This study is the first study to focus on migraine comorbidity in female college students with PMS in China, aiming to elucidate the prevalence and influencing factors within this specific population. The study contributes to raising societal awareness of women’s health issues, facilitating early identification and intervention, and enhancing women’s quality of life. Simultaneously, it furnishes a scientific foundation for clinicians to refine their diagnostic and therapeutic strategies, possessing significant clinical and societal implications.
This study found that the prevalence of PMS among female college students was 21.6%, which is lower compared to previous studies conducted domestically and internationally. Studies have reported a prevalence of 30%–95% for PMS in women of childbearing age in developed nations, including Europe and the United28. In China, the prevalence of PMS is reported to be between 41% and 57%29. The observed result may be attributed to the varying screening scales employed and the unique characteristics of the studied population. Our study employed the Premenstrual Symptoms Screening Tool, whereas Xiaoding Zhou30 and Ays31 utilized the Bancroft Premenstrual Syndrome Scale. A separate survey conducted by Chinese scholar Mingqi Qiao employed an interview-based methodology29. The symptoms investigated in the latter two studies differed from those in our study, as did the screening criteria employed. Furthermore, our study primarily focused on college students, with an average age of 18.99 years, and 99% of participants were under 25, differing significantly from previous research29. The study identified age as a significant determinant of PMS incidence. The incidence of PMS increased with age within the range of 18 to 45 years32. Therefore, we conclude that the two aforementioned factors are the primary reasons for the relatively low occurrence of PMS among female college students.
In China, there is a scarcity of research on migraine among female college students, particularly focusing on those with PMS. This study revealed a significantly higher prevalence of migraine among female college students with PMS (18.8%) compared to those without PMS (5.7%). The discrepancies observed may be attributed to various factors within the context of PMS, including hormonal fluctuations, stress, and emotional state. Based on these results, we conducted an in-depth study to investigate the influencing factors.
This study identified sleep quality and anxiety as independent factors influencing migraine among female college students with PMS. The study demonstrated that a one-point increase in the PSQI score was associated with a 8% increase in the odds of developing migraine. On the one hand, sleep disorders can induce or exacerbate migraine symptoms by disrupting estrogen secretion. During PMS, the patient’s estrogen levels fluctuate significantly, leading to a complex physiological state. Fluctuations in estrogen levels may disrupt the balance of neurotransmitters, thereby inducing migraine attacks33. Prior research has demonstrated a positive correlation between 5-hydroxytryptamine (5-HT) levels in peripheral blood and estrogen levels. Additionally, calcitonin gene-related peptide (CGRP) levels fluctuate with estrogen levels34. Notably, a recent review conducted by Singh et al. highlighted that CGRP not only serves as a crucial mediator in migraine onset but also demonstrates significant female-selective effects. It induces migraines in females at lower doses and for longer durations compared to males. This reveals the crucial role of CGRP in explaining the higher prevalence of migraines in females, a mechanism highly pertinent to the female student population with PMS in our study35. However, sleep disorders are not unique to individuals with PMS; it is also a well-recognized risk factor for migraines in the general population. For example, a large-scale cross-sectional study carried out by Ray M. Merrill et al. revealed that sleep disorders in the general population could substantially elevate the risk of migraines36. This result aligns with our study’s findings, indicating that sleep quality universally affects migraines across diverse populations. For individuals with PMS, the decline in sleep quality further exacerbates estrogen fluctuations, creating a cumulative effect: poor sleep intensifies hormonal fluctuations37, thereby increasing the risk of migraines. This is where the differences between the PMS group and the general population become evident. Studies have established a correlation between sleep disorders and estrogen secretion. The mean estradiol (E2) level in young women with regular sleep (7.7 h per night) was 60% lower compared to those with irregular sleep (7 h per night)38. These findings further support the results of our study. On the other hand, sleep disorders can lower pain thresholds, whereas adequate sleep can elevate them. Animal studies have demonstrated that moderate sleep deprivation over a period of five consecutive days significantly enhances pain sensitivity in healthy mice, whereas augmenting sleep has been shown to ameliorate chronic pain39. An imaging study have demonstrated that sleep deprivation enhances pain responsiveness in key sensory regions of the cerebral cortex40. Sivertsen conducted a large cross-sectional population study and found a close correlation between the frequency and severity of insomnia and pain sensitivity. The study revealed that sleep deprivation significantly heightened pain sensitivity, whereas enhanced sleep led to a reduction in pain sensitivity41. Therefore, among female college students with PMS, poor sleep quality increases their risk of migraine and exacerbates PMS-related symptoms. Consequently, encouraging female college students with PMS to prioritize and enhance their sleep quality may mitigate the risk of migraine.
This study revealed that female college students with PMS and migraine exhibited significantly higher anxiety, depression, and stress scores compared to those with PMS but without migraine. Notably, after eliminating the influence of other confounding variables, we identified anxiety as an independent risk factor for migraine among female college students with PMS. Specifically, for each increment in anxiety level, the risk of experiencing a migraine increased by 24%. However, in this specific cohort, depression and stress exhibited no significant direct effect, necessitating further investigation into the underlying causes. Numerous studies have established a bidirectional relationship between migraine and anxiety, suggesting that anxiety elevates the risk of migraine, and migraine similarly increases the risk of anxiety42.. This is strongly supported by large-scale epidemiological data, which shows that individuals with any mental illness are 162% more likely to experience migraine, with anxiety being one of the most strongly associated conditions36. The potentiated risk may originate from shared underlying pathological mechanisms. Anxiety may originate from functional and structural modifications within the amygdala. Enhanced connectivity between the amygdala and anterior cingulate cortex (ACC) significantly heightens anxious individuals’ focus on environmental threat cues43. Furthermore, certain genetic predispositions to anxiety disorders are linked to heightened responsiveness of the amygdala and hippocampus to threatening cues. Furthermore, specific risk genotypes for anxiety disorders were associated with increased amygdala and hippocampus responsiveness to threat stimuli44. Recent research indicates that anxiety disorders and migraine exhibit shared molecular and cellular mechanisms regulating serotonin and glutamatergic neurotransmitter36 systems45. Therefore, common genetic factors may underlie the comorbidity of migraine and anxiety disorders. Further confirmation of this genetic overlap was provided by a study employing innovative statistical methods to elucidate its underlying basis36. Second, alterations in neurotransmitter and biochemical factor concentrations serve as common regulatory mechanisms and mediators for anxiety and migraine. For instance, serotonin dysfunction plays a crucial role in the pathogenesis of migraine and is also essential for maintaining and regulating normal mood46. Furthermore, several brain regions and their associated neurotransmitters in the ventricles and brainstem are implicated in the etiology of both anxiety disorders and migraine. In individuals with PMS, the common pathological mechanisms are exacerbated by periodic hormonal fluctuations. For example, reduced estrogen levels decrease the activity of the 5-hydroxytryptamine transporter (SERT), leading to lower 5-HT concentrations in the synaptic cleft. This alteration not only induces migraines34 but also worsens anxiety46. Anxiety, in turn, interacts with the hypothalamic–pituitary–adrenal (HPA)47 and hypothalamic-pituitary–gonadal (HPG) axes48, further exacerbating hormonal fluctuations and creating a “hormonal fluctuation-anxiety-migraine” synergistic amplification cycle. This is the primary reason for the stronger association between anxiety and migraines in the PMS population. The findings of this study offer significant insights for personalized migraine prevention strategies. Hence, clinical practitioners should pay more attention to the anxiety among female college students with PMS. For the key population, we recommend a range of interventions, including behavioral strategies, psychological counseling, and, if necessary, pharmacological treatment, to alleviate anxiety and prevent the onset and progression of migraine.
In this study, we opted for a nomogram to predict the migraine risk in female college students with PMS instead of a simple scoring model or risk index. The key reason is that the nomogram better meets the clinical and public health requirements of this population and can address the shortcomings of other tools. This nomogram serves as a user-friendly visualization tool. It integrates the identified independent risk factors, namely sleep quality and anxiety, to calculate the personalized probability of migraine risk. This intuitive format satisfies the practical needs of front-line staff in student health services, as these situations frequently entail time limitations and differing levels of staff familiarity with statistical models. Clinical practitioners specializing in student health, including school doctors and mental health counselors, frequently require rapid assessment of migraine risk during routine work, such as when students present with symptoms of PMS. The nomogram obviates the need for manual calculations or dependence on complex formulas. Staff can utilize a ruler or digital interface to align a student’s PSQI score and anxiety level with the corresponding scales on the nomogram. After summing the scores, they can promptly obtain a personalized risk probability for rapid evaluation of an individual student’s risk. This approach efficiently identifies high-risk individuals who are most likely to benefit from targeted interventions, such as sleep hygiene education or cognitive behavioral therapy for anxiety. This enables early prevention and personalized management within the campus medical system.
This study has several limitations. First, as a cross-sectional investigation employing questionnaire surveys, it is susceptible to reporting and recall biases. Specifically, in this study, PMS diagnosis relied on a retrospective self-assessment questionnaire, akin to the approach used in Vetvik et al.’s study49. It primarily assessed whether respondents self-reported experiencing PMS symptoms. However, biases may exist in the assessment of symptom severity, menstrual cycle regularity, and their functional consequences. Second, the lack of detailed investigation into specific migraine characteristics (e.g., attack frequency, severity, and duration) restricts a more thorough understanding of the disorder’s interplay with PMS, sleep, and anxiety. Furthermore, other potential confounders, such as physical activity levels, dietary habits, and academic stress, were not measured and could influence the onset of migraines. Third, the generalizability of our findings may be limited. Our sample was drawn solely from northeastern Sichuan, a region with unique socioeconomic and cultural characteristics. Although our multi-center sampling enhances internal validity, caution is needed when extrapolating prevalence rates to other regions of China. This regional specificity might also explain the relatively low Cronbach’s α coefficient for the ID-Migraine scale, warranting further validation in broader populations. Nonetheless, the core associations we identified between sleep quality, anxiety, and migraine are likely grounded in shared physiological mechanisms, thus still offering valuable insights for similar populations elsewhere.
Most importantly, the cross-sectional study design could not establish a definitive causal relationship between migraine and their risk factors, potentially resulting in reverse causation issues in the findings. While we interpret the findings as poor sleep quality and anxiety increasing the risk of migraine, reverse causality remains a plausible alternative explanation. For instance, it is equally possible that frequent migraine attacks themselves disrupt sleep patterns, leading to poor sleep quality, and/or that the experience of recurrent pain and disability contributes to the development or exacerbation of anxiety. Although multivariate logistic regression adjusted for potential confounders, it cannot resolve this inherent ambiguity regarding temporal sequence in cross-sectional data.
Future longitudinal studies are crucial to elucidate the temporal relationships and causal pathways, and for enhancing the precision of PMS diagnosis. Prospective cohort studies that track the onset of sleep disturbances, anxiety symptoms, and migraine attacks across multiple menstrual cycles could help determine which factor typically arises first. Additionally, interventional studies, such as randomized controlled trials treating insomnia or anxiety in students with PMS and assessing the subsequent impact on migraine incidence and severity, would provide stronger evidence for a causal role and inform effective management strategies.