This study shows that systemic PACAP-38 concentrations in women may vary depending on the presence of migraine and on their hormonal status. Specifically, women with migraine and a RMC had higher interictal PACAP-38 concentrations across the menstrual cycle compared to healthy control participants. In contrast, women with migraine under COC treatment had similar PACAP-38 concentrations to those observed in healthy women (C-RMC and C-COC). Among women with migraine, PACAP-38 concentrations during the physiological menstruation were higher compared to the endometrial bleeding from COC withdrawal during HFI, which might point to a distinct impact of endogeneous and exogeneous sex hormones on PACAP-38 release.

Prior investigations into interictal PACAP concentrations in patients with migraine have yielded conflicting results. For instance, Cernuda-Morollón and colleagues reported no differences in total PACAP levels in plasma between women with migraine and healthy controls [32]. Two recent pediatric studies observed higher concentrations of PACAP-38 in plasma in children with migraine during the interictal period compared to their healthy counterparts [33, 34]. In contrast, some studies detected lower PACAP concentrations in interictal patients with migraine compared to healthy individuals [35, 36]. These contradictory findings may stem from differences in study populations, definitions of the interictal period, and measurement protocols [37]. Furthermore, none of these previous studies specifically evaluated the hormonal status of participants. Tuka et al. described similar interictal PACAP-38 concentrations in women with migraine reporting a “menstruation cycle sensitivity” compared to patients without, but they did not consider their specific hormonal profiles [36]. Our findings highlight the importance of considering the specific hormonal status when evaluating interictal PACAP-38 levels. Indeed, in our cohort, only women with a RMC, but not those undergoing COC treatment, exhibited elevated interictal PACAP levels. This underscores the intricate relationship between hormonal fluctuations and PACAP dynamics in women with migraine, emphasizing the need for nuanced consideration in future research endeavors.

To date, only one study unrelated to migraine has examined PACAP plasma levels in women with various hormonal profiles, including women with a RMC, during pregnancy, after menopause, and with hormonal disorders including premature ovarian failure (POF) and idiopathic hypogonadotropic hypogonadism (IHH) [38]. Women with a RMC were assessed during menstruation, while women with POF and IHH received COC treatment and were measured during the HFI. In this previous study, PACAP levels during the physiological menstruation were significantly higher than during the HFI, which closely mirrors our findings in the migraine groups, but not in the healthy groups. However, the comparability of our findings with those of this previous study is limited due to two main factors. First, it is unknown whether the women in the previous study had migraine or other headache disorders, making it challenging to assess how this potential confounding factor may have influenced the results. Second, the women with COC in the previous study had pathological conditions, namely POF and IHH, which inherently disrupt sex hormone levels. Future studies should aim to replicate our findings in populations specifically diagnosed with migraine, while also considering other hormonal statuses such as pregnancy or postmenopause.

In our cohort, we did not observe any significant correlation between the absolute concentrations of sex hormones and those of PACAP-38. This suggests that hormonal fluctuations, rather than the absolute hormone concentrations, may exert influence on PACAP-38 levels in plasma. While serial measurements throughout the menstrual cycle present methodological challenges, such an approach would offer valuable insights into the impact of hormonal changes on PACAP-38 concentrations. Furthermore, the absence of correlation between PACAP-38 and CGRP plasma levels, which were measured in our parent study, hints at potentially distinct hormone-dependent modulation of these pathways [12]. Nevertheless, despite the lack of correlation, the overall patterns of plasma PACAP-38 and CGRP levels exhibited similarities across groups and visits. Specifically, higher concentrations of both neuropeptides were consistently observed in women with migraine and a RMC, particularly during menstruation, while lower concentrations were noted in women with migraine undergoing COC treatment and in healthy controls. This observation underscores the need for a dedicated study to investigate the intricate relationship between PACAP-38, CGRP, and fluctuating sex hormones.

The measurement of plasma PACAP-38 levels is subject to several methodological limitations that warrant consideration when interpreting our findings. The wide array of preanalytical and measurement methods reported in the literature poses challenges for direct comparison with previous studies. However, our findings show comparable concentrations in plasma to previous studies involving children with migraine [34], patients with heart failure [39], and both pregnant and non-pregnant healthy women [40].

Moreover, it remains unclear what proportion of PACAP-38 in plasma originates from the trigeminovascular system versus other sources [41]. Given the multitude of processes influencing PACAP concentrations, it is difficult to ascertain whether these dynamics are specific to migraine or due to other confounding factors. Additionally, similar to CGRP, PACAP-38 has a short half-life in plasma (approximately 5 min) and is rapidly degraded by endopeptidases [41, 42]. In our preanalytical phase, efforts were made to stabilize PACAP-38 through rapid removal and storage as well as the addition of protease inhibitors [43]. Given the exploratory nature of our study, the sample size calculation was based on the original publication [12], and its applicability to PACAP-38 analysis may require further validation in future studies.

A further limitation of our study pertains to the definition of the interictal period, set as 12 h free of migraine, which is shorter than in other investigations and may introduce bias [44]. Nevertheless, pragmatically, extending the migraine-free period would have increased the likelihood of visit cancellations. Furthermore, while the peak incidence of migraine attacks occurs at day -1 of the menstrual cycle [45], our perimenstrual visits took place slightly later. This scheduling was a result of ensuring interictal measurements and practical considerations, such as participants informing us on the first day of menstruation and being unable to visit immediately. An earlier perimenstrual timepoint might provide different insights and should be considered in future studies. In addition, we did not specifically inquire about the presence or absence of aura during the perimenstrual attacks in patients who have had more than two attacks with aura in their lifetime and thus fulfilled the ICHD-3 criteria for migraine with aura. This could influence the interpretation of our results [46] and should be considered in future research. Moreover, we cannot precisely determine whether periovulatory visits occurred on the day of ovulation or in the days immediately preceding or following it. Future studies should incorporate ovulation tests to measure the LH surge and confirm the timing of ovulation.

In conclusion, this exploratory study provides evidence of a complex interplay between PACAP-38 and sex hormones in women with migraine. The elevated PACAP-38 levels observed in participants with migraine and a regular menstrual cycle might contribute to the higher susceptibility to migraine in menstruating women.

These findings underscore the need for further longitudinal research to elucidate the role of PACAP-38 in migraine pathophysiology and its modulation by hormonal changes.