Google Trends is an open access portal that allows researchers to explore how the public’s quest for information on specific topics varies with time. The data made available by Google Trends is the “volume” (number) of searches for a specific search term entered by the public into the Google search engine per unit time (eg, per week), provided as a percentage of the highest search volume for that term over the period of interest (eg, last 5 years). The data are anonymous and collated geographically, and, given the public use of Google to search for health information [1,2], has been used to establish unexpected seasonality in the symptom burden or incidence of a variety of chronic conditions [3-5]. To describe such population-level investigations of disease processes using web-based data sources, Eysenbach [6] has coined the term “infodemiology.”

“Seasonality” in symptom burden refers to an annual periodicity, or modulation, in some measurable aspect of those symptoms. Much of this modulation may result from seasonal variation in environmental factors that convey the risk of disease. Respiratory viral illnesses are one of the best examples of this [7]. Humidity, temperature, and wind speed are all seasonally modulated, and each factor influences the spread of air-borne pathogens [8]. Mammals additionally have some seasonal modulation of their physiology (eg, body weight, fur thickness, and estrus). While this is not commonly thought of for humans, some studies suggest that even our physiology has some seasonality. Examples of this include higher long-bone growth in children during summer, retention of extracellular water starting in spring, continuing into the summer for patients on dialysis and increased immune system reactivity during winter [9-11]. Outside of infectious diseases, seasonality has also been observed in depression, cardiovascular disease, and overall mortality [12-14]. Recognizing and trying to understand the driving forces behind disease seasonality helps deliver insights that might lead to more effective prevention or treatment of seasonally modulated conditions.

Google Trends has become a popular tool for investigation of disease seasonality. An early use in this area was rapid real-time surveillance of influenza-like illness [15], something that continues to be worked on to augment conventional public health surveillance measures [16]. Others have sought to uncover unexpected seasonality in common conditions such as nocturnal leg cramps, ankle swelling, dental carries, and various mental health disorders [4,5,17-19]. However, a variety of things can confound the use of big data sources such as Google Trends search volume for health information as a proxy for symptom burden [20]. Search terms, for instance, might have dual meanings. Shingles is a disease, but they are also roofing tiles, whose use and related searches might be seasonal in Northern (snow experiencing) climates. Medical conditions can also be more or less newsworthy (eg, when celebrities are involved), and news coverage can sometimes drive search volume more than personal experience with the condition [21]. Influenza surveillance, for instance, has been inconsistent in its predictive ability when compared to hospital-based viral detection [22].

In our use of Google Trends to explore disease seasonality, we have come across an important potential confounder, which has yet to be described. This confounder is the searches for health information carried out by students who are taking courses at the undergraduate level. Such searches can be expected to be low in volume during the summer and winter break (in most countries) and high in volume during the final examination season. We have repeatedly observed such a biphasic seasonal pattern, which we will refer to as “academic cycling,” in many academic-oriented search terms (ie, fairly technical terms that are less commonly used in lay conversation such as proper diagnoses). Such academic cycling spans all fields of study. Some examples from health care, mathematics, and physics are shown in Figure 1. This same academic cycling pattern is clearly present in some of the infodemiology literature, but, even when it appears to be the main driver of the variation in search volume, it is either not acknowledged as such or not accounted for when its presence is recognized [18,23,24]. In this study, we (1) used the fast Fourier transform (FFT) on Google Trends search volume data with strong academic cycling, (2) identified the frequency domain pattern of that academic cycling, (3) searched for and removed that pattern from the frequency domain of search terms where seasonality is of clinical interest, and (4) recreated the time series data for the terms of clinical interest, with the academic cycling component removed. In so doing, we seek to empower researchers with strategies to investigate whether the seasonal trend they see in their Google Trends data is true, disease-related seasonality, or merely a confounding search pattern introduced by academic, school year–driven search volume.

One of the pillars of signal processing is the recognition that time-series data can be represented as the sum of many different sinusoidal waves, each with its own amplitude and phase difference. FFT is a software tool that does just that, representing a given time series (such as our 5-year Google Trends search volume) in the “frequency domain,” by showing what sinusoidal waves would need to be added together to produce the same curve [25]. FFT also lets users go backwards from the frequency domain representation and recreate the time series again (the “inverse FFT”). The advantage of this frequency domain representation is that we can think of our data as having a variety of driving forces and, if the frequencies of those driving forces are unique and can be identified, we can potentially remove them in the frequency domain and put the time series back again without the contribution of the unwanted component. A simple example of this, if one is listening to the radio, would be removing high- or low-frequency “noise” from the radio waves to hear voices more clearly. A more complex use of the same technique might be adding or removing an antipiracy frequency domain watermark from a piece of music or an image [26]. FFT has previously been applied to Google Trends data in order to identify the dominant frequency in time series data describing urinary tract infections and chronic lifestyle diseases [27,28].

We first demonstrated our filtering process in detail using the term [“thermodynamics”], which was chosen because of its strong academic cycling and helped each step to be visualized. The initial step involved preprocessing of the Google Trends data before FFT could be applied and involved shifting the time-series data down by subtracting the mean value. The resulting transformed data had the same shape as the original time series, but the data were now represented by positive and negative numbers that had a mean value of 0. Although not strictly necessary, we also chose to filter out high-frequency “noise” with FFT to make patterns more visible to the naked eye. These 2 preprocessing steps were applied to both the term of interest and to the control terms that represent the academic cycling that we wish to remove. We then identified how much of the academic cycling component was present in the term of interest by using a least squares regression analysis, subtracted that component in the frequency domain, and recreated the time series with inverse FFT. Following this demonstration, we applied the same technique to a selection of clinically relevant examples.

Google Trends time series data are freely downloadable and presented as the relative search volume (RSV) for the specified search terms per unit time (month, week, day, or hour). An RSV of 0 indicates little to no search volume, and an RSV of 100 indicates the highest volume for that term in the period of interest. We used weekly data for the 5-year period from July 3, 2016, to June 30, 2021. We restricted our analysis to the United States since it was the country with the largest internet search volume and since a single geographic region was needed for most residents to have a shared experience of the changing of the seasons and school year. Our 5-year window was selected to capture 5 full academic years. Although Google Trends provides the option of having search terms represent “topics” (in which case Google Trends aggregates a variety of searches they feel capture the same topic area), this option is not available for all search terms. Hence, for consistency, unless otherwise indicated, we did not use the “topic” search feature. Our search term nomenclature is in accordance with previous literature [29].

Post filtering, for bacterial genus terms, we selected the 6 terms (top 10%) with the strongest annual cycling component (ie, genus names with the highest amplitude frequency domain peaks at 52 weeks) and displayed them graphically. To do this, since these terms generally had a low search volume, and hence a relatively high amount of noise (ie, more seemingly random fluctuations), we graphed the average monthly volume to help average out random fluctuations and make any annual patterns more visible. In order to demonstrate how much the academic searches were driving the search volume for bacterial genus terms, we also calculated the squared Spearman rank correlation coefficient between the time series for each bacterial term and the time series for our control term (ie, [“gram stain” + “gram positive” + “gram negative”]). The squared Spearman rank correlation coefficient was used to estimate the amount of variation in the test data set, which was explained by the variation in the control.

Our filtering technique successfully removed academic cycling from a wide variety of Google Trends data where it is evident. Although the terms [“depression”], [“hypertension”], and [“myocardial infarction”] all had annual cycling prefiltering, this was only evident in searches for [“depression”] once academic cycling was removed. Of 56 pathogenic bacterial genus names, largely because of low search volumes, only 5 displayed substantial annual cycling prefiltering ([“Clostridium”], [“Escherichia”], [“Mycobacterium”], [“Staphylococcus”], and [“Streptococcus”]), and none of these 5 genus names displayed seasonality after academic cycling was removed. After filtering all genus terms, 10% of them with the strongest seasonality (ie, strongest 1-year periodicity in the frequency domain) were [“Aeromonas” + “Plesiomonas”], [“Moraxella”], [“Haemophilus”], [“Ehrlichia”], [“Legionella”], and [“Vibrio”], each of which had search volume peaks consistent with what the clinical literature would predict.

Owing to the relatively low search volume, few of our 56 bacterial genus terms displayed obvious academic cycling, with only 5 having a squared Spearman rank correlation coefficient of ≥0.5 with their corresponding control term. Academic cycling was clearly present, however, when the bacterial genus terms were averaged together and in the term [“Staphylococcus”] (Figure 3). Academic cycling explained three-quarters of the variation in the search volume for “Staphylococcus” (ie, R²=0.74), and half of the variation in our aggregate of 56 other bacterial terms (R²=0.55).

The top 10% of genus terms with the most annual cycling (ie, highest 52-week frequency domain peaks) after filtering out academic cycling are shown in Figure 4. [“Aeromonas” + “Plesiomonas”] searches increased during midsummer, [“Moraxella”] and [“Haemophilus”] searches increased during late winter, and “Ehrlichia” search volume spiked in late spring. [“Legionella”] searches had a slow, sustained peak throughout the summer months and during early fall, and [“Vibrio”] searches had a sharp spike during midsummer. All of these had no visible academic cycling and were essentially unaffected by our filter as demonstrated in Multimedia Appendix 1. Only 5 bacterial genus terms had obvious academic cycling, as demonstrated by a squared Spearman rank correlation coefficient of ≥0.5 for comparison with our control terms. These terms were [“Clostridium”], [“Escherichia”], [“Mycobacterium”], [“Staphylococcus”], and [“Streptococcus”], none of which displayed seasonality after filtering (Figure 5).

Academic cycling is evident in searches for information on all 3 of these common conditions (Figure 6). However, after filtering, only [“depression”] displays what appears to be a strong seasonal pattern in the time domain (corresponding to a dominant 52-week peak in the frequency domain), with searches peaking during winter. The terms [“hypertension”] and [“myocardial infarction”] have small peaks at 52 weeks. This could represent a lesser degree of seasonality or perhaps some residual academic cycling that we failed to remove.