Association of migraines with brain tumors: a nationwide population-based study

DOI: https://doi.org/10.1186/s10194- 018- 0944- 1

Abstract-Summary This study aimed to investigate the risk of migraine occurrence in the preceding years among patients diagnosed with brain tumors and unaffected controls.

Conditional logistic regressions were performed to calculate the odds ratio (OR) and the corresponding 95% confidence interval (CI) to present the association between brain tumors and having previously been diagnosed with migraines.

We found that among patients with and those without brain tumors, 554 (4.89%) and 235 (2.08%) individuals, respectively, were identified as having a prior migraine diagnosis.

Compared to unaffected controls, patients with brain tumors experienced an

independent 2.45-fold increased risk of having a prior migraine diagnosis.

The risks were even higher among men (odds ratio (OR) = 3.04, 95% confidence interval (CI) = 2.29 ~ 4.04) and after patients who had received a prior migraine diagnosis within 3 years were excluded (OR = 1.91, 95% CI = 1.59 ~ 2.29).

This is the first report demonstrating the occurrence of brain tumors to be associ- ated with a prior migraine history, for both men and women, in a population- based study.

Extended: We found that there was a significant association between brain tumors

and migraines regardless of sex.

Compared to unaffected controls, patients with brain tumors experienced a 2.45- fold increased risk of having a prior migraine diagnosis, after the factors of age, sex, monthly income, geographic region, and urbanization level were considered in the conditional logistic regression analyses.

Future studies are required to elucidate the mechanism and possible causal link, if any, connecting migraines to brain tumors, together with the potential effects of migraine treatment on the risks of consequent brain neoplasms.

10

1 Public Health

Introduction Although brain neoplasms in headache patients are less common, headaches fre- quently occur (about 33% ~ 71%) in patients with brain tumors [26–28].

Approximately 58.78% of patients with brain neoplasms also reported head- aches in a 2-year prospective study, and half of them declared headaches to be their first complaint [29].

As headaches are the most common complaint and are frequently reported as the initial symptom among patients with malignant brain tumors, it was speculated that headaches or migraines should be treated as risk factors for the development of brain tumors or merely be considered as the “first sign” of brain neoplasms.

That large, prospective cohort study found that a headache history was not asso-

ciated with increased risks of subsequent brain neoplasm occurrence.

Our nationwide, population-based nested case-control study was carried out to investigate the risks of migraine occurrence in preceding years among patients diag- nosed with a brain tumor and unaffected matched controls.

Methods As for selection of cases for this case-control study, we first identified 12,355 patients who had received a first-time diagnosis of a brain tumor (International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 191) between January 1, 2006 and December 31, 2013.

11,325 controls (one control per case) were randomly selected and matched with cases in terms of sex, age, monthly income (NT$0  ~  NT$15,840; NT$15,841 ~ NT$25,000; ≥NT$25,001, the government-stipulated minimum wage for full-time employees in Taiwan was NT$15,840, and the average exchange rate in 2017 was US$1 ≈ New Taiwan (NT)$30), geographical location (northern, central, eastern, and southern Taiwan), urbanization level (five levels, with 1 being the most urbanized and 5 the least), and index year.

While for cases, we assigned the year of the index date as the year in which the cases received their first brain tumor diagnosis, for controls, the year of the index date was simply a matched year in which controls had an ambulatory care visit.

Results The conditional logistic regression analyses (conditioned on sex, age, monthly income, geographical location, and urbanization level) suggested that the odds of prior migraines was 2.45-times greater (95% CI  =  2.09  ~  2.96, p  <  0.001) for patients with a brain tumor than for controls.

After excluding patients who had received a first-time migraine diagnosis within 1 year prior to the index date, we found a significant association still existed between brain tumors and migraines (OR = 1.99, 95% CI = 1.68–2.34).

The association between brain tumors and migraines remained after excluding patients who had received a first-time migraine diagnosis within 2 (OR  =  1.87, 95% CI  =  1.57–2.23) or 3  years (OR  =  1.91, 95% CI  =  1.59–2.29) prior to the index date.

1.1

Epidemiology

11

Discussion In this nested case-control study, we found that among patients with and those with- out brain tumors, 554 (4.89%) and 235 (2.08%) individuals, respectively, were iden- tified as having a prior migraine diagnosis.

Several methodological strengths of our study should be noted, including the use of a large-scale population-based study to evaluate the risks of brain tumors associ- ated with a prior diagnosis of migraines to fairly exempt our study from selection and non-response biases.

In terms of sex differences, Kurth and others specifically examined apparently healthy women only in a prospective study and reported a null finding of migraines being associated with subsequent brain tumor risks.

Because the healthcare utilization was recorded chronologically in this nested case-controls study, there was no apparent reason to consider that patients with and without malignant brain tumors afterwards would present distinctively on prior healthcare visits for migraine.

Conclusions Although epidemiologic evidence suggests the association, replication of studies to validate migraines per se as a risk factor for brain neoplasm development is impera- tive, both scientifically (i.e., to characterize the potential causal link) and clinically (i.e., to direct patient management).

Future studies are required to elucidate the mechanism and possible causal link, if any, connecting migraines to brain tumors, together with the potential effects of migraine treatment on the risks of consequent brain neoplasms.

Acknowledgement A machine generated summary based on the work of Chen, Chao-Hung; Sheu, Jau- Jiuan; Lin, Yi-Chun; Lin, Herng-Ching. 2018 in The Journal of Headache and Pain.

The prevalence of headache disorders in children and adolescents in Ethiopia: a schools-based study