Menopause hormone therapy linked to 17% fewer type 2 diabetes cases across 14 studies

Hormone therapy for menopause is back in the news. On October 5 the American College of Physicians issued a guideline telling doctors to start with estrogen in combination with progestogen for women who still have a uterus, or estrogen alone for women who do not, as the first treatment for hot flashes and night sweats. It is a strong recommendation based on high-certainty evidence. For women weighing that option, a natural next question is what else the hormones do to the body, including to blood sugar.

A review in Diabetologia has pooled the evidence on that question. It brought together 49 studies, 33 randomized trials and 16 cohort studies, involving 1,472,277 participants from 15 countries. Across 14 of the studies, women using hormone therapy were about 17% less likely to be diagnosed with type 2 diabetes than women who were not, and women who already had diabetes had slightly lower blood sugar, cholesterol and blood pressure.

The authors, led by Chen Liang with Gita Mishra as senior author and funded by Australia’s National Health and Medical Research Council, are cautious about what that adds up to. They rated the evidence on new diabetes as very low certainty, and concluded that “the effect size was generally too small to be clinically significant”.

What the hormone therapy review pooled

The review asked two separate questions. The first was whether hormone therapy changes the chance of developing type 2 diabetes in women who do not have it, along with their insulin and blood sugar. The second was whether it changes blood sugar, blood fats and blood pressure in women who already have diabetes. To answer both, the team searched six research databases up to July 2025 for trials and cohort studies that set hormone users against non-users. The women in them were past menopause or undergoing the menopausal transition, the years of changing periods that lead up to it.

For each pooled result, the team graded how far the evidence could be trusted, using a standard system that rates certainty from high to very low. A rating of very low means the true effect could be substantially different from the estimate. The studies varied widely in length, with trials lasting from 1 week to 13.2 years and cohort studies following women for six months to 17 years, and only seven of the 49 were judged to be at low risk of bias.

That matters most for the estimate of new-onset type 2 diabetes, which drew on 1,450,134 women, nearly all of the review’s total. Most of that number must come from observational cohorts, in which women chose whether to take hormones: even a large trial such as the Women’s Health Initiative’s estrogen plus progestin arm enrolled 15,641 women. Women who take hormone therapy may differ from those who do not in health, income and access to care, and those differences could explain part of a lower diabetes rate.

What the review found for women with and without diabetes

In women without diabetes, the most solid result was for insulin. Hormone therapy users had slightly lower fasting insulin, a result graded high certainty, and a slightly lower score on a standard measure of insulin resistance, the state in which the body needs more insulin to keep blood sugar in check. Insulin resistance is the main route to type 2 diabetes, so the direction is consistent with the lower rate of diabetes diagnoses.

On diabetes itself, hormone users had about 17% fewer new cases, with a plausible range from 8% to 25% fewer. Because it rests on very low-certainty evidence, the review reports only that hormone use went with a lower risk of new-onset diabetes, not that it prevents it.

In women who already had type 2 diabetes, hormone therapy went with lower fasting blood sugar and a lower HbA1c, a blood test that reflects average blood sugar over the previous two to three months. Across 17 studies with 15,720 women, HbA1c was about half a percentage point lower in hormone users. Total cholesterol, low-density lipoprotein cholesterol (the LDL or “bad” kind) and systolic blood pressure were also slightly lower. The review rated the cholesterol and blood pressure results low certainty, and the HbA1c and fasting blood sugar results very low.

Those changes are modest. Systolic blood pressure, for instance, was about 3.4 points lower on average. The authors call them small beneficial associations with glucose metabolism, and they found no sign that hormone therapy made blood sugar worse in women with diabetes.

The analyses in women with diabetes were also small. Apart from the HbA1c result, each rested on a few hundred to about 1,400 women: the cholesterol result on 907, and systolic blood pressure on 630 women in nine studies. Pooling many small studies can steady an estimate, but it cannot fix weaknesses the studies share, which is part of why the certainty ratings stayed low.

How the review compares with the big hormone trials

The idea that hormones might protect against diabetes is not new. It comes mainly from the Women’s Health Initiative, a large US hormone trial. In its estrogen plus progestin arm, 15,641 women aged 50 to 79 were randomly assigned to the hormones or a placebo. Over 5.6 years, diabetes needing treatment developed in 3.5% of the women on hormones, against 4.2% in the placebo group, about a fifth fewer. The gap barely changed when the researchers allowed for changes in body weight and waist size, and in the first year there was a fall in insulin resistance among the women taking hormones, which pointed to an effect on how the body handles sugar that did not depend on weight.

Large cohorts pointed the same way. In the French E3N study of 63,624 postmenopausal women, those who had ever used hormone therapy had an 18% lower risk of new-onset diabetes than never-users, and estrogen taken by mouth was associated with a larger reduction than estrogen through the skin. Taking account of changes in body weight during follow-up did not substantially modify this association, the E3N team reported, echoing the American trial. The route of estrogen is one of the differences between studies that later reviews have tried to account for.

A 2017 review in Endocrine Reviews, by researchers including JoAnn Manson, concluded from the trials that hormone therapy delays the onset of type 2 diabetes, and proposed reasons, including effects of estrogen on insulin secretion, insulin sensitivity and where the body stores fat. So the new pooled estimate agrees with what was already suspected. Its contribution is to gather trials and cohorts together, extend the question to women who already have diabetes, and grade how much confidence the combined evidence deserves, which turns out to be little for the diabetes outcome.

Why guidelines do not use hormone therapy to prevent diabetes

A small benefit for blood sugar sits alongside known harms. The evidence review behind the 2022 US Preventive Services Task Force recommendation put numbers on both, per 10,000 women. Women using estrogen plus progestin compared with placebo had 78 fewer cases of diabetes over 5.6 years. Participants using estrogen only, who were women without a uterus, had 134 fewer over 7.1 years.

The harms came in similar numbers. For every 10,000 women on estrogen plus progestin over 5.6 years, risks were increased for invasive breast cancer, with 51 more cases, along with 52 more strokes, 120 more blood clots and 260 more cases of gallbladder disease. Estrogen alone also went with more strokes, blood clots and gallbladder disease.

Weighing benefits against harms, the task force recommends against the use of combined estrogen and progestin, and of estrogen alone, to prevent chronic conditions in women after menopause. The 2017 review reached a similar conclusion for diabetes specifically, describing hormone therapy as “neither approved nor appropriate for the prevention of type 2 diabetes”. Its authors added that it should not be withheld from women at higher risk of diabetes who seek treatment for menopausal symptoms.

Treating symptoms is where guidance supports hormones. The North American Menopause Society’s 2022 position statement says that for women younger than 60 years or within 10 years of menopause without reasons to avoid it, the balance of benefit and risk is favorable for bothersome hot flashes and for preventing bone loss. For women who start later or after 60, it judges the balance less favorable, because of greater absolute risks of coronary heart disease, stroke, blood clots and dementia. It also notes that risks differ by type, dose, duration, route and timing of treatment. The new review’s authors note that the effects of hormone therapy may differ by the age at which women start, and call for more research on that point.

What the diabetes review cannot settle

The authors name three limitations themselves: “the predominantly single-reviewer screening and assessment, the high risk of bias of the included studies and restricted geographical representation”. That is, a single person usually decided which studies qualified, many of those studies had design flaws, and they came from a limited range of countries.

Beyond those, a review is only as good as the studies it pools. The trials, which can test cause, were smaller and often short, while the cohorts that supplied most of the women cannot rule out the differences between users and non-users described above. And hormone therapy is not one treatment: the pooled studies mixed different estrogens, doses and routes, with and without a progestogen, which may not act alike.

Do not stop or change a prescribed medicine without talking to your doctor.

Whether hormone therapy suits a particular woman, with or without diabetes, depends on her symptoms, age and other risks, and is assessed by her doctor.

The pooled evidence points to slightly better blood sugar control with hormone therapy and no sign that it makes blood sugar worse, though the evidence on diabetes itself is weak.