THE EFFECT OF COMPUTER USE ON THE OCCURRENCE OF MIGRAINE

Background/Aim. Risk factors concerning computer use related migraine onsets are still unknown. Study aims: determine prevalence of headaches in computer users; and effects of computer use and behavior on prediction of migraine presence. Methods. A cross-sectional study included 1500 subjects from the general population who were asked a questionnaire to assess the presence and type of headache and questions regarding computer-assisted behavior. Results. 67.9% of the subjects had a headache, of which 23.9% had a migraine. Results of multinominal regression analysis showed that significant predictors of migraine group, compared to group without headache, were sex (female) and family anamnesis, positive for migraine as well as spending more time on computer, making shorter and rare pauses in which physical activities are less included. Furthermore, members of migraine group, compared with group with other types of headache, were younger and had family anamnesis, rarely make pauses during computer use and their pauses were shorter. Conclusion. Improper and overlong computer use could be considered as a risk factor for migraine occurrence particularly in young people who have positive family anamnesis for migraine.


Introduction
The number of people suffering from frequent headaches is increasing every day 1 .
Headache is the most common neurological symptom today, which can have a significant impact on reducing quality of life 2,3 . This also affects work life, as headaches can reduce productivity at work 3 , and pupils and students may experience learning disabilities 4 .
Nowadays, working on a computer is becoming more and more common within the professional as well as in the private sphere of life 5,6 . In addition to the many benefits that use of a computer can bring, it can also lead to numerous damage to the health of users, of which, in the past research, besides spinal pain, shoulder pain, visual impairment 7 , fatigue, depression, obesity, headache has become the most mentioned 6,[8][9][10] . Some studies show that the use of computers can be considered as a significant risk factor for migraines 6,8,11 .
Although the use of computers has recently been increasingly associated to the onset of headache attack, not all the risk factors for working on a computer that contribute to its occurrence are known yet 8 .
Consequently, all strategies for preventing computer harm are not yet well known 8,12 , and further research is needed 8 . Objectives of the study: To determine the prevalence of headaches in computer users, to determine the use and behavior of computer users in order to affirm the risk factors for migraine.

Methods
This cross-sectional study was approved by the Ethics Committee of the Faculty of Medicine in Novi Sad (No 01-39/81/1). The study was conducted over a one-year period in primary, secondary schools, pre-schools, colleges and enterprises in the municipalities of Sombor, Apatin, Novi Sad and Mali Idjos, with prior approval of the management of these institutions.
The minimum number of subjects was calculated on the basis of G-power software, with an a priori set test power of 0.95. The baseline for initial values were the headache prevalence results from a previous study 13 , as this is the only study in these areas that presents headache prevalence results. Consistent with the prevalence of headaches in the aforementioned study, it was calculated that the minimum sample size should be 1040 subjects to control the type II error. A total of 1506 subjects that use a computer were included in the survey. The respondents in the study were adult pupils, students, teaching and non-teaching staff in schools and colleges, workers in firms. The exclusion criteria were as follows: persons with poor mental development and dementia, severe and lifethreatening acute and chronic illnesses, which excluded 6 subjects. Data collection was performed using a set of questions that was modeled after other studies 2,6,8,[13][14][15] .
Immediately before the questionnaire was distributed, the study participants were

Data analysis
Multinominal regression analysis was used in order to test the prediction of belonging to the migraine group, compared to the controls and other headaches (non-migraine) group.
Category predictors were sex, employment, family anamnesis and if participants make pause during the computer use, and continuous predictors were age, pause frequency (on a scale from 1 = after 30 min to 5 = never), pause duration (on a scale from 1 = never to 5 = 2 hours and more), correct position during computer use (on a scale from 1 = never to 4 = always), and total hours of computer use per day. Due missing data on some questions, the total number of answers was not the same across variables Using the chi-square test, the difference in the prevalence of migraine and other types of headaches between school children and adults was determined. Analysis was performed in SPSS v.23 for Windows.

Results
The total sample was first divided into two groups. The first group consisted of respondents who had a headache (1019 or 67.9%). The second group consisted of subjects who did not have a headache (control group), which consisted of 481 or 32.1% of respondents. The headache group was divided into two subgroups. The first subgroup consisted of the subjects with migraine (243, or 23.9%), and the second group the subjects with other types of primary and secondary headaches (776, or 76.1%). Table 1 shows the demographic characteristics, usage patterns, and behavior while working on a computer.
Using the chi-square test, it was determined that there are statistically significant differences in the prevalence of migraine and other types of headaches between school children and adults (χ2 (4) = 10.55, p = 0.032). The prevalence of migraine and other types of primary and secondary headaches is significantly higher in school children than in adults. Results of multinominal regression analysis showed that model is significant (χ 2 (20) = 1044.93, p < 0.001), with R 2 ranged from 0.59 (Cox & Snell) to 0.68 (Nagelkerke) and 77.3% of overall correct classification. Significant prediction of membership to migraine group compared to controls showed sex (with more females in the migraine group) and family anamnesis (with more participant with family anamnesis in the migraine group) from the demographics characteristics (Table 2). Regarding variables about computer use, results showed that migraine group compared to controls spend more total hours per day, rarely make pauses during computer use (after 3 hours or never), and make shorter pauses (up to 10 min), and they are more relaxing in pauses than engaging in physical activity.
Compared to participants with other headshakes, participants from the migraine group are younger and had family anamnesis more often, while there is no significant sex differences nor differences in employment (Table 2). Furthermore, although there are no differences in spent total hours per day, participants with migraine rarely make pauses during computer use and make shorter pauses, again. However, they report that they seat more correct during computer use, compared to the participants with other headaches. There are no differences in type of activity during pause between migraine and other headshakes group.
Most of the subjects with both migraine and other types of headaches reported having a headache after 2 to 6 hours of computer work (Table 3).

Discussion
Due to the increasing number of people suffering from headaches, many studies conducted so far have focused on discovering significant triggers for headache attacks [17][18][19] .
One of the triggers analyzed, which has recently become increasingly significant, is the use of computers [4][5][6]8,11,14,15,17,[18][19][20] . As in many other studies 6,9,11,21 , our study also has a high prevalence of headaches among computer users. In our study, as many as 67.9% of respondents who are computer users, experienced a headache in the previous year. A similar prevalence of headache among computer users (64.5%) was observed in a study conducted in Sweden 11 . Also, in a survey conducted in Iceland, 65.2% of computer users had a headache 11 . A slightly higher prevalence of headaches (74.9%) among computer users was observed in a study conducted in Finland 11 and in a study conducted in Brazil, where 80.6% had a headache 6 . A significantly lower prevalence of headache (26%) in computer users has been observed in a study conducted in Australia 8 . In our study, 23.9% of the subjects were affected by migraine. Slightly lower prevalence of migraine, 19.3% is observed in a study on the prevalence of headache in adolescents and its association with the use of computers and video games 6 . The prevalence of migraine, 30.2% is observed in the study by Saueressig et al 9 . These differences in the prevalence of headaches among computer users between the different surveys may be due primarily to different demographic characteristics of the respondents (due to differences in gender and age structure), different methodology, since it is observed that the inclusion criteria for determining the presence of headaches differ from study to study (from three months to one year). Given that, headache prevalence is expected to be higher in studies where the inclusion criterion for headache was the presence of headache for at least the previous year.
This study, in addition to determining the prevalence of migraine among computer users, was also conducted to identify behaviors during computer use to determine risk factors contributing to the onset of migraine attacks.
In addition to the already known fact that migraines are more common in women and in those with a positive family history of headache, the results of our study also indicate the importance of length of work and certain computer behaviors as a risk factor for the presence of migraines. Specifically, the subjects with migraine in comparison with the control group without headache spend significantly more time during the day working on the computer, rarely take a break, and when they do, they are of shorter duration and more often physically inactive during the break. In comparison with the group of patients with other types of primary and secondary headaches, the subjects with migraine belong to the younger age category and have a positive family history. Although there are no significant differences in daily computer exposure during the day, subjects with migraine compared with subjects with other types of primary and secondary headaches are significantly less likely to take a break, and when they do, they have a significantly shorter duration. However, it is noted that migraine sufferers are more likely to occupy a proper position while working on a computer than those suffering from other types of headaches.
The length of computer exposure during the day is the most studied factor to date, which has proven to be significant in the onset of migraine attacks 6,9,14,15 . Specifically, Saueressing et al point out that the chance of a migraine is even 2.54 times higher for computer users who use the computer for more than 3 hours during the day 9 . In their research, Xavier MK et al points out that computer users who used a computer for more than 4 hours a day were more likely to experience primary headaches, especially migraines 6 . Milde-Busch et al warns that even shorter exposure times to computers / the Internet (as little as 30 minutes) may result in an increased risk of migraine attacks in his research warns 15 . Also, a confirmation of the importance of the length of work on a computer during the day for the onset of migraine is also found in the results of the research conducted by Montagni et al. They cite two potential "scenarios" that could explain the effects of computer screens on migraine. The first is the brightness and frequency of the screen that can directly trigger the attack and the second is the screen exposure time, which can reduce the threshold for headache, which is then induced by other factors 14 . In contrast to our and the results of the aforementioned studies, different results, namely that the length of computer exposure does not play a significant role in the onset of headache attacks, were obtained by Smith et al in the study of the prevalence of neck pain and headache in computer users. However, the results of their study show the importance of length of work on the computer for the onset of pain in the cervical spine 8 . It is well known that ergonomic recommendations aimed at preventing the harmful effects of a computer on the health of users require the proper positioning of the body while operating the computer 6,18,22,23 . Prolonged irregular position of the body when working on a computer in an environment that is not designed according to ergonomic rules is thought to be stressful for trapezius muscle, which in addition to pain in the neck and shoulder can lead to headaches 24,25 . In our study, taking the right position when working on a computer proved to be a significant predictor between migraines and other types of headaches. There may not be a single factor contributing to the presence of migraines, but rather that there are several factors during computer work (prolonged work, irregular position and visual system overload) that jointly lead to headache attacks, as pointed out by Katherine et al 6 .
In order to determine how much and if computer users that were suffering from headache were aware of the computer as a trigger in our study, the respondents were asked if the occurrence of a headache attack could be affected by computer work. 85.2% of migraine sufferers and 72.2% sufferers of other primary and secondary headaches reported that the computer could be a trigger. More than half of migraine sufferers and of other primary and secondary headaches reported that the headache usually occurs after 2-6 hours of computer work.
As in other studies 4,8 , the results of our study show that computer users with headache generally do not adhere to existing ergonomic recommendations to prevent the harmful effects of computers on health. Given that the number of headache sufferers is increasing every day, there is a need to develop and implement measures to prevent the onset of headache attacks, especially migraines. It is recommended that computer users receive adequate ergonomic training to prevent headache 26 .

Conclusion
The results of our study indicate a high prevalence of headaches in computer users.
In addition to being female and having a positive family history of headache, length of computer work and individual computer behaviors were significant predictors of belonging to the migraine group compared to the headache-free group. Respondents with migraine compared to the group without headache significantly spend more time during the day working on the computer, less often take a break, and when they do, they are of shorter duration and are more often physically inactive during the break. Respondents with migraine compared to subjects with other types of headaches are younger and have a positive family history of headaches. Although there are no significant differences in the length of work on a computer during the day, subjects with migraine compared to those with other types of headaches are much less likely to take a break, and when they do, they are of a significantly shorter duration. So, the length of time you work on your computer is probably a risk factor for the presence of different types of headache, but the main difference between the types of headache is in the dynamics of working on the computer, that is, taking breaks. Computer users with headaches generally do not adhere to existing ergonomic recommendations for the prevention of the harmful effects of computers on health, and there is a need to develop and implement preventative measures, which can be achieved by training users on ergonomic principles for proper and adequate use of computers.

Study contributions
In addition to confirming the results of previous research that the length of work on the computer is probably a risk factor for the presence of different types of headaches, this study found that the main difference between the types of headaches is the dynamics of work at the computer, that is, the key is to take a break.

Disadvantages of the study
This research has several limitations. One of them refers to the way data was collected which is a survey, and therefore not the best way to obtain reliable data. The    nnumber of subjects