Of fracture thresholds and bone mineral density reference data: does one size really fit all?

Men who fracture have greater mean bone mineral density (BMD) than women who fracture, and to some this suggests that BMD fracture "thresholds" are greater in men than women, justifying use of a male BMD reference standard. Others disagree and argue that because BMD distribution in men is right-shifted (higher) compared with women, anything that occurs with equal probability for men and women will occur at a greater mean BMD in men. If the latter is true, it supports using a common (female) BMD reference standard. We directly tested this latter principle and, indirectly, the validity of using a constant BMD reference standard, in 51,326 women (3722 major fractures) and 4691 men (276 fractures), by comparing mean BMD in fracture vs nonfracture groups defined by sex, age, race and body mass index because these 4 factors affect both mean BMD and risk of fractures. Among those who fractured, mean BMD for all measurement sites were significantly greater in men vs women, youngest vs oldest, whites vs Asian, and heaviest vs the lightest (all p < 0.001). However, the same BMD pattern was seen in those who did not fracture and the absolute difference in BMD between those who fractured and those who did not was essentially constant regardless of sex, age, race, or weight class (all p-interaction nonsignificant). This finding suggests that use of a common reference standard for BMD (i.e., young white women) for men, and indeed all subgroups defined by osteoporosis risk factors, is reasonable and consistent with empiric observations.