Menstrual Cycle Linked to Changes in Insulin Sensitivity and Glucose Control in Women With Type 1 Diabetes Using Automated Insulin Delivery

1-minute summary

A real-world observational study published in Diabetes Care found that insulin requirements, glucose levels, carbohydrate intake and insulin sensitivity varied systematically across menstrual-cycle phases in menstruating adults with type 1 diabetes (T1D) using automated insulin delivery (AID) systems.

The study analyzed data from 77 participants contributing 380 menstrual cycles. Overall, insulin sensitivity was greatest during the early follicular phase and declined during the luteal phase, reaching its lowest level around the midluteal phase.

Correspondingly, glycemic control worsened during the luteal phase. Time in range (TIR; glucose 70–180 mg/dL) declined from 78.9% during the early follicular phase to 75.7% during the late luteal phase. Time in tight range (70–140 mg/dL) showed a similar decline.

At the population level, modeled insulin sensitivity was approximately 2.6% above the population mean during the early follicular phase and approximately 2.6% below the mean during the midluteal phase.

However, there was considerable variation between individuals. Approximately 84.7% of participants demonstrated the same overall direction of change—higher insulin sensitivity during the follicular phase followed by lower sensitivity during the midluteal phase—while others showed different patterns.

The researchers suggest that future AID algorithms may need to account for menstrual-cycle-related physiological changes rather than relying exclusively on fixed learning periods.

Why this matters

Menstrual-cycle hormones can influence insulin sensitivity and glucose regulation. For people with T1D, even relatively modest changes in insulin sensitivity can affect insulin requirements and glucose control.

Modern AID systems continuously adjust insulin delivery based on glucose measurements, but the algorithms may not specifically incorporate predictable hormonal changes associated with the menstrual cycle.

The study suggests that menstrual-cycle phase may be an additional physiological variable that could improve individualized insulin delivery.

Importantly, the researchers also found substantial differences between individuals. Therefore, a single standardized insulin adjustment for every person at a particular menstrual-cycle phase is unlikely to work universally.

Study at a glance

The participants contributed a total of approximately 10,107 days of data, with a mean of 4.9 menstrual cycles per participant.

Participant characteristics and baseline diabetes-management data

The study included menstruating adults with T1D who reported regular menstrual cycles and were using AID systems.

Both commercial and open-source AID systems were represented among participants with available device information.

What the researchers found

1. Glycemic control varied across the menstrual cycle

The investigators found a gradual change in glucose control across menstrual-cycle phases.

Time in range was highest during the early follicular phase and progressively decreased toward the luteal phase.

Time in range by menstrual-cycle phase

The overall phase-related difference was statistically significant (P = .004).

Thus, compared with the early follicular phase, glycemic control was less favorable during the luteal phase.

2. Time in tight range also decreased

A similar pattern was observed when the researchers used the narrower glucose target of 70–140 mg/dL.

Time in tight range

The variation across the cycle was statistically significant (P < .001).

The decline in tight-range control was therefore more pronounced than the change in the broader 70–180 mg/dL target.

3. More insulin was required during the luteal phase

The researchers observed that total daily insulin requirements increased during the luteal phase.

Carbohydrate intake also increased during this part of the cycle.

The combination of:

• greater carbohydrate intake,
• higher insulin requirements,
• higher mean glucose levels, and
• lower TIR

suggests that the additional insulin delivered during the luteal phase did not completely compensate for the accompanying reduction in insulin sensitivity.

4. Insulin sensitivity changed across the menstrual cycle

One of the most important findings was the modeled change in insulin sensitivity.

Using a hierarchical Bayesian state-space model, investigators estimated a latent insulin-sensitivity parameter from CGM and insulin-delivery data.

Population-level insulin sensitivity

The credible interval for the early follicular estimate was 0.3% to 5.0%, while the midluteal estimate had a credible interval of −5.2% to −0.1%.

This represents an approximately 5.2-percentage-point difference in modeled insulin sensitivity between the early follicular and midluteal phases at the population level.

The important point is not that every participant experienced exactly a 5.2% change, but that the population-level model identified a consistent shift toward lower insulin sensitivity during the luteal phase.

5. Most participants followed the overall pattern—but not everyone

Individual responses varied substantially.

Approximately 84.7% of participants showed a pattern consistent with the overall population finding:

Higher insulin sensitivity during the early follicular/periovulatory period → lower insulin sensitivity during the midluteal phase.

However, the remaining participants did not demonstrate the same pattern.

This individual variability is clinically important because it means that automatically increasing insulin doses for every person during the luteal phase could lead to inappropriate insulin delivery in some individuals.

6. Changes in time above and below range

The study also found phase-related changes in glucose excursions.

During the early follicular phase, time below range was higher relative to the individual’s cycle mean. This subsequently declined during the luteal phase.

Conversely, time above range increased during the luteal phase.

In the supplied study report:

This pattern is consistent with the overall finding of greater insulin sensitivity earlier in the cycle and reduced sensitivity later.

7. The effect was not explained simply by contraception or PCOS

Subgroup analyses examined participants according to hormonal contraception use and PCOS or other cycle-related conditions.

The general pattern of:

higher insulin sensitivity in the early follicular phase → lower sensitivity around the midluteal phase

remained present, although estimates were less precise in these smaller subgroups.

This suggests that inclusion of participants using hormonal contraception did not substantially account for the main population-level finding.

Researchers’ perspective

Martina Rothenbühler, PhD, emphasized that insulin requirements, glucose levels and carbohydrate intake differed according to menstrual-cycle phase.

She explained:

“There is substantial individual variability; changes in insulin sensitivity don’t look the same for everyone.”

She also highlighted a potential limitation of current automated insulin-delivery algorithms:

“Rigid learning windows of automated insulin delivery algorithms might not make sense with regard to fast or important changes in hormones.”

The researchers therefore suggest that future AID systems may need to recognize menstrual-cycle-related physiological changes when estimating insulin requirements.

Clinical significance

Menstrual cycle may be an important variable in T1D management

For menstruating individuals with T1D, changes in insulin sensitivity across the menstrual cycle may contribute to fluctuations in:

• insulin requirements,
• mean glucose,
• time in range,
• time in tight range,
• time above range, and
• carbohydrate requirements.

This provides a physiological explanation for why some individuals notice predictable changes in glucose control at particular points in their menstrual cycle.

AID systems may need greater personalization

Current AID systems are designed to continuously adjust insulin delivery in response to glucose patterns.

However, the study suggests that the underlying insulin sensitivity of an individual may change according to hormonal phase.

A future algorithm could potentially incorporate:

CGM data + insulin requirements + menstrual-cycle phase + individual historical response

rather than relying exclusively on recent glucose data.

However, such an approach would need to be validated prospectively before being incorporated into routine clinical practice.

Why individual variability is important

The study does not show that every woman with T1D should increase insulin during the luteal phase.

Instead, the findings indicate that:

Many—but not all—individuals experience reduced insulin sensitivity during the luteal phase.

Therefore, individualized pattern recognition is likely to be more appropriate than applying a universal menstrual-cycle insulin adjustment.

Any insulin-dose modification should be individualized and undertaken with appropriate diabetes-management guidance.

Limitations

1. Observational study

This was an observational study using real-world data rather than a randomized intervention trial.

Therefore, it can demonstrate associations between menstrual-cycle phase and glucose/insulin patterns but cannot establish causation.

2. Self-reported menstrual-cycle data

Participants self-tracked their menstrual cycles. This introduces the possibility of inaccuracies in determining cycle phase.

3. Self-selected population

Participants voluntarily contributed their data, which may introduce selection bias.

Individuals who were particularly interested in menstrual-cycle-related glucose changes may have been more likely to participate.

4. Heterogeneity of AID systems

Participants used different AID systems, including commercial and open-source systems. Algorithm characteristics may therefore have differed between individuals.

5. Individual variability

Although a large majority showed the overall population pattern, approximately 15% did not demonstrate the same direction of change.

This limits the applicability of a universal menstrual-cycle-based insulin adjustment.

6. Limited representation

The study focused on menstruating individuals with T1D and regular cycles. The findings may not directly apply to people with irregular cycles, menopause, pregnancy or other endocrine conditions.

7. Observational estimates of insulin sensitivity

The insulin-sensitivity measure was model-derived, rather than directly measured using a gold-standard metabolic clamp technique. The findings should therefore be interpreted as modeled physiological estimates.

What doctors should know

• Menstrual-cycle phase can be associated with meaningful changes in glucose control and insulin requirements in menstruating individuals with T1D.
• In this study, TIR declined from the early follicular phase toward the luteal phase.
• Time in tight range showed a similar but somewhat larger deterioration.
• Total daily insulin requirements and carbohydrate intake increased during the luteal phase.
• Population-level modeled insulin sensitivity was approximately 2.6% higher than average during the early follicular phase and approximately 2.6% lower during the midluteal phase.
• Approximately 84.7% of participants showed the overall population pattern.
• Substantial individual variability remains, meaning that a uniform insulin adjustment based solely on menstrual-cycle phase is unlikely to be appropriate.
• Clinicians should consider discussing menstrual-cycle-related glucose patterns with menstruating patients with T1D.
• AID manufacturers may need to consider hormonal-cycle information when developing future adaptive algorithms.
• The study does not establish a specific insulin dose adjustment that should be applied during any particular menstrual phase.

Practical clinical interpretation

A useful way of conceptualizing the findings is:

Early follicular phase

↓
Higher insulin sensitivity
↓
Lower insulin requirements
↓
Better TIR

Transition toward luteal phase

↓
Reduced insulin sensitivity
↓
Higher insulin requirements
↓
Higher mean glucose
↓
Lower TIR / higher time above range

This pattern was observed at the population level but was not universal among individuals.

Bottom line

A real-world observational study published in Diabetes Care found that menstrual-cycle phase was associated with systematic changes in insulin sensitivity, insulin requirements and glycemic control among menstruating adults with type 1 diabetes using automated insulin delivery systems.

Across 380 menstrual cycles contributed by 77 participants, insulin sensitivity was highest during the early follicular phase and lowest around the midluteal phase. TIR and tight-range glucose control also declined during the luteal phase, while insulin requirements and carbohydrate intake increased.

The findings highlight a potentially important limitation of current AID systems: hormonal changes can alter insulin requirements in ways that may not be fully captured by algorithms trained primarily on recent glucose and insulin-delivery patterns.

However, the substantial variability between individuals means that these findings should support personalized monitoring rather than routine, universal insulin-dose changes based solely on menstrual-cycle phase.

Original research / DOI

Hossmann S, Friedman M, Lizoain A, et al. Effects of Menstrual Cycle on Insulin Sensitivity in Type 1 Diabetes: An Observational Study of Individuals Using an Automated Insulin Delivery System. Diabetes Care. 2026;49(9):1673-1680.

DOI: 10.2337/dc26-0692
PMID: 42482329
PMCID: PMC13493352.

The article was published in the September 2026 issue of Diabetes Care.

Sugar, Red Meat, and MASLD: Reframing Dietary Risk Beyond Fat

Key Points

• Excess fructose—particularly from sugar-sweetened beverages, high-fructose fruit juices, and foods containing high-fructose corn syrup—can promote hepatic de novo lipogenesis and contribute to metabolic dysfunction-associated steatotic liver disease (MASLD).
• Red and processed meat consumption has also been associated with metabolic and liver-related risks, although the strength of the association may vary according to sex, quantity consumed, processing, and preparation.
• Whole fruits should not be equated with fructose-containing beverages because their fiber-rich matrix alters fructose absorption and metabolic handling.
• A Mediterranean-style dietary pattern emphasizing vegetables, whole fruits, legumes, nuts, fish, fiber and minimally processed foods has demonstrated benefits in hepatic fat reduction and insulin sensitivity.
• Dietary prevention of MASLD should focus less on eliminating dietary fat indiscriminately and more on reducing liquid sources of fructose, limiting red and processed meat, and improving overall dietary quality.

Reframing the Dietary Risk of Fatty Liver Disease

When fatty liver disease is discussed, dietary fat is often assumed to be the primary nutritional culprit. However, the evidence summarized in the source article points toward a more nuanced interpretation.

For metabolic dysfunction-associated steatotic liver disease, excessive intake of fructose—particularly from sugar-sweetened beverages—and high consumption of red and processed meat represent important modifiable dietary exposures.

This distinction has substantial clinical relevance because MASLD is estimated to affect approximately 30% of adults worldwide. The disease can progress silently for years, making prevention and modification of metabolic risk factors particularly important.

Major dietary factors associated with MASLD

Fructose and Hepatic Fat Production

The concern surrounding fructose extends beyond its caloric content. Fructose has metabolic characteristics that can directly influence hepatic fat production.

After absorption, dietary fructose is transported through the hepatic portal circulation to the liver, where it can stimulate de novo lipogenesis, the biochemical process through which the liver converts carbohydrate substrates into fatty acids.

The article explains that this process involves activation of carbohydrate-responsive element-binding protein (ChREBP) and occurs through pathways that are less dependent on insulin signaling than those governing glucose metabolism. Consequently, fructose can exert a relatively strong stimulatory effect on hepatic fat synthesis.

Fructose metabolism may also increase uric acid production and mitochondrial oxidative stress. These effects can interfere with fatty-acid oxidation. When fatty-acid synthesis and delivery exceed the liver’s capacity to oxidize or export fat, triglycerides can accumulate within hepatocytes.

The resulting metabolic environment may include:

Fructose excess → increased hepatic lipogenesis → triglyceride accumulation → insulin resistance/hyperinsulinemia → hepatic inflammation → progression of MASLD.

Evidence from a controlled feeding study

The article cites a controlled feeding study involving adolescents with nonalcoholic fatty liver disease. Reduction in dietary sugar was associated with a substantial decline in hepatic de novo lipogenesis.

Thus, reducing added sugar may influence not only dietary calorie intake but also the biochemical pathways responsible for hepatic fat production.

Sugar-Sweetened Beverages Deserve Particular Attention

Among the modifiable dietary exposures, sugar-sweetened beverages deserve particular consideration.

This category includes conventional soft drinks as well as some fruit juices containing substantial quantities of fructose. The concern also extends to products sweetened with high-fructose corn syrup (HFCS), an ingredient used extensively in numerous ultra-processed foods and beverages, particularly in the United States.

For patients attempting to reduce fructose exposure, simply avoiding soda may therefore be insufficient. Ingredient-label awareness can help identify HFCS and other added sugars in products that may not immediately appear to be high-sugar foods.

Examples include:

• Sweetened snacks
• Desserts
• Breakfast cereals
• Sauces
• Packaged foods
• Sweetened beverages
• Certain processed fruit products

The practical message is therefore to examine both beverages and packaged foods rather than focusing exclusively on table sugar or soft drinks.

Whole Fruit Is Different From Fruit Juice

An important nutritional distinction is that whole fruit should not automatically be grouped together with fructose-containing beverages.

Whole fruits such as apples, mangoes and grapes naturally contain fructose, but they also contain fiber and a complex food matrix that slows digestion and absorption.

According to the source article, these fruits have not been associated with hepatic steatosis or incident MASLD in the cited evidence.

Whole fruit versus fructose-containing beverages

Therefore, dietary recommendations should emphasize whole fruit rather than fruit juice, rather than advising patients to avoid fruit altogether.

The Gut–Liver Axis

The relationship between diet and MASLD also involves the gut–liver axis.

Changes in the intestinal microbiome, or gut dysbiosis, may impair intestinal barrier function and increase intestinal permeability. This can facilitate movement of bacterial products, including lipopolysaccharide, from the gut into the portal circulation.

Once these signals reach the liver, they can activate inflammatory pathways involving Toll-like receptors and nuclear factor-kappa B (NF-κB).

These pathways may contribute to:

• Hepatic insulin resistance
• Steatosis
• Inflammation
• Oxidative stress
• Progression toward metabolic dysfunction-associated steatohepatitis

Gut dysbiosis can also influence several metabolically relevant compounds and pathways, including short-chain fatty acids, bile acids, endogenous ethanol, choline metabolism and trimethylamine-N-oxide (TMAO). These metabolites may affect lipid metabolism, oxidative stress and inflammatory signaling.

This provides another rationale for dietary patterns rich in fiber, prebiotic substrates and minimally processed plant foods.

Red and Processed Meat as Additional Dietary Risk Factors

The article identifies red and processed meat as another important dietary consideration in MASLD.

A 2018 cross-sectional analysis published in the Journal of Hepatology reported that high total meat consumption was independently associated with NAFLD and insulin resistance.

Findings reported from the 2018 analysis

However, red meat is nutritionally complex. It can provide important nutrients including:

• Protein
• Iron
• Zinc
• Vitamin B12

At the same time, it can contribute substantial quantities of saturated fatty acids, cholesterol, heme iron and advanced glycation end products. These compounds and dietary components may contribute to oxidative stress, insulin resistance and metabolic dysfunction.

Processed meats may introduce additional concerns because of their sodium, nitrate and nitrite content, potentially extending their adverse metabolic implications beyond liver disease to conditions such as type 2 diabetes and cardiovascular disease.

The Association Is Not Necessarily Linear

Importantly, the evidence does not demonstrate an identical relationship between red-meat intake and MASLD across all populations.

A 2026 analysis from southern Italy, published in Nutrients, reportedly found no significant overall association between red-meat consumption and the outcome examined. An increased risk was observed specifically among men consuming approximately 75–90 g of red meat per day.

This finding suggests that factors such as:

• Sex
• Quantity consumed
• Food processing
• Cooking or preparation methods
• Overall dietary pattern

may modify the relationship between red meat and metabolic liver disease.

Consequently, the practical dietary recommendation presented in the article is not that all red meat must be eliminated, but that red meat—particularly processed meat—should be consumed in moderation.

The Mediterranean Diet

Against this background, the Mediterranean diet emerges as a dietary pattern with substantial evidence supporting its use in MASLD.

Rather than focusing on the elimination of a single macronutrient, this dietary pattern emphasizes overall food quality.

Core components of the Mediterranean dietary pattern

The pattern combines plant-based foods, fish, legumes, nuts and healthy unsaturated fats while reducing refined sugars and red and processed meat.

Evidence for Reduction in Hepatic Fat

The source article cites a randomized crossover trial published in the Journal of Hepatology in 2013.

In that study, adherence to a Mediterranean diet was associated with a substantially greater reduction in intrahepatic lipid content than a low-fat diet, and the effect occurred independently of weight loss.

Randomized crossover trial

The article also cites meta-analyses reporting improvements in the fatty liver index and HOMA-IR, while longitudinal data from the PREDIMED study suggest that stronger adherence to the Mediterranean dietary pattern may be associated with delayed progression of MASLD over five years.

Why Might the Mediterranean Diet Help?

Several components of the Mediterranean diet may act synergistically.

Monounsaturated fatty acids and omega-3 fatty acids may improve lipid metabolism and insulin sensitivity.

Polyphenols provide antioxidant and potentially anti-inflammatory effects.

Dietary fiber supports intestinal microbial diversity and may improve the gut–liver axis.

Together, these dietary components may:

1. Reduce hepatic oxidative stress.
2. Improve insulin sensitivity.
3. Influence lipid metabolism.
4. Reduce excessive hepatic fat accumulation.
5. Support a healthier gut microbiome.
6. Reduce inflammatory signaling.

The source article states that the Mediterranean dietary pattern is supported as an important component of MASLD care by several major professional organizations, including the American Gastroenterological Association (AGA), American Association for the Study of Liver Diseases (AASLD), American College of Gastroenterology (ACG), American Heart Association (AHA), American Diabetes Association (ADA), and the joint EASL-EASD-EASO guideline.

A Shift From Fat Avoidance to Dietary Quality

The central message is not that dietary fat is irrelevant, but that an exclusive focus on fat may obscure other important nutritional drivers of MASLD.

Instead, dietary prevention can be framed around three practical priorities:

1. Reduce liquid sources of fructose

Patients should minimize:

• Sugar-sweetened soft drinks
• High-fructose fruit juices
• Other beverages containing substantial added sugars

2. Limit red and processed meat

Particular attention should be paid to processed meat products, while overall red-meat intake should be moderated.

3. Adopt a Mediterranean-style dietary pattern

The dietary foundation should consist primarily of:

• Whole fruits
• Vegetables
• Legumes
• Nuts
• Fish
• Fiber-rich foods
• Minimally processed foods
• Healthier unsaturated fats

Clinical Takeaway

MASLD is common and frequently asymptomatic, making lifestyle-based prevention particularly important.

The evidence presented in the source article suggests that dietary counseling should move beyond the traditional message of simply avoiding fatty foods. A more comprehensive strategy would focus on reducing excessive fructose exposure, particularly from beverages; moderating red and processed meat; and increasing consumption of whole, fiber-rich and minimally processed foods.

Whole fruit should not be unnecessarily restricted merely because it contains naturally occurring fructose. The distinction between whole fruit and fiber-free fructose-containing beverages is particularly important when providing practical dietary advice.

A Mediterranean-style eating pattern provides a broader framework that incorporates these principles while emphasizing foods associated with improved metabolic health.

Three practical dietary priorities for MASLD

For a disease that is both increasingly prevalent and frequently silent, these dietary interventions represent relatively accessible measures that can potentially be incorporated into routine preventive care.

Expert Information

Robert Glatter, MD, FACEP, FAAEM, is identified in the source as an attending physician in the Department of Emergency Medicine at Lenox Hill Hospital and an assistant professor of emergency medicine at the Donald and Barbara Zucker School of Medicine at Hofstra/Northwell.

The source reports no relevant financial disclosures for Dr. Glatter.