The Burden of Migraine in the United States: Current and Emerging Perspectives on Disease Management and Economic Analysis

Migraine Epidemiology: Incidence and Prevalence

Estimating migraine incidence within a given population is challenging. Because of the episodic nature of the migraine disorder, and because migraine is frequently underreported and undertreated, it can be difficult to pinpoint the age of migraine onset or to retrospectively assess its incidence. As a result, migraine incidence has been assessed in few longitudinal studies and research evaluating migraine incidence is less common than prevalence studies [13,14].

Historically, it has also been challenging to quantify migraine prevalence. This has been because of variations in methodologies employed to obtain data, a lack of universal clinical standards for migraine diagnosis, and variations in age and sex distributions in evaluated populations [15,16]. The International Classification of Headache Disorders (ICHD) initially published in 1988 (ICHD-1) and revised in 2005 (ICHD-2), has standardized migraine diagnosis and provided a foundation for a series of community-based studies that have enabled us to obtain a more encompassing picture of the disorder's incidence and prevalence [17]. Linet and colleagues conducted a study of migraine with and without preceding visual aura through a Washington County, Maryland telephone survey of individuals aged 12 to 29 years (n = 10,169). A total of 392 men and 1018 women reported a history of migraine [18]. They found that migraine occurs more frequently in women compared with men, presents earlier in life in males than in females, and that migraine with aura presents at an earlier age than migraine without aura.

The 1989 American Migraine Study I (AMS I) concluded that approximately 23.6 million Americans, including 17.6% of women and 5.7% of men, suffered from migraine disorder [19]. The American Migraine Study II (AMS II), conducted in 1999, found that the number of migraineurs had grown to 27.9 million in 1999 commensurate with the corresponding increase in US population [20]. The 1-year prevalence of migraine based on data collected in 1999 in AMS II was 18.2% among females and 6.5% among males. The American Migraine Prevalence and Prevention (AMPP) study in 2004 found that the unadjusted 1-year prevalence was 11.7% (17.1% in females and 5.6% in males) [21]. Nevertheless, as shown in Figure 1, only a minority of those meeting the ICHD-2 criteria for migraine reported having received a diagnosis of migraine.

Studies of consultation patterns show that in any given year, 48% of migraine sufferers will see a doctor for their headaches. These current consulters have more frequent and more severe headaches than nonconsulters [22]. As a consequence, prevalence estimates of the burden of migraine derived from health-care settings are prone to at least two forms of bias. Specifically, these studies underestimate prevalence (as many people remain undiagnosed) and overestimate disease severity (as consulters have more severe disease).

Prevalence related to age and sex. As noted above, during the adult years, migraine affects significantly more women than men. Migraine's impact is felt most strongly between the ages of 25 and 50 years, with its highest prevalence between the ages of 30 and 49 years. Peak prevalence occurs at approximately 40 years of age (Fig. 2). This is a particularly stressful burden, because for most people, these represent the most productive years of life [20,23,24]. To our knowledge, no studies have been published in the United States on the relationship between aging and migraine. Nevertheless, findings from European long-term follow-up studies demonstrate that migraine continues to be burdensome with aging. Although researchers in Denmark have found that migraine incidence remits with advancing age [25], relief from migraine does not always accompany aging; a 40-year longitudinal study conducted in Sweden found that more than 50% of migraine sufferers who had their first attacks when young were still having migraines at age 50 [26].

Migraine Management: Underdiagnosis and Undertreatment

Numerous studies report that the majority of migraine sufferers have either never been diagnosed and/or have not received adequate treatment for the disorder. In a follow-up analysis of the AMS I, Lipton and colleagues estimated that only 66% of migraine sufferers had ever consulted a doctor for headache [27]. Consultation rates were higher for females and for patients who experienced severe migraine-related disability, and tended to increase with increasing age [15,27]. Nonetheless, even among patients who had consulted physicians, only 65.7% of female and 57.9% of male migraineurs had been properly diagnosed [27].

Because safe and effective treatment is available for migraine, ensuring that migraineurs receive appropriate therapy is beneficial for individual sufferers, the health-care system, and society. Nevertheless, in a national sampling of the US population (n = 20,468), Celentano and colleagues found that only 28% of male and 40% of female US migraineurs used acute prescription migraine medication [28]. Among those with a physician diagnosis of migraine, only 49% of men and 64% of women reported current treatment with prescription medication [28]. This was the same rate found in the 2004 AMPP study. Just 56.2% of those with migraine had ever received a medical diagnosis, and 49% of migraineurs reported using over the counter (OTC) medications only to treat their migraine attacks [21,23]. Along the same lines, a large-scale US community survey conducted by Stewart and colleagues found that the majority of migraine sufferers who reported more than three headaches each month and/or significant disability associated with their headaches did not take acute prescription medication [15].

As with acute medication, prophylactic medication for migraine is also underused. The 2004 AMPP study found that over one-third (38.8%) of patients met the study guidelines for being offered (25.7%) or considered for (13.1%) prophylactic medication, but only 12.4% indicated that they were taking a migraine preventive medication [23]. Nevertheless, 17.2% were using medications that could be used to prevent migraine for other medical reasons (for example, a beta-blocker for hypertension) [21].

In some cases, a self-management approach to migraine using OTC therapy or other modalities is appropriate—for example, when patients only experience infrequent, mild migraine attacks. Nevertheless, Lipton and colleagues found that 61% of nonconsulters reported severe or very severe pain, while 67% reported severe disability or the need for bed rest with their headaches [27,29]. This makes it clear that in many cases, self-management is not efficacious.

Migraine Human Impact: Quality of Life

The concept of quality of life (QoL) proposes that health is defined not only by the absence of disease, but by the presence of physical, mental, and social well-being. HRQoL is a subset of overall QoL that encompasses individuals' health state, functional status (both physical and mental), and overall well-being [30].

A body of research now demonstrates that migraine poses a substantial and unique HRQoL burden on its sufferers. Patients with migraine experience not only diminished HRQoL compared with normal, healthy individuals but also decreased HRQoL comparable with or in some cases greater than that experienced by individuals with more serious diseases. For example, Turner-Bowker and colleagues conducted a survey by mail and the Internet to compare the burden of migraine with other chronic conditions in 7557 participants using the SF (Short Form)-8 Health Survey, an abbreviated generic survey derived from the SF-36 Health Survey. Results showed that the health-related impact of migraine was comparable with that experienced by patients with congestive heart failure, hypertension, or diabetes [30].

The effect of migraine on HRQoL was examined in a population-based study. Lipton and colleagues found that both the Physical and Mental Component scores of the SF-12 were reduced in a general sample of the US population with migraine; this effect was independent of the influence of depression [31]. Furthermore, migraine has been shown to impact patients' HRQoL during periods between attacks as a result of apprehension and worries associated with the anticipation of the next attack. Migraine patients have reported significantly more sleepiness (P = 0.007), less vigor (P < 0.05), and shorter duration of activity during the afternoon (P = 0.018) and evening (P = 0.006) compared with patients without migraine, and less than half of migraine patients return to normal functioning between migraine attacks [32].

These studies and others [33] confirm that migraine is a disabling condition that leads to compromised HRQoL. Nevertheless, as the next section will discuss, the majority of migraine costs are borne by society and the health-care system in terms of lost productivity and increased health-care utilization.

The Cost of Migraine

When costs associated with migraine are analyzed, they are usually categorized as direct and indirect costs. Direct costs are those related to the use of medical resources, including physician visits for diagnosis and/or treatment, emergency room (ER) visits, diagnostic procedures, and medication. Indirect costs are generally assessed in terms of temporary disability, reduced functionality, lost productivity, and associated costs to employers. Tables 2 and 3 provide a summary of direct and indirect costs of migraine published since 2002. As evidenced in Table 2 and 3, cost and productivity estimates varied widely across the studies because of methodological differences, such as: overall severity of disease in population sampled, operational definition of migraine, and medications and services included in aggregate totals.

The Minimal-Therapist-Contact Approach

Economic considerations associated with migraine received little attention prior to the development of triptans, the first effective treatment for acute migraine. Prior to the introduction of triptans, ergotamine was the mainstay of acute migraine pharmacological treatment. To our knowledge, no published economic evaluations of the ergot alkaloids are available. A few rare exceptions exist, most notably studies in the 1980s that examined the efficacy and cost-effectiveness of the minimal-therapist-contact approach to migraine, in which the patient was taught cognitive–behavioral therapy “self-help” techniques in a few sessions, compared with a comparable clinic-based approach. Richardson and McGrath compared the efficacy of the minimal-therapist-contact approach with a clinic-based cognitive–behavioral program and found that the frequency of headache, duration of intensity, and peak intensity were significantly reduced in both groups for up to 6 months [50]. Another study also found that both approaches resulted in a significant reduction in headache frequency, duration, and peak intensity [51]. In the absence of effective acute treatments, the minimal-therapist-contact approach was found to be more cost-effective [50,51].

Early Studies of the Economic Burden of Migraine

Starting in 1992, research on the economic burden of migraine in the United States began to appear. For example, to assess the costs associated with migraine, both direct and indirect, Osterhaus and colleagues surveyed patients who had participated in clinical trials of an antimigraine compound [52]. Migraine frequency and costs associated with health-care resource use, decreased productivity, and missed workdays were measured. By 1994, a number of articles had assessed the economic impact of migraine on both the health-care system and society, and found that cost-effective interventions could significantly reduce the burden of migraine both for society and individual sufferers [15].

At that time, researchers began to examine medical and pharmacy claim databases to assess the migraine-related costs associated with health-care resource utilization and medication. Methodologies applied included comparisons of resource use among migraineurs compared with nonmigraineurs, drug costs compared with the burden imposed by other resource use, as well as dollar cost of the burden of migraine. These reviews often concluded that more data and improved assessments of the associated costs and severity of migraine were needed to understand the true burden of the disease on society [15,42,53,54]. Today, researchers continue to strive to find methodologies that will fully capture the multifaceted economic impact of migraine.

Introduction of Sumatriptan

Early examinations of the economic burden of migraine coincided with the introduction of the first triptan, sumatriptan, used for treatment of acute migraine attacks. At the time, sumatriptan was an exceptional drug in that it was expensive relative to existing treatment, but also highly effective. This led to a new methodology for the evaluation of cost-effectiveness of migraine care, one that was based upon a wider range of criteria than drug cost alone. In particular, research began to examine the extent to which the acquisition cost of a drug such as sumatriptan was offset by reductions in both direct costs, such as overall health-care expenditures, and indirect costs, such as productivity loss. For example, Legg et al. evaluated the cost-effectiveness of sumatriptan by measuring health-care costs (emergency department [ED] visits, physician office visits, hospital visits, and medications) and number of days without migraine-related disability before and following treatment with sumatriptan [55]. In other randomized controlled trials of sumatriptan, the average disability time per treated attack was approximately 2 hours, compared with 4 hours in untreated patients. By evaluating disability-free time intervals as a study end point, researchers attempted to establish a clearer picture of the overall savings to society when sumatriptan became available [55–57].

Introduction of other Triptan Drugs

As other triptans, such as rizatriptan, almotriptan, and eletriptan, were developed and introduced, economic comparisons of triptans were prepared. Some studies employed efficacy data from randomized clinical trials and combined this with an economic analysis of drug cost. Other studies examined the additional costs associated with the triptan-associated adverse event of chest pain. Meta-analyses of overall efficacy and adverse event profiles across multiple clinical trials were evaluated. Last, others compared drug costs and aggregate numbers needed to treat in an effort to assess the wide-ranging cost-effectiveness of triptans [58–63]. Studies of the economic impact of acute treatments continue [64], and the value of triptans continues to be compared to nontriptan acute treatments [65].

The Costs of Medication Overuse and the Emergence of Migraine Prophylaxis

By the late 1990s, both payers and clinicians began to be concerned that triptans were being overused in the treatment of acute migraine, and in the early 2000s, researchers in Europe began studying medication-overuse headache (MOH). In a 2002 prospective study of 98 patients, Limmroth and colleagues found that triptan overuse led to MOH more quickly and at lower dose levels than did overuse of ergots or analgesics. They also found that patients were more likely to describe a daily migraine-like MOH with triptan overuse, as opposed to tension-type headache with ergots and analgesics [66].

Two longitudinal, retrospective review studies of medical and pharmacy claims in HMOs found that limits on monthly triptan access reduced drug costs. Nevertheless, the two studies (conducted in 1999 and 2003) had conflicting findings as to the effect of these limits on other health-care resource costs, with the first study finding that additional migraine-related costs did not significantly change, and the second study finding reduced costs [67,68]. Nevertheless, in one of the above-cited studies on sumatriptan limitation, Goldfarb and colleagues noted that as sumatriptan use was reduced, prescriptions for drugs that could be used for migraine prevention increased by 33.9% [67]. Indeed, as early as the mid-1990s, the costs of various prophylactic treatments for migraine began to be compared [69].

In part spurred by the concern about the possible over use of triptans, there has been a focus on prophylactic treatment of migraine. Current clinical guidelines for prophylactic treatment are contained in Table 4. Following this clinical focus on prophylactic treatment, a new body of research has emerged to examine the economic impact of prophylactic treatment for migraine. There is sufficient evidence and consensus exists to recommend topiramate, divalproex sodium, amitriptyline, timolol, and propranolol as effective first-line agents for migraine prophylaxis [70]. According to Silberstein and colleagues, migraine preventive therapy is effective in reducing overall resource utilization, including the use of other migraine medications and the frequency of physician and emergency room visits [71].

It was not overlooked that the economic benefits of costly migraine treatments were likely to be greatest in patients with the most severe disease. In 2002, Adelman and colleagues observed that in patients responsive to acute medications, preventive therapy was only cost-effective if the patient experienced a high frequency of migraine or comorbid disease [72]. Brown and colleagues found that the costs of topiramate for migraine prevention could be offset by gains in reduced headache frequency, disability hours, and lost productivity. Nevertheless, their cost-effectiveness analysis modeled the use of topiramate for patients with frequent (two or more per month) attacks or for patients who experienced substantial disruption of daily activities [71,73].

Although the threshold for prophylactic treatment of migraine has remained relatively high, recent clinical findings suggest that lowering this threshold may need to be considered. This reconsideration of prophylactic treatment has important implications for the standards of economic evaluation in migraine and suggests that the need for a reassessment of the standards of economic evaluation of migraine preventive therapy may be necessary.

Long-Term Impact of Migraine and the Cost of Prophylaxis

Recent clinical studies suggest that contrary to previous assumptions, migraine may be a progressive disease that potentially leads to chronic headache conditions, impaired long-term psychosocial functioning, and permanent neurologic damage. Epidemiologic studies have recently indicated that 4% to 5% of the general population suffers from chronic daily headache (CDH), which is characterized by 15 or more headache days per month, and that transformed migraine—the transformation of episodic migraine to the chronic form of the disorder—accounts for approximately half of these patients. Although medication overuse and baseline frequency of attacks explained some CDH, a subgroup analysis indicated that migraine in and of itself was potentially a progressive disorder [71,74].

In addition, evidence is accumulating, indicating that for some migraineurs, repeated headache episodes may result in permanent changes to central nervous system (CNS) structure or function, in particular free radical formation, neuronal injury, and iron deposition [75,76]. Migraine may also be a risk factor for subclinical brain lesions. A Dutch population-based study found a higher rate of posterior circulation infarcts and deep white matter lesions among subjects aged 30 to 60 years with a history of migraine compared with nonmigraine controls [77]. Nevertheless, the precise relationship of white matter lesions to migraine is not yet known [75,77]. Repeated episodes of oligemia during migraine aura have been suggested as the cause of subtle neuronal damage to the visual system, and it has been suggested that a similar mechanism may cause damage in the auditory system [75,76].

Not only has migraine been implicated in causing permanent changes to CNS structure and function, it has also been implicated as a possible cause of cardiovascular and cerebrovascular disease. In fact, migraine with aura has been shown to increase the risk of major cardiovascular disease and stroke by twofold [78]. It is generally accepted that physiological changes responsible for head pain and aura associated with migraine are of neurovascular origin. Similarly, the vasculature may play a role in adverse cardiovascular events stemming from migraine. Potential mechanisms for migrainous infarction include microcirculatory vasoconstriction, intracerebral large vessel spasm, and vascular changes because of cortical spreading depression. Stroke occurring remote of the migraine attack is thought to be caused by endothelial-related hypercoagulability, arterial dissection, and cardioembolism. Endothelial dysfunction as a cause of or subsequent to migraine may have a role in ischemic heart disease [79]. Indeed, deleterious clinical outcomes associated with migraine extend far beyond just symptoms of migraine attack.

In addition to the long-term physical effects of migraine, clinical experience suggests that prolonged psychosocial impairment because of headache is more likely among patients who suffer from chronic forms of the disease, with social functioning improvement often lagging behind migraine symptom improvement [75].

Patients that experience migraine-like symptoms also experience substantial impairment. Patients with probable migraine, defined as a headache that meets all criteria except migraine with or without aura, experience increased depression, increased disability, and reductions in HRQoL compared with those without migraine. More importantly, decrements in these measures because of probable migraine were similar to migraine. This finding expands our current understanding and suggests that probable migraine is a form of migraine, and should be considered when developing a treatment plan [3].

If confirmed, these findings would suggest that early, aggressive intervention for migraine prevention, targeted to reduce headache frequency and/or severity, may reduce risk factors for long-term CNS and vascular damage, as well as prevent the progression of migraine to the more debilitating state of transformed migraine [71,75]. Currently, most physicians do not think of migraine as they do illnesses such as diabetes, hypertension, asthma, or rheumatoid arthritis. Nevertheless, migraine may need to be re-examined from a similar perspective, as a chronic illness for which preventative therapy is important to delay or reduce more serious forms or complications [75]. Similarly, economic evaluations to date have focused on the cost-effectiveness of migraine as an episodic rather than a chronic and progressive disease. Should these more recent findings be established, the standards of economic evaluation of migraine will undoubtedly need to be re-envisioned once again to take into account the long-term effects and costs of the disease.

Future Studies of Migraine

The traditional goal of prevention is to prevent headache attacks and thereby limit the current pain and disability experienced by headache sufferers. In the near term, we should additionally assess the ability of migraine prevention to reduce the incidence and associated costs of transformed migraine. In the longer term, the value of prevention may be found to be in its ability to reduce the incidence and associated costs of cardiovascular, cerebrovascular, and long-term CNS damage consequent to frequent and severe migraines. The economics of migraine prevention will change as future studies more clearly define the burden of transformed migraine as well as disease progression.