Migraine and the Environment


  • Deborah I. Friedman MD, MPH,

    1. From the Department of Ophthalmology and Neurology, University of Rochester School of Medicine and Dentistry, Rochester, NY (D.I. Friedman); Department of Public Health and Preventive Medicine, SUNY Upstate Medical University, Syracuse, NY, USA (T. De Ver Dye).
    Search for more papers by this author
  • Timothy De Ver Dye PhD

    1. From the Department of Ophthalmology and Neurology, University of Rochester School of Medicine and Dentistry, Rochester, NY (D.I. Friedman); Department of Public Health and Preventive Medicine, SUNY Upstate Medical University, Syracuse, NY, USA (T. De Ver Dye).
    Search for more papers by this author

  • This work was supported by the National Eye Institute 1K23EY015525 and a Research to Prevent Blindness Challenge Grant. The author has no conflicts of interest relevant to the content of this review.

D.I. Friedman, 601 Elmwood Avenue, Box 659, Rochester, NY 14642, USA.


Migraineurs often describe environmental triggers of their headaches, such as barometric pressure change, bright sunlight, flickering lights, air quality, and odors. Environmental aspects of indoor space and workplaces are also implicated in migraine experience. Comprehensive migraine treatment programs emphasize awareness and avoidance of trigger factors as part of the therapeutic regimen. As migraine has a substantial economic impact, remediation of correctable environmental triggers may benefit employee attendance and productivity among migraineurs. Few controlled studies in the literature, however, confirm environmental influences on migraine and headaches. Although some are controversial, migraineurs worldwide consistently report similar environmental triggers. This article addresses commonly mentioned environmental triggers with a discussion of their pathophysiology and proposed preventive measures.

Objective: To examine the epidemiological evidence for commonly-mentioned environmental migraine triggers, discuss their possible role in the pathophysiology of migraine and propose preventive measures to avoid or minimize exposure.

Background: Migraineurs often describe environmental triggers of their headaches, such as barometric pressure change, bright sunlight, flickering lights, air quality and odors. Environmental aspects of indoor space and workplaces are also implicated in the migraine experience. As migraine has a substantial economic impact, remediation of correctable environmental triggers may improve attendance and productivity among migraineurs in the workplace.

Methods: We reviewed the literature addressing indoor and outdoor environmental factors which are commonly implicated as migraine triggers.

Results: Although some factors are controversial, migraineurs worldwide consistently report similar environmental triggers. However, few studies confirm environmental influences on migraine and headaches. Research to date indicates that migraineurs have lower thresholds for light-induced discomfort, sine grating distortion and illusions, noise tolerance and olfactory sensitivity compared to the general population.

Conclusion: There are conflicting studies supporting the validity of patient-reported environmental migraine triggers. Prospective studies are needed to determine the extent that external stimuli influence the migraine process. Decreased thesholds for light, noise, olfactory and visual stimuli in migraineurs may be minimized by modifying the work, home and classroom settings.

Migraine is a common problem worldwide with significant morbidity and economic impact.1 The direct costs of migraine are directly related to the severity of migraine pain and disability, and rise dramatically with prescription medication usage.2,3 The indirect costs exceed the costs of medical care, however, and work-related disability is the most important determinant of the economic impact of migraine.4 Migraineurs often miss work (absenteeism) or have reduced productivity at work (presenteeism). A study from 1999 estimated that migraine costs American employers over $13 billion yearly.2

Epidemiologic data suggest that successful therapy of the most severely affected migraineurs may significantly impact the overall economic burden of migraine.5 Migraine therapy employs preventive and symptomatic measures with pharmacologic and non-pharmacologic treatments are often used in both strategies. With careful examination of headache diaries and lifestyle influences, approximately 50-75% of migraineurs are able to identify factors that provoke their headaches.6,7 Awareness and avoidance of specific migraine triggers are incorporated into the treatment strategy to decrease the frequency of migraine in a given individual. Triggers for migraine include various foods and beverages, stress or relief of stress, and hormonal factors (such as menstruation and pregnancy).6,8,9

Additionally, many migraine sufferers relate that various environmental conditions trigger their headaches, most commonly changes in the weather, exposure to bright lights, high altitude, smoke, and certain odors.9,10 Patients frequently claim that they are “better than the weatherman” in predicting the weather, particularly impending rain or snowstorms.11 Some authors postulate that migraine developed as an evolutionary defense mechanism.12,13 For example, if weather is truly a migraine trigger, it may be an evolutionary advantage to be able to predict bad storms.13 Likewise, the heightened sensitivity to odors may serve as a warning signal for various environmental toxins. Proponents of this theory theorize that the preponderance of migraine in women is a remnant of their role in the hunter-gatherer society and their need for greater biological fitness to gather foodstuffs and bear children.

The indoor environment, particularly at the work place, also contributes to the migraine disability. Migraineurs may request medical documentation for accommodations at work or school to avoid the strong perfumes of co-workers, the flicker of the computer screen or bright fluorescent lighting in the classroom. The Job Accommodation Network's Searchable Online Accommodation Resource (SOAR), a website maintained by the US Department of Labor, recommends modification of lighting triggers, noise triggers, smell/fragrance triggers, and other aspects of the work site for employees with migraine headaches. Examples include: add fluorescent light filters to existing fluorescent lights to create a more natural lighting, change lighting completely, provide an anti-glare filter for computer monitor, provide a liquid crystal display monitor that has a better refresh rate, move employee to a more private area or away from high traffic areas, provide sound absorption panels, implement a fragrance-free policy, provide flexible leave when the employee is experiencing a migraine, and provide the employee with a dark, private area to go to when experiencing a migraine.14

This paper reviews the relationship of migraine to aspects of the outdoor and indoor environments. The pathophysiology of migraine triggers apropos various triggers and sensitivities is discussed.


Studies of various populations worldwide have investigated the link between migraine and the weather. The 2 methods of assessing migraine triggers are self-report and using diary information. Thirty-eight patients with migraine and 17 patients with tension-type headache meeting the International Headache Society (IHS) classification criteria were interviewed about migraine triggers using a standard questionnaire.15 Weather was one of the most frequent precipitating factors for migraine (71% of patients) and helped to differentiate migraine from tension-type headache (35%). Further details regarding weather conditions were not elucidated. Another survey of 115 female Mexican-American migraineurs between ages 15-45 years living in San Diego county found weather changes to be a precipitating factor in 54.4% (odds ratio 1.8).16 One hundred and seventy-two consecutive migraine patients, 52 patients with tension-type headaches, and 53 patients with mixed (migraine plus tension-type) headaches at a Pittsburgh headache clinic were queried on the frequency of various purported migraine triggers. Weather changes were endorsed by 45.5% of migraineurs, 48% of subjects with tension-type headaches, and 42% of those with mixed headaches. In contrast to the previous study, weather changes did not discern between migraine and tension-type headaches. A cross-sectional population study of 1400 schoolchildren ages 6 to 14 years in the United Arab Emirates surveyed 159 children who met the diagnostic criteria for migraine (78 boys and 83 girls).17 Hot climate was cited as a trigger by 59 children.

A cross-sectional study interviewing 1064 German citizens and 1506 Canadian citizens revealed that 19.2% of Germans thought that the weather affected their health “to a strong degree”, and 35.2% felt that the weather had “some influence on their health.”18 Headache and migraine were the most frequent weather-related symptoms experienced (61%). Sixty-one percent of the Canadian subjects considered themselves weather-sensitive but migraine and headaches were not commonly reported (<10%).

A prospective cohort study in the United States using a database of 1207 patients meeting IHS criteria for migraine identified weather as the fourth most frequent trigger, occurring in 53.2% of patients.7 Thirty percent of patients cited heat as a trigger. A cross-sectional study in Vienna, Austria used semi-structured interviews identified weather as the most common trigger, occurring in 82.5% of patients with migraine or tension-type headache.19

Despite the relatively high percentage of migraineurs attributing their headaches to weather changes, several studies suggest that the perception of weather as a trigger may be overestimated. Studies incorporating daily headache diaries have explored various environmental factors as headache triggers (Table 1). One of the earliest studies reviewed the headache data from 60 headache patients and compared it with meteorological statistics for the workday of the headache and the following day.20 A significant correlation existed between people who were off work due to headaches and bad weather, with a tendency for headaches to occur before windy weather and a rise in humidity. Another small study of 6 patients who were observed for 5 years found no correlation between their complaints and atmospheric pressure, temperature, humidity, and ionization.21 However, headaches were less frequent in July when the weather was “biologically favourable.” These findings prompted a study of 310 patients seeking acute care at the Princess Margaret Migraine Clinic in London, including those with migraine, tension-type headache, or “migrainous neuralgia” who were queried about the date and time of the onset of their headache attack.22 No correlation was found between the attack and a change in wind velocity, wind direction, barometric pressure, temperature, or humidity.

Table 1.—. Types of Studies of Various Environmental Factors
FactorTypes of supporting studiesTypes of refuting studiesConfounding factors
  1. ED, emergency department.

Prospective cohort
Diary with meteorological records
Diary and ED studies using official weather data 
Air travel/altitudeQuestionnaire
Case report
NoiseRetrospective studies
Noise challenge study
Hearing threshold measurements
 Stressful distractor during noise challenge study
Electromagnetic fields and sferics Diary using official atmospheric data 
Outdoor light exposureSurvey
Prospective population study
Case–control study
Seasonal variation and daily light exposureQuestionnaire
Retrospective study
Light threshold studies
Retrospective ED study
Indoor lighting, fluorescent lights, computer screenCase–control
Case report
Functional MRI
Retrospective study
Head posture, office stress, noise, refractive error, bifocals, cervical disc disease, myopia, ergonomic factors
Busy visual environmentStudies of sine grating patterns  
Population studies
Olfactory hypersensitivity studies
Air pollutionPopulation study with self report
Cross sectional study
Mold QuestionnaireCo-existing allergies, asthma, sensitivity to odors

Forty-four adults with migraine from Edinburgh recorded the dates of their migraine attacks, which were compared with meteorological records.23 Analysis of 960 migraine attacks found a relationship between migraines and the morning barometric pressure reading (<1005 mb) as well as a rise in barometric pressure over the preceding 24 hours (>15 mb). Falls in barometric pressure were not correlated with migraine attacks. The highest attack rate was between September and November, and the lowest in February and March. Geomagnetic activity and meteorological factors were assessed in an Italian study of 40 migraine patients meeting IHS criteria.24 Geomagnetic activity, which influences air ionization, was monitored at 3 hour intervals by physicists. Humidity and temperature were also analyzed. Subjects recorded the date and hour of onset and duration of every headache attack from March through June, 1988. There was a significant correlation between headaches and geomagnetic activity during March only. There was no correlation with headaches and temperature or humidity.

A study of the effect of weather on headache compared patients' perceptions about the influence of weather on their migraine headaches with headache diaries.25 Adult patients (n = 77) at 3 headache centers in New York and Connecticut provided headache diaries for 2-24 months (mean = 7 months), noting when they were out of town. Their headache calendars were compared with National Weather Service data from 3 reporting sites near the study centers. Variables of interest (n = 43) included temperature, dew point, humidity, barometric pressure, wind speed, and the presence of rain, snow, haze, and thunderstorms. Based on a matrix, 3 composite variables were analyzed: (factor 1) stable weather pattern that is a function of temperature and humidity, unchanged from the previous or following day; (factor 2) change in weather pattern based on temperature, barometric pressure, or humidity from the previous 1-2 days; and (factor 3) absolute barometric pressure followed by the change in barometric pressure 2 days prior or over the following 2 days. Additionally, 61 of the patients completed questionnaires regarding their beliefs on the effect of weather.

Weather was considered the most important headache trigger by 25% and among the top 3 triggers among 60%. Rain, low barometer reading, high humidity, bright sunshine, and falling barometer were cited as triggers by approximately 30% of subjects. Hot temperature was implicated by 25%. However, analysis of the diaries indicated that 51% were significantly sensitive to any weather factors. The most frequent weather pattern associated with headache was factor 1 in 69% (stable weather). There was no correlation between the patients' beliefs and their true susceptibility based on the analysis. Thus, although weather is a trigger factor for migraine, most patients in this study could not accurately predict their own sensitivity to the weather.

A case-crossover study of 4039 emergency department (ED) visits for migraine in Ottawa, Canada compared the date and time upon arrival with the ED to hourly meteorological conditions from Environment Canada.26 They considered precipitation-related weather events (fog, snow, rain, thunder), atmospheric pressure, temperature, wind speed, and relative humidity within 24 hours of presentation. No association was found, except for a decrease in ED visits 8-12 hours after wind gusts in excess of 19 km per hour.

With so many migraineurs citing weather and barometric pressure changes as migraine triggers, one would expect air travel to produce migraines as well. A questionnaire-based study among 906 consecutive adult airline passengers was administered through an Israeli travel clinic. Fifty-two travelers (5.7%) reported flight-associated headaches.27 The majority of all participants (56%) were men but most travelers with flight-associated headaches were women (65.5%). Nineteen percent of travelers had a history of migraine, 19% reported head pain on every flight, and 45% reported that their flight-associated headaches were unilateral. Travelers with flight-associated headaches were more likely to have non-flight headaches, which tended to be bilaterally located. An equal number of subjects (n = 9) experienced headaches when descending below sea level as with climbing to a high altitude. There is one report of altitude-induced migraine headaches occurring in a previously asymptomatic airline captain after taking pravastatin and flying at specific altitudes.28


Noise as a migraine trigger is largely reported from retrospective data and is not specifically identified in many large studies. Physiological responses to a noise challenge were measured in 24 migraine subjects and 44 subjects with tension-type headache.29 Subjects were asked to solve a difficult or insolvable anagram under various conditions incorporating stress and noise. Seventy-nine percent of subjects developed a headache during the challenge phase. A subsequent study exposed subjects to annoying levels of white noise of varying duration.30 Half of the subjects developed a headache, and patients with headaches had a lower tolerance for noise and found the noise stimulus more aversive than non-headache controls. One study of sound-induced discomfort found that subjects with migraine had significantly lower noise thresholds than controls.31 Another study concurred that although the hearing threshold is not different between migraine patients and control subjects, the hearing discomfort threshold is significantly lower in migraine patients between attacks.32

The Large Analysis and Review of European housing and health Status (LARES) study analyzed noise annoyance in the housing environment and correlated it to medically diagnosed illness, including migraine.33 Examples of common sources of environmental noise annoyance are traffic noise (road, railway, aircraft, parking cars) and neighborhood noise (voices, staircase noise, footsteps). Increased risk of migraine was confirmed in adults with severe and chronic annoyance from traffic and neighborhood noise.


Sferics (or “atmospherics”) are short-duration, low-intensity electromagnetic impulses that are generated by lightning and similar electric discharges in the atmosphere.34 They reflect in the ionosphere and may travel thousands of kilometers around the earth, having an effect in areas remote to the site of a regional storm. The magnetic component of sferics penetrates into buildings and into the human body from all directions. As many sources contribute to their local intensity, there may be large daily and seasonal variations. The only demonstrated effect of sferics on the brain is a change in the pattern of the electroencephalogram in some subjects.35

To study whether electromagnetic change with weather change affects headaches, 1-year diary information from 21 patients living in the Munich area (having migraine, tension-type headache, or both) was compared with a 24-hour average of summed sferic impulses recorded in Munich.34 There was no correlation in 11 of the subjects. The headaches preceded sferics by 1 to 6 days in 4 subjects, and the sferic preceded the headache in six others. However, the correlation in these subjects was not statistically significant. Only 1 patient had a clear correlation between the occurrence of headaches and the occurrence of sferics in all parameters measured (ie, the headaches and sferics occurred at the same time).

Another German group studied the diaries of 37 women with migraine without aura (MWOA) and tension-type headaches (36 patients had both headache types) who attributed their headaches to weather conditions.36 Forty-nine percent of subjects perceived that weather conditions triggered their headaches, with the headache usually following a weather change. Sferics only correlated with thunderstorms during the summer. Within the cohort, no relationship was found between headaches and thunderstorms, barometric pressure, humidity or wind velocity during summer or autumn. Headaches, particularly tension-type headaches, were linked to moist and warm weather conditions. Increases in the daily maximum of the variation in sferic frequency, a measure of sferic activity, were associated with a higher rate of migraines during autumn.


Exposure to bright sunlight is a commonly reported migraine trigger. Less often, flickering lights (strobe light, variable light intensity while driving by trees, fluorescent lights) are implicated.37 A British study surveyed 1044 women with and without migraine regarding visual sensitivity.38 Women with migraine reported sensitivity to glare, flicker, contrasting patterns, fluorescent lights, road stripes, and colors more frequently than controls. For any given stressor, subjects having migraine with aura (MWA) were more sensitive than participants having MWOA.38 In the pediatric population, glare and bright lighting were identified as triggers by 38.8% of migraineurs and 54.9% of subjects with tension-type headache,39 and in 28.3% of children with migraine.17 A study in northern California surveyed 263 patients with migraine headaches, “non-migrainous vascular headaches,” and “muscle contraction” (tension-type) headaches.40 Exposure to the sun was a precipitant in 30% of subjects with migraine, 17% with non-migrainous vascular headaches, and 7% with muscle contraction headaches, with a statistically significant difference between migrainous and non-migrainous headaches. No data were collected about the duration of exposure or temperature. More than one-third of 494 patients identified bright sunlight as a migraine trigger in a retrospective study.41 Light was endorsed as a trigger in 26.9% of a large US migraine population study.7

Several investigations suggest a seasonal variation of migraine although only 1 study of children collected information prospectively. Headache patients from a 3-county area north of the Arctic Circle in Norway completed a questionnaire (n = 1052, representing a 75% response rate). Subjects were queried about their headache characteristics to classify them by IHS criteria. They were also asked about seasonal variation of their headaches. Responses were provided by 289 patients with migraine and 653 patients without migraine. Among migraineurs, 20.6% reported seasonal variation, compared with 17.8% of the non-migraine group. Significantly more patients in the non-migraine group reported that their headaches were worse during the polar night season. During the midnight sun season, a significantly higher proportion of migraineurs than non-migraineurs (11.7% vs 5.5%) had worse headaches. However, as suggested by the previously mentioned studies of migraine and weather, survey data based on recall are of questionable reliability.

A retrospective review of hospital admissions in South Carolina over a 20-year period disclosed 214 primary admissions for migraine.42 Hospital admissions varied seasonally by sex. Women were most frequently admitted in the spring (statistically significant) and men were most frequently admitted in autumn. Reliance on physician diagnosis rather than IHS diagnostic criteria, the patient's distance from the hospital, and other factors create potential sources of bias in this study. Thus, the link between daily light exposure and migraine is not well substantiated in the literature. The relationship to melatonin, serotonin or other neurotransmitters, and seasonal affective disorder is uncertain.

One hundred and fifteen Italian children with migraine were followed using diaries for 12 consecutive months and recorded a total of 2517 migraine events. Migraines were significantly more frequent in the late autumn-early winter, peaking between November and January for the entire cohort.43 There was a nadir of events in July.

Eighty-nine female migraine subjects in Norway prospectively recorded every migraine attack in detail for 12 months to determine whether or not there was seasonal variation.44 Overall, there was no seasonal fluctuation although a trend for more migraine attacks in the light season was present in patients having MWA. However, the seasonal variation in these subjects was attributable to sleep disturbances with insomnia-related attacks.

Migraine sufferers may be more sensitive to light in general, and those with chronic headache may be more sensitive to environmental lighting even when they are headache-free. The reduced threshold for cortical excitability increases the susceptibility to cortical activation in these individuals, which may inhibit neuronal discharge in the brainstem and facilitate trigeminovascular activation.45 It is postulated that migraineurs have defective adaptation to light stimulation, possibly related to hyperexcitability of the occipital cortex or at the thalamic level.46 One case–control study showed light sensitivity in 44.7% of headache sufferers compared with 27.8% of controls in the absence of headache.47 Bright light precipitated headaches in 29% and aggravated it in 73%. A university-based headache clinic study in Illinois found that headaches were reported least frequently during the winter compared with other times of the year.41 Women were more likely than men to report seasonal variation in their migraines.

Threshold tolerance to light stimulation was investigated in subjects meeting IHS criteria for migraine compared with controls.46 Algometry was performed on selected trigeminal cervical sites and pain perception thresholds were measured bilaterally over the emergence of the trigeminal nerve branches, the temporalis muscles, and greater occipital nerves. Participants were then exposed to computer-controlled, progressively brighter light stimulation. Individuals reported the onset of light-induced discomfort and the algometry was repeated. Light stimulation was tolerated at 20,000 lux by only 5 of the 23 healthy volunteers. Migraineurs had discomfort at a lower light intensity (mean 1747 lux, median 680 lux) than normal controls (mean 6429 lux, median 2510 lux; P < .0001). Subjects with light colored irides had discomfort at lower levels than those with dark colored irides, but the difference in iris pigmentation did not affect the results. After experiencing light-induced discomfort, migraineurs had a significant and sustained lowering of pain perception thresholds at all trigeminal cervical sites while controls subjects showed a tendency for thresholds to increase. One patient developed a migraine after the light stimulation but no subjects reported visual pain. This study implicates impaired trigeminal cervical nociception as a contributing mechanism, consistent with the known role of the trigeminovascular pathways in the generation of migraine pain.


Commonly reported migraine triggers in the indoor environment include bright lights, fluorescent lights, glare, flicker (eg, computer screen, driving along a tree-lined street), neon lights, and busy visual patterns.6 In the office environment, fluorescent lighting and flicker from the computer screen are often cited. Working at the computer screen precipitated headaches in 14.5% and aggravated it in 31.3% in one case–control study of chronic headache patients.47 The exact role of isolated visual stimuli at the workplace may be difficult to accurately ascertain; among video display terminal (VDT) users, the light source and brightness, flicker frequency, visual clarity, posture, and work-related stress may contribute to headaches.

Adverse conditions related to vision, musculoskeletal discomfort, and headaches were the most prevalent symptoms in among 1545 Massachusetts clerical workers working at VDTs.48 Another study of 92 VDT users who kept a diary of postural and visual symptoms over 5 consecutive work days found that screen legibility significantly influenced ocular discomfort, and vertical head movements affected headache symptoms.49 The contribution of posture to headaches was emphasized in 24 computer operators with occipital headaches, neck or shoulder pain.50 Prolonged cervical hyperextension combined with repetitive head rotation seemed to trigger their discomfort. Excessively high placement of the computer screen, prolonged copying of laterally placed text, wearing bifocals, and using a chair that pitched the user forward were also identified as provocative factors. About half of the operators had radiographically confirmed cervical disc disease and another 15% had electromyographic evidence of a cervical radiculopathy.

Workers (n = 814) spending at least 4 hours daily at the computer were compared with age, sex, and education level-matched controls (n = 325). Computer users had a statistically significantly higher prevalence of eye irritation, blurred vision, tearing, and headaches than control subjects.51 Improving the spectacle correction for distance vision reduced the symptoms in this cohort. An optometric study of 324 VDT workers found that mildly myopic women using the computer screen about 5 hours daily seemed to be at greatest risk for developing headaches (42%) or eyestrain (65%).52

The largest study of VDT-related symptoms was conducted in 25,000 workers at a Japanese information technology company.53 The study was performed over a 3-year period and considered confounders such as sex, age, daily working hours, and number of monthly holidays, daily sleeping hours, glasses, and contact lens use. Three major factors were analyzed: mental symptoms, physical symptoms, and sleep-related symptoms. Mental and sleep-related symptoms were higher for workers using VDTs for more than 5 hours compared with those using VDTs for less than 5 hours. Physical symptoms were more prominent with increasing duration of daily VDT use for all time intervals examined. Physical symptoms included headache, eyestrain, arthralgia, stiff shoulders, low back pain, and general fatigue. Eyestrain (38% of men, 53% of women) and stiff shoulders (13% of men, 25% of women) occurred most frequently. Headache was the third most common physical symptom, reported by about 13% of men and 33% of women with consistent percentages over the 3-year study period. Although the sample size was considerable and there was high internal consistency among respondents, the lack of a control group to assess the prevalence of headaches in other types of workers makes it difficult to draw specific conclusions. Additionally, it is not possible to determine whether visual stimulation, posture or some other factor contributed to the headaches in these workers.

A case report of a computer systems analyst who developed migrainous headaches each time he worked at his personal computer screen described resolution of his symptoms when the screen frequency was changed from 60 Hz to 75 Hz.54

Visual field defects have been demonstrated using flickering stimuli in otherwise healthy individuals who have migraine with visual aura.55,56 Subjects were tested off medications, at least 4 days (mean 31 days) after a migraine attack. One of 28 subjects with MWA, 4 of 25 subjects with MWOA, but no control subjects had at least 1 eye with significantly decreased generalized depression of sensitivity across the visual field.56 Superior arcuate field defects were found in 6 MWA and 5 MWOA subjects. Three MWA and 8 MWOA subjects had inferior arcuate field defects. Foveal defects were present in 2 MWA and 3 MWOA subjects. Higher migraine frequency and longer history of migraine correlated with general sensitivity loss but not to specific visual field defects. The etiology of the visual field defects, which are sometimes transient in migraineurs, is uncertain.

Migraineurs often report discomfort while exposed to a busy visual environment, such as the grocery store aisle, or other types of visual patterns. Sine wave patterns often induce illusions in migraineurs, and the patterns are used to study visual cortex activity in migraine patients. Migraineurs consistently show higher intensity of sine grating-induced illusions at all spatial frequencies compared with control subjects.57 Red light maximized the discomfort produced by illuminated grating patterns with variable luminance and color in migraineurs but not controls.58 Using both red and blue illuminants (letter E) on a sine grating background, subjects having MWA show statistically higher thresholds for letter detection than the group with MWOA or controls.59

Another case–control study investigated the effect of pattern glare, as subjects viewed printed sine wave gratings with and without the use of colored filters.60 Perceptual distortions and illusions were more commonly reported in migraine patients viewing a low spatial frequency pattern (3 cpd) whereas control subjects had more pattern glare using a high spatial frequency pattern (12 cpd). Pattern glare was present in 16 of the 25 migraine subjects and only 6 of 25 control subjects, yielding a positive predictive value of 73% and a negative predictive value of 68% for identifying individuals with migraine. Using the sine grating paradigm, pattern glare was reduced using a colored filter in 13 migraine subjects and a gray filter in 4 migraine subjects; most controls did not benefit from using a filter. Filters in the green to blue range were most commonly selected by subjects with migraine.

Functional blood oxygen level dependent functional MRI (fMRI-BOLD) was used to study occipital and brainstem activation in patients with visually triggered migraine used.61 A red-green checkerboard stimulus was observed by migraine subjects (off preventive medication) and controls. Four MWA patients developed visual symptoms and headaches, and 8 patients (6 MWA, 2 MWOA) developed headaches only during the testing. In 75% of subjects who developed symptoms, fMRI showed activation of the red nucleus and substantia nigra before occipital cortex signal elevation or the onset of symptoms. Patients who developed severe headache also had activation of the locus ceruleus and periaqueductal gray. This suggests that brainstem structures, as well as the occipital cortex, are involved during visually triggered migraine attacks.


Over 40% of migraineurs cite odors as a migraine trigger8,39,62 and osmophobia (aversion to odors) is present in a similar percentage during a migraine headache.62 There is little published literature on this topic. Seven hundred and twenty-seven migraineurs were queried specifically about osmophobia, taste abnormality, and perfume or odor as a migraine trigger.63 Osmophobia was defined as a “change in smell,” (increased, decreased, or different) either perceived or experienced with or without the actual source of the smell being present. Patients graded their responses on a 1 to 3 scale (occasional to very frequent). About one-quarter of patients experienced osmophobia, another quarter had taste abnormality, and perfume or odor was cited as a headache trigger in 45.4% (22% occasionally, 10.2% frequently, and 12.6% very frequently). A subsequent study by the same investigator determined that osmophobia and taste abnormalities were highly specific, although insensitive, for diagnosing migraine.63 A French population study cited smoking or smoke as a migraine trigger in 44% of subjects and powerful odors in 33.5%.6

Olfactory hypersensitivity was studied in 74 patients with migraine and 30 control subjects using everyday odorants. Olfactory hypersensitivity was determined using a validated rating index of odor intolerance. Twenty-six migraine patients, but no controls, reported interictal olfactory hypersensitivity. Migraine subjects with olfactory hypersensitivity were more likely to have odor-triggered migraines, increased attack frequency, and visual hypersensitivity than those without olfactory hypersensitivity.64

However, sensitivity to odors is common in the general population. Cacosmia, or feeling ill from the odor of xenobiotic substances (chemical odor intolerance), may be manifested as headaches, itching eyes, nasal congestion, dry or sore throat, cough, dizziness, itching or rash.65 Cacosmia seems to be proportional to educational level and is more prevalent at higher socioeconomic status.65 It is commonly present in workers exposed to solvents, office workers, and college students, and is part of the symptom cluster of Multiple Chemical Sensitivity and Sick Building syndromes.

Forty-six percent of 151 healthy food store workers studied in Italy reported feeling ill after being exposed to at least 1 of 10 materials.65 The most problematic odors were car exhaust and pesticides (19.2%), combustible gas (17.2%), tobacco smoke and asphalt (16.6%), perfumes (6%), carpeting and freshly printed paper (4.6%), and detergents (4%). Headache was the most common symptom (7.3%), occurring 1.5 times more than the next frequent symptom of eye irritation or burning. There was no association between odor intolerance and psychological symptoms, smoking status, job, age or sex in this cohort of relatively unskilled workers.

A population-based study of 1387 individuals from Sweden found general odor intolerance in one-third of subjects, usually producing respiratory symptoms, and occurring twice as frequently in women than men.66 Similar frequencies of chemically sensitive individuals were found in other studies from Arizona (thought to have the highest percentage of atopic individuals of any state) and North Carolina.67 Despite the high frequency of self-described chemically sensitive individuals in the population, most of them are generally not disabled by the effect of offensive odors.

Chemical sensitivity seems to develop in 2 stages, the loss of tolerance (induction) and subsequent triggering of symptoms by small quantities of previously tolerated substances.67 The induction may involve an acute or chronic initial exposure to pesticides, solvents, indoor air contaminants, etc. (Other migraine triggers, such as alcohol, monosodium glutamate, and chocolate fall into this category.) Subsequently, low levels of exposure to those compounds become a trigger and leads to avoidance behavior. In this regard, chemical sensitivity is similar to allergic or immunological response. Clinically, migraine and asthma are comorbid conditions, associated in the same individual more frequently than would be expected by chance alone.68-70 Moreover, children born to mothers who have migraine are twice as likely to have asthma or eczema in the first 7 years of life than those born to mothers without migraine or asthma/eczema (6% vs 3.2%).71 The risk of having asthma/allergies is greater if the mother had both migraine and asthma/eczema.71

Advances in the pathogenesis of migraine may be germane to the physiology of chemical sensitivity and explain the overlap of symptoms between and genetic predisposition to the two disorders. Neurogenic (or sterile) inflammation is characterized by edema, vasodilation, and leukocytic infiltration. Neurogenic inflammation is observed in the region of the trigeminal ganglion and dural afferents in experimental animals and humans with migraine.72-74 It is accompanied by plasma protein extravasation and release of various neuropeptides, neurotransmitters and neuromodulators, including vasoactive intestinal polypeptide, substance P, calcitonin gene-related protein, neurokinin A, and glutamate. Stimulation of trigeminal afferents produces vasodilation and changes in blood flow.

Olfactory chemical irritants may have similar effects on the trigeminovascular system. While the sense of smell is transmitted by the olfactory nerve, the trigeminal nerve innervates the nasal mucosa. Chemical irritants are transmitted by the sensory trigeminal afferent nerves, and substance P release has been verified for exposures to nicotine, formaldehyde, capsaicin, ether, and cigarette smoke.75 Meggs postulated that neurogenic inflammation may be the unifying pathogenesis for chemical sensitivity syndrome, sick building syndrome, multiple chemical sensitivity syndrome, and various reactive airway disorders.75 This is in keeping with a recent hypothesis by Burstein linking migraine triggers to the parasympathetic neurons that converge in the superior salivatory nucleus and sphenopalatine ganglion, and produce neurogenic inflammation in meningeal neurons.76 Anatomical projections from the brainstem and trigeminal nuclei to the hypothalamus, limbic area, thalamus, brainstem and spinal cord account for the pain, emotional changes, nausea, anorexia, lethargy, and other aspects of a migraine attack.76

Paradoxically, while many migraineurs experience osmophobia during their attack and can identify olfactory triggers, a study of olfactory function found decreased odor thresholds in 18% of the 67 migraineurs studied, compared with 1% of the general US population with hyposmia or anosmia.77


Cigarette smoke is a reported trigger for many migraineurs and appears to be a more frequent trigger for women than for men with migraine.41,78 Paradoxically, The Collaborative Prenatal Project of the National Institute of Neurological and Communicative Disorders and Stroke (now NINDS) identified 508 pregnant women with migraine and 3192 women with no history of migraine.79 Migraineurs smoked more heavily and had a longer smoking history than those without headaches. A smaller study comparing women with migraine with women with cluster headaches found that 14 of 27 migraineurs had never smoked, vs 3 of 27 cluster headache patients.80 Migraineurs smoked significantly less than cluster headache patients (13 vs 6 cigarettes daily) and had a shorter history of smoking than women with cluster headaches 8 vs 21 years). For comparison, 23.1% of all women and 14.6% of pregnant women in the United States smoked in the mid-1990s.81 Overall, there are little data regarding the smoking habits of migraineurs.

A large population study of over 42,000 subscribers to the Kaiser Permanente Medical Care program in northern California studied the self-reported effects of indoor tobacco smoke on non-smokers between ages 15-105 years (mean 37 years in women and 38 years in men). There was a statistically significant increase (OR 1.06-1.21) in “severe headaches” occurring more than 10 days per year in women and men, correlating with total tobacco smoke exposure time. The increased risk was found for exposure at home, small spaces, and large indoor areas. However, there was no increase in the odds of having migraine except in women exposed to smoke in large indoor areas (OR 1.37). Because the symptoms were collected by self-report, and migraine is commonly underestimated by headache suffers (perhaps confused with “sinus headache” in this scenario), it is possible that many of the subjects with severe headache actually had migraine.82

A cross-sectional study of Parisian office workers was performed to determine the effect of different ventilation systems on health, using workers from a group of office buildings with natural ventilation as the control population.83 A random sample of 2137 workers yielded a response rate of 53.5% (n = 1144), of which 857 surveys were satisfactorily completed and used in the analysis. Workers in buildings ventilated with an HVAC (heating, ventilating and air conditioning) system were more likely to participate than those from the fan-cooled and naturally ventilated buildings. The home environment (including ventilation system, exposure to smoke, pets, carpet, etc.) was also considered and air samples were obtained at work. After adjusting for confounders, workers in the HVAC system were more likely to experience work-related migraines than those in the fan-cooled work environment. Work-related migraine correlated with high mold content in the dust in the air samples. Although 42.3% of participants were active smokers, 28% were exposed to passive smoking at home, and 23.9% were exposed to passive smoking at work, no correlation between smoking and migraine was reported.

A small Spanish study reported 11 patients having headaches related to smoking certain brands of cigarettes which they termed “tobacco brand related-headache.”84 Headaches were migraine-like in 7 cases, cluster-like in three, and non-specific in 1 subject. The headaches resolved after changing brands or smoking cessation. The authors speculated that this headache may be a specific form of chemical odor intolerance.


Outdoor air quality was assessed in 10,497 headache visits to a Montreal hospital ED using air pollution data from Environment Canada.85 Parameters investigated included atmospheric pressure, gases (SO2, NO2, CO, and O3), and particulate matter in the ambient air. There was a significant association between ED visits for headaches and SO2, NO2 for current day and 1-day lagged exposure, as well as CO and small particulate matter on current day exposure. Atmospheric pressure also correlated with ED visits.

Entering “headache and mold” to a major Internet search engine yielded over 840,000 entries, many of them from mold remediation companies. However, there is scant scientific basis for the claims of “toxic mold syndrome” as a cause of migraine or headache disorders. According to the “Mold Inspector” web site, visible mold in the setting of frequent severe headaches (“migraine or otherwise”) suggests “large and possibly permanent damage to (her) brain and overall health.”86 The top 5 mold health symptoms listed on that site are memory loss or memory difficulties, learning difficulties, feeling lost or disconnected from what is happening around you, headaches, and seizures. Respiratory symptoms comprise the next eight, followed by allergic symptoms, gastrointestinal symptoms, fatigue, and treatment-resistant dandruff. A 2004 article in the Pittsburgh Post-Gazette noted the adverse health effects of Stachybotrys altra, a slimy, greenish-black mold that grows on material with a high fiber and low nitrogen content such as paper, wood, jute, carpet, or rotting hay, as “the one that's generated the most headlines, along with lawsuits, in recent years.”87 However, a report from the Centers for Disease Control and Prevention refutes the association between S. altra and severe illness based on lack of evidence and poor aerosolization of the fungus.88 The Post-Gazette correctly pointed out that the most common symptoms of mold exposure are allergic and respiratory. The person featured in the articles experienced morning headaches, sore throat, and fatigue. Her symptoms abated when she went to work and resolved after moving out of her fungus-infested house.

“Toxic Mold and Tort News Online” reports that allergies are the most common symptom of toxic mold exposure, arising from fungal spores and their airborne metabolites.89 Atopic individuals may be sensitive to molds as well as other environmental allergens. Headaches, according to this source, may be a reaction to mold odor or a direct “toxicity” of the mold on the nervous system. A listing of Toxic Mold Lawyers by region and state is also available on this site.

Toxic mold syndrome is a controversial diagnosis that is predominantly in the domain of allergists and immunologists. The most common symptoms in 65 individuals at an allergy and asthma center complaining of “toxic mold exposure” were rhinitis (62%), cough (52%), headache (34%), central nervous system symptoms (25%), and fatigue (23%).90 Skin testing provoked a skin reaction to molds in 33 of 62 patients. They concluded that the effects of mold were allergic, rather than toxic. A Finnish study surveyed teachers before and 1 year after mold remediation of school buildings and compared their findings with the general population.91 Prior to remediation, allergic rhinitis, sinusitis, conjunctivitis, and fatigue were higher than in the general population. Fatigue and headaches improved significantly after remediation. The greatest risk factor for having headaches was working at the same school for more than 10 years (OR = 5.4). Another study of 522 Danish teachers from 15 public schools (8 “water damaged” and 7 “non-damaged”) assessed mold growth and content in the dust and air of the buildings. The teachers' health was measured by questionnaire and pulmonary testing. Personal and psychosocial factors were included as confounders. Mucous membrane irritation was more common in women and correlated with mold exposure. Headache and difficulty concentrating were more common with high mold exposure, especially in women. Mold exposure did not correlate with pulmonary symptoms or objective evidence of lung impairment.

Aerosol samples and symptoms were assessed for 22 Norwegian waste collection workers who were exposed to either separated (organic) (n = 11) or unsorted (n = 11) household waste during the summer and in winter.92 Overall, exposure to bacteria, endotoxins, and dust was low. Workers noted various work-related symptoms (58%) and respiratory symptoms (41%). Tiredness (24%), headache (22%), nose irritation (17%), cough (15%), and eye irritation (10%) were the most frequent complaints. Headache, unusual tiredness, any respiratory symptom, and any work-related symptoms were significantly more frequent (P < .05) in the collectors handling organic waste than mixed waste. Exposure to fungal spores was more common in workers with headache but there was no association with bacterial, endotoxin or dust exposure. The authors commented that headache and tiredness were subjective and may be multifactorial in origin.

Molds produce alcohols and sulfur-containing compounds that exude musty and pungent odors.88 Contact with the nasal mucosa and conjunctiva may stimulate the trigeminal nerve causing local discomfort. Trigeminal stimulation may be the mechanism for headache associated with molds in this context. The consensus from the medical community is that mold-related illness is exaggerated and most claims are characterized by absence of objective evidence of disease or distinctive pathology.88,93-95 Unfortunately, most “epidemiologic studies” regarding indoor mold exposure lack a control group, are based on self-reporting rather than examination, and do not implicate any particular type of mold. The odor of mold may trigger headaches but there is no convincing evidence that mold is related to migraine. However, studies of the effects of mold infestation following Hurricane Katrina in Louisiana may provide additional insight into this area.


Many environmental factors are reported to trigger migraines and other types of headaches. Migraineurs are frequently more sensitive to various environmental stimuli than individuals without migraine. Their heightened sensitivity may be related to abnormal activation in the cerebral cortex and the brain stem. Most of the medical literature regarding migraine triggers consists of studies performed using patient interviews and surveys, and is subject to recall bias and selection bias. For example, the few studies that have investigated the relationship between migraine and weather using headache diaries and official weather reports suggest that patients overestimate the influence of weather as a migraine trigger. Other environmental factors, such as pollution, exposure to environmental allergens, indoor lighting, indoor air quality, and exposure to chemicals or odors, are potential confounders as they may provoke headaches in susceptible individuals. Rigorous prospective studies with diary confirmation are needed.

Despite the paucity of sound epidemiological evidence, many environmental factors are consistently identified by patients as migraine triggers. Population studies in many different countries find similar triggers in migraineurs. Exposure to these triggers is modifiable to some extent. Preventive measures include developing a perfume-free policy in the work places of sensitive migraineurs, banning indoor smoking, and prohibiting smoking near outdoor access doors of public buildings and places of employment. Migraine sufferers would benefit by changing indoor lighting sources to avoid fluorescent lights; this is contrary to the Department of Environmental Protection's recommendation to replace incandescent bulbs with energy saving compact fluorescent light bulbs. Minimizing time workers spend at the VDT, using ergonomically designed computer chairs, adjusting the flicker frequency on the computer screen, and using VDT filters may be helpful in preventing headaches in susceptible migraineurs. Computer users may also benefit from tinted or filtered spectacle lenses and the correction of refractive errors. Addressing preventable trigger factors, coupled with optimization of symptomatic and preventive migraine treatments, will decrease the personal, economic, and societal cost of migraine. This will require effort at the personal level, from employers as well as public policy makers.