Mental health

US recession saw more Google stress searches

"Google searches for stress-related illnesses rose during the recession," the Mail Online reports. The news comes from research looking at how US Google search trends for health complaints changed during the period researchers dubbed the "Great Recession".

The recession, which affected most western countries, was sparked by a global banking crisis and lasted from 2008 to 2011, although many would argue we are still suffering from its effects today.

Researchers found that higher than expected numbers of Google searches for illnesses that may sometimes be associated with stress, such as headaches and stomach ulcers, occurred during this time.

This is an interesting approach and adds to the body of evidence suggesting that Google search data can be used to gain an insight into wider health trends. For example, researchers can use Google to map flu prevalence based on search activity.

However, the study has a number of limitations, including the fact that searching for a health condition does not necessarily mean that the person was experiencing the problem themselves. It is also possible that the effect of increased internet usage with time in general has not been fully accounted for.

We do not know whether Google searches in the UK would show the same pattern. However, some of the most searched for conditions between 2008 and 2011 on the NHS Choices website were conditions that may in part be influenced by stress, such as depression and back pain.

Where did the story come from?

The study was carried out by researchers at the Santa Fe Institute, the University of Southern California and other US institutions, and was funded by a grant, a National Science Foundation (NSF) Graduate Research Fellowship and the Omidyar Foundation.

It was published in the peer-reviewed American Journal of Preventive Medicine.

The Mail Online's reporting fails to note there are many limitations in the approach this study used that may affect the interpretation of the results. They also do not note that this study relates to Google searches in the US during the recession, not in the UK.

What kind of research was this?

This research examined Google search queries to see how specific health concerns changed during the US recession between 2008 and 2011.

The authors say that most previous research examining the health effects of the economic situation comes from surveys. Surveys can provide useful data, but they are cost intensive and time consuming to complete. They also often include non-specific measures, such as general self-ratings of health, rather than looking at specific concerns.

The researchers say that by monitoring hundreds of systematically selected Google query trends, their novel approach "takes the traditional self-rated health questionnaire to the next level by identifying precise health concerns by the query content and their prevalence by the query volume".

This is arguably an appropriate approach to take because of the sheer pervasiveness of Google. It is estimated that more than one billion Google searches occur in the course of just one day.

What did the research involve?

Search trends in the US were downloaded from Google Trends in the form of weekly relative search volume (RSV) time series.

These automatically generated search figures reflect the proportion of search queries inputted by the investigator (the person Googling) relative to all queries each week, normalised on a 100-point scale.

For example, an RSV of 50 means 50% of the search volume of the highest search proportion in that week. This is said to correct for overall increases in Google searches over time as a result of changes in internet access or people's disposable time.

The researchers focused on health concerns that previous research has suggested may have both a psychosomatic element and be caused by economic concerns. These included complaints such as chest pain, headache and other pain, and stomach problems.

The researchers used the term "Great Recession" to describe the period between December 2008 and 2011. They then compared the cumulative difference between the observed and expected search query volume based on linear projections from pre-existing trends.

Higher than expected searches during the recession years were called "excess" searches. The 100 queries with the greatest excess searches were ranked and clustered into themes based on search query content.

What were the basic results?

Overall trends showed that compared with the period 2006-08, there was a general 26% increase in searches for all health concerns during the Great Recession.

Among the top 100 health concerns, there were around 205 million excess health concern queries during the Great Recession. These were considered "excess" in the sense that they were higher in number than would be expected if previous pre-recession trends had continued.

The specific concerns with the greatest excess queries were stomach ulcer symptoms, with a 228% increase above expected, accounting for around 1,480,000 excess queries. Headache symptoms had a 193% increase above expected, accounting for around 1,520,000 excess queries. 

Queries tended to follow themes of looking for symptoms and diagnostic procedures or tests, such as heart monitor tests. Other top concerns were hernia (37% above expected), chest pain (35%) and heart rhythm problems (32%).

Other types of pain with excess queries were back, stomach and tooth pain (each with 19% excess), and joint pain (11% excess).

Cancer searches were also up by 32% during the period, with "congestion" (breathing problems) searches up by 26% and pregnancy searches up by 22%.

The confidence intervals on the estimates of the excess searches were noticeably large, indicating there is a large margin of error in the estimates given above.

For example, all health-related search queries were estimated to have an excess of 26% overall, but the true value was reported to be anywhere between 3% and 138%, as expressed using a 95% confidence interval.

How did the researchers interpret the results?

The researchers conclude that, "Google queries indicate that the Great Recession coincided with substantial increases in health concerns, hinting at how population health specifically changed during that time."


This research found that there were increased Google searches during the US "Great Recession" between December 2008 and 2011 for a range of health concerns, including headaches, stomach ulcers and other stomach problems, chest pain, heart rhythm problems and various other pains, including back pain and toothache. The researchers consider that this could potentially indicate worsening population health.

However, although this is an interesting approach, it is limited by several factors. A person may search Google for information on a health complaint for many different reasons. It does not necessarily mean that a person was experiencing that problem themselves.

Even if they were experiencing the general symptom searched for, it cannot tell us what their actual diagnosis was, how long they had been suffering from it, or any associated health problems they had. It certainly cannot tell us what the direct cause was.

It is worth pointing out that not everyone with a health problem chooses to search for information about it on Google as a first port of call. Many people may just visit a doctor and obtain related health information through other sources, such as printed literature or media sources other than Google.

Although the research has tried to account for increases in overall internet usage with time, including improved internet access or more disposable time, it is still difficult to tell whether this effect has been completely accounted for.

This research also only relates to searches in the US and did not investigate whether Google searches in the UK showed the same pattern.

However, the study highlights how Google search data could be used to provide a useful insight into wider health trends at a population level.

Bearing the limitations in mind, the suggestion that economic problems can trigger health problems is certainly plausible, although it cannot be proven by this research.

NHS Attribution