Keywords
Migraine Disorders - Photophobia - Hyperacusis - Olfaction Disorders - Precipitating
Factors - Food - Sensation Disorders
Palavras-chave
Transtornos de Enxaqueca - Fotofobia - Hiperacusia - Transtornos do Olfato - Fatores
Desencadeantes - Alimentos - Transtornos de Sensação
INTRODUCTION
Migraine is a genetically-determined neurological disorder that predisposes individuals
to recurrent episodes of typically intense, pulsating, unilateral headache accompanied
by phonophobia, photophobia, and nausea.[1]
Various behavioral, nutritional, and emotional factors are perceived as triggers for
migraine attacks; however, there is limited compelling data to support these associations.[2] On the other hand, neurofunctional imaging studies[3] have demonstrated brain alterations that could lead to behavioral, dietary, and
mood changes in the prodromal phase, which precedes migraine headache attacks. Thus,
differentiating migraine triggers from prodromal symptoms remains a major challenge
in the neurological practice.
It is a fact that the perception of dietary triggers occurs in approximately 25% of
individuals with migraine.[2] There are several theories to explain this perception:
-
there is a craving associated with the prodrome, wherein brain changes lead individuals
to overconsume certain types of foods, especially sweets, with the ingestion of these
foods being caused by the migraine itself rather than being a true trigger;[4]
-
there is an increase in nitric oxide availability, for example, during watermelon
consumption, through activation of the L-arginine-nitric oxide pathway;[5]
-
there is an allergic mechanism mediated by immunoglobulins G (IgG) or E (IgE) against
certain foods;[6] and
-
the composition of oral and intestinal flora with nitrate, nitrite, and nitric oxide-reducing
bacteria could increase the likelihood of migraine.[7]
Despite the widely-acknowledged association between food and migraine, there is scant
literature on this topic, and, to date, there are no studies clinically characterizing
individuals who perceive dietary triggers. The aim of the present study was to explore
the clinical characteristics of migraine patients associated with food triggers, as
well as to identify predictors of dietary triggers.
METHODS
Study design and patients
The present was a prospective cross-sectional study with a group comparison. The study
population comprised a non-random, convenience sampling of the first 524 migraine
patients diagnosed by a neurologist according to the criteria of the third edition
of the International Classification of Headache Disorders (ICHD-3).[1] They were cared for at the Headache Outpatient Clinic of the Pontifícia Universidade
Católica do Paraná (PUCPR), in the city of Londrina, Brazil, and invited to participate
in this research from December 2018 to June 2023.
Inclusion and exclusion criteria
Patients aged 18 to 69 years with episodic or chronic migraine, with or without aura,
were included according to the diagnostic criteria of the ICHD-3.[1] Participants with severe and uncontrolled systemic and/or psychiatric diseases,
as well as pregnant women, were excluded from the study.
Data collection
The patients provided a comprehensive medical history to a headache specialist, who
recorded clinical, personal, and anthropometric data. Headache attacks were characterized
according to the diagnostic criteria of the ICHD-3,[1] obtaining information on the presence of aura, nausea, phonophobia, photophobia,
osmophobia, and presence of dietary triggers. After excluding the subjects with incomplete
clinical data for the present study (22 patients), only 502 patients comprised the
final sample. They were divided into two groups, according to the presence or absence
of dietary triggers.
We considered that the patient was on preventive treatment when continuously using,
for at least 30 consecutive days, topiramate, valproate, divalproex, amitriptyline,
nortriptyline, venlafaxine, propranolol, atenolol, metoprolol, or flunarizine. Medication
overuse was considered when the patient ingested a triptan, opioid, dihydroergotamine,
or combination analgesic for ≥ 10 days per month, or non-opioid analgesics or nonsteroidal
anti-inflammatory drugs for ≥ 15 days per month, according to ICHD-3 criteria.[1]
The patients also filled out self-administered validated questionnaires to assess
migraine-related disability through the Migraine Disability Assessment (MIDAS)[8] and migraine impact through the 6-Item Headache Impact Test (HIT-6).[9] The presence of allodynia was measured using the 12-Item Allodynia Symptom Checklist
(ASC-12),[10] anxiety symptoms were assessed using the State-Trait Anxiety Inventory (STAI 1 and
2),[11] and depression was measured using the Beck Depression Inventory (BDI).[12] Finally, the participants completed the hyperacusis scale.[13]
Food triggers
The participants were asked about their perception of whether any food triggered a
migraine attack. The interviewer read a list of foods, and the patient mentioned the
food they considered triggering. This list included the following foods: cheese, chocolate,
citrus fruits, red wine, white wine, beer, spirits, coffee, cold cuts, salami, monosodium
glutamate, milk, dairy products, soft drinks, ice cream, nuts, chestnuts, almonds,
sweetener, tomato, excessive carbohydrates, and fermented products (yogurts and similar
foods). Excess carbohydrates were considered when the patient mentioned consuming
foods high in carbohydrates, such as sweets and pasta, in greater quantities than
usual. Participants were also given the option to add other foods not mentioned in
the list.
Ethical aspects
The present study was approved by the Ethics in Research Involving Human Subjects
Committee at PUCPR, under protocol number 4.293.039, and the Presentation Certificate
to Ethics Assessment, under registry number 98316718.7.0000.0020. All participants
signed the informed consent form.
Statistical analysis
Once the information was organized in the database, we used the IBM SPSS Statistics
for Windows (IBM Corp., Armonk, NY, United States) software, version 28.0, for the
statistical analysis. The categorical data were expressed as absolute numbers and
percentages, and the continuous data, as mean and standard deviation values. The Chi-squared
test with Yates correction, the Fisher's exact test and the Mann–Whitney test for
differences between averages of unpaired samples were used, assuming a significance
level of 0.05. The Student's t-test was used to analyze the clinical assessment scales. Binary logistic regression
(enter method) was performed to identify variables associated with the perception
of food triggers in two models: model 1 included demographic characteristics combined
with clinical features of migraine that had a p-value < 0.1 in the univariate analyses; the second model included only these clinical
features.
RESULTS
Of the 502 migraine patients, 58.4% (293/502) reported that foods were triggers for
headache attacks, while 41.6% (209/502) did not have this trigger. There were no statistical
differences in terms of age, sex, ethnicity, and body mass index between patients
with and without food triggers. The sample data are shown in [Table 1].
Table 1
Sociodemographic characteristics of 502 patients with migraine according to the presence
or absence of food triggers
|
Characteristics
|
Comparison groups
|
p-value
|
|
With food trigger
(n = 293)
|
Without food trigger
(n = 209)
|
|
Sex: n (%)
|
Male
|
38 (13.0)
|
33 (15.8)
|
0.371
|
|
Female
|
255 (87.0)
|
176 (84.2)
|
|
Age (years)
|
Mean(± SD)
|
35(± 13)
|
34(± 12)
|
0.413
|
|
Range
|
18–65
|
18–69
|
|
Ethnicity: n (%)
|
Caucasian: n (%)
|
230 (78.5)
|
154 (73.7)
|
0.210
|
|
Non-Caucasian: n (%)
|
63 (21.5)
|
55 (26.3)
|
|
Body mass index: n (%)
|
Underweight
|
16 (5.5)
|
7 (3.3)
|
0.748
|
|
Healthy weight
|
128 (43.7)
|
93 (44.5)
|
|
Overweight
|
91 (31.0)
|
61 (29.2)
|
|
Obesity
|
51 (17.4)
|
42 (20.1)
|
|
No information
|
7 (2.4)
|
6 (2.9)
|
Abbreviation: SD, standard deviation.
Note: p-values based on the Chi-squared and Mann–Whitney tests.
Alcohol was the most frequently reported food trigger (44%; 129/293), followed by
chocolate (42%; 123/293), cheese (27.7%; 81/293), excessive carbohydrates (27.7%;
81/293), coffee (21.8%; 64/293), cold cuts (16.0%; 47/293), citrus fruits (11.9%;
35/293), monosodium glutamate (7.5%; 22/293), and sweeteners (7.2%; 21/293) ([Table 2]).
Table 2
Distribution of food triggers in 293 patients with migraine
|
Food triggers
|
Frequency
|
|
n
|
%
|
|
Alcohol
|
129
|
44.0
|
|
Chocolate
|
123
|
42.0
|
|
Cheese
|
81
|
27.7
|
|
Excess carbohydrates
|
81
|
27.7
|
|
Coffee
|
64
|
21.8
|
|
Cold cuts
|
47
|
16.0
|
|
Citrus fruits
|
35
|
11.9
|
|
Monosodium glutamate
|
22
|
7.5
|
|
Sweetener
|
21
|
7.2
|
|
Other foods
|
174
|
59.4
|
The clinical characteristics of migraine are presented in [Table 3]. Aura and prophylactic medication use were more frequent among patients with food
triggers compared to those without food triggers respectively: 137/293 (46.8%) versus
67/209 (32.1%) (p = 0.001) and 113/293 (38.6%) versus 60/209 (28.7%) (p = 0.022). There was no statistical difference in the prevalence of chronic migraine
or medication overuse between the groups. The age at migraine onset was lower (19 ± 10
years versus 21 ± 11 years; p = 0.006) and the duration of illness was longer (18 ± 14 years versus 13 ± 11 years;
p = 0.001) among individuals with food triggers compared to those without food triggers,
respectively. There was no difference in the number of headache days or days of disabling
pain. Photophobia and osmophobia were higher among individuals with food triggers
compared to those without food triggers respectively: 278/293 (94.9%) versus 184/209
(88%) (p = 0.005) and 228/293 (77.8%) versus 117/209 (56%) (p < 0.001).
Table 3
Clinical characteristics of migraine in patients with and without food triggers
|
Parameters
|
Comparison groups
|
p-value
|
|
With food trigger
(n = 293)
|
Without food trigger
(n = 209)
|
|
Classification: n (%)
|
Episodic
|
155 (52.9)
|
118 (56.5)
|
0.430
|
|
Chronic
|
138 (47.1)
|
91 (43.5)
|
|
Aura: n (%)
|
No
|
156 (53.2)
|
142 (67.9)
|
0.001
|
|
Yes
|
137 (46.8)
|
67 (32.1)
|
|
Prophylactic medication: n (%)
|
No
|
180 (61.4)
|
149 (71.3)
|
0.022
|
|
Yes
|
113 (38.6)
|
60 (28.7)
|
|
Medication overuse: n (%)
|
No
|
161 (54.9)
|
124 (59.3)
|
0.475
|
|
Yes
|
115 (39.3)
|
71 (34.0)
|
|
No information
|
17 (5.8)
|
14 (6.7)
|
|
Frequency of headaches (days)
|
Mean(± SD)
|
11(± 9)
|
11(± 10)
|
0.323
|
|
Range
|
0–31
|
0–39
|
|
Frequency of disability headache (days)
|
Mean(± SD)
|
4(± 5)
|
4(± 5)
|
0.389
|
|
Range
|
0–30
|
0–30
|
|
Age at onset of migraine (years)
|
Mean(± SD)
|
19(± 10)
|
21(± 11)
|
0.006
|
|
Range
|
3–60
|
4–61
|
|
Duration of migraine (years)
|
Mean(± SD)
|
18(± 14)
|
13(± 11)
|
0.001
|
|
Range
|
0–58
|
0–57
|
|
Phonophobia: n (%)
|
No
|
35 (11.9)
|
37 (17.7)
|
0.070
|
|
Yes
|
258 (88.1)
|
172 (82.3)
|
|
Photophobia: n (%)
|
No
|
15 (5.1)
|
25 (12.0)
|
0.005
|
|
Yes
|
278 (94.9)
|
184 (88.0)
|
|
Osmophobia: n (%)
|
No
|
65 (22.2)
|
92 (44.0)
|
< 0.001
|
|
Yes
|
228 (77.8)
|
117 (56.0)
|
Abbreviation: SD, standard deviation.
Note: p-values based on the Chi-squared and Mann–Whitney tests.
Patients with food triggers had a higher mean migraine impact score (64 ± 6) compared
to those without food triggers (62 ± 9) (p = 0.002). There were no differences in terms of anxiety, depression, disability,
allodynia, or hyperacusis scores (p > 0.05) ([Table 4]).
Table 4
Analysis of scales validated in individuals with migraine according to the presence
of a food trigger
|
Questionnaires and scores
|
Comparison groups
|
p-value
|
|
With food trigger
(n = 293)
|
Without food trigger
(n = 209)
|
|
State-Trait Anxiety Inventory (STAI 1): mean ± SD
|
45(± 10)
|
44(± 12)
|
0.618
|
|
Range
|
24–69
|
20–80
|
|
|
State-Trait Anxiety Inventory (STAI 2): mean ± SD
|
45(± 11)
|
45(± 11)
|
0.696
|
|
Range
|
22–98
|
22–79
|
|
|
Beck Depression Inventory (BDI): mean ± SD
|
13(± 11)
|
12(± 9)
|
0.881
|
|
Range
|
0–130
|
0–47
|
|
|
Migraine Disability Assessment (MIDAS): mean ± SD
|
34(± 40)
|
29(± 40)
|
0.175
|
|
Range
|
0–330
|
0–314
|
|
|
6-Item Headache Impact Test (HIT 6): mean ± SD
|
64(± 6)
|
62(± 9)
|
0.002
|
|
Range
|
47–76
|
5–78
|
|
|
12-Item Allodynia Symptom Checklist (ASC-12): mean ± SD
|
6(± 4)
|
6(± 4)
|
0.08
|
|
Range
|
0–18
|
0–22
|
|
|
Hyperacusis Scale: mean ± SD
|
20(± 10)
|
19(± 11)
|
0.140
|
|
Range
|
0–42
|
0–42
|
|
Abbreviation: SD, standard deviation.
Note: p-value based on the Student's t-test.
The binary logistic regression aimed at identifying clinical predictors of food triggers
is presented in 2 models in [Table 5]. Increasing age was associated with protection, while longer migraine duration,
presence of photophobia and osmophobia, and higher migraine impact were associated
with increased odds of food triggers. Notably, the presence of photophobia was associated
with 3-fold higher odds of presenting food triggers (odds ratio [OR] = 3.31; 95% confidence
interval [95%CI] = 1.26–8.66; p = 0.015 in model 1, and OR = 3.22; 95%CI = 1.26–8.18; p = 0.014 in model 2), and osmophobia was associated with 2-fold higher odds of presenting
food triggers (OR = 2.64; 95%CI = 1.58–4.41; p < 0.001 in model 1, and OR = 2.45; 95%CI = 1.50–4.02; p < 0.001 in model 2).
Table 5
Analysis of food trigger predictors using binary logistic regression
|
Models
|
B
|
SE
|
Wald
|
df
|
OR (95%CI)
|
p-value
|
|
Model 1
|
|
|
|
|
|
|
|
Age
|
−0.085
|
0.042
|
4.048
|
1
|
0.919 (0.846–0.998)
|
0.044
|
|
Sex
|
−0.045
|
0.355
|
0.016
|
1
|
0.956 (0.476–1.918)
|
0.899
|
|
Ethnicity
|
−0.528
|
0.290
|
3.312
|
1
|
0.590 (0.334–1.041)
|
0.069
|
|
BMI
|
Overweight
|
0.124
|
0.270
|
0.211
|
1
|
1.132 (0.667-1.919)
|
0.646
|
|
Obesity
|
0.113
|
0.327
|
0.120
|
1
|
1.120 (0.590-2.123)
|
0.730
|
|
Underweight
|
0.695
|
0.569
|
1.496
|
1
|
2.005 (0.658-6.110)
|
0.221
|
|
Aura
|
0.296
|
0.240
|
1.517
|
1
|
1.344 (0.839–2.154)
|
0.218
|
|
Prophylactic Medication
|
0.261
|
0.253
|
1.067
|
1
|
1.299 (0.791–2.132)
|
0.302
|
|
Age at onset of migraine
|
0.068
|
0.043
|
2.503
|
1
|
1.070 (0.984–1.164)
|
0.114
|
|
Duration of migraine
|
0.097
|
0.040
|
5.755
|
1
|
1.102 (1.018–1.193)
|
0.016
|
|
Photophobia
|
1.196
|
0.492
|
5.916
|
1
|
3.306 (1.261–8.663)
|
0.015
|
|
Phonophobia
|
−0.082
|
0.337
|
0.059
|
1
|
0.921 (0.475–1.784)
|
0.808
|
|
Osmophobia
|
0.970
|
0.262
|
13.700
|
1
|
2.638 (1.578–4.410)
|
< 0.001
|
|
HIT-6
|
0.035
|
0.017
|
4.014
|
1
|
1.036 (1.001–1.072)
|
0.045
|
|
ASC-12
|
−0.017
|
0.031
|
0.296
|
1
|
0.983 (0.926–1.045)
|
0.586
|
|
Model 2
|
|
|
|
|
|
|
|
Aura
|
0.271
|
0.232
|
1.368
|
1
|
1.311 (0.833–2.064)
|
0.242
|
|
Prophylactic Medication
|
0.331
|
0.247
|
1.801
|
1
|
1.392 (0.859–2.258)
|
0.180
|
|
Age at onset of migraine
|
−0.015
|
0.012
|
1.593
|
1
|
0.985 (0.963–1.008)
|
0.207
|
|
Duration of migraine
|
0.022
|
0.010
|
4.833
|
1
|
1.022 (1.002–1.042)
|
0.028
|
|
Photophobia
|
1.168
|
0.476
|
6.021
|
1
|
3.216 (1.265–8.178)
|
0.014
|
|
Phonophobia
|
−0.080
|
0.328
|
0.059
|
1
|
0.923 (0.486–1.756)
|
0.808
|
|
Osmophobia
|
0.898
|
0.252
|
12.665
|
1
|
2.454 (1.497–4.023)
|
< 0.001
|
|
HIT-6
|
0.037
|
0.017
|
4.693
|
1
|
1.038 (1.004–1.073)
|
0.030
|
|
ASC-12
|
−0.022
|
0.030
|
0.543
|
1
|
0.978 (0.923–1.037)
|
0.461
|
Abbreviations: 95%CI, 95% confidence interval; ASC-12, 12-item Allodynia Symptom Checklist;
B, Beta coefficient; BMI, body mass index; df, degrees of freedom; HIT-6, 6-Item Headache
Impact Test; OR, odds ratio; SE, standard error; Wald, Wald statistic.
DISCUSSION
The main finding of the present study was the understanding that migraine patients
who perceive food triggers are clinically different from those without food triggers,
especially concerning sensory symptoms such as photophobia and osmophobia.
Migraine is characterized by headache accompanied by sensory hypersensitivity of different
modalities, such as visual (photophobia), auditory (phonophobia), olfactory (osmophobia),
and somatosensory (allodynia), with neurofunctional studies[14] demonstrating that these symptoms may even precede pain. In the present study, patients
with photophobia and osmophobia had a higher chance of presenting with food triggers.
One hypothesis for this finding could be that, in at least a subset of patients, the
perception of food triggers may result from a sensory stimulus, especially osmophobia,
since the interpretation of food flavor occurs through the combination of gustatory
and olfactory stimuli. This hypothesis is supported by a study[15] describing various odors capable of triggering headaches, with 20.5% of individuals
with migraine perceiving the odor of culinary products as a trigger for migraine attacks.
Each year lived with migraine increases the odds of perceiving food triggers by 1.1
time, which may occur due to the greater number of painful experiences over time,
but there are studies[16] demonstrating that the time lived with migraine can modify olfactory perception,
with a positive correlation between interictal hypersensitivity and disease duration.
The same thing occurs with allodynia, the presence and severity of which are associated
with disease duration.[17] In the present study, the impact of migraine assessed by the self-administered HIT-6
scale was also associated with a slight increase in the odds of presenting food triggers.
Limitations
The strength of the present study, which makes it innovative, was the comparison of
migraine patients in the presence and absence of food triggers, using a relevant sample
and controlling for confounding variables through regression analysis. However, some
inherent limitations to its design were observed, thus precluding the inference of
causality between the associations. Patients from a single center were included, and
the sample was heterogeneous regarding treatment and medication overuse, which may
have altered the perception of food triggers. Additionally, the definition of food
triggers was based solely on the patient's perception. Therefore, it was not possible
to reliably differentiate food triggers from premonitory symptoms of migraine, such
as changes in appetite. The validated scales had a considerable amount of missing
data and may have failed to show significance due to the smaller valid sample size.
In conclusion, the presence of food triggers was significantly associated with photophobia
and osmophobia, suggesting that, at least in some patients, the presence of food triggers
may be due to sensory hypersensitivity, especially to osmophobia, since olfactory
stimulation is inherent to the act of eating. Osmophobia might be another mechanism
by which patients perceive foods as triggers for their migraine attacks. Finally,
there is a need for interventions and prospective studies to determine this association.
Bibliographical Record
Aline Vitali-Silva, Valéria Aparecida Bello, Regina Célia Poli-Frederico, Carlos Eduardo
Coral de Oliveira, Edna Maria Vissoci Reiche, Beatriz Bagatim Bossa, Debora Villas
Boas Rezende, Bárbara Ferreira Khouri, Raimundo Pereira Silva-Néto. Relationship between
food triggers and sensory hypersensitivity in patients with migraine. Arq Neuropsiquiatr
2024; 82: s00441793934.
DOI: 10.1055/s-0044-1793934
