A study in African American families found little correlation between S217L and prostate cancer (they note a p=0.03 difference in heterozygote sporadic cases). Sporadic cases: L/L: 3, S/L: 69, S/S: 83. Familial cases: L/L: 3, S/L: 30, S/L: 55. Controls: L/L: 15, S/L: 99, S/S: 182. Combined numbers with a recessive hypothesis is used below.
The most recent meta-analysis of ELAC2 polymorphisms Ser217Leu and Ala541Thr, this concludes that these are associated with prostate cancer and may be “low-penetrance susceptibility markers of prostate cancer”. They note that there is publication bias for results supporting pathogenicity (p = 0.006) but that no single paper is necessary for their conclusion of pathogenicity (ie removal of any single paper doesn’t change this result).