Sage Journals: Discover world-class research

Abstract

Objectives:

Several studies have suggested an association between migraine and insulin resistance (IR) without adequately addressing the issue according to migraine type. We assessed IR in subjects with migraine with aura (MwA) and migraine without aura (MwoA) to estimate the consistency of the possible association.

Methods:

In a case-control study we included case subjects with MwA and MwoA, who were consecutively selected from those referred to our Regional Headache Center from September 2011 to February 2013, and age-matched control subjects selected using general practitioners’ databases. IR was calculated by means of the homeostatic model assessment of IR (HOMA-IR), β-cell function (HOMA-B), and the quantitative insulin sensitivity check index (QUICKI) measuring glucose and insulin values in a blood sample collected in the morning after overnight fasting. Data regarding anthropometric measures, comorbidity risk factors, and migraine characteristics were also recorded.

Results:

We recruited 50 case subjects with MwA (38 women) and 50 with MwoA (40 women) and 50 control subjects (40 women). Proportions of arterial hypertension, cigarette smoking, hypercholesterolemia, use of oral contraceptives, and mean values of the body mass index (BMI) were similar in the three groups. We found significantly different glucose values among and within groups considering case subjects with MwA and MwoA and control subjects (4.9 ± 0.6 vs 4.7 ± 0.5 vs 4.6 ± 0.5 mmol/l; p = 0.018) in the absence of any difference in insulin (53.1 ± 24.0 vs 56.7 ± 34.4 vs 53.8 ± 24.4 pmol/l; p = 0.811), HOMA-IR (1.6 ± 0.8 vs 1.7 ± 1.0 vs 1.6 ± 0.7; p = 0.765), HOMA-B (121.4 ± 71.1 vs 149.2 ± 93.8 vs 162.8 ± 109.7; p = 0.107), and QUICKI (0.36 ± 0.03 vs 0.37 ± 0.03 vs 0.37 ± 0.03; p = 0.877) values. The logistic regression model showed increased odds of MwA in subjects exposed to the highest tertile of glucose values. This association was confirmed in the adjusted model, in which case subjects with MwA were compared with those with MwoA but not with control subjects.

Conclusions:

In contrast to what has been shown by the majority of the available studies, the results of our study do not support the association of migraine with IR. As our study was not population-based and several patients had low disease activity, these findings need further confirmation.

Keywords

Migraine migraine with aura insulin resistance comorbidity

Introduction

Migraine, and particularly migraine with aura (MwA), has been associated with an increased risk of cardiovascular diseases including stroke, silent brain infarction, cardiac disease, and retinal artery disease (1 –4). Not many studies have addressed the burden of vascular risk factors, including diabetes, in subjects with migraine (5 –9). Results on the association between migraine and diabetes are controversial. A recent population-based study reported a higher prevalence, especially in migraineurs with aura (9). However, the available data mostly point to the lack of an association, or even to a negative association between migraine and diabetes (5,6,8). A reduced prevalence of migraine among older patients with diabetes has also been reported (7). Insulin resistance (IR) is a physiologic state in which a normal amount of insulin produces, among other manifestations, subnormal disposal of glucose in peripheral tissue (10). IR is present in patients with diabetes but also in nondiabetic individuals (10). Since IR represents a risk factor for cardiovascular diseases in the general population (11,12), its role might be relevant in patients with migraine. As a further point of interest, some studies have hypothesized that insulin might be implicated in migraine pathogenesis (13,14).

The existing data on IR in migraineurs largely point to an association between the two conditions (15 –23). Some of the reference studies were not specifically designed to address the role of IR in migraine and did not include an adequate number of patients with MwA to infer the role of IR. Since MwA appears to be a risk factor for cardiovascular disease, while migraine without aura (MwoA) does not (3,4), it is reasonable to hypothesize a positive association between IR and MwA, but not between IR and MwoA.

The present study aims to assess IR in patients with migraine to detect whether the claimed association between migraine and IR is present both in MwA and MwoA or only in MwA.

Methods

We evaluated consecutive patients with migraine referring to our Regional Headache Center from September 2011 to February 2013 and matching the diagnostic criteria of the International Classification for Headache Disorders, second edition (ICHD-II). The Regional Headache Center is located in a tertiary-level hospital and the visit is covered for all or most of the due amount (depending on personal financial status) by the National Health System; patients are referred to the Center by primary care physicians or specialists from several medical areas. Migraineurs were classified as MwA or MwoA by two experienced neurologists (SS and FP). To be included in the study patients had to be aged 18 years or older with an age of onset of their migraine attacks lower than 50 years and with at least three migraine attacks in the last three months. To avoid possible confounders, case subjects were excluded if they had a history of diabetes mellitus, thyroid dysfunction, chronic metabolic or inflammatory diseases, infectious diseases, cancer, drug or alcohol dependency, previous cardiovascular diseases, were taking medication that interfered with glucose metabolism (antidiabetics and corticosteroids), or presented any coexisting nonmigraine headache or chronic migraine.

Control subjects were residents recruited using general physicians’ databases covering the same area and were matched with subjects with MwA and MwoA by age (±2 years). Exclusion criteria for migraineurs were also applied to control subjects. All study participants were Caucasian.

The study was approved by the local ethics committee and written informed consent was obtained from all participants. At the first visit a structured questionnaire was completed to collect data referring to age, gender, ethnicity, family history of migraine and of cardiovascular diseases, migraine preventive treatments, and the use of drugs to treat acute migraine attacks. Case and control subjects were screened for the presence of cardiovascular risk factors such as arterial hypertension, hypercholesterolemia, and active cigarette smoking. Arterial hypertension was diagnosed in the presence of blood pressure values ≥140/90 mmHg in at least two independent measurements or if subjects were taking blood pressure lowering agents. Hypercholesterolemia was diagnosed in the presence of values of total blood cholesterol ≥240 mg/dl or if subjects were taking cholesterol lowering agents. Weight and height were measured in all subjects and body mass index (BMI) was calculated using the formula: BMI [kg/m²] = weight [kg]/(height [m])².

In case subjects the frequency and duration of the migraine attacks in the previous three months were also recorded. Attack frequency was classified as three or fewer episodes per month, four to eight episodes per month, and more than eight episodes per month; attack duration was classified as less than 24 hours, 24 to 48 hours, and more than 48 hours. The overall burden of migraine on daily life was evaluated by the Migraine Impact and Disability Assessment Scale (MIDAS) and by the Headache Impact Test, 6th version (HIT-6) (24,25).

Blood collection was performed by venous puncture after a 12-hour overnight fast. Migraine patients were headache free for at least 48 hours to avoid any interference of acute phase mechanisms with glucose levels. Glucose level was determined using the enzymatic method and expressed in mmol/l. Insulin level was measured by an ultrasensitive enzyme-linked immunosorbent assay and expressed in pmol/l. IR was assessed by calculating: 1) the homeostatic model assessment of insulin resistance (HOMA-IR) index, a simple, inexpensive, and reliable surrogate measure of IR obtained from mathematical manipulation of the values of fasting serum insulin and fasting glucose (26), 2) the homeostatic model assessment of β-cell function (HOMA-B), a good surrogate measure of β-cell function (27), and 3) the quantitative insulin sensitivity check index (QUICKI) as an index of insulin sensitivity (28). HOMA-IR was calculated as (fasting insulin [in µU/ml] × fasting glucose [in mmol/l])/22.5 (26). HOMA-B was calculated as (20 × fasting insulin [in µU/ml])/(fasting glucose [in mmol/l]–3.5) (27). QUICKI index was calculated as 1/(log insulin [in µU/ml] + log glucose [in mg/dl]) (28). In line with the International System of Units, we used the conversion factors 0.0555 to transform glucose and 7.175 to transform insulin from gravimetric (mg/dl and µU/ml, respectively) to international units (mmol/l and pmol/l, respectively).

Statistical analysis

Continuous variables are presented as means ± standard deviation (SD) and ordinal variables as medians with interquartile range (IQR). The Pearson χ² test was used to compare the distribution among categorical variables. Variables were compared using a two-sample t-test, one-way analysis of variance (ANOVA), Kruskal-Wallis test, or Wilcoxon test, as appropriate. HOMA-B and HOMA-IR were managed with nonparametric statistics as the Kolmogorov-Smirnov test of normality revealed a non-normal distribution. Logistic regression models were used to evaluate the associations between migraine and glucose or insulin or HOMA-IR or HOMA-B or QUICKI grouped into tertiles. The odds ratios (ORs) values were calculated using Wald’s method and each risk value was quoted with a 95% confidence interval (CI). The considered potential confounders were age and BMI as continuous variables and gender, arterial hypertension, cigarette smoking, and hypercholesterolemia as dummy variables (presence/absence). All the potential confounders were selected as possible factors associated with metabolic syndrome and IR. The Spearman or Pearson correlation was used to evaluate the relationship of glucose, insulin, HOMA-IR, HOMA-B, and QUICKI values with attack frequency and duration, migraine duration, MIDAS and HIT-6 scores, where appropriate. For study power calculation, we selected the smallest reported difference in HOMA-IR (0.6) in the available case-control studies (15 –20). The study had 98% power based on the following parameters: α (type I error) = 0.05; K (number of groups) = 3; σ (within-group standard deviation) = 0.7; and ϕ (non-centrality parameter value) = 2.47. The sample size estimated with a power of 98% was of 50 subjects for each group. P values < 0.05 were considered statistically significant. There were no missing data. All analyses were carried out using SAS/STAT statistical software.

Results

During the study period we included 50 case subjects with MwA (38 women), 50 with MwoA (40 women), and 50 age-matched control subjects (40 women). To enroll 50 case subjects with MwA, 91 consecutive subjects were evaluated; among them 30 did not meet our inclusion criteria and 11 refused to participate. The identified case subjects with MwoA numbered 104; among them 41 did not meet the inclusion criteria and 13 refused to participate. To enroll 50 control subjects we had to exclude 21 of 71 consecutive healthy volunteers. There were no differences in the baseline characteristics between subjects who agreed to participate in the study and those who refused (data not shown). No differences in the baseline characteristics were observed among the three groups (Table 1). Most of the included subjects had fewer than four attacks/month with attacks lasting less than 24 hours (Table 1). The median MIDAS score was lower (p = 0.011) in subjects with MwA (13; IQR 4–32) than in subjects with MwoA (31; IQR 12–42) while median HIT-6 scores were similar in the two groups (67; IQR 59–68 vs 63; IQR 61–67; p = 0.494). Duration of migraine was similar in subjects with MwA and MwoA (17.4 ± 10.9 years vs 15.8 ± 11.3 years; p = 0.481). As shown in Table 1, proportions of arterial hypertension, cigarette smoking, hypercholesterolemia, use of oral contraceptives, and mean ± SD values of BMI were similar in subjects with MwA, MwoA, and in control subjects.

Table 1.

Characteristics of case subjects and control subjects.

		Migraine with aura (n = 50)	Migraine without aura (n = 50)	Controls (n = 50)	p
		Mean ± SD
Age (years)		35.8 ± 9.8	37.0 ± 9.3	35.2 ± 10.6	0.638
BMI (kg/m²)		23.1 ± 3.3	23.5 ± 4.0	22.1 ± 2.8	0.090
Disease duration (years)		17.4 ± 10.9	15.8 ± 11.3	–	0.481
		n (%)
Gender (women)		38 (76.0)	40 (80.0)	40 (80.0)	0.853
Arterial hypertension		8 (16.0)	5 (10.0)	1 (2.0)	0.054
Cigarette smoking		14 (28.0)	17 (34.0)	12 (24.0)	0.538
Hypercholesterolemia		6 (12.0)	3 (6.0)	1 (2.0)	0.131
Oral contraceptives^a		6 (12.0)	4 (8.0)	4 (8.0)	0.662
Episode frequency (per month)	<4	33 (66.0)	29 (58.0)	–	0.706
	4 to 8	11 (22.0)	14 (28.0)	–
	>8	6 (12.0)	7 (14.0)	–
Episode duration (hours)	<24	29 (58.0)	33 (66.0)	–	0.708
	24 to 48	12 (24.0)	10 (20.0)	–
	>48	9 (18.0)	7 (14.0)	–

BMI: body mass index. ^aIn women. Continuous variables were compared by one-way analysis of variance (ANOVA) (age and BMI) or two-sample t-test (disease duration). Categorical variables were compared by the Kruskal-Wallis (gender, arterial hypertension, hypercholesterolemia, oral contraceptives) or the Wilcoxon test (episode frequency and duration).

No correlations were found in case subjects with MwA and MwoA of glucose, insulin, HOMA-IR, HOMA-B, and QUICKI values with migraine attack duration and frequency, migraine duration, and MIDAS and HIT-6 scores (data not shown).

We found significantly different glucose values among and within groups considering subjects with MwA, MwoA, and controls but we did not find any difference in insulin, HOMA-IR, HOMA-B, and QUICKI values (Table 2). In Table 3, we have summarized the OR estimates associated with the variables of interest in the study. When comparing case subjects with MwA with control subjects, our data demonstrated that the risk of MwA for subjects exposed to the highest tertile group of glucose values (OR 2.71; 95% CI 1.02–7.22) was increased; this finding was not confirmed when potential confounders were included in the model (OR 2.65; 95% CI 0.88–8.00). The same analysis, when comparing case subjects with MwA to case subjects with MwoA, showed that the risk of MwA for subjects exposed to the highest tertile group of glucose values was increased (OR 3.23; 95% CI 1.21–8.60); the finding was confirmed also in the adjusted model (OR 5.32; 95% CI 1.62–17.42).

Table 2.

Comparison of mean values of glucose, insulin, HOMA-IR, HOMA-B, and QUICKI among case subjects with MwA, MwoA, and control subjects.

Migraine with aura (n = 50)	Migraine without aura (n = 50)	Controls (n = 50)	F	p
Mean ± SD
Glucose (mmol/l)	4.9 ± 0.6	4.7 ± 0.5	4.6 ± 0.5	4.12	0.018
Insulin (pmol/l)	53.1 ± 24.4	56.7 ± 34.4	53.8 ± 24.4	0.21	0.811
HOMA-IR	1.6 ± 0.8	1.7 ± 1.0	1.6 ± 0.7	–	0.765
HOMA-B	121.4 ± 71.1	149.2 ± 93.8	162.8 ± 109.7	–	0.107
QUICKI	0.36 ± 0.03	0.37 ± 0.03	0.37 ± 0.03	0.13	0.877

HOMA-B: homeostatic model assessment for beta-cell function; HOMA-IR: homeostatic model assessment for insulin resistance; QUICKI: Quantitative Insulin Sensitivity Check Index.

One-way ANOVA was used to compare the distribution among normally distributed variables (glucose, insulin, and QUICKI) and the Kruskal-Wallis test among non-normally distributed variables (HOMA-IR and HOMA-B).

F: Variance.

Table 3.

Logistic regression analysis according to tertiles of glucose, insulin, HOMA-IR, HOMA-B, and QUICKI.

Parameter	Tertile 1^a	Tertile 2			Tertile 3
n	n	OR (95% CI) unadjusted	OR (95% CI) adjusted	n	OR (95% CI) unadjusted	OR (95% CI) adjusted
Migraine with aura vs controls
Glucose	38	31	1.30 (0.47–3.59)	1.48 (0.51–4.30)	31	2.71 (1.02–7.22)	2.65 (0.88–8.00)
Insulin	31	38	1.50 (0.58–3.89)	1.28 (0.47–3.48)	31	1.14 (0.42–3.09)	0.67 (0.21–2.14)
HOMA-B	28	37	0.71 (0.28–1.80)	0.69 (0.25–1.91)	35	0.56 (0.21–1.54)	0.44 (0.15–1.34)
HOMA-IR	31	38	1.71 (0.66–4.46)	1.63 (0.58–4.53)	31	1.48 (0.54–4.03)	0.88 (0.27–2.87)
QUICKI	31	38	1.16 (0.45–3.00)	1.86 (0.61–5.60)	31	0.68 (0.25–1.85)	1.14 (0.35–3.73)
Migraine without aura vs controls
Glucose	27	33	0.85 (0.34–2.14)	0.74 (0.27–2.04)	40	0.84 (0.32–2.23)	0.66 (0.22–1.93)
Insulin	31	32	0.75 (0.29–1.94)	0.66 (0.24–1.82)	37	0.80 (0.31–2.07)	0.66 (0.22–1.93)
HOMA-B	36	34	0.79 (0.30–2.11)	0.88 (0.31–2.50)	30	1.25 (0.47–3.30)	1.35 (0.47–3.85)
HOMA-IR	31	31	0.96 (0.37–2.48)	0.54 (0.19–1.52)	38	0.74 (0.29–1.92)	0.78 (0.27–2.24)
QUICKI	38	31	0.77 (0.29–2.10)	0.69 (0.23–2.12)	31	1.04 (0.40–2.69)	1.29 (0.45–3.69)
Migraine with aura vs migraine without aura
Glucose	37	30	1.53 (0.56–4.20)	1.78 (0.61–5.21)	33	3.23 (1.21–8.60)	5.32 (1.62–17.42)
Insulin	30	36	2.00 (0.77–5.18)	1.70 (0.56–5.16)	34	1.43 (0.53–3.84)	2.05 (0.74–5.68)
HOMA-B	32	33	0.90 (0.35–2.35)	0.81 (0.30–2.21)	35	0.45 (0.17–1.20)	0.40 (0.14–1.16)
HOMA-IR	32	35	2.31 (0.87–6.10)	2.43 (0.87–6.80)	33	1.54 (0.58–4.12)	1.79 (0.59–5.45)
QUICKI	33	35	1.50 (0.57–3.95)	1.35 (0.47–3.92)	32	0.65 (0.24–1.74)	0.56 (0.18–1.69)

Reference category. Tertile limits for glucose: tertile 1, 3.7–4.4; tertile 2, 4.5–4.9; tertile 3, 5.0–6.4 mmol/l. Tertile limits for insulin: tertile 1, 2.0–5.5; tertile 2, 5.6–8.7; tertile 3, 8.8–23.0 pmol/l. Tertile limits for HOMA-B: tertile 1, 23.2–88.9; tertile 2, 89.0–156.8; tertile 3, 156.9–558.0. Tertile limits for HOMA-IR: tertile 1, 0.44–1.14; tertile 2, 1.15–1.80; tertile 3, 1.81–5.05. Tertile limits for QUICKI: tertile 1, 0.302–0.349; tertile 2, 0.350–0.374; tertile 3, 0.375–0.443 pmol/l. Values are adjusted for confounding variables: age, gender, arterial hypertension, cigarette smoking, hypercholesterolemia, and body mass index. HOMA-B: homeostatic model assessment for beta-cell function; HOMA-IR: homeostatic model assessment for insulin resistance; QUICKI: Quantitative Insulin Sensitivity Check Index; OR: odds ratio; CI: confidence interval.

Discussion

This is the first study evaluating IR according to migraine type in an adequately powered sample of subjects. Unexpectedly, and at variance with the available studies, our results did not suggest an association between MwA and MwoA and IR. Comparisons between our data and those from other studies showed that some of them were not specifically designed to evaluate the association of IR with migraine, had different inclusion and exclusion criteria, found associations between migraine and markers of diabetes other than IR, or were based on a nonrandom convenience sample.

We have summarized in Table 4 the results of the six case-control studies that used HOMA-IR and were suitable for indirect comparisons (15 –20). Four of them, carried out by the same group of authors, found higher mean HOMA-IR values in migraineurs than in controls (15 –18). Those studies were not primarily aimed at addressing IR, and HOMA-IR was reported as collateral information. When comparing the values we obtained in the present study with those found in those same studies, HOMA-IR values were similar in case subjects and lower in control subjects (15 –18). This difference is crucial. Since in the four above-mentioned studies the control subjects were selected among patients’ friends and volunteers, we cannot exclude that a bias in the selection of controls may have amplified the difference between HOMA-IR values in migraineurs and in controls (15 –18). In the remaining two studies (19,20) mean HOMA-IR values in migraineurs were above the threshold level of 2.6, indicating IR (29). In the first of the two studies (19) the mean HOMA-IR value of control subjects was significantly lower than that of migraineurs and was very close to the value found in our study. Intriguingly, in the other study (20), even controls had a high mean HOMA-IR value, resulting in a nonsignificant difference between groups. Those high values indicating the presence of IR both in migraineurs and control subjects are difficult to explain, since both controls and migraineurs were young, nonobese, and nondiabetic, which suggests caution in receiving those data. In addition, in this same study, other indices of IR such as the impaired insulin sensitivity Stumvoll’s index and the oral glucose insulin sensitivity-180 were higher in the same migraineur subjects than in controls (20). It is difficult to explain why those tests, all reliable indicators of IR, led to different results in the same subjects. Three other studies on the same topic provided results not comparable with ours because of differences in reference parameters or in the selection of cases (21 –23). The authors of two of the three quoted studies found higher fasting insulin values in migraineurs than in controls without directly assessing IR (22,23). Isolated measurement of insulin without taking into account the parallel glucose levels cannot be considered as a reliable marker of IR. The authors of the third study, which lacked a control group and included subjects with metabolic syndrome, found IR, as assessed by the HOMA-IR, in 11.1% of migraineurs (21).

Table 4.

Available studies assessing insulin resistance in case subjects with migraine and in control subject.

First author, year (ref)	Subjects included (n)			Body mass index (kg/m²)				HOMA-IR
Any migraine		Controls	Any migraine		Controls	Any migraine		Controls
MwoA	MwA	MwoA	MwA	p	MwoA	MwA	p
Rainero, 2005 (20)	22	8	15	21.8 ± 2.3	21.6 ± 1.6	21.8 ± 2.3	NS	3.1 ± 1.0	3.1 ± 1.3	3.1 ± 0.9	NS
Guldiken, 2008 (19)	46	14	25	26.1 ± 5.6		25.1 ± 3.7	NS	2.8 ± 1.6	2.8 ± 1.0	1.8 ± 0.7	<0.01^a
Bernecker, 2010 (15)	27	13	44	22.3 ± 3.2		21.3 ± 2.6	NS	1.9 ± 1.6		0.9 ± 0.9	<0.01
Gruber, 2010 (18)	29	19	72	22.0 ± 3.0	23.9 ± 3.4	22.6 ± 3.1	NS	1.5 ± 1.1	2.3 ± 2.0	1.1 ± 1.0	<0.01^b
Bernecker, 2011 (16)	30	20	74	22.6 ± 3.1		22.5 ± 3.0	NS	1.73 ± 1.5		1.0 ± 1.0	<0.01
Bernecker, 2011 (17)	31	17	48	23.6 ± 5.2		21.8 ± 4.0	NS	2.1 ± 1.9		0.9 ± 0.9	<0.0001
Sacco, 2013 (present study)	50	50	50	23.5 ± 4.0	23.1 ± 3.3	22.1 ± 2.8	0.11	1.7 ± 1.0	1.6 ± 0.8	1.6 ± 0.7	0.79

MwoA: migraine without aura; MwA: migraine with aura; NS: not significant. ^ap for any migraine vs controls; no difference between MwA and MwoA (p > 0.05); ^bp for any migraine vs controls; p < 0.001 for cases with MwA compared to control subjects.

Intriguingly, the failure to find an association between migraine and IR in the present study is in agreement with the results of studies that evaluated the possible link between migraine and obesity and did not find any evidence of an association (30 –39). Those same studies claimed an association between obesity and increased migraine frequency and severity (30 –39). This might have implied that, since IR was associated with obesity, the association between IR and migraine could be relevant in patients with increased migraine frequency and severity. However, our results do not support this hypothesis since we did not find any correlation between IR and migraine frequency, severity, and duration.

The lack of clear evidence of an association between migraine and IR is not in conflict with the reported association between migraine and vascular disease. In fact, in the general population IR has been associated with vascular disease by favoring atherothrombosis (11) while in migraineurs atherothrombosis is not the primary mechanism leading to vascular events (40 –44). Likewise, our results are in line with studies reporting that diabetes is negatively associated with migraine (5,7).

Our data showed higher mean glucose values in case subjects with MwA (22,23). However, the difference in mean glucose values between case subjects with MwA and other groups, although statistically significant, was small. The analysis of the risk of migraine associated with glucose levels showed an increased risk of MwA vs MwoA in subjects exposed to the highest glucose values. Unfortunately, the design of our study does not allow any reliable conclusions to be drawn about the mechanisms that might underlie the association between high mean glucose values and MwA. The difference in mean glucose values among groups, although statistically significant, is very small and within the range of normal values: a computational result without any clinical implications. In addition, it has recently been shown in an experimental setting that increased cerebral glucose availability makes the tissue resistant to spreading depression (45).

The strengths of our study are the inclusion of patients both with MwA and MwoA, the investigation of a larger group of migraineurs than in comparable studies, and the use of the HOMA method, a widely accepted measure of IR. We are aware that HOMA methods may not be an ideal means of assessing IR; the gold standard is the euglycemic hyperinsulinemic clamp, but the procedure is technically complex and invasive (46,47). Likewise, tests based on oral or intravenous glucose tolerance are somewhat easier to manage, but imply major costs and substantial participant effort (46,48). As case subjects were not randomly selected from the general population, but were recruited from those referred to our Regional Headache Center, we are aware that our selection procedure may have introduced a selection bias. The main reason for referral to our center is the burden of the disease. Patients with infrequent attacks are usually not referred. Since fewer than three attacks in the last three months was an exclusion criterion in our study, most of the patients not referred to our center would not have been included. Frequency and duration of the migraine attacks were retrospectively assessed and we cannot exclude a recall bias. The majority of the cases included in our study had fewer than four attacks per month and attacks lasting less than 24 hours. We cannot therefore exclude that the lack of association between migraine and IR in our study might be attributed to the low disease activity in case subjects with migraine. While the power of the study was high to address possible differences among groups, the study may be underpowered to address tertile subgroups and the correlation between HOMA-IR and migraine frequency and duration, so the related data must be interpreted with caution. We excluded subjects with chronic migraine to avoid confounding factors linked to drug use and differences in pathogenic mechanisms with respect to episodic migraine, while we are aware that an evaluation of IR in this category of patients is warranted. Finally, we cannot exclude that migraineurs, to voluntarily prevent attacks, may follow a more regular lifestyle, which could, in turn, affect IR.

In conclusion, while most of the available studies suggest an association between IR and migraine, our study, involving a larger number of subjects and investigating the association according to migraine type, did not support this notion. Even considering all the possible limitations and pitfalls of our study, we believe that the claimed association of migraine with IR should be reviewed. There is the need for a large population-based study with a randomly selected sample stratified according to frequency, severity, and duration of migraine attacks, and presence or absence of aura to establish the role of IR in migraine.

Clinical implications

There is not adequate evidence to reliably consider migraine and insulin resistance as associated conditions.

There is not a correlation between insulin resistance and migraine, frequency, severity, and duration.

It is unlikely that insulin resistance contributes to the increased vascular risk in migraineurs.

Footnotes

Funding

This work was supported by the “ex 60%” grant of the Italian Ministero dell’Istruzione, dell’Università e della Ricerca in the year 2010.

Conflict of interest

None declared.

References

Bigal

Kurth

Santanello

. Migraine and cardiovascular disease: A population-based study. Neurology 2010; 74: 628–635.

Sacco

Cerone

Carolei

. Comorbid neuropathologies in migraine: An update on cerebrovascular and cardiovascular aspects. J Headache Pain 2008; 9: 237–248.

Sacco

Ricci

Carolei

. Migraine and vascular diseases: A review of the evidence and potential implications for management. Cephalalgia 2012; 32: 785–795.

Schürks

Rist

Bigal

. Migraine and cardiovascular disease: Systematic review and meta-analysis. BMJ 2009; 339: b3914–b3914.

Aamodt

Stovner

Midthjell

. Headache prevalence related to diabetes mellitus The Head-HUNT study. Eur J Neurol 2007; 14: 738–744.

Benseñor

Goulart

Lotufo

. Cardiovascular risk factors associated with migraine among the elderly with a low income: The Sao Paulo Ageing & Health Study (SPAH). Cephalalgia 2011; 31: 331–337.

Berge

Riise

Fasmer

. Does diabetes have a protective effect on migraine? Epidemiology 2013; 24: 129–134.

Burch

Rist

Winter

. Migraine and risk of incident diabetes in women: A prospective study. Cephalalgia 2012; 32: 991–997.

Stam

Weller

Janssens

. Migraine is not associated with enhanced atherosclerosis. Cephalalgia 2013; 33: 228–235.

10.

Krentz

. Insulin resistance. BMJ 1996; 313: 1385–1389.

11.

Kernan

Inzucchi

Viscoli

. Insulin resistance and risk for stroke. Neurology 2002; 59: 809–815.

12.

Rewers

Zaccaro

D'Agostino

. Insulin sensitivity, insulinemia, and coronary artery disease: The Insulin Resistance Atherosclerosis Study. Diabetes Care 2004; 27: 781–787.

13.

Bic

Blix

Hopp

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

14.

McCarthy

Hosford

Riley

. Single-nucleotide polymorphism alleles in the insulin receptor gene are associated with typical migraine. Genomics 2001; 78: 135–149.

15.

Bernecker

Pailer

Kieslinger

. GLP-2 and leptin are associated with hyperinsulinemia in non-obese female migraineurs. Cephalalgia 2010; 30: 1366–1374.

16.

Bernecker

Pailer

Kieslinger

. Increased matrix metalloproteinase activity is associated with migraine and migraine-related metabolic dysfunctions. Eur J Neurol 2011; 18: 571–576.

17.

Bernecker

Ragginer

Fauler

. Oxidative stress is associated with migraine and migraine-related metabolic risk in females. Eur J Neurol 2011; 18: 1233–1239.

18.

Gruber

Bernecker

Pailer

. Hyperinsulinaemia in migraineurs is associated with nitric oxide stress. Cephalalgia 2010; 30: 593–598.

19.

Guldiken

Demir

. Insulin resistance and high sensitivity C-reactive protein in migraine. Can J Neurol Sci 2008; 35: 448–451.

20.

Rainero

Limone

Ferrero

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

21.

Bhoi

Kalita

Misra

. Metabolic syndrome and insulin resistance in migraine. J Headache Pain 2012; 13: 321–326.

22.

Cavestro

Rosatello

Micca

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

23.

Hamed

Ezz Eldin

. Vascular risk factors, endothelial function, and carotid thickness in patients with migraine: Relationship to atherosclerosis. J Stroke Cerebrovasc Dis 2010; 19: 92–103.

24.

Kosinski

Bayliss

Bjorner

. A six-item short-form survey for measuring headache impact: The HIT-6. Qual Life Res 2003; 12: 963–974.

25.

Stewart

Lipton

Dowson

. Development and testing of the Migraine Disability Assessment (MIDAS) Questionnaire to assess headache-related disability. Neurology 2001; 56(Suppl 1): S20–S28.

26.

Matthews

Hosker

Rudenski

. Homeostasis model assessment: Insulin resistance and B-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419.

27.

Wallace

Levy

Matthews

. Use and abuse of HOMA modeling. Diabetes Care 2004; 27: 1487–1495.

28.

Katz

Nambi

Mather

. Quantitative insulin sensitivity check index: A simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab 2000; 85: 2402–2410.

29.

Ascaso

Pardo

Real

. Diagnosing insulin resistance by simple quantitative methods in subjects with normal glucose metabolism. Diabetes Care 2003; 26: 3320–3325.

30.

Bigal

Liberman

Lipton

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

31.

Bigal

Lipton

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

32.

Bigal

Tsang

Loder

. Body mass index and episodic headaches: A population-based study. Arch Intern Med 2007; 167: 1964–1970.

33.

Bond

Roth

Nash

. Migraine and obesity: Epidemiology, possible mechanisms and the potential role of weight loss treatment. Obes Rev 2011; 12: e362–e371.

34.

Evans

Williams

Rapoport

. The association of obesity with episodic and chronic migraine. Headache 2012; 52: 663–671.

35.

Ford

Pearson

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

36.

Peterlin

Rosso

Rapoport

. Obesity and migraine: The effect of age, gender and adipose tissue distribution. Headache 2010; 50: 52–62.

37.

Peterlin

Rapoport

Kurth

. Migraine and obesity: Epidemiology, mechanisms, and implications. Headache 2010; 50: 631–648.

38.

Sacco

Olivieri

Bastianello

. Comorbid neuropathologies in migraine. J Headache Pain 2006; 7: 222–230.

39.

Ainalem

Qiu

. Body mass index and adult weight gain among reproductive age women with migraine. Headache 2011; 51: 559–569.

40.

Ahmed

Bairey Merz

McClure

. Migraines, angiographic coronary artery disease and cardiovascular outcomes in women. Am J Med 2006; 119: 670–675.

41.

Kurth

Gaziano

Cook

. Migraine and risk of cardiovascular disease in men. Arch Intern Med 2007; 167: 795–801.

42.

MacClellan

Giles

Cole

. Probable migraine with visual aura and risk of ischemic stroke: The stroke prevention in young women study. Stroke 2007; 38: 2438–2445.

43.

Rist

Buring

Kase

. Migraine and functional outcome from ischemic cerebral events in women. Circulation 2010; 122: 2551–2557.

44.

Rose

Carson

Sanford

. Migraine and other headaches: Associations with Rose angina and coronary heart disease. Neurology 2004; 63: 2233–2239.

45.

Hoffmann

Sukhotinsky

Eikermann-Haerter

. Glucose modulation of spreading depression susceptibility. J Cereb Blood Flow Metab 2013; 33: 191–195.

46.

Caro

. Clinical Review 26. Insulin resistance in obese and nonobese man. J Clin Endocrinol Metab 1991; 73: 691–695.

47.

Howard

Bergman

Wagenknecht

. Ability of alternative indices of insulin sensitivity to predict cardiovascular risk: Comparison with the “minimal model”. Ann Epidemiol 1998; 8: 358–369.

48.

Matsuda

De Fronzo

. Insulin sensitivity indices obtained from oral glucose tolerance testing. Diabetes Care 1999; 22: 1462–1470.

Insulin resistance in migraineurs: Results from a case-control study

Abstract

Objectives:

Methods:

Results:

Conclusions:

Keywords

Introduction

Methods

Statistical analysis

Results

Discussion

Clinical implications

Footnotes

Funding

Conflict of interest

References