QMigraine affects 10-15% of the population.
QPatients are dealing with different trigger factors
(food, smells, ect.) and try to manage or avoid them.
However, most of the studies, analyzing triggers of
migraine, are based only on small groups of patients.
QIn a preliminary project we have already shown that
a study using a web platform for data collection pro-
duces reliable results with the advantage of reach-
ing considerably more patients participating .
QObjective: Obtain trigger factors and distribu-
tions of headaches during the week with special fo-
cus on weekends and correlate findings with gender
QMigraine patients register using a smartphone app
(see figure 1) or web platform and afterwards regu-
larly submit information about their migraine attacks.
A NOVEL APPROACH COLLECTING MIGRAINE ATTACK DATA USING SMARTPHONE
APPS AND WEB FORMS
WOGENSTEIN F1, DRESCHER J1, RILL S1, SCHEIDT J1, GAUL C2, KROPP P3
1INSTITUTE OF INFORMATION SYSTEMS, UNIVERSITY OF APPLIED SCIENCES HOF, ALFONS-GOPPEL-PLATZ 1, 95028 HOF, GERMANY
2MIGRAINE AND HEADACHE CLINIC KÖNIGSTEIN, ÖLMÜHLWEG 31, 61462 KÖNIGSTEIN, GERMANY
3INSTITUTE OF MEDICAL PSYCHOLOGY AND MEDICAL SOCIOLOGY, UNIVERSITY OF ROSTOCK, GEHLSHEIMER STR. 20, 18147 ROSTOCK, GERMANY
QTo examine the thesis of an increase of migraine at-
tacks over the weekends, attack data for full-time
and part-time workers as well as for pupils and stu-
dents within a time period of 64 full weeks (1st of
June 2015 until 21st of August 2016) were analyzed.
QSince the participants are not as close guided as in
clinical studies, data cleaning is necessary before
further analyses. Especially, according to  two at-
tacks have to fulfill the condition that they are sepa-
rated by a 24 hours period without headache.
QTo decide if the distribution of the migraine attacks
over the weekdays significantly differs from a uni-
form distribution, chi-square tests were performed.
Q3,200 migraine patients were recruited (Germany,
Austria and Switzerland). Since the participation is
not limited to specific areas, the places of residence
of the registered migraineurs are well distributed
within the three countries (see figure 2).
QThe hypothesis of weekend migraine can not gener-
ally be confirmed.
QOur study shows a reduced migraine frequency on
Sundays and Mondays for working people and from
Fridays to Sundays for female pupils and students.
This supports the results of , were a decreased
migraine frequency on Sundays was found. Only for
male full-time workers an increased migraine attack
rate on Saturdays is visible. Torelli et al. concluded
to similar results in .
QThe present project shows that reliable results are
achievable using data acquired using web platforms
or smartphone apps. The large quantity of data will
later on enable us to investigate sub-groups of pa-
tients with very specific triggers, including certain
weather conditions, geomagnetic activity or other
environmental impacts. As more data is collected
it will enable us to examine more granular groups,
e.g., in age or location.
QNevertheless, the design of this study still has some
shortcomings in comparison to traditional epidemi-
ological studies. First of all, the participants are not
as closely introduced and medically guided as in
clinical studies. Furthermore, the fact that it is pos-
sible to join and leave the project at any time may
create some bias.
Figure 2: Spatial distribution of all participants within Germany,
Austria and Switzerland.
 Scheidt J, Koppe C, Rill S, Reinel D, Wogenstein F, Drescher J. Influence of
temperature changes on migraine occurrence in Germany. International Jour-
nal of Biometeorology 2013; 57:649-54.
 Göbel H. Paper-Pencil Tests for Retrospective and Prospective Evaluation of
Primary Headaches on the Basis of the IHS Criteria. Headache 1994; 34:564–
 Tfelt-Hansen P, Pascual J, Ramadan N, Dahlöf C, D’Amico D, Diener HC et al.
Guidelines for controlled trials of drugs in migraine: Third edition. A guide for
investigators. Cephalalgia 2012; 32:6-38.
 Alstadhaug KB, Salvesen R, Bekkelund S. Weekend migraine. Cephalalgia
 Torelli P, Cologno D, Manzoni GC. Weekend Headache: A Possible Role of
Work and Life-style. Headache 1999; 39:398-408.
Figure 3: Distribution of migraine attacks by day of the week for
full-time and part-time workers as well as for pupils and students.
QThe migraine type is diagnosed by a validated ques-
tionnaire . For each migraine attack necessary
information is collected, e.g., pain intensity and lo-
cation, associated symptoms, medication and the
disturbance of activities or work. In order to cluster
participants later on, we also collect demographic
information as well as spatial data for each migraine
QTo respect the patients’ privacy, all data are saved
QIn order to motivate migraineurs to take part in the
project, participants have the ability to use the plat-
form as an online migraine diary. They can review
their attack data, get a better overview about their
attacks using the provided statistics and create
printable reports for their treating doctor.
Figure 1: Android App showing the migraine calendar (sample
QWithin 64 weeks nearly 30,000 migraine attacks
were reported, about 20,000 remained after the fil-
QThe attack frequencies of each day of the week are
shown in figure 3 (a) to (f). Uniform distributions are
indicated by dashed lines in each of the plots. Four
of the plots – (a), (b), (c) and (e) – show a curve shape
being significantly different from a uniform distribu-
QFemale and male full-time workers as well as fe-
male part-time workers have a reduced or average
presence of migraine attacks between Sundays
and Tuesdays and an excess of attacks between
Wednesdays and Saturdays. For female pupils and
students, the number of migraine attacks decreases
significantly over the week. The other two distribu-
tions – (d) and (f) – do not show a significant devia-
tion from a uniform distribution.