Week 6 on semaglutide: fasting glucose up, A1c down?

Same pattern around week 5. My 3am dip then 4:30 climb. Are you sleeping worse? Poor sleep spikes my mornings more than food.

My sleep tracker showed hte same rise after 4am. I changed nothing except eating dinner earlier and the 6am number dropped ~12 points.

I had A1c fall from 8.1 to 7.2 over 12 weeks while fasting went from 140s to 160s for a few weeks. Endo ordered cortisol and thyroid, both normal. What helped me understand it was CGM: my overnight lows weren't real lows, just compression lows from sleeping on sensor arm. Once I moved sensor, the 3am dip disappeared. Not saying that's your case, but worth checking sensor vs fingerstick. Did your A1c drop match time in range or mostly post-meal?

Compression lows fooled me for two sensors. My fasting looked worse until I fingersticked at 4am. Have you compared your CGM to a meter those mornings?

My dawn numbers shifted when I took metformin later.

A1c 7.8 to 7.1 in 8 weeks is a big move. My fasting was the last thing to budge, like week 10.

I saw the opposite: fasting dropped first, post-meals stayed high. Weird how it varies. What was your pre-breakfast reading vs 2h after? My CGM showed a sharp 5-7am rise even when bedtime was 110. I ended up logging it as 'dawn + delayed gastric emptying' but never got a clean answer. Following this thread because your week 6 timing is exactly when mine went sideways.

Following too — seems like week 7 is where things tend to go sideways for a lot of folks.

Week 6 was my worst stretch too — CGM showing a 3am dip to about 92 then a march up to 150-160 by 6am, and I was sure something was broken. What actually moved it for me had nothing to do with my meds: I was eating dinner at 8:30pm and going to bed at 10:30, so I flipped dinner to 6:15, kept it protein-heavy, and did a 20-minute walk after, and fasting came down 15-20 points in about two weeks. Bad sleep nights alone were worth +12 for me, so I quit checking at 5am when I'd been up since 3. Contrarian take: I stopped chasing the fasting number and started tracking my waist with a tape every Sunday instead — down 2.5 inches while fasting was still bouncing around, and that kept me from spiraling every morning.

Two variables I rarely see logged: dinner-to-bed gap and sleep fragmentation. My fasting tracked the nights I ate late + slept <6.5h, even while the 90-day average kept dropping. Moving dinner earlier and keeping pre-draw conditions identical (same time, same fast length) cut my morning scatter by ~20 points in 10 days. Are you comparing lab fastings or CGM wake values? If CGM, what’s your overnight sleep and last meal timing?

one thing i never see logged: what time you actually take the "fasting" reading. with slowed emptying, my dinner is basically still landing at 7am, so that morning number is more like a leftover-dinner number than a true fasting one. if i test the second i wake vs after 20 min of moving around + coffee, i get a 20-30 pt spread. also a1c is a ~3 month average so 6 weeks barely nudges it while fastings react fast. are you testing before your feet hit the floor, or after you're up and around a bit?

not sure if this is your deal, but the variable that wrecked my fastings was caffeine timing, not dinner. i was sipping zero-sugar cola/energy drinks until like 2am while gaming, then wondering why 7am was 150. cut everything except water/decaf after midnight for two weks and my morning numbers dropped maybe 10-15 pts. a1c still lagged behind. what's your caffeine cutoff, and do you log it separately from actual food?

something that finally showed up in my logs: my fasting number correlates way more with the previous day’s step count than with what I ate. Under 4k steps = next morning runs 10–15 points higher. I started putting my daily steps next to my fasting reading and the pattern was stupidly obvious. Have you looked at activity the day before? Might be another variable to add to your chart.

been lurking this thread since it started. one thing that ate me alive for a week: meter and strip variance. same drop of blood, two meters, 18 points apart. started double-sticking back to back and realized like half my "week 6 spike" was a sketchy strip batch. also my left index consistently reads a bit higher than my right, no clue why but it's reliable. did you ever run a control solution or compare against a lab draw? that's the first thing i'd rule out before blaming the med or your sleep.

same, my 'fasting' was just whenevr i finally woke up lol

one weird one for me was late-night liquid intake. i was doing the soda swap thing with fizzy water, but drinking a ton after midnight meant i was up peeing 2–3x. on those nights my morning number was consistently 10–15 higher than on nights i actually slept through. food was basically the same. have you got sleep/wake-ups logged next to the fastings? that was my missing column.

If you have CGM data, look at the 2–4am slope, not just the 7am stick. Mine climbed overnight even when bedtime was flat; logging awakenings showed it tracked sleep fragmentation, not dinner. Also remember A1c is a ~90-day weighted average—six weeks of improved postprandials can lower it while dawn-driven morning numbers stay high. So the discordance isn't necessarily paradoxical. What does your overnight curve look like before the morning rise starts?

One thing I didn’t see mentioned: evening meal composition, especially late carbs vs. protein/fat. I started logging dinner macros next to my morning fasting number. My fastings ran 15–20 points higher on nights I had a big pasta dinner vs. steak and veg, even when calories were similar. A1c kept drifting down over months. Not claiming cause/effect, just a pattern. Are you tracking dinner timing/macros, or mostly glucose? That was the missing variable for me.

Mine’s the same: A1c down, morning number acting like it missed the memo.

Cafe owner here, so portion control is a joke when I’m quality-checking pastries. The thing that explained my split numbers was “bite creep”: not meals, but spoon-licks, broken cookie halves, and latte milk tastes between 4–7pm

Mine did the opposite of what I expected around week 5: fasting crept 118→142 while A1c slid 0.9. The only thing that lined up wasn’t food—it was a run of 5-hour sleep nights plus a mild cold. Once sleep normalized, morning numbers drifted back over ~10 days. Fasting is a spot price; A1c is the rolling average. If you’re already logging CGM, add sleep duration and any illness/stress flag for the same window. Does your morning number track more with sleep debt or with the previous day’s activity?

My dawn numbers were the last to RSVP too — A1c showed up early, fasting rolled in fashionably late. 🙃

Week one newbie here, so glad y’all share this stuff — makes me less nervous!

My fastings love to be the last to get the memo too. Solidarity, friend.

I live out of hotels and airports, and week 6 hit during a three-city stretch. My morning numbers ran ~15–20 higher, but my CGM 14-day average was basically flat. The pattern wasn’t dinner — it was sleep and timing: short nights plus 5 a.m. airport coffee made fastings look ugly, while a boring hotel-room protein snack and blackout curtains calmed them down. If you’re comparing weeks, log travel and sleep next to glucose. Anyone else’s fastings act more like a jet-lag gauge than a food report?

My mornings pulled that same fake-out around week 6 too — hang in there, it’s such a head-scratcher.

In the trials, A1c is the weeks-long average, so it can fall while individual fastings bounce around. What helped me stop panicking was overlaying my CGM overnight curve with boring stuff: last bite time, alcohol, sleep length, and whether I woke at 3am. My “fasting” rise was usually a slow 2–4am climb after a late dinner plus a dawn surge, not a sign the whole day was worse. Are you wearing a CGM? If so, does the rise start before or after you wake?

Something that helped me was separating the dawn window from the all-day average. I tracked fasting at the same time each morning, plus a weekly average of pre-dinner readings. Around week six, my fastings climbed while the weekly average kept sliding. When I added notes for sleep, evening meal timing, and hydration, the climb lined up with short sleep and later dinners, not with the

One thing that helped me stop spiraling: I logged my bedtime number right next to the morning one for a week. When fasting popped, my bedtime was drifting up too, but my 2–3am check was flat — so the rise was happening just before wake-up, not overnight. That made me trust the longer trend line instead of one bad morning. Short sleep and late dinners moved my morning number more than I expected. Do you have any overnight data, or just the wake-up checks?

I’m a spreadsheet nerd, so I added sleep hours and “dinner-to-bed gap” next to my morning numbers. My week-6 fasting highs clustered on nights under 6 hrs sleep and meals less than 2 hrs before bed. 7-day fasting avg went 132 → 148 → 126 when I ate earlier and slept better, while A1c kpt drifting down. Not claiming cause, it was just the only pattern that lined up for me. Did yours correlate with sleep or late dinners at all?

Airport rat here. I added a travel flag to my log: H = home, A = airport/road, Z = new time zone within 48h. My fastings run higher on A/Z mornings, while the weekly trend looks boring-flat. Week 6 looked dramatic for me too—three flights, hotel AC, vending-machine dinner—until I had four normal home mornings and things chilled out. Are your 155s clustered on hotel mornings, or showing up at home too? That would tell me if it’s travel-logistics noise or a real pattern.

The variable I accidentally discovered: what time I actually test. On 5 a.m. flight days my fasting looks scary-high, but if I wait till my normal 7:30 at the hotel it drops into my usual range. Same sleep, same dinner. Now I log “test o’clock” next to the number, and it stopped me from spiraling over what was basically a schedule artifact. Anyone else see a bump just from sampling earlier? Or is that just my body being dramatic