Managing by Reaction

For the first year after Landon's diagnosis, I managed by reaction. A high came through on the CGM and I corrected it. A low woke me at 3am and I treated it. I watched the numbers move, made decisions in real time, and fell back asleep hoping the next alarm would wait a few hours.

I didn't realize it then, but I was managing noise. What I was missing — what nobody really teaches you when you leave the hospital — is pattern recognition.

A pattern is different from a data point. A data point tells you what happened. A pattern tells you why it keeps happening. And once you start seeing patterns in your child's glucose data, you stop chasing individual numbers and start getting ahead of them. Not always. Not perfectly. But enough that the work gets a little lighter.

It took me longer than I expected to learn how to do this. Here's what finally made it click.

Overnight Data Is Your Clearest Window

The overnight window — roughly midnight to 7am — is the most useful place to start because the noise is lowest. No meals, no activity, no stress. Just basal insulin working on its own.

What you're looking for isn't perfection. You're looking for drift.

If Landon's glucose trends slowly upward through the night — starting at 120, creeping to 180, 200 by morning — that tells me his basal may not be keeping up. If it trends slowly downward, the basal may be doing more than it needs to, or a bedtime snack is wearing off earlier than expected. A sharp drop in the middle of the night is a different signal — often the tail of a dinner bolus that lasted longer than expected, or basal that's running too strong for that part of the night.

The key word is slowly. A gradual, consistent drift over several hours is a pattern. A single jagged spike at 2am is probably noise — a restless moment, sensor compression, nothing. The shape of the line over hours matters far more than any individual reading.

And the three-night rule: don't adjust based on one overnight. Look for the same shape — same timing, same direction — on three or more nights before concluding something needs to change. Kids get sick. They have growth spurts. They have random nights that don't look like anything. One bad night isn't a signal.

Meals Are Harder — Because You Need Both Sides of the Story

Meal-related patterns took me much longer to understand, and for a while I couldn't figure out why. The answer was that I was only looking at one side.

The CGM shows you what happened to Landon's blood sugar after eating. But without knowing what he actually ate — and how much insulin we gave, and when we gave it — the curve alone doesn't tell you much. A spike that comes back to range within two hours might mean the dose was close. Or it might mean he only ate half the meal. A glucose line that stays elevated at the two-hour mark might mean we underdosed. Or it might mean he ate twice as many carbs as we accounted for.

The meal log connects the two sides. Once you have both — the glucose curve and what actually went into his body — you can start to see the real story. Which meals his body handles consistently well. Which ones run him high no matter what. What happens to the same meal on a day when he was sick versus a normal day.

The comparison that helped me most: pick a meal Landon eats regularly and look at how his glucose responded across several different days. Same meal, same approximate amount — what does the curve look like each time? If the shape is consistent, you've learned something real about that meal. If it's wildly different every time, something else is the variable. Active insulin left over from an earlier correction. Activity before eating. An illness starting. Narrowing it down is how you learn.

Activity Has a Longer Tail Than You Think

Exercise was the thing that surprised me most when we started actually paying attention.

I knew activity lowered glucose. What I didn't expect was how long that effect lasted. A big afternoon at the playground — running, climbing, the kind of physical output that completely wipes out a toddler — doesn't just lower glucose during the activity. It lowers it for hours afterward. The muscles are still restoring glycogen, still pulling glucose out of the bloodstream, long after everyone is home and dinner is done and Landon is asleep.

Which means the 3am low that follows a big Tuesday afternoon at the park isn't a basal problem. It's the afternoon catching up with us.

Once I understood this, I stopped making panicked setting changes after nights that followed high-activity days. I just noted it, treated the low, and waited to see if the same afternoon produced the same night again. It usually did.

The other thing worth knowing: not all activity works the same way. Sustained aerobic movement — running, swimming, extended play — tends to pull glucose down. Short, intense bursts — sprinting, jumping, high-intensity games — can actually spike glucose temporarily because the body releases stress hormones that push glucose up. Landon gets both, sometimes in the same afternoon. I'm still learning to read it, but most of the time now I can tell which kind of afternoon he had just by looking at what his glucose is doing before bed.

How to Actually Do This

None of it happens automatically. Patterns emerge from data, and data requires logging. Not perfect logging — just enough consistency that you can compare similar situations over time. A few things that genuinely helped me:

• Start with one window. Don't try to analyze everything at once. I spent weeks just watching overnights before I touched anything else. Once that felt solid, I moved to breakfast. Trying to see all the patterns at once is overwhelming.
• Look at a week at a time, not a day. A single day is almost always too noisy to learn from. Seven days together start to show structure. Fourteen is better.
• Flag the outliers before they skew your thinking. Sick days, travel days, days where the sensor was clearly off - note them and set them aside. Build your baseline from the typical days.

What It Actually Gives You

Pattern recognition doesn't make T1D management easy. But it changes the nature of the work. You go from reacting to individual numbers to understanding the forces underneath them — the meals that consistently run high, the activity that catches up at 3am, the overnight drift that needs a basal conversation with your endo team, not a middle-of-the-night juice box.

It also makes appointments more useful. Instead of showing up with a wall of CGM data and hoping something stands out, you arrive with specific questions. Those conversations move faster, and they lead somewhere.

You're already collecting the data every day. Learning to read it is what turns it into something you can actually use.

— Jordan, GlucoLab founder and T1D dad