Neurol Sci DOI 10.1007/s10072-014-1719-3
REVIEW ARTICLE
Adiponectin and migraine: systematic review of clinical evidence Giuseppe Lippi • Tiziana Meschi • Camilla Mattiuzzi Loris Borghi • Giovanni Targher
•
Received: 17 February 2014 / Accepted: 10 March 2014 Ó Springer-Verlag Italia 2014
Abstract Although the pathogenesis of migraine is very complex and has not been thoughtfully elucidated, general consensus exists to date that this condition should be considered a primary neurovascular disorder with an important inflammatory component. Owing to epidemiological evidence of increased risk of migraine in overweight and obese subjects and to the inverse relationship that exists between serum adiponectin concentration and obesity, we performed an electronic search on Medline, Scopus and Web of Science, using the keywords ‘‘migraine’’ and ‘‘adiponectin’’ with no language or date restriction to explore the existence of an association between serum adiponectin and migraine. According to our search criteria, five studies were finally included in this systematic review, four cross-sectional (totaling 300 patients with migraine and 177 controls) and
G. Lippi Laboratory of Clinical Chemistry and Hematology, Academic Hospital of Parma, Parma, Italy G. Lippi (&) U.O. Diagnostica Ematochimica, Azienda Ospedaliero-Universitaria di Parma, Via Gramsci, 14, 43126 Parma, Italy e-mail: [email protected]; [email protected] T. Meschi L. Borghi Department of Clinical and Experimental Medicine, Academic Hospital of Parma, Parma, Italy C. Mattiuzzi Service of Clinical Governance, General Hospital of Trento, Trento, Italy G. Targher Section of Endocrinology, Diabetes and Metabolism, Department of Medicine, University and Azienda Ospedaliera Universitaria Integrata of Verona, Verona, Italy
one interventional. Collectively, the results of our analysis suggest that a link between serum adiponectin and migraine remains elusive, at the best. The four cross-sectional studies failed to find any significant association, whereas the outcome of the single interventional study reported a rather modest variation of serum adiponectin concentration in a very limited sample size. Further larger studies are needed to firmly establish the existence of a relationship between adiponectin metabolism and migraine. Keywords Review
Migraine Headache Obesity Adiponectin
Introduction Migraine, which is conventionally defined as recurrent headache manifesting in attacks lasting 4–72 h [1], is a highly prevalent condition. According to recent epidemiological data, the prevalence of migraine approximates 15 % in the general population, which makes it the third most common disease worldwide in both genders, behind dental caries and tension-type headache [2]. It is also noteworthy that approximately 3 % of worldwide disability in terms of years lived with a disability are attributable to migraine [3]. The recent document issued by the Headache Classification Committee of the International Headache Society (IHS) classifies migraine in three major forms that are migraine without aura, migraine with aura and chronic migraine, wherein aura is defined as unilateral fully reversible visual, sensory, or additional central nervous system symptoms that typically onset gradually and are followed by headache and associated migraine symptoms [1]. Although the pathogenesis of migraine is very complex and has not been thoughtfully elucidated, there is general
123
Neurol Sci
consensus to date that migraine should be considered as a primary neurovascular disorder with an important inflammatory component. Regardless of its specific classification, a migraine attack is typically triggered by a local reduction of cerebral blood flow, which is then followed by reactive cerebral hyperemia, inflammation, and increased sensitization of pain pathways [1, 4]. According to this model, the role of proinflammatory cytokines has been recently reevaluated and emphasized, since local release of a variety of immune mediators may ultimately contribute to amplify vasodilatation and sensitization of nerve endings, which would ultimately magnify the feeling of pain [5]. Another significant association that has been recently highlighted is that between migraine and metabolism abnormalities. A recent systematic review concluded that the risk of both episodic and chronic migraine is significantly increased in obese subjects, a risk that may be even stronger in the reproductive age [5]. It is also noteworthy that the risk of migraine increased with increasing obesity status, from normal to overweight to obese. The prevalence of insulin resistance (i.e., the leading causal factor of conditions involved in metabolic syndrome), has also found to be particularly frequent in patients with migraine [6]. Among the various metabolic abnormalities, which are typically observed in overweight and obese subjects, serum adiponectin concentrations have been consistently reported to be inversely associated with obesity [7]. Adiponectin is a 247-aminoacid cytokine with anti-inflammatory effects that is actively secreted by the adipose tissue, and is involved in a variety of metabolic processes including glucose homeostasis and fatty acid oxidation [8]. Before secretion, adiponectin is subjected to a number of post-translational modifications, which culminate in secretion of different forms into the circulation that typically include trimers, hexamers, and high molecular weight oligomers, ranging in size from 75 to approximately 500 kDa. A smaller form has also been described, globular adiponectin, which originate by cleavage of fulllength molecule by proteases [8]. Beside the established relationship with obesity and body mass index (BMI), the concentration of various isoforms of adiponectin, especially that of the high-molecular weigh oligomers, was found to be inversely associated with inflammation, cardiovascular disease, metabolic syndrome, diabetes [7], as well as with non-alcoholic fatty liver disease (NAFLD) [9]. An intriguing relationship has also been recently observed between serum adiponectin and neurological disorders, supported by the evidence that this cytokine may modulate cognitive impairment and contribute to impair glucose metabolism and mitochondrial dysfunction, two common hallmarks of Alzheimer’s disease [10]. As such, the intriguing association existing between obesity and migraine, combined with the evidence that the concentration of serum adiponectin is constantly decreased in overweight/obese patients, prompted us to perform a
123
literature search for investigating the epidemiological relationship between adiponectin levels and migraine in published epidemiological studies.
Search criteria We performed an electronic search on Medline, Scopus and Web of Science using the keywords ‘‘migraine’’ and ‘‘adiponectin’’, with no language or date restriction. The references of retrieved articles were also accurately scrutinized for identifying additional publications about this topic. Only cross-sectional or interventional studies using standardized criteria for diagnosing migraine (e.g., those of the IHS) [1] and reporting original data about the epidemiological association between migraine and adiponectin were considered. Inter-study heterogeneity was evaluated by v2 based statistics and I2 test [11].
Overview on epidemiological data The electronic search performed according to the previously defined criteria yielded 29 items after exclusion of duplicates. Accurate reading of title, abstract and full text (when available) allowed to exclude 24 items, which did not reported original data about the epidemiological association between adiponectin and migraine (Fig. 1). Finally, five studies were included in this systematic review: four cross-sectional (totaling 300 patients with migraine and 177 controls) and one interventional (Table 1) [12–16]. The inter-study variability of the four cross-sectional studies was high, and attributable to heterogeneity (I2, 89 %; p \ 0.001). Demographic information was also reported, whenever available. 29 potentially relevant items identified and screened
11: Review articles 3: Editorials or letters 10: No data on adiponectin value
5 items finally selected
1 interventional study
4 cross - sectional studies
Fig. 1 Flow diagram of study selection about the association between adiponectin and migraine
Neurol Sci Table 1 Synthesis of epidemiological studies exploring the association between adiponectin and migraine Author
Study design
Study population
Outcome
References
Peterlin et al.
Crosssectional Crosssectional Crosssectional Crosssectional
Peterlin et al.
Interventional
Concentration of total, HMW and MMW adiponectin higher in chronic daily headache suffers than in episodic migraineurs and controls Concentration of total adiponectin not significantly different between migraineurs and controls Concentration of total adiponectin not significantly different between migraineurs and controls Concentration of total adiponectin not significantly higher in subjects with adverse childhood experiences of migraine Significant reduction in adiponectin in patients who responded to the therapy
[12]
Tietjen et al. Bernecker et al. Tietjen et al.
13 women with episodic migraine, 12 women with chronic daily headache and 12 nonmigraineurs female controls 125 women with migraine and 50 females controls 50 migraineurs and 74 controls 100 women with migraine and 41 female controls 20 women with episodic migraine (11 responders and 9 non responders to sumatriptan/naproxen)
The very first article exploring the association between adiponectin and migraine was published by Peterlin et al. [12]. The study included 13 nonobese women with episodic migraine (mean age 34 ± 9 years; BMI 23.9 ± 5.1 kg/m2), 12 nonobese women with chronic daily headache (mean age 34 ± 9 years; BMI 24.4 ± 3.8 kg/m2) and 12 nonmigraineurs female controls (mean age 33 ± 9 years; BMI 24.8 ± 5.0 kg/m2). After adjustment for waist-to-hip ratio and BMI, the concentration of total serum adiponectin was found to be significantly higher in chronic daily headache suffers (10.1 ± 4.0 lg/mL) than in both episodic migraineurs (8.6 ± 3.5 lg/mL) and controls (7.5 ± 2.4 lg/mL) (p \ 0.05). The concentration of high molecular weight adiponectin was also found to be significantly higher (p \ 0.005) in patients with chronic daily headache (6.1 ± 2.8 lg/mL) than in those with episodic migraine (4.2 ± 1.7 lg/mL) and in control subjects (3.9 ± 1.5 lg/mL). Similar results were found for middle molecular weight adiponectin, the concentration of which was found to be higher (p \ 0.01) in chronic daily headache suffers (2.0 ± 1.2 lg/mL) than in episodic migraine sufferers (1.5 ± 0.7 lg/mL) or control subjects (1.1 ± 0.4 lg/mL). Conversely, no differences were observed in low molecular weight adiponectin levels among the three groups (p = 0.22 for trend). Tietjen et al. performed a cross-sectional study including 125 women with migraine (mean age 37 ± 9 years) and 50 females controls matched for age, race, education and income [13]. The concentration of total serum adiponectin was found to be not significantly different between migraineurs and controls (2.6 ± 1.2 vs. 2.5 ± 2.0 lg/mL; p = 0.86). Bernecker et al. studied 124 subjects, 50 with migraine (42 females and 8 males; mean age 37 ± 12; BMI 22.6 ± 3.1 kg/m2) and 74 without (46 females and 28 males; mean age 36.0 ± 9.0 years; BMI 22.5 ± 3.0 kg/m2) [14], and reported that the concentration of total adiponectin was not significantly different between migraineurs and controls (16.6 ± 7.4 vs. 16.2 ± 10.0 lg/mL; p = ns).
[13] [14] [15]
[16]
In a following investigation, Tietjen et al. measured total serum adiponectin levels in 100 predominantly overweight women with migraine (mean age 37 ± 8 years; BMI 29.9 ± 7.5) and in 41 nonmigraineurs women (mean age 37 ± 10 years; BMI 25.9 ± 5.2 kg/m2) [15]. Interestingly, the concentration of this biomarker was not associated with history of adverse childhood experiences related to migraine after adjustment for age, oral contraceptive use, BMI, hypertension, hyperlipidemia, and smoking (OR, 0.75; 95 % CI, 0.37–1.52; p = 0.27). The evidence that the significant relationship between adiponectin and ACE score was lost after adjustment for BMI is noteworthy, and probably attributable to well established inverse relationship existing between adiponectin and BMI. These results could hence be interpreted to mean that migraine and elevated BMI (and lower adiponectin with it) share a common pathogenesis in childhood adversity. More recently, Peterlin et al. assessed ictal adiponectin concentrations before and after acute abortive treatment with sumatriptan/naproxen sodium vs. placebo in 20 overweight women with episodic migraine (mean age 35 ± 9 years; BMI 26.2 ± 6.4 kg/m2; 11 responders and 9 nonresponders) [16]. In patients who responded to abortive treatment, total adiponectin concentration was marginally, but significantly reduced after 30 min (12.5 ± 3.4 lg/mL; p = 0.03), 60 min (12.3 ± 3.2 lg/mL; p = 0.017), and 120 min (12.6 ± 3.2 lg/mL; p = 0.016) compared with pre-treatment values (13.4 ± 3.8 lg/mL). A significant decrease in high molecular weight to low molecular weight adiponectin ratio was also observed in treatment responders at 60 min (2.3 ± 1.1; p = 0.002) and 120 min (2.7 ± 1.4; p = 0.02) as compared to baseline (3.7 ± 1.9).
Conclusions Several lines of evidence attest that a decreased concentration of adiponectin is inversely correlated with atherosclerosis, obesity and other metabolic diseases, such as
123
Neurol Sci
metabolic syndrome, diabetes, NAFLD and lipodystrophy [8]. The biological mechanisms underlining these intriguing relationships are still unclear, but are probably mediated by the modulatory, beneficial, effects of adiponectin on insulin sensitivity and inflammation. Among the various mechanisms that are involved in the putative biological link between migraine and obesity, a speculative association has also been hypothesized between adiponectin and migraine [17]. At a glance, our analysis of the current scientific literature suggests that the epidemiological link between adiponectin and migraine remains elusive, at the best. Although the overall number of cross-sectional studies that have been published so far is limited and does not allow to draw any definitive conclusions (n = 4), these trials failed to find any significant association (Table 1). Interestingly, Peterlin et al. found that the association between adiponectin and chronic daily headache (but not migraine) was mostly attributable to high-molecular weight oligomers [12], whereas in the three other studies only total adiponectin rather than the different oligomers was assessed [13–15]. The outcome of the single intervention trial is also uncertain. In this trial, including a rather limited sample size (11 responders), the abortive treatment with sumatriptan/naproxen was indeed effective to lower plasma adiponectin concentration by approximately 8 %, a value that was however very close to the intra-assay imprecision of the method used for measuring this cytokine (i.e., 6 %) [17]. The outcome is hence weak, and requires confirmation in other studies with larger populations of migraineurs. Controversial data have also been published about adiponectin metabolism in the central nervous system. Neumeier et al. reported that although adiponectin mRNA was undetectable in the brain, a little amount of this cytokine could still be measured by immuno-blotting in the cerebrospinal fluid, thus suggesting that adiponectin may effectively cross the blood–brain barrier and/or the blood– cerebrospinal fluid barrier [18]. These results were replicated by Ebinuma et al. who developed a specific enzyme linked immunosorbent assay (ELISA) for selective measurement of adiponectin multimers in 19 samples of normal cerebrospinal fluid [19] and reported that the mean total concentration of adiponectin was approximately 1/1,000 lower than that typically detectable in serum (i.e., 7.2 ± 7.2 lg/L). Interestingly, low molecular weight adiponectin represented the large majority of adiponectin immunoreactivity found in cerebrospinal fluids (i.e., mean value 90 %, range 62–100 %). At variance with these results, Spranger et al. showed that neither nonglycosylated nor glycosylated adiponectin can efficiently cross the blood–brain barrier, so that direct effects of this cytokine on central nervous system pathways was deemed to be unlikely [20].
123
These findings, combined with the evidence that the significant association between adiponectin and chronic daily headache found by Peterlin et al. [12] was mostly attributable to high and medium molecular weight oligomers of the protein contribute to raise serious doubts about the existence of any biological and causal relationship between adiponectin and migraine, since the levels of molecular forms of adiponectin were found to be very low in cerebrospinal fluids, up to undetectable [19], and thereby it remains questionable whether they may exert any potential biological effect in this environment.
References 1. Headache Classification Committee of the International Headache Society (IHS) (2013) The international classification of headache disorders, 3rd edition (beta version). Cephalalgia 33:629–808 2. Steiner TJ, Stovner LJ, Birbeck GL (2013) Migraine: the seventh disabler. Headache 53:227–229 3. Leonardi M, Raggi A (2013) Burden of migraine: international perspectives. Neurol Sci 34(Suppl 1):S117–S118 4. Eising E, Datson N, van den Maagdenberg AM, Ferrari MD (2013) Epigenetic mechanisms in migraine: a promising avenue? BMC Med 11:26 5. Chai NC, Scher AI, Moghekar A, Bond DS, Peterlin BL (2014) Obesity and headache: part I––a systematic review of the epidemiology of obesity and headache. Headache 54:219–234 6. Casucci G, Villani V, Cologno D, D’Onofrio F (2012) Migraine and metabolism. Neurol Sci 33(Suppl 1):S81–S85 7. Mather KJ, Goldberg RB (2014) Clinical use of adiponectin as a marker of metabolic dysregulation. Best Pract Res Clin Endocrinol Metab 28:107–117 8. Oh DK, Ciaraldi T, Henry RR (2007) Adiponectin in health and disease. Diabetes Obes Metab 9:282–289 9. Targher G, Bertolini L, Rodella S, Zoppini G, Scala L, Zenari L, Falezza G (2006) Associations between plasma adiponectin concentrations and liver histology in patients with nonalcoholic fatty liver disease. Clin Endocrinol (Oxf) 64:679–683 10. Song J, Lee JE (2013) Adiponectin as a new paradigm for approaching Alzheimer’s disease. Anat Cell Biol 46:229–234 11. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327:557–560 12. Peterlin BL, Alexander G, Tabby D, Reichenberger E (2008) Oligomerization state-dependent elevations of adiponectin in chronic daily headache. Neurology 70:1905–1911 13. Tietjen GE, Khubchandani J, Khan MA, Herial N (2010) Adiponectin and inflammatory cytokines in young women with migraine. Headache 50:71 14. Bernecker C, Pailer S, Kieslinger P et al (2011) Increased matrix metalloproteinase activity is associated with migraine and migraine-related metabolic dysfunctions. Eur J Neurol 18:571–576 15. Tietjen GE, Khubchandani J, Herial NA, Shah K (2012) Adverse childhood experiences are associated with migraine and vascular biomarkers. Headache 52:920–929 16. Peterlin BL, Tietjen GE, Gower BA et al (2013) Ictal adiponectin levels in episodic migraineurs: a randomized pilot trial. Headache 53:474–490 17. Peterlin BL (2009) The role of the adipocytokines adiponectin and leptin in migraine. J Am Osteopath Assoc 109:314–317
Neurol Sci 18. Neumeier M, Weigert J, Buettner R et al (2007) Detection of adiponectin in cerebrospinal fluid in humans. Am J Physiol Endocrinol Metab 293:E965–E969 19. Ebinuma H, Miida T, Yamauchi T et al (2007) Improved ELISA for selective measurement of adiponectin multimers and identi-
fication of adiponectin in human cerebrospinal fluid. Clin Chem 53:1541–1544 20. Spranger J, Verma S, Go¨hring I et al (2006) Adiponectin does not cross the blood–brain barrier but modifies cytokine expression of brain endothelial cells. Diabetes 55:141–147
123
REVIEW ARTICLE
Adiponectin and migraine: systematic review of clinical evidence Giuseppe Lippi • Tiziana Meschi • Camilla Mattiuzzi Loris Borghi • Giovanni Targher
•
Received: 17 February 2014 / Accepted: 10 March 2014 Ó Springer-Verlag Italia 2014
Abstract Although the pathogenesis of migraine is very complex and has not been thoughtfully elucidated, general consensus exists to date that this condition should be considered a primary neurovascular disorder with an important inflammatory component. Owing to epidemiological evidence of increased risk of migraine in overweight and obese subjects and to the inverse relationship that exists between serum adiponectin concentration and obesity, we performed an electronic search on Medline, Scopus and Web of Science, using the keywords ‘‘migraine’’ and ‘‘adiponectin’’ with no language or date restriction to explore the existence of an association between serum adiponectin and migraine. According to our search criteria, five studies were finally included in this systematic review, four cross-sectional (totaling 300 patients with migraine and 177 controls) and
G. Lippi Laboratory of Clinical Chemistry and Hematology, Academic Hospital of Parma, Parma, Italy G. Lippi (&) U.O. Diagnostica Ematochimica, Azienda Ospedaliero-Universitaria di Parma, Via Gramsci, 14, 43126 Parma, Italy e-mail: [email protected]; [email protected] T. Meschi L. Borghi Department of Clinical and Experimental Medicine, Academic Hospital of Parma, Parma, Italy C. Mattiuzzi Service of Clinical Governance, General Hospital of Trento, Trento, Italy G. Targher Section of Endocrinology, Diabetes and Metabolism, Department of Medicine, University and Azienda Ospedaliera Universitaria Integrata of Verona, Verona, Italy
one interventional. Collectively, the results of our analysis suggest that a link between serum adiponectin and migraine remains elusive, at the best. The four cross-sectional studies failed to find any significant association, whereas the outcome of the single interventional study reported a rather modest variation of serum adiponectin concentration in a very limited sample size. Further larger studies are needed to firmly establish the existence of a relationship between adiponectin metabolism and migraine. Keywords Review
Migraine Headache Obesity Adiponectin
Introduction Migraine, which is conventionally defined as recurrent headache manifesting in attacks lasting 4–72 h [1], is a highly prevalent condition. According to recent epidemiological data, the prevalence of migraine approximates 15 % in the general population, which makes it the third most common disease worldwide in both genders, behind dental caries and tension-type headache [2]. It is also noteworthy that approximately 3 % of worldwide disability in terms of years lived with a disability are attributable to migraine [3]. The recent document issued by the Headache Classification Committee of the International Headache Society (IHS) classifies migraine in three major forms that are migraine without aura, migraine with aura and chronic migraine, wherein aura is defined as unilateral fully reversible visual, sensory, or additional central nervous system symptoms that typically onset gradually and are followed by headache and associated migraine symptoms [1]. Although the pathogenesis of migraine is very complex and has not been thoughtfully elucidated, there is general
123
Neurol Sci
consensus to date that migraine should be considered as a primary neurovascular disorder with an important inflammatory component. Regardless of its specific classification, a migraine attack is typically triggered by a local reduction of cerebral blood flow, which is then followed by reactive cerebral hyperemia, inflammation, and increased sensitization of pain pathways [1, 4]. According to this model, the role of proinflammatory cytokines has been recently reevaluated and emphasized, since local release of a variety of immune mediators may ultimately contribute to amplify vasodilatation and sensitization of nerve endings, which would ultimately magnify the feeling of pain [5]. Another significant association that has been recently highlighted is that between migraine and metabolism abnormalities. A recent systematic review concluded that the risk of both episodic and chronic migraine is significantly increased in obese subjects, a risk that may be even stronger in the reproductive age [5]. It is also noteworthy that the risk of migraine increased with increasing obesity status, from normal to overweight to obese. The prevalence of insulin resistance (i.e., the leading causal factor of conditions involved in metabolic syndrome), has also found to be particularly frequent in patients with migraine [6]. Among the various metabolic abnormalities, which are typically observed in overweight and obese subjects, serum adiponectin concentrations have been consistently reported to be inversely associated with obesity [7]. Adiponectin is a 247-aminoacid cytokine with anti-inflammatory effects that is actively secreted by the adipose tissue, and is involved in a variety of metabolic processes including glucose homeostasis and fatty acid oxidation [8]. Before secretion, adiponectin is subjected to a number of post-translational modifications, which culminate in secretion of different forms into the circulation that typically include trimers, hexamers, and high molecular weight oligomers, ranging in size from 75 to approximately 500 kDa. A smaller form has also been described, globular adiponectin, which originate by cleavage of fulllength molecule by proteases [8]. Beside the established relationship with obesity and body mass index (BMI), the concentration of various isoforms of adiponectin, especially that of the high-molecular weigh oligomers, was found to be inversely associated with inflammation, cardiovascular disease, metabolic syndrome, diabetes [7], as well as with non-alcoholic fatty liver disease (NAFLD) [9]. An intriguing relationship has also been recently observed between serum adiponectin and neurological disorders, supported by the evidence that this cytokine may modulate cognitive impairment and contribute to impair glucose metabolism and mitochondrial dysfunction, two common hallmarks of Alzheimer’s disease [10]. As such, the intriguing association existing between obesity and migraine, combined with the evidence that the concentration of serum adiponectin is constantly decreased in overweight/obese patients, prompted us to perform a
123
literature search for investigating the epidemiological relationship between adiponectin levels and migraine in published epidemiological studies.
Search criteria We performed an electronic search on Medline, Scopus and Web of Science using the keywords ‘‘migraine’’ and ‘‘adiponectin’’, with no language or date restriction. The references of retrieved articles were also accurately scrutinized for identifying additional publications about this topic. Only cross-sectional or interventional studies using standardized criteria for diagnosing migraine (e.g., those of the IHS) [1] and reporting original data about the epidemiological association between migraine and adiponectin were considered. Inter-study heterogeneity was evaluated by v2 based statistics and I2 test [11].
Overview on epidemiological data The electronic search performed according to the previously defined criteria yielded 29 items after exclusion of duplicates. Accurate reading of title, abstract and full text (when available) allowed to exclude 24 items, which did not reported original data about the epidemiological association between adiponectin and migraine (Fig. 1). Finally, five studies were included in this systematic review: four cross-sectional (totaling 300 patients with migraine and 177 controls) and one interventional (Table 1) [12–16]. The inter-study variability of the four cross-sectional studies was high, and attributable to heterogeneity (I2, 89 %; p \ 0.001). Demographic information was also reported, whenever available. 29 potentially relevant items identified and screened
11: Review articles 3: Editorials or letters 10: No data on adiponectin value
5 items finally selected
1 interventional study
4 cross - sectional studies
Fig. 1 Flow diagram of study selection about the association between adiponectin and migraine
Neurol Sci Table 1 Synthesis of epidemiological studies exploring the association between adiponectin and migraine Author
Study design
Study population
Outcome
References
Peterlin et al.
Crosssectional Crosssectional Crosssectional Crosssectional
Peterlin et al.
Interventional
Concentration of total, HMW and MMW adiponectin higher in chronic daily headache suffers than in episodic migraineurs and controls Concentration of total adiponectin not significantly different between migraineurs and controls Concentration of total adiponectin not significantly different between migraineurs and controls Concentration of total adiponectin not significantly higher in subjects with adverse childhood experiences of migraine Significant reduction in adiponectin in patients who responded to the therapy
[12]
Tietjen et al. Bernecker et al. Tietjen et al.
13 women with episodic migraine, 12 women with chronic daily headache and 12 nonmigraineurs female controls 125 women with migraine and 50 females controls 50 migraineurs and 74 controls 100 women with migraine and 41 female controls 20 women with episodic migraine (11 responders and 9 non responders to sumatriptan/naproxen)
The very first article exploring the association between adiponectin and migraine was published by Peterlin et al. [12]. The study included 13 nonobese women with episodic migraine (mean age 34 ± 9 years; BMI 23.9 ± 5.1 kg/m2), 12 nonobese women with chronic daily headache (mean age 34 ± 9 years; BMI 24.4 ± 3.8 kg/m2) and 12 nonmigraineurs female controls (mean age 33 ± 9 years; BMI 24.8 ± 5.0 kg/m2). After adjustment for waist-to-hip ratio and BMI, the concentration of total serum adiponectin was found to be significantly higher in chronic daily headache suffers (10.1 ± 4.0 lg/mL) than in both episodic migraineurs (8.6 ± 3.5 lg/mL) and controls (7.5 ± 2.4 lg/mL) (p \ 0.05). The concentration of high molecular weight adiponectin was also found to be significantly higher (p \ 0.005) in patients with chronic daily headache (6.1 ± 2.8 lg/mL) than in those with episodic migraine (4.2 ± 1.7 lg/mL) and in control subjects (3.9 ± 1.5 lg/mL). Similar results were found for middle molecular weight adiponectin, the concentration of which was found to be higher (p \ 0.01) in chronic daily headache suffers (2.0 ± 1.2 lg/mL) than in episodic migraine sufferers (1.5 ± 0.7 lg/mL) or control subjects (1.1 ± 0.4 lg/mL). Conversely, no differences were observed in low molecular weight adiponectin levels among the three groups (p = 0.22 for trend). Tietjen et al. performed a cross-sectional study including 125 women with migraine (mean age 37 ± 9 years) and 50 females controls matched for age, race, education and income [13]. The concentration of total serum adiponectin was found to be not significantly different between migraineurs and controls (2.6 ± 1.2 vs. 2.5 ± 2.0 lg/mL; p = 0.86). Bernecker et al. studied 124 subjects, 50 with migraine (42 females and 8 males; mean age 37 ± 12; BMI 22.6 ± 3.1 kg/m2) and 74 without (46 females and 28 males; mean age 36.0 ± 9.0 years; BMI 22.5 ± 3.0 kg/m2) [14], and reported that the concentration of total adiponectin was not significantly different between migraineurs and controls (16.6 ± 7.4 vs. 16.2 ± 10.0 lg/mL; p = ns).
[13] [14] [15]
[16]
In a following investigation, Tietjen et al. measured total serum adiponectin levels in 100 predominantly overweight women with migraine (mean age 37 ± 8 years; BMI 29.9 ± 7.5) and in 41 nonmigraineurs women (mean age 37 ± 10 years; BMI 25.9 ± 5.2 kg/m2) [15]. Interestingly, the concentration of this biomarker was not associated with history of adverse childhood experiences related to migraine after adjustment for age, oral contraceptive use, BMI, hypertension, hyperlipidemia, and smoking (OR, 0.75; 95 % CI, 0.37–1.52; p = 0.27). The evidence that the significant relationship between adiponectin and ACE score was lost after adjustment for BMI is noteworthy, and probably attributable to well established inverse relationship existing between adiponectin and BMI. These results could hence be interpreted to mean that migraine and elevated BMI (and lower adiponectin with it) share a common pathogenesis in childhood adversity. More recently, Peterlin et al. assessed ictal adiponectin concentrations before and after acute abortive treatment with sumatriptan/naproxen sodium vs. placebo in 20 overweight women with episodic migraine (mean age 35 ± 9 years; BMI 26.2 ± 6.4 kg/m2; 11 responders and 9 nonresponders) [16]. In patients who responded to abortive treatment, total adiponectin concentration was marginally, but significantly reduced after 30 min (12.5 ± 3.4 lg/mL; p = 0.03), 60 min (12.3 ± 3.2 lg/mL; p = 0.017), and 120 min (12.6 ± 3.2 lg/mL; p = 0.016) compared with pre-treatment values (13.4 ± 3.8 lg/mL). A significant decrease in high molecular weight to low molecular weight adiponectin ratio was also observed in treatment responders at 60 min (2.3 ± 1.1; p = 0.002) and 120 min (2.7 ± 1.4; p = 0.02) as compared to baseline (3.7 ± 1.9).
Conclusions Several lines of evidence attest that a decreased concentration of adiponectin is inversely correlated with atherosclerosis, obesity and other metabolic diseases, such as
123
Neurol Sci
metabolic syndrome, diabetes, NAFLD and lipodystrophy [8]. The biological mechanisms underlining these intriguing relationships are still unclear, but are probably mediated by the modulatory, beneficial, effects of adiponectin on insulin sensitivity and inflammation. Among the various mechanisms that are involved in the putative biological link between migraine and obesity, a speculative association has also been hypothesized between adiponectin and migraine [17]. At a glance, our analysis of the current scientific literature suggests that the epidemiological link between adiponectin and migraine remains elusive, at the best. Although the overall number of cross-sectional studies that have been published so far is limited and does not allow to draw any definitive conclusions (n = 4), these trials failed to find any significant association (Table 1). Interestingly, Peterlin et al. found that the association between adiponectin and chronic daily headache (but not migraine) was mostly attributable to high-molecular weight oligomers [12], whereas in the three other studies only total adiponectin rather than the different oligomers was assessed [13–15]. The outcome of the single intervention trial is also uncertain. In this trial, including a rather limited sample size (11 responders), the abortive treatment with sumatriptan/naproxen was indeed effective to lower plasma adiponectin concentration by approximately 8 %, a value that was however very close to the intra-assay imprecision of the method used for measuring this cytokine (i.e., 6 %) [17]. The outcome is hence weak, and requires confirmation in other studies with larger populations of migraineurs. Controversial data have also been published about adiponectin metabolism in the central nervous system. Neumeier et al. reported that although adiponectin mRNA was undetectable in the brain, a little amount of this cytokine could still be measured by immuno-blotting in the cerebrospinal fluid, thus suggesting that adiponectin may effectively cross the blood–brain barrier and/or the blood– cerebrospinal fluid barrier [18]. These results were replicated by Ebinuma et al. who developed a specific enzyme linked immunosorbent assay (ELISA) for selective measurement of adiponectin multimers in 19 samples of normal cerebrospinal fluid [19] and reported that the mean total concentration of adiponectin was approximately 1/1,000 lower than that typically detectable in serum (i.e., 7.2 ± 7.2 lg/L). Interestingly, low molecular weight adiponectin represented the large majority of adiponectin immunoreactivity found in cerebrospinal fluids (i.e., mean value 90 %, range 62–100 %). At variance with these results, Spranger et al. showed that neither nonglycosylated nor glycosylated adiponectin can efficiently cross the blood–brain barrier, so that direct effects of this cytokine on central nervous system pathways was deemed to be unlikely [20].
123
These findings, combined with the evidence that the significant association between adiponectin and chronic daily headache found by Peterlin et al. [12] was mostly attributable to high and medium molecular weight oligomers of the protein contribute to raise serious doubts about the existence of any biological and causal relationship between adiponectin and migraine, since the levels of molecular forms of adiponectin were found to be very low in cerebrospinal fluids, up to undetectable [19], and thereby it remains questionable whether they may exert any potential biological effect in this environment.
References 1. Headache Classification Committee of the International Headache Society (IHS) (2013) The international classification of headache disorders, 3rd edition (beta version). Cephalalgia 33:629–808 2. Steiner TJ, Stovner LJ, Birbeck GL (2013) Migraine: the seventh disabler. Headache 53:227–229 3. Leonardi M, Raggi A (2013) Burden of migraine: international perspectives. Neurol Sci 34(Suppl 1):S117–S118 4. Eising E, Datson N, van den Maagdenberg AM, Ferrari MD (2013) Epigenetic mechanisms in migraine: a promising avenue? BMC Med 11:26 5. Chai NC, Scher AI, Moghekar A, Bond DS, Peterlin BL (2014) Obesity and headache: part I––a systematic review of the epidemiology of obesity and headache. Headache 54:219–234 6. Casucci G, Villani V, Cologno D, D’Onofrio F (2012) Migraine and metabolism. Neurol Sci 33(Suppl 1):S81–S85 7. Mather KJ, Goldberg RB (2014) Clinical use of adiponectin as a marker of metabolic dysregulation. Best Pract Res Clin Endocrinol Metab 28:107–117 8. Oh DK, Ciaraldi T, Henry RR (2007) Adiponectin in health and disease. Diabetes Obes Metab 9:282–289 9. Targher G, Bertolini L, Rodella S, Zoppini G, Scala L, Zenari L, Falezza G (2006) Associations between plasma adiponectin concentrations and liver histology in patients with nonalcoholic fatty liver disease. Clin Endocrinol (Oxf) 64:679–683 10. Song J, Lee JE (2013) Adiponectin as a new paradigm for approaching Alzheimer’s disease. Anat Cell Biol 46:229–234 11. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327:557–560 12. Peterlin BL, Alexander G, Tabby D, Reichenberger E (2008) Oligomerization state-dependent elevations of adiponectin in chronic daily headache. Neurology 70:1905–1911 13. Tietjen GE, Khubchandani J, Khan MA, Herial N (2010) Adiponectin and inflammatory cytokines in young women with migraine. Headache 50:71 14. Bernecker C, Pailer S, Kieslinger P et al (2011) Increased matrix metalloproteinase activity is associated with migraine and migraine-related metabolic dysfunctions. Eur J Neurol 18:571–576 15. Tietjen GE, Khubchandani J, Herial NA, Shah K (2012) Adverse childhood experiences are associated with migraine and vascular biomarkers. Headache 52:920–929 16. Peterlin BL, Tietjen GE, Gower BA et al (2013) Ictal adiponectin levels in episodic migraineurs: a randomized pilot trial. Headache 53:474–490 17. Peterlin BL (2009) The role of the adipocytokines adiponectin and leptin in migraine. J Am Osteopath Assoc 109:314–317
Neurol Sci 18. Neumeier M, Weigert J, Buettner R et al (2007) Detection of adiponectin in cerebrospinal fluid in humans. Am J Physiol Endocrinol Metab 293:E965–E969 19. Ebinuma H, Miida T, Yamauchi T et al (2007) Improved ELISA for selective measurement of adiponectin multimers and identi-
fication of adiponectin in human cerebrospinal fluid. Clin Chem 53:1541–1544 20. Spranger J, Verma S, Go¨hring I et al (2006) Adiponectin does not cross the blood–brain barrier but modifies cytokine expression of brain endothelial cells. Diabetes 55:141–147
123
