The Poincaré Plot in HRV: What SD1 and SD2 Actually Show

A picture of your heartbeat, one pair at a time

Most heart rate variability numbers are single figures: RMSSD, SDNN, an HF power reading. The Poincaré plot is different. It is a picture. It takes the same raw beat-to-beat intervals and draws them as a scatter of dots, and the shape of that scatter tells you at a glance how your rhythm is behaving. The two numbers that summarize the shape, SD1 and SD2, are what people usually mean when they talk about “nonlinear” HRV.

Here is the whole trick. Line up your heartbeat intervals in order. Take the first two, and plot a single dot: the first interval sets the horizontal position, the second sets the vertical position. Then slide over by one beat and plot the next pair, the second interval against the third. Keep going. Every dot is one pair of consecutive beats, so a five-minute reading becomes a cloud of a few hundred dots. If your rhythm barely changes, each interval is almost the same as the next, so the dots pile up in a tight little ball on the diagonal. If your rhythm is lively, the dots spread out into a long, comet-shaped cloud. The eye reads recovery before you even look at a number.

If the idea of “one interval against the next” is new, the RR intervals and tachogram explainer walks through where these intervals come from in the first place.

What SD1 measures

SD1 is the width of the cloud, measured across the diagonal line. That perpendicular spread is created entirely by how much each beat differs from the one right before it, which is the fast, one-beat-to-the-next change your vagus nerve produces. So SD1 is a vagal-tone marker, in the same family as RMSSD and pNN50. When your parasympathetic “rest and digest” side is engaged, successive beats swing further apart, the cloud fattens, and SD1 goes up.

There is no hand-waving in that link. SD1 is not merely like RMSSD, it is RMSSD in disguise:

What SD2 measures

SD2 is the length of the cloud, measured along the diagonal. It grows with your overall variability, both the fast changes and the slower drifts in heart rate across the reading. Because it blends everything, SD2 relates closely to SDNN, the broad summary of total variability. A long comet means plenty of variation over the whole reading; a short one means the rhythm stayed hemmed in.

The relationship is again a clean formula, built from the two numbers you already know: SD2 is derived from SDNN and SD1 together, as SD2 = √(2·SDNN² − SD1²). In plain terms, once you know your total variability (SDNN) and your fast beat-to-beat width (SD1), the long axis of the cloud is fixed. So the Poincaré plot is not new information so much as a new view of RMSSD and SDNN, one that happens to be very good at showing shape and at flagging odd beats.

SD1, SD2 and the SD1/SD2 ratio at a glance

Here is how the Poincaré descriptors line up against the metrics you may already track. Treat the numbers as orientation for short at-home readings, not as diagnostic cut-offs.

Because SD1 mirrors RMSSD exactly, you can borrow the same recovery bands the app uses for RMSSD and divide by √2. That gives a rough SD1 orientation for a resting reading:

SD2 is best read qualitatively rather than against a hard band: a longer comet is a broader rhythm, a shorter one is a hemmed-in rhythm, and what matters is the direction your own SD2 moves over weeks. The SD1/SD2 ratio is the easiest of the three to over-read. It describes the shape of the cloud, the balance between fast and slow variation, and it does not map neatly onto “good” or “bad” on its own. A very round cloud (ratio near 1) can appear when fast vagal activity is high relative to slower drifts, and a very stretched cloud (small ratio) when slow drifts dominate. Read it as texture, alongside the metrics that do carry a clear direction.

Work out your own SD1 and SD2

If you have an RMSSD and an SDNN from a reading, you already have everything the Poincaré descriptors need. Enter them and see the geometry:

Notice that SD1 is fixed the moment you type your RMSSD, because it is just RMSSD ÷ √2. SD2 then depends on how much bigger your SDNN is than that width. If you enter an SDNN smaller than your SD1, the formula has nothing left to give and SD2 goes to zero, a sign the two inputs came from different readings or a very short, noisy one.

Reading the plot in POTS and long COVID

In POTS, long COVID and post-viral dysautonomia, the autonomic system tends to sit biased toward “fight or flight,” and the Poincaré plot shows it plainly: a small, tight cluster with a low SD1 (suppressed vagal change) and often a shortened SD2 (less overall variability). It is the same story RMSSD and SDNN tell, drawn as a shape. For how these conditions overlap and why the rhythm locks down, see POTS, long COVID and MCAS.

The encouraging part is that the cloud is not fixed. Vagal tone is trainable, and as people pace, sleep and recover, the comet usually widens and lengthens over months, sometimes before the symptoms ease. That mirrors the RMSSD climb we describe in recovery from post-viral dysautonomia, and it is why SD1 feeds directly into the PNS index, the app’s parasympathetic balance score. The Poincaré plot and the frequency-domain view are two angles on the same rhythm; the VLF, LF and HF power guide covers the other one.

How to actually use it day to day

• Watch SD1 as your fast recovery signal. It moves with RMSSD, so a widening cloud on your usual morning reading is a good sign, and a collapsing one after a poor night or a flare is expected. Follow 7 to 14 days, not one plot.
• Use SD2 as the wider view. A steadily lengthening comet over weeks says your overall variability capacity is coming back, which tracks with genuine recovery.
• Let the shape flag bad data. Stray dots far from the cloud usually mean a missed or extra beat. Clean readings, taken the same way each time, make the trend trustworthy. The measuring-well guide covers the how.
• Read it beside the rest. SD1 and SD2 next to a steady resting heart rate and a shrinking stand-test rise tell a far stronger story than any single picture. The complete HRV guide shows how the metrics fit together.

The bottom line

The Poincaré plot is the most intuitive HRV view you have: each dot is one pair of consecutive beats, a healthy rhythm draws a long wide comet, and a strained one collapses to a tight ball. SD1 is the width, an exact restatement of RMSSD that tracks fast vagal change, and SD2 is the length, a broad measure of overall variability close to SDNN. The SD1/SD2 ratio describes shape, not health, so read it gently. In POTS and long COVID the cloud tends to run small and tight and usually widens with recovery. Measure the same way each morning, follow the trend across weeks, and let the shape sit beside your other numbers rather than standing alone.