None of the studies below found harm, and all of them are observational rather than randomized. The main study of men already treated for prostate cancer followed men with type 2 diabetes on hormone therapy [1]. Against DPP-4 inhibitors, GLP-1 use was linked to lower all-cause mortality, HR 0.60. Heart events did not differ. That gap is more likely to reflect who got the drug than what it did.
Two questions sit under this one. The first is safety and survival in men who already have prostate cancer. The second is whether the drugs change the risk of getting it at all. They use different studies and different outcomes. The broader cancer-incidence picture is in whether GLP-1 drugs reduce cancer risk.
Men already on hormone therapy
Androgen deprivation therapy is the standard hormone treatment for advanced prostate cancer. It causes weight gain, insulin resistance and cardiovascular risk. Those are the problems GLP-1 drugs treat. That makes these men a sensible group to study.
One retrospective cohort matched them against two other diabetes drug classes [1]. There were 659 men per group against DPP-4 inhibitors. There were 1,009 per group against SGLT2 inhibitors. Most people starting these drugs already carry another diagnosis, as registry data on new users show.
Against DPP-4 inhibitors, all-cause mortality was HR 0.60, 95% CI 0.46 to 0.79. Major kidney events were 0.63. Thrombotic events were 0.53.
Selection is a known problem in cancer care. A man whose cancer is advancing may be less likely to start a new weekly injection. If so, users and non-users differ in expected survival before any drug acts. That alone can produce a mortality gap.
The SGLT2 comparison is the stricter test. SGLT2 inhibitors are an active comparator with their own heart and kidney benefits. Against them, mortality was 0.76, upper bound 0.99. Heart, kidney and clot events showed no difference. How the two classes compare generally is in GLP-1 against SGLT2.
The authors say the neutral findings need caution. Limited events may have reduced power. That is the right reading of groups this size.
A different outcome: getting prostate cancer
The incidence studies ask something else. They follow men without prostate cancer and count new diagnoses. They say nothing about survival after a diagnosis.
A 2025 meta-analysis pooled five studies [2]. GLP-1 use was linked to a 28% lower risk of prostate cancer, RR 0.72, 95% CI 0.610 to 0.832. Heterogeneity was moderate, I² 51%. The comparators were placebo or other diabetes drugs.
A Danish nationwide cohort is more cautious [4]. It compared 14,206 GLP-1 starters with 21,756 basal insulin starters. The main estimate was HR 0.91, 95% CI 0.73 to 1.14. That interval includes no effect. A per-protocol analysis gave 0.80, with an upper bound of 1.01.
A 2025 review of 77 publications calls these drugs promising for aggressive or advanced disease [3]. It summarizes prostate cancer biology and incretin pathways. It is not a treatment trial.
What this means for a man considering one
Nothing here shows harm. Nothing here shows a survival benefit either. The decision sits with the oncologist and the prescriber, who know the stage and the other drugs. Erectile function is a separate worry for many men, covered in whether GLP-1 drugs cause erectile dysfunction.
Observational mortality figures need the same caution everywhere. The same problem shows up in the evidence past age eighty, where no trial goes either.