Three Weird Trends in Mental Health

“My eyes eat the words but my brain won’t swallow them," my friend Ray told me at lunch last month.

He went on to explain that whenever he tried to read, his eyes would scan a page but he had no comprehension of what he was reading. With intense concentration, he could eventually understand what he read, but the effort left him exhausted.

Aware that I’m a neuroscientist and former therapist, Ray wanted to know what was wrong with him.

I asked “How long has this been going on?”

“Since early November,” he replied, then added “I get stupid every November.”

“Stupid?”

“Yeah. I lose things, I can’t concentrate, I feel spaced out. I can’t even operate my dumb coffee machine half the time. And…I can hardly read.”

“That’s got to be scary,” I observed. “How are you feeling, emotionally?”

“Anxious all the time. I wake up early but don’t want to get out of bed. I’m afraid of people because they’ll see how dumb I’ve gotten. I used to love work, now I dread it.”

I thought for a moment then asked, “You say this starts every November. Does that mean it gets better after a while?

“I think by July, I don’t notice it much. Then, before Thanksgiving, it seems to come back.”

I had a hunch what was going on with Ray, but didn’t share my thoughts because I no longer practice therapy. Instead I told Ray I wasn’t sure what the trouble was, and gave him the name of a good Psychiatrist.

But in my heart of hearts, I suspected Ray suffered from Seasonal Affective Disorder (SAD), also called “winter blues” or seasonal mood disorder.

His cognitive difficulties are common with depression, as are his anxiety and lack of energy. The fact that his symptoms abated in the summer, suggested that he was not suffering from something organic, like early dementia, but something cyclical.

SAD affects about 1.4% of people in Florida but over 9% in Alaska. This makes sense given that Florida has many more hours of sunlight than Alaska, and researchers believe that variations in sun exposure play a large role in the disorder. For example, exposure to extra light for 60-90 minutes a day can alleviate symptoms.

And Psychiatrist Jon Draud, MD writing in Psych Congress Network, points out that cognitive impairments, of the kind that Ray described are common with SAD.

Digging deeper into SAD, I stumbled on a fascinating phenomena. The volume of Internet searches of “Seasonal Affective Disorder” follow an extremely regular, cyclical pattern shown below, peaking in winter and bottoming out in summer. Although people suffering from SAD may not be directly responsible for this search volume, the patterns are highly suggestive.

SAD may not be the only seasonal mental disorder

Dr. John Ayers of USC Medical School found that searches for most mental disorders, including ADHD, Anorexia, Bulimia, Anxiety and OCD follow a winter-peak-summer-trough seasonal pattern, probably from people seeking information about their symptoms and treatment options.

Zeroing in on depression, Dr. Albert Yang and coworkers in Taiwan discovered that internet searches for the disorder are highly seasonal.

Google searches for “depression” show pronounced cycles in Sweden, almost no yearly pattern in equatorial Malaysia and a cycle in Australia that is precisely opposite that of Sweden. The chart also shows that searches for depression in both hemispheres closely track searches for SAD. (The Australia pattern is shifted exactly 6 months because summer there is winter in the Northern Hemisphere)

Although it’s conceivable that the pronounced cycles in searches for “depression” are due solely to people curious about SAD, the very low volume of searches of SAD relative to depression (SAD searches are less than 2% of "depression" queries) and the large magnitude of periodic swings in “depression” search volume, hint that a significant amount of “vanilla” depression has a seasonal component.

The seasonal pattern of searches for almost all mental disorders also suggests that SAD sufferers are not responsible for the large periodic swings in search interest for depression.

The questions and implications of these findings are profound.

Are most mental disorders related somehow to exposure to sunlight? If so, what is the underlying biology?

Should light therapy be used for disorders other than just SAD?

Analysis of Internet search volume can uncover weird, unexpected results.

Internet search involves truly enormous numbers. Roughly 3.3 billion people are on the Web and these people make 1.2 trillion searches per year. Thus Internet search behavior may be able to uncover pervasive psychological phenomena that are not easily seen in academic studies.

A very large academic study of a mental disorder typically encompasses a few thousand subjects. But the yearly Google searches in the US alone for "depression" top 32 million!

This means that small, but highly significant mental health patterns can be found through search volume that would elude traditional analysis.

Consider the results for “Seasonal Affective Disorder” and “Depression.”

Both show a seasonal pattern, but there is a consistent December dip for depression that is absent for SAD. This dip probably is not caused by a Christmas holiday hiatus in Internet search, because data from Australia show the same dip in July for the Southern Hemisphere.

What’s this dip all about?

I haven’t a clue, nor does anyone else.

Internet search data yield highly counterintuitive results

One implication of seasonal variation in search volumes for mental disorders is that there is a direct relationship between the amount of people suffering from a disorder and the amount of searches for information on that disorder.

But a comprehensive comparison of the prevalence (percent of population stricken) of different mental disorders and search volume for those same disorders suggests that the situation is far more complicated than that.

Here, I’ve created a scatter plot of the monthly prevalence of common disorders vs. monthly Google search volume for those same disorders.

If anything, there is an inverse relationship between disorder prevalence and search interest in that condition.

For example, the most prevalent mental problem, anxiety (18% monthly prevalence or 57 million people in the US) is associated with 2.2 million Google searches per month in the US, yielding a 3.7% ratio of “searches-to-sufferers.” But schizophrenia, which afflicts 1.1% of the US population (3.52 million) has a search-to-sufferer ratio (with 2 million monthly searches) of 56%!!

Hmmm.

Are the mental disorder prevalence data themselves wrong? Do we have a lot more schizophrenia and narcissism out there than we suspect and a lot less anxiety? Is search volume correlated with the amount of people suffering from mental disorders some of the time but not others?

Who knows?

But the truth may be out there, waiting for more analysis of Internet searches to uncover it!!

http://www.psychcongress.com/blogs/jon-w-draud/june-2-2011-1234pm/seaso…

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2965678/

http://www.ncbi.nlm.nih.gov/pubmed/23597817

http://www.nimh.nih.gov/health/statistics/prevalence/index.shtml

http://www.ncbi.nlm.nih.gov/pubmed/1503129

Nolen-Hoeksema, Susan (2014). Abnormal Psychology (6th ed.). New York, New York: McGraw-Hill Education. p. 179. ISBN 978-1-259-06072-4.