Effect of Galcanezumab on PHQ-9 and GAD-7 in Patients with Migraine: A Real-world Study in Japan

Abstract

Objective

Patients with migraine are at increased risk of developing psychiatric comorbidities, particularly depression and anxiety. Recently, anti-calcitonin gene-related peptide monoclonal antibodies (CGRP mAbs) have been suggested to be effective not only for migraine but also for comorbid depression. However, their effectiveness in depression and anxiety remains inconclusive, particularly in Asian populations. This study evaluated the effect of galcanezumab on depressive and anxiety symptoms in patients with migraine.

Methods

We conducted a single-center, retrospective, real-world study of 27 Japanese patients with migraine who received galcanezumab at Keio University Hospital, using the Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7) scores at two time points (baseline and 3-5 months after initiating galcanezumab) to evaluate depressive and anxiety symptoms.

Results

During galcanezumab treatment, the mean monthly migraine days (MMD) significantly decreased from 13.6±8.2 to 8.5±9.3 (p＜0.001). PHQ-9 scores significantly decreased from 5.8±4.4 to 4.5±4.2 (p=0.010) and GAD-7 scores from 4.0±3.9 to 3.1±3.5 (p=0.030). No significant correlation was observed between changes in PHQ-9/GAD-7 scores and changes in MMD, numerical rating scale scores, or associated symptoms. Furthermore, Fisher's exact tests showed no significant association between improvements in PHQ-9/GAD-7 scores and MMD reduction.

Conclusion

Galcanezumab improved PHQ-9 and GAD-7 scores in Japanese patients with migraine. These improvements may be partly attributable to the direct effects of galcanezumab, independent of its effect on migraine frequency.

Introduction

Migraine is one of the most prevalent neurological disorders that significantly affect patients' well-being and quality of life. Numerous studies have demonstrated that individuals with migraine are at increased risk of developing psychiatric comorbidities, particularly depression (1) and anxiety (2,3). Conversely, comorbid depression and anxiety can reduce the effectiveness of both acute and preventive migraine treatments (4-6), highlighting the importance of managing these psychiatric conditions in patients with migraine.

The development of anti-calcitonin gene-related peptide monoclonal antibodies (CGRP mAbs) has revolutionized prophylactic treatment options for patients with migraine refractory to conventional therapies (7-9). As of January 2025, three classes of CGRP mAbs had become available in Japan: those targeting the CGRP ligand (galcanezumab and fremanezumab) and those targeting the CGRP receptor (erenumab) (8,10-12).

Recent studies have suggested that CGRP mAbs may be beneficial for not only migraine but also comorbid depression. Improvements in depressive symptoms, as measured by the Patient Health Questionnaire-9 (PHQ-9), a nine-item self-report scale for assessing depression severity, have been reported in patients treated with CGRP mAbs (13-16). However, whether or not these improvements are statistically significant remains unclear. Furthermore, it remains to be determined whether the improvement in depressive symptoms is independent of the overall migraine treatment response. In addition, real-world evidence regarding the effects of CGRP mAbs on anxiety symptoms is limited. Although some studies have assessed anxiety using the Generalized Anxiety Disorder-7 (GAD-7) scale or Hamilton Anxiety Rating Scale (HARS), the findings remain inconclusive (15,17,18).

The present study evaluated the effectiveness of galcanezumab in alleviating depressive and anxiety symptoms in Japanese patients with migraine using real-world data from the PHQ-9 and GAD-7 assessments. We focused on galcanezumab because it was the first CGRP mAb to become available in Japan and was most widely adopted in the institutional setting during the study period. To our knowledge, this is the first study in an Asian population to examine the effect of CGRP mAbs on both depressive and anxiety symptoms in a real-world setting.

Materials and Methods

Study design and ethics

This single-center, observational, retrospective cohort study was approved by the Ethics Committee of Keio University School of Medicine, Tokyo, Japan (approval number: 20211144). The patients were informed of this observational study via the institute's website and could choose to opt out of the study. The requirement for informed consent was waived by the Ethics Committee of the Keio University School of Medicine in accordance with national regulations (Ethical Guidelines for Medical and Biological Research Involving Human Subjects). All study procedures were performed in accordance with the relevant guidelines and regulations.

Inclusion criteria

The inclusion criteria were as follows: age ≥18 years, Asian descent, a confirmed migraine diagnosis (including probable migraine) based on the International Classification of Headache Disorders, 3rd Edition (ICHD-3), 3- to 5-month treatment with galcanezumab (dose: 240 mg/120 mg/120 mg) as the first CGRP mAb treatment from the Headache Group of Keio University Hospital, receipt of the first CGRP mAb dose between August 2021 (when the drug became available at the hospital) and January 2023, and PHQ-9 and GAD-7 assessments between 3 and 5 months after initiating galcanezumab. Patients were diagnosed with migraine by a headache specialist. In our practice, PHQ-9 and GAD-7 are routinely assessed at the initiation of CGRP mAb treatment. Therefore, we included patients who later switched to another CGRP mAb. The timing of the switch was determined through discussions between the patient and physician (19-21).

Data collection

We retrospectively collected demographic data (age, sex, and body mass index), the medical history, the family history of headache, age at headache onset, migraine characteristics (unilateral and pulsating pain), and presence of aura. The headache specialist explained the ICHD-3-based criteria for migraine diagnosis to all patients, who then tracked headache and migraine days (including probable migraine days) in their headache diaries. Based on their diary entries, patients completed a questionnaire on monthly migraine days (MMD), monthly headache days (MHD), monthly acute medication intake days, pain intensity [0-10; numerical rating scale (NRS)], and associated symptoms during headache attacks (photophobia, phonophobia, and nausea/vomiting, rated as none, mild, moderate, or severe, respectively) at each dose administration. A month was defined as 28 days. A headache specialist interviewed each patient and periodically reviewed their headaches to verify the questionnaire's completeness and accuracy. According to the ICHD-3, patients were classified as having episodic migraine or chronic migraine (CM). Patients completed the PHQ-9 (22) and GAD-7 questionnaires (23,24) to determine depression and anxiety levels, respectively. Furthermore, we collected data on migraine preventive medications (lomerizine, propranolol, valproate, amitriptyline, and topiramate), including any treatment failures before CGRP mAb treatment.

Statistical analyses

All statistical analyses were performed using the IBM SPSS Statistics software program, ver. 29.0.2.0 (IBM, Armonk, USA). Continuous variables were summarized as the mean±standard deviation (SD), except for Figure, where they were presented as the mean±standard error. The Wilcoxon's signed-rank test was used to compare the pre- and post-treatment scores. Pearson's correlation coefficients were calculated, and a test of no correlation was performed to assess the relationship between the PHQ-9 and GAD-7 scores and clinical parameters, including MMD, NRS, and associated symptoms (photophobia, phonophobia, nausea/vomiting). Associations between improvements in PHQ-9 and GAD-7 scores and treatment responsiveness (≥25%, ≥50%, and ≥75% reduction in MMD) were evaluated using Fisher's exact test. The associations with the use of antidepressants or benzodiazepines were similarly assessed. The Mann-Whitney U test was conducted to evaluate the association between the number of prior treatment failures and improvements in PHQ-9 and GAD-7 scores. Statistical significance was set at a p value of ＜0.05.

Figure.

Effect of galcanezumab on MMD, PHQ-9, and GAD-7 scores. During galcanezumab treatment, MMD (A), PHQ-9 (B), and GAD-7 (C) scores significantly decreased (p<0.001, p=0.010, p=0.030, Wilcoxon’s signed-rank test, respectively). Among the PHQ-9 subdomains, only the somatic factor (sleep, fatigue, and appetite) showed a significant reduction in total subdomain score (p=0.008, Wilcoxon’s signed-rank test) (D). Although not statistically significant, both the emotional and physical subdomains of the GAD-7 showed a tendency toward score reduction (p=0.069 and p=0.079, respectively, Wilcoxon’s signed-rank test) (E). Data are presented in the graphs as mean ± standard error.

Results

Participant demographics and baseline parameters

We identified 72 patients who received galcanezumab as the first CGRP mAb treatment before January 2023. Of these, 27 were included in our analysis because they underwent PHQ-9 and GAD-7 assessments between 3 and 5 months after initiating galcanezumab treatment. As the PHQ-9 and GAD-7 are routinely assessed at the initiation of CGRP mAb treatment at our institution, our cohort included patients who later switched to another CGRP mAb. Specifically, 8 patients switched to erenumab, and 19 switched to fremanezumab as their second treatment. For one patient, the PHQ-9 and GAD-7 were assessed 2 months after receiving the third and final doses of galcanezumab.

Baseline characteristics before initiating galcanezumab administration are presented in Table 1. Most patients were women (85%), with a mean age of 45.6±12.6 years old. The average age at migraine onset was 22.6±10.5 years old. The mean numbers of MMD and MHD were 13.6±8.2 and 16.5±8.3, respectively. Furthermore, 5 patients (19%) had migraine with aura, while 16 (59%) had CM.

Table 1.

Patient Demographic and Clinical Characteristics.

	n=27
Patient characteristics
Age, mean±SD	45.6±12.6
Onset age	22.6±10.5
Male	4, 15%
Female	23, 85%
Family history	13, 48%
Body mass index	21.0±2.9
Migraine characteristics
Unilateral pain	17, 63%
Pulsating pain	16, 59%
Aura	5, 19%
CM	16, 59%
MHD	16.5±8.3
MMD	13.6±8.2
AMD	8.9±6.9
NRS	6.3±1.3
Associated symptoms
Photophobia	21, 78%
Phonophobia	18, 67%
Nausea/Vomiting	20, 74%
Total prior treatment failures	1.7±1.1
History of psychiatric disorders	10, 37%
Depression	5, 19%
Adjustment disorder	1, 4%
Panic disorder	1, 4%
Insomnia	3, 11%
Narcolepsy	1, 4%
Mental health
PHQ-9 total score	5.8±4.4
GAD-7 total score	4.0±3.9

CM: chronic migraine, MHD: monthly headache days, MMD: monthly migraine days, AMD: acute medication intake days, NRS: numerical rating scale

Most patients exhibited associated symptoms: 21 (78%) had photophobia, 18 (67%) had phonophobia, and 20 (74%) experienced nausea and/or vomiting. The total number of prior treatment failures was 1.7±1.1. In addition, 10 patients (37%) had a history of psychiatric disorders: 5 (19%) had depression, 1 (4%) had adjustment disorder, 1 (4%) had panic disorder, 3 (11%) had insomnia, and 1 (4%) had narcolepsy. Among them, 6 patients were taking antidepressants, and 6 were taking benzodiazepines. For the overall study cohort, the baseline PHQ-9 and GAD-7 scores were 5.8±4.4 and 4.0±3.9, respectively.

To examine the correlation between PHQ-9/GAD-7 and MMD/NRS at baseline, Pearson's correlation coefficients were calculated to perform a test of no correlation (Supplementary material 1). The correlation coefficients for the PHQ-9 with MMD and NRS were 0.130 (p=0.518) and 0.254 (p=0.202), respectively, while for GAD-7, they were 0.088 (p=0.664) and 0.316 (p=0.108), respectively. Thus, no significant correlation was observed between these parameters at baseline.

Effects of galcanezumab on MMD, PHQ-9, and GAD-7 scores

The comparison of MMD, PHQ-9, and GAD-7 scores before initiating galcanezumab and upon completion of galacanezumab treatment is presented in Figure A-C. During galcanezumab treatment, the mean number of MMD significantly decreased from 13.6±8.2 to 8.5±9.3 (p＜0.001, Wilcoxon's signed-rank test). In addition, PHQ-9 scores significantly decreased from 5.8±4.4 to 4.5±4.2 (p=0.010, Wilcoxon's signed-rank test). Similarly, GAD-7 scores significantly decreased from 4.0±3.9 at baseline to 3.1±3.5 after completing galcanezumab (p=0.030, Wilcoxon's signed-rank test).

To explore which symptom subdomains contributed to improvements in PHQ-9 and GAD-7 scores, we categorized the PHQ-9 items into four factors based on a previously established model (25): affective (anhedonia + depressed mood; items 1 and 2), somatic (sleep + fatigue + appetite; items 3, 4, and 5), internalizing (worth/guilt + suicidality; items 6 and 9), and sensorimotor (concentration + psychomotor; items 7 and 8). Items 1-9 in the English version of the PHQ-9 correspond to items A-I in the Japanese version. The GAD-7 items were similarly grouped into two factors: emotional experience of anxiety (items 1, 2, 3, and 7) and physical experience of restlessness (items 4, 5, and 6) (23). Among PHQ-9 subdomains, the somatic factor (items 3, 4, and 5) showed a significant reduction in total score, from 3.5±1.9 to 2.6±1.9 (p=0.008, Wilcoxon's signed-rank test) (Figure D). In contrast, the affective, internalizing, and sensorimotor subdomains did not show statistically significant changes (p=0.439, 1.000, and 0.107, respectively; Wilcoxon's signed-rank test). Although not statistically significant, both the emotional and physical subdomains of the GAD-7 showed a tendency toward score reduction, from 2.4±2.4 to 1.9±2.3 (p=0.069) and from 1.6±1.8 to 1.3±1.4 (p=0.079), respectively (Wilcoxon's signed-rank test) (Figure E).

A correlation analysis between PHQ-9/GAD-7 and clinical parameters

Next, we examined the correlation between changes in PHQ-9/GAD-7 scores and changes in MMD, NRS, and associated symptoms during galcanezumab treatment. Associated symptoms, including photophobia, phonophobia, and nausea/vomiting, were rated as none (0), mild (1), moderate (2), or severe (3), and symptom severity changes were calculated accordingly. Pearson's correlation coefficients were calculated to test of no correlation (Table 2). For changes in PHQ-9 scores, the correlation coefficients with changes in MMD, NRS, photophobia, phonophobia, and nausea/vomiting were 0.291 (p=0.141), -0.163 (p=0.415), -0.110 (p=0.593), -0.159 (p=0.439), and 0.003 (p=0.988), respectively; for GAD-7 scores, they were 0.121 (p=0.548), 0.029 (p=0.884), -0.163 (p=0.426), -0.080 (p=0.699), and 0.092 (p=0.654), respectively. Thus, no significant correlation was observed between changes in PHQ-9/GAD-7 scores and changes in MMD, NRS, or associated symptoms during galcanezumab treatment.

Table 2.

Correlation Analysis between PHQ-9/GAD-7 and Clinical Parameters.

Variables Compared (n=27)		Correlation (r)	p value	95% CI (Lower)	95% CI (Upper)
PHQ-9	MMD	0.291	0.141	-0.100	0.604
	NRS	-0.163	0.415	-0.512	0.231
	Photophobia	-0.110	0.593	-0.477	0.290
	Phonophobia	-0.159	0.439	-0.514	0.244
	Nausea/Vomiting	0.003	0.988	-0.385	0.390
GAD-7	MMD	0.121	0.548	-0.271	0.479
	NRS	0.029	0.884	-0.354	0.405
	Photophobia	-0.163	0.426	-0.518	0.239
	Phonophobia	-0.080	0.699	-0.453	0.317
	Nausea/Vomiting	0.092	0.654	-0.306	0.463

The association between PHQ-9/GAD-7 and responsiveness to galcanezumab

To assess whether the improvement in PHQ-9/GAD-7 scores within each severity group was associated with responsiveness to galcanezumab, we performed Fisher's exact test (Table 3). PHQ-9 severity was categorized as minimal (0-4), mild (5-9), moderate (10-14), moderately severe (15-19), and severe (≥20), whereas GAD-7 severity was classified as minimal (0-4), mild (5-9), moderate (10-14), or severe (≥15). Severity improvement was defined as a reduction in the severity level, and patients with minimal scores (0-4) were excluded from the analysis. Fisher's exact test revealed no significant association between PHQ-9 severity improvement and 25% response (≥25% reduction in MMD) (n=15, p=1.000). Similarly, no significant association was found for 50% (p=1.000) or 75% (p=1.000) response. Likewise, GAD-7 severity improvement was not significantly associated with 25% (n=10, p=0.500), 50% (p=1.000), or 75% (p=1.000) response.

Table 3.

Association between Improvement in PHQ-9/GAD-7 Severity Categories and Response to Galcanezumab.

Variables Compared		p value
PHQ-9 severity improvement (n=15)	25% Responder	1.000
	50% Responder	1.000
	75% Responder	1.000
GAD-7 severity improvement (n=10)	25% Responder	0.500
	50% Responder	1.000
	75% Responder	1.000

Although the minimal clinically important differences for PHQ-9 and GAD-7 are generally considered to be 5 and 4 points, respectively (26,27), we conducted an exploratory analysis to examine whether even minimal symptom improvement (≥1-point improvement) was associated with response to galcanezumab (Table 4). Patients with a baseline score of 0 were excluded from the analysis. Fisher's exact test revealed no significant association between ≥1-point improvement in PHQ-9 scores and 25% (p=0.635), 50% (p=0.073), or 75% response (p=0.058). Similarly, a ≥1-point improvement in GAD-7 scores was not significantly associated with 25% (p=0.621), 50% (p=0.400), or 75% (p=0.657) response.

Table 4.

Association between ≥1-point Improvements in PHQ-9/GAD-7 Scores and Response to Galcanezumab.

Variables Compared		p value
PHQ-9 improvement (≥1 point, n=26)	25% Responder	0.635
	50% Responder	0.073
	75% Responder	0.058
GAD-7 improvement (≥1 point, n=23)	25% Responder	0.621
	50% Responder	0.400
	75% Responder	0.657

Changes in PHQ-9 and GAD-7 by MMD responder group

To visualize PHQ-9/GAD-7 changes during galcanezumab treatment, we plotted the changes in PHQ-9 and GAD-7 scores by MMD responder group (Supplementary material 2). Among 50-100% responders, PHQ-9 scores showed a decreasing trend, whereas no clear pattern was observed among 0-49% responders. In contrast, changes in GAD-7 scores did not show any consistent trend within each responder group.

Association between PHQ-9/GAD-7 and medication use

Next, we explored the association between medications other than galcanezumab and improvements in PHQ-9/GAD-7 severity categories (Table 5). First, the total number of prior treatment failures was not significantly associated with improvements in PHQ-9 severity (p=0.463, Mann-Whitney U test) or GAD-7 severity (p=1.000, Mann-Whitney U test). The use of antidepressants was not significantly associated with improvements in PHQ-9 severity (p=0.315) or GAD-7 severity (p=1.000), as assessed using Fisher's exact test. Similarly, the use of benzodiazepines was not significantly associated with improvements in PHQ-9 (p=0.608) or GAD-7 (p=1.000) severity (Fisher's exact test).

Table 5.

Association between Medication Use and Improvement in PHQ-9/GAD-7 Severity.

Variables Compared		p value
PHQ-9 severity improvement (n=15)	Total prior treatment failures	0.463
	Antidepressants	0.315
	Benzodiazepines	0.608
GAD-7 severity improvement (n=10)	Total prior treatment failures	1.000
	Antidepressants	1.000
	Benzodiazepines	1.000

Discussion

In this study, PHQ-9 scores significantly decreased during galcanezumab treatment, consistent with previous studies suggesting that CGRP mAbs are effective for not only migraine but also comorbid depression. For example, a post-hoc analysis of the HALO-CM trial reported reductions in PHQ-9 scores after 12 weeks of fremanezumab treatment in patients with moderate to severe depression (PHQ-9 ≥10), with changes of -10.9 (quarterly dosing) and -9.8 (monthly dosing); however, these differences from placebo were not statistically significant (16). In contrast, the FOCUS trial demonstrated significantly greater PHQ-9 score improvement with monthly fremanezumab than with placebo (14). Similarly, treatment with erenumab resulted in a significant reduction in PHQ-9 scores from 16.5 at baseline to 8.5 after 11 months (p=0.001) (13).

However, patients in these studies had higher baseline PHQ-9 scores than those in the present study. In our cohort, the baseline PHQ-9 score was 5.8±4.4, with 44% of participants (n=12) classified as having minimal depressive symptoms (PHQ-9 score of 0-4). Although most of our cohort exhibited minimal to moderate depression, galcanezumab was still effective. Nonetheless, the mean reduction in PHQ-9 was approximately 1 point, which is smaller than the reductions reported in most previous studies, probably owing to the relatively low baseline PHQ-9 scores in our population. Furthermore, the CONQUER study reported a significantly greater reduction in PHQ-9 scores with galcanezumab than with placebo (-2.1, -1.2; p=0.009) in patients with migraine (15). In that study, the proportion of patients with possible major depressive disorder (PHQ-9 ≥10) at baseline was 12.9% compared with 19% (n=5) in our cohort. These findings support the effectiveness of galcanezumab even in patients with relatively mild depressive symptoms.

Our study also demonstrated a significant reduction in GAD-7 scores following galcanezumab treatment. However, the CONQUER study did not report significant changes in GAD-7 scores compared to placebo during galcanezumab treatment (15). In this study, the mean reduction in GAD-7 scores from baseline was 0.9, which aligns with the changes observed in the placebo groups of the CONQUER study (15), suggesting that the improvement in GAD-7 scores may be partly due to a placebo effect. In contrast, other studies using HARS have reported significant improvements in anxiety symptoms following treatment with galcanezumab (17) and erenumab (18). These findings suggest that the sensitivity of anxiety measurement tools may influence the detection of the effects of treatment on anxiety symptoms.

Furthermore, improvements in PHQ-9 and GAD-7 scores in our study were not significantly correlated with changes in MMD, NRS scores, or associated migraine symptoms. Moreover, Fisher's exact tests revealed no significant association between improvements in PHQ-9/GAD-7 severity/scores and responsiveness to galcanezumab, as defined by MMD reduction. These results suggest that galcanezumab may improve depressive and anxiety symptoms independently of its effects on migraine frequency.

Supporting this hypothesis, a study assessing the Hospital Anxiety and Depression Scale-Depression subscale (HADS-D) reported that treatment with CGRP mAb was associated with reductions in depressive symptoms independent of MMD reduction (28). Similarly, reductions in depressive symptoms measured using the Beck Depression Inventory-II (BDI-II) were independent of migraine frequency improvements (29). These findings support the idea that CGRP mAbs may have a direct effect on depressive symptoms independent of migraine control, which is biologically plausible, as both migraine and major depressive disorder are associated with elevated CGRP levels (30-32). Thus, CGRP mAbs may influence both migraine and depressive symptoms through shared neurobiological pathways.

A subdomain analysis of the PHQ-9 revealed that galcanezumab specifically reduced scores in the somatic subdomain, with no significant effects observed in the affective, internalizing, or sensorimotor subdomains. Notably, baseline scores in the somatic subdomain were higher than those in the other subdomains, with a maximum mean value of approximately 1 point, limiting the potential for further reduction. Although a higher baseline severity may be necessary to detect subdomain-specific changes, these findings suggest that galcanezumab primarily alleviates somatic symptoms associated with depressive distress. In contrast, in the GAD-7 analysis, both the emotional experience of anxiety and physical experience of restlessness showed a decreasing trend, although the changes were not statistically significant. This may reflect a modest and broadly distributed reduction in anxiety symptoms rather than a subdomain-specific effect.

Interestingly, although a ≥1-point improvement in PHQ-9 scores was not significantly associated with 50% or 75% response, a trend toward a positive association was observed, as shown in Table 4 and Figure S1. This finding suggests a possible nonlinear relationship between the response to galcanezumab and minimal improvement in depressive symptoms, rather than substantial clinical improvement. Furthermore, among the PHQ-9 subdomains, only somatic factors (sleep, fatigue, and appetite), which may be directly influenced by migraine itself, showed a significant reduction in the total subdomain score. This suggests that improvements in migraine symptoms may partly contribute to a reduction in PHQ-9 scores, in addition to any potential direct effect of galcanezumab on depressive symptoms. However, definitive conclusions could not be drawn because of the small sample size.

However, evidence regarding the effects of CGRP mAbs on anxiety symptoms remains limited. CGRP has been implicated in anxiety pathophysiology (33,34), and our real-world data provide additional insights into the potential role of CGRP mAbs in modulating anxiety symptoms.

To our knowledge, this is the first Asian study to examine the effects of CGRP mAbs on depressive and anxiety symptoms, which is particularly important because the association between migraine and psychiatric comorbidities may vary across ethnic groups. For example, genetic studies have demonstrated that the genetic architecture of depression differs between East Asian and European populations (35). Furthermore, migraine epidemiology in the Asia-Pacific region differs from that in Europe (36). Moreover, migraine and depression share common genetic risk factors (37,38), suggesting that the interaction between these conditions may be influenced by ethnicity-specific genetic variations.

Several limitations associated with the present study warrant mention. First, the sample size was relatively small compared with that of previous studies, limiting the statistical power to detect significant associations. Second, most of our cohort had minimal or moderate depressive and anxiety symptoms at baseline, which may have reduced their ability to observe significant treatment effects. Third, concomitant medications, such as antidepressants or anxiolytics, may have influenced the observed changes in psychiatric symptoms, introducing potential confounding effects. Furthermore, selection bias is a potential concern because our cohort included patients who later switched to another CGRP mAb. Galcanezumab was the first CGRP mAb introduced in Japan in 2021, with fremanezumab and erenumab becoming available several months later (19,21). During this transition, some patients switched to fremanezumab, including responders, to reduce hospital visits owing to its quarterly administration. Therefore, the potential bias introduced by switching was expected to be minimal, as our cohort included both responders and non-responders. However, another source of selection bias may persist; patients who experienced favorable outcomes in terms of headache, depression, or anxiety symptoms were more likely to continue treatment and complete follow-up assessments.

In conclusion, galcanezumab appears to be effective for not only migraine symptoms but also depressive and anxiety symptoms in Japanese patients, potentially independent of migraine improvement. Our findings highlight the need for larger studies focusing on Asian populations, particularly Japanese populations, to further investigate the effects of CGRP mAbs on psychiatric comorbidities, which could have important implications in comprehensive migraine management.

The patients were informed of this observational study via the institute's website and could choose to opt out of the study. The requirement for informed consent was waived by the Ethics Committee of the Keio University School of Medicine in accordance with national regulations (Ethical Guidelines for Medical and Biological Research Involving Human Subjects).

Author’s disclosure of potential Conflicts of Interest (COI).

Tsubasa Takizawa is a consultant/advisor and/or serves as an advisory board for Eli Lilly, Otsuka, Amgen, Pfizer, and Teijin, and has received speaker honoraria from Eli Lilly, Daiichi Sankyo, Otsuka, Amgen, Kowa, Kyowa Kirin, Eisai, UCB Japan, Takeda, Sawai, and Santen Pharmaceutical and grants from Eli Lilly, Pfizer, and Tsumura outside the submitted work. Jin Nakahara has received honoraria and research scholarships from Amgen and Daiichi Sankyo.

Kota Oshima and Yuna Hattori contributed equally to this work.

Supplementary Material

Table S1

Correlation between PHQ-9/GAD-7 and MMD/NRS at baseline

Trend in PHQ-9 and GAD-7 score changes by MMD responder group

Among 50–100% responders, PHQ-9 scores showed a decreasing trend, whereas no clear pattern was observed among 0–49% responders. Changes in GAD-7 scores did not show a consistent trend across responder groups.

Funding Statement

This study was supported by JSPS KAKENHI (grant numbers 22K15693 and 24K10606 to TT) and Keio University Academic Development Funds for Joint Research to TT.