Common treatments for diabetes don’t appear to mask or reduce levels of a prostate cancer marker in the blood. Rather, the lower rates of prostate cancer detected in men with diabetes might be because they are less likely to undergo prostate biopsies, a large population-based study from Sweden suggests.
The study, “Association Between Antidiabetic Medications and Prostate-Specific Antigen Levels and Biopsy Results,” was published in JAMA Network Open.
It is well-documented that people with diabetes have a lower rate of prostate cancer diagnoses, but the reasons for this are unclear. Researchers debate whether this is because people with diabetes actually get cancer less often, or if they are just diagnosed less frequently.
One idea proposed is that medications used to treat diabetes might lower blood levels of prostate specific antigen (PSA). PSA is often tested as a marker for prostate cancer — elevated levels of this protein are considered a sign of possible cancer — prompting biopsies to confirm a diagnosis. Men with low PSA levels have little reason to undergo a prostate biopsy.
To test this idea, researchers analyzed data from the Stockholm PSA and Biopsy Register, which collects data on people undergoing PSA tests and prostate biopsies in Sweden.
They identified individuals who were treated with the diabetes medications metformin (4,583 men), sulfonylurea (1,104), or insulin (987 men). Collectively, these people were referred to as the “exposed” group, referring to exposure to a medication. They were then age-matched with people in the database who had no record of using these medications, and who served as an “unexposed” group.
On average, PSA levels were lower in the exposed than in the unexposed group before the first recorded prescription for a diabetes medication (1.2 vs. 1.6 ng/mL). That is, prior to treatment, people with diabetes tended to have lower PSA levels than those without diabetes.
The researchers acknowledged that, since only medication use was assessed, it’s possible that people in the unexposed group had diabetes that was being managed without medication, but they considered this unlikely to have a sizable impact on the results.
After the researchers accounted for initial PSA levels and other baseline (study start) differences between the groups, however, no difference in PSA levels between these two groups were evident.
“Our findings of no difference in PSA levels after initial prescriptions of antidiabetic drugs do not support the hypothesis that the use of common antidiabetic medications leads to lower PSA levels that may mask the presence of prostate cancer in men with diabetes,” the researchers concluded.
“Our results, which adjusted for differences in premedication PSA levels, lend support to lower PSA levels being associated with disease rather than with medication.”
Individuals taking metformin or sulfonylurea were significantly more likely to undergo PSA testing — a 6% and 7% greater likelihood, respectively — than those not using these medications. In contrast, those on insulin were 21% less likely to undergo PSA testing; overall, the exposed group was 7% less likely to take a PSA test than the exposed group.
Men prescribed metformin or insulin were also significantly less likely to undergo prostate biopsy after their PSA test came back positive — 24% and 33% less likely, respectively. There was no statistically significant difference in prostate biopsy rates among those prescribed sulfonylurea.
Overall, people in the exposed group were 41% less likely to undergo a prostate biopsy than those in the unexposed group.
Among the 53,357 men who underwent PSA testing and had abnormally high PSA levels (over 3 ng/mL), those prescribed one of these three diabetes drugs were significantly less likely to have a biopsy: 13% less likely among men on metformin, 12% among those using sulfonylurea, and 17% for men on insulin.
Of the 32,123 people who did have a prostate biopsy, 41% were found to have prostate cancer. In this group, no difference existed in rates of prostate cancer detection among exposed as compared to unexposed individuals.
“Our findings do indicate the potential for detection bias (eg, fewer biopsies undertaken when PSA is elevated), which may explain some of the lower risk in men with diabetes,” the researchers concluded.