Skip to content
MyFreud

Cyberchondria: Why Searching Symptoms Backfires

Searching a symptom feels like gathering information. A meta-analysis of 20 studies found the stronger the health anxiety, the more the searching escalates it.

5 min read

Pop-art illustration of a woman resting her chin on one hand and looking down at a phone she is holding, against a split teal and orange background.

Key takeaways

  • Cyberchondria is the escalation of health worry through online searching, and it is measurably distinct from simply looking things up.
  • A systematic review of 20 studies with 7,373 participants found a medium correlation between health anxiety and online health information seeking, and a large one between health anxiety and cyberchondria.
  • The mechanism is a search engine ranking by relevance, not by likelihood. The rare and serious explanation is the one written about most, so it is the one you find.
  • Searching is a reassurance-seeking compulsion. It relieves the worry for minutes and lowers the threshold for the next search, which is why the sessions get longer.
  • The useful intervention is not willpower. It is deciding in advance what would genuinely warrant a doctor, and then treating that as the only question worth answering.

Cyberchondria is the escalation of health worry through online searching. It is not the same as looking something up, and the difference is that the searching makes the worry bigger rather than smaller.

If you have ever started with a twinge and finished, forty minutes later, reading about something rare and terrible, you already know the shape of it.

What the research actually found

A systematic review published in 2019 pooled 20 studies covering 7,373 participants and ran two meta-analyses. It found a medium positive correlation between health anxiety and online health information seeking, and a large positive correlation between health anxiety and cyberchondria itself. [mcmullan-2019-cyberchondria]

Two honest caveats. These are correlations, so they cannot prove the searching causes the anxiety rather than the other way round; the likeliest reading is that each feeds the other. And cyberchondria is a research description rather than a diagnosis. What gets diagnosed is health anxiety, which the manuals split into illness anxiety disorder and somatic symptom disorder depending on whether physical symptoms are prominent. [scarella-2019-iad-cyber]

Our guide to health anxiety covers the wider condition this behaviour sits inside.

The search engine is not doing what you think

The most useful thing to understand is mechanical rather than psychological.

Search results are ordered by relevance and by what has been written, not by how likely a diagnosis is. A symptom that turns out to be nothing in the overwhelming majority of cases generates very little writing, because there is nothing to say. The rare and serious explanation generates a great deal, because it is important, dramatic, and worth documenting carefully.

So the ranking you see is close to the inverse of the ranking you want. You are not looking at a list of probable causes. You are looking at a list of well-documented ones, which is a different thing, and nothing on the page tells you the difference.

A forty-minute search, from the inside Illustrative
0 25 50 75 100 Level Minute 0 Minute 5 Minute 15 Minute 30 Minute 40 Anxiety Certainty Pages read
0 25 50 75 100 Level Day 1 Day 2 Day 3 Day 5 Day 7 Anxiety Certainty Searches that day

A schematic of the pattern described in this article and in the sources cited, not measured data.

The line that matters is certainty, and it goes the wrong way. More reading produces less confidence, not more, which is the opposite of what the activity promises and the reason it does not end on its own.

Why it is a compulsion rather than research

Research has a stopping condition. You wanted to know something, you found it out, you stop.

Searching a symptom has no stopping condition, because the question being asked is not really “what causes this” but “am I safe”, and no page can answer that. So the search ends when exhaustion or an interruption ends it, and the relief that follows a reassuring paragraph lasts about as long as any other reassurance: minutes.

That is the same loop that runs in reassurance-seeking generally, and it behaves the same way. Each round teaches the brain that the worry was worth taking seriously and that checking is what resolved it, so the threshold drops and the next session starts sooner.

What it feels like you are doingWhat is actually happening
Gathering informationPerforming a reassurance ritual
Narrowing it downWidening the list, since every page adds a cause
Being responsiblePostponing the one useful action, which is asking
Nearly finishedAt a point with no stopping condition
Checking one more thingRestarting the loop with a lower threshold

What helps, in the order it helps

Decide the threshold before you feel anything. Write down what would genuinely warrant a doctor for you: a lump, blood, a symptom lasting more than a set number of weeks, anything getting steadily worse. Then treat everything not on that list as not a question. The list is easier to write on a calm afternoon than during a search.

Put a delay in the way, and let the urge fall. Not a ban, a delay. The urge behaves like the contaminated feeling in obsessive-compulsive disorder: left alone, it drops on its own, and the point of waiting is to find that out.

Count the repeats. People are usually startled by how many times they have searched the same symptom. It reframes the activity from research, which you would only need to do once, into ritual.

Take the real question to a real person. Cognitive behavioural therapy has the best evidence for health anxiety, with a meta-analysis of treatment trials supporting it, and part of what it does is exactly this work on checking. [olatunji-2014-meta-cyber] Our guide to CBT for health anxiety covers what it involves.

If the worry has attached itself to physical sensations specifically, our guide to somatic symptom disorder covers that presentation, which is treated slightly differently.

When to seek help

Speak to a doctor if searching about your health is taking substantial time most days, if it is affecting your sleep or work, or if you have been to appointments and found the reassurance wore off within days.

Say that last part out loud when you go. “Reassurance stops working after a day or two” is a clinically meaningful sentence and it changes what gets offered, whereas describing the symptom again tends to produce another round of tests.

And do go about the symptom itself if it meets your own written threshold. The goal here is not to stop seeking medical care; it is to stop substituting a search engine for it. Go urgently if you have thoughts of harming yourself, or if the worry has reached a point where you cannot function.

How MyFreud can help

The pattern this runs on is invisible day to day and obvious across a month. MyFreud gives you daily mood tracking that takes seconds, which is how you find out whether the searching is settling or slowly taking more of your week.

Download MyFreud and start today: App Store or Google Play.

Frequently asked questions

What is cyberchondria?

Cyberchondria is the escalation of health-related worry that follows from searching symptoms online. It is more than looking something up: it describes a pattern in which searching increases distress rather than settling it, becomes hard to stop, and interferes with the rest of the day. Researchers treat it as related to health anxiety but measurably its own thing, with its own scales.

Does googling symptoms actually make anxiety worse?

For people already prone to health anxiety, the evidence points that way. A 2019 systematic review and meta-analysis of 20 studies found a medium positive correlation between health anxiety and online health information seeking, and a large positive correlation between health anxiety and cyberchondria. It is correlational rather than proof of direction, but it fits what the loop predicts: the more anxious you are, the more you search, and the more you search, the worse it gets.

Why do searches always turn up something serious?

Because search results are ranked by relevance and interest, not by how likely a diagnosis is. Common, dull explanations generate few pages; rare and frightening ones generate many. A symptom that is almost always benign will still return the serious possibility near the top, simply because that is what has been written about. You are not finding the most probable answer, you are finding the most published one.

How do I stop searching?

Not by trying to want it less. Decide in advance what would actually change your decision, which is usually a short list a doctor would recognise, and treat everything else as not a question. Put a delay in the way, because the urge falls on its own. And notice that the search is a reassurance ritual rather than research, which is easier to see if you count how many times you have searched the same symptom.

Is cyberchondria a diagnosis?

No. It is a research and clinical description rather than a category in the diagnostic manuals. What can be diagnosed is health anxiety, which the manuals call illness anxiety disorder or somatic symptom disorder depending on the presentation, and cyberchondria is usually one of its behaviours rather than a separate condition.

References

  1. 1.McMullan RD, Berle D, Arnáez S, Starcevic V ( 2019). The relationships between health anxiety, online health information seeking, and cyberchondria: systematic review and meta-analysis. Journal of Affective Disorders. doi.org . doi:10.1016/j.jad.2018.11.037
  2. 2.Scarella TM, Boland RJ, Barsky AJ ( 2019). Illness anxiety disorder: psychopathology, epidemiology, clinical characteristics, and treatment. Psychosomatic Medicine. doi.org . doi:10.1097/PSY.0000000000000691
  3. 3.Olatunji BO, Kauffman BY, Meltzer S, Davis ML, Smits JAJ, Powers MB ( 2014). Cognitive-behavioral therapy for hypochondriasis/health anxiety: a meta-analysis of treatment outcome and moderators. Behaviour Research and Therapy. doi.org . doi:10.1016/j.brat.2014.05.002