Appetite 55 (2010) 582–588

Contents lists available at ScienceDirect

Appetite

journal homepage: www.elsevier.com/locate/appet

Research report Sugar alters the level of serum insulin and plasma glucose and the serum cortisol:DHEAS ratio in female migraine sufferers§ Anna Kokavec a,*, Susan J. Crebbin b a La Trobe University, School of Psychological Science, Bendigo 3550, Australia b La Trobe University, School of Psychological Science, Bundoora 3086, Australia

ARTICLE INFO ABSTRACT

Article history: Early work has highlighted that a large percentage of migraineurs may have an altered glucidic Received 28 April 2010 methabolis due to carbohydrate-induced hyperinsulinism. The aim of this study was to assess the effect Received in revised form 7 September 2010 of sucrose on biomarkers of energy metabolism and utilization in migraineous females. A total of 16 Accepted 8 September 2010 participants (8 = Migraine, 8 = Non-migraine) at the mid-point of their menstrual cycle underwent a 15- h fast prior to ingesting 75 g sucrose dissolved in 175 g water. Blood sampling for the assessment of Key words: serum insulin, serum cortisol and serum dehydroepiandrosterone sulfate (DHEAS) and plasma glucose Migraine was conducted upon arrival at 09:00 h and then at regular 15-min intervals across a 150-min Sucrose experimental period. The results showed a signiﬁcant alteration in serum insulin and plasma glucose Cortisol DHEAS following sucrose ingestion in the migraine and non-migraine groups. In addition, signiﬁcant group Insulin differences were observed in the level of serum insulin, serum DHEAS, and the cortisol:DHEAS ratio with Glucose migraine participants on average recording a higher sucrose-induced serum insulin level and lower Hypoglycemia DHEAS level and cortisol:DHEAS ratio when group data was compared. It was concluded that while sucrose consumption may potentiate serum insulin in migraineurs this does not result in the development of sucrose-induced hypoglycemia in migraine or non-migraine participants. Crown Copyright ß 2010 Published by Elsevier Ltd. All rights reserved.

Introduction elevated insulin level may be specific to migraineurs (only). There is usually evidence of hypercortisolism in hyperinsulinemic Reactive hypoglycemia is a relatively uncommon meal-induced patients (Fruehwald-Schultes, Kern, Born, Fehm, & Peters, 2001), hypoglycemia disorder (Vexiau, Legoff, & Cathelineau, 1983). and similarly cortisol can be elevated in migraine sufferers (Ziegler, Patients who present with evidence of reactive hypoglycemia also Hassanein, Kodanaz, & Meek, 1979). Furthermore, conditions such report ingesting excessive quantities of refined carbohydrate and as tinnitus, poor eyesight, and vertigo are often reported by hyperinsulinism is usually responsible for the cause of the migraine patients and hyperinsulinemia is known to incite the hypoglycemia. Dietary restriction of refined carbohydrates is development of these conditions (Kazmierczak & Doroszewska, successful in the treatment of reactive hypoglycemia, although 2001). some patients can require pharmacological assistance (Hofeldt, Dexter and colleagues (1978) performed a standard oral glucose 1989). tolerance test (100 g glucose) on 74 migraineurs and identified Early studies have highlighted that a large percentage of glucose tolerance curves consistent with a classification of diabetes migraineurs present with signs of a reactive hypoglycemia (Dexter, and reactive hypoglycemia in 8% and 72% of migraine patients, Roberts, & Byer, 1978), due to diet-induced hyperinsulinism respectively. Subsequent dietary therapy with a low sucrose six- (Roberts, 1967). Moreover, research conducted by Cavestro and meal regimen resulted in greater than 75% reduction in migraine colleagues (2007), which compared glucose and insulin concen- frequency in all diabetic patients (100%), and most migraineurs trations in a total of 84 migraineurs, 25 headache patients, and 26 (63%) who previously demonstrated reactive hypoglycemia curves. non-migraine/headache controls confirmed that while an eleva- It has been known for some time that fasting-induced tion in glucose may be associated with headaches in general, an hypoglycemia can promote the development of migraine (Critchley, 1933). Later studies showed that food deprivation (Martin & Seneviratne, 1997), and the missing of one or more meals could trigger a migraine attack (Blau & Cumings, 1966). The majority of § This study was funded by a research grant provided by the Hunter Medical Research Institute (Newcastle, Australia). migraine attacks seem to occur between 6AM and midday (Solomon, * Corresponding author. 1992), with studies showing that more than 48% of migraine attacks E-mail address: [email protected] (A. Kokavec). occurring between 4 and 9 a.m. (Fox & Davis, 1998).

A significant association between polymorphisms in the insulin of oral contraceptive usage or hormone replacement therapy; or receptor gene and migraine pathogenesis has been suggested (6) were post-menopausal. In addition, all non-migraine control (McCarthy et al., 2001). Impaired insulin sensitivity has been participants were assessed using the International Headache implicated in vascular diseases and comorbidity between migraine Society criteria and only those who did not satisfy any of the and conditions such as stroke and hypertension is known to exist. headache criteria listed (IHS, 2004) were selected for participation. An evaluation of insulin sensitivity in 30 non-diabetic normoten- Subject participation was obtained by informed consent and no sive migraine patients has revealed that insulin sensitivity is also financial or other incentives were provided to any participant in impaired in migraine. Thus, insulin resistance may have some role return for participation in this study. Approval for the study was in the comorbidity reported between these two conditions granted by La Trobe University Human Ethics Committee (HEC No. (Rainero et al., 2005). 08-025), who determined that the procedures were consistent Early suggestions that migraine may be triggered by food with ethical guidelines for human research set by the National allergies (Monro, Carini, & Brostoff, 1984), fasting hypoglycemia Health and Medical Research Council of Australia (2001). (Blau & Cumings, 1966; Critchley, 1933; Fox & Davis, 1998; Marsters, Mortimer, & Hay, 1986; Martin & Seneviratne, 1997; Equipment and measures Peroutka, 2002), hyperinsulinism (Cavestro et al., 2007), insulin resistance (Rainero et al., 2005) or diet-induced reactive hypogly- Serum insulin, serum cortisol and serum DHEAS was assessed cemia (Dexter et al., 1978; Roberts, 1967), which can effectively be by Cobas Elecsys 20101 immunoassay (Roche Diagnostics, managed by altering dietary behaviour, remain to be clarified. Indianapolis, IN, USA). The sensitivity (lower detection limit) of Thus, the aim of this study is to attempt to address this discrepancy the Elecsys: insulin assay is .20 mU/mL; cortisol assay is 0.5 nmol/ in the migraine literature by assessing the effect of a sucrose load L; and DHEAS assay is 0.003 mmol/L. The intraassay coefficient of on biomarkers of glucoregulation in migraineous females. variation of the: Elecsys insulin assay was 1.9% at 6.36 mU/mL and 1.9% at 20.9 mU/mL; Elecsys cortisol assay was 3.2% at 202 nmol/L, Method 3.1% at 377 nmol/L, and 2.2% at 546 nmol/L; and Elecsys DHEAS assay was 2.8% at 3.18 mmol/L, 2.4% at 10.7 mmol/L, and 1.7% at Participants 26.7 mmol/L. Values in the 5th–95th percentile range for healthy fasted individuals was: 2.6–24.9 mU/mL for serum insulin; .96– A total of 18 participants aged between 38 and 54 years were 6.95 mmol/L for serum DHEAS in females aged 45–54 years (and recruited via a general advertisement placed in local and 3.28 mmol/L at 50th percentile), with effects of the menopause on community newspapers. All participants who replied to the the results tested and found to be negligible; and 171–536 nmol/L advertisement were interviewed and 12 individuals (8 females, for serum cortisol. Staff at Analytic Reference Laboratories (St. 4 males) were found to be suitable for allocation to a migraine Kilda Road, Melbourne, VIC), who were blind to the experimental group using the International Headache Society (IHS, 2004) criteria procedure, performed all biochemical analyses. listed for migraine. At least two participants did not meet the IHS A total of 75 g refined sugar in the form of commercially criteria for migraine but were found to be suitable for allocation to available sucrose, dissolved in 175 mL water, was used as the the non-migraine group and a further six non-migraineous carbohydrate load. Plasma glucose was monitored using a 1 controls were recruited from the local area and were known to Caresens (i-SENS Inc, Korea) blood glucose monitoring system. the researchers (Table 1). Semi-quantitative urinalysis was performed with LabstixTM (Bayer All participants (migraine and non-migraine) were excluded if Australia Limited) to measure ketones (sensitivity was 0.5– they: (1) had any history of psychiatric disorder, neurological 1.0 mmol/L acetoacetic acid). disease, or any major physical complaint including diabetes; (2) had a history of illicit drug use; (3) satisfied the DSM-IV diagnostic Experimental procedure criteria for alcohol abuse and/or dependence (American Psychiatric Association, 2000); (4) smoked tobacco or consumed any All participants were required to undergo a structured substance containing nicotine within 48 h; (5) had a recent history interview prior to participation in the study. During the structured

Table 1 Summary of headache symptoms as listed by headache criteria provided by the International Headache Society (HIS, 1988) for the migraine group (N = 7).

Criteria N %

Migraine (without aura) A. At least five headache attacks lasting 4–72 h, which has at least two of the four following characteristics: 7 100 1. Unilateral location 572 2. Pulsating quality 457 3. Moderate or severe intensity (inhibits or prohibits daily activities) 7 100 4. Aggravated by walking stairs or similar routine physical activity 7 100 B. During headache at least one of the two following symptoms occur: 1. Phonophobia and photophobia 7 100 2. Nausea and/or vomiting 7 100 Migraine (with aura) A. At least two attacks fulfilling with at least three of the following: 114 1. One or more fully reversible aura symptoms indicating focal cerebral cortical and/or brain stem functions 2. At least one aura symptom develops gradually over more than 4 min, or two or more symptoms occur in succession 3 43 3. No aura symptom lasts more than 60 min; if more than one aura symptom is present, accepted duration is proportionally increased 1 14 4. Headache follows aura with free interval of at least 60 min (it may also simultaneously begin with the aura) 3 43 B. At least one of the following aura features establishes a diagnosis of migraine with typical aura: 1. Homonymous visual disturbance 686 2. Unilateral paresthesia and/or numbness 343 3. Unilateral weakness 114 4. Aphasia or unclassifiable speech difficulty 457

Headache Classification Committee of the HIS. (1988). Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia, 8, 1–96. 584 A. Kokavec, S.J. Crebbin / Appetite 55 (2010) 582–588 interview participants were asked to provide self-report informa- SD = 4 years) were finally selected to participate in the current tion with respect to: demographic factors (e.g. age, gender); study. A decision was made to only include female participants due lifestyle factors (diet, alcohol and tobacco usage); headache to suggestions that the level of steroid hormones can be influenced history; medical history (individual and relevant family aspects, by age and sex differences (Orentreich, Brind, Rizer, & Vogelman, menstrual disorders, use of oral contraceptives, mood disorders, 1984; Uhart, Chong, Oswald, Lin, & Wand, 2006). psychosis); any physical or emotional trauma. At the completion of All participants were white of Anglo-Saxon origin, and no female the interview the migraine and headache status of all participants was obese as the weight of all participants when assessed was within was assessed in accordance with the IHS (2004) criteria for the the medically recommended range for age and height. The majority diagnosis of migraine. of females lived at home with family (87.5%), all claimed to be non- A total of three testing sessions were conducted at 7-day intervals smokers and irregular social consumers of alcohol. during the same month. Allocation to any particular testing session Ketones were detected by urinalysis in all participants prior to for migraine and non-migraine participants was determined by the participation in their respective experimental trial, which confirmed timing of a participant’s menstrual cycle. Blood glucose level can that all participants had complied with the fasting conditions. vary across the menstrual cycle (Brand-Miller, 2004) and each The ingestion of 75 g sucrose did not negatively influence the well participant was allocated to the testing session scheduled as near as being of non-migraine participants and no reports of headache, possible to the mid-point of their menstrual cycle. nausea or general malaise were received. Alternatively, following On the day prior to testing strenuous exercise was not the ingestion of the sucrose load most migraine participants permitted and all participants were required to maintain a 15-h reported feeling nauseous (n = 6), mild headache (n = 6), or general fast commencing at 18:00 h Eastern Standard Time (EST). Testing malaise (n = 7). One migraine participant was forced to withdraw was scheduled the following morning at 09:00 h EST. Upon arrival from the study due to onset of migraine symptoms severe enough a catheter was fitted to the left forearm of each participant in order to require triptan medication and as a result the data from this to facilitate repeated blood sampling and minimize stress participant has not been included in the final analysis. associated with the blood sampling procedure. At 09:00 h Demographic data collected during the structured interview participants were asked to provide a 50 mL urine sample for revealed that the age of onset for the migraine group was onset of urinalysis prior to having 10 mL of blood drawn from their left menstruation between age 13–16 years (n = 5) and early 20s forearm for assessment of serum insulin, cortisol and DHEAS and (n = 1). However, one participant reported migraine symptoms plasma glucose. The purpose of the urinalysis was to assess for the since age 5 (n = 1). At least five of the seven women in the migraine presence of ketone bodies, a valid indicator of fasting (Kaplan & group reported their migraine symptoms were linked to their Pesce, 1996) in order to confirm that all participants had complied menstrual cycle (>71%). In most cases the migraine could occur in with the fasting conditions. Following the blood taking procedure the week before and after menstruation and be severe enough to participants immediately ingested 75 g sucrose, which had been last for 1–2 days. In addition, one participant also reported dissolved in 175 mL water. Sucrose ingestion was monitored to migraine during menstruation. On average migraine severe ensure that the rate of consumption was the same for all enough to require medication occurred 2–3 times per month participants. Further blood sampling for the assessment of serum (n = 5), 3–4 times per year (n = 1) and >5 times per month (n = 1). insulin, cortisol and DHEAS and plasma glucose was performed at Migraine severe enough to require hospitalization was reported by regular 15-min intervals and all physical activity was restricted most participants with the number of hospitalizations being 1 during the 150-min testing period. No subject experienced (n = 1), 2 (n = 1), 3 (n = 2) and 6 (n = 1). At least one participant gastrointestinal distress during the course of any of the three trials. claimed she had never been hospitalized and another participant reported her migraine was severe enough to go to hospital but she Data analysis usually consumed a sedative in order to allow her to sleep. Migraine triggers in addition to menstrual cycle reported included Hormonal data from the three testing sessions was collated and fasting (n = 2), increased emotional stress (n = 3), and heat (n = 1). descriptive group statistics for migraine and non-migraine Nutritional triggers were reported by all migraine participants participants was computed. Group comparisons across time points (n = 7) and all reported that the ingestion of monosodium were made using a 2 Â 11 multivariate analysis of variance glutamate and red wine could instantly trigger a migraine (MANOVA). The independent variables in the MANOVA design (n = 7). There was variability with respect to other foods with were Group (Migraine, Non-Migraine), and Time (0-min, 15-min, cheese (n = 2), peanuts (n = 1), chocolate (n = 4) and bananas 30-min, 45-min, 60-min, 75-min, 90-min, 105-min, 120-min, 135- (n = 1) being noted to promote the development of migraine. min, 150-min). The dependent variables in the MANOVA design Furthermore, many participants reported disliking these foods were measured level of: serum insulin (mU/mL), serum cortisol despite not having noticed a connection between the particular (nmol/L), serum DHEAS (mmol/L), plasma glucose (mmol/L), and food and development of migraine symptoms. the calculated serum cortisol:DHEAS ratio (mg/dL). Calculation of Descriptive group serum insulin and plasma glucose data for sample size revealed that for a 2 Â 11 ANOVA design a minimum of migraine and non-migraine participants is presented graphically in 6 participants per cell is required in order to detect a moderate- Figs. 1 and 2. Additionally, descriptive serum cortisol and serum large effect at power = .80 when a is set at .05 (Hinkle, Wiersma, & DHEAS data for migraine and non-migraine participants and the Jurs, 2003). Thus, a total of seven migraine participants was serum cortisol:DHEAS ratio is presented in Figs. 3–5. calculated as being sufficient to reveal any significant group The MANOVA produced two significant main effects, and one difference between migraine and non-migraine participants. non-significant interaction. Utilizing the more conservative Wilks’ Results were classed as significant if the probability of achieving Lambda statistic results for the Time factor produced a robust and a result in the population was <.05. largely expected result (F(50, 637.3) = 3.24, p < .01). Study of univariate F ratios revealed significant Time effects for serum Results insulin (F(10, 143) = 11.50, p < .01) and plasma glucose (F(10, 143) = 9.87, p < .01). No significant contributions to the Time effect A total of eight migraineous females (age range = 38–54 years, were observed for serum cortisol (F(10, 143) = .53, p > .86) or M = 46 years, SD = 5 years) and eight age-matched non-migrai- serum DHEAS (F(10, 143) = .02, p > .99). Thus, the consumption of neous female controls (age range = 42–54 years, M = 48 years, sucrose does not appear to result in any significant alteration in [(Fig._1)TD$IG] A. Kokavec, S.J. Crebbin / Appetite[(Fig._3)TD$IG] 55 (2010) 582–588 585

60 400 Migraine Non-Migraine 350 50

300

40 U/mL 250 μ

30 200

150 20 Mean serum cortisol nmol/L

Mean serum insulin 100

10 50 Migraine Non-Migraine 0 0 0 15 30 45 60 75 90 105 120 135 150 0 15 30 45 60 75 90 105 120 135 150 Time in minutes Time in Minutes

Fig. 1. Mean serum insulin level before (0-min) and after sucrose ingestion (15– Fig. 3. Mean serum cortisol level before (0-min) and after sucrose ingestion (15– 150-min) in migraine (n = 7) and non-migraine (n = 8) participants. Data is shown 150-min) in migraine (n = 7) and non-migraine (n = 8) participants. Data is shown as the mean Æ S.E.M. (N = 15). as the mean Æ S.E.M. (N = 15).

steroid hormones (only) in fasted migraine or non-migraine The non-significant interaction between Time Â Group (F(50, participants despite the level of serum insulin and plasma glucose 637.3) = .34, p > .99) showed that the two factors were indepen- becoming significantly elevated. dent. Univariate data confirmed this result, with non-significant Results obtained for the Group factor were also significant (F(5, findings for serum insulin (F(10, 143) = .22 p > .99), serum 139) = 6.77, p < .01). Univariate data showed that serum insulin cortisol (F(10, 143) = .41, p > .94), serum DHEAS (F(10, 143) = .03, (F(1, 143) = 5.21, p < .03), and serum DHEAS (F(1, 143) = 13.82, p > .99) and plasma glucose (F(10, 143) = .56, p > .84) being p < .01) were significant contributors to the Group effect with the obtained. level of serum insulin noted to be significantly higher in the Inspection of the raw plasma glucose data revealed that the migraine group. Alternatively, the level of DHEAS was observed to level of plasma glucose was >4 mmol/L in all participants prior to be significantly lower in migraine participants. In addition, the sucrose ingestion. However, the level of post-sucrose plasma cortisol:DHEAS ratio was also significant (F(1, 143) = 5.49, p < .03), glucose was noted to drop below 4 mmol/L in two non-migraine with a higher ratio being noted in non-migraine participants. No participants (2.1 and 3.5 mmol/L) at 11:15 and 11:30 h, respec- significant contributions to the Group effect were observed for tively, and one migraine participant (3.5 mmol/L) at 11:30 h. serum cortisol (F(1, 143) = 2.76, p < .10) or plasma glucose (F(1, Inspection of the serum insulin and plasma glucose data did not 143) = 2.37, p < .17). An antagonistic relationship usually exists reveal evidence of any sucrose-induced hyperinsulinism in between DHEAS and insulin (Baulieu, 1996), and the migraine data migraine or non-migraine participants. The level of serum insulin for these two biochemical parameters appears consistent with this was noted to be <5.1 mU/mL in all three participants who suggestion. demonstrated a plasma glucose level <4 mmol/L. [(Fig._2)TD$IG] [(Fig._4)TD$IG]

9 6

8 5 7

6 4 mol/L μ 5 3 4

3 2 Mean plasmaglucose mmol/L

2 Mean serum DHEAS 1 1 Migraine Migraine Non-Migraine Non-Migraine 0 0 0 15 30 45 60 75 90 105 120 135 150 0 15 30 45 60 75 90 105 120 135 150 Time in Minutes Time in Minutes

Fig. 2. Mean plasma glucose level before (0-min) and after sucrose ingestion (15– Fig. 4. Mean serum DHEAS level before (0-min) and after sucrose ingestion (15– 150-min) in migraine (n = 7) and non-migraine (n = 8) participants. Data is shown 150-min) in migraine (n = 7) and non-migraine (n = 8) participants. Data is shown as the mean Æ S.E.M. (N = 15). as the mean Æ S.E.M. (N = 15). 586[(Fig._5)TD$IG] A. Kokavec, S.J. Crebbin / Appetite 55 (2010) 582–588

25 However, we did not ﬁnd any evidence of hypercortisolism in any of the participants when serum cortisol was assessed. Further- more, the consumption of sucrose did not result in any signiﬁcant 20 alteration in serum cortisol over time. In contrast, qualitative

g/dL assessment of the group data suggested a non-signiﬁcant trend μ for an initial sucrose-induced decrease in serum cortisol 15 concentration in migraine participants. In the brain steroid hormones act as modulators of synaptic

10 events (Majewska, 1992). Glucocorticoids such as cortisol can induce a metabolic vulnerability to neurons (Virgin et al., 1991), and astrocytes (Tombaugh, Yang, Swanson, & Sapolsky, 1992; 5 Vielkind, Walencewicz, Levine, & Bohn, 1990), by inhibiting local cerebral glucose utilisation and interfering with glucose transport

Mean Cortisol:DHEAS ratio Migraine Non-Migraine into neurons and glia (Horner, Packan, & Sapolsky, 1990; Virgin 0 et al., 1991). The role of DHEAS may be to antagonize the effect of 0 15 30 45 60 75 90 105 120 135 150 Time in Minutes cortisol in the brain. A low level of DHEAS may allow even a low level of cortisol to act unopposed and the amount of DHEAS in Fig. 5. Mean serum cortisol:DHEAS ratio mg/dL before (0-min) and after sucrose plasma is a major source of DHEAS in the brain (Guazzo et al., ingestion (15–150-min) in migraine (n = 7) and non-migraine (n = 8) participants. 1996). Therefore, the significantly reduced level of DHEAS noted in Data is shown as the mean Æ S.E.M. (N = 15). migraine sufferers could suggest some increased risk for hypoglycemia. However, the provision of additional energy substrates such as glucose to neurons and astrocyctes can attenuate the hypoxic Discussion and hypoglycemic damage induced by cortisol (Sapolsky, 1986; Sapolsky, Packan, & Vale, 1988; Tombaugh et al., 1992). The results showed that the consumption of 75 g sucrose prior Recent findings suggest that the salivary cortisol:DHEAS ratio is to food can significantly alter the level of serum insulin and significantly higher in migraine patients when compared to plasma glucose in the absence of any change in serum cortisol or controls (Patacchioli et al., 2006). However, the serum data in serum DHEAS concentration in female migraine and non- the present study is at odds with the salivary data with the serum migraine participants. However, significant group differences cortisol:DHEAS ratio calculated to be significantly higher in non- were observed on some (but not all) biochemical parameters with migraine participants. Moreover, on average the cortisol:DHEAS the level of fasting serum DHEAS and cortisol:DHEAS ratio noted ratio was 1:17–19 and 1:10–18 mg/dL across the 150-min study in to be significantly lower (in general). Alternatively, the level of non-migraine and migraine participants, respectively. However, it serum insulin following sucrose ingestion (only) was significantly should be noted that the migraineurs studied by Patacchioli and higher in migraine participants when compared to non-migraine colleagues, unlike some in the present study, were all chronic controls. migraine sufferers of both genders, which could have had some In the present study we did not observe any significant group influence on the cortisol data. differences in the level of plasma glucose following sucrose Early work has suggested that DHEAS may be a highly specific ingestion when migraine and non-migraine participants were individual marker under hereditary control (Thomas et al., 1994). compared. Instead, qualitative assessment of the plasma glucose Similarly, migraine is thought to be due to genetic factors data suggested a non-significant trend for a higher level of plasma (McCarthy et al., 2001) and it may be more than merely glucose over time in the migraine group, which is at odds with the coincidence that adrenarche is related to both DHEAS synthesis suggestion that migrainerus can show signs of carbohydrate- (Baulieu, 1996) and onset of migraine symptoms. In the present induced reactive hypoglycemia (Dexter et al., 1978). study we noted a significantly lower level of serum DHEAS in The serum insulin and plasma glucose data in the current study migraine participants with the average level of fasting serum supports the suggestion that a significantly elevated insulin level DHEAS in the migraine group assessed as being in the lower 50% combined with an elevated plasma glucose level can occur in expected range for fasted individuals. A low level of DHEAS has migraineurs (Cavestro et al., 2007). We observed significant group been associated with obesity (Jakubowicz, Beer, Beer, & Nestler, differences in serum insulin concentration with the level of serum 1995) and increased risk of age related diseases such as cancer insulin observed to be higher on average following sucrose intake (Ebeling & Koivisto, 1994), cardiovascular disease (Barrett-Connor, at all time points in migraine participants. Furthermore, a non- Khaw, & Yen, 1986; Mitchell, Sprecher, Borecki, Rice, & Laskas- significant trend for a higher level of post-sucrose glucose at 75- zewski, 1994), and diabetes (Small, Gray, Beastall, & MacCuish, min in migraine participants was also noted. 1989). Thus, migraine sufferers may be at increased risk of DHEAS is the most abundant steroid in the body with the level developing these chronic conditions and research aimed at of plasma DHEAS being 5–20 times higher that of plasma cortisol in investigating whether a connection between migraine and DHEAS young adults. There is a high correlation between the level of exists is encouraged. DHEAS and cortisol in cerebrospinal fluid (CSF) and saliva. The study of activity in the endocrine system when food intake Moreover, the level of DHEAS and cortisol in CSF represents is varied demonstrates the metabolic effects of feeding and fasting approximately 0.2% and 5% of levels measured in plasma, (Dallman, Akana, Strack, Hanson, & Sebastian, 1995). The respectively. Thus, the steroidal environment in the brain is relationship between cortisol and insulin (Kaplan & Pesce, 1996) dissimilar to that in the blood (Guazzo, Kirkpatrick, Goodyer, and DHEAS and insulin (Baulieu, 1996), is usually antagonistic. Shiers, & Herbert, 1996). After a meal rich in nutrients, insulin a pancreatic peptide hormone Migraineurs reportedly show significantly raised salivary plays the major role in the utilization of glucose, amino acid and cortisol levels in the early morning and late evening (Patacchioli free fatty acids. Alternatively, in the absence of adequate nutrition et al., 2006). Indeed, similar to that observed in hyperinsulinemic when insulin level is low, cortisol raises the concentration of patients (Fruehwald-Schultes et al., 2001), hypercortisolism is a glucose by stimulating gluconeogenesis (Dallman et al., 1993). common finding in migraine patients (Ziegler et al., 1979). While the relationship between cortisol and insulin is usually A. Kokavec, S.J. Crebbin / Appetite 55 (2010) 582–588 587 antagonistic (Kaplan & Pesce, 1996), elevations in cortisol and Conclusions insulin can occur simultaneously in response to food intake. The role of cortisol in this instance is to modulate the effects of insulin The results of this study have demonstrated that the on glucose utilization when glycogen stores are depleted, in order consumption of 75 g sucrose under fasting conditions can promote to ensure that energy stores are replenished (Goldstein et al., a significant increase in serum insulin in female migraine sufferers. 1993). Similar to cortisol, the relationship between insulin and Alternatively, a significantly lower level of serum DHEAS and DHEAS is also antagonistic. A direct effect of DHEAS is to inhibit cortisol:DHEAS ratio was also noted. However, while these glucose-6-phosphate dehydrogenase, an important factor associ- findings together could suggest some link between steroid ated with lipogenesis. Similar to cortisol, the level of DHEAS is hormones, hypoglycemia and migraine pathogenesis the signifi- elevated during fasting and lowered after a meal when the level of cance of the DHEAS findings to migraine pathology should be insulin is usually elevated (Baulieu, 1996). Therefore, the interpreted with caution given the small sample size. A non- significantly lowered level of DHEAS observed in this study could significant trend for a sucrose-induced elevation of plasma glucose be highlighting that lipogenesis is promoted in migrainous concentration over time was noted and when this is considered females. together with the plasma insulin data early claims that sucrose It has been suggested that insulin sensitivity is impaired in may promote a reactive hypoglycemia due to hyperinsulinism do migraine (Rainero et al., 2005) and the insulin and glucose data in not at first appear to be supported by the data in this study. this study appears to some extent to support this claim. Significant However, the steroid data could suggest there may be a need to group differences were observed following sucrose ingestion with increase glucose in order to attenuate cortisol-induced hypogly- respect to serum insulin in the migraine group and this was not cemia due to the significantly lowered DHEAS. It was suggested accompanied by any significant group difference in plasma that the increased appetite for sucrose, which is often observed in glucose, which could suggest some impairment in insulin migraine sufferers could to some degree be related to increase 5- sensitivity in migraineous females. However, whether sucrose HT activity. However, future research using a larger sample size promotes insulin insensitivity will need to be researched further with some controls for nutritional status could be directed towards before any firm conclusions can be made. It is probably pertinent at investigating this further and clarifying whether this is indeed the this point to report that a significant association between case. polymorphisms in the insulin receptor gene and migraine pathogenesis have been confirmed (McCarthy et al., 2001), which further suggests that a link may exist between migraine and References pancreatic function. American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental The level of transmitter that serotoninergic neurons release is Disorders (4th Edition). Washington DC: American Psychiatric Association. text altered with food intake. Moreover, it has been reported that revision. carbohydrates with a higher glycemic index may have a greater Barrett-Connor, E., Khaw, K. T., & Yen, S. S. C. (1986). A prospective study of dehydroepiandrosterone sulfate, mortality and cardiovascular disease. New England Journal serotoninergic effect than carbohydrates with a low glycemic of Medicine, 315, 1519–1524. index (Lyons & Truswell, 1988). Comparisons with starch show Baulieu, E. (1996). Dehydroepiandrosterone (DHEA): a fountain of youth? Journal of that sucrose can produce a higher glucose and insulin peak Clinical Endocrinology and Metabolism, 81, 3147–3151. Baumann, P., & Gaillard, J. (1976). Insulin coma therapy: decrease of plasma tryptophan (Yamada, Sugimoto, Kimura, Takeuchi, & Horisaka, 1989) and in man. Journal of Neural Transmission, 39, 309–313. increased serotonin (5-HT) level (Smolders, Loo, Sarre, Ebinger, & Blau, J. N., & Cumings, J. N. (1966). Method of precipitating and preventing some Michotte, 2001). migraine attacks. British Medical Journal, 2, 1242–1243. Brand-Miller, J. C. (2004). Postprandial glycemia, glycemic index, and the prevention of Migraine sufferers often immediately prior to an attack report type 2 diabetes. American Journal of Clinical Nutrition, 80, 243–244. cravings for food high in refined sugar (Jacome, 2001). Sucrose Cavestro, C., Rosatello, A., Micca, G., Ravotto, M., Marino, M. P., Asteggiano, G., et al. significantly increases the insulin level, which in turn promotes the (2007). Insulin metabolism is altered in migraineurs: a new pathogenic mecha- uptake of tryptophan (Baumann & Gaillard, 1976), and ultimately nism for migraine? Headache, 47, 1436–1442. Critchley, M. (1933). Migraine. Lancet, 1, 123–126. influences the synthesis of 5-HT. It is well accepted that a Dallman, M. F., Akana, S. F., Strack, A. M., Hanson, E. S., & Sebastian, R. J. (1995). The connection exists between not only the serotonergic system and neural network that regulates energy balance is responsive to glucocorticoids and mood disturbance but also between the serotonergic system and insulin and also regulates HPA axis responsivity at a site proximal to CRF neurons. Annals New York Academy of Sciences, 771, 730–742. migraine itself. Comorbidity with mood disturbance is a common Dallman, M. F., Strack, A. M., Akana, S. F., Bradbury, M. J., Hanson, E. S., Scribner, K. A., finding in the migraine population (Neuhauser & Lempert, 2004; et al. (1993). Feast and famine: critical role of glucocorticoids with insulin in daily Radat, Irachabal, Swendsen, & Henry, 2002) and sucrose can energy flow. Frontiers in Neuroendocrinology, 14, 303–347. Dexter, J. D., Roberts, J., & Byer, J. A. (1978). The five hour glucose tolerance test and promote the release of 5-HT (Smolders et al., 2001). Thus, when effect of low sucrose diet in migraine. Headache, 18, 91–94. these findings are considered together it is tempting to suggest Ebeling, P., & Koivisto, V. A. (1994). Physiological importance of dehydroepiandroster- that the preference for refined sugar observed in migraine sufferers one. The Lancet, 343, 1479–1481. Fox, A. W., & Davis, R. L. (1998). Migraine chronobiology. Headache, 38, 436–441. could, to some extent, be linked to a physiological need to increase Fruehwald-Schultes, B., Kern, W., Born, J., Fehm, H. L., & Peters, A. (2001). Hyperinsu- 5-HT synthesis and/or release. linemia causes activation of the hypothalamus–pituitary–adrenal axis in humans. It is important that the limitations of this study are addressed. International Journal of Obesity, 25, S38–S40. Goldstein, R. E., Wasserman, D. H., McGuinness, O. P., Lacy, D. B., Cherrington, A. D., & In particular, the number of participants while being satisfactory Abumrad, N. N. (1993). Effects of chronic elevation in plasma cortisol on hepatic from a statistical perspective is still relatively small when we carbohydrate metabolism. American Journal of Physiology, 264, E119–E127. consider that the synthesis of DHEAS (as already stated) is Guazzo, E. P., Kirkpatrick, P. J., Goodyer, I. M., Shiers, H. M., & Herbert, J. (1996). Cortisol, influenced by heredity factors (Baulieu, 1996). It is possible that in dehydroepiandrosterone (DHEA), and DHEA sulfate in the cerebrospinal fluid of man: relation to blood levels and the effects of age. Journal of Clinical Endocrinology the present study the difference in DHEAS observed between and Metabolism, 81, 3951–3960. groups may simply be due to sampling factors. Therefore, Hinkle, D. E, Wiersma, W., & Jurs, S. G. (2003). Applied statistics for the behavioral sciences some caution in the interpretation of the significantly lower level (4th ed.). Boston: Houghton Mifflin Company. Hofeldt, F. D. (1989). Reactive hypoglycemia. Endocrinology & Metabolism Clinics of of DHEAS in the migraine group is recommended. Another North America, 18, 185–201. consideration that should be noted is the impact that fasting Horner, H., Packan, D., & Sapolsky, R. (1990). Glucocorticoids inhibit glucose transport (alone) may have had on some data and the well accepted effect in cultured hippocampal neurons and glia. Neuroendocrinology, 52, 57–64. International Headache Society. (2004). The primary headaches. Cephalalgia, Suppl. of fasting on the development of headache in some migraineurs 1(6), 24–136. (e.g. Critchley, 1933). Jacome, D. E. (2001). Hypoglycemia rebound migraine. Headache, 41, 895–898. 588 A. Kokavec, S.J. Crebbin / Appetite 55 (2010) 582–588

Jakubowicz, D. J., Beer, N. A., Beer, R. M., & Nestler, J. E. (1995). Disparate effects of Roberts, H. J. (1967). Migraine and related vascular headaches due to diabetogenic weight reduction by diet on serum dehydroepiandrosterone-sulfate levels in hyperinsulinism. Observations on pathogenesis and rational treatment in 421 obese men and women. Journal of Clinical Endocrinology and Metabolism, 80, patients. Headache, 7, 41–62. 3373–3376. Sapolsky, R. M. (1986). Glucocorticoid toxicity in the hippocampus: reversal by Kaplan, L. A., & Pesce, A. J. (1996). Clinical Chemistry: Theory, Analysis, Correlation. St. supplementation with brain fuels. Journal of Neuroscience, 6, 2240–2246. Louis Mosby. Sapolsky, R. M., Packan, D. R., & Vale, W. W. (1988). Glucocorticoid toxicity in the Kazmierczak, H., & Doroszewska, G. (2001). Metabolic disorders in vertigo, tinnitus, hippocampus: in vitro demonstration. Brain Research, 453, 367–371. and hearing loss. International Tinnitus Journal, 7, 54–58. Small, M., Gray, C. E., Beastall, G. H., & MacCuish, A. C. (1989). Adrenal androgens and Lyons, P. M., & Truswell, A. S. (1988). Serotonin precursor inﬂuenced by type of insulin-dependent diabetes mellitus. Diabetes Research, 11, 93–95. carbohydrate meal in healthy adults. American Journal of Clinical Nutrition, 47, Smolders, I., Loo, J. V., Sarre, S., Ebinger, G., & Michotte, Y. (2001). Effects of dietary 433–439. sucrose on hippocampal serotonin release: microdialysis study in the freely-

Majewska, M. D. (1992). Neurosteroids: endogenous bimodal modulators of the GABAA moving rat. British Journal of Nutrition, 86, 151–155. receptor. Mechanism of action and physiological signiﬁcance. Progress in Neurobi- Solomon, G. D. (1992). Circadian rhythms and migraine. Cleveland Clinical Journal of ology, 38, 379–395. Medicine, 9, 326–329. Marsters, J. B., Mortimer, M. J., & Hay, K. M. (1986). Glucose and diet in the fasting Thomas, G., Frenoy, N., Legrain, S., Sebag-Lanoe, R., Baulieu, E. E., & Debuire, B. (1994). migraineur. Headache, 26, 243–247. Serum dehydroepiandrosterone sulfate levels as an individual marker. Journal of Martin, P. R., & Seneviratne, H. M. (1997). Effects of food deprivation and a stressor on Clinical Endocrinology & Metabolism, 79, 1273–1276. head pain. Health Psychology, 16, 310–318. Tombaugh, G. C., Yang, S. H., Swanson, R. A., & Sapolsky, R. M. (1992). Glucocorticoids McCarthy, L. C., Hosford, D. A., Riley, J. H., Bird, M. I., White, N. J., Hewett, D. R., et al. exacerbate hypoxic and hypoglycemic hippocampal injury in vitro: bio- (2001). Single-nucleotide polymorphism alleles in the insulin receptor gene are chemical correlates and a role for astrocytes. Journal of Neurochemistry, 59, associated with typical migraine. Genomics, 78, 135–149. 137–146. Mitchell, L. E., Sprecher, D. L., Borecki, I. B., Rice, T., & Laskaszewski, P. M. (1994). Uhart, M., Chong, R. Y., Oswald, L., Lin, P. I., & Wand, G. S. (2006). Gender differences in Evidence for an association between dehydroepiandrosterone sulfate and non- hypothalamic–pituitary–adrenal (HPA) axis reactivity. Psychoneuroendocrinology, fatal, premature myocardial infarction in males. Circulation, 89, 89–93. 31, 642–652. Monro, J., Carini, C., & Brostoff, J. (1984). Migraine is a food-allergic disease. Lancet, Vexiau, P., Legoff, B., & Cathelineau, G. (1983). Insulin and cortisol secretion during 2(8405), 719–721. OGTT in patients with reactive hypoglycaemia with or without clinical symptoms. Neuhauser, H., & Lempert, T. (2004). Vertigo and dizziness related to migraine: a Hormones & Metabolism Research, 15, 419–421. diagnostic challenge. Cephalalgia, 24, 83–91. Vielkind, U., Walencewicz, A., Levine, J. M., & Bohn, C. (1990). Type II glucocorticoid Orentreich, N., Brind, J. L., Rizer, R. L., & Vogelman, J. H. (1984). Age changes and sex receptors are expressed in oligodendrocytes and astrocytes. Journal of Neuroscience differences in serum dehydroepiandrosterone sulfate concentrations throughout Research, 27, 360–373. adulthood. Journal of Clinical Endocrinology and Metabolism, 59, 551–555. Virgin, C. E., Jr., Ha, T. P. T., Packan, D. R., Tombaugh, G. C., Yang, S. H., Horner, H. C., et al. Patacchioli, F. R., Monnazzi, P., Simeoni, S., De Filippis, S., Salvatori, E., Coloprisco, G., (1991). Glucocorticoids inhibit glucose transport and glutamate uptake in hippo- et al. (2006). Salivary cortisol, dehydroepiandrosterone-sulphate (DHEA-S) and campal astrocytes: implications for glucocorticoid neurotoxicity. Journal of Neu- testosterone in women with chronic migraine. Journal of Headache and Pain, 7, rochemistry, 57, 1422–1428. 90–94. Yamada, J., Sugimoto, Y., Kimura, I., Takeuchi, N., & Horisaka, K. (1989). Serotonin- Peroutka, S.J. (2002). Serum glucose regulation and headache. Headache, 42, 303–308. induced hypoglycemia and increased serum insulin levels in mice. Life Science, 45, Radat, F., Irachabal, S., Swendsen, J., & Henry, P. (2002). Analgesic abuse and psychiatric 1931–1936. comorbidity in headache patients. Encephaleology, 28, 466–471. Ziegler, D. K., Hassanein, R. S., Kodanaz, A., & Meek, J. C. (1979). Circadian rhythms of Rainero, I., Limone, P., Ferrero, M., Valfre, W., Pelissetto, C., Rubino, E., et al. (2005). plasma cortisol in migraine. Journal of Neurology, Neurosurgery & Psychiatry, 42, Insulin sensitivity is impaired in patients with migraine. Cephalalgia, 25, 593–597. 741–748.