Introduction

Migraine is a chronic neurovascular disease characterized by recurrent, debilitating headache attacks separated by pain-free periods. According to the Global Burden of Diseases Study (2015), migraine ranks as one of the leading causes of disability-adjusted life-years (DALY) [1••]. Migraine is associated with a considerable economic burden to the patient, the family, and the society, due to both healthcare resource utilization and loss of productivity [2].

Migraine is associated with a wide range of neurological, psychiatric, cardiovascular, and endocrine comorbidities that significantly influence disease progression as well as therapeutic strategies. Concerning cardiovascular comorbidity, migraine patients, and particularly migraine with aura (MA), are at increased risk of developing stroke [3], angina and myocardial infarction [4], patent foramen ovale, arterial hypertension, and cervical artery dissection [5•] in comparison with non-migraineurs. The increased risk for several cardiovascular diseases in migraineurs suggested that a common metabolic risk factor underlies these pathologies and prompted several investigation in order to identify common metabolic abnormalities. In addition, recent studies suggested a comorbidity between migraine and several endocrine diseases, as hypothyroidism, endometriosis, and polycistic ovarian syndrome (PCO) [6•, 7, 8]. Understanding the mechanisms underlying these comorbidities may significantly improve our understanding of migraine pathogenesis and may open new therapeutic strategies.

The purpose of this review was, therefore, to evaluate recent advances in the comorbidity between migraine and metaboloendocrine disorders and to discuss relevant pathogenic mechanisms.

Methods

We performed a computerized search of PubMed and Cochrane Library data to identify English-language articles published in the last 5 years (January 2013–December 2017). Search terms included migraine, insulin resistance, diabetes, obesity, hypertension, metabolic syndrome, and hypothyroidism. Relevant articles were selected for the review, according to specific headings. Out of 321 articles screened, 56 were selected for review.

Insulin Resistance

Insulin resistance (IR) can be broadly defined a condition characterized by a subnormal physiological response to normal insulin concentrations. As a result, increased quantities of insulin are produced in order to maintain adequate intracellular glucose concentrations. IR is the predisposing factor for several diseases like type 2 diabetes mellitus, obesity, atherosclerotic vascular disease, heart disease, stroke, PCO, and neuropathological processes underlying cognitive aging and dementia. Insulin resistance is a highly prevalent syndrome in developed as well as developing countries, and researchers have estimated that 3–16% of the general population in the Western Countries are insulin-resistant [9].

Various indices of insulin resistance have been proposed in the last 2 decades. Hyperinsulinemic euglycemic clamp is known to be the “gold standard” for the measurement of IR, but this is an expensive and time-consuming method. At present, there are two groups of IR indices that are currently used in clinic-based and epidemiological studies: (1) indices calculated using fasting plasma concentrations of insulin, glucose, and triglycerides, and (2) indices calculated using plasma concentrations of insulin and glucose obtained during a standard (75 g glucose) 120′ oral glucose tolerance test (OGTT). The former group includes homeostasis model assessment of insulin resistance (HOMA-IR), which is proportional to the product of the fasting insulin and fasting glucose level, β-cell function (HOMA-B), QUIKI Index, Matsuda, and Stumvoll indexes. From OGTT is possible to calculate the Impaired Fasting Glucose (IFG), defined as a fasting glucose level above a predetermined cutoff (commonly 100 mg per deciliter) and the impaired glucose tolerance (IGT) test, defined as a glucose level above a predetermined cutoff, commonly 140 mg/dl. At present, there is no widely accepted definition of insulin resistance. According to WHO, insulin resistance is defined as the highest quartile of the HOMA-IR index in nondiabetic subjects. In the general population, there is a considerable individual variation in sensitivity to insulin, with the most insulin-sensitive subjects being as much as six times as sensitive to the hormone as those identified as most resistant.

In 2005, our research group provided for the first time evidence that migraine is associated with an altered insulin sensitivity. In detail, we performed the OGTT test in a group of 30 young, nonobese, nondiabetic, normotensive migraine patients and in 15 healthy controls. Fasting-based indexes of insulin sensitivity were not significantly different between patients and controls. However, during the OGTT, glucose plasma concentrations were significantly higher in migraineurs than in controls. Several indexes of insulin sensitivity, like ISI Stumvoll and OGIS-180, showed a condition of insulin resistance in migraineurs [10]. Our results were confirmed by Cavestro et al. that in a larger group of 84 unselected migraine patients found that, during the OGTT, both glucose and insulin concentrations were significantly higher in patients than in healthy controls [11].

More recently, several additional clinical studies confirmed the presence of insulin resistance in both episodic and chronic migraine [12, 13•, 14]. Hyperinsulinemia is associated with a 5.67-fold higher risk for migraine, when comparing the lowest with the higher quartile of HOMA [15]. Only a few studies, in contrast to what has been shown by the majority of the published studies, did not support the association of migraine and IR [16, 17]. However, these studies evaluated IR in migraineurs only using fasting glucose and insulin concentrations and not during a dynamic test like OGTT. Table 1 shows the results of all the studies that examined IR in patients with migraine. Taken together, these studies suggest that migraine is characterized by an alteration of insulin receptor response that is clearly evident during a stress condition.

Table 1 Studies investigating insulin resistance in migraine
Full size table

The insulin receptor is a member of the ligand-activated receptor and tyrosine kinase family of transmembrane signaling proteins. These proteins are involved in several physiological functions like cell differentiation, growth, and metabolism. The insulin receptor is encoded by a single gene (INSR) that is located on chromosome 19p13.2. The protein is proteolytically processed to generate alpha and beta subunits that form a heterotetrameric receptor. Binding of insulin or other ligands, like IGF-I and IGF-II, to the receptor induces structural changes leading to auto-phosphorylation of various tyrosine residues within the intracellular domain of the β-chain. These changes facilitate the recruitment of specific adapter proteins such as the insulin receptor substrate proteins (IRS), PI3K, and Akt. The activation of the glucose transporter proteins promotes the glucose influx in different cells. In addition, insulin receptors regulate several complex physiological actions like the synthesis and storage of carbohydrates, lipids, and proteins. The INSR gene is widely expressed in the body, and interestingly, the brain was also identified as an insulin-sensitive organ. Indeed, insulin receptors are abundantly distributed throughout the brain, and insulin action produces multiple behavioral and metabolic effects within the central nervous system. Figure 1 shows the insulin receptor and its complex role in metabolic signaling.

Fig. 1
figure 1

Insulin receptor and mechanisms of signaling. PIP2 phosphatidylinositol 4,5-bisphosphate, PIP3 phosphatidylinositol (3,4,5)-trisphosphate, IRS-1 insulin receptor substrate 1, PI3K phosphatidylinositol-3-kinase, PKB/AKT protein kinase B also known as Akt, GLUT-4 glucose transporter type 4, PGC-1α peroxisome proliferator-activated receptor gamma coactivator 1-alpha, GSK3 glycogen synthase kinase 3, PDE3B phosphodiesterase 3B, mTOR mechanistic target of rapamycin, FOXO forkhead box protein O, GS glycogen synthase, cAMP cyclic adenosine monophosphate, PKA protein kinase A, HSL hormone sensitive lipase, ACC acetyl-CoA carboxylase, ACL ATP-citrate lyase

Full size image

In recent years, biological mechanisms underlying insulin resistance in migraine have been investigated. In experimental animals, insulin-induced hypoglycemia significantly prolongs the duration of cortical spreading depression (CSD) that is believed to cause the migraine aura and trigger headaches [18•]. Clinical studies showed that fasting, a condition that frequently provokes headache attacks in migraineurs, is characterized by an insulin receptor activation. Intriguingly, intravenous injection of insulin was shown to be effective in provoking migraine aura. Two genetic studies found a significant association between polymorphisms of the INSR gene and migraine [19, 20]. Hyperinsulinemia in migraine is correlated with increased NO stress [15]. Finally, Bernecker et al. recently showed that in female migraineurs, hyperinsulinemia is associated with a significant increase of both leptin and glucagon-like peptide 2 (GLP-2) concentrations [12] .

The detection of insulin resistance in migraine patients may have several clinical applications. Aerobic exercise reduces insulin resistance and therefore could help to reduce the frequency of headache attacks in migraineurs. In addition, strategies improving insulin sensitivity may be of benefit in migraine. In recent years, several insulin-sensitizing drugs have become commercially available and thence could be of interest to test these drugs in migraine prophylaxis. Recently, a small exploratory study showed that administration of alpha-lipoic acid (400 mg b.i.d. for 6 months), a drug that lowers insulin resistance, is associated with a reduction in the number of attacks and the days of treatment in migraineurs with IR [21]. In addition, a controlled trial showed that in obese migraineurs, bariatric surgery significantly reduces frequency and duration of headache attacks, significant increase in the number of migraine-free days [22]. Randomized controlled studies are needed in order to verify the safety and efficacy of these therapeutic strategies in migraine prevention.

Diabetes Mellitus

Diabetes mellitus (DM) is a chronic, lifelong disease caused by a deficiency in production of insulin by the pancreas, or by the ineffectiveness of the produced insulin. Such a deficiency results in increased concentrations of glucose in the blood, which in turn damages many of the body’s systems. There are two main forms of diabetes: type 1 diabetes (T1DM), an autoimmune disorder that develops most frequently in children and adolescents, and Type 2 diabetes (T2DM) which results from the body’s inability to respond properly to the action of insulin produced by the pancreas. T2DM occurs most frequently in adults and accounts for about 90% of all diabetes cases worldwide.

Publication of several studies showing an impaired insulin sensitivity in migraine leads to the speculation that migraine patients are at increased risk for DM. After initial conflicting results, all recent studies showed that there is no significant association between diabetes and migraine. Both case-control clinical studies [23•] and prospective epidemiological studies [24, 25] clearly demonstrated that migraineurs are not at increased risk of developing T2DM or pre-diabetes. Recently, data from the Nord-Trøndelag Health Surveys showed that type 1 diabetes is associated with a lower prevalence of migraine (OR = 0.47, 95% CI 0.26–0.96), and any other type of headache (OR = 0.55, 95% CI 0.34–0.88) when compared to subjects without DM [26]. The protective effects of type 1 diabetes on migraine risk were observed also in children, suggesting that vascular complication of T1DM may be protective for migraine [27]. These finding is in accord with an old clinical study of Blau and Pyke showing that in migraineurs, the onset of diabetes significantly reduces or abolishes the frequency of migraine attacks [28].

Biological mechanisms underlying this protective effect of diabetes on the risk of developing migraine attacks are, at present, unclear. A recent study in normal rat brain showed that glycemic state and cerebral glucose availability are important modulators of the susceptibility to as well as the duration of cortical spreading depression (CSD), the neurobiological mechanism of migraine aura. Hyperglycemia renders the cerebral cortex more resistant to CSD initiation and hastened CSD recovery, whereas hypoglycemia had the opposite effect on CSD durations [18••, 29]. In diabetic patients, the prevalence of migraine varies strongly depending on age and is significantly influenced by pharmacological treatment: Patients using diabetic drugs have an overall reduced prevalence of medically treated migraine when compared with the nondiabetic population (OR = 0.72, 95% CI = 0.68–0.75) [30].

Obesity

Overweight and obesity are nowadays a major public health problem worldwide. Obesity has reached epidemic proportions globally, with more than 1 billion adults overweight—at least 300 million of them clinically obese—and is a major contributor to the global burden of chronic disease and disability.

The relationship between migraine and obesity was firstly evaluated in a clinic-based studies, showing that obese patients were significantly more likely than age-matched normal-weight controls to have migraine. Then, several large cross-sectional studies investigated this relationship in population-based samples. However, in these studies, obesity was not specifically associated with increased prevalence of migraine but was related only to headache attack frequency [31,32,33]. Furthermore, migraineurs in the severely obese group were nearly twice as likely to report severe headache pain, compared to those in the normal weight group [34]. To resolve the discrepancies observed between clinical and epidemiologic studies, a recent meta-analytic analysis was performed on 12 different studies, encompassing data from 288,981 participants. The age- and sex-adjusted pooled risk of migraine in obese subjects was increased by 27% compared with those of normal weight (OR 1.27; 95% CI 1.16–1.37, p < 0.001). In underweight individuals, the pooled risk of migraine was marginally increased by 13% compared with those of normal weight (OR 1.13; 95% CI 1.02, 1.24, p < 0.001) [35••]. This meta-analysis efficaciously showed that the risk of migraine is significantly increased in both obese and underweight individuals.

A large array of physiological, psychological, and behavioral mechanisms may contribute to the co-occurrence of migraine and obesity. Several neurotransmitters and peptides that are involved in migraine pathophysiology, as serotonin and orexins, also regulate eating behavior. In recent years, several studies investigated the role of the adipokines in migraine pathophysiology, suggesting a role for these mediators in the comorbidity with obesity. Adipose tissue is no longer considered to be an inert tissue that stores fat but acts as an endocrine organ, releasing several hormones that regulate nutritional intake, sensitivity to insulin and inflammatory processes. These hormones, called adipokines, have been recently investigated in migraine. In a pilot study of women with episodic migraineurs, adiponectin concentrations were associated with migraine severity and predictive of acute treatment response, representing novel potential biomarkers and drug targets for episodic migraine [36]. In addition, plasmatic leptin concentration resulted significantly increased in migraine with aura patients [37]. Finally, a role for adiponectin and resistin was shown in chronic migraine [38••]. Intriguingly, in this study, a positive correlation between leptin concentrations and both indices of insulin resistance and markers of inflammation was found.

Recently, a randomized controlled study compared the effects of weight loss versus standard migraine education in a group of 110 women with a comorbid migraine and overweight or obesity. Both strategies yielded similar results, with a significant reduction in migraine headaches [39]. Further controlled studies are needed in order to evaluate safety and efficacy of various strategies for weight loss in migraine prevention.

Hypertension

A large body of work has investigated the relationship between hypertension and migraine, but this topic still needs to be elucidated. From a clinical perspective, several studies have shown that young migraineurs generally have low-pressure values, while with advancing age (mainly after menopause in women), arterial pressure values progressively increase and patients frequently develop arterial hypertension.

Several studies have investigated the association between migraine and both diastolic and systolic hypertension with conflicting results. Many epidemiologic studies found no evidence of an association between migraine and hypertension or demonstrated an inverse relationship between the two diseases [40,41,42,43]. In a recent clinic-based study, hypertension was more frequent in patients with headache in respect to the general population. In the analysis of clinical subgroups, positive associations were found for episodic and chronic tension-type headache, but less so for migraine [44]. In a further multi-ethnic urban population study, hypertension was found to be associated with migraine, both with and without aura (MO). This association was particularly evident for those subjects with hypertension of long duration (> 9 years) [45•]. Finally, hypertension has been identified as one of the most important factors of episodic migraine chronification and highly increases the cerebrovascular and cardiovascular risk of migraineurs. Conclusively, many studies support the hypothesis that migraine patients have an increased risk of developing hypertension, while hypertensive subjects do not seem to have an increased risk of migraine or any other types of primary headaches.

The association between migraine and hypertension may be driven by several factors, like as environmental factors, shared biological factors, or a common genetic vulnerability. Environmental factors that may be associated with both migraine and hypertension include dietary factors, physical inactivity, and chronic stress, all factors also linked to metabolic syndrome. Migraine and arterial hypertension may share common pathogenic mechanisms as endothelial dysfunction, deficiency of autonomic cardiovascular regulation, and renin angiotensin system involvement. Finally, preventive therapy for migraine takes advantage of several antihypertensive agents as beta-blockers, and more recently angiotensin-converting-enzyme inhibitors and angiotensin II receptor blockers. Further clinical and experimental studies are required to better elucidate the role of blood pressure abnormalities in migraine pathophysiology and comorbidity.

Metabolic Syndrome

Metabolic syndrome (MetS) is characterized by a cluster of metabolic abnormalities including insulin resistance, hypertension, dyslipidemia, obesity, microalbuminuria, and a proinflammatory state. These medical conditions are interrelated and share common underlying mediators, pathogenic mechanisms, and pathways. MetS has become one of the major health-related challenges worldwide. MetS is a highly prevalent medical condition, affecting approximately 25% of the general population, and is associated with a 2-fold risk for atherosclerotic cardiovascular disease and 5-fold increase for T2DM. Over the last two decades, different definitions and diagnostic criteria for MetS have been proposed, like the EGIR (European Group for the Study of Insulin Resistance), the ATPIII (Adult Treatment Panel III), the ACEE (American Association of Clinical Endocrinologist), and the WHO criteria. Evidence for insulin resistance is an absolute requirement in the WHO definition, and two additional criteria, including obesity, dyslipidemia, hypertension, and microalbuminuria, have to be met.

In the past decade, an increased attention has been devoted to the evaluation of migraine as a risk factor for MetS. The results of different studies are conflicting, partially due to the different MetS criteria used. First studies provided evidence that patients with migraine are at increased risk of developing MetS [46,47,48], whereas more recent studies found a correlation with only some of the metabolic syndrome components, like body mass index and waist circumference. In addition, a correlation between the presence of MetS and age, gender, number of triggers, years of headache, and duration of headache attacks was found in migraineurs.

Recently, the prevalence of metabolic syndrome was carefully investigated in migraine subgroups, like migraine with aura and chronic migraine with medication overuse headache (MOH). Streel et al. found that migraine with aura subjects, when compared to controls, is at higher risk of MetS (OR 3.45; 95% CI 1.63–7.29) in contrast to migraine without aura individuals. Singularly considering MetS components, MA was positively associated with low HDL-cholesterol (OR 2.26; 95% CI 1.08–4.74), hyperglycemia (OR 2.77; 95% CI 1.30–5.88), and abdominal obesity (OR 2.03; 95% CI 1.07–3.86) [49]. The association between MA and MetS is influenced by smoking: In comparison with headache-free controls, the adjusted incident risk ratio (IRR) for migraine with aura is 2.10 (95% CI 1.53–2.89) in smokers and 1.39 (95% CI 1.03–1.86) among non-smokers [49]. These studies suggest that migraine with aura patients, being at higher risk of MetS, can benefit from a systematic screening of metabolic risk factors.

Intriguingly, a Chinese clinical-based study showed that chronic migraine in women is significantly associated with MetS (OR = 5.342, p = 0.032), but when the chronic migraine patients are comorbid with medication overuse headache (MOH), the risk for MetS exponentially increased (OR = 12.68, p = 0.007). Furthermore, MOH was associated with abdominal obesity and hypertension among the components of MetS (OR = 4.205, p = 0.043; OR = 3.234, and p = 0.039; respectively) [50••]. This study suggests an important contribution of analgesic overuse in the risk of developing MetS in migraineurs. Table 2 summarizes the results of all the studies that investigated the relationship between metabolic syndrome and migraine.

Table 2 Studies investigating metabolic syndrome in migraine
Full size table

Hypothyroidism

Migraine and hypothyroidism are frequent diseases in the general populations with a lifetime prevalence of 12 and 2%, respectively. Both disorders share several demographic and socioeconomic characteristics. Since 1998, numerous studies showed an association between migraine and hypothyroidism, with hypothyroidism being significantly more prevalent in subjects with chronic migraine compared to those with episodic migraine [54, 55]. However, only recently, this association was deeply investigated. Using data from the Fernald Medical Monitoring Program (FMMP), a medical surveillance program of residents who lived near Cincinnati (OH), Martin et al. clearly demonstrated that headache disorders are risk factors for the development of new onset hypothyroidism, with migraine patients showing an increased risk of 41% of developing this disorder [6•]. In addition, reviewing the clinical characteristics of headache attacks in a cohort of 213 patients with hypothyroidism, Lima Carvalho et al. showed that approximately 60% of these patients fulfilled the ICHD-3beta criteria for migraine [56]. Finally, we recently demonstrated in a case-control study that patients with subclinical hypothyroidism have a significantly increased risk of developing migraine [57].

Taken together, the results of these studies suggest that migraine and hypothyroidism are linked by a bidirectional relationship. Both genetic and immune mechanisms may explain this association. Discovering the biological basis of this phenomenon may lead to better understanding of migraine pathophysiology.

Conclusions

The aim of this paper was to review and to stimulate discussion on the relationship between migraine and metabolic and endocrine disorders. Recently, several studies clearly showed that there is a complex relationship between migraine and metaboloendocrine diseases like obesity, diabetes, hypertension, metabolic syndrome, and hypothyroidism. This association is more evident in patients with migraine with aura and chronic migraine. The pathophysiological mechanisms underlying this relationship are still under investigation. However, insulin resistance seems to be the more relevant mechanism explaining this association.

Insulin sensitivity is clearly impaired in migraine, even in young, nonobese, nondiabetic, normotensive patients. Migraineurs, particularly under stress, have significantly increased plasmatic concentrations of both insulin and glucose, leading to impairment in complex metabolic patterns. Abnormalities in these pattern may explain the increased risk for vascular disorders observed in several epidemiological studies. In recent years, several studies clearly showed that insulin may modulate several critical brain functions like metabolism, cognition, and motivated behaviors. Therefore, a condition of insulin resistance within the central nervous system may have an important role in explaining the complex pathogenesis of migraine.

In conclusion, at present, it is not possible to answer the question if migraine is primarily a metaboloendocrine disorder. Additional research is needed to better investigate the biological mechanisms underlying metabolic alterations in migraine patients and to define the best clinical practices in patients. Increased physical activity, weight loss, and management of cardiovascular risk factors may be of relevance for prevention and treatment of migraine attacks.