“Dr Google” will see you now – time trends in online searches on headache

Headaches are a large group of neurological disorders, with a high personal and social burden (1). Evidence on temporal changes in headache prevalence remains conflicting, with some studies reporting an increase while others report relative stability (2).

Google Trends is a web-based facility that analyses unbiased samples of Google search queries. It calculates the amount of searches for a given term relative to the total number of searches entered on the Google search engine.

Search volume for medical subjects has been pointed out as a valuable epidemiological indicator of disease prevalence over time. There is a high correlation between Google search volume and laboratory-supported epidemiological data of seasonal disorders such as influenza (3).

To analyse the evolution of the search for migraine and other headaches, we extracted Google Trends world data since its inception (January 2004) to September 2016. We used migraine and headache as keywords and compared search trends with other high burden neurological disorders. We also included asthma, a non-neurological condition that has a similar prevalence to migraine in children and young adults.

The Google Trends algorithm excludes queries made by very few people or the same person over a short period of time. It normalises data against total search volume, scaling results from 0 to 100 (details on support.google.com/trends). The Matlab polyfit function was used to fit the data with a second-degree polynomial (Figure 1).

Figure 1.

Plots of the normalised search results for each disorder. Gray lines represent the fitting of a second degree polynomial in the data. R2 for headache = 0.9814, for stroke = 0.5904, for migraine = 0.8578, for MS = 0.8875, for asthma = 0.8321. Based on our fitted curves, the minimum value for headache searches was in September 2004, for migraine in February 2009, for stroke in June 2010, for asthma in January 2012, and for MS in September 2016.

We observed an increase in online headache searches since September 2004. The relative number of online headache searches increased by 158% between January 2004 and January 2016, and a 15% increase was also observed for migraine. During this time period, searches for stroke (6%), multiple sclerosis (MS) (40%) and asthma (28%) decreased.

In January 2016, the volume of headache searches was 266% higher than for MS, 141% higher than for stroke and 227% higher than for asthma. Migraine searches were 22% lower than for stroke, but 18% higher than for MS and 5% higher than for asthma. These results contrast with those in January 2004, where the volume of headache and migraine searches was between 8 and 38% less than other disorders.

Although this exploratory analysis is limited by biases attributable to different internet access rates through time, different cultures or search engines, we showed a temporal increase in online searches of headache and migraine not accompanying the trend of other common disabling disorders.

Different interpretations could be made. Google Trends results were found to be associated with some diseases’ prevalence (3), thus we can hypothesise that an increase in headache prevalence could be an explanation. Simultaneously, as the age of onset of most frequent primary headaches somehow overlaps MS and asthma, and the searches on these disorders diverge from migraine and headache, our findings probably cannot be explained as the result of Generation Z’s online relationship with medical topics.

The reason we are using more online services to access specifically headache-related information remains to be elucidated. The design of studies to answer this question is a scientific challenge, but the data forwarded should promote our efforts in order to improve the quality of headache information available and to identify potential online intervention strategies.