Testosterone and Diabetes: What Therapy Does to Blood Sugar

The findings often arrive in pairs. Your GP sends you for follow-up because of raised glucose, and because you mention fatigue and no drive, testosterone gets measured too. Both values come back abnormal, and suddenly there are two diagnoses on the table that appear to have nothing to do with each other.

They have a great deal to do with each other. The link runs in both directions, and since an Australian trial in more than a thousand men, we know it can be influenced in both directions too.

Why the two are connected

Men who are overweight frequently have low testosterone concentrations, and low concentrations are associated with an increased risk of type 2 diabetes [1]. Several mechanisms sit behind this, and they reinforce one another.

Fat tissue, especially around the abdomen, is the main site of aromatase. This enzyme converts testosterone into estradiol, which lowers the testosterone level. At the same time, abdominal fat promotes insulin resistance. And less testosterone means less muscle mass, which means less tissue taking up glucose.

That produces a loop: waist up, testosterone down, insulin resistance up, muscle mass down, waist further up. Tackling that loop at one point only means working against the other points.

The trial that reversed the direction

For a long time it was unclear whether testosterone changes anything about glucose, or whether the low level is merely a consequence of the excess weight. The T4DM trial tested this properly [1].

It enrolled 1,007 men aged 50 to 74 with a waist circumference of at least 95 centimetres, a testosterone level up to 14.0 nmol/L, and either impaired glucose tolerance or newly diagnosed type 2 diabetes. Every participant was enrolled in a lifestyle programme. On top of that they received either testosterone undecanoate or placebo over two years.

The result after two years:

In other words: of a hundred men who would have progressed into diabetes despite the lifestyle programme, roughly forty were spared. Notably, the effect was independent of baseline testosterone [1].

Two limitations belong firmly alongside this. First, these men did not have pathological hypogonadism, only values in the lower range. This was prevention, not treatment of a deficiency. Second, the trial ended after two years, and whether the effect persists beyond that is unknown.

The price in this trial was hematocrit

One result of the trial is rarely quoted, although it is the most important one in practice. The safety threshold of a hematocrit above 54 percent was reached by 1 percent of men in the placebo group and by 22 percent in the testosterone group [1].

That is more than one in five, and considerably more than routine practice with gels suggests. The reason almost certainly lies in the preparation: 1,000 milligrams of testosterone undecanoate as a depot injection every three months produces high peaks, and peaks drive red cell production. Why that happens and what to do about it is covered in hematocrit and TRT.

The conclusion is not that therapy is dangerous, but that the delivery form and the monitoring interval matter especially in this group.

What is measurable in existing diabetes with a deficiency

T4DM was about prevention. For men who already have both, diabetes and a documented deficiency, there is a meta-analysis of five randomised trials with 351 participants and a mean follow-up of 6.5 months [2]:

• HbA1c: −0.87 percentage points
• Fasting glucose: −1.10 mmol/L
• Fasting insulin: −2.73
• Triglycerides: −0.35 mmol/L

HbA1c is the most meaningful of these, and a fall of 0.87 percentage points is in the range you would expect from adding a diabetes drug. Blood pressure and the other lipids did not change significantly.

The authors stress the narrow basis themselves: five trials, 351 men, six months. That is an indication, not a certainty, and it does not replace diabetes treatment.

The measurement trap in diabetes and obesity

This is where assessment often goes wrong. Excess weight and type 2 diabetes lower SHBG, the transport protein that binds most of the testosterone in your blood. When SHBG falls, total testosterone falls automatically, while the free, actually active testosterone can sit in the normal range.

That is why free testosterone, or at least SHBG, belongs in the panel for these patients rather than drawing conclusions from the total alone [4]. How the two relate is described in free vs. total testosterone.

The reverse holds equally: a normal total value can hide a real deficiency when SHBG is high. Without that second number, assessment in diabetes is incomplete.

GLP-1 agonists change the picture

Many men with type 2 diabetes now take a GLP-1 agonist such as semaglutide, whether for glucose or for weight. That is relevant to the hormone question.

A 2025 meta-analysis pooled four studies covering 219 patients. On GLP-1 agonists, bioavailable testosterone rose and HbA1c fell; results for free testosterone and SHBG remained inconclusive. The authors state explicitly that this association is not yet reflected in any clinical guideline and that larger studies are needed [3].

In practice: if you are starting a GLP-1 agonist and losing substantial weight, your testosterone level may improve on its own. Starting hormone therapy at that moment can mask an improvement that would have happened anyway. The order is worth getting right. What weight loss realistically achieves is covered in increasing testosterone naturally.

Safety in a high-risk group

Men with type 2 diabetes almost always belong to the cardiovascular risk group, and that is precisely the group studied in the largest safety trial of testosterone therapy. TRAVERSE enrolled 5,246 men with existing cardiovascular disease or elevated risk. On the primary endpoint of cardiovascular death, heart attack and stroke, testosterone was not inferior to placebo, 7.0 against 7.3 percent. More frequent were atrial fibrillation, acute kidney injury and pulmonary embolism [5].

That is the most robust foundation available, and it supports the therapy. It is also the reason why, in men with diabetes, the blood count and the heart rhythm stay in view alongside the glucose.

What this means in practice

A sensible order for men with both:

1. Weight and movement first. In T4DM every participant was in the lifestyle programme; testosterone worked in addition, not instead [1].
2. Treat the diabetes to guideline standard. Hormone therapy is not a diabetes drug, even though it lowers HbA1c [2].
3. Measure correctly. Two morning values, plus free testosterone or SHBG [4].
4. Treat a documented deficiency with symptoms, using more frequent smaller doses rather than three-monthly depots.
5. Monitor hematocrit closely, particularly in this group [1].

One frequent companion deserves its own line: erectile problems are common in diabetes, and the effect of hormone therapy on them is smaller in men with diabetes and obesity than in men without [6]. That is not an argument against therapy but one for treating the metabolic situation alongside it. The distinction is covered in erectile dysfunction or low testosterone.

If you are unsure whether your symptoms fit a hormonal cause, the online self-test gives a first indication. Interpreting glucose and hormone values together belongs in a conversation, and that can be arranged by telemedicine.

FAQ

Does testosterone lower blood sugar?

In men with type 2 diabetes and a documented deficiency, yes, measurably. A meta-analysis of five randomised trials found HbA1c down by 0.87 percentage points and fasting glucose down by 1.10 mmol/L [2]. The evidence base is narrow, though, at 351 men and six months of follow-up.

Can testosterone prevent diabetes?

In a two-year trial with 1,007 men it clearly reduced the risk: 12 against 21 percent on glucose tolerance testing, on top of a lifestyle programme [1]. The words "on top of" matter: the comparison group had the same programme. Whether the effect persists beyond two years is unresolved.

My total testosterone is low and my doctor says that is normal in diabetes. Is it?

Partly. Excess weight and diabetes lower SHBG, which lowers total testosterone while free testosterone can be normal. That is why free testosterone or SHBG belongs in the panel [4]. A low total alone neither proves a deficiency in your case nor does a normal one rule it out.

I take semaglutide. Should I wait before starting hormone therapy?

Often that is worth doing. On GLP-1 agonists, bioavailable testosterone rose and HbA1c fell across four pooled studies [3]. If you lose substantial weight, the value may improve on its own. A second measurement after some months of weight loss is the better basis for deciding.

Is therapy dangerous in diabetes?

Not cardiovascularly: in the largest trial, covering 5,246 men at cardiovascular risk, testosterone was not inferior to placebo on the primary endpoint [5]. What needs attention is hematocrit, which reached the safety threshold in 22 percent of treated men in the diabetes prevention trial, against 1 percent on placebo [1].

Further Reading

• Hematocrit and TRT: when the number matters
• Free vs. total testosterone: what counts?
• Increasing testosterone naturally: what actually works