Impact of monthly headache days on anxiety, depression and disability in migraine patients: results from the Spanish Atlas

Identifying highly disabled patients or at high risk of psychiatric comorbidity is crucial for migraine management. The burden of migraine increases with headache frequency, but the number of headache days (HDs) per month after which disability becomes severe or the risk of anxiety and depression is higher has not been established. Here, we estimate the number of HDs per month after which migraine is associated with higher risk of anxiety and depression, severe disability and lower quality of life. We analysed 468 migraine patients (mean age 36.8 ± 10.7; 90.2% female), of whom 38.5% had ≥ 15 HDs per month. Our results show a positive linear correlation between the number of HDs per month and the risk of anxiety (r = 0.273; p < 0.001), depression (r = 0.337; p < 0.001) and severe disability (r = 0.519; p < 0.001). The risk of anxiety is higher in patients having ≥ 3HDs per month, and those with ≥ 19HDs per month are at risk of depression. Moreover, patients suffering ≥ 10HDs per month have very severe disability. Our results suggest that migraine patients with ≥ 10HDs per month are very disabled and also that those with ≥ 3HDs per month should be screened for anxiety.

Variables. The questionnaire included variables related to healthcare utilization (diagnostic tests, medical visits, emergency visits, and hospital admissions), service utilization incurred by patients (visits to private specialists, and other complementary treatments for migraine) and data relating to patient work productivity losses in the previous year. In addition, three validated self-administered scales were used: HADS, MIDAS and HANA.
Patients were divided into four categories according to the frequency of HDs. Patients with 0-4 HDs per month were classified as low-frequency episodic migraine (LFEM), those with 5-8 HDs per month entered into the category of medium-frequency episodic migraine (MFEM), and when the number of HDs per month was between 9-14 HD per month the patients were classified as high-frequency episodic migraine (HFEM). Finally, patients with ≥ 15 HD per month were included into the CM group.
HADS is an instrument designed for screening potential anxiety and depression rather than grading the severity of the anxiety and depression in the general population. The HADS questionnaire included 14 items, seven of which evaluate anxiety (HADS-A) and seven that evaluate depression (HADS-D). Each item is scored on a scale of 0-3, resulting in an overall score of 0-21 for both HADS-A and HADS-D to detect possible cases of depression and anxiety. According to the score obtained in the HADS scale it is possible to distinguish between no case: 0-7; borderline case: 8-10, and case: 11-21 for both anxiety and depression 17 . The risk of anxiety and depression in HADS is considered from value 8 onwards 18 .
Additionally, disability was measured using the MIDAS scale, a 5-item questionnaire designed to evaluate disability within the past three months 19 . A score of 0-270 is used to indicate the overall level of disability due to headaches based on the following grading system: grade I, little or no disability (score of 0-5); grade II, mild disability (score of 6-10); grade III, moderate disability (score of [11][12][13][14][15][16][17][18][19][20]; and grade IV, severe disability (score of ≥ 21). The highest category is subdivided into grade IV-A, severe disability (scores of 21-40) and grade IV-B, very severe disability (scores of 41-270) 19 . www.nature.com/scientificreports/ HANA is a migraine-specific quality of life instrument measuring two dimensions of the chronic impact of migraine: frequency and bothersomeness. This scale contains the following seven domains: (i) anxiety/worry; (ii) depression/discouragement; (iii) self-control; (iv) energy; (v) function/work; (vi) family/social activities; and (vii) overall impact of migraine. The HANA validation studies confirmed internal consistency and reliability 20 .

Statistical analysis.
A descriptive analysis of the population was completed through the mean and standard deviation for quantitative variables and the size and percentage for qualitative variables.
The Mann-Whitney test was used to compare the homogeneity of distribution between the number of HDs and variables with two categories. The Kruskal-Wallis test was used to compare the homogeneity of distribution between the number of HD and variable with more than two categories.
Partial correlation was used to evaluate the association between the frequency of HDs and the risk of anxiety, depression (HADS), disability (MIDAS), and quality of life (HANA) controlling by age, gender, education level, smoking, alcohol, BMI and health insurance coverage.
A scatter plot was used to represent the means of HADS anxiety, HADS depression, MIDAS and HANA versus the frequency of HD per month. To this graphical representation we added the linear trend lines for both the total and each of the established categories (LFEM, MFEM, HFEM and CM).
Predictions were made from simple linear regression equations at 95% confidence. In addition, the mean number of visits to the emergency services due to migraine in the previous 12 months was 5.5 (± 11.0), while the mean number of hospital admissions due to migraine in the same period was 4.8 (± 19.5). 94.6% had public health coverage (only or in combination with private).

Results
The diagnostic delay in our sample was 6.6 years (SD = 6.6, Median = 5.0) and disease duration was 20.6 years (SD = 11.6, Median = 19.0). 81.8% of patients took analgesics or anti-inflammatory drugs during a migraine attack and 23.7% took antidepressants as preventive treatment.
According to the HADS scale, 69.0% were at risk of anxiety and 39.3% susceptible to depression. The average MIDAS score was 50.3 (± 52.6) and the average HANA score was 96.5 (± 34.5).
Headache frequency and the risk of anxiety, depression and severe disability. With respect to mental health, those at risk of anxiety reported a higher frequency of headache (13.2 vs 9.0; p < 0.001), as did those at risk of depression (14.8 vs 9.9; p < 0.001). The HDs frequency was higher in patients with severe disability (MIDAS Grade I: 7.3 vs Grade II: 5.2 vs Grade III: 6.8 vs Grade IV-A: 10.3 vs Grade IV-B: 16.5; p < 0.001) [ Table 2].
Those patients with higher frequencies of HDs had a higher HADS anxiety mean score (LFEM: 7.9, MFEM: 9.8, HFEM: 9.9 and CM: 11.1; p < 0.001) and HADS depression (LFEM  Table 3]. There was a positive linear correlation between the number of HDs and HADS anxiety (r = 0.273; p < 0.001), HADS depression (r = 0.337; p < 0.001), MIDAS scales (r = 0.519; p < 0.001) and HANA (r = 0.490; p < 0.001) and all are controlled by age, gender, education level, smoking, alcohol, BMI and health insurance coverage. Therefore, as the frequency of headaches increased, the risk of anxiety and depression increased, and the quality of life and disability of patients worsened [ Table 4].
There It was found that from the third day with a headache per month (MFEM), patients became part of the population at risk of anxiety, while the risk of depression appeared from day 19 (CM). From day 12 (HFEM), the values of the HANA were higher than the mean established at 96.5. Disability (MIDAS) was severe after suffering from three or more headache/days; and became very severe after 10 or more HDs/month.

Discussion
In the present study, the headache day threshold to identify migraine patients with severe disability is 10 HDs per month, and after 12 monthly HDs quality of life worsens above average. Our data are aligned with previous observations 4,5,25 which have suggested that the headache day threshold after which migraine is associated with severe disability, lower quality of life disability, and high healthcare utilization in patients with migraine is below 15 HDs per month. Although our analysis was not predefined using a multiple regression analysis, we found a similar threshold reported by Torres-Ferrús et al. 4 in a cohort of patients from a headache unit.
Anxiety and depression are common among migraine patients [8][9][10][11] and impact on migraine-related disability and quality of life 26 . In our sample, we found that after three HDs per month anxiety impacts on migraine patients www.nature.com/scientificreports/ and interfere in everyday life. Headache frequency was previously associated with higher scores on the HADS 27 . However, anxiety may impact patients with LFEM, as observed here. Contrary to our observation, Zebenholzer et al. 15 did not find an increased risk of anxiety in migraine patients with LFEM. Interictal anxiety is an important component of the burden of EM especially in those patients with severe migraine attacks 28 . The uncertainty of not knowing when the next migraine will occur and the severity of the attack may cause anxiety symptoms such as restlessness, persistent worrying and inability to concentrate, among others 29 . Moreover, migraine patients often endorse higher levels of neuroticism and are more likely than average to experience anxiety 30,31 . Early recognition and management of anxiety may be of great value to improve the life of migraine patients, and, according to our findings, screening of anxiety should be considered as part of the routine clinical evaluation in patients with migraine, including those with LFEM. In this respect, the use of preventive medication should be considered when the frequency of attacks per month is two or higher, but particularly in those patients with severe disability or comorbidities such as anxiety 6,32 . Additionally, we studied the number of HDs after which patients have a higher risk of depression. Surprisingly, the headache day threshold from which a migraine patient has a high risk for depression is 19 monthly HDs. Different studies have shown that the risk of depression is increased in patients with frequent migraine attacks 13,15,33 . Furthermore, depression is a risk factor for migraine chronification 34 . In contrast with our results, previous observations suggest that migraine patients are at risk of depression when the cut-off point is somewhat below 15 HDs per month 5,6,25 . Multiple reasons explain why the headache threshold is so different for anxiety and depression in the present study. First, depression onset is not associated with minor episodic stressors, but to chronic stress or major life events 35 . Nineteen days per month means that patients are spending more than half their life with pain and disability, and gives weight to the consideration of depression as secondary. That realization surely will lead to hopelessness and the inability to put in place appropriate pain-coping strategies that characterize depression in migraine patients 36 . Also, the relationship between migraine and depression is complex and bidirectional 9 . Migraine may aggravate depressive symptoms and is associated with a lower improvement in mood 37 . Finally, we cannot exclude that the high headache day threshold for depression observed here is because the HADS scale only captures a significant change in those patients in whom depressive symptoms are very severe.
Among the strengths of the present study was the use of a large sample including 468 migraine patients who were evaluated using validated scales to measure anxiety, depression, quality of life and disability. However, this study is subject to several limitations. First, the diagnosis of anxiety and depression was based on the patient's response and self-reported validated questionnaires and not in-person interviews, so the clinical diagnosis cannot be verified. In addition, although HADS is widely used for detecting anxiety and depressive disorders 12 , this scale is a screening tool rather than a diagnostic test. Another limitation is that the survey was completed through an online platform, and respondents had to have access to internet, as well as knowledge of technology. In addition, www.nature.com/scientificreports/ a high proportion of patients were invited to participate in the survey through the association AEMICE, so patients with severe forms of migraine may be over-represented. The increase in the number of HDs per month is associated with high disability, and a decrease in the quality of life. Anxiety may have a significant impact on migraine patients when the number of HDs is relatively low whilst depression strikes later within the headache day threshold. Our results support the need to redefine the diagnostic criteria of CM to include those patients with less than 15 monthly HDs but with high disease burden and treatment needs.

Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.