Tracking ME/CFS: What to Record and What Patterns to Look For

By Andre Minoli · Medically reviewed by Mary Carol Jennings, MD, MPH

Most symptom-tracking advice quietly assumes that causes and effects meet in the same day. You did the thing, you felt the thing, you wrote both down on one line. Nearly every tracking habit people build — and most of the tools they build it in — inherits that assumption.

Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) breaks it. The defining symptom of the illness is post-exertional malaise (PEM): a worsening that arrives, per Johns Hopkins, typically 12 to 48 hours after the exertion, and then lasts days or weeks. By the time you feel it, the day that caused it is two days back and reads, in memory, as a perfectly ordinary Tuesday.

That single fact reshapes what a useful ME/CFS log looks like. Here’s what to record, and what to go looking for once you have a few weeks of it.

The symptom the whole log is built around

Under the National Academy of Medicine (formerly the Institute of Medicine) criteria that Hopkins summarizes, a diagnosis rests on a substantial drop in what you used to tolerate, unrefreshing sleep, and PEM — plus either cognitive impairment or orthostatic intolerance, the early arrival of symptoms when sitting or standing for a while.

For tracking purposes, treat PEM as the outcome variable. Severity is worth logging for its own sake, but the question your log exists to answer is narrower and more useful: what did I do 12 to 48 hours before each crash? Everything below is in service of being able to answer that later.

Log four kinds of exertion, not one

This is the most common gap, and the most fixable. Ask someone to record their activity and they’ll write down the walk, the errand, the stairs. Hopkins is explicit that PEM follows exertion “whether that be physical, mental, emotional or upright activity, no matter how minor” — and three of those four almost never make it into a log.

• Physical. The one everybody records. Keep it, but record effort, not achievement — twenty minutes of cooking may cost more than a slow half-mile.
• Mental. A tax form, a long email thread, a difficult phone call, a day of concentrated work. Cognitive exertion is real exertion and it bills at the same lag.
• Emotional. A hard conversation, a hospital appointment, a confrontation, grief. This is the one people leave out because it doesn’t feel like doing anything.
• Upright. Time spent sitting or standing, which for many people with orthostatic intolerance is a genuine cost with no visible activity attached. Hopkins names the ordinary things that get expensive: standing in line, shopping, showering, hot environments.

Upright time is the one worth adding first if you add only one. It explains the crashes that otherwise look causeless — the ones where you “didn’t do anything,” except stand at a counter for forty minutes.

Read the log backwards, not down the page

Same-day reading will tell you almost nothing here. The technique is a lookback: take each crash, and read the two days before it, not the day of.

Do that across five or six crashes and one of two things happens. Either a specific kind of exertion keeps showing up in the window — a pattern worth naming — or nothing does, and you’re looking at a rolling load problem instead: no single day did it, but three moderate days stacked. Rolling load is common and invisible unless you look at multi-day stretches rather than individual entries, which is exactly the failure mode of a log you only ever read one line at a time.

This is the same discipline as testing any trigger hypothesis — count the times the suspect appeared and nothing happened, not just the times it did — but stretched across a longer clock, and with a much higher price for guessing wrong.

Learn your own warning signals

Here’s a finding worth knowing about. In a retrospective study of 197 ME/CFS patients, a minority — about one in seven — reported new or atypical symptoms appearing before their usual symptoms worsened: most often mood changes, nausea, unusual skin sensations, or headache, typically more than three hours ahead of the crash proper.

Two honest caveats. It was a minority of patients, so not having a warning signal is entirely normal. And while that group also reported less intense PEM, a retrospective study can’t tell us whether noticing the signal caused the milder crash.

What makes it worth tracking anyway is the study’s other observation: the warning symptoms were nearly constant for each individual patient. Whatever yours is, if you have one, it’s likely to be the same thing every time — which means it’s learnable, but only if you’re writing down odd symptoms that don’t seem to belong to anything. Leave room in your log for “weird thing today,” and read it back against your crashes after a couple of months.

Don’t go looking for your ceiling

Everywhere else on this blog, the advice for finding your activity limit is to compare the days before your crashes with the days before your quiet stretches, form a hypothesis about where the line sits, and test it.

In ME/CFS, skip the testing step. Probing a limit whose penalty is a multi-day or multi-week crash is an experiment with a genuinely awful cost-benefit ratio, and in a condition where crashes can ratchet your baseline downward, a failed test isn’t always something you get back. The NHS frames the goal as making the best use of the energy you have without making your symptoms worse — note the direction of that sentence. You are not searching for the edge. You’re establishing a level you can live under, with the log as evidence of where that is, and working the plan out with a clinician who knows the illness.

Practically, that means reading your log for the quiet stretches as carefully as the crashes. What did a sustainable week actually contain? That number — not the number that broke you — is the one worth knowing.

The patterns to look for

Once you have six to eight weeks, four things are worth checking for specifically:

• A consistent lag. Do your crashes sit a reliable distance behind their triggers — always next-day, always two-day? Personal lags tend to be stable, and knowing yours turns the lookback from a search into a check.
• Stacking. Crashes that follow several moderate days rather than one big one. If your worst days follow your busiest weeks, you’re budgeting the wrong unit.
• The invisible-exertion crash. Crashes with no physical trigger in the window. Look for upright time, cognitive load, or an emotionally expensive day instead.
• Baseline drift. Compare this month’s typical day with the same month last quarter. A slow downward step after a bad crash is important information, and it’s the pattern memory is worst at seeing — recency quietly rewrites what “normal” used to feel like.

That last one is also the strongest thing you can hand a clinician. “I’ve been worse since March” is easy to discount; three months of entries showing where the step happened is much harder to wave away.

Keep the logging cheap

One closing caution, and it matters more here than in any other condition we’ve written about: logging is itself a small cognitive exertion. A tracking habit elaborate enough to cost you energy is a bad trade, and the days you most want data from are precisely the days you’ll have least capacity to record it.

So build for the bad days. One tap for severity, a word or two for what you did, and nothing else mandatory. Gaps are fine — an unrated day is unknown, not a good day, and any honest tool should treat it that way rather than quietly drawing a line through the hole.

That’s roughly the shape Flare - Chronic Pain Tracker is built around: entries that take two seconds on a bad day, activities logged alongside severity so the two can be read against each other later, and a timeline that makes a two-day lookback a matter of scrolling rather than remembering. It won’t tell you why you crashed. It will make sure the Tuesday that caused it is still there to read on Thursday.

If you’re newer to this, the general tracking guide covers the habit itself — this post is the ME/CFS-shaped version of it. Everything here is about recording and reading your own data; what to do with a pattern once you’ve found one is a conversation for you and your care team.

Flare - Chronic Pain Tracker is a symptom-tracking tool, not a medical device. This article is general information, not medical advice — talk to your doctor about decisions affecting your health.