- Guest: Dr. Cecilia Nobili (Pediatric Diabetology Resident, Regina Marcherita Children’s Hospital, Turin, Italy)
- Host: John Pemberton
Available on:
Why this episode exists
The menstrual cycle affects roughly half of people with type 1 diabetes, yet it remains largely invisible in clinical guidelines, research literature, and algorithm design. Women report predictable patterns of insulin resistance in the days before their period, increased hypoglycemia risk when bleeding starts, and intense frustration managing glucose levels that swing wildly despite doing “everything right.” Dr. Cecilia Nobili, a physician-researcher living with type 1 diabetes herself, bridges the gap between lived experience and clinical evidence. Her observational study of 170 women reveals how different insulin delivery systems handle monthly hormonal shifts, which phases create the biggest management burden, and why this represents a genuine gender gap in diabetes care. This episode provides the roadmap that should exist in every clinic.
The menstrual cycle roadmap
- Early follicular phase (days 1-7): Bleeding starts, estrogen rises, progesterone drops — insulin sensitivity increases sharply, creating high risk of hypoglycemia if insulin doses aren’t reduced quickly.
- Mid follicular phase (days 8-12): Relatively stable glucose patterns for most women — this is the “easy” window where usual insulin strategies work best.
- Periovulatory phase (days 13-15): Ovulation occurs around day 14 (varies by cycle length) — some women notice glucose changes here, but this phase is highly individual.
- Mid luteal phase (days 16-21): Progesterone begins climbing, insulin resistance starts building — early intervention here prevents worse highs later.
- Late luteal phase (days 22-28): Peak progesterone drives maximum insulin resistance — carb cravings intensify, time in range drops, pre-bolusing becomes critical, and frustration peaks.
- The progesterone effect: Progesterone is the primary driver of insulin resistance during the luteal phase — it creates the “perfect storm” of needing more insulin while craving high-carb, high-fat foods.
- Not universal: Roughly 60% of women on MDI experience clinically significant glucose deterioration (≥5% drop in time in range), but patterns vary — some women have minimal cycle impact, others experience dramatic swings.
Key takeaways
1) This is a gender gap in diabetes care
The menstrual cycle is conspicuously absent from clinical guidelines, diabetes technology algorithms, and structured education programs — despite affecting half the type 1 diabetes population monthly for decades. Dr. Nobili calls this what it is: a gender gap in care. Women consistently report that menstrual cycle glucose management is harder than managing illness, yet intercurrent illness protocols are standard while menstrual cycle protocols don’t exist. Algorithms weren’t designed with female physiology in mind (likely because most engineers are male). This isn’t about blaming individuals — it’s about recognizing a systematic blind spot and demanding that research, technology, and clinical practice catch up.
2) 60% of women on MDI experience clinically significant deterioration
In Dr. Nobili’s observational study of 170 women, 60% of those using multiple daily injections experienced a drop in time in range of 5% or more between the early follicular phase (when bleeding starts) and the late luteal phase (days before the period). This deterioration happens monthly, is predictable, and compounds over time. Women on MDI face the double burden of manually adjusting basal insulin, carb ratios, and correction factors simultaneously while navigating intense carb cravings and emotional changes — without structured guidance on how to do it.
3) AID systems cut deterioration in half — but aren’t perfect
Hybrid closed-loop systems reduce the percentage of women experiencing clinically significant glucose deterioration from 60% (on MDI) to 30-40% — even though these algorithms weren’t designed to account for menstrual cycles. This is a massive quality-of-life improvement. However, 30-40% of women on AID still experience meaningful deterioration, and many report frustrating hypoglycemia in the early follicular phase when insulin sensitivity spikes. The systems work better than MDI, but they’re not solving the problem completely — and they could be doing much more.
4) Hypoglycemia is the hidden burden
Dr. Nobili expected women to report the luteal phase (high glucose, insulin resistance) as the most burdensome. Instead, many women ranked the early follicular phase — when bleeding starts and insulin sensitivity surges — as equally or more burdensome. Hypoglycemia forces you to stop everything: you can’t ignore it, you can’t delay treating it, and it disrupts work, sleep, driving, and daily life. For women on AID systems that use recent total daily dose to set basal rates (780G, Omnipod 5, CamAPS FX), the algorithm “remembers” the high insulin needs from the luteal phase and delivers too much insulin when sensitivity returns — creating a hypo tsunami just as the period starts.
5) Control-IQ shows the most stability
Observational data from Dr. Nobili’s study suggests the Tandem t:slim Control-IQ system maintains relatively stable time in range across menstrual cycle phases, even without switching profiles. This likely reflects its design: Control-IQ doesn’t heavily weight recent total daily dose in its algorithm, so it doesn’t “learn” the luteal phase insulin resistance and carry it forward into the follicular phase. Women may still experience hyperglycemia during the luteal phase, but they avoid the sharp hypoglycemia rebound that other systems create. This is observation, not causation — but it’s a clinically relevant pattern worth discussing with patients.
6) 780G and Omnipod 5 may need target adjustments
Systems that prioritize the previous 3-5 days of total daily dose (Medtronic 780G, Omnipod 5) handle the luteal phase well but create hypoglycemia risk when bleeding starts. Practical solution: as soon as the period begins, raise the glucose target for 3-4 days. On the 780G, switch to a higher target range. On Omnipod 5, increase the target from 6.1 mmol/L (110 mg/dL) to 7.8 mmol/L (140 mg/dL). This gives the algorithm time to “forget” the high insulin doses and recalibrate to the new insulin sensitivity without causing hypos. Alternatively, lengthen the active insulin time slightly to reduce correction aggression.
7) CamAPS FX boost function is ideal for luteal phase
The CamAPS FX “boost” function increases insulin delivery by 30% without the algorithm learning from it — perfect for temporary insulin resistance. Turn boost on during the mid and late luteal phases (days 16-28) to handle rising progesterone-driven insulin needs. Turn boost off when bleeding starts. Because the algorithm doesn’t incorporate boost into its learning, you avoid the hypoglycemia rebound that happens with systems that “remember” high insulin doses. Dr. Nobili doesn’t yet have enough CamAPS users in her dataset to confirm this observationally, but the mechanism makes it a strong theoretical option.
8) MDI strategies: basal, carb ratios, or both
For women on injections, the luteal phase requires more insulin — but how you deliver it matters. Options include increasing basal insulin (Lantus, Tresiba, Levemir) by 10-20%, strengthening carb ratios (e.g., 1:10 → 1:8 or 1:7), or making correction factors more aggressive. Many women find that adjusting both basal and bolus works best. The key is anticipating the change in the mid luteal phase (day 16-18) rather than waiting until glucose is already high. When bleeding starts, reverse all adjustments immediately to avoid hypoglycemia.
9) Pre-bolusing matters more during luteal phase
Insulin resistance during the luteal phase means insulin acts more slowly relative to carbohydrate absorption. Pre-bolusing 15-20 minutes before meals (even on AID systems) becomes critical. Mixed meals — combining protein, fat, and vegetables with carbohydrates — slow absorption and better match insulin action, which peaks around 2 hours. This is particularly important when carb cravings are high and you’re reaching for pizza, chocolate, or ice cream. Walking 10-15 minutes after eating amplifies insulin action and smooths post-meal glucose peaks.
10) One bad day is not catastrophic
Women with type 1 diabetes often feel intense anxiety about glucose spikes during the luteal phase — worried that every high will harm their long-term health. Dr. Nobili’s message: one high day, or even three to five difficult days per month, is not catastrophic. What matters is time in range averaged over weeks and months, not perfection every day. Reframe CGM data as information, not judgment: “What happened? What’s the likely driver? What’s one tweak for next time?” This reduces the paralysis that comes from fear of imperfection and allows women to make proactive adjustments without guilt.
11) Track your cycle — know your pattern
Normal cycle length is 28-35 days. Cycles shorter or longer than this warrant discussion with a healthcare provider. Tracking cycle length, bleeding heaviness (how many tampons/pads per day), and glucose patterns across phases gives you the data to predict what’s coming and adjust proactively. Many smartphone apps track cycles and could theoretically link to AID algorithms — but this integration doesn’t exist yet. For now, manual tracking is essential. Know when ovulation typically happens (around day 14, but varies), when the luteal phase starts (day 16-18), and when to expect bleeding.
12) This should be in the algorithm
Dr. Nobili’s central argument: menstrual cycle tracking apps already exist on smartphones. Linking these apps to AID algorithms is technically trivial compared to the machine learning and fully closed-loop systems presented at diabetes conferences. Industries should factor menstrual cycles into algorithm design — anticipating insulin resistance in the luteal phase and recalibrating quickly when bleeding starts. This isn’t a “nice to have” feature; it’s addressing a predictable, recurring physiological reality for half the user base. Until this happens, the burden falls entirely on women to manually compensate for what the algorithm should be doing automatically.
Practical checklist: managing menstrual cycles with T1D
Tracking and preparation:
- Track cycle length (normal: 28-35 days) and note if bleeding is earlier or later than expected
- Monitor bleeding heaviness (tampons/pads per day) — abnormal patterns should be discussed with a doctor
- Review CGM data across 2-3 cycles to identify your personal glucose pattern (some women have minimal impact, others have dramatic swings)
- Mark key phases in your calendar: mid luteal (day 16-18), late luteal (day 22-28), early follicular (when bleeding starts)
Luteal phase (days 16-28) — managing insulin resistance:
- On MDI: Increase basal insulin by 10-20% starting day 16-18, strengthen carb ratios (e.g., 1:10 → 1:8), make correction factors more aggressive
- On 780G or Omnipod 5: Consider lowering glucose target or shortening active insulin time if not already at optimal settings
- On CamAPS FX: Turn on boost function (30% more insulin) during mid and late luteal phases
- On Control-IQ: May not need adjustments — monitor and adjust only if seeing consistent highs
- Pre-bolus 15-20 minutes before meals (even on AID systems)
- Choose mixed meals (protein + fat + veg + carbs) over high-GI foods when possible
- Walk 10-15 minutes after eating to amplify insulin action
- Keep moderate carbohydrate intake (30-40% of total) despite cravings
Early follicular phase (when bleeding starts) — preventing hypoglycemia:
- On MDI: Reduce basal insulin and relax carb ratios immediately back to baseline (or slightly below if breastfeeding)
- On 780G or Omnipod 5: Raise glucose target for 3-4 days to prevent hypos while algorithm recalibrates (e.g., target 7.8 mmol/L instead of 6.1 mmol/L)
- On CamAPS FX: Turn off boost function immediately when bleeding starts, consider using ease-off function for 2-3 days
- On Control-IQ: May not need adjustments, but monitor closely for unexpected lows
- Keep fast-acting carbs readily available — hypo risk is highest in first 2-3 days of bleeding
- Reduce pre-bolus timing or bolus after meals if experiencing frequent hypos
General principles:
- Act proactively in mid luteal phase (day 16-18) rather than waiting for highs to appear
- Reverse all adjustments immediately when bleeding starts to avoid hypoglycemia
- Accept that 3-5 difficult days per month is normal — one bad day is not catastrophic
- Reframe CGM as information, not judgment: “What happened? What’s one tweak for next time?”
- Discuss patterns with your diabetes team — menstrual cycles should be part of routine clinic conversations
Guest
Dr. Cecilia Nobili is a pediatric diabetology resident in Turin, Italy, and a physician-researcher living with type 1 diabetes. Diagnosed at age 25 during the COVID-19 lockdown, she transformed her personal experience with trial-and-error diabetes management into clinical and research expertise. Dr. Nobili leads a multi-center observational study examining how menstrual cycles impact glucose control across different insulin delivery systems, funded by a Breakthrough T1D research grant. Her work addresses a critical gender gap in diabetes care by bridging lived experience, patient-reported outcomes, and clinical evidence. She is a graduate of the Spare Science School and an advocate for integrating menstrual cycle management into diabetes technology algorithms and clinical guidelines.
Connect with Dr. Nobili:
- LinkedIn: linkedin.com/in/cecilia-n-383228107
Disclaimer
The content available in The Glucose Never Lies® guides is for informational purposes only. Reading or listening to the content does not constitute medical advice and is not a substitute for individualized care, and does not create a clinician–patient or therapeutic relationship with The Glucose Never Lies® or any guest. Always discuss any changes to your diabetes management with your healthcare team. Menstrual cycle patterns vary significantly between individuals — work with your diabetes team to identify your personal patterns and develop safe, effective adjustment strategies.
Episodes on women’s health and diabetes
- 31 — Pregnancy with Type 1 Diabetes (Prof. Eleanor Scott)
