Increased risk of uterine leiomyoma among women with migraine in reproductive age.

This study is a nationwide population-based cohort study conducted through a secondary analysis of claims and health examination data from the Korean National Health Insurance Service (NHIS) database. The NHIS is a single public insurer that covers approximately 97% of the Korean population; the remaining 3%, who is in the lowest income bracket, is also covered by the NHIS through the Medical Aid program. All individuals aged 20 years or older who are registered in the NHIS database are advised to undergo national health screening every 2 years. Consequently, the NHIS database contains a substantial volume of representative information on personal demographics, healthcare utilization, and national health-screening results for almost the entire population of Korea. The NHIS database has been used to create cohort data for a variety of epidemiological studies 30 , and its validity has been demonstrated 31 , 32 . This study was approved by the institutional review board of the Samsung Medical Center (IRB File No.: 2022-07-101) and adhered to the Helsinki Declaration. The IRB of the Samsung Medical Center waived the requirement for written informed consent from participants due to the public and anonymized nature of the data, in accordance with confidentiality guidelines (Fig. 1 ). Fig. 1 A Kaplan-Meier plot showing cumulative incidence of uterine leiomyoma according to the presence or absence of migraine. A Kaplan-Meier plot showing cumulative incidence of uterine leiomyoma according to the presence or absence of migraine. This study initially included 2,755,790 women aged 20–39 years who underwent national health screening between January 1, 2009, and December 31, 2012. Among these, women with missing data for one or more variables listed in Table  1 ( n  = 108,553), those who had been diagnosed with a uterine leiomyoma before health screening ( n  = 76,049), and those who were diagnosed with a uterine leiomyoma and died within 1 year of the health-screening date ( n  = 12,077) were excluded from the study. Finally, 2,559,111 women were included in the analysis (Fig.  2 ). This is an observational study based on pre-existing data, and no statistical power calculation was conducted prior to the study. The sample size was determined based on the available data. Fig. 2 Flowchart of the study population. Flowchart of the study population. Table 1 Baseline characteristics of the study population. Data are presented as mean ± standard deviation for continuous variables and n (%) for categorical variables.abbreviations: HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol. a Geometric means. b Absolute standardized mean Difference. Total ( n  = 2,559,111) Women without uterine leiomyoma ( n  = 2,404,683) Women with uterine leiomyoma ( n  = 154,428) p value ASD b Age 29.7 ± 5.2 29.5 ± 5.2 31.7 ± 5.0 < 0.001 0.435 Income (lowest 25%) 739,241 (28.9) 696,330 (29.0) 42,911(27.8) < 0.001 0.026 Smoking status < 0.001 Never 2,316,189 (90.5) 2,175,991 (90.5) 140,198 (90.8) 0.010 Ex-smoker 91,513 (3.6) 86,118 (3.6) 5395 (3.5) 0.005 Current 151,409 (5.9) 142,574 (5.9) 8,835 (5.7) 0.009 Alcohol consumption < 0.001 Non 1,392,189(54.4) 1,308,052 (54.4) 84,137 (54.5) 0.002 Mild 1,105,323 (43.2) 1,038,417 (43.2) 66,906 (43.3) 0.003 Heavy 61,599 (2.4) 58,214 (2.4) 3,385 (2.2) 0.015 Regular exercise 244,418 (9.6) 228,742 (9.5) 15,676 (10.2) < 0.001 0.022 Body mass index (kg/m2) 21.3 ± 3.3 21.3 ± 3.3 21.6 ± 3.2 < 0.001 0.090 < 18.5 382,625 (15.0) 363,938 (15.1) 18,687 (12.1) 0.088 18.5–23 1,580,302 (61.8) 1,485,066 (61.8) 95,236 (61.7) 0.002 23–25 295,721 (11.6) 275,783 (11.5) 19,938 (12.9) 0.044 25–30 240,440 (9.4) 223,558 (9.3) 16,882 (10.9) 0.054 ≥ 30 60,023 (2.4) 56,338(2.3) 3,685 (2.4) 0.003 Abdominal obesity 156,590 (6.1) 146,542 (6.1) 10,048(6.5) < 0.001 0.017 Blood pressure Systolic 111.2 ± 11.5 111.2 ± 11.5 112.2 ± 11.8 < 0.001 0.090 Diastolic 69.8 ± 8.5 69.8 ± 8.5 70.5 ± 8.7 < 0.001 0.085 Comorbid condition Hypertension 59,093 (2.3) 54,185 (2.3) 4,908 (3.2) < 0.001 0.057 Diabetes 23,952 (0.9) 22,303 (0.9) 1,649 (1.1) < 0.001 0.014 Dyslipidemia 94,704 (3.7) 88,724 (3.7) 5,980 (3.9) < 0.001 0.009 Cancer 10,578 (0.4) 9,486 (0.4) 1,092 (0.7) < 0.001 0.043 Endometriosis 9,501(0.4) 8,210(0.3) 1,291(0.8) < 0.001 0.065 Laboratory findings Fasting glucose (mg/dL) 88.2 ± 13.2 88.1 ± 13.2 88.8 ± 13.4 < 0.001 0.051 Totalcholesterol (mg/dL) 177.9 ± 30.8 177.8 ± 30.9 179.5 ± 30.4 < 0.001 0.056 Triglyceride a (mg/dL) 71.8 (71.7–71.8) 71.66 (71.6–71.7) 73.83 (73.7–74) < 0.001 0.046 HDL-C (mg/dL) 63.3 ± 23.6 63.3 ± 23.5 62.6 ± 23.8 < 0.001 0.031 LDL-C (mg/dL) 99.2 ± 30.9 99.0 ± 30.8 101.0 ± 31.0 < 0.001 0.064 Baseline characteristics of the study population. Data are presented as mean ± standard deviation for continuous variables and n (%) for categorical variables.abbreviations: HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol. a Geometric means. b Absolute standardized mean Difference. The starting point of our study is the date of the first health examination for all participants, rather than using the individual migraine diagnosis dates. Study subjects were linked to claims data and divided into subgroups based on the presence or absence of migraine at baseline, defined using ICD-10 diagnostic code G43. To investigate the association between migraine and the risk of uterine leiomyomas, this study focused on participants who were newly diagnosed with uterine leiomyomas during the follow-up period. The main outcome of the study was the incidence of uterine leiomyoma, defined as two outpatient visits or one hospitalization with ICD-10 code D25 (i.e., the diagnostic code of uterine leiomyoma) per year. Using at least one inpatient diagnosis or a minimum of two outpatient diagnoses as the criteria for defining medical conditions is a common practice in healthcare and medical research to ensure the validity of the diagnosis 33 , 34 . The follow-up period was from the date of the first national health examination to the date of diagnosis of new-onset uterine leiomyoma or the end of the study in December 2018, whichever came first. We excluded individuals with previously diagnosed uterine leiomyoma (wash-out), those who developed uterine leiomyoma or died within the first year of entry (lag period), and those with missing data for at least one variable. Death before and events after December 31st, 2018 were right-censored. Well-trained examiners took anthropometric measurements, such as weight, height, waist circumference (WC), and systolic and diastolic blood pressure. The body mass index was calculated as weight (in kilograms) divided by height (in meters squared); using the World Health Organization Asia–Pacific criteria, it was classified as underweight (< 18.5 kg/m 2 ), normal (18.5–23 kg/m 2 ), overweight (23–25 kg/m 2 ), obese (25–30 kg/m 2 ), and severely obese (≥ 30 kg/m 2 ). Based on the Korean Society for the Study of Obesity guidelines, abdominal obesity was defined by a WC of ≥ 85 cm. Information on health-related behaviors, such as smoking, alcohol consumption, and regular exercise, was obtained using a self-reported questionnaire. The smoking status was classified as never, ex-smoker, and current smoker. Alcohol consumption was classified as none (0 g/day), mild (< 30 g/day), and heavy (≥ 30 g/day). Regular exercise was defined as ≥ 30 min of moderate physical activity performed at least 5 times per week or ≥ 20 min of vigorous physical activity performed at least 3 times per week. The income level was dichotomized at the lowest 25%. The baseline comorbidities were based on claims data and health-screening results. Diabetes was defined as the presence of a fasting glucose level of ≥ 126 mg/dL or at least one claim for the prescription of antidiabetic medications under ICD-10 codes E11–14 per year. Hypertension was defined as the presence of a systolic/diastolic blood pressure of ≥ 140/90 mmHg or at least one claim for the prescription of antihypertensive agents under ICD-10 codes I10–I13 and I15 per year. Dyslipidemia was defined as the presence of a total cholesterol level of ≥ 240 mg/dL or at least one claim for the prescription of lipid-lowering agents under ICD-10 code E78 per year. Endometriosis was defined as the presence of at least one claim under ICD-10 code N80. ICD-10 codes beginning with C and the registration code V193 were used to define cancer. Definition of endometriosis and cancer were used in analysis of baseline characteristics. Blood samples were obtained on the day of the health screening after making the participants fast for 8 h overnight. Independent t-tests and chi-squared tests were used to compare continuous and categorical variables of the baseline characteristics, respectively, between women with and without uterine leiomyoma. The results were presented as means and standard deviations for continuous variables and as numbers and percentages for categorical variables. The incidence rate of uterine leiomyoma was calculated by dividing the number of incident cases by the total follow-up period, multiplied by the study population. The proportional hazard assumption was tested using log-log plot for Cox regression, and the assumption was satisfied. A Cox proportional hazards regression model was used to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for the risk of incident uterine leiomyoma according to the presence or absence of migraine. Three different models were applied: (1) Model 1 was not adjusted; (2) Model 2 was adjusted for sociodemographic factors (age, income) and lifestyle factors (smoking status, alcohol consumption, and exercise); and (3) Model 3 was further adjusted for covariates in Model 2 as well as comorbid conditions (obesity, diabetes, hypertension, dyslipidemia, endometriosis, and cancer) and anthropometric measurement (BMI). Kaplan–Meier curves were used to graphically describe the cumulative incidence of uterine leiomyoma over time. All statistical analyses were performed using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA); the Kaplan–Meier plots were visualized using R (version 3.6.4; The R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was defined as a p value < 0.05 in a two-sided test.

The baseline characteristics of the study participants, stratified by the presence or absence of uterine leiomyoma, are summarized in Table  1 . Among the 2,559,111 women included, 154,428 were diagnosed with uterine leiomyoma. The mean ages of women with and without uterine leiomyoma were 31.7 years and 29.5 years, respectively. Women with a low income were less common in the uterine leiomyoma group. Furthermore, the body mass index and percentage of abdominal obesity were higher in the uterine leiomyoma group. The women in this group were less likely to be current smokers or heavy drinkers, and were more likely to engage in regular exercise. Comorbidities, including hypertension, diabetes, dyslipidemia, endometriosis, and cancer, were more common in the uterine leiomyoma group. Furthermore, laboratory variables, including fasting glucose, total cholesterol, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol, were less metabolically favorable in this group. With the concern of a large sample size, where even minor differences in distribution can result in statistically significant p-values, we have included ASD (Absolute Mean Standardized Difference) in Table  1 as a balance test between the groups of women with and without uterine leiomyoma. A significant distribution difference can be considered when ASD is greater than 0.1. Except for age, the ASD values for all covariates are less than 0.1, indicating no significant distribution differences in specific covariates between the groups. The mean follow-up duration was 7.1 ± 1.4 years. The incidence rates of uterine leiomyoma in women with and without migraine were 9.78 and 8.50 per 1,000 person-years, respectively. Based on the non-adjusted Model 1, the risk of incident uterine leiomyoma was significantly higher in the migraine group than in the non-migraine group (HR: 1.15, 95% CI: 1.12–1.19). After adjusting for age, income, smoking, alcohol consumption, and exercise in Model 2, the migraine group still had a higher risk of incident uterine leiomyoma than the non-migraine group (adjusted HR [aHR]: 1.10, 95% CI: 1.07–1.14). The results remained consistent after further adjustment for the body mass index, abdominal obesity, and various comorbidities in Model 3 ([aHR]: 1.10, 95% CI: 1.07–1.13). In the subgroup analysis, we observed that the migraine with aura group had a higher HR of uterine leiomyoma development than the migraine without aura group (HR = 1.19 [95% CI, 1.06–1.33] vs. HR = 1.09 [95% CI, 1.06–1.13]; Table  2 ). The Kaplan–Meier plot of the cumulative incidence of uterine leiomyoma in women with and without migraine is shown in Fig.  1 . Uterine leiomyoma occurred more frequently in the migraine group than in the non-migraine group (Fig.  1 A). In the subgroup analysis, the migraine with aura group showed a higher incidence of uterine leiomyoma than the migraine without aura group (Fig.  1 B) (log-rank [overall] p  < 0.001). Table 2 Hazard rations and 95% confidence intervals for incident uterine leiomyoma according to the presence or absence of migraine. Model 1 non-adjusted.Model 2 adjusted for age, income, smoking status, alcohol consumption, and regular exercise.model 3 adjusted for age, income, smoking status, alcohol consumption, regular exercise, body mass index, abdominal obesity, hypertension, diabetes, dyslipidemia, cancer, and endometriosis. Subjects ( n ) Events ( n ) Duration (person-years) Incidence rate (per 1,000 person-years) Hazard ratio (95% confidence interval) Model 1 Model 2 Model 3 No Migraine 2,491,940 149,822 1,762,0643 8.50 1 (Ref.) 1 (Ref.) 1 (Ref.) Migraine 67,171 4,606 470,908 9.78 1.15 (1.12–1.19) 1.10 (1.07–1.14) 1.10 (1.07–1.13) Migraine withoutaura 62,985 4,301 441,803 9.74 1.15 (1.12–1.19) 1.10 (1.07–1.13) 1.10 (1.06–1.13) Migraine withaura 4,186 305 29,105 10.48 1.24 (1.11–1.39) 1.19 (1.07–1.33) 1.19 (1.06–1.33) Hazard rations and 95% confidence intervals for incident uterine leiomyoma according to the presence or absence of migraine. Model 1 non-adjusted.Model 2 adjusted for age, income, smoking status, alcohol consumption, and regular exercise.model 3 adjusted for age, income, smoking status, alcohol consumption, regular exercise, body mass index, abdominal obesity, hypertension, diabetes, dyslipidemia, cancer, and endometriosis.

We found that the risk for uterine leiomyoma was higher in women with migraine than in those without, and a strong association between the two persisted after adjusting for other potential confounders. This study is the first to suggest a potential association between uterine leiomyoma and migraine in a large population. Leiomyoma development is thought to result from chronic tissue injury and inflammation; these conditions cause myofibroblast cells to excessively produce components of the extracellular matrix (ECM), resulting in pathological fibrosis. Recently, vitamin D has been identified as a potent antitumor agent that inhibits leiomyoma cells and reduces the size of leiomyoma. Previous studies have demonstrated a correlation between vitamin D deficiency and an increased risk of leiomyoma 35 , 36 . One study found that 1,25(OH) 2 D 3 inhibited the growth of both myometrium and leiomyoma cells 37 . Further studies on the pathogenesis of fibrosis have revealed that 1,25(OH) 2 D 3 reduced transforming growth factor-β3(TGF-β3)-induced protein expression in immortalized human uterine leiomyoma cells. TGF-β3, a profibrotic cytokine, is elevated in leiomyoma cells and upregulates the synthesis of ECM proteins (namely fibronectin, collagen, andproteoglycan) 38 . One study revealed that treatment with 1,25(OH) 2 D 3 significantly reduced leiomyoma volume in the Eker rat model by the following mechanisms: (1) suppression of cell-growth genes that regulate proliferation and apoptosis and (2) reduction of estrogen and progesterone receptor levels 39 . Thus, based on recent reports, vitamin D is suggested as a potential option for the effective, noninvasive, and safe treatment of leiomyoma. Randomized controlled trials have also been conducted to evaluate the effects of various doses of vitamin D on leiomyoma 40 , 41 . The pathogenesis of migraine remains unclear; however, substantial evidence supports the hypothesis that neurogenic inflammation and neuroinflammation play a key role in its development 42 – 44 . Vitamin D is known to have an anti-inflammatory effect, and suppresses the levels of pro-inflammatory cytokines 29 , 45 . Recent studies have demonstrated that the 25-hydroxyvitamin D receptor (VDR) and 1α-hydroxylase (1α-OHase), responsible for the production of the active form of vitamin D, are expressed in many areas of the central nervous system, particularly in the hypothalamus 46 . In this context, vitamin D administration may potentially be effective in reducing migraine attacks. The association between migraine and uterine leiomyoma may be influenced by sex hormones, particularly estrogen. Migraine, more prevalent in women than in men, follows a lifetime course mirroring female reproductive milestones, increasing after menarche, peaking during reproductive age, and declining post-menopause 47 . The “estrogen withdrawal hypothesis” suggests that fluctuations in estrogen levels, particularly prolonged elevation followed by decline, may trigger migraine without aura 48 . Supporting this, an animal model demonstrated that estrogen level fluctuations modulate the expression of specific peptides in trigeminal ganglia, potentially contributing to migraine pathogenesis 49 . While fluctuating estrogen levels are associated with migraine without aura, maintaining high and stable plasma estrogen concentrations, as observed during pregnancy or with the use of oral contraceptives and hormone replacement therapy, is suggested to increase susceptibility to migraine with aura 50 , 51 . An additional in-vitro study of female rat trigeminal ganglia provided evidence that chronically elevated plasma estrogen concentrations may enhance the likelihood of migraine through gene expression and intracellular signaling, crucial in the hormonal regulation of pain pathways 52 . The well-established association between estrogen levels and myoma development further supports our hypothesis. Estrogen, in collaboration with progesterone, stimulates myoma growth 53 , and epidemiological studies have identified factors contributing to increased lifetime exposure to estrogen, such as early menarche, late menopause, and obesity, which are associated with an elevated risk of developing uterine leiomyoma. The incidence and growth rate of myoma generally decrease after menopause 54 . Considering the significant role of estrogen in pain modulation and myoma development, it seems plausible that this could provide insight into the observed association between uterine leiomyoma and migraine in our study. We found a significant association between migraine and the development of uterine leiomyoma, with migraine with aura showing a stronger association. However, it is essential to acknowledge a limitation in our study—the absence of estrogen concentration measurements. Consequently, these hypotheses remain speculative and will require confirmation in future studies. To the best of our knowledge, this is the first study to investigate the risk of uterine myoma among women with migraine. Moreover, the study population was derived from the nationwide NHIS database, allowing our findings to be generalized to a wider population in South Korea. However, several limitations must be considered when interpreting our findings. First, although more than 75% of women aged 20–39 participated in the health check-up program, not all women in this age group underwent health check-ups. Since those who did not participate or had one or more missing variables were excluded from the analysis, this may have introduced a generalizability bias. Second, reliance on ICD-10 codes from insurance claims to diagnose uterine leiomyoma and migraine may lead to misclassification, as some cases might be coded differently or missed if individuals did not seek medical care, potentially causing selection bias. Third, while the estimated incidence of uterine leiomyoma by age 50 is up to 70%, only 6% of our cohort (median age 31) was diagnosed 55 , Table  1 . This discrepancy may stem from asymptomatic cases not captured in hospital visits and the younger age of our population, consistent with reports of lower incidence (under 10%) in white women at age 30, though data specific to Asian women are lacking 56 . Fourth, the NHIS database lacks detailed personal medical data, preventing adjustment for confounders like hormonal, genetic, lifestyle, and environmental factors (e.g., diet, vitamin D, or phthalate exposure), though racial variation likely has minimal impact given our predominantly Korean cohort 57 – 60 . Fifth, the NHIS database did not provide disease-specific data, such as the characteristics (i.e., episode vs. chronic), duration, frequency, and severity of migraine, as well as the symptoms associated with leiomyoma. Therefore, we could not investigate whether the characteristics of migraine and leiomyoma had different effects on the relationship between the two conditions. Sixth, defining comorbidities from a single health screening may over- or under-report chronic conditions. Lastly, due to the nature of NHIS data, critical clinical information, such as serum vitamin D levels or vitamin D supplement usage, could not be included in the data analyses. Future studies should take these aspects into consideration to provide a more comprehensive understanding of the role of vitamin D in the relationship between uterine leiomyoma development and migraine.

This study provides population-based evidence that the risk of incident uterine leiomyoma are significantly increased in women with migraine. Our findings suggest that clinicians should consider the potential development of leiomyoma in patients with migraine. Vitamin D is thought to be a potential linker in the association between uterine leiomyoma and migraine; however, there is still insufficient evidence to support this due to lack of data. Further studies are required to confirm the association between the two as well as the pathophysiological mechanisms underlying this association.

Gynecological diseases and headache disorders are the leading causes of global disability in women (especially those aged 15–49 years) 1 .Uterine leiomyoma(also known as fibroids) are the most common benign gynecological tumors, affecting nearly 70% of womenof reproductive age 2 – 4 . At least 50% of leiomyoma are asymptomatic and may be detected incidentally by ultrasonography conducted for screening or other indications 5 , 6 . Conversely, a significant number of women with leiomyoma often suffer from conditions that negatively impact their well-being, such as heavy or prolonged uterine bleeding, pelvic pain or pressure, urinary symptoms, constipation, sexual dysfunction, or pregnancy complications. Heavy menstrual bleeding, the most common symptom of uterine leiomyoma, significantly reduces the quality of life for women and can lead to medical conditions such as anemia, which has long-term negative effects on health 3 , 5 , 7 – 9 . Despite the high prevalence and substantial burden of leiomyoma, their etiology and molecular mechanisms are not fully understood. Possible risk factors for uterine leiomyoma include age, race, family history of leiomyoma, reproductive status, body mass index, hypertension, and environmental factors (such as diet, vitamin level, exercise, smoking, alcohol, and stress) 8 , 10 – 12 . In addition, recent evidence increasingly highlights that vitamin D levels are significant risk factors for the development of uterine leiomyoma. Specifically, some studies have demonstrated that vitamin D can reduce leiomyoma cell proliferation in vitro and inhibit tumor growth in in vivo animal models. Migraine is a common disabling primary headache disorder characterized by recurrent headaches with unilateral location, pulsating pain, worsening with physical activity, and often associated with nausea, photophobia and phonophobia 13 . Migraine is associated with a significant burden of illness, and affected individuals have a lower health-related quality of life 14 – 16 . Women are three times more commonly affected by migraine than men; in fact, migraine is the third most prevalent cause of global disability among women 1 , 17 . The lifetime prevalence of migraine in women is 43%; the peak prevalence is in the reproductive years(i.e., between the ages of 25 and 55 years) 10 , 11 , 18 . Possible risk factors for migraine include advanced age, female gender, low socioeconomic status, stress, sleep disorders, obesity, and excessive caffeine or medication usage 19 . While the exact pathogenic mechanism of migraine remains controversial, one proposed mechanism is neurogenic inflammation, which leads to cranial vasodilation, plasma protein extravasation, and the release of pro-inflammatory mediators 20 . The role of vitamin D—as an anti-inflammatory hormone and antioxidant—in the pathogenesis of migraine, as well as its potential benefits in migraine management, has been discussed in several studies 21 – 24 . Recent research has indicated that the prevalence of vitamin D deficiency and insufficiency is significantly higher in individuals with chronic migraine compared to healthy controls 21 . Moreover, several clinical trials have demonstrated that vitamin D supplementation can lead to a significant reduction in the frequency of headache attacks among migraine patients 25 – 29 . Several studies have explored the association between migraine and gynecological diseases (especially endometriosis and the polycystic ovary syndrome). Although uterine leiomyoma are some of the most common gynecological diseases in women of reproductive age, little is known about their association with migraine. Reports that vitamin D is involved in the pathophysiology of both uterine leiomyoma and migraine have prompted our interest in the link between these two. Therefore, this study aimed to investigate the association between uterine leiomyoma and migraine using a nationwide representative dataset. We hypothesized that migraine is associated with a higher incidence of uterine leiomyoma.