The latter is affected by gait, balance, muscle strength, vision, and environment, among other factors. It turns out the last factor can and should be evaluated during a home visit (something that isn't performed enough in this day & age of assembly-line patient care). However, the first three factors can be roughly tested via TUG analysis . Just time how long it takes your patient to get up out of a chair (ideally w/o using his/her upper extremities), walk 3 meters (~10 feet), turn around, walk back, and sit down.
In this particular population of community dwelling Australians average 75yo, average TUG time was just over 9 seconds, although more than one third took over 10 seconds to complete the test. Over 10yrs of observation, those participants who required more than 10 seconds to complete the TUG test had a greater risk of non-vertebral fracture, regardless of hip BMD, compared to those who completed their mobility in less than 10 seconds.
Of note, those with TUG test times less than 10 seconds AND normal BMD (T score greater than -1.0), had the lowest risk of fracture. Those with TUG test times greater than 10 seconds AND abnormally low BMD (T score less than -1.0), had the greatest risk of fracture. And those with a composite of the two results (one high and one low) had near identical rates of fracture intermediate between the two extremes.
So while DEXA (or DXA) is important, the TUG test time is also an independent risk factor in osteoporotic fracture assessment. The best part? It's free and easy to perform in any setting. Better yet, it's something family members can quietly observe over the upcoming holiday family get-togethers and make appropriate referrals to their elders' family physician for further evaluation if need be.