Fasting on Yom Kippur — Part 3: Why Fasting Is Risky for Someone With Type 1 Diabetes

BS"D

Blood sugar is the part we have tools for. The real danger is a rise in acid that no home device can measure — and a number nobody can give you.

In this post:

• It Is Not the Blood Sugar Anymore
• What Insulin Is Really For
• The Trade a Healthy Body Never Has to Make
• The First Blind Spot: Nobody Can Measure the Acid
• The Second Blind Spot: Nobody Knows How Much Insulin the Body Needs
• This Is Not Theoretical
• So Can a Person With Type 1 Fast?
• The Problem With Symptoms — and Why It Does Not Matter
• The Settings Have to Be Right Before You Start
• If You Take Jardiance or Another SGLT2 Inhibitor
• A Word on the Low-Carb Post

It Is Not the Blood Sugar Anymore

When a person with Type 1 diabetes thinks about the risk of fasting, what comes to mind is blood sugar. Going low. Going high. Getting through the day with numbers in the middle.

Watch your numbers. Watch them closely, closer than on an ordinary day.

But that is not the reason fasting is risky.

Blood sugar is the part we have tools for. A sensor on your arm shows you every number, and an AID system (a pump with automated insulin delivery) is on the job all day long. A low can be seen coming. A low can be treated. For twenty-five years we have been publishing guidance for this fast. Now more than ever, the blood sugar part can be trusted to the AID system.

The risk sits somewhere else. And the way we handle the blood sugar can actually make the real problem worse.

What Insulin Is Really For

The way we think of insulin is simple: it is the thing that lowers blood sugar. Take more, the number comes down. Take less, it goes up. That is how we use it all year, and it is how the pump talks to us about it.

Lowering blood sugar is one of the jobs insulin does. But it is not the only one.

Insulin is also what tells the body it does not need to break down its own fat for energy. When there is enough insulin, that process stays under control. When there is less available insulin, the body starts burning fat hard and quick, and burning fat produces ketones. Ketones are acids.

Which brings us to the scary word: DKA, diabetic ketoacidosis. Ketones, acid, and diabetes — all three parts are in the name.

But a lot of ketones is not the same as DKA. The body can absorb a good deal of acid and keep the blood normal — through the kidneys, through breathing, through its own buffers. People on a strict keto diet carry ketone levels that would look alarming on a strip and are perfectly fine — on that diet, ketones are the goal.

So what actually tips it into DKA? Two things together. Production has to be running uncapped, which is what happens without insulin, since insulin is the brake on fat breakdown. And it has to run long enough or hard enough to overwhelm the body's ability to neutralize it. That is the difference between a lot of ketones and DKA — not the number on the strip, but whether anything is holding the production down.

Take less insulin than your body needs, and you are bordering DKA. That sentence has nothing to do with your blood sugar number. A person can be at 110 and be on the way to DKA. That has a name — euglycemic DKA, DKA at a normal blood sugar — and it is real.

This is what I have been saying for years, and this is what I mean by it: insulin is life.

The Trade a Healthy Body Never Has to Make

A person without diabetes who fasts does not get there. Their body keeps insulin at exactly the level it needs — enough to hold the brake on fat, all day — and it can afford to, because if the sugar drops, glucagon comes in and pushes it back up. Two hormones, one lowers and one raises, and the body plays them against each other minute by minute. It never has to take the brake off to avoid a low.

We have neither one working on time. Insulin comes from a pump or a pen, and there is no glucagon behind it to catch a fall. So every unit we keep on board during a fast is a real risk of going low, and the only way to remove that risk is to take insulin off. That is exactly what an AID pump does, by itself, every time the number drifts down. Which is taking the brake off.

One day there may be a dual-chamber pump that delivers insulin and glucagon together. Whether that makes fasting safer or harder is something we will find out when it arrives.

That is a fast day for us. No food, so no boluses. Basal coming down, by you or by the pump, because there is nothing coming in and the numbers are drifting. Insulin dropping for twenty-five hours. That is why DKA is the risk on the table on a fast day, and it is the whole reason we write about this every year. The mechanism itself I laid out in last year's fasting guidance.

And it is what I meant at the top. The fix for the blood sugar is the thing that walks us toward the real problem.

The First Blind Spot: Nobody Can Measure the Acid

DKA is acid in the blood. So the obvious question is why we do not simply measure it, the way we measure glucose.

There is no home monitor for it. Not for sale, not in development that I know of, not at any price. In a hospital they draw blood and check it in a lab. At home you have nothing.

Ketones are the closest thing we have, and people treat them as if they were the same measurement. They are not. A ketone meter measures the cause. It does not measure the condition. How to check — the meter, the strips, and what is permitted on Yom Kippur — I went through in last year's post on ketone testing.

And there is a bigger problem with using ketones as the warning. Every human body makes ketones during a long fast, with diabetes or without. By the middle of a fast day, everyone in shul has them. So if we tell people to check and to act on a positive result, we have told most people with Type 1 to break the fast over something that is normal for the day. None of us would fast at all.

Which sounds like a reason to drop ketone checking altogether. It is not. Go back to what decides DKA. A person without diabetes does not usually run to a high level on a fast day, because their own insulin is still holding the brake. Ours is not — the only brake we have is the insulin we take.

So for us a genuinely high level on a fast day cannot be explained away as what fasting does. It might be. It might not be. That is a safek of pikuach nefesh — a doubt, not a certainty — and halacha does not ask you to resolve it. It asks you to act. A safek is not a reason to keep watching.

This is where the guidance we had from Dr. Zwickler comes in. Checking during the fast has a real place — not so a mild reading can end somebody's fast for nothing, but in the hours before you feel anything at all, when there is still something to be done about it. His numbers are in the 5785 guidance, and how to check is in last year's post on ketone testing.

The Second Blind Spot: Nobody Knows How Much Insulin the Body Needs

This one is worse.

You know one number about your insulin: how much of it holds your blood sugar where you want it. That is what a basal rate is. That is what a carb ratio and a correction factor are. Every setting in your pump answers that one question.

There is a second number. How much insulin does your body need so that acid does not start building up faster than the body can handle. That number is not the same as the first one, and nobody has it.

Your doctor does not have it. The pump does not have it, and was never built to look for it. There is no blood test that produces it, no report that prints it, no company working on a screen that displays it. It is not that you were never told. It does not exist.

So every reduction on a fast day — yours or the pump's — is measured against the only number we have, the blood sugar one. And the number that would tell us we have gone too far is the number nobody can see.

You will not be warned. There is no reading that changes color, no alarm, no line on a graph. Whatever the crossing point is, it is crossed quietly.

This Is Not Theoretical

I have seen DKA come out of a fast more than once over the years. Some of those cases resolved at home and some ended in a hospital.

One of them was a young teenager, more than thirty years ago. He started the fast, and by the afternoon he was in DKA. Afterwards he went to Rav Fishel Hershkowitz z"l, who told him: never again. Not even to start. He has not fasted since.

That is the other cost of DKA, and people do not think about it. Get there once and you may never be allowed to fast again. So the care this post is asking for is not only for this Yom Kippur. It is for every one after it.

And a few weeks ago a member sent me his own account of something that makes the point sharper than I could. He went into DKA with normal blood sugar readings. Not high. Normal. His sensor was showing him numbers he had no reason to worry about while his blood was turning acidic, and by the time he felt what was happening he was very sick. That is euglycemic DKA, the thing I named above, in a real person. I published his letter in full in the ketone post I was going to write.

His case had a specific cause, which I come back to near the end of this post. Read it for the part that applies to everyone: the glucose number was never going to warn him. It is the clearest illustration I have of what the two blind spots mean in a real person on a real day.

So Can a Person With Type 1 Fast?

Yes. We still believe a person with Type 1 diabetes can fast, and Part 1 of this series laid out where that position comes from and what it rests on.

But hold on to what you just read. DKA during a fast is real, and neither of the two instruments that would catch it exists. We cannot measure the acid, and we cannot know the exact amount of insulin required.

So watch two things. A ketone check, in the hours before you feel anything, can tell you it may be on the way — that is what the guidance from Dr. Zwickler is for. What your body tells you is the other half, and it is the half no device can give you. When it speaks, it is telling you it has arrived.

The Problem With Symptoms — and Why It Does Not Matter

Some of what DKA feels like is also what a long fast feels like.

Take nausea. Plenty of people feel nauseous on a fast day for no reason other than that they are fasting. Nausea is also the first symptom of DKA. Looking at your own nausea in the middle of the afternoon, you cannot tell which one it is. Neither can I, and neither can your doctor over the phone.

People treat that as a reason to wait and see.

It is the opposite. Same rule as the ketones: halacha does not ask you to know. The very fact that you cannot tell the two apart is what settles it.

So: at the slightest hint of nausea, the full fast is over. Not after it gets worse. Not after another hour to see whether it passes.

There is a hard reason for that timing. What usually follows nausea is vomiting, and vomiting is the door closing. Insulin is the only thing that reverses DKA, and insulin cannot be taken safely without something to eat or drink. Once you cannot hold anything down, you cannot take the insulin, and the thing you needed to stop is now unstoppable at home. That is how a fast day ends in an emergency room.

And do not spend one second on a strip. You already know what you need, and it is insulin. A reading cannot change that, and the minutes it costs you are the minutes that matter most.

While the nausea is still mild, sipping something sweet slowly is usually manageable, and then insulin can go with it.

The other symptom worth knowing is the breathing. Heavy, deep breathing, or shortness of breath — people describe it both ways, and either one counts. Some feel no nausea at all and notice the breathing first. On a fast day that is not ordinary, and it should be treated exactly the way nausea is.

I am deliberately not giving you a longer list, and I should say why, because everything above could be read as "any symptom at all, break the fast." That is not the rule. A safek is a real doubt, not any possibility you can imagine. Headache, weakness, tiredness — half of the people fasting have those by mid-afternoon every year. They are what fasting feels like, and it is very unlikely that they are DKA. Nausea and heavy breathing are the other way around: on a fast day they are much more likely to be DKA than not, and that is a real doubt. A remote possibility does not end a fast. A real doubt does.

And a person who breaks a fast on a maybe has not gone against Hashem's instructions. He did what halacha asked of him. He tried to fast, and he did.

The Settings Have to Be Right Before You Start

Symptoms are the last line. Everything before that is making sure you never get there.

For years my guidance was built on temporary basal rates. In the first version we sent out an Excel file: you entered your own basal settings, the file gave you the reductions for the whole day, and you built a new profile in the pump and switched it on as the fast began. Later I revised it and told people to keep adjusting the temp basal through the day instead. Some of that is still in the 5785 guidance.

That is old wisdom now. Most people are on an AID system, and several of those systems will not even let you set a temporary basal rate while automation is running. Not because the pump is holding back the reduction — because the pump is already making it. That is the whole change. In those years you did the arithmetic: minus twenty percent at this hour, minus thirty at the next. Now the pump does it, every five minutes, without asking you, and it will keep going further and further, because the only thing it looks at keeps telling it to.

In those years I had one hard warning, and I gave it to everyone: never below fifty percent. Now there is nobody to give it to. The pump is doing the reducing, and the pump is not listening to me. I have a separate post coming on exactly that.

I have said for years that a good fast last year is no prediction for this year. People have fasted well for ten years and then landed in DKA on the eleventh. This year it is sharper than that. If your good fasts were on temp basals or a special profile, you were the one deciding the insulin, hour by hour, and you were the safety margin. Handing that to an AID running on its own is not the same fast with a better tool. It is a different fast, and your years of experience are not experience with it.

Since most people want to stay in automation for the fast, the pump is the one deciding how much insulin you get for twenty-five hours. And the pump sees one thing: blood sugar. It does not see ketones. It does not see acid. It does not see whether there is still enough insulin on board to hold the brake on fat. It is the second blind spot again, except now there is a machine sitting in it, making every decision against the only number it can see.

So the settings have to be right before the fast starts. Not adjusted during it.

The practical piece — what to set, on which pump, and what to leave alone — is coming as its own post in this series, bez"H, before Yom Kippur.

One thing belongs here rather than there, because it is not a setting. If you are the person who keeps taking insulin off — down another notch at noon, another at two, another at four, each one reasonable on its own because the graph is drifting down — you are the person this post was written for. Every one of those reductions was made against the number you can see. Nobody was checking the other one.

If You Take Jardiance or Another SGLT2 Inhibitor

This is a small group, and most people with Type 1 can skip it. But for those it applies to it matters a great deal, and it is what put the member whose letter I mentioned into the situation he was in.

These medications change how the body handles glucose and make DKA possible, and faster, even at normal blood sugar readings. Fasting is exactly the kind of day where that combination shows up.

If you take one, call the doctor who prescribed it and ask how many days before the fast you should stop taking it. Ask now, not on Erev Yom Kippur. Do not work it out from this post and do not stop a medication because a blog told you to — the answer belongs to the doctor who put you on it.

A Word on the Low-Carb Post

In Part 2 I wrote that people who have been eating low carb for the weeks before a fast tend to have an easier one, and I explained why. A few of you will read this post and feel that the two contradict each other. If low carb means running on fat, and running on fat means ketones, does that not walk you toward the danger described here?

It does not. Go back to what decides DKA. It is not where your ketone level started. It is whether anything is holding the brake on fat, and for a person with Type 1 that brake is the insulin you take. Someone who began the fast already in mild ketosis from low carb, and someone who began at zero, are in exactly the same position on the one thing that matters: is there enough insulin on board. The starting number does not buy the first person a head start, because the head start is not what causes DKA. Losing the brake is.

Low carb changes where you start on the ketone scale. It does not change the thing that decides DKA — and that thing was already the whole question for every person with Type 1 who fasts, on any diet.

· · ·

Fasting with Type 1 diabetes is possible. It is possible with a clear head about what we can see and what we cannot. We can see the blood sugar, and we handle it better than we ever have. We cannot see the acid and we cannot see the insulin required, and no device is coming to fix that this year.

So watch your numbers, get the settings right beforehand, and then listen to your body — and act the moment it says something, even if you are not sure.

Review your own plan with your doctor before the fast.

Wishing you a kesivah vechasimah tovah, and an easy fast.

— Rabbi Hirsch Meisels

P.S. More in this series before Yom Kippur, bez"H — the pump settings, and the automation question.