Prostate Medications and Blood Sugar: A New Zealand Cohort Study
Does finasteride affect blood sugar control in men with type 2 diabetes? Using New Zealand's national pharmaceutical data, we compared long-term outcomes between finasteride and tamsulosin users.
💬 Discuss with AI
Click to copy a prompt for ChatGPT, Claude, or your preferred LLM
The Overlap Between Prostate and Metabolic Health
If you’re a man managing both an enlarged prostate and type 2 diabetes, you face a common scenario: which medication should you take? Both finasteride and tamsulosin are first-line treatments for prostate problems, but they work completely differently. Does it matter which one you choose for your blood sugar?
The question isn’t hypothetical. Benign prostatic hyperplasia (BPH) and type 2 diabetes mellitus (T2DM) frequently coexist - both conditions become more common with age. And research has shown real biological links between prostate enlargement and glucose metabolism:
- Sex steroid metabolism: Androgens influence both prostate growth and insulin sensitivity
- Inflammation: Chronic inflammation is implicated in both BPH progression and insulin resistance
- Insulin-like growth factors: IGF signaling affects both prostate cell proliferation and glucose homeostasis
Given this overlap, the clinical question is urgent: Do medications used to treat BPH affect glycemic control in patients who already have diabetes?
The Medications in Question
Finasteride (a 5-alpha reductase inhibitor)
- Blocks conversion of testosterone to DHT
- Shrinks prostate over time
- Potential metabolic concerns based on animal studies and small clinical trials
Tamsulosin (an alpha-blocker)
- Relaxes smooth muscle in prostate and bladder neck
- Provides rapid symptom relief
- Some evidence suggests possible effects on glucose metabolism
Both are first-line treatments for BPH, but they work through completely different mechanisms. Understanding their metabolic effects is crucial for the large number of men who have both conditions.
Existing Evidence: A Mixed Picture
Previous research on BPH medications and glucose metabolism has produced contradictory findings:
Studies suggesting 5-ARIs impair glucose control:
- Animal studies showing hepatic insulin resistance
- Small RCT showing decreased hepatic insulin sensitivity with dutasteride
- Case reports of metabolic changes
Studies suggesting alpha-blockers improve glucose control:
- Clinical trial showing doxazosin improved glycemic control in hypertensive diabetics
- Potential mechanisms involving improved insulin sensitivity
But there’s a critical gap: No one had directly asked the simple question - in men who already have diabetes and need BPH treatment, does it matter which medication they choose? That’s what we set out to answer.
Our Study
I collaborated with Dr. Maxim S. Petrov at the University of Auckland and Dr. Jaelim Cho at Yonsei University to address this gap using New Zealand’s national pharmaceutical dispensing and hospitalization databases.
Study Design
Population: Men with type 2 diabetes who were new users of either finasteride or tamsulosin
Outcome: Poor glycemic control, defined as:
- HbA1c ≥ 8.5% (69 mmol/mol), or
- Initiation of insulin therapy, or
- Hospitalization for hyperglycemia or diabetes complications
Design: Retrospective cohort study with:
- New-user design: No prior use of study medications
- Active comparator: Head-to-head comparison of the two drug classes
- Propensity score weighting: Balanced baseline characteristics
- Cox proportional hazards models: Time-to-event analysis
Key Methodological Features
1. Lag period: We excluded outcomes in the first 6 months to address:
- Protopathic bias - the risk that disease symptoms trigger treatment decisions. We addressed this by excluding the first 6 months of follow-up. Why? If someone’s blood sugar started worsening before they began finasteride, that wouldn’t be the drug’s fault. The lag period prevents us from mistaking existing disease for a drug effect.
- Treatment stabilization: Allowed time for medication effects to manifest
2. Multiple definitions: We tested various HbA1c thresholds and outcome definitions to ensure robustness
3. Sensitivity analyses: We examined different follow-up periods, subgroups, and model specifications
What We Found
Our study included 1,839 men with T2DM:
- 1,259 finasteride users
- 580 tamsulosin users
Incidence Rates
During follow-up (median 2.4 years for finasteride, 1.6 years for tamsulosin):
- Tamsulosin users: 564 events per 10,000 person-years
- Finasteride users: 409 events per 10,000 person-years
At first glance, this suggests better outcomes with finasteride. But this crude comparison doesn’t account for baseline differences between the groups.
Adjusted Analysis
After applying inverse probability of treatment weighting to balance the cohorts:
| Comparison | Weighted HR | 95% CI | Interpretation |
|---|---|---|---|
| Tamsulosin vs. Finasteride | 1.27 | 0.95–1.72 | No significant difference |
The hazard ratio of 1.27 suggests a possible 27% higher risk of poor glycemic control with tamsulosin compared to finasteride. But look at the confidence interval: 0.95-1.72. This range crosses 1.0, meaning the true risk could be lower or higher. In other words, we can’t tell if tamsulosin actually causes worse control, or if we’re seeing random variation.
Sensitivity Analyses
We conducted multiple sensitivity analyses:
1. Different HbA1c thresholds (8.0%, 9.0%, 9.5%): Results were consistent — no significant difference
2. Extended lag periods (9 months, 12 months): Results remained consistent
3. Subgroup analyses (by age, diabetes duration, baseline HbA1c): No significant interactions
4. Competing risks analysis: Accounting for mortality as a competing event didn’t change conclusions
Across all analyses, the consistent finding was no statistically significant difference in glycemic control between the two medications.
Interpretation
Clinical Meaning
What does it mean when a study finds no difference between two medications?
-
Therapeutic equivalence: For glycemic control purposes, finasteride and tamsulosin appear equivalent
-
Prescribing freedom: Clinicians can choose between these medications based on:
- Efficacy for urinary symptoms
- Side effect profiles
- Patient preferences
- Cost and availability
— without concern for differential effects on diabetes control
-
Patient reassurance: Men with diabetes can take either medication without worrying about adverse effects on their blood sugar management
Why the Null Finding Makes Sense
Several factors support the plausibility of equivalent metabolic effects:
-
Different mechanisms: While both medications affect BPH symptoms, they work through entirely different pathways (hormonal vs. smooth muscle relaxation)
-
Compensatory mechanisms: The body has robust homeostatic mechanisms for glucose regulation that may compensate for any minor drug effects
-
Prior diabetes diagnosis: These patients already had diagnosed and (presumably) managed diabetes, making them less susceptible to medication-induced metabolic changes
Comparison to Other Studies
Our findings differ from some smaller studies that suggested either medication class might affect glucose metabolism. Several factors explain this:
- Study population: We focused on patients with established T2DM, not patients at risk of developing diabetes
- Sample size: Our larger sample provided more statistical power to detect true differences (or confirm their absence)
- Real-world setting: Database studies capture routine clinical practice, not selected clinical trial populations
- Active comparator design: Instead of comparing to people taking no BPH medication, we compared finasteride users to tamsulosin users. Both groups have the same disease and both are visiting doctors regularly - so we isolate the effect of one medication versus another.
Strengths and Limitations
Strengths
- National data: New Zealand’s pharmaceutical dispensing data is comprehensive and accurately tracked
- Relevant population: We studied the exact population of clinical interest — men with both BPH and T2DM
- Clinical outcome: Poor glycemic control is a meaningful endpoint that affects quality of life and complication risk
- Rigorous methods: New-user, active-comparator design with careful confounding control
Limitations
- Observational data: Despite propensity score methods, unmeasured confounding is possible
- HbA1c measurement: We relied on laboratory test timing in the database; not all patients may have had regular testing
- Adherence: We assumed continuous use based on dispensing records; actual adherence may vary
- Generalizability: New Zealand’s population may differ from other countries in ways that affect generalizability
- Finasteride only: We didn’t have sufficient dutasteride users for separate analysis
Clinical Implications
For Patients with Both BPH and T2DM
If you have both conditions and are considering treatment:
- Either medication is reasonable — choose based on symptom relief and side effects
- Monitor your diabetes — continue regular HbA1c testing as recommended by your doctor
- Don’t switch medications for glycemic control reasons alone
- Discuss urinary symptoms with your doctor — both medications are effective, but individual responses vary
For Primary Care Physicians and Urologists
When managing patients with BPH and diabetes:
- Prescribe confidently — neither medication appears to worsen glycemic control
- Coordinate care — ensure diabetes management remains optimized regardless of BPH treatment
- Consider other factors:
- Sexual side effects (more common with 5-ARIs)
- Orthostatic hypotension (more common with alpha-blockers)
- Onset of action (faster with alpha-blockers)
- Need for prostate cancer screening (PSA monitoring with 5-ARIs)
For Endocrinologists
When reviewing diabetes management:
- BPH medications are unlikely culprits if glycemic control worsens
- Focus on other factors — diet, exercise, medication adherence, other medications
- Collaborate with urology when both conditions require attention
The Broader Context
This study contributes to our understanding of polypharmacy in older adults. As people age, they often accumulate multiple conditions requiring multiple medications. Understanding drug-disease and drug-drug interactions becomes increasingly important.
The reassuring finding from this study is that two commonly co-prescribed medication classes — 5-ARIs and alpha-blockers — don’t appear to interfere with diabetes management. This simplifies clinical decision-making and reduces concerns about therapeutic conflicts.
Future Research Directions
While this study provides important reassurance, several questions remain:
-
Longer-term effects: Our median follow-up was 2-3 years; what about decade-long use?
-
Dose-response relationships: Do higher doses or better adherence change the risk profile?
-
Dutasteride: We lacked power to analyze dutasteride separately; does the more potent 5-ARI have different effects?
-
Combination therapy: Many men take both a 5-ARI and an alpha-blocker; what are the metabolic effects of combination therapy?
-
Diabetes prevention: While this study examined glycemic control in patients with established diabetes, the question of whether these medications affect diabetes incidence in at-risk populations remains relevant
Conclusion
For men navigating both benign prostatic hyperplasia and type 2 diabetes, medication choices can feel overwhelming. This study provides clarity: finasteride and tamsulosin appear equivalent in their effects on glycemic control.
The null finding is good news. It means clinicians can focus on the factors that really matter — symptom relief, side effect profiles, and patient preferences — without worrying about differential metabolic effects.
In the complex landscape of managing multiple chronic conditions, this is one less thing to worry about. Your doctor can now choose the medication based purely on what works best for your urinary symptoms and which side effects you can tolerate - without second-guessing its impact on your blood sugar.
This post summarizes a study conducted with Dr. Maxim S. Petrov at the University of Auckland and Dr. Jaelim Cho at Yonsei University. The full paper is published in PubMed.