Buzkova P, et al. - In the Cardiovascular Health Study which is an ongoing cohort study of 5,888 elderly people, experts evaluated risk factors of non-fatal outcomes amid a competing risk of mortality. Cox and Fine-Gray (FG) methods gave directionally concordant yet quantitatively distinctive strengths of the relationship for the age, current smoking, and gender. On the contrary, the FG approach assessed a robust relationship of hip fracture with gender. The two approaches gave approximately indistinguishable outcomes for the race. For diabetes and kidney function, the estimates were different in direction, and the FG HRs implied influences that were in the inverse direction of well-understood and broadly trusted correlations. In conclusion, even in the presence of competing risk of mortality, for non-fatal outcomes such as hip fracture, cause-specific Cox models give relevant estimates of hazard. In the population of people who have not yet had an incident hip fracture and remain alive which is typically the group of clinical attention, the Cox approach estimates hazard. Moreover, in populations of clinical interest, the Fine-Gray method estimates hazard in a hypothetical population that can produce misleading conclusions about risk factors.