Insulin use is not significantly predictive for prostate cancer mortality in diabetic patients: a 12-year follow-up study

OBJECTIVE To evaluate whether insulin use in diabetic patients could be predictive for prostate cancer mortality. PATIENTS AND METHODS A total of 39 135 diabetic men aged >= 40 years from a nationally representative cohort were followed prospectively from 1995 to 2006 for prostate cancer mortality. Cox proportional hazards models were used to estimate the hazard ratios for the following independent variables: age, diabetes type, diabetes duration, body mass index, smoking, insulin use and area of residence. The models were created for patients aged >= 40 years and >= 65 years, separately; and before and after excluding patients with a duration between onset of diabetes and prostate cancer mortality <5 years. RESULTS A total of 105 diabetic men died of prostate cancer during follow-up. Age was the only significant risk factor. Insulin use was associated with an insignificantly higher risk of prostate cancer mortality ranging from 24% to 49%. When stratified by the duration of insulin use < 5 and >= 5 years, a lack of significant association was also observed. CONCLUSIONS Insulin use in diabetic patients does not significantly predict the mortality from prostate cancer. Further confirmation in other ethnicities is needed.