Humanistic and economic burden of depression and anxiety among adults with migraine: A systematic review

Depression and anxiety are well‐recognized comorbid health conditions among adults with migraine due to their humanistic and economic burden. This review was conducted to systematically assess the humanistic and economic burden of comorbid depression and/or anxiety disorder among adults with migraine.

migraine-related disability than those with episodic migraine (Buse et al., 2012). Coexisting depression and anxiety among adults with migraine are not only a major clinical problem but they are also associated with higher disability (Lantéri-Minet, Radat, Chautard, & Lucas, 2005), and poor health-related quality of life (HRQoL; Lantéri-Minet et al., 2005;Lipton et al., 2000Lipton et al., , 2003. Moreover, populationbased studies conducted in the United States (US) and the United Kingdom found that comorbidity of migraine and depression each exerts a significant association with poor HRQoL in two domains; the mental and physical health component scores (Lipton et al., 2000).
Further, as the result of the combined disability, depression and anxiety is likely to increase healthcare costs and utilization (Bonafede et al., 2017;Pesa & Lage, 2004). Migraine is one of the most expensive chronic conditions, with an estimated annual cost of $17 billion in the US alone (Goldberg, 2005). Depression and anxiety among adults with migraine are common and can lead to significantly higher healthcare costs (Bonafede et al., 2017;Pesa & Lage, 2004).
Yet, depression and anxiety are largely unrecognized and untreated in adults with migraine (Hamelsky & Lipton, 2006). Unrecognized and untreated depression is partially attributed to the competing demand to treat migraine coupled with the overlap of the symptoms of depression and migraine (e.g., insomnia). The association between migraine and comorbid depression and/or anxiety appears to be bidirectional; some epidemiological studies suggest that adults with migraine are at increased risk for depression or anxiety when compared to subjects without migraine and vice versa (Dresler et al., 2019). Furthermore, biomedical evidence shows shared genetic causes for migraine and these mental disorders suggesting common genetic pathways (Dresler et al., 2019).
As the comorbidity between migraine and depression and/or anxiety is highly prevalent in the clinical practice and may influence both the humanistic and financial outcomes of affected individuals and healthcare systems, these health outcomes emphasize the importance of being carefully investigated. Therefore this has urged the need for this systematic review which aimed to provide a comprehensive overview of the current evidence focusing on the comorbidity of depression and anxiety disorders in adults with migraine to obtain a clearer picture of the humanistic and financial burden on those patients. By summarizing a large body of literature, it should also be possible to reach a broader understanding of both the mental illnesses burden and the most appropriate directions for future research and practice in this area. As a result, this study aimed to bring the research attention to comorbid depression and anxiety and their associated humanistic (disability and HRQoL) and the economic (healthcare cost and utilization) burden on patients with migraine. This review's findings can help the practitioner understand the importance of providing comprehensive care to patients with such complex conditions to reduce the burden on patients and the healthcare system. This systematic review questions can be summarized as 1. Does coexisting depression and/or anxiety associated with poor HRQoL and higher disability in adults with migraine? 2. Does coexisting depression and/or anxiety associated with higher healthcare costs and utilization of adults with migraine? 3. Have the questions above been sufficiently addressed in the literature?
This review attempted to identify all studies that covered the humanistic and economic outcomes of coexisting depression and anxiety among adults with migraine published in the English language in peer-reviewed journals from inception until January 2020.
The study PICOS question is described as follows: Population: migraine subjects without anxiety/depression; Interventions: noninterventional studies; Comparisons: migraine subjects with anxiety/ depression; Outcomes: humanistic (disability, HRQoL) or economic outcomes; and the Study design: cross-sectional studies.
Articles were included in the review if they met the following inclusion criteria: (a) cross-sectional studies; (b) adults with migraine (i.e., not pediatric populations), (c) assess either humanistic burden (disability, HRQoL) and/or economic burden (direct or indirect healthcare costs or healthcare utilization), (d) used validated measure to diagnose migraine, (e) used validated diagnostic measures of depression and anxiety with/out clinical or threshold score, and (f) used validated self-report measures of HRQoL and disability.
Exclusion criteria were (a) qualitative study design, (b) non-English language studies, (c) studies that have not used validated measures to assess (migraine diagnosis, mental health, HRQoL, and disability outcomes) were not included in this systematic review; and (d) conference abstracts and full-text articles without raw data available for extraction were excluded.

| Design
The present systematic review was reported according to the preferred reporting items for the systematic reviews and metaanalyses (PRISMA) guidelines (Moher, Liberati, Tetzlaff, Altman, & Group, 2009).

| Search strategy
The comprehensive systematic review presented below was based on a computerized systematic literature search conducted using three databases (MEDLINE and CINAHL via EBSCO, EMBASE, and Cochrane Database of Systematic Reviews via OVID). The search strategy used the Medical Subject Headings (MeSH) terms and keywords relating to depression, anxiety, migraine, disability, HRQoL, and healthcare cost and utilization for each database. The free test search included the following combinations of keywords: "anxiety OR depression" AND ALWHAIBI AND ALHAWASSI | 1147 "health-related quality of life (HRQoL)" OR "disability" OR "healthcare cost" OR "healthcare utilization" AND "migraine." For the controlled language search included the following exploded MeSH terms: "Depressive Disorder, major," "Anxiety Disorder," "Migraine Disorders," "Quality of Life," "Healthcare Costs," "Healthcare Expenditures" as recommended for each database. Other terms have also been used such as "mood disorder" or "HRQoL" or "SF-36" or "SF-12" or "disability or humanistic burden" or "direct medical cost" (title/abstract) or "indirect medical cost" (title/abstract) or "out of pocket costs" or "out of pocket expenses" or "economic burden." Appendix A shows the complete search strategy used in MEDLINE.
Articles were also identified by a manual search of bibliographies from all retrieved articles.
Abstracts and articles were read for relevance to the research questions. The database search was conducted independently by two review authors (M. A. and T. M. A.); any disagreement was resolved by consensus. The protocol for the systematic review was submitted for registration on the International Prospective Register of Systematic Reviews (PROSPERO).

| Study selection
The initial literature screening involved examining the titles and abstracts of extracted articles for potential relevance for inclusion independently by the study authors (M. A. and T. M. A.). Full-text studies potentially related to the review questions were then assessed against the inclusion/exclusion criteria independently by the study authors; any disagreements were discussed and resolved until consensus was reached. Figure 1 shows how the database search and article selection process was carried.

| Data extraction and analysis
The authors independently extracted the included data. Study characteristics (author and year of publication), methodological details (study design, location, study participants, sample size, mean age, and percentage women), main outcome measures, and findings were extracted from the eligible articles and entered into a Microsoft Excel® spreadsheet designed to facilitate data sorting and assessment of study quality.

| Quality assessment
The included studies were assessed for their quality using the Appraisal Tool for Cross-Sectional Studies (AXIS tool; Downes, Brennan, Williams, & Dean, 2016). The authors independently performed a quality assessment for each included study. This tool contains 20 criteria for three domains: (a) quality of reporting, (b) study design quality, and F I G U R E 1 PRISMA flow diagram for identification and selection of articles for inclusion in this review. HRQoL, health-related quality of life; PRISMA, preferred reporting items for the systematic reviews and meta-analyses (c) introduction of biases in the study. For each criterion met, the study received one point; the maximum score is 20, which indicates the highest quality of the included studies in the present review.

| Selection process
A total of 640 articles were identified from all searches. Based on the title and abstract screening, 48 potentially eligible full-text articles were retrieved. Of these, 23 studies had met the inclusion criteria and included in the final review ( Figure 1).

| General characteristics of included studies
The description of the included studies is displayed in Tables 1-3.

| Quality assessment
Overall, the included studies scored high for clearly stating the study aims, design, target population, study risk factors and outcomes measurement, result explanation, discussion, and conclusion justified by the results (Table 4). The main issues were mainly related to the failure to address the sample calculation or addressing the nonresponse rate (Table 4).

| MAIN FINDINGS
For this review, the results were categorized into a humanistic and economic burden.

| Depression and/or anxiety comorbidity
In the included studies, depression and anxiety were mainly evaluated using the Beck Depression Inventory (BDI) in four studies.
Other scales were also used, such as Hospital Anxiety and Depression Score (HADS), Patient Health Questionnaire-4 (PHQ-4), and the Center for Epidemiologic Studies-Depression Scale.

| Humanistic burden of comorbid depression and anxiety among adults with migraine
The humanistic burden was measured using the HRQoL and disability. HRQoL instruments evaluate the burden of a disease on a person's overall well-being (Theofilou, 2013). Thirteen studies gave data for the HRQoL outcome for adults with migraine and comorbid depression and/or anxiety (Table 1). The results from these studies showed that depression and/or anxiety were associated with poor HRQoL. HRQoL instruments used were either generic or diseasespecific. Most studies mentioned in this review assessed the HRQoL among individuals with migraine had used generic tools (Table 1) (Tables 1). A national population-based study in Canada using national data from the Canadian Community Health Survey evaluated the association between depression, anxiety, and HRQoL (Jette et al., 2008). This study found that among adults with migraine; those with depression or anxiety had lower HRQoL compared to those without these conditions. Studies in China, France, Germany, Italy, India, and South Korea have also demonstrated poor HRQoL among individuals with migraine and depression and/or anxiety (Kim & Park, 2014;Sharma et al., 2013;Wang et al., 2016). A nationwide population-based survey in France among subjects with migraine (n = 1,957), in which 28.0% had anxiety, 3.5% de-

| Economic burden of comorbid depression and anxiety among adults with migraine
The economic burden was measured using healthcare costs and healthcare utilization ( Abbreviations: ED, emergency department; ER, emergency room; ICD-9-CM codes, International Classification of Diseases, ninth revision, Clinical Modification.
a Healthcare expenditures were adjusted using the consumer price index and were inflated to the year 2012 constant dollars.
depression and or anxiety are significantly associated with healthcare utilization patterns. Wu et al. (2016) reported higher odds of emergency department visits in adults with migraine and comorbid depression than those without depression (OR = 1.36, 95% CI: 1.08, 1.71). This study suggested that depression increases healthcare utilization among adults with migraine; however, the association between comorbid depression and/or anxiety and healthcare utilization has not been evaluated yet.
(1) Were the aims/objectives of the study clear?
(2) Was the study design appropriate for the stated aim(s)?
(3) Was the sample size justified? (4) Was the target/reference population clearly defined? (is it clear who the research was about?) (5) Was the sample frame taken from an appropriate population base so that it closely represented the target/reference population under investigation? (6) Was the selection process likely to select subjects/participants that were representative of the target/reference population under investigation? (7) Were measures undertaken to address and categorize nonresponders? (8) Were the risk factor and outcome variables measured appropriate to the aims of the study? (9) Were the risk factor and outcome variables measured correctly using instruments/measurements that had been trialled, piloted, or published previously? (10) Is it clear what was used to determine statistical significance and/or precision estimates? (e.g., p-values, confidence intervals) (11) Were the methods (including statistical methods) sufficiently described to enable them to be rejected? (12) Were the basic data adequately described?

| DISCUSSION
The present review highlights the substantial disease burden of comorbid depression and/or anxiety among adults with migraine in terms of HRQoL, disability, and healthcare costs and utilization.
Consistent evidence from the included studies on the association between depression and/or anxiety comorbidity and higher humanistic and economic burden among adults with migraine was found.
Twenty studies were identified in this systematic review; predominately evaluated the humanistic burden while three studies had evaluated the economic outcomes of depression and anxiety among adults with migraine.
When the humanistic burden of depression and/or anxiety among adults with migraine was measured using either a generic or a diseasespecific measure of HRQoL, studies were consistent in documenting a relationship between the presence of either of these conditions and poor HRQoL (Brna et al., 2007;D'Amico et al., 2015;Dindo et al., 2015;Hung et al., 2008;Jette et al., 2008;Kim & Park, 2014;Kolotylo & Broome, 2000;Lantéri-Minet et al., 2005;Lipton et al., 2000;Prisnie et al., 2018;Sharma et al., 2013;Stuginski-Barbosa et al., 2012;Wang et al., 2016). Some studies highlighted that these psychopathological conditions were associated with both the mental and physical health component scores of HRQoL (Lipton et al., 2000).
However, most studies have found a significant association between these illnesses and the lower mental rather than the physical health component scores of HRQoL reinforcing that mental health is a substantial concern for patients with migraine (Brna et al., 2007;Dindo et al., 2015;Lantéri-Minet et al., 2005;Lipton et al., 2000;Sharma et al., 2013). Additionally, while most of the studies focused on depression comorbidity, anxiety usually coexists with depression yielding more negative outcomes than individuals with migraine and depression only. Therefore, future studies on the association between depression and anxiety comorbidity and HRQoL among individuals with migraine are warranted.
Further, when the burden of depression and/or anxiety was evaluated in terms of disability, studies reported a consistent relationship with higher disability among adults with migraine (Breslau et al., 2003;Dindo et al., 2015;Kolotylo & Broome, 2000;Lantéri-Minet et al., 2005;Rammohan et al., 2019;Seng et al., 2017). However, the definitions of disability varied among studies. Most of the included studies used the Migraine Disability Assessment scale. Only two studies had examined the relationship between anxiety and migraine-related disability (Seng et al., 2017), with none of the included studies had examined the association between anxiety and depression comorbidity and disability among adults with migraine. In all included studies, depression and anxiety were both associated with severe migraine-related disability. Lantéri-Minet et al. (2005) in their study among 1,834 adults with migraine using data from the FRAMIG 3, a national population-based survey for migraine in France, have found that the migraine-related disability was associated with anxiety-depression comorbidity but not with either disease alone. Consequently, there is a substantial association between depression and/or anxiety and the increased risk of disability urge for further multicenter, longitudinal and large sample size investigations and research to understand this association, and provoke factors.
The economic burden associated with the mental conditions (depressions and anxiety) and migraine among adults is a crucial finding of this study. The current evidence among adults with migraine reported that individuals with comorbid depression and/or anxiety have a higher economic burden in terms of healthcare costs and medical services utilization (Bonafede et al., 2017;Pesa & Lage, 2004). Only three studies estimated the excess economic burden of depression and/or anxiety among individuals with migraine. These studies found that direct medical care costs were higher for those with depression and/or anxiety. Given the higher economic and humanistic burden of disability and HRQoL among adults with migraine and the higher mental comorbidity, information on patient-mental health is poorly diagnosed at the office visits. Minen et al. (2016) in their study among primary care physicians (PCPs) have found that one-third of PCPs do not routinely assess depression, and about half of the physicians do not regularly screen for anxiety. The findings from this review support the need for regular screening tools for mental health as part of comprehensive clinical care for adults with migraine. Adults with migraine may benefit from routine mental health screening to capture the mental health of patients early before it may compromise the health outcomes of adults with migraine.

| Strengths and limitations
This systematic review utilized a vigorous search strategy to summarize the current knowledge of the humanistic and economic burden of coexisting depression and anxiety among adults with migraine. Although the included studies' qualities were evaluated, the results should be interpreted with caution considering some limitations from each included study. First, there was no consistency in defining cases of depression or anxiety from different studies. For example, some studies defined depression or anxiety as symptomology and others that have used a threshold to define probable cases of depression or anxiety. Second, several factors are known to influence both HRQoL and the healthcare costs and utilization such as the severity of migraine, migraine-related disability, and income have not been taken into account by adjusting for them in the analyses in many studies. These factors may confound the association between depression or anxiety and the humanistic and economic burden. However, among studies that adjusted for these confounders, the association between depression and anxiety and the health outcomes for adults with migraine did not differ. Yet, confounders can affect the estimates and therefore provide biased estimates and should, therefore, be adjusted in the analysis. Third is the generalizability of the included studies, many studies had recruited subjects from a single headache center and it is likely to be descriptive only of the most severe migraine. Therefore, a population-based longitudinal study to measure the impact of depression and anxiety on the humanistic and economic outcomes of subjects with migraine is needed.

| CONCLUSIONS
In summary, this review documents an association between depression and anxiety and the humanistic and economic outcomes on subjects with migraine. Depression and anxiety are associated with poorer humanistic and economic outcomes seen as a decreased HRQoL, increased disability and healthcare utilization, and cost.
These findings emphasize the importance of screening and treating mental health disorders for adults with migraine to improve patients' quality of life and reduce the economic burden on healthcare systems.