The supplement question is one I came at sideways, because tracking glucose for 18 months meant I noticed something hormonal before I had a name for it. Around last autumn I started seeing mid-cycle glucose spikes that didn’t match my food log. Not dramatic, maybe 8-10 points above my usual fasting range, but consistent across three weeks so I started logging them separately. Protein intake, fibre, steps within 30 minutes post-meal - all unchanged. I’d ruled those out before I raised it with my endocrinologist. She connected it immediately to estrogen. Estrogen directly affects insulin sensitivity - when it drops in the luteal phase, and especially when levels become irregular in perimenopause, glucose can climb even without any dietary explanation. The Libre was essentially giving me a hormonal signal before I’d developed the classic symptoms people describe. That’s the part I hadn’t read anywhere: in T2D, perimenopause may register in blood glucose before it shows up elsewhere. On supplements specifically - I’m cautious. Magnesium glycinate has a reasonable evidence base for sleep and some glucose metabolism data behind it, and the interaction profile with antihypertensives looks manageable. Most of the rest I’ve looked at is thin, or the studies are so confounded I can’t pull anything useful out. The “check with your GP” instruction is genuinely frustrating when your GP isn’t curious, but for BP medication interactions that particular caveat earns its place. Worth pressing your prescriber specifically on that question rather than the general supplements conversation.
the “mid-cycle glucose spikes” framing is the one piece I’d press on. mid-cycle in most cycle-tracking convention means the ovulation window, when estrogen is peaking, not dropping. the insulin sensitivity hit you’re describing is the luteal pattern (post-ovulation, estrogen falling while progesterone climbs), which is a distinct window. matters for charting because if you’re tagging the wrong phase the magnitude question and any supplement-or-not call gets confounded. the underlying mechanism holds and matches what I see in my own logs. luteal bump ran ~10 mg/dL at 7.5mg tirz and narrowed to ~5-6 after titrating up. still present, not flat. that directional shift suggests the effect is partially dose-modifiable but the ceiling question is unresolved and probably can’t be answered from a single dose. perimenopause-on-T2D overlap is barely documented as its own territory afaik, and the “CGM picked up the hormonal signal before vasomotor symptoms” piece is one of the more interesting patient-level observations I’ve seen. another tracker in this community posted ~6 wks of vasomotor-to-CGM correlation data that lined up with this directionally. on magnesium glycinate, agree the evidence base elsewhere is thin. the antihypertensive interaction caveat is the one part of “ask your GP” that earns its place imo, since it’s a specific drug-class question rather than the usual generic punt.
night sweat fragmentation is the confound nobody named here yet: cortisol elevation from broken sleep increases fasting glucose independently of the estrogen-insulin sensitivity pathway, so even clean CGM data might be showing two mechanisms at once with no device-level way to separate them. if the spikes cluster in the early fasting window or overnight rather than postprandially, that’s a mild pointer toward the cortisol route rather than pure luteal hormonal effect. probably additive rather than either/or, and both mechanisms suppressing insulin sensitivity from different angles makes dose titration targets less stable than they look when you’re optimizing for just one. iirc follicular-to-luteal sleep architecture differences, longer onset latency and reduced slow-wave, show up in actimetry before vasomotor symptoms become clinically obvious, which maps to the “CGM ahead of symptoms” framing in my OP. my own logs don’t have sleep quality tagged against the glucose data, which is the gap i’d go back and fill.