saw the 14-years-on-MyFitnessPal-back-at-start post and it stuck with me for a different reason than most of the replies. the scale didn’t lie to that person. it just answered a question that’s noisier than people treat it as. so here’s the n=1 comparison I keep coming back to, because I tracked both for 18 months and they did not move together. scale weight as the long-term signal - moves on water, sodium, glycogen, where you are in your cycle, whether you slept
- a single number with no context about what’s underneath it
- in my own logs the day-to-day noise band was easily 3-4 lb, which is wider than a month of real fat loss for a lot of people
- 14 years of it can net to zero and still hide two completely different metabolic states at the same body weight. that’s the part the back-at-start framing misses A1c / CGM as the long-term signal - A1c is a 3-month average, so it lags but it doesn’t bounce on a salty dinner
- my dx A1c was 8.4. it’s 5.6 now. that number didn’t reverse on me once in 18 months the way the scale did week to week
- CGM is the granular version: my 14-day average dropped 148 to 102 between weeks 4 and 12, and my MAGE (glucose variability) fell steeper and earlier than the mean did
- the hepatic glucose improvement showed up in weeks 4-8, before fat mass moved meaningfully on DEXA that last point is the whole thing for me. for T2D specifically the scale was reporting on body composition while the metabolic improvement was already happening underneath it on a different clock. if I’d only had the scale I’d have called weeks 4-8 a slow start. CGM said otherwise. literature-wise none of this is novel, postprandial excursions being a more sensitive marker than fasting glucose or mean A1c is reasonably well established. what I haven’t seen said cleanly in these threads is that the scale and the glucose data answer different questions and people keep grading the drug on the louder, noisier one. the thing that actually convinced me something real was happening early was graphing my CGM 14-day average against dose-week on the symptom chart in careclinic and watching the line drop before the scale did anything. seeing the two curves diverge on one screen reframed which number I trusted. none of this means weight doesn’t matter. it means if you’ve got 14 years of scale data netting to zero, that’s one variable with a wide noise band, not a verdict on whether your metabolism changed. for a lot of people back at their old weight, an A1c or a fasting insulin from then vs now would tell a different story than the scale does. fwiw I read the A1c line before the scale number on every onederland post now. for anyone tracking both over a year-plus: did your weight curve and your glucose data ever actually disagree, and which one did you end up trusting?
edit: forgot to add