (ie, adhered to medication regimens on fewer than 80% of days) or 1 = “Consistent” (ie, adhered to medication regimens on 80% of more days during the past month). Eighty percent adherence to medication regimens is a common operational definition of adequate medication adherence across a number of health conditions.1-3 Headache Management Self-Efficacy Scale.—The 25-item HMSE scale55 assessed patients’ confidence April 2011 that they could prevent and manage headaches. Respondents rated the extent to which they agreed with each item using a 7-point scale (1 = “Strongly disagree” to 7 = “Strongly agree”). The HMSE evidenced good internal consistency in the current study (a = 0.90). Demographic Characteristics.—Patients provided data on their age, gender, socioeconomic status, health insurance status (health maintenance organization, preferred provider organization, private, social security disability insurance), and selfselected the racial category with which they most identified. Data Analytic Procedures.—Data screening analyses evaluated the distributions of study variables (eg, skew, kurtosis) and identified potential univariate and bivariate outliers. Differences between Caucasians and African Americans at treatment initiation were examined using analysis of variance (ANOVA) and chi-square. Chi-square analyses tested if race was associated with type of acute or preventive medication agents prescribed to patients. Several exploratory univariate logistic regression analyses sought to identify individual predictors of the dichotomized medication adherence measure (0 = “Inconsistent,” 1 = “Consistent”). Specifically, separate regression analyses examined if medication adherence was related to patients’ race, age, education, gender, being diagnosed with MDD (yes/no), headache frequency at pretreatment, headache episode severity at pretreatment, and HMSE at pretreatment. Three exploratory interaction terms (ie, “Race ¥ MDD Status,” “Race ¥ Headache Frequency,” and “Race ¥ Headache Severity”) were also examined in separate univariate logistic regression analyses to determine if these interactions predicted medication adherence. The regression analyses were exploratory in that they did not test a formal theoretical model that might explain preventive medication adherence in headache patients. Instead, these regression analyses tested if individual variables (or 2-way interactions involving race) predicted adherence. All data analyses were conducted using spss Version 18 (PASW Statistics). All inferential statistical analyses that tested a priori study hypotheses used a ⱕ 0.05, 2-tailed. Headache 525 RESULTS Demographics and Headache Diagnoses.—As shown in Table 1, the average patient was female (89%), 36.6 years of age (SD = 10.4, Min = 18, Max = 66), had completed 13.9 years of education, and reported an annual income of less than $60,000 (87%). Ninety-nine percent of patients had some form of health insurance. Seventy percent of patients received a headache diagnosis of migraine and 54% received a diagnosis of tension-type headache at their initial visit. Patients experienced 1 or more headaches on 17.2 days during the past month (Min = 1, Max = 30, SD = 7.7). Thirty-one percent of all patients were diagnosed with MDD. Table 1 shows the demographic characteristics of the 173 Caucasian and 111 African American patients who completed pretreatment assessments and the 79 Caucasian and 32 African American patients who completed 6-month follow-up assessments (although 2 of the 79 Caucasians who completed 6-month follow-up assessments did not provide adherence data). Chi-square tests of association and 1-way ANOVA characterized associations among race, demographic variables, headache diagnoses, and headache characteristics at pretreatment and 6-month follow-up. Table 1 shows that, at pretreatment, African American patients were older, slightly more likely to be female, reported fewer years of education, were more likely to be diagnosed with MDD, reported more headache days per month, reported more severe headache episodes, and were less likely to be diagnosed with migraine as their primary headache disorder compared with Caucasians (all Ps < .05). Table 1 also shows that among the 79 Caucasians and 32 African Americans who completed 6-month follow-up assessment instruments, African Americans were more likely to be female and report annual incomes below $60,000 (P < .05). Table 2 shows that, at the end of the 6-month follow-up period, preventive agents prescribed frequently to Caucasian patients included anticonvulsants (33.1%), antidepressants (31%), and “other” preventive therapies (eg, low-dose aspirin, nonsteroidal anti-inflammatories; 33.1%). Triptans or Table 1.—Sociodemographic Characteristics and Headache Diagnoses of Caucasian and African American Patients Characteristic Age (years) Years of education Income below $60,000 Being female ⱖ80% medication adherence Diagnosed with major depressive disorder Headache days/month Headache episode severity Diagnosed with migraine (episodic or chronic, with or without aura) Diagnosed with TTH (episodic or chronic) 6-month Pretreatment Overall follow-up Pretreatment sample pretreatment sample sample African c2 or t sample Caucasians Americans pretreatment Caucasians sample† (n = 284) (n = 79) (n = 173) (n = 112) 6-month follow-up sample African Americans (n = 32) c2 or t 6-month follow-up sample‡ 36.6 ⫾ 10.4 13.9 ⫾ 2.2 88.8% 87.8% n/a 30.5% 35.6 ⫾ 10.2 14.3 ⫾ 2.1 88.7% 85.2% n/a 24.1% 38.4 ⫾ 10.5 13.5 ⫾ 2.3 88.8% 92.1% n/a 40.4% 2.2** 2.7*** 0.1 2.8* n/a 8.3*** 36.5 ⫾ 10.5 40.8 ⫾ 11.2 14.4 ⫾ 2.1 13.7 ⫾ 2.1 81.3% 100% 84.4% 100% 82% 69% 24.7% 34.4% 1.7 2.8 6.7** 5.4** 1.6 1.1 17.2 ⫾ 7.7 2.64 ⫾ 0.3 69.6% 16.6 ⫾ 7.9 2.61 ⫾ 0.3 76.0% 18.2 ⫾ 7.3 2.70 ⫾ 0.4 59.8% 2.1** 2.2** 8.4*** 11.9 ⫾ 7.6 2.54 ⫾ 0.3 – 14.3 ⫾ 7.7 2.51 ⫾ 0.3 – 1.5 0.4 – 53.7% 57.3% 48.2% – – *P < .10; **P < .05; ***P < .01. †Caucasians and African Americans compared at baseline. ‡Caucasians and African Americans compared at 6-month follow-up. TTH = tension-type headache. 2.3 – 526 April 2011 Table 2.—Preventive and Acute Medications Prescribed to Patients by Race Caucasians (%) Treatment Preventive medication Antidepressant Anticonvulsant b-blocker Calcium blocker Other (NSAS, low-dose aspirin, Botox) Acute medication Simple and compound analgesic Compound analgesics with codeine/barbiturate Triptans or ergotamine Adjunctive treatments (antiemetics, etc) African Americans (%) 1-month follow-up 6-month follow-up 1-month follow-up 6-month follow-up 19.9 27.5 9.4 2.3 12.9 31.0 33.1 11.3 2.8 33.1 27.3 24.5 10.9 4.5 6.4 42.1 30.7 13.6 5.7 26.1 25.1 2.3 41.5 25.7 28.9 2.8 42.3 26.1 23.6 2.7 32.7 25.5 25.0 1.1 35.2 25.0 No between-group or within-group pairwise comparisons were significant at P < .05. ergotamines were the most widely prescribed acute medications for Caucasians (42.3%). Among African American patients, the most widely prescribed preventive therapies were antidepressants (42.1%), anticonvulsants (30.7%), and “other” preventive therapies (26.1%). The most common acute medications prescribed to African Americans were triptans or ergotamines (35.2%). The large proportion of African American patients prescribed antidepressants is noteworthy given that African Americans (40.4%) were significantly more likely than Caucasians (24.1%) to be diagnosed with MDD. No significant race differences were found in preventive or acute medication prescription practices (all Ps > .10). Adherence Data Collection by Race.— Significantly fewer African Americans (28.1%; 32/114) provided adherence data compared with Caucasians (44.5%; 77/173) via 30-day daily diaries completed after the 6-month follow-up, c2(1) = 7.9, P < .005. To examine whether patients who provided adherence data differed from patients who did not provide adherence data at 6-month follow-up, a series of within-race analyses were conducted. As shown in Table 3, within-race analyses showed that African Americans who provided adherence data at 6-month follow-up were comparable with African Americans who did not provide adherence data in age, educa- tion, annual income, being diagnosed with MDD, headache frequency, headache severity, headache disability, headache-specific quality of life, or HMSE assessed at pretreatment assessed at 6-month follow-up at pretreatment (all Ps > .10). Table 3 also shows that Caucasians who provided adherence data at 6-month follow-up were comparable with Caucasians who did not provide adherence data at 6-month follow-up (all Ps > .05). Predictors of Medication Adherence.—A series of exploratory univariate logistic regression analyses were conducted that identified predictors of medication adherence group: 0 = “Inconsistent” (<80%), 1 = “Consistent” (ⱖ80%). Each analysis included 109 African American and Caucasian patients who provided adherence data at 6-month follow-up. Because no demographic variable predicted adherence group, none was employed as a covariate in the univariate regression analyses. For regression analyses that tested race-related interactions (ie, “Race ¥ MDD Status,” “Race ¥ Headache Frequency,” and “Race ¥ Headache Severity”), the 2 main effects were entered first followed by the interaction term. Rates of adherence did not vary by the 4 study sites (P > .60). Table 4 displays the odds ratios, 95% confidence intervals, and significance levels for each individual predictor of adherence. A logistic regression found Headache 527 Table 3.—Demographic and Behavioral Differences Between Patients Who Did and Did Not Provide Adherence Data at 6-Month Follow-Up Characteristic Age (years) Years of education SES factor score Diagnosed with MDD Headache days/month Headache episode severity Headache disability inventory Headache specific quality of life Headache management self-efficacy Caucasians who provided adherence data (n = 77) Caucasians who did not provide adherence data (n = 96) 37.23 ⫾ 10.5 14.4 ⫾ 2.11 .24 ⫾ .92 27.5% 16.5 ⫾ 7.6 2.57 ⫾ 0.31 1.8 ⫾ 0.52 44.8 ⫾ 13.9 93.8 ⫾ 23.9 34.1 ⫾ 9.8 14.2 ⫾ 2.1 0.1 ⫾ 0.96 32.4% 15.9 ⫾ 8.2 2.66 ⫾ 0.23 1.9 ⫾ 0.52 44.5 ⫾ 15.6 97.1 ⫾ 21.1 c2 or t African Americans who provided adherence data (n = 32) African Americans who did not provide adherence data (n = 79) c2 or t -1.9 -0.72 -1.6 0.7 -0.40 1.5 0.83 -0.09 -0.89 40.73 ⫾ 11.4 13.7 ⫾ 2.1 -0.10 ⫾ 0.9 42.9% 18.9 ⫾ 7.3 2.63 ⫾ 0.38 2.1 ⫾ 0.5 47.9 ⫾ 16.1 93.9 ⫾ 21.1 37.3 ⫾ 9.9 13.4 ⫾ 2.4 -0.21 ⫾ 1.1 34.4% 18.9 ⫾ 7.7 2.67 ⫾ 0.26 2.0 ⫾ 0.7 48.3 ⫾ 17.7 90.5 ⫾ 23.7 -1.5 -0.49 -0.51 0.7 -0.04 0.06 -0.41 0.09 -0.65 MDD = major depressive disorder. that the proportions of African American (69%) and Caucasian (82%) patients who adhered consistently to preventive medication agents during the past month did not differ significantly, OR = 0.49, 95% CI = 0.19, 1.26, P = .14.* Table 4 also shows that patients who reported greater HMSE were also significantly more likely to adhere consistently to preventive headache agents, OR = 1.02, 95% CI = 1.011.05, P = .03. Finally, patients with MDD were also less likely to adhere consistently to preventive headache agents, OR = 0.34, 95% CI = 0.13-0.87, P = .03. None of the 2-way interactions involving Race (ie, Race ¥ MDD Status, Race ¥ Headache Frequency, or Race ¥ Headache Severity) predicted adherence. A post-hoc point-biserial correlational analysis found that HMSE and being diagnosed with MDD (the 2 strongest predictors of adherence) were significantly and negatively correlated, r(109) = -0.29, P < .002. *The logistic regression analysis was reconducted controlling for patients’ age, years of education completed, gender, and being diagnosed with MDD at pretreatment (yes/no). Similar to the logistic regression analysis conducted without covariates, this logistic regression analysis found that the proportion of African American and Caucasian patients who adhered consistently to preventive medication agents during the past month did not differ significantly, ORAdj = 0.47, 95% CI = 0.16, 1.33, P = .15. Specifically, headache patients diagnosed with MDD also reported lower levels of HMSE. DISCUSSION This is the first exploratory study to compare rates of adherence and identify predictors of adherence to preventive headache agents in Caucasian and African American headache patients. Comparable Table 4.—Predictors of Medication Adherence Group in African American and Caucasian Headache Patients Predictor variable OR 95% CI for OR P Age (years) Years of education Being female Being African American Headache days per month Headache episode severity Headache management self-efficacy Being diagnosed with MDD Race ¥ headache episode severity Race ¥ headache days per month Race ¥ major depression diagnosis 1.00 0.99 0.29 0.49 1.02 2.02 1.02 0.34 9.87 1.09 0.35 0.96-1.05 0.80-1.24 0.04-2.39 0.19-1.26 0.96-1.08 0.56-7.28 1.01-1.05 0.13-0.87 0.60-161.54 0.95-1.25 0.05-2.91 .89 .98 .25 .14 .63 .28 .03 .03 .11 .25 .36 MDD = major depressive disorder. 528 proportions of African American and Caucasian headache patients reported consistent adherence (ⱖ 80%) to preventive headache medications in the past month. This finding was unexpected given that African Americans with chronic health conditions (eg, HIV/AIDS, diabetes) often report poorer medication adherence compared with Caucasians.26,28 While both groups reported reasonably good adherence to preventive medication regimens, it is unclear whether the ⱖ80% adherence rate recorded by both groups (ie, 69% for African Americans and 82% for Caucasians) are sufficient to manage patients’ headaches effectively. The lack of race-related differences in medication adherence in this study may be explained, at least in part, by the differentially higher rate of treatment dropout in African Americans compared with Caucasians. Medication adherence was assessed only at 6-month follow-up. Less than 30% of African Americans provided adherence data in the current study. It is possible that only the most adherent African American and Caucasian patients remained in treatment and provided adherence data at 6-month follow-up, masking any potential race-related differences in medication adherence. Had medication adherence been assessed earlier in the study, racerelated differences in medication adherence may have emerged. Headache patients diagnosed with MDD at pretreatment were less likely to adhere consistently to headache medication regimens 6 months later. Adherence-improvement interventions that focus on depressed patients and that intervene simultaneously on the patient’s adherence difficulties and depression appear warranted. Currently, the use of antidepressant pharmacotherapies (eg, tricyclics, SSRIs) is indicated for patients who present to headache clinics with chronic headache disorders and comorbid depression. Cognitive-behavioral interventions56 that eliminate maladaptive and self-defeating thoughts and behaviors that prevent consistent adherence to preventive headache medications may also be helpful for depressed headache patients. In fact, cognitivebehavioral interventions have already been shown to simultaneously reduce depressive symptoms and improve adherence to antiretroviral therapy in April 2011 persons living with HIV/AIDS.57 Interpersonal psychotherapy,58 which links depression to challenging life events (eg, job loss, a bereavement, or an interpersonal dispute), should also be considered a potential psychotherapeutic approach to alleviate headache activity and depressive symptoms in depressed headache patients. At a minimum, all practitioners should monitor closely rates of preventive medication adherence in their headache patients with comorbid depressive disorders. Headache patients who reported less HMSE at pretreatment were also less likely to adhere consistently to preventive headache agents 6 months later. This finding was expected given the strong relationship between self-efficacy and adherence observed across other health disorders.38-46 This finding suggests that headache practitioners should proactively monitor patients’ adherence behaviors if they appear to lack the self-confidence or skills required to adhere consistently to preventive headache agents. Given that headache patients who were depressed also tended to report lower levels of HMSE, interventions that target depressed headache patients are likely to capture headache patients who also have low selfefficacy to adhere consistently to preventive headache agents. Three study hypotheses were not supported in the current study. As noted above, adherence was unrelated to patient race. However, it is worth noting that a greater proportion of African American patients (40%) were diagnosed with MDD compared with Caucasians (24%). Because depression appears to be associated with nonadherence to preventive headache agents, a disproportionately large number of African American headache patients may be at risk for nonadherence because of their MDD diagnoses. Additional research is needed to examine associations among race, MDD status, and adherence. Finally, adherence was unrelated to the “Race by Headache Severity” and “Race by Headache Frequency” interactions. While African Americans, in general, may have greater pain sensitivity and lower pain thresholds,48,49 these tendencies do not appear to result in greater rates of adherence in African American headache patients compared with Caucasians. Headache While this study identified predictors of medication adherence in headache patients, and a previous report using this dataset identified predictors of treatment appointment nonattendance in this sample,31 predictors of the 2 health-related behaviors are likely to differ. For example, medication adherence is likely to be related to factors such as negative medication side effects, medication costs, and beliefs that medications are inefficacious. Conversely, treatment appointment nonattendance may be related to factors such as long waiting times during previous treatment appointments, poor patient–physician relationships, lack of transportation or child care, and severe headache activity at the time of the appointment. Future research should continue to identify reasons for poor medication adherence and inconsistent treatment appointment attendance in headache patients. The present study had several limitations. All participating headache treatment facilities were located in large urban areas of Ohio. The extent to which study findings generalize to other geographic regions is unclear. The current study utilized a selfrecorded daily diary measure of adherence; no objective measures of adherence were used (eg, MEMS or prescription refills). While daily self-recorded measures can alleviate problems related to retrospective recall, they are still prone to problems related to demand characteristics and social desirability and may overestimate rates of medication adherence.45 This study used a relatively small nonprobability sample and African American patients were recruited into the study more aggressively than were Caucasians (to ensure adequate representation of African American patients). Because no data were collected that characterized the demographic compositions of the 4 participating headache subspecialty clinics, it was not possible to determine if patients enrolled in the study were representative of the clinics from which they were recruited. Finally, while patients who provided adherence data did not differ from those who did not provide adherence data, study findings are likely to generalize slightly more so to Caucasian patients who are female, older, less educated, more likely to be depressed, and have headaches that are more frequent, severe, and disabling and African American patients who are female and 529 poorer (given patterns of attrition from baseline to 6-month follow-up). An important limitation that warrants reiteration is that adherence was assessed only once and near the completion of the study’s final follow-up period. Patients who dropped out of the study may have evidenced different (and potentially lower) rates of adherence than patients who remained in the study, further limiting the generalizability of study findings. As such, it is likely that this study characterized and modeled medication adherence in relatively adherent headache patients, limiting the external validity of study findings. Additionally, some patients may have been managing their headaches during the study through relaxation, meditation, biofeedback, and herbal therapies. While practices such as these may have altered patients’ headache activity and adherence to pharmacotherapies, they were not assessed in the study. Measures used in the current study were not specific to medication adherence behaviors and beliefs. For example, the self-efficacy measure assessed patients’ abilities to manage their headaches and was not specific to medication adherence selfefficacy. The relationship between self-efficacy and preventive medication adherence might have differed if a medication-specific measure of self-efficacy was used. The relatively small number of patients enrolled in the study did not permit us to examine if race differences existed in adherence to different classes of preventive medications, such as antidepressants, anticonvulsants, or b-blockers. Finally, study findings generalize only to headache patients being treated in headache subspecialty clinics (not patients with headaches in primary care settings, etc). CONCLUSIONS In spite of these limitations, this study is perhaps the first to investigate race-related differences in rates and predictors of adherence to preventive pharmacotherapies. This research found that while African American and Caucasian patients did not differ significantly in adherence to preventive headache agents, almost one-third of African Americans failed to achieve the 80% adherence rate recommended for persons taking medications for chronic health conditions. Research on rates and predictors of medication 530 adherence in chronic health conditions such as HIV/ AIDS, diabetes, and hypertension is extensive and rather sophisticated (eg, the use of MEMS caps and pharmacy refills to assess adherence and studies correlating different types of adherence assessment methodologies to clinical health outcomes). Far less research has examined medication adherence and how medication adherence relates to treatment outcomes in persons with headache disorders. While the limitations associated with this exploratory study prevent us from making any definitive conclusions regarding how medication adherence differs by race in headache patients, perhaps this study can serve as a starting point from which additional and more rigorous scientific studies of medication adherence in headache patients can be initiated. Medication adherence research in the area of headache appears to be lagging in amount, sophistication, and theoretical rigor relative to this type of research in other chronic health conditions. It is time to narrow the research gap on this important topic. Acknowledgments: This research was supported by Grant K01 NS046582 from the National Institute of Neurological Disorders and Stroke (B. Heckman, P.I.). We would like to acknowledge our collaborators Dr. Ken Holroyd and the Project INSIGHT Research Team for their helpful comments on the manuscript and for their important contributions to the conduct of the study. REFERENCES 1. DiMatteo MR. Variations in patients’ adherence to medication recommendations: A quantitative review of 50 years of research. Med Care. 2004;42:200-209. 2. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353:487-497. 3. Donnan PT, MacDonald TM, Morris AP. The DARTS/MEMO collaboration: Adherence to prescribed oral hypoglycemic medication in a population of patients with type two diabetes: A retrospective study. Diabetes Med. 2002;19:279284. 4. Saran R, Bragg-Gresham JL, Rayner HC, et al. Nonadherence in hemodialysis: Association with mortality, hospitalization, and practice patterns in the DOPPS. Kidney Int. 2003;64:254-262. April 2011 5. Peterson LA, Wright SM, Person ED, et al. Impact of race on cardiac care and outcomes in veterans with acute myocardial infarction. Med Care. 2002;40:186-196. 6. Weiden PJ, Kozma C, Grogg A, Locklear J. Partial compliance and risk of rehospitalization among California Medicaid patients with schizophrenia. Psychiat Serv. 2004;55:886-891. 7. Ghali JK, Kadakia S, Cooper R, Ferlinez J. Precipitating factors leading to decompensated of heart failure: Traits among urban blacks. Arch Intern Med. 1988;148:2013-2016. 8. Dunbar-Jacob J, Mortimer-Stephens MK. Treatment adherence in chronic disease. J Clin Epidemiol. 2001;54:S57-S60. 9. Levine AJ, Hinkin CH, Castellon SA, et al. Variation in patterns of highly active antiretroviral therapy (HAART) adherence. AIDS Behav. 2005;3:355-362. 10. Schneider J, Kaplan SH, Greenfield S, Li W, Wilson IB. Better physician-patient relationships are associated with higher reported adherence to antiretroviral therapy in patients with HIV infection. J Gen Intern Med. 2004;19:1096-1103. 11. Dunbar-Jacob J, Bohachick P, Mortimer MK, Sereika SM. Medication adherence in person with cardiovascular disease. J Cardiovasc Nurs. 2003;3:209-218. 12. Apter AJ, Boston RC, George M, et al. Modifiable barriers to adherence to inhaled steroids among adults with asthma: It’s not just black and white. J Allergy Clin Immunol. 2003;111:1219-1226. 13. Silberstein SD, for the US Headache Consortium. 2000 Practice parameter: evidence-based guidelines for migraine headache (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. http:// www.neurology.org/content/55/6/754.full.pdf+html. 14. Brandes JL, Saper JR, Diamond M, et al. Topiramate for migraine prevention a randomized controlled trial. JAMA. 2004;8:965-973. 15. Zissis NP, Harmoussi S, Vlaikidis N, Thomaidis T, Georgiadis G, Karageorgiou K. A randomized, double-blind, placebo-controlled study of venlafaxine xr in out-patients with tension-type headache. Cephalalgia. 2007;27:315-323. 16. D’Amico D, Solari A, Usai S, Santoro P, Bernardoni P, Frediani F. Improvement in quality of life and activity limitations in migraine after prophylaxis: Headache 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. A prospective longitudinal multicenter study. Cephalalgia. 2006;26:691-696. Ozyalcin SN, Talu GK, Kiziltan E, Yucel B, Ertas M, Disci R. The efficacy and safety of venlafaxine in the prophylaxis of migraine. Headache. 2005;45:144-152. Bendsten L, Rigmore Jensen R. Mirtazapine is effective in the prophylactic treatment of chronic tension-type headache. Neurology. 2004;62:17061711. Packard RC, O’Connell P. Medication compliance among headache patients. Headache. 1986;26:416419. Steiner TJ, Catarci T, Hering R, Whirmarsh T, Couturier EGM. If migraine prophylaxis does not work, think about compliance. Cephalalgia. 1994; 14:463-464. Mullerners WM, Whitemarsh TE, Steiner TJ. Noncompliance may render migraine prophylaxis useless, but once-daily regimens are better. Cephalalgia. 1998;18:52-56. Dirmaier J, Watzke B, Och U, Lehnert H, Pieper L, Wittchen HU. Diabetes in primary care: Prospective associations between depression, nonadherence, and glycemic control. Psychother Psychosom. 2010;79:172-178. Fisher L, Glasgow RE, Stryker LA. The relationship between diabetes distress and clinical depression with glycemic control among patients with Type 2 diabetes. Diabetes Care. 2010;33:1034-1036. Epub 2010 Feb 11. Park J, Nachman S. The link between religion and HAART adherence in pediatric HIV patients. AIDS Care. 2010;15:1-6. Cha E, Erlen JA, Kim KH, Sereika SM, Caruthers D. Mediating roles of medication-taking self-efficacy and depressive symptoms on self-reported medication adherence in persons with HIV: A questionnaire study. Int J Nurs Stud. 2008;45:1175-1184. Scher AI, Stewart WF, Liberman J, Lipton RB. Prevalence of frequent headaches in a population sample. Headache. 1998;38:497-506. Weng FL, Israni AK, Marshall MJ, et al. Race and electronically measured adherence to immunosuppressive medication after deceased donor renal transplantation. J Am Soc Nephrol. 2005;16:18391848. Golin CE, Liu H, Hays RD, et al. A prospective study of predictors of adherence to combination antiretroviral medication. J Gen Intern Med. 2002;17:756-765. 531 29. Gifford AL, Bormann JE, Shively MJ, Wright BC, Richman DD, Bozzette SA. Predictors of selfreported adherence and plasma HIV concentration in patients on multidrug, antiretroviral regimens. JAIDS. 2002;23:386-395. 30. Monane M, Bohn RL, Gurwitz JH, Glynn RJ, Levin R, Avorn J. Compliance with antihypertensive therapy among elderly Medicaid enrollees:The roles of age gender, and race. Am J Public Health. 1996;86:1805-1808. 31. Heckman BD, Holroyd KA, Tietjen G, et al. Whites and African American in specialty clinic respond equally well to treatment. Cephalalgia. 2009;2:650661. 32. Heckman BD, Catz S, Heckman TG, Miller JG, Kalichman SC. Adherence to antiretroviral therapies among HIV-infected rural persons. AIDS Care. 2004;16:219-230. 33. Jayawant SS, Feldman SR, Camacho FT, Yentzer B, Balkrishnan R. Prescription refills and healthcare costs associated with topical metronidazole in Medicaid enrolled patients with rosacea. J Dermatolog Treat. 2008;19:267-273. 34. Wu J, Lennie TA, Jong MJ, et al. Medication adherence is a mediator of the relationship between ethnicity and event-free survival in patients with heart failure. J Card Fail. 2010;16:142-149. 35. Gellad WF, Haas JS, Safran DG. Race-ethnicity and nonadherence to prescription medications among seniors: Results of a national study. J Gen Intern Med. 2007;22:1572-1578. 36. Fleck DE, Keck PE, Corey KB, Strakowski SM. Factors associated with medication adherence in African American and white patients with bipolar disorder. J Clin Psychiatry. 2005;66:646-652. 37. Bogart LM, Howerton D, Lange J, et al. Providerrelated barriers to rapid HIV testing in US urban non-profit community clinics, community-based organizations (CBOs) and hospitals. AIDS Behav. 2010;14:697-707. 38. Anderson ES, Winett RA, Wojcik JR. Selfregulation, self-efficacy, outcome expectations, and social support: Social cognitive theory and nutrition behavior. Ann Behav Med. 2007;34:304-312. 39. Anderson ES, Winett RA, Wojcik JR, Williams DM. Social cognitive mediators of change in a group randomized nutrition and physical activity intervention: Social support, self-efficacy, outcome expectations, and self regulation in the guide-to-health trial. J Health Psychol. 2010;15:21-32. 532 40. Ayotte BJ, Margrett JA, Hick-Patrick J. Physical activity in middle-aged and young-old adults: The roles of self-efficacy barriers, outcome expectancies, self-regulatory behaviors and social support. J Health Psychol. 2010;15:173-185. 41. Blamey R, Jolly K, Greenfield S, Jobanputra P. Patterns of analgesic use, pain and self-efficacy: A cross-sectional study of patients attending a hospital rheumatology clinic. BMC Musculoskelet Disord. 2009;10:120-137. 42. Cook PF, Emiliozzi S, El-Hajj D, McCabe MM. Telephone nurse counseling for medication adherence in ulcerative colitis: A preliminary study. Patient Educ Couns. 2010;81:182-186. Epub 2010 Jan 15. 43. Liang SY, Yates P, Edwards H, Tsay SL. Factors influencing opioid-taking self-efficacy and analgesic adherence in Taiwanese outpatients with cancer. Psychooncology. 2008;21:1100-1107. 44. Mak WW, Law RW, Woo J, Cheung FM, Lee D. Social support and psychological adjustments to SARS: The mediating role of self-care self-efficacy. Psychol Health. 2009;24:161-174. 45. Safren SA, Traeger L, Skeer MR, et al. Testing a social-cognitive model of HIV transmission risk behaviors in HIV-infected MSM with and without depression. Health Psychol. 2010;29:215-221. 46. Simoni JM, Frick PA, Huang B. A longitudinal evaluation of a social support model of medication adherence among HIV-positive men and women on antiretroviral therapy. Health Psychol. 2006;25:7481. 47. Bandura A. Self-efficacy. In: Ramachaudran VS, ed. Encyclopedia of Human Behavior. Vol. 4. 71-81. New York: Academic Press; 1994. 48. Sheffield D, Biles PL, Maixner W, Sheps DS. Race and sex differences in cutaneous pain perception. Psychosom Med. 2000;62:517-523. 49. Riley JL, Wade JB, Myers CD, Sheffield D, Papas RK, Price DD. Racial differences in the experience of chronic pain. Pain. 2002;100:291-298. April 2011 50. Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders, 2nd Edition. Cephalalgia. 2004;24:1-160. 51. Heckman BD, Holroyd KA, O’Donnell FJ, et al. Race differences in adherence to headache treatment appointments in persons with headache disorders. J Natl Med Assoc. 2008;100:243-247. 52. Heckman BD, Holroyd KA, O’Donnell FJ, Tietjen G, Utley C, Stillman M. Do psychiatric comorbidities influence headache treatment effectiveness? Results of a naturalistic longitudinal treatment study? Pain. 2000;146:56-64. 53. Heckman BD, Berlin K, Watakakosol R, Pierre V. Psychometric characteristics and measurement invariance of psychosocial headache measures in Caucasian and African American headache patients. Cephalalgia. 2010;29:650-651. 54. Spitzer R, Williams J, Kroenke K. Validity and utility of the PRIME-MD Patient Health Questionnaire in assessment of 3,000 obstetric-gynecologic patients: The PRIME-MD patient health questionnaire obstetrics-gynecology study. Am J Obstet Gynecol. 2000;183:759-769. 55. French DJ, Holroyd KA, Cornelia P, Malinoski PT, O’Donnell F, Hill KR. Perceived self-efficacy and headache related disability. Headache. 2000;40:647656. 56. Beck A. Cognitive Therapy and the Emotional Disorders. New York: Penguin; 1993. 57. Safren SA, O’Clerigh C, Tan JY, et al. A randomised controlled trial of cognitive behavioral therapy for adherence and depression (CBT-AD) in HIVinfected individuals. Health Psychol. 2009;28:1-10. 58. Weissman MM, Markowitz JC, Klerman G. Clinican’s Quick Guide to Interpersonal Psychotherapy. New York: Oxford University Press; 2007.