Does insulin resistance predict prostate cancer? Results from the Reduction by Dutasteride of Prostate Cancer (REDUCE) Trial

Abstract Purpose Prior studies testing the association between insulin resistance (IR) and prostate cancer (PC) risk are inconsistent. We examined the association between Homeostatic Assessment of Insulin Resistance (HOMA‐IR; calculated from fasting baseline insulin and glucose) and PC in REDUCE, a 4‐year randomized trial of dutasteride vs. placebo for PC prevention. Experimental design All patients had prestudy negative biopsies and underwent study mandated biopsies at 2 and 4 years regardless of prostate‐specific antigen. Multivariable logistic regression models were used to investigate the associations between log‐transformed or categorized HOMA‐IR scores and PC risk. Multinominal regression was used to assess associations between HOMA‐IR scores and tumor grade (low grade [grade group 1]; high‐grade [grade groups 2–5]). Results Among 5430 REDUCE participants (1212 with PC; 856 low‐ and 356 high‐grade), higher HOMA‐IR was associated with lower PC risk (log‐HOMA‐IR: OR, 0.89; 95% CI, 0.80–0.99; p = .03; categorized HOMA‐IR: p ‐trend = .04). When stratified by grade, HOMA‐IR was significantly associated with reduced low‐grade PC risk (log‐HOMA‐IR: OR, 0.84; 95% CI , 0.74–0.94; p = .003; categorized HOMA‐IR: p ‐trend = .002) but was unrelated to high‐grade PC (log‐HOMA‐IR: OR, 1.02; 95% CI, 0.86–1.21; p = .81; categorized HOMA‐IR: p ‐trend = .26). Results were similar in placebo and treatment arms. Conclusions In summary, higher HOMA‐IR was associated with a reduced risk of low‐grade PC but was not associated with high‐grade disease. The mechanisms to explain these findings are unclear.