Sage Journals: Discover world-class research

Abstract

Objective: Impaired insulin metabolism has been implicated in migraine. However, to date only some putative effects, especially regarding the involvement of adipocytokines and glucagon-like peptides (GLPs), have been described. The aim of the present study was to investigate adipocytokines and GLPs in non-obese female migraineurs.

Methods: Various parameters of the insulin metabolism and body measurements were determined in 84 non-obese female subjects.

Results: We found highly significantly increased insulin levels with an odds ratio of 10.62 for migraine. Leptin and GLP-2 levels were also increased and correlated with insulin. Logistic regression analysis of leptin and GLP-2 revealed odds ratios of 3.79 and 4.26 for migraine, respectively, when comparing the lowest with the highest quartile of the test variable in the complete study cohort.

Discussion: We show that non-obese female migraineurs suffer from hyperinsulinemia, which is associated with elevated leptin and GLP-2 levels. Increased leptin and GLP-2 are risk factors for migraine. Our data suggest that migraine is associated with a higher risk for insulin resistance and its clinical consequences.

Keywords

migraine leptin GLP-2 insulin

Introduction

Migraine has been found to be associated with higher risks for various metabolic disorders, including dyslipidemia and hyperinsulinemia and their clinical outcomes ischemic stroke, cardiovascular disease and diabetes (1,2). An association of migraine with obesity, somewhat controversially, has also been explored. Specifically, in post-menopausal women suffering from episodic migraine or episodic headache, no association was found between obesity and the prevalence of migraine (3–5). Bigal et al. suggested that migraine prevalence is not associated with obesity when body mass index (BMI) is self-reported (6). However, attack frequency, severity and clinical features of migraine increased with BMI. In contrast, three studies have reported an increased prevalence of headache in those with obesity; two of these studies, by Ford and Peterlin, respectively, used self-reported severe headache or migraine but evaluated obesity by measured anthropometric indices to calculate the BMI, while the third study was a large meta-analysis which predominantly utilized both self-reported headache and height and weight (4,7,8). Obesity may be involved in the progression of migraine from an episodic to a chronic form (9–11). However, the underlying mechanisms as well as the impact of aura symptoms on these pathophysiological mechanisms are not clear to date (12–15). Impaired insulin sensitivity is of particular interest, as it has emerged as a risk factor for coronary artery disease, hypertension and stroke. Indeed, there is evidence for an association of migraine and impaired insulin metabolism. Studies by Rainero et al. and Cavestro et al. suggest a pathophysiological role of insulin resistance in migraine, as non-obese migraineurs revealed impaired insulin sensitivity (16,17). These studies, as well as previous studies evaluating adipocytokines in migraineurs, suggest that adipocytokines and insulin signalling may be related to the pathophysiological mechanisms of migraine (10,18–20). Adipocytokines are adipocyte-derived hormones including leptin, adiponectin and resistin, which play an important role in energy homeostasis and have been associated with obesity and diabetes. First, leptin has been shown to counteract obesity by decreasing appetite; however, leptin has also been shown to have vasoactive (12,13,15) and inflammatory properties (21–24). Second, adiponectin may function as an insulin sensitizer and has protective effects against insulin resistance as well as anti-inflammatory effects (14). However, adiponectin also exerts pro-inflammatory properties, depending on the form or multimer involved: low molecular weight (LMW) adiponectin may haven an anti-inflammatory effect, and high molecular weight (HMW) adiponectin has been shown to induce cytokine secretion (25,26). In one study Peterlin et al. found elevated HMW adiponectin in patients with chronic daily headache, as compared to controls, but HMW adiponectin was not significantly elevated in episodic migraineurs interictally (10). Finally, resistin was originally shown to be able to promote insulin resistance (27) and has now been implicated in the regulation of inflammatory processes (28–30). The effects of resistin on obesity have been studied with inconsisten results; increased, decreased and unchanged resistin levels in obesity have all been described (22–24). Beyond adipocytokines, the insulin signalling related proteins, GLP-1 and GLP 2 are also involved in insulin signaling and glucose metabolism (31–36). GLPs are produced and secreted from intestinal cells. GLP-1 stimulates insulin secretion and inhibits glucagon secretion. This leads to inhibition of hepatic glucose production and lowers blood glucose levels (37). GLP-2 may counteract the glucagonostatic effects of GLP-1 by stimulating glucagon secretion (38). Taken together, adipocytokines and GLPs may play a pathophysiological role in migraine-related impaired insulin and glucose metabolism. However, only some putative effects of adipocytokines have been described to date. Most of the studies suggest that adiponectin levels are elevated and may protect against migraine attacks (10,14). Only one study has evaluated adiponectin in episodic migraineurs to date and found no differences in total adiponectin or its multimers interictally; however, in those with chronic daily headache, total adiponectin was elevated, predominantly due to elevations in the HMW multimer (10). A recent study by Guldiken et al. suggests that low fat mass and leptin levels may be related to the pathogenesis of migraine (19). To our knowledge, the role of resistin, GLP-1 and GLP-2 in the pathophysiology of migraine has not been investigated to date.

We here hypothesize that adipocytokines and GLPs are involved in migraine and migraine-associated hyperinsulinemia. Therefore, the aim of the present study was to investigate the adipocytokines adiponectin, resistin, leptin and leptin-receptor, as well as GLP-1 and GLP-2 and their associations with insulin and glucose metabolism in non-obese female migraineurs.

Patients and methods

Patients

A total number of 84 female subjects participated in the study (Table 1). Patients were clinically assessed by a neurologist at the Department of Neurology, Medical University Graz, Austria, according to the International Headache Society (IHS) criteria (39). Forty female migraine patients, 13 with symptoms of aura and 27 without symptoms of aura were included. The mean annual frequency of migraine attacks was 30.68 ± 29.84 (range 12.17–91.25). The median annual frequency of migraine attacks was 24.33. Forty-four female controls, free of headache were recruited from among patients’ friends and volunteers. All study participants were Caucasian. Exclusion criteria were obesity (body mass index [BMI] >29.9), diabetes, thyroidal dysfunction as well as metabolic and cardiovascular diseases, chronic illnesses, infectious diseases, pregnancy and nursing. Acute or systemic low-grade inflammation, indicated by high-sensitivity c-reactive protein (CRP) levels (>3 mg/l) was an exclusion criteria. Patients with depression and chronic daily headache were also excluded from the study. The study was approved by the ethics committee of the Medical University of Graz, Austria.

Table 1.

Baseline characteristics of female migraine patients and female healthy controls

Characteristics	Controls (N = 44)	Migraine (N = 40)
Age (years)	35.78 ± 8.73	36.69 ± 11.21
Mean annual frequency of attacks (range)	NA	30.68 ± 29.84 (12.17–91.25)
Median annual frequency of migraine attacks	NA	24.33
Median time from last attack to blood collection, in days (range)	NA	8 (7–62)
Familiar predisposition, % (N)	11	21
Symptoms of aura(N)		13
BMI (kg/m²)	21.27 ± 2.56	22.27 ± 3.24
Waist circumference (cm)	73 ± 6	75 ± 9
Waist-to-hip ratio	0.76 ± 0.05	0.78 ± 0.05
Waist-to-height ratio	0.43 ± 0.04	0.45 ± 0.05
Systolic BP (mmHg)	115 ± 12	115 ± 14
Diastolic BP (mmHg)	73 ± 9	74 ± 10
Insulin (pmol/l)	29.14 ± 27.27	60.65 ± 47.12***
Glucose (mmol/l)	5.05 ± 0.58	4.79 ± 0.52*
HOMA ndex	0.92 ± 0.91	1.87 ± 1.61**
Total adiponectin (µg/ml)	18.77 ± 11.27	17.01 ± 7.54
Leptin (ng/ml)	9.99 ± 5.62	15.07 ± 9.63**
Leptin receptor (ng/ml)	24.50 ± 8.74	22.12 ± 7.53
Resistin (ng/ml)	5.73 ± 1.59	5.86 ± 2.26
Glucagon (pg/ml)	692 ± 392	708 ± 548
GLP-1 (ng/ml)	6.01 ± 2.08	5.82 ± 2.07
GLP-2 (ng/ml)	9.18 ± 2.19	10.72 ± 3.06**

NA = not applicable; BMI = body mass index; BP = blood pressure; HOMA = homeostatic model assessment; GLP = glucagon-like peptide.

*p < .05 compared to controls. **p < .01 compared to controls. ***p < .001 compared to controls.

Blood collection

Blood collection was performed by venous puncture after an overnight fast, after written informed consent was given. At the time of blood draw, probands were free of common infectious diseases. None of the study participants were on medication. The median time period between patients’ last migraine attack and blood collection was eight days (range 7–62 days). Blood was immediately centrifuged and serum was stored at −80°C until analysis. Additionally, body measurements (including body weight, length, waist and hip circumference) and blood pressure were taken.

Laboratory procedures

Fasting glucose and insulin were determined routinely on Roche analysers (Roche Diagnostics, Mannheim, Germany). The homeostasis model assessment (HOMA) index was calculated as follows: fasting glucose (mmol/l) × fasting insulin (UE/ml)/22.5. High-sensitivity CRP levels, estradiol (Estradiol Maia, Adaltis, Germany; sensitivity: 5 pg/ml) and progesterone (RIA Progesteron, Immunotec, Canada; sensitivity: 0.03 ng/ml) were determined routinely. Serum adipocytokines, including total adiponectin (effective range 1.5–45 µg/ml), leptin (effective range 0.5–50 ng/ml), leptin receptor (effective range 1.2–100 ng/ml) and resistin (effective range 0.1–50 ng/ml), as well as glucagon (effective range 50–10000 pg/ml), GLP-1 (effective range 0.206–50 ng/ml) and GLP-2 (effective range 0.412–100 ng/ml) were analysed by commercial enzyme-linked immosorbent assay (ELISA) as recommended by the manufacturer (BioVendor, Czech Republic). All biomarker levels were above the detection limit of the appropriate assay.

Statistical analysis

Data are presented as means ± standard deviation. Continuous variables were compared using Student’s t-test for independent samples or Mann-Whitney U-test, depending on the distribution of data. Correlations between variables were determined by linear regression analysis according to Pearson. p values >.05 were considered statistically significant. Subsequent multiple testing by linear stepwise regression analysis was used to determine the significance of test variables. The associations between test variables and migraine were analysed using logistic regression to generate odds ratios (ORs) and 95% confidence interval (CI). Test variables were each initially examined in quartiles, with the lowest quartile defined as the reference group. The multivariable model included age, BMI, waist circumference and waist-to-height-ratio as possible a priori confounders. Analyses were performed using SPSS 16.0 for Windows (Chicago, IL, USA).

Results

Baseline characteristics of the study cohort are given in Table 1. Migraine patients and healthy controls were balanced according to BMI (mean BMI 21.67 ± 2.86; median BMI 21.39, BMI range 18.5–29.9) and age (mean age 35.92 ± 9.57; median age 35.29; age range 18–64 years). Migraineurs and controls further showed comparable waist circumference, waist-to-hip-ratio, waist-to-height-ratio and blood pressure. Analyses of insulin and glucose metabolism revealed differences between female migraineurs and controls (Table 1). Insulin was highly significantly increased (p < .001) and fasting glucose levels were significantly decreased (p < .05) in migraineurs compared to healthy controls. The HOMA index was significantly higher in migraine patients compared to controls (p < .01). Investigations of adipocytokines revealed that leptin levels were significantly increased in migraine patients compared to healthy controls (p < .01). GLP-2 levels were also significantly increased in migraineurs compared to healthy controls (p < .01). Leptin receptor, total adiponectin, resistin, glucagon and GLP-1 showed no significant differences between migraine patients and controls. Adjustment of total adiponectin levels for anthropometric indices like BMI, waist circumference, waist-to-hip ratio, waist-to-height ratio and age revealed no significant differences (Table 2).

Table 2.

Total adiponectin levels adjusted for anthropometric indices

Normalisation parameter	Controls (N = 44)	Migraineurs (N = 40)	p value
Body mass indes	0.880 ± 0.558	0.787 ± 0.376	.377
Waist circumference	0.253 ± 0.159	0.229 ± 0.108	.414
Waist-to-hip ratio	24.34 ± 15.19	21.92 ± 10.09	.395
Waist-to-length ratio	42.77 ± 26.91	38.10 ± 17.79	.354
Age	0.534 ± 0.353	0.503 ± 0.275	.655

We next examined effects of aura on increased insulin, leptin and GLP-2 levels. Insulin levels were significantly higher in patients with migraine with aura (MA) (84.29 ± 60.04 pmol/l) than in those with migraine with aura (MWA) (49.28 ± 35.44 pmol/l). Leptin and GLP-2 showed no differences between patients with MA and those with MWA.

To determine possible correlations, we performed linear regression analyses (Table 3). We found significant correlations of insulin with age (r = 0.252, p = 0.02), BMI (r = 0.518, p < .0001), waist circumference (r = 0.440, p < .0001), waist-to-hip-ratio (r = 0.327, p = .002), waist-to-height-ratio (r = 0.484, p < .0001), systolic blood pressure (r = 0.367, p = .001), diastolic blood pressure (r = 0.257, p = 0.017), resistin (r = 0.213, p = 0.05), leptin (r = 0.484, p < .0001) and GLP-2 (r = 0.219, p = .044). Furthermore, a significant negative correlation was found between insulin and leptin receptor (r = −0.264, p = .015). Linear regression analyses of insulin and the annual frequency of migraine attacks and the time period from patients’ last migraine attack to blood collection showed no significant correlations (Table 3). Multiple testing by subsequent multiple stepwise regression analyses of insulin, adipocytokines and body measurement parameters revealed the strongest correlation for insulin and leptin (beta = 0.618, p < .0001).

Table 3.

Linear regression analyses

	Insulin		GLP-2		Leptin		Total adiponectin
	r¹	p ²	r	p	r	p	r	p
Age (years)	0.252	0.020	0.025	0.817	0.259	0.017	−0.124	0.261
BMI (kg/m²)	0.518	<0.0001	0.135	0.217	0.644	<0.0001	−0.156	0.157
Waist circumference (cm)	0.440	<0.0001	0.217	0.046	0.623	<0.0001	−0.126	0.255
Waist-to-hip ratio	0.327	0.002	0.156	0.153	0.442	<0.0001	−0.131	0.237
Waist-to-height ratio	0.484	<0.0001	0.202	0.064	0.613	<0.0001	−0.143	0.193
Systolic BP (mmHg)	0.367	0.001	0.154	0.159	0.296	0.006	0.004	0.971
Diastolic BP (mmHg)	0.257	0.017	0.173	0.113	0.235	0.030	0.130	0.237
Estradiol (pg/ml)	0.095	0.391	−0.186	0.089	0.021	0.848	−0.085	0.442
Progesterone (ng/ml)	0.180	0.102	0.032	0.770	0.068	0.537	−0.077	0.486
Total adiponectin (µg/ml)	0.026	0.815	−0.034	0.756	−0.095	0.389
Leptin (ng/ml)	0.484	<0.0001	0.218	0.045			−0.095	0.389
Leptin-receptor (ng/ml)	−0.264	0.015	−0.020	0.855	−0.506	<0.0001	0.246	0.024
Resistin (ng/ml)	0.213	0.050	0.006	0.953	0.092	0.400	−0.085	0.444
Glucagon (pg/ml)	−0.135	0.220	0.544	<0.0001	−0.153	0.161	−0.128	0.247
Glucose (mmol/l)	−0.049	0.650	−0.179	0.102	0.024	0.830	−0.054	0.626
GLP-1 (ng/ml)	−0.036	0.740	0.519	<0.0001	0.016	0.884	−0.075	0.496
GLP-2 (ng/ml)	0.219	0.044			0.218	0.045	−0.034	0.756
Annual frequency of attacks	−0.080	0.610	0.062	0.701	−0.031	0.848	−0.166	0.301
Time from last attack (days)	0.077	0.630	0.117	0.474	0.091	0.576	0.129	0.427

GLP-2 = glucagon-like peptide-2; BMI = body mass index; BP = blood pressure.

¹Pearson correlation coefficient. ²Univariate ANOVA (analysis of variance).

We found significant correlations of GLP-2 with waist circumference (r = 0.217, p = .046), leptin (r = 0.218, p = .045), glucagon (r = 0.544, p < .0001) and GLP-1 (r = 0.519, p < .0001). Multiple testing by subsequent multiple stepwise regression analyses of GLP-2 revealed the strongest correlations for GLP-2 with glucagon (beta = 0.592, p < .0001) and leptin (beta = 0.309, p = .001).

Linear regression analyses with leptin revealed significant correlations with age (r = 0.259, p = .017), BMI (r = 0.644, p < .0001), waist circumference (r = 0.623, p < .0001), waist-to-hip ratio (r = 0.442, p < .0001), waist-to-height ratio (r = 0.613, p < .0001), systolic blood pressure (r = 0.296, p = .006), diastolic blood pressure (r = 0.235, p = .030), leptin receptor (r = −0.506, p < .0001) and GLP-2 (r = 0.218, p = .045).

Multiple testing by subsequent multiple stepwise regression analyses of leptin revealed the strongest correlation for leptin with leptin receptor (beta = −0.367, p < .0001), waist circumference (beta = 0.473, p < .0001) and systolic blood pressure (beta = 0.202, p = .011). A significant correlation of total adiponectin was found with leptin receptor (r = 0.246, p = .024).

Furthermore, to test for possible correlations with female sex hormones, we performed linear regression analyses of insulin, GLP-2, leptin and total adiponectin with estradiol and progesterone (Table 3). We found no correlations between insulin and estradiol (r = 0.095, p = .391) or progesterone (r = 0.180, p = .102). Linear regression analyses of GLP-2 with estradiol (r = −0.186, p = .089) and progesterone (r = 0.032, p = .770) showed no statistical significance. Leptin showed also no statistical significant correlations with estradiol (r = 0.021, p = .848) or progesterone (r = 0.068, p = .537) levels. Finally, total adiponectin showed also no correlations with estradiol (r = −0.085, p = .442) or progesterone (r = −0.077, p = .486).

To determine the impact of increased insulin, leptin and GLP-2 on migraine, logistic regression analyses were performed (Table 3). The logistic regression model conditioned on insulin revealed that women in the highest quartile compared with the lowest quartile of insulin had an OR for migraine of 10.62 (95% CI: 2.50–44.98; p = .001). Multiple modelling to adjust for age and BMI as possible confounders revealed an OR for migraine of 9.84 (95% CI: 2.0–48.47; p = .005). Multiple modelling adjusted for age and waist circumference, as well as for age and waist-to-height-ratio, as possible confounders revealed an OR for migraine of 10.71 (95% CI: 2.22–51.56; p = .003) and 9.56 (95% CI: 2.03–44.99; p = .004), respectively (Table 4). The logistic regression model for leptin revealed an OR for migraine of 3.79 (95% CI: 1.03–13.91; p = .045) when comparing the lowest with the highest quartile of leptin. Multiple modelling to adjust for age and BMI, age and waist circumference, as well as for age and waist-to-height-ratio, revealed no significant models (Table 5). The logistic regression model conditioned on GLP-2 revealed that women in the fourth quartile compared with those in the first quartile had an OR for migraine of 4.26 (95% CI: 1.13–16.05; p = .032) and that women in the third quartile had an OR for migraine of 5.20 (95% CI: 1.36–19.77; p = .016). Multiple modelling to adjust for age and BMI, age and waist circumference, as well as for age and waist-to-height-ratio, revealed no significant models (Table 6).

Table 4.

Logistic regression of insulin quartiles and odds of migraine

	Quartile 1	Quartile 2	Quartile 3	Quartile 4
Insulin (mean ± SD)	4.78 ± 4.30	21.99 ± 5.13	47.74 ± 8.88	97.92 ± 37.96
Controls (N)	15	14	11	4
Migraineurs (N)	6	7	10	17
OR (95% CI) model 1	1.0 (ref)	1.25 (0.33–4.63)	2.27 (0.63–8.14)	10.62 (2.50–44.98)¹
OR (95% CI) model 2	1.0 (ref)	1.20 (0.31–4.64)	2.2 (0.59–8.14)	9.84 (2.0–48.47)²
OR (95% CI) model 3	1.0 (ref)	1.25 (0.33–4.70)	2.27 (0.62–8.28)	10.71 (2.22–51.56)³
OR (95% CI) model 4	1.0 (ref)	1.20 (0.32–4.55)	2.20 (0.60–7.99)	9.56 (2.03–44.99)⁴

SD = standard deviation; OR = odds ratio; CI = confidence interval. Model 1 is conditioned on insulin quartiles and unadjusted. Model 2 is conditioned on insulin quartiles and adjusted for age and BMI. Model 3 is conditioned on insulin quartiles and adjusted for age and waist circumference. Model 4 is conditioned on insulin quartiles and adjusted for age and waist-to-height-ratio. ¹ p = .001. ² p = .005. ³ p = .003. ⁴ p = .004.

Table 5.

Logistic regression of leptin quartiles and odds of migraine

	Quartile 1	Quartile 2	Quartile 3	Quartile 4
Leptin (mean ± SD)	4.77 ± 1.88	8.47 ± 1.06	13.26 ± 1.69	23.80 ± 7.85
Controls (N)	13	12	12	6
Migraineurs (N)	8	9	9	15
OR (95% CI) model 1	1.0 (ref)	1.21 (0.34–4.18)	1.22 (0.35–4.18)	3.79 (1.03–13.91)¹
OR (95% CI) model 2	1.0 (ref)	1.10 (0.30–4.00)	1.029 (0.27–3.90)	2.29 (0.51–10.10)²
OR (95% CI) model 3	1.0 (ref)	1.19 (0.33–4.26)	1.13 (0.30–4.30)	2.66 (0.57–12.32)²
OR (95% CI) model 4	1.0 (ref)	1.18 (0.33–4.16)	1.08 (0.29–4.01)	2.31 (0.50–10.71)²

SD = standard deviation; OR = odds ratio; CI = confidence interval. Model 1 is conditioned on leptin quartiles and unadjusted. Model 2 is conditioned on leptin quartiles and adjusted for age and BMI. Model 3 is conditioned on leptin quartiles and adjusted for age and waist circumference. Model 4 is conditioned on leptin quartiles and adjusted for age and waist-to-height-ratio. ¹ p = .045. ²Model not significant.

Table 6.

Logistic regression of GLP-2 quartiles and odds of migraine

	Quartile 1	Quartile 2	Quartile 3	Quartile 4
GLP-2 (mean ± SD)	7.16 ± 0.86	8.58 ± 0.41	10.20 ± 0.59	13.61 ± 2.39
Controls (N)	16	11	8	9
Migraineurs (N)	5	10	13	12
OR (95% CI) model 1	1.0 (ref)	2.90 (0.77–10.88)	5.20 (1.36–19.77)¹	4.26 (1.13–16.05)²
OR (95% CI) model 2	1.0 (ref)	2.20 (0.54–8.96)	4.59 (1.18–17.79)	3.74 (0.94–14.85)³
OR (95% CI) model 3	1.0 (ref)	2.44 (0.61–9.72)	4.77 (1.24–18.38)	3.75 (0.92–15.21)³
OR (95% CI) model 4	1.0 (ref)	2.25 (0.55–9.13)	4.55 (1.17–17.67)	3.55 (0.87–14.42)³

GLP = glucagon-like peptide; SD = standard deviation; OR = odds ratio; CI = confidence interval. Model 1 is conditioned on GLP-2 quartiles and unadjusted. Model 2 is conditioned on GLP-2 quartiles and adjusted for age and BMI. Model 3 is conditioned on GLP-2 quartiles and adjusted for age and waist circumference. Model 4 is conditioned on GLP-2 quartiles and adjusted for age and waist-to-height-ratio. ¹ p = .016. ² p = .032. ³Models not significant.

Discussion

We investigated the insulin and glucose metabolism and possible associations with adipocytokines and GLPs in non-obese female migraineurs in the headache-free period.

First, we demonstrate that non-obese female migraineurs show significantly increased insulin levels compared to healthy, BMI-matched controls. Sub-dividing of migraineurs into patients with MA and patients with MWA revealed that the increase in insulin was significant in both sub-groups compared to controls; however, patients with MA had significantly higher insulin levels than patients with MWA. Calculated HOMA indices as a combined parameter of insulin and glucose levels were significantly increased in migraine patients compared to healthy controls. The mean HOMA index of 1.87 ± 1.61 of migraineurs indicates a mild insulin-resistant status. To investigate the impact of elevated insulin levels on migraine, we performed logistic regression analyses, revealing an OR of 10.62 for migraine when comparing the lowest with the highest quartile of insulin. Summarising, we show here that non-obese female migraineurs suffer from hyperinsulinemia. Furthermore, we found that hyperinsulinemia is a risk factor for migraine. Our results are in line with two recent studies delineating impaired insulin sensitivity in migraineurs (16,17). Our findings are further supported by the hypothesis of Bic et al. that insulin resistance may play an important role in the pathophysiology of migraine (40).

To determine the underlying pathophysiological mechanisms, we next investigated possible interactions of increased insulin with adipocytokines and GLPs. We found significantly increased leptin and GLP-2 levels in non-obese female migraineurs compared to healthy female controls. Sub-group analyses between patients with MA and MWA showed no differences in leptin and GLP-2 levels. The elevated leptin and GLP-2 levels did show a significant correlation with increased insulin. As adipocytokines may be influenced by female sex hormones (10,41,42), we determined estradiol and progesterone levels. Regression analyses revealed that insulin and adipocytokines showed no correlation with estradiol or progesterone.

We have to note that our study cohort included pre- and postmenopausal women between 18 and 64 years. Taken together, our study does not support an association of increased insulin and adipocytokines with female sex hormones in migraineurs. However, our study design may not be suitable to examine this issue in detail. A population-based epidemiological study would be more suitable.

Next, to determine the impact of adipocytokines on migraine, logistic regression was performed, revealing that increased leptin is a risk factor for migraine, with an OR of 3.79 comparing the highest with the lowest quartile of leptin. Logistic regression of GLP-2 revealed an OR of 4.26 for migraine. Analyses of total adiponectin and resistin showed no differences between non-obese migraineurs and healthy controls. Taken together, these results show that hyperinsulinemia is correlated with increased leptin and GLP-2, which are both associated with migraine in non-obese females.

Limited data exist on the association of adipocytokines and migraine (18–20). The adipocyte-derived hormone leptin regulates energy homeostasis as circulating leptin levels act on higher centres in the brain, thereby decreasing appetite (43–46). Dysfunction in leptin regulation due to leptin deficiency or resistance has been associated with obesity and diabetes (47). Furthermore, leptin receptors are expressed on endothelial cells, thereby activating the sympathetic nervous system, which exerts vascular effects (12,13,15). A study by Guldiken et al. investigated fat mass and leptin levels in migraine patients, both male and female (19). In migraine patients with a mean BMI of about 26 ± 5, fat mass levels were found to be lower and accompanied by reduced leptin levels. One may speculate that the different study designs may at least in part cause the differences between Guldiken et al.’s study and ours. We investigated non-obese females with a mean BMI of about 22 ± 3, to avoid obesity-related effects on insulin metabolism. Leptin is secreted by adipose tissue and the amount of plasma leptin levels is proportional to the amount of energy stored as fat, so that obese subjects have higher leptin levels than lean subjects (43).

The role of resistin and adiponectin in the pathophysiology of migraine remains unclear to date (20). Resistin has been shown to be involved in the progression of insulin resistance (48). Adiponectin is negatively correlated with obesity and insulin resistance (49,50). It has been hypothesized that baseline high levels of adiponectin, in particular globular adiponectin, may be protective of migraine (14). HMW adiponectin has been shown to induce cytokine secretion and thereby exert pro-inflammatory effects (25,26). However, we did not find any differences in total adiponectin levels between migraine patients and healthy controls. This finding is similar to that of Peterlin et al., who reported no difference in total adiponectin or in the multimers of adiponectin interictally in episodic migraineurs (10). Additionally, our current study is the first to evaluate the adipocytokine resistin, and found no difference in resistin levels in migraineurs as compared to controls.

As described above, in addition to adipocytokines we also investigated the insulinotropic hormones GLP-1 and GLP-2, and found an association between GLP-2, hyperinsulinemia and migraine. The hormones GLP-1 and GLP-2 are secreted from gut endocrine cells and promote nutrient absorption and disposal. GLP-1 improves glucose-stimulated insulin secretion and restores glucose competence in glucose-resistant beta cells (32). GLP-1 analogues (incretin mimetics) are therefore used as a new class of anti-diabetic medication (34). GLP-2 may counteract the glucagonostatic effects of GLP-1 by stimulating glucagon secretion (38). GLP-2 further regulates intestinal hexose transport, increases the barrier function of the gut epithelium and enhances the surface area of the mucosal epithelium (33). GLP-2 also exerts cytoprotective and reparative effects regarding intestinal injury. It is also involved in cytokine expression and promotes anti-apoptotic effects (51). Furthermore, GLP-1 and GLP-2 have been shown to promote cell proliferation, which contributes to the protective and regenerative actions of these peptide hormones in subjects with diabetes and intestinal disorders (31).

In summary, we show that non-obese female migraineurs suffer from hyperinsulinemia, independent of aura symptoms. Hyperinsulinemia is associated with elevated leptin and GLP-2 levels, which are risk factors for migraine. These findings represent a new pathophysiological mechanism which may be of clinical relevance.

References

Bigal

Kurth

Santanello

Lipton

. Migraine and cardiovascular disease: possible mechanisms of interaction. Neurology 2009; 72: 1864–1871.

Guldiken

Taskiran

. Migraine in metabolic syndrome. Neurologist 2009; 15: 55–58.

Mattsson

. Migraine headache and obesity in women aged 40–74 years: a population-based study. Cephalalgia 2007; 27: 877–880.

Peterlin

Rosso

Rapoport

Scher

. Obesity and migraine: the effect of age, gender and adipose tissue distribution. Headache 2010; 00: 1–1. accessed 18 February 2010.

Winter

Berger

Buring

Kurth

. Body mass index, migraine, migraine frequency and migraine features in women. Cephalalgia 2009; 29: 269–278.

Bigal

Liberman

Lipton

. Obesity and migraine: a population study. Neurology 2006; 66: 545–550.

Ford

Pearson

Zhao

Strine

Mokdad

. Body mass index and headaches: findings from a national sample of US adults. Cephalalgia 2008; 28: 1270–1276.

Keith

Wang

Fontaine

Cowan

Allison

. BMI and headache among women: results from 11 epidemiologic datasets. Obesity (Silver Spring) 2008; 16: 377–383.

Bigal

Lipton

. Obesity is a risk factor for transformed migraine but not chronic tension-type headache. Neurology 2006; 67: 252–257.

10.

Peterlin

Alexander

Tabby

Reichenberger

. Oligomerization state-dependent elevations of adiponectin in chronic daily headache. Neurology 2008; 70: 1905–1911.

11.

Scher

Stewart

Ricci

Lipton

. Factors associated with the onset and remission of chronic daily headache in a population-based study. Pain 2003; 106: 81–89.

12.

Dunbar

. Intracerebroventricular leptin increases lumbar and renal sympathetic nerve activity and blood pressure in normal rats. Diabetes 1997; 46: 2040–2043.

13.

Haynes

Morgan

Walsh

Mark

Sivitz

. Receptor-mediated regional sympathetic nerve activation by leptin. J Clin Invest 1997; 100: 270–278.

14.

Peterlin

Bigal

Tepper

Urakaze

Sheftell

Rapoport

. Migraine and adiponectin: is there a connection? Cephalalgia 2007; 27: 435–446.

15.

Shek

Brands

Hall

. Chronic leptin infusion increases arterial pressure. Hypertension 1998; 31: 409–414.

16.

Cavestro

Rosatello

Micca

. Insulin metabolism is altered in migraineurs: a new pathogenic mechanism for migraine? Headache 2007; 47: 1436–1442.

17.

Rainero

Limone

Ferrero

. Insulin sensitivity is impaired in patients with migraine. Cephalalgia 2005; 25: 593–597.

18.

Berilgen

Bulut

Gonen

Tekatas

Dag

Mungen

. Comparison of the effects of amitriptyline and flunarizine on weight gain and serum leptin, C peptide and insulin levels when used as migraine preventive treatment. Cephalalgia 2005; 25: 1048–1053.

19.

Guldiken

Demir

Turgut

Tugrul

. Low leptin levels in migraine: a case control study. Headache 2008; 48: 1103–1107.

20.

Peterlin

. The role of the adipocytokines adiponectin and leptin in migraine. J Am Osteopath Assoc 2009; 109: 314–317.

21.

Festa

Hanley

Tracy

D’Agostino

R Haffner

. Inflammation in the prediabetic state is related to increased insulin resistance rather than decreased insulin secretion. Circulation 2003; 108: 1822–1830.

22.

Lee

Chan

Yiannakouris

. Circulating resistin levels are not associated with obesity or insulin resistance in humans and are not regulated by fasting or leptin administration: cross-sectional and interventional studies in normal, insulin-resistant, and diabetic subjects. J Clin Endocrinol Metab 2003; 88: 4848–4856.

23.

Rajala

Lin

Ranalletta

. Cell type-specific expression and coregulation of murine resistin and resistin-like molecule-alpha in adipose tissue. Mol Endocrinol 2002; 16: 1920–1930.

24.

Silha

Krsek

Skrha

Sucharda

Nyomba

Murphy

. Plasma resistin, adiponectin and leptin levels in lean and obese subjects: correlations with insulin resistance. Eur J Endocrinol 2003; 149: 331–335.

25.

Haugen

Drevon

. Activation of nuclear factor-kappaB by high molecular weight and globular adiponectin. Endocrinology 2007; 148: 5478–5486.

26.

Tsatsanis

Zacharioudaki

Androulidaki

. Adiponectin induces TNF-alpha and IL-6 in macrophages and promotes tolerance to itself and other pro-inflammatory stimuli. Biochem Biophys Res Commun 2005; 335: 1254–1263.

27.

Sheng

Jin

. Resistin is expressed in human hepatocytes and induces insulin resistance. Endocrine 2008; 33: 135–143.

28.

Fantuzzi

. Adipose tissue, adipokines, and inflammation. J Allergy Clin Immunol 2005; 115: 911–919.

29.

Qatanani

Szwergold

Greaves

Ahima

Lazar

. Macrophage-derived human resistin exacerbates adipose tissue inflammation and insulin resistance in mice. J Clin Invest 2010; 00: 1–1. accessed 18 February 2010.

30.

Reilly

Lehrke

Wolfe

Rohatgi

Lazar

Rader

. Resistin is an inflammatory marker of atherosclerosis in humans. Circulation 2005; 111: 932–939.

31.

Brubaker

Drucker

. Minireview: Glucagon-like peptides regulate cell proliferation and apoptosis in the pancreas, gut, and central nervous system. Endocrinology 2004; 145: 2653–2659.

32.

Buteau

. GLP-1 receptor signaling: effects on pancreatic beta-cell proliferation and survival. Diabetes Metab 2008; 34(Suppl 2): 73–77.

33.

Drucker

. Glucagon-like peptide 2. J Clin Endocrinol Metab 2001; 86: 1759–1764.

34.

Frias

Edelman

. Incretins and their role in the management of diabetes. Curr Opin Endocrinol Diabetes Obes 2007; 14: 269–276.

35.

Knauf

Cani

Ait-Belgnaoui

. Brain glucagon-like peptide 1 signaling controls the onset of high-fat diet-induced insulin resistance and reduces energy expenditure. Endocrinology 2008; 149: 4768–4777.

36.

Lim

Huang

Flora

LeRoith

Rhodes

Brubaker

. Insulin regulates glucagon-like peptide-1 secretion from the enteroendocrine L cell. Endocrinology 2009; 150: 580–591.

37.

Holst

. Glucagonlike peptide 1: a newly discovered gastrointestinal hormone. Gastroenterology 1994; 107: 1848–1855.

38.

Meier

Nauck

Pott

. Glucagon-like peptide 2 stimulates glucagon secretion, enhances lipid absorption, and inhibits gastric acid secretion in humans. Gastroenterology 2006; 130: 44–54.

39.

International Headache Society. The International Classification of Headache Disorders. 2nd ed. Cephalalgia 2004; 24: 1–160.

40.

Bic

Blix

Hopp

Leslie

. In search of the ideal treatment for migraine headache. Med Hypotheses 1998; 50: 1–7.

41.

Kalish

Barrett-Connor

Laughlin

Gulanski

. Association of endogenous sex hormones and insulin resistance among postmenopausal women: results from the Postmenopausal Estrogen/Progestin Intervention Trial. J Clin Endocrinol Metab 2003; 88: 1646–1652.

42.

Lin

Sagawa

Yura

Itoh

Fujii

. Simultaneous increases of leptin and gonadotropin-releasing hormone following exogenous estrogen administration in women with normally menstrual cycle. Endocr J 2005; 52: 449–454.

43.

Friedman

Halaas

. Leptin and the regulation of body weight in mammals. Nature 1998; 395: 763–770.

44.

Gannage-Yared

Khalife

Semaan

Fares

Jambart

Halaby

. Serum adiponectin and leptin levels in relation to the metabolic syndrome, androgenic profile and somatotropic axis in healthy non-diabetic elderly men. Eur J Endocrinol 2006; 155: 167–176.

45.

Dyck

. Adipokines as regulators of muscle metabolism and insulin sensitivity. Appl Physiol Nutr Metab 2009; 34: 396–402.

46.

Farooqi

O’Rahilly

. Leptin: a pivotal regulator of human energy homeostasis. Am J Clin Nutr 2009; 89(Suppl): 980–984.

47.

Clement

Vaisse

Lahlou

. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature 1998; 392: 398–401.

48.

Haluzik

Haluzikova

. The role of resistin in obesity-induced insulin resistance. Curr Opin Investig Drugs 2006; 7: 306–311.

49.

Arita

Kihara

Ouchi

. Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity. Biochem Biophys Res Commun 1999; 257: 79–83.

50.

Hotta

Funahashi

Arita

. Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol 2000; 20: 1595–1599.

51.

Drucker

. Minireview: the glucagon-like peptides. Endocrinology 2001; 142: 521–527.

GLP-2 and leptin are associated with hyperinsulinemia in non-obese female migraineurs

Abstract

Keywords

Introduction

Patients and methods

Patients

Blood collection

Laboratory procedures

Statistical analysis

Results

Discussion

References