Abstract
Background and Objectives
Headache is an adverse event associated with the use of proton pump inhibitors (PPIs). Recently, migraine has emerged more specifically as a potential adverse event with PPI use. The objectives of this work were to capitalize on existing data to evaluate the association between migraine and severe headache prevalence and use of acid-suppression therapy, including PPIs, H2 receptor antagonists (H2RAs), and generic antacids; to compare risk from PPIs vs H2RAs; and to assess for potential mitigation by a dietary factor affected by acid-suppression therapy.
Methods
Data from adults in the 1999–2004 National Health and Nutrition Examination Survey were used for this cross-sectional analysis. Acid-suppression therapy use was identified from self-report confirmed by product packaging review. Respondents who endorsed migraine or severe headache in the past 3 months were classified in the migraine or severe headache group. Dietary intake of magnesium was determined using one 24-hour recall interview. Multivariable logistic regression models were generated to analyze the relationship between acid-suppression therapy use and migraine or severe headache, and an interaction test was conducted to evaluate whether migraine or severe headache prevalence differed in relation to nutritional magnesium intake across acid-suppression therapy users and nonusers.
Results
In 11,818 US adults, the use of acid-suppression therapy was associated with higher odds of migraine or severe headache for all types of acid-suppression therapy and use of any type, as compared with those who did not use acid-suppression therapy: use of PPIs (70% higher), H2RAs (40% higher), and generic antacids (30% higher). Differences between acid-suppression therapy were not significant. An interaction was observed for H2RA use and magnesium intake (p = 0.024).
Discussion
These observations in US adults agree with previous findings that migraine or severe headache is a potential adverse event of PPIs, the most efficacious and most frequently used type of acid suppressing medication, and further suggest that other classes of acid suppressing medications (H2RAs and generic antacids) may also be implicated for migraine and severe headache. Future prospective analyses are needed to investigate migraine risk associated with acid suppressing medications while current evidence is sufficient to evaluate patients with migraine in light of recent deprescribing advice for PPIs.
Introduction
Acid-suppression therapies are widely used medications for treatment and prevention of gastric acid–related upper gastrointestinal disorders, including gastroesophageal reflux disease (GERD), Barrett esophagus, and peptic ulcer disease.1 Several classes of acid-suppression therapy are currently available in the United States. Proton pump inhibitors (PPIs) include omeprazole, esomeprazole, and lansoprazole and function by inhibiting H+/K+ ATPase pumps, the final step of acid secretion in the stomach.2 Histamine H2-receptor antagonists (H2RAs) include cimetidine, famotidine, and nizatidine and work by competitive inhibition of H2 receptors on parietal cells, which prevent transport of the pumps to the cell membrane, ultimately resulting in decreases of gastric acid secretion.3,4 By acting directly on the H+/K+ ATPase pump—the final step of acid secretion in the stomach—PPIs are more potent acid-suppression therapy and are more commonly prescribed.2,5,6
While they are considered to be generally well-tolerated, PPIs and H2RAs are also found to be overprescribed, particularly in light of increasing recognition of risks associated with long-term use of PPIs.1,5,7 Headache is listed among the most common adverse reactions in adults for both PPIs and H2RAs,4,8 and several studies have pointed to the possibility of an association between migraine and acid-suppression therapy. PPI use was associated with higher risk of migraine for adults in the UK Biobank study.9 Another study observed that the odds of reporting migraines as an adverse event to the US Food and Drug Administration (FDA) Adverse Event Reporting System (AERS) for PPIs was 2.2X the rate reported for H2RAs.10 In a study of the Taiwan National Health Insurance Database, odds of headache diagnosis increased after PPI use, with the highest odds at 1.41X of headache within 7 days of PPI exposure.11
Meanwhile, evidence is increasingly pointing toward a relationship between dietary intake and migraine.12 The reduction of gastric pH elicited by acid-suppression therapy has nutritional implications particularly when used long-term, which may be relevant to migraine, including inhibiting digestion, slowing gastric emptying, decreasing absorption of nutrients, and resulting in changes to the gut microbiome. One nutrient of particular interest is magnesium, the inadequate consumption of which has been associated with migraine risk13 and whose nutritional status may be influenced by PPI use. An FDA communication from 2011 highlighted the risk of hypomagnesemia as a potential adverse effect of PPIs during prolonged use based on AERS data.14 Although the FDA report was concerned primarily with serious adverse events, a recent meta-analysis also found that patients using PPI had 1.43X risk of hypomagnesemia,15 indicating the potential for the presence of subclinical hypomagnesemia. Similarly, the use of H2RAs has been associated with an increased risk of hypomagnesemia in the Rotterdam Study at odds similar to that of PPIs.16 However, other evidence suggests a lower risk and extent of hypomagnesemia with H2RAs compared with PPIs.17,18
Given these points, there exists biologic plausibility for PPIs to increase migraine to a greater degree than other classes of acid-suppression therapy. Their more potent acid-suppressing effects have the potential to more negatively influence various aspects with known or suspected ties to migraine, including absorption of nutrients and drugs and species diversity of the gut microbiome.19,20 Furthermore, a potential genetic susceptibility to PPIs influence on migraine has been identified, such that individuals with a rapid metabolizer phenotype for the major enzyme in PPI clearance have lower prevalence of migraine.19
Given the wide usage of acid-suppression therapy and the potential implications for migraine, this area warrants further investigation. The observed difference between rates of adverse events in PPIs and H2RAs10 introduces an important clinical question as to whether certain acid-suppression therapy pose a higher risk for precipitating or exacerbating migraine. This study sought to use an existing data set from the National Health and Nutrition Evaluation Survey (NHANES) to evaluate risk of migraine or severe headache in American adults in relation to their usage of 3 classes of acid-suppression therapy: PPIs, H2RAs, and antacids. Our hypothesis was that the use of PPIs, H2RAs, and antacids would be associated with higher rates of migraine or severe headache and that PPIs would have a greater risk. In addition, we sought to explore a potential interaction between use of acid-suppression therapy and nutritional intake of magnesium, a nutrient whose intake level is inversely associated with the risk of migraine and absorption may be reduced with PPI use.
Methods
Data Source and Subject Selection
NHANES data from the continuous survey data releases 1999–2004 were analyzed. NHANES is a repeated, cross-sectional analysis that has been conducted continuously since 1999. Data from 1999 to 2004 was used because these data survey years included a specific question about headache and migraine that was not continued in later years of the survey. Although it is possible that an individual could participate in NHANES more than once, such occurrence is unlikely, and each survey release represents new participants. The survey includes interview and physical measurement components and is conducted in a nationally representative sample of approximately 5,000 persons each year, and persons are located in 15 counties that are randomly selected each year.
In the combined data releases from 1999 to 2004, there were 31,127 NHANES participants. Participants were excluded from this analysis in the following order: age <20 years (n = 15,794), pregnant (n = 760), breastfeeding (n = 112), missing alcohol consumption information (n = 18), excessive alcohol intake (n = 113), missing body mass index information (n = 1,596), missing dietary information (n = 610), dietary intake <500 kcal or >5,000 kcal in daily reporting (n = 303), and missing headache/migraine information (n = 3). The sample used for analysis was n = 11,818.
Standard Protocol Approvals, Registrations, and Patient Consents
This analysis used deidentified, publicly available data and was thus exempt from Institutional Review Board review.
Outcome Classification
As part of a series of questions about pain, NHANES included the question of “During the past 3 months did you have severe headaches or migraines?” which was asked of anyone aged 20 years and older. Individuals who responded “yes” were classified as having migraine or severe headache and individuals who responded “no” were classified as not having migraine or severe headache.
Acid-Suppression Therapy Use
As part of the in-person interview, participants are asked to provide detailed information about dietary supplement and prescription medication use. This information was used to classify intake of antacids, H2 antagonists, and PPIs. The prescription medications questionnaire collects information about the use of prescription medications during a 30-day period before the participant's interview date. The drug codes were used to identify participants taking prescription antacids, H2 antagonists, and PPIs. For antacids, there is an additional question on the dietary supplement subsection, which also collects use in the 30 days before the participant's interview, and this was combined with the information from prescription medications for a single category of antacid use. The majority of antacid use was captured by supplement intake. Individuals were classified as user vs nonuser for each of the types of acid suppression therapy; each type was considered separately and as a combined category (use of any of the 3 types) in analysis.
Magnesium Intake
Dietary magnesium intake was estimated using data collected from a 24-hour dietary recall conducted in-person. The dietary intake data are used to estimate total intake of energy, nutrients, and nonnutrient food components from foods and beverages that were consumed during the 24-hour period before the interview (midnight to midnight). In 1999–2000, only a single dietary interview was conducted, and 2 dietary interviews were conducted for data releases 2001–2002 and later. For consistency across the data releases, the single (1999–2000) or first (2001–2004) were used in analysis. Information on nutrient intake is calculated by NHANES using the US Department of Agriculture food composition database. Supplemental magnesium intake was estimated using data collected from the dietary supplements questionnaire. Participants provide the packaging for supplements taken as part of the in-person interview and NHANES uses this information to record intake of individual nutrients within the supplements. Analyses were conducted for dietary and total (dietary plus supplemental) magnesium. Magnesium intake was considered in 50/mg day increments and as categories based on meeting recommended intake. Individuals were also classified as to whether they met recommended intake for magnesium based on diet or total intake based on Dietary Recommended Intake (DRI) cutoffs: 400 mg/d for men age 19–30 years; 420 mg/d for men age 31+ years; 310 mg/d for nonpregnant and nonlactating women age 19–30 years; and 320 mg/d for nonpregnant and nonlactating women age 31+ years.
Adjustment Variables
Demographic information was collected as part of the NHANES household interview, and there is complete data for individuals age, sex, and race and ethnicity. Age is included in the analysis as categories for descriptive purposes and in years as an adjustment variable. Possible sex responses in NHANES include men and women. Women were further divided into menopausal vs nonmenopausal based on response to a question on the reproductive health questionnaire (if menopause the reason for not having regular menstrual periods). Possible race and ethnicity responses include non-Hispanic White, non-Hispanic Black, non-Hispanic other, Mexican American, and other Hispanic. Owing to smaller sizes when considering medication intake, non-Hispanic Black and non-Hispanic other were combined to a single category, and Mexican American and other Hispanic were combined to a single category. Household income and household size information was collected as part of the interview, and this information was used by NHANES to calculate poverty-to-income ratio (PIR), which was classified into 5 categories: <1, 1 to <1.5, 1.5 to <3, 3+, and missing (where 3 represents an income that is 3 times the federal poverty income level).
Other adjustment variables included body mass index (BMI), dietary quality, caffeine, and alcohol intake. BMI was calculated from measured height and weight (kg/m2) and classified into categories of underweight (<18.5 kg/m2), normal weight (18.5 to <25 kg/m2), overweight (25 to <30 kg/m2), and obese (30+ kg/m2) for descriptive purposes. Healthy Eating Index 2010 (HEI-2010) was calculated to as a measure of dietary quality,21 which provides a score scaled 0 to 100, with 100 indicating the highest dietary quality. Caffeine intake was ascertained from the 24-hour recall and classified into categories: <5 mg/d, 5 to <100 mg/d, 100 to <400 mg/d, and 400+ mg/d. Alcohol intake was classified based on self-reported drinks in the past month: <1 drinks/month, 1 to <20 drinks/month, and 20+ drinks/month for descriptive purposes and 2 categories (because of sample size) of <20 vs 20+ drinks/month for analysis.
Statistical Analysis
Population weights (provided by NHANES), appropriate subsample designations, and stratified clustering and sampling procedures (designations of the sampling strata and primary sampling units) were used in analyses as specified in the NHANES data analytic procedures.22 Means and standard errors (continuous variables) and frequencies and percentages (categorical variables) were calculated to describe the study sample, and significance between groups was determined by the t-test (continuous) or chi-square (categorical). Logistic regression was used to calculate odds ratios for migraine or severe headache in relation to acid-suppression therapy use. Logistic regression models evaluating migraine or severe headache in relation to magnesium intake (in 50 mg/d units) were stratified by user vs nonusers of the acid-suppression therapy to visualize potential interaction between acid-suppression therapy and magnesium, and a cross-product term was used in logistic regression models to statistically test for interaction. Unadjusted and multivariable logistic regression models were generated. Multivariable models included age (in years), sex and menopause (categories), race (categories), PIR (categories), BMI (continuous), caffeine (ordinal categories), alcohol (categories), and HEI-2010 score (continuous). To explore potential clinical implications, an additional analysis comparing migraine or severe headache prevalence in H2 antagonist users vs PPI users was conducted. Analyses were conducted using Stata (version 15.1, StataCorp, TX), and a two-tailed alpha of 0.05 (p < 0.05) was used for statistical significance.
Data Availability
The data used in this study are publicly accessible through the National Center for Health Statistics of the US Centers for Disease Control (wwwn.cdc.gov/nchs/nhanes/).
Results
Individuals in the migraine or severe headache group tended to be younger, female, and have lower poverty-to-income ratio (Table 1). Body mass index distribution across categories was similar across groups. Individuals in the migraine or severe headache group appeared to have slightly lower overall diet quality, similar consumption of caffeine, and lower intake of alcohol compared with the no headache group. Dietary magnesium intake was lower in the migraine or severe headache group, but the percentages meeting magnesium intake recommendations were similar across groups, likely due to the higher representation of women with a lower DRI than men in the migraine and severe headache group.
Table 1.
Descriptive Characteristics of the Study Sample Age 20 Years and Older in NHANES (1999–2004) by Migraine or Severe Headache Status
| Personal characteristic | Headache, NO | Headache, YES |
| n (%)a or mean [SE] | n (%)a or mean [SE] | |
| Age group, y | ||
| 20 to <40 | 2,744 (0.37) | 941 (0.47) |
| 40 to <60 | 2,827 (0.38) | 912 (0.42) |
| 60+ | 3,907 (0.25) | 487 (0.11) |
| Sex | ||
| Male | 5,087 (0.52) | 799 (0.33) |
| Female, no menopause | 1,947 (0.26) | 941 (0.44) |
| Female, menopause | 2,444 (0.22) | 600 (0.23) |
| Race and ethnicity | ||
| Non-Hispanic White | 4,951 (0.74) | 1,063 (0.69) |
| Non-Hispanic non-White | 2,071 (0.14) | 571 (0.16) |
| Mexican American and other Hispanic | 2,456 (0.12) | 706 (0.15) |
| PIR | ||
| <1 | 1,386 (0.11) | 503 (0.18) |
| 1 to <1.5 | 1,267 (0.10) | 342 (0.12) |
| 1.5 to <3 | 2,397 (0.23) | 595 (0.24) |
| 3+ | 3,614 (0.49) | 683 (0.38) |
| Missing | 814 (0.07) | 217 (0.08) |
| BMI (kg/m2) | ||
| <18.5 | 135 (0.02) | 56 (0.03) |
| 18.5 to <25 | 2,899 (0.34) | 678 (0.30) |
| 25 to <30 | 3,507 (0.35) | 748 (0.30) |
| 30+ | 2,937 (0.30) | 858 (0.36) |
| Caffeine, mg/d | ||
| <5 | 1,910 (0.16) | 456 (0.16) |
| 5 to <100 | 2,736 (0.26) | 660 (0.25) |
| 100 to <400 | 3,870 (0.43) | 969 (0.44) |
| 400+ | 962 (0.14) | 255 (0.15) |
| Alcohol intake, drinks/mo | ||
| <1 | 4,920 (0.45) | 1,389 (0.54) |
| 1 to <20 | 2,824 (0.34) | 699 (0.34) |
| 20+ | 1,734 (0.21) | 252 (0.12) |
| Diet | ||
| Dietary magnesium, mg | 285.1 [2.99] | 259.7 [3.34] |
| Dietary and supplemental magnesium, mg | 312.5 [3.36] | 283.4 [3.99] |
| HEI-2010 score | 47.58 [0.39] | 44.78 [0.49] |
| Meet magnesium recommendation w/diet | ||
| No | 7,689 (0.79) | 1,912 (0.80) |
| Yes | 1,789 (0.21) | 428 (0.20) |
| Meet magnesium recommendation w/diet and supplement | ||
| No | 7,120 (0.72) | 1,794 (0.75) |
| Yes | 2,358 (0.28) | 546 (0.25) |
Abbreviations: BMI = body mass index; HEI-2010 = Healthy Eating Index 2010; PIR = poverty-to-income ratio.
Survey-weighted percentage.
Higher acid-suppression therapy use was associated with higher odds of migraine or severe headache for all types of acid-suppression therapy and use of any type (Table 2). After adjustment for personal characteristics, higher odds of migraine or severe headache were observed for PPI use (70% higher), H2 antagonist use (40% higher), supplement antacid (30% higher), and use of any type of acid-suppression therapy (47% higher).
Table 2.
Migraine or Severe Headache Status in Relation to Acid-Suppression Therapy Use in NHANES Participants 20 Years and Older (1999–2004)
| Acid-suppression therapy use | Headache, NO n (%)a |
Headache, YES n (%)a |
Unadjusted OR (95% CI) |
Adjustedb OR (95% CI) |
| PPI | ||||
| No | 8,886 (0.94) | 2,138 (0.92) | 1.00 (ref) | 1.00 (ref) |
| Yes | 592 (0.06) | 202 (0.08) | 1.39 (1.11–1.74) | 1.70 (1.32–2.18) |
| H2 antagonist | ||||
| No | 9,252 (0.98) | 2,265 (0.98) | 1.00 (ref) | 1.00 (ref) |
| Yes | 226 (0.02) | 75 (0.02) | 1.13 (0.82–1.57) | 1.40 (1.00–1.95) |
| Supplement antacid | ||||
| No | 8,516 (0.88) | 2,067 (0.86) | 1.00 (ref) | 1.00 (ref) |
| Yes | 962 (0.12) | 273 (0.14) | 1.19 (0.99–1.44) | 1.30 (1.05–1.61) |
| PPI, H2 antagonist, or supplement antacid | ||||
| No | 7,807 (0.81) | 1,836 (0.78) | 1.00 (ref) | 1.00 (ref) |
| Yes | 1,671 (0.19) | 504 (0.22) | 1.25 (1.09–1.43) | 1.47 (1.25–1.72) |
Abbreviations: NHANES = National Health and Nutrition Evaluation Survey; PPI = proton pump inhibitor.
Survey-weighted percentage.
Adjusted for age (in 5 y), sex and menopause (male, female no menopause, female menopause), race (non-Hispanic White, non-Hispanic other, Hispanic), poverty-to-income ratio (categories), body mass index (continuous), caffeine (ordinal categories), alcohol (<20 drinks/mo; 20+ drinks/mo), and HEI-2010 (continuous).
When comparing odds of migraine or severe headache in H2RA users vs PPI users and designating H2RA as the comparator and PPI as the active group using the same adjustment variables as the main analysis, no association was observed (not shown; OR = 1.20, 95% CI 0.81–1.82).
In acid-suppression therapy users and nonusers together, there were no associations of magnesium intake with migraine or severe headache (Table 3): dietary magnesium (OR = 0.99, 95% CI 0.97–1.01); meeting requirement with diet (OR = 1.08, 95% CI 0.90–1.31); total magnesium (OR = 1.00, 95% CI 0.98–1.01); and meeting requirement with total magnesium (OR = 1.06, 95% CI 0.88–1.27). When stratified by acid-suppression therapy use and nonuse and tested for interaction, a significant interaction was observed for H2 antagonist use and magnesium intake. Higher total magnesium and meeting requirements with total magnesium was associated with higher odds of migraine or severe headache in H2 antagonist users, but not in nonusers.
Table 3.
Odds of Migraine or Severe Headache in Relation to Dietary Intake of Magnesium, Stratified by Acid-Suppression Therapy Use, in NHANES Participants 20 Years and Older (1999–2004)
| Group and magnesium classification | Not user OR (95% CI) |
User OR (95% CI) |
p-interactiona |
| PPI | |||
| Dietary magnesium, in 50 mg/d | 0.99 (0.97–1.01) | 0.96 (0.83–1.10) | 0.983 |
| Meet requirement w/diet | 1.00 (0.85–1.18) | 0.60 (0.28–1.30) | 0.249 |
| Dietary and supplement magnesium, in 50 mg/d | 1.00 (0.98–1.01) | 0.99 (0.93–1.06) | 0.885 |
| Meet requirement w/diet and supplement | 0.98 (0.83–1.16) | 0.70 (0.42–1.16) | 0.270 |
| H2 antagonist | |||
| Dietary magnesium, in 50 mg/d | 0.99 (0.97–1.01) | 1.29 (1.04–1.62) | 0.055 |
| Meet requirement w/diet | 0.95 (0.81–1.12) | 2.58 (0.91–7.31) | 0.095 |
| Dietary and supplement magnesium, in 50 mg/d | 0.99 (0.97–1.01) | 1.22 (1.02–1.45) | 0.024 |
| Meet requirement w/diet and supplement | 0.94 (0.81–1.09) | 2.80 (1.22–6.44) | 0.025 |
| Supplemental antacid | |||
| Dietary magnesium, in 50 mg/d | 0.99 (0.97–1.01) | 0.98 (0.90–1.08) | 0.848 |
| Meet requirement w/diet | 0.96 (0.79–1.17) | 0.98 (0.63–1.54) | 0.889 |
| Dietary and supplement magnesium, in 50 mg/d | 0.99 (0.98–1.01) | 1.01 (0.93–1.10) | 0.899 |
| Meet requirement w/diet and supplement | 0.95 (0.79–1.13) | 1.00 (0.65–1.55) | 0.894 |
| PPI, H2 antagonist, or supplement antacid | |||
| Dietary magnesium, in 50 mg/d | 0.99 (0.97–1.01) | 0.99 (0.92–1.06) | 0.835 |
| Meet requirement w/diet | 0.97 (0.79–1.20) | 0.95 (0.66–1.37) | 0.794 |
| Dietary and supplement magnesium, in 50 mg/d | 0.99 (0.97–1.01) | 1.02 (0.98–1.07) | 0.258 |
| Meet requirement w/diet and supplement | 0.94 (0.77–1.15) | 1.02 (0.75–1.39) | 0.910 |
Abbreviations: NHANES = National Health and Nutrition Evaluation Survey; PPI = proton pump inhibitor.
Adjusted for age (in 5 y), sex and menopause (male, female no menopause, female menopause), race (non-Hispanic White, non-Hispanic other, Hispanic), poverty-to-income ratio (categories), body mass index (continuous), caffeine (ordinal categories), alcohol (<20 drinks/mo; 20+ drinks/mo), and HEI-2010 (continuous).
Discussion
This analysis of a study of American adults observed higher odds of migraine or severe headache with use of PPIs, and the results of this study agree with the limited previous work in other populations on this topic. Pisanu and others observed that PPI use was associated with a 25% increase in odds of migraine at baseline in the UK Biobank study9—this is near the low end of the 95% confidence interval observed here.
This study further observed that use of H2RAs and supplemental antacids were independently associated with higher odds of migraine or severe headache, as was use of any class of acid-suppression therapy. The effect size was larger with PPIs (1.70X) than with H2RAs (1.40) as compared with those who did not use acid-suppression therapy. In comparing odds of migraine or severe headache with H2RA users against PPI users, a statistical difference was not observed. However, the directionality of the relationship is as would be expected, with a lower OR for H2RA than PPIs, which is consistent with that of Makunts et al., who noted that migraine was reported as an adverse event to the FDA AERS at 2.2X the rate for PPIs as compared with H2RAs.10
It is possible that these observed associations are related to comorbidity between gastrointestinal (GI) conditions and migraine disease and symptoms. Numerous studies have observed associations between the presence of migraine and GI conditions, including Helicobacter pylori infection, irritable bowel syndrome, celiac disease, peptic ulcers, and gastroparesis.23-26 Furthermore, higher migraine frequency and severity is associated with higher rates of GERD.26,27 Thus, it is possible that individuals with more severe migraine may use acid-suppression therapy at higher rates to treat their greater GI symptoms. In a longitudinal analysis, the UK Biobank study observed higher rates of IBS and peptic ulcers among people with migraine at baseline; however, use of PPIs was also associated with new incidence of migraine at follow-up, indicating that the associations between PPI use and migraine onset may not be strictly detecting treatment of comorbid GI conditions.25 Therefore, it is unlikely that all variance in the relationship between acid-suppression therapy and migraine is explained by comorbid GI conditions. The possibility of adverse events related to migraines warrants further investigation prospectively, and future investigations should include H2RAs alongside PPIs.
We observed an interaction between magnesium intake level and use of H2RA, with a large effect size, such that meeting the DRI for magnesium through diet and supplements by those who reported using H2RAs was associated with a 2.8X higher odds of migraine or severe headache than those who did not meet their magnesium DRI. However, magnesium is generally considered to be protective against migraine28 and thus the expectation would be to observe a decrease in migraine odds with higher intake, which is the opposite of what was observed here. There is reason to believe that an interaction may be possible for H2RAs and not PPIs. The mechanism of PPIs is highly specific to the final stage of acid release, whereas the mechanism of H2RAs is further upstream where inhibition of histamine binding to the H2 receptor also prevents elevation in intracellular cyclic AMP and activation of protein kinase A (PKA).29 Transport of the H+/K+ ATPase pump to the plasma membrane, which ultimately produces the acid-suppressing properties of H2RAs, is only one outcome of PKA activation, leaving opportunities for other interactions with magnesium. In addition to the gastric mucosa, the H2 receptor is also present in the nervous system,30 and histamine is capable of eliciting and aggravating headache.31 The small sample size of H2RA users (n = 75) precludes further conclusions at this stage, and this relationship warrants further evaluation.
Beyond the possible interaction with magnesium, various mechanisms have been proposed to explain the relationship between acid-suppression therapy and migraine. Genetic susceptibility has recently been identified on the basis of PPI metabolizer rate by Pisanu and colleagues using the UK Biobank cohort.9 An association was observed in the overall population between PPI use and migraine prevalence. Among PPI users, those with a rapid metabolizer phenotype of the major enzyme involved in PPI clearance (CYP2C19) had a lower prevalence of migraine, indicating a potential genetic basis for differences in migraine response to PPIs. This may not be relevant for second-generation PPIs, which are not as strongly influenced by CYP2C19 enzyme activity.19
The influence of acid-suppression therapy on digestive and absorptive function in the GI tract could also play a role in their relationship with migraine. PPIs have been documented to affect nutritional status in numerous ways, by slowing gastric emptying of solid foods, inhibition of protein digestion, and decreased absorption of nutrients including vitamin B12, calcium, and magnesium.19,32 The altered GI pH may influence absorption of the divalent cations by reducing solubility of their salts and/or by influencing function of the transient receptor potential melastin 6 and 7 (TRPM6/7) transport proteins responsible for their active transport into intestinal cells. Because these mechanisms are mediated through an increased gastrointestinal pH, similar mechanisms could be expected for H2RAs and consistent use of antacids, although with a lesser effect. In addition, acid-suppression therapy influence the gut microbiome negatively through decreases in species diversity and influence on specific genera and species,20,33 and early research indicates that the gut microbiome also has potential influence over migraine.34,35
This is a cross-sectional study and attributing temporal direction of associations is limited, and there is a resulting lack of clarity in whether observed relationships are a comorbidity or adverse event. Future prospective research will be valuable to confirm and determine temporality of the observed relationships in this study. The use of a single question to identify migraine or severe headache is a limiting factor which may result in some misclassification of the outcome; however, the results of the American Migraine Prevalence and Prevention study demonstrated strong agreement between those who report severe headache and the International Classification of Headache Disorders diagnostic criteria for migraine or probable migraine.36 The timeframes for collection of data are different, with prescriptions reported for the past one month and migraine or severe headache for the past 3 months. Given the typical use, it is likely that many prescriptions were taken for longer than one month, but the available data cannot confirm this, and this may weaken the association between exposure and outcome. Furthermore, both dietary and drug intake data may be subject to some recall bias, although the NHANES data collection methods were explicitly designed to minimize this, with interviewers recording medication information directly from containers during the at-home interview and using the multiple-pass recall method for dietary interviews. Dietary data were collected as a single-day dietary recall, which does not necessarily capture an individual's usual intake; there are precision gains made by collecting a larger representation of overall dietary intake with a 24-hour intake that outweigh some of the loss of usual intake information that comes from instruments designed to collect usual intake, such as food frequency questionnaires, that limit response options. The data from this study were collected from 1999 to 2004. For most of this era, PPIs were available by prescription-only; omeprazole was the first PPI approved for over-the-counter (OTC) use at nonprescription strengths in the United States in June 2003. Because ranitidine was first available as on OTC drug in 2004, most of its use was captured on the prescription drug survey and included in the analysis; by contrast, famotidine was approved for OTC use in 1995. Given that the NHANES survey asked specifically about prescription medication, those taking OTC PPIs and H2RAs may have been missed. The sample size is small, particularly for H2RAs. Finally, it is possible that people taking the different classes of acid-suppression therapy may have different traits that we were not able to detect or control in our analyses. Finally, a smaller subset of patients may be more predisposed to profound hypomagnesemia that may go unrecognized, potentially because of drug-drug interactions and other factors that are not fully understood.37
This study observed a positive association between migraine or severe headache and use of acid-suppressing medications, including PPIs, H2RAs, and antacids. Therefore, the results of this study in a US-based population confirm results from similar studies on PPIs in other populations and headache outcomes.9,11 By observing an association for H2RAs and antacids with migraine or severe headache, this study also points to potential clinical implications for any of acid-suppression therapy. Combined, these results suggest that there is a need for more intentionally designed prospective work to inform the extent to which associations between migraine and acid-suppression therapy are merely detecting comorbidities or to what extent migraine is an adverse event associated with the medications. Meanwhile, current evidence warrants that clinicians be attuned to recent clinical practice updates7 for evaluating the need for continued use of acid-suppression therapy in patients with migraine or severe headache. Although obtaining serum magnesium concentrations may not be routine practice among neurologists in deciding whether to use magnesium for headache prevention, such testing could be considered in patients taking acid-suppression therapy.
Appendix. Authors
| Name | Location | Contribution |
| Margaret Slavin, PhD | Department of Nutrition and Food Science, University of Maryland, College Park; Department of Nutrition and Food Studies, George Mason University | Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data |
| Cara L. Frankenfeld, PhD | MaineHealth Institute of Research, Scarborough, ME | Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data |
| Alexander B. Guirguis, PharmD | VA Connecticut Healthcare System, Headache Center of Excellence, West Haven, CT | Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data |
| Elizabeth K. Seng, PhD | VA Connecticut Healthcare System, Headache Center of Excellence, West Haven, CT; Ferkauf Graduate School of Psychology, Yeshiva University; Department of Neurology, Albert Einstein College of Medicine, New York | Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data |
Study Funding
The authors report no targeted funding.
Disclosure
M. Slavin reports no disclosures relevant to this manuscript; C. Frankenfeld reports no disclosures relevant to this manuscript; A. Guirguis reports no disclosures relevant to this manuscript; E. Seng has consulted or served on an advisory board for GlaxoSmithKline, Click Therapeutics, and Abbvie and received research funding from the NINDS (NS096107 PI: Seng), NCCIH (R01AT011005-01A1 MPIs: Seng and Shallcross), and the Veteran's Health Administration (the Headache Center of Excellence Research and Evaluation Center and VA HSR&D, IRP 20-002 PI: Damush). Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.
TAKE-HOME POINTS
→ Acid-suppression therapy use was associated with migraine or severe headache for all types of acid-suppression therapy, individually and use of any type.
→ An interaction was detected between nutritional magnesium intake and use of H2RAs on odds of migraine or severe headache.
→ It is unlikely that associations between acid-suppression therapy and migraine are fully explained by comorbidities; future prospective studies are needed to better understand this relationship.
References
- 1.Freedberg DE, Kim LS, Yang YX. The risks and benefits of long-term use of proton pump inhibitors: expert review and best practice advice from the American Gastroenterological Association. Gastroenterology. 2017;152(4):706-715. doi: 10.1053/j.gastro.2017.01.031 [DOI] [PubMed] [Google Scholar]
- 2.Ahmed A, Clarke JO. Proton pump inhibitors (PPI). In: StatPearls. StatPearls Publishing; 2022. Accessed July 20, 2022. ncbi.nlm.nih.gov/books/NBK557385/ [PubMed] [Google Scholar]
- 3.Kung YM, Hsu WH, Wu MC, et al. Recent advances in the pharmacological management of gastroesophageal reflux disease. Dig Dis Sci. 2017;62(12):3298-3316. doi: 10.1007/s10620-017-4830-5 [DOI] [PubMed] [Google Scholar]
- 4.Nugent CC, Falkson SR, Terrell JM. H2 blockers. In: StatPearls. StatPearls Publishing; 2022. Accessed July 20, 2022. ncbi.nlm.nih.gov/books/NBK525994/ [PubMed] [Google Scholar]
- 5.Bustillos H, Leer K, Kitten A, Reveles KR. A cross-sectional study of national outpatient gastric acid suppressant prescribing in the United States between 2009 and 2015. PLoS ONE. 2018;13(11):e0208461. doi: 10.1371/journal.pone.0208461 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Strand DS, Kim D, Peura DA. 25 Years of proton pump inhibitors: a comprehensive review. Gut Liver. 2017;11(1):27-37. doi: 10.5009/gnl15502 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Targownik LE, Fisher DA, Saini SD. AGA clinical practice update on de-prescribing of proton pump inhibitors: expert review. Gastroenterology. 2022;162(4):1334-1342. doi: 10.1053/j.gastro.2021.12.247 [DOI] [PubMed] [Google Scholar]
- 8.Yibirin M, De Oliveira D, Valera R, Plitt AE, Lutgen S. Adverse effects associated with proton pump inhibitor use. Cureus. 2021;13(1):e12759. doi: 10.7759/cureus.12759 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pisanu C, Welander NZ, Rukh G, Schiöth HB, Mwinyi J. Association between migraine prevalence, treatment with proton-pump inhibitors and CYP2C19 phenotypes in UK Biobank. Biomed Pharmacother. 2021;143:112234. doi: 10.1016/j.biopha.2021.112234 [DOI] [PubMed] [Google Scholar]
- 10.Makunts T, Alpatty S, Lee KC, Atayee RS, Abagyan R. Proton-pump inhibitor use is associated with a broad spectrum of neurological adverse events including impaired hearing, vision, and memory. Sci Rep. 2019;9(1):17280. doi: 10.1038/s41598-019-53622-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Liang JF, Chen YT, Fuh JL, et al. Proton pump inhibitor-related headaches: a nationwide population-based case-crossover study in Taiwan. Cephalalgia. 2015;35(3):203-210. doi: 10.1177/0333102414535114 [DOI] [PubMed] [Google Scholar]
- 12.Gazerani P. A bidirectional view of migraine and diet relationship. Neuropsychiatr Dis Treat. 2021;17:435-451. doi: 10.2147/NDT.S282565 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Slavin M, Li H, Khatri M, Frankenfeld C. Dietary magnesium and migraine in adults: a cross-sectional analysis of the national health and nutrition examination survey 2001-2004. Headache. 2021;61(2):276-286. doi: 10.1111/head.14065 [DOI] [PubMed] [Google Scholar]
- 14.Center for Drug Research. Low Magnesium Levels Can Be Associated with Long-Term Use of Proton Pump Inhibitor Drugs (PPIs). U.S. Food and Drug Administration; 2011. Accessed July 22, 2022. fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-low-magnesium-levels-can-be-associated-long-term-use-proton-pump [Google Scholar]
- 15.Cheungpasitporn W, Thongprayoon C, Kittanamongkolchai W, et al. Proton pump inhibitors linked to hypomagnesemia: a systematic review and meta-analysis of observational studies. Ren Fail. 2015;37(7):1237-1241. doi: 10.3109/0886022X.2015.1057800 [DOI] [PubMed] [Google Scholar]
- 16.Kieboom BCT, Kiefte-de Jong JC, Eijgelsheim M, et al. Proton pump inhibitors and hypomagnesemia in the general population: a population-based cohort study. Am J Kidney Dis. 2015;66(5):775-782. doi: 10.1053/j.ajkd.2015.05.012 [DOI] [PubMed] [Google Scholar]
- 17.Hess MW, Hoenderop JGJ, Bindels RJM, Drenth JPH. Systematic review: hypomagnesaemia induced by proton pump inhibition. Aliment Pharmacol Ther. 2012;36(5):405-413. doi: 10.1111/j.1365-2036.2012.05201.x [DOI] [PubMed] [Google Scholar]
- 18.Danziger J, William JH, Scott DJ, et al. Proton-pump inhibitor use is associated with low serum magnesium concentrations. Kidney Int. 2013;83(4):692-699. doi: 10.1038/ki.2012.452 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kinoshita Y, Ishimura N, Ishihara S. Advantages and disadvantages of long-term proton pump inhibitor use. J Neurogastroenterol Motil. 2018;24(2):182-196. doi: 10.5056/jnm18001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Imhann F, Bonder MJ, Vich Vila A, et al. Proton pump inhibitors affect the gut microbiome. Gut. 2016;65(5):740-748. doi: 10.1136/gutjnl-2015-310376 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Guenther PM, Kirkpatrick SI, Reedy J, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr. 2014;144(3):399-407. doi: 10.3945/jn.113.183079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.National Center for Health Statistics (U.S.). National Health and Nutrition Examination Survey: Analytic Guidelines, 1999-2010. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics; 2013. [Google Scholar]
- 23.Hormati A, Akbari N, Sharifipour E, et al. Migraine and gastric disorders: are they associated? J Res Med Sci. 2019;24(1):60. doi: 10.4103/jrms.JRMS_464_18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Cámara-Lemarroy CR, Rodriguez-Gutierrez R, Monreal-Robles R, Marfil-Rivera A. Gastrointestinal disorders associated with migraine: a comprehensive review. World J Gastroenterol. 2016;22(36):8149-8160. doi: 10.3748/wjg.v22.i36.8149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Welander NZ, Olivo G, Pisanu C, Rukh G, Schiöth HB, Mwinyi J. Migraine and gastrointestinal disorders in middle and old age: a UK Biobank study. Brain Behav. 2021;11(8):e2291. doi: 10.1002/brb3.2291 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Aamodt AH, Stovner LJ, Hagen K, Zwart JA. Comorbidity of headache and gastrointestinal complaints. The Head-HUNT Study. Cephalalgia. 2008;28(2):144-151. doi: 10.1111/j.1468-2982.2007.01486.x [DOI] [PubMed] [Google Scholar]
- 27.Katić BJ, Golden W, Cady RK, Hu XH. GERD prevalence in migraine patients and the implication for acute migraine treatment. J Headache Pain. 2009;10(1):35-43. doi: 10.1007/s10194-008-0083-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dolati S, Rikhtegar R, Mehdizadeh A, Yousefi M. The role of magnesium in pathophysiology and migraine treatment. Biol Trace Elem Res. 2020;196(2):375-383. doi: 10.1007/s12011-019-01931-z [DOI] [PubMed] [Google Scholar]
- 29.Shin JM, Kim N. Pharmacokinetics and pharmacodynamics of the proton pump inhibitors. J Neurogastroenterol Motil. 2013;19(1):25-35. doi: 10.5056/jnm.2013.19.1.25 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Obara I, Telezhkin V, Alrashdi I, Chazot PL. Histamine, histamine receptors, and neuropathic pain relief. Br J Pharmacol. 2020;177(3):580-599. doi: 10.1111/bph.14696 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Alstadhaug KB. Histamine in migraine and brain. Headache. 2014;54(2):246-259. doi: 10.1111/head.12293 [DOI] [PubMed] [Google Scholar]
- 32.Sanaka M, Yamamoto T, Kuyama Y. Effects of proton pump inhibitors on gastric emptying: a systematic review. Dig Dis Sci. 2010;55(9):2431-2440. doi: 10.1007/s10620-009-1076-x [DOI] [PubMed] [Google Scholar]
- 33.Elias E, Targownik LE. The clinician's guide to proton pump inhibitor related adverse events. Drugs. 2019;79(7):715-731. doi: 10.1007/s40265-019-01110-3 [DOI] [PubMed] [Google Scholar]
- 34.Arzani M, Jahromi SR, Ghorbani Z, et al. Gut-brain axis and migraine headache: a comprehensive review. J Headache Pain. 2020;21(1):15. doi: 10.1186/s10194-020-1078-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Slavin M, Li H, Frankenfeld C, Cheskin L. What is needed for evidence-based dietary recommendations for migraine: a call to action for nutrition and microbiome research. Headache. 2019;59(9):1566-1581. doi: 10.1111/head.13658 [DOI] [PubMed] [Google Scholar]
- 36.Buse DC, Loder EW, Gorman JA, et al. Sex differences in the prevalence, symptoms, and associated features of migraine, probable migraine and other severe headache: results of the American Migraine Prevalence and Prevention (AMPP) Study. Headache. 2013;53(8):1278-1299. doi: 10.1111/head.12150 [DOI] [PubMed] [Google Scholar]
- 37.Atkinson NSS, Reynolds DJM, Travis SPL. “Lemonade Legs”: why do some patients get profound hypomagnesaemia on proton-pump inhibitors? Intest Res. 2015;13(3):227-232. doi: 10.5217/ir.2015.13.3.227 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data used in this study are publicly accessible through the National Center for Health Statistics of the US Centers for Disease Control (wwwn.cdc.gov/nchs/nhanes/).