Determinants of migraine headache among regular undergraduate students, of Wollo University, Dessie, Ethiopia: cross–sectional study

Background The prevalence of migraine headaches varied from 2.4 to 48.5% worldwide among university students. As per the knowledge of the investigators, the study of migraine headaches is not done in this study area. Hence, this study aimed to assess determinates of migraine headaches among regular undergraduate students, of Wollo University, Dessie, Ethiopia, 2020. Method Institution-based cross-sectional study design was conducted from November 15–30 /2020 by using a self-administered questionnaire among 371 regular undergraduate students at Wollo University. A multistage sampling technique was employed to represent our study population. Identity migraine test, Migraine Disability Assessment Questionnaire, perceived stress scale, and Oslo social support scale was used. Descriptive statistics, binary and multiple logistic regression were employed. The adjusted odds ratio at a 95% confidence interval and level of significance (p-value < 0.05) was used to interpret the findings. Results From 371 participants, the prevalence of migraine headache in this study was 34% (95% CI: 29.2, 38.5). Variables statistically associated with migraine headache, were a family history of headache (AOR = 3.83, CI: 2.313, 6.366), suicidal thoughts in the past 3 months (AOR = 10.76, CI: 2.117, 54.74), and had low perceived stress 62.6% (AOR = 0.374, CI: 0.205, 0.683). Conclusion In this study, the prevalence of migraine headaches was very high. Family history of headache, suicidal thoughts, and low perceived stress were determinates for migraine headache. So special emphasis shall be given to those participants who had high perceived stress, a family history of migraine headache, and suicidal thoughts.

Several data on the epidemiology of migraines among young people and students are available in western countries [22][23][24][25]. Despite, in developing countries in Africa, especially in Ethiopia few data were available. The prevalence of migraine has been estimated to be 19.8% among Nigerian students [26], in Kenya 33.8% [27], and in Benin 11.3% [28]. Moreover, in studies conducted on students at the University of Gondar the prevalence of primary headache was 584 (81.11%), of them, 94(13.06%) were having a migraine [29].
Factors trigger migraine headaches such as stress, exposure to the sun, sleeplessness, eating habits, changes in weather conditions, temperature, frequent traveling, food items, oral contraceptives, and physical activities. Of them, the most common migraine triggers are lack of sleep and fatigue along with smoking [30][31][32]. Furthermore, having a family history of migraine, those students who enrolled in the second academic year, fatigue, anxiety, depression, stress, prolonged tension, changes in sleep or lack of sleep, hangovers, medications to treat other disorders, overuse of pain medications, panic attacks, strokes, dehydration, influenza, hypertension, strong odors, allergies, excessive intake of caffeine, caffeine withdrawal, smoked meats, skipping a meal, mechanical factors such as neck strain and cigarette smoking [18,33].
Migraine headaches hurt individuals' performance, and high costs are imposed on society due to its absenteeism in the workplace as well as education [34][35][36]. And also migraine headache has impacts on psychological problems such as depression, Obsessive-compulsive disorder, Bipolar and anxiety disorders [37]. Migraine headaches are under-diagnosed and treated conditions in undergraduate students. This leads to the compromised academic performance of students. Based on our knowledge, there are no studies done previously, about migraine headaches among Wollo university students. Therefore, this study aimed to assess the determinates of migraine headaches among regular undergraduate students in Wollo University, Dessie, Ethiopia.

Study design, period, and setting
An institutional-based cross-sectional study was conducted among regular undergraduate Wollo university students from November 15-30, 2020. Wollo University is one of the thirteen 3rd generation universities which were established in 1997 E.C. It is located in the Amhara region, 401 KM away from the capital city of Ethiopia. The university was starting to work with the enrollment of 749 students in one faculty of education and currently, it has more than 7000 students with six colleges.

Study participants, sample size determination, and sampling technique
There were a total of 1852 regular undergraduate students registered for the academic year of 2020, across different colleges. The sample size was determined using the single population proportion formula by taking the prevalence of migraine headache 13% [29], and 95% confidence interval and, 5% margin of error was used and finally sample size was 192. After adding a 5% none response rate the sample size was 202. Since we were used the multi-stage sampling technique we multiply our sample size by 2 and we got 404. A two-stage cluster sampling technique was employed to select the study participants. In the first stage by using a simple random sampling technique we choose three colleges out of the six colleges. In the second stage, again by applying a simple random sampling technique, we select 6 departments from each chosen college and on the final stage, we proportionally allocated the sample to select 404 participants (Fig. 1). We excluded students with visual impairment, those unable to fill the self-administered questionnaire due to illness, and extension students.

Measurements
The dependent variable is migraine headache (Yes or No). The independent variable includes socio-demographic characteristics (age, sex, marital status, family history migraine headache, monthly income, living place before joining to campus), academic factors (Academic year, academic CGPA), other potential factors (stress, sunlight, noise, fatigue, heat, menstruation in women, too much work late sleep at night, exercise, smells and some foods) and mental illness-mood and emotional changes, suicidal ideation, and attempt.

Having migraine headache
From the Identity migraine test tool used, three questions having one or more symptoms on headacherelated disability, nausea, and photophobia [38].

Migraine disability assessment questionnaire score (MIDAS)
Score 0-5 is grade 1 or little or no disability, score 6-10 is grade 2 or mild disability, score 11-20 is grade 3 or moderate disability, score 21 and above is the severe disability [39].
Individual scores on the PSS can range from 0 to 40 with higher scores indicating higher perceived stress. Scores ranging from 0 to 13 would be considered low stress, 14-26 would be considered moderate stress and 27-40 would be considered high perceived stress [40].

OSLO social support scale
A score ranging from 3 to 8 would be considered poor social support, 9-11 moderate social support, and 12-14 strong social support [41,42].

Data collection tools and instrument
The data was collected using a self-administered questionnaire which contains the socio-demographic variables and different tools were used (Identity migraine test) to measure migraine headache, and its disability (MIDAS), perceived stress (PSS), and level of social support (OSSS-3). Identity migraine testis the most widely used and validated of the migraine tools is the ID Migraine, with three questions on headache-related disability, nausea, and photophobia [38]. MIDAS is used to assign a disability grade indicating the overall intensity of illness during a 3 month recall period: little or no disability (grade 1, score 0-5), mild disability (grade 2 scores 6-10), moderate disability (grade 3, score [11][12][13][14][15][16][17][18][19][20], severe disability (grade 4 score 21 and above) [10,39].
The Perceived Stress Scale (PSS) is a classic stress assessment instrument. The tool, while originally developed in 1983, remains a popular choice for helping us understand how different situations affect our feelings and our perceived stress. The questions in this scale ask about your feelings and thoughts during the last month. In each case, you will be asked to indicate how often you felt or thought a certain way. Which has ten questions, for each question has five alternatives [40]. The OSSS-3 consists of three items assessing the level of social support. It has been recommended for epidemiological and population-based surveys. Consequently, it can be considered a suited instrument for the current thesis [41,42].

Data collection techniques and quality control
Data were collected by two BSc psychiatry nursing professionals using a self-administered questionnaire having eight parts. The first part is socio-demographic characteristics, the second part is containing questions regarding academic achievement, the third part contains questions used to assess the presence of migraine headache, the fourth part contains questions used to assess migraine disability, the fifth part contains questions related to stress, the sixth part contain factors may intensify migraine headache, the seventh part contains questions related to social support, the eighth part contains questions related to mental health and substance.
We took many measures to assure the quality of data. Before proceeding to data collection the training was given for data collectors about the tools used and how to approach the respondents and to clearly explain the objectives and the purpose of the study. The questionnaire was pre-tested 1 week before the actual data collection time on 5% of the sample size on undergraduate students at Woldya University and appropriate modification was made. Finally, the questionnaire was written in simple and understandable language, using double data entry, and checked for completeness before data entry and incomplete data was discarded.

Data processing and analysis
Data clean-up and cross-checking were done before analysis. Data were checked, coded, and entered in EPI data 3.1 version then it was exported to SPSS version 21 for analysis.
Both descriptive and analytical statistical procedures were utilized. Descriptive statistics like percentage mean and standard deviation were used for the presentation of demographic data and prevalence of migraine headache. Tables also were used for data presentation. Binary logistic regression was used to identify factors associated with migraine headaches. Variables with P-value of less than 0.25 were considered as a candidate for multivariate analysis. Multiple logistic regression models were fitted to control the possible effect of confounders and finally, the variables which had an independent association with migraine headache were identified based on AOR, with 95%CI and p-value less than 0.05.

Socio-demographic characteristics of the respondent
We approached 404 participants, of these 371 respondents participated with a response rate of 91.8%. The mean age of the respondents was 23 years of age with the standard division of ±2.836. More than half of the respondents were male190 (51.2%), orthodox followers 255 (68.7%), Amhara by ethnicity 258 (69.5%), and single 337(90.8%), while 193 (52%) of them were grown in rural areas. Less than half of the students120(32.3%) visits a place of worship daily. The average monthly income was 770 ETB and the majority of the students 271(78.3%) had no family history of migraine headache (Table 1).

Academic related characteristics of students
The majority of the students 233(62.8%) are in the third year, besides 74(19.9%) were in the fourth year and the rest 64(17.3%) are in the fifth year. Besides 78(21%), 64(17.3%), and 50(13.5%) are accounting, pharmacy, and plant science from the selected department respectively. Moreover more than half of 195(52.6%) students have a cumulative grade point of 3-3.5. ( Table 2).

Prevalence of migraine headache among students
In this study, the lifetime prevalence of migraine headache among regular undergraduate WU students was 34% (95CI: 29.2, 38.5) (Fig. 2).

Assessments of substance and mental health
In our study 15(4%) of the students had a mental illness, 9(2.4%) had a history of suicidal attempts, 12(3.2%) had a suicidal thought in the last 3 months, 59(15.9%) had substance use (alcohol is the leading 27(7.3%)) and 17(4.6%) had a medical condition (Table 4).

Variables having a p-value less than 0.25 in bi-variable analysis
Variables statistically associated with migraine headache in bivariate analysis (p-value of < 0.25) was female students, marital status of other than single, Muslim in religion, having a family history of headache, never going to place of worship, monthly income of greater than or equal to 2000, having low perceived stress, and have suicidal thoughts. However, age, ethnicity,  living residence before campus, study year, social support, and cumulative GPA of the students do not show any association with migraine (Table 5). Those variables with a P-value of < 0.25 in bivariate analysis were entered into multivariable logistic regression for further analysis to control the confounding factors. In multiple logistic regression analysis family history of migraine headache, history of suicidal thought in the past 3 months was significantly associated with migraine headache, while having low perceived stress positively affects the migraine headache.

Factors independently associated with migraine headache
On the final model analysis, those who had a family history of headaches are more than 3.83 times (AOR = 3.83, CI: 2.313, 6.366) had a chance of having migraine headaches than those who had no family history of headache. Similarly, those who had suicidal thoughts in the past 3 months are more than 10 less likely to have migraine headaches than those who had moderate and high perceived stress (Table 6).

Discussion
In our study, we tried to assess the prevalence, impact, and associated factors of migraine headache among regular undergraduate students in Wollo university. The prevalence of migraine headaches was high. Having a family history of headaches and had suicidal thoughts were the risk factors for migraine headaches, while had low perceived stress was the protective factor. In our study, the prevalence of migraine headaches among regular graduate Wollo University students was found to be 34% at 95% CI (29.2, 38.5). This finding is higher as compared with the study conducted in Gondar university 13.06% [29,43]. This may be due to different stressors like covid-19 epidemics during the data were collected. Likewise, in the study conducted inParakou Beninethe, lifetime prevalence was 14.3% (95% CI 12.3 to 16.4%) [22], in Zahedan among medical students was 7.14% [21,42], in Greece, was 2.4% [18], and in Turkey was 12.4% [10]. This difference might be due to time variation, cultural and geographical differences, and environmental factors.
In our study the prevalence was slightly higher than the study conducted in Nigeria among nursing students was 26.8% [37], in Brazil, found a prevalence of 25% of migraines [10], and in King Abdul-Aziz University was 26.3% [37]. This might be due to sampling size differences, cultural and geographical differences, and environmental factors. However, our finding is in line with the study conducted in Kenya 33.8% [13]. In our study the MIDAS showed that 237(63.9%) had no disability (grade I), 65(17.5%) had a moderate disability (grade III), 45(12.1%) had a mild disability (grade II) and the rest 24(6.5%) had a severe disability (grade IV). This is consistent with an epidemiological study conducted    [44].
Based on our study one of the factors that strongly associate with migraine is had a family history of headaches which increased the risk of migraine more than 3.83 times (AOR = 3.83, 95% CI: 2.313, 6.366) than those who had no family history of migraine headache. This is consistent with a study conducted at King Abdul-Aziz University [37]. But our finding is lower as compared with the study conducted in southeast China [10].
Moreover, in this study those who had suicidal thoughts in the past 3 months are more than 10.76 times (AOR = 10.76, 95% CI: 2.117, 54.74) had a chance of having migraine headache than those who had no suicidal thoughts in the past 3 months. Our finding was supported by the studies done in Taiwan [7]. However, our finding is somewhat high as compared with the study conducted in Rom [12]. This difference may be due to using a different type of method and geographical variation.
In this study having low perceived stress 61.4% (AOR = 0.386, 95% CI: 0.207, 0.72) were less likely to had migraines than those who had moderate and high perceived stress. This is consistent with the study conducted on Jazan medical students [20]. Besides in our study, we observed no relation to gender as in the study done in Iran-Zahedan, but there was a female predominance in most of the other studies in Riyadh, [36] in the USA, [12] in Nigeria, Brazil, and Turkey [9]. This difference may be due to using a different type of method, geographical variation, and cultural differences.

Conclusion and recommendations
This study was aimed to assess the determinates of migraine headaches among regular undergraduate Wollo University students. Therefore the prevalence of migraine headaches was found to be very high. The predictors of migraine headaches were a family history of headaches, suicidal thoughts in the past 3 months, and had low perceived stress. So we recommend for different stakeholders, attention shall be given to those study participants who had a family history of migraine headache, suicidal thoughts, and high perceived stress.

Strength and limitations
Conducting this study is important because as university students the curriculum requires constant concentration and hard work, students even a one-day absence of student can affect his/her school success, so we think that our research is important as it confirms the opportunity to evaluate migraine prevalence and its determinates, in a population where the disease could exert a relevant negative impact. The use of validated tools, doing multivariate logistic regression for analysis of data to control possible confounders could be our strength.
Besides this study had some limitations, which include using the cross-sectional method does not show cause and effect relationship. And also suicidal ideation and suicide attempts were evaluated using yes or no questionnaires rather than through psychiatric standardized evaluation tools. Furthermore, we did not classify migraine headaches, with aura and without aura. Therefore, we recommend further studies employing using psychiatric standardized evaluation tools to clarify suicide risk among patients with migraine headaches.