Fasting glucose stuck at 118 while A1c keeps dropping?

Same pattern. A1c moves first; fasting hangs around 120 for another 6 weeks. CGM tracking finally showed less 4-6am rise once sleep got better.

Sleep is the wild card fro me. Bad apnea night = 128 fasting. Good night = 104. Same food, same meds.

Wait, 104 on a good night? I never get below 110 no matter what. Are you waking at same time or just using CGM overnight avg?

Same boat. My A1c went 8.1 to 6.9 by week 14 but fasting stayed 118-130. Endo said liver glucose output can lag. Annoying but not unheard of.

Similar window here, 13 weeks of tracking. A1c 7.7→6.5, weight down 21 lb, fasting still bouncing around 112-126. What finally changed the read was pulling the overnight CGM data instead of leaning on the single morning fingerstick — there's a slow climb starting around 3am, not a sudden dawn spike. Morning fingerstick also runs 8-12 points higher than CGM at the same time, which is enough to drive a person nuts. Comparing both side by side killed the urge to overreact to one number. The morning plateau is still the stubborn one. It's the last thing to budge, and A1c has been flat for a month.

What did your CGM say around 2-3am? If there's a dip before the rise, that changes the picture. I had that pattern and my clinician adjusted my evening meds. Fasting went from 122 to 106 over a month. Still not perfect but trending. Also OP, 17 lb in 12 weeks is no small thing.

My fasting was 118-124 every single morning around week 11 while my A1c had already slid from 8.0 to 6.7, and my CGM showed that same 4:15-4:45 ramp — so I get how backwards it feels. For me it wasn’t just lag; it was a late-ish dinner plus trash sleep, so my liver was doing its early-morning dump while my afternoon numbers looked great at 95-105. The surprise was that eating dinner 30 min earlier barely moved it, but a 15-20 min walk after my last meal and a locked-in 10:30 bedtime (actual sleep, not just lying there) knocked my wake-up number into the 100-108 range by week 14, then under 100 around week 16. If you’re already logging food and sleep, try tagging your CGM with hours slept and time of last bite, not just “slept 8 hours” — my “8” was usually 6.5 and the 4:30 rise was

Weeks 10-14 my CGM had a split personality: weekdays it would ramp at 3:45-4:30 and I’d wake up 112-120, but Saturdays/Sundays I’d sleep until 7:30 and fasting was 98-104 — same dinners, same-ish carbs, only difference was no 5:45 alarm and no stressful commute. The weird reversal was that tweaking my evening meal did almost nothing; adding 45 minutes of sleep and a 10-minute walk after my last meal dropped my weekday fasting about 10 points by week 16, while my A1c had already done its big drop earlier. So now I compare weekday vs weekend CGM traces instead of panicking over the 6am number; if your 4:30 rise tracks with your wake time or stress, that’s a totally different lever than food.

I was logging food and sleep too, but the missing column for me was last-bite-to-bed time and overnight wakeups. My mornings sat 116-122 from weeks 9-13 while my A1c kept dropping, so I assumed it was pure lag. The surprise: a 6:15 dinner with rice or potatoes gave me 104-110 by week 14, while my “healthy” 8:45 low-carb dinner had me at a 4:15 rise and a 119. Sleep was the other half—my tracker said 8 hours in bed but 6h05 actual, usually two wakeups; 10 days of 7h15 actual moved the rise closer to 6 and dropped my AM number about 10 points. What’s your CGM dong at 2-3am—flat then climbing, or already drifting up?

I’d look at your before-bed number and the delta to morning, not just the morning value. In year two mine did the same: A1c kept sliding, fasting parked at 118. But my 10pm reading told the story. If I was 105 at 10 and 120 at 5, that was the dawn ramp. If I was 125 at 10 and 120 at 5, it was just a high baseline. My morning number also tracked evening movement way more than dinner, weirdly. What’s your 10pm number?

My dorm microwave and I are both slow to warm up; fasting was last for me too 😂

My spreadsheet calls my 4:30 rise the stubborn tab too—solidarity.

hi, I’m brand new here (week one), so sorry if this is obvious. Does the 4:30 rise happen on nights you wake up more, even if total sleep looks okay? I’ve only got a few days of CGM data, but my morning number seems to track how broken my sleep was more than dinner. Also, are you logging the time you actually get up vs. when you take the fasting reading? I realized my “fasting” was sometimes 90 minutes after I woke, so it wasn’t the same morning number. Just trying to learn.

Fasting glucose is the system’s baseline, not its response—so it often moves last. A1c can drop while morning stays sticky because post-meal wins are easier. Check the shape, not just the 4:30 rise: does your CGM show a dip around 2–3am first? If yes, it’s a rebound solving a different equation. If it’s flat then ramps at 4:30, that’s more dawn/liver output. I only started trusting my morning number after logging the 3am nadir plus sleep stage. What’s your 3am value?

my dawn rise shows up before my alarm too — stubborn little tab 😂

One thing that moved my morning number more than my fasting window: dinner timing. If I finish eating after 8, my 4:30 rise starts earlier and lands 10–15 points higher than when I’m done by 6:30. A 15-minute walk after that earlier dinner flattened my whole overnight line. Alcohol, even one drink, did the opposite. Have you lined up your CGM overnight curve with dinner time? That comparison was way more useful to me than the morning number alone.

Mine does that early-morning creep too—stubborn as a mule. Hang in there, friend!

Mine creeps up around

Mine creeps up pre-alarm too,

same, my cgm shows a little 4am bump and i just

Mine did that same stubborn morning dance for ages—A1c moved first. There’s an old thread here on it.

One thing that surprised me: my morning number tracked my dinner timing more than my overnight curve. If I ate after 8 or had a bigger carb-heavy dinner, I’d wake up 118–125 even with good sleep. Earlier dinner, a short walk after, and no late grazing got me to 105–112 most days. A1c and weight moved first for me too, which felt backwards. Are you logging dinner time and what the meal was, or just total daily carbs/calories? That was the missing variable in my own spreadsheet.

night shift angle: what counts as “morning” for you? When I worked nights, my fasting number tracked my sleep cycle, not the clock. Test after fragmented daytime sleep, looked awful. After a full chunk, way better. Log sleep start/end and wake time next to the number. Are you testing at the same point in your sleep-wake cycle daily, or just before your shift? That’ll tell you lag vs circadian mess.

tbh night shifter here. My morning number tracked the gap between my last bite and actual wake-up, not the clock. If I ate at 2am and woke at noon, the 8am “fasting” reading was useless. Split sleep also jacked mine up more than anything. The 4:30 rise might just be your body switching fuel before you get up. What’s your wake time and last bite time? Log that gap for a week. Mine looked stuck until I lined those up. A1c can drop while mornings lag.

contrarian angle: my fasting number was a lagging sleep marker, not a lagging food one. I exported CGM data and my 12-week trendline showed A1c down 1.3 while the 4:30 rise barely moved. The only variable that predicted it was sleep efficiency and wake-after-sleep-onset, not dinner. What’s your CGM delta from 3:00 to 4:30? If it’s flat then jumps, that window is setting the morning number. Also, is your “fasting” draw at 7am vs 9am? Mine can be 118 at 7 and 98 by 9 with zero food.

Honestly, cafe owner here: 4:30 is my “did I eat that sample?” alarm too. Morning numbers are so weird. Solidarity! ☕

Road warrior here. My hidden variable was hotel wine/airport beer: even 1–2 drinks around 8pm gae me the same 4:30 CGM bump and a fasting number 10–15 points higher, with identical dinner and sleep. A1c kept trending down because my daytime numbers were great. Do you log alcohol on travel nights, or are you already dry? If not, that might be the “something else” to peek at.

Something that finally cracked mine: I logged when I actually woke against the CGM rise. My 4:30 bump wasn’t dinner, it was a full bladder plus a too-warm bedroom. I stopped fluids after 7, cracked a window, and my fasting slowly came down ~8 points over two weeks. A1c dropping while fasting lags was true for me too—morning number was last to move. Do you wake up at 4:30, or just see the line? If you’re awake, is it bathroom, stress dream, room temp? That might be the hidden variable.