The Role of Diet and Nutrition in Migraine Triggers and Treatment

A Systematic Literature Review

Nada Ahmad Hindiyeh, MD; Niushen Zhang, MD; Mallory Farrar, PharmD; Pixy Banerjee, MPharm; Louise Lombard, MNutr; Sheena K. Aurora, MD


Headache. 2020;60(7):1300-1316. 

In This Article


Study Selection

A total of 1601 articles were identified, of which 320 were selected for full-text screening based on the inclusion criteria. Forty-three articles were finally included in this systematic literature review (Figure 1). Of the studies included in the review, 11 outlined the evidence on diet patterns, 20 outlined the evidence on diet-related triggers, and 12 outlined evidence on diet as interventions in people with migraine.

Figure 1.

PRISMA diagram. [Color figure can be viewed at]

Study Characteristics and Finding

Majority of the included studies (81.4%) were observational studies. Quality appraisals of RCT and observational studies are presented in Supplementary Table S1, Supplementary Table S2 and Supplementary Table S3 and explained separately in the sections for diet patterns, diet-related triggers, and dietary interventions. All the RCT studies in the review (n = 8) assessed the effect of diet interventions.

Diet Patterns

Study characteristics and a narrative summary of the results of studies evaluating diet patterns are summarized in Table 2. Diet patterns are features of a patient's diet that are observed either more or less frequently in people with migraine compared to those without. Diet patterns include multiple factors associated with diet in people with migraine. The overall quality of studies assessing diet patterns was low and it was hard to generalize a complete consensus. Most of the studies were completed at 1 site and none were rigorous RCT or prospective observational surveys. Most were point-in-time surveys and were, therefore, susceptible to recall bias.

General Diet Patterns. There was conflicting evidence regarding the general diet patterns in people with migraine. The Prospective Analysis of Factors Related to Migraine Attacks Swiss study conducted in 327 patients stated that limited evidence exists on the effect of nutrition in the precipitation of migraine. No unfavorable impact of any nutritional factor was noted.[21] A daily diary data analysis from a National Institutes of Health study reported that night time snacking and eating a late dinner reduced the odds of headache by 40% (hazard ratio [HR]: 0.60; 95% confidence interval [CI]: 0.40, 0.90; P = .013) and 21% (HR: 0.79; 95% CI: 0.55, 1.15; P = .22), respectively, in people with migraine.[22] A case-control study conducted in Iran observed that females with migraine were more likely than females without migraine to report no regular diet program (37.6% vs 17.6%; P = .004), irregular schedule of meals (37.6% vs 23.5%; P = .046), and less than 3 meals per day (29.4% vs 9.4%; P = .001).[23]

Several cross-sectional studies and surveys also reported the association of diet patterns with migraine. In a United States cross-sectional study, people with migraine with aura were reported to be more likely to have a low intake of chocolate (P = .005), cheese (P = 0.008), ice cream (P = .003), hot dogs (P < .001) and processed meats (P = .009), as compared to those with migraine without aura.[24] In a Swedish population-based survey, an increased prevalence of migraine was observed among those who skipped breakfast.[25] The National Health and Nutrition Examination Survey (NHANES) reported that the diet quality was significantly higher in women of normal weight (body mass index [BMI]: 18.5–24.9 kg/m2) without migraine compared to women of normal weight with migraine (P <. 0001).[26]

Specific Diet Patterns. Unlike general diet patterns, specific diet patterns showed a more consistent relationship with migraine. Smoking and alcohol consumption had an association with migraine, with a retrospective analysis of annual health survey data reporting that the correlation of migraine prevalence with daily smoking was positive (Spearman co-efficient, rs = 0.49) and the correlation with alcohol consumption was negative (rs = −0.52).[27] The Prospective Analysis of Factors Related to Migraine Attacks study reported that the risk of headache and migraine, as well as the risk of headache persistence, was reduced by the consumption of beer on days before headache onset.[21]

A cohort study conducted in Rome reported that increased consumption of whole-grain bread (P = .04) and whole grain pasta (P = .004), and decreased consumption of white bread (P = .004) was associated with a statistically significant reduction in migraine attack frequency and pharmacological rescue drug usage per month.[28]

Several cross-sectional and survey studies also reported correlation between specific diet patterns and migraine. A study reported a significant positive association of consumption of fried food with migraine headache.[29] Similarly, another study showed that the migraine attack frequency is inversely proportional to adherence toward "healthy" eating pattern (high consumption of fruits, fish, vegetable pickles, vegetables, and legumes) (P for trend = .04) and directly proportional to adherence toward "western" eating pattern (high consumption of cola, salted nuts, processed meat, and fast foods) (P for trend = .02).[30] An inverse relationship between dietary sodium intake and the odds of probable migraine history was reported; however, in women, the relationship was limited to those with lower BMI (P = .007).[31]

The association of alcohol intake and migraine varied across several cross-sectional studies. A cross-sectional study conducted in the United States reported that people with migraine had a low intake of total alcohol (P < .001) compared to those without any headache history.[24] The NHANES cross-sectional survey and Swedish population-based survey reported a positive association of alcohol consumption and migraine in women;[25,26] however, no association between heavy alcohol use and migraine was noted in men.[25]

Diet-related Triggers

Study characteristics and a narrative summary of the results of studies evaluating diet-related triggers are summarized in Table 3. Diet-related triggers for migraine identified included alcohol, caffeine, fasting, and a wide variety of specific foods. No RCT studies assessing diet as triggers were identified. Comparison between studies is limited, as some studies focused on 1 specific trigger (such as alcohol), while others assessed a variety of triggers. In addition, some studies presented a list to patients to assess triggers vs patients recalling triggers on their own. No study involved patient or assessor blinding. Due to the survey or questionnaire nature of these studies, patients were susceptible to recall bias. Most of the studies were of poor-to-medium quality due to the lack of complete reporting of study design or patient characteristics (Supplementary Table S2).

Alcohol. Alcohol was reported to be significantly associated with migraine compared to nonmigraine headaches in a Korean prospective observational study (odds ratio [OR] = 2.5, 95% CI = 1.3, 5.0; P < .001).[14] Another prospective cohort study reported alcohol to be associated with migraine in 3.9% (N = 126) patients.[32]

Several cross-sectional studies reported the association of alcohol with migraine. Alcohol use was found to be associated with migraine in 17.5%, 20.5% and 35.6% participants in 3 questionnaire studies.[33–35] One study reported red wine (77.8%) as the most common trigger among alcoholic beverages.[33] Alcohol was found to be associated with migraine in a Chinese cross-sectional study.[36] However, the association was less likely to be present in women compared to men.[36] Also, a Japanese population-based survey reported no association of alcohol and migraine after age and gender adjustment,[37] and a cross-sectional study reported that alcohol was associated with migraine in only a small percentage of patients.[38]

Fasting. A cohort cross-over migraine diary study showed that during the month of Ramadan, there was an increase in the average number of migraine days among observant people with migraine compared to the following (control) month (9.4 ± 4.3 vs 3.7 ± 2.1; P < .001).[39]

This finding was supported by a patient survey conducted in Iraq which reported that fasting in the month of Ramadan is a trigger for headaches in people with migraine.[40] Another prospective questionnaire study also reported fasting as a major trigger among people with migraine who reported any triggers.[12]

Food Allergens. One study assessed the effect of food allergens on migraine. The prospective observational study conducted in Turkey showed that food allergens, such as fish, egg whites, egg yolks, nuts, orange, and strawberry, were not related to migraine attacks; however, people with migraine were found to have more pollen allergies compared to the people without migraine.[41]

Caffeine. A prospective cohort study conducted in Turkey (N = 126) revealed that caffeine was a trigger in 6.3% people with migraine.32 Caffeine was found to be a a trigger for migraine-associated vertigo attacks in 69.6% of patients in a Turkish retrospective, observational study.[42]

The association between caffeine and migraine was also reported in another prospective observational, cross-sectional study.[43]

Dietary Triggers. Dietary factors (43.6%), including hunger (53.9%) and consumption of milk and cheese (10.3%) and chocolate (18.3%), were found to be associated with migraine in a prospective cohort study conducted in Turkey. However, no statistical difference in dietary triggers was noted between migraine with and without aura (P = .753) or between genders (P = .081).[32] In another Korean prospective, observational study, overeating was significantly associated with migraine compared to nonmigraine headaches (OR = 2.4; 95% CI = 1.1, 5.7; P = .001).[14] A Turkish prospective cohort study included 23 people with migraine-associated vertigo. The most common triggers that may have triggered vertigo attacks included cheese or cheese products (100%) and excessive intake of nuts (56.5%), fresh or dry fruits (39.1%), dairy products (39.1%), processed food (30.4%), baked yeast foods (21.7%), and processed meat (21.7%).[42]

A retrospective, observational study assessing the triggers or precipitants of migraine attacks reported the frequency of food triggers in people with migraine to be 26.9%. Food was identified as one of the very frequently occurring (>66% of headaches) triggers. Food was a more common trigger in people with migraine vs those with probable migraine (P = .017), in episodic migraine compared to chronic migraine (P = .025) and in migraine with aura compared to migraine without aura (P = .010).[44]

Several cross-sectional studies and questionnaire surveys also reported the association of dietary factors with migraine. In a questionnaire survey assessing the effect of diet-related triggers, the number of people with at least 1 trigger was significantly higher in people with migraine attacks than those with no attacks within the last year (P < .001).[45] A cross-sectional survey reported that migraine was commonly associated with dietary triggers, with only 2.4% of people with migraine (N = 123) not experiencing susceptibility to any dietary trigger.[46] In an Italian cross-sectional study, people with migraine were reported to be better in recognizing triggers, such as particular foods and alcohol.[47] A questionnaire survey reported food and seasonings to be associated with migraine.[35] Another questionnaire survey reported missing meals and the use of certain foods including chocolate, cheese and hot dogs to be associated with migraine.[34] Chocolate and foods rich in monosodium glutamate were reported to be the most common dietary factors associated with migraine in a prospective, cross-sectional study.[43] Chocolate was found to be significantly associated with migraine compared to tension-type headache.[43] A Croatian population-based survey reported a significant positive association of various food items (chocolate, cheese, alcoholic drinks, fried fatty foods, vegetables, and coffee) with migraine with aura compared to migraine without aura.[48]

Diet Interventions

Study characteristics and a narrative summary of the results of studies evaluating diet interventions are summarized in Table 4. Diet interventions refer to the adjustment of a patient's diet by adding or eliminating a specific type of food that might influence the frequency and severity of migraine attacks. The quality of studies included was higher for diet interventions and included 8 RCTs which mostly had a low risk of bias in the quality assessment, allowing for possible claims of causation.

Specific Diets. In a randomized, controlled study, in the first month after restriction to low glycemic index diet, monthly attack frequency significantly decreased from baseline in both diet and medication (control) groups (P < 0.05). The mean frequency and severity of attacks as measured by the visual analog scale (VAS) decreased significantly after 3 months in the diet group compared with those in the medication group (P < .05).[50] In another randomized, cross-over dietary interventional trial, a low-lipid (lipid content <20% of the total daily energy intake) or a normal-lipid (between 25% and 30% of the total daily energy intake) diet was assigned randomly for 3 months and then the diets were crossed over for the following 3 months. A significant correlation between low-lipid diet and decrease in migraine attacks (2.9 ± 3.7; P < .001 vs baseline and P < .05 vs normal-lipid diet) was established. The low-lipid diet was effective in reducing the mean (± standard deviation [SD]) severity of attacks (1.7 ± 0.9 vs 1.2 ± 0.9, P = .001) and the number of severe pain attacks (1.8 ± 1.6 vs 0.4 ± 1.3, P = .01) vs the normal-lipid diet.[51]

Another randomized, cross-over study compared 2 treatments: dietary instruction with a placebo supplement. Each treatment period was 16 weeks, with a 4-week washout period before the cross-over to the alternate treatment. During the diet period, patients were prescribed a low-fat vegan diet for 4 weeks, after which they were asked to follow an elimination diet to identify possible specific migraine -triggering foods. During the elimination diet period, patients continued the low-fat vegan diet along with the elimination of common trigger foods, chosen based on previous studies. The elimination diet was continued until no further improvement was noted or until the midpoint of the period (typically 10 to 21 days), after which the omitted foods were reintroduced one at a time. Improvement in headache pain, as measured by the patient's global impression of change and change in pain question scale (5-point Likert-style scale ranging from "much worse" to "much better"), was significantly greater after the diet period (P < .001). Pain relief medication use decreased significantly during the diet period compared to the placebo supplement period (19 vs 3 absolute percentage point decrease in medicated headaches, P = .004). Improvement in average headache intensity and average headache frequency was not significantly higher in the diet period compared to the supplement period (P = .20 and P = .61, respectively).[52]

An association of water intake on migraine was reported in an RCT, with an observed reduction of headache hours and headache intensity with higher water intake (1.5 liters) vs control (no water intake recommendations, continued normal water intake); however, the effects were not statistically significant. Headaches reduced by 21 hours (95% CI: −48, 5) within 2 weeks in the higher water intake group compared to the control group. The mean headache intensity was measured using the visual analog scale (VAS; 0–100 mm scale, higher scores indicating severe headache). The observed difference in mean improvement in headache intensity at 12 weeks was 13 mm (95% CI: −32, 5) on the VAS in the group with high water intake vs the control group.[53]

Ketogenic diet administration for 1 month was also significantly related to the reduction in the mean attack frequency and duration compared to baseline (all P < .001) as shown in a small, prospective observational study.[54]

Evidence of the association between specific diets and migraine was noted in cross-sectional studies. Greater intake of omega-3 polyunsaturated fatty acids was found to be statistically significantly associated with lower prevalence of severe headache or migraine.[55] Also, highest adherence to dietary approaches to stop hypertension (DASH) diet displayed a 30% lower prevalence in severe headaches compared to those with the lowest adherence.[56]

Elimination Diets. The importance of an Immunoglobulin G (IgG)-based elimination diet was shown in a double-blind, randomized, controlled, cross-over study which included baseline (usual diet) for 6 weeks, first diet (elimination [IgG-negative food] or provocation [IgG-positive food] diets) phase of 6 weeks, and second diet (interchange of elimination or provocation diets) phase of 6 weeks (with a 3-week washout phase with usual diet at the end of the first diet phase). There was a significant reduction with the elimination diet when compared to the baseline levels in the mean (±SD) number of attacks (4.8 ± 2.1 vs 2.7 ± 2.0; P < .001), maximum attack duration (2.6 ± 0.6 vs 1.4 ± 1.1 days; P < .001), mean attack duration (1.8 ± 0.5 vs 1.1 ± 0.8 days; P < .01), maximum attack severity (VAS 8.5 ± 1.4 mm vs VAS 6.6 ± 3.3 mm; P < .001) and number of attacks with acute medication (4.0 ± 1.5 vs 1.9 ± 1.8; P < .001).[57]

In a single-blind, randomized clinical trial, a significant difference in the median number of migraine-like headache days at 4 weeks (incident rate ratio [IRR] 1.23, 95% CI 1.01, 1.50; P = .04) was observed between true diet and sham diet, which included IgG antibody reactivity-related elimination diet (identified using enzyme-linked immunosorbent assay). However, the difference in the median number of migraine-like headache days over 12 weeks was not significantly different between the true and sham diet groups (IRR 1.15, 95% CI 0.94, 1.41; P = .18).[58] Another double-blind, cross-over study evaluated the importance of elimination of foods with a high IgG antibody level (elimination diet) by showing a significant reduction in the attack count, number of attacks with acute medication, and total medication intake with elimination diet compared to provocation diet that consisted of food with high IgG antibody levels (P = .006).[59]

Another randomized, cross-over study showed the importance of diet implementation (restriction of triggers like wheat, orange, egg, caffeine, cheese, chocolate, and milk) in people with migraine. The migraine-triggering foods were excluded from the diet of 2 groups of patients and then the diet restriction was relaxed in group 1 after the second month and continued in group 2. Assessments were made before the start of diet restriction, and at 2 and 4 months. Monthly attack frequency, attack duration, and attack severity were found to decrease significantly after 2 months of diet implementation compared to the period before diet implementation in group 1 (P = .011, P = .041, and P = .003, respectively) and group 2 (P = .015, P = .037, and P = .003, respectively). In the 4th month evaluation, the significant decrease was maintained only in patients who continued the diet restriction (group 2) (P < .05).[60]

In another prospective cohort study, the specific elimination diet (elimination of IgG-positive food) for 1 to 6 months was shown to be effective in 56 people with migraine who were positive on immunoassay for food IgG. Remission (no migraine) was achieved in 43 patients; the intensity and frequency were decreased in 4 patients, while 9 patients had no changes. The IgG food reactivity was significantly different between people with migraine and the control group (people without migraine) (P < .01), suggesting people with migraine may be more sensitive to IgG food reactivity.[61]