Functional healing was measured by the timed-up-and-go test (TUG) over Weeks 6 through 20. During this assessment, the clinician timed the participant while they stood up from a seated position in a chair, walked three meters, turned around, walked three meters back to the chair, and returned to the seated position. A TUG value of ten seconds or less was considered normal for a healthy elderly person. Higher TUG values after hip fracture have been shown to be a predictor of future falls.

Least squares mean (LSM) estimates were based on a repeated measures model fitted with the log-transformed TUG values at weeks 2, 6, 12, 16, 20, 24, 36, 52 as the dependent variable and adjusted for treatment, randomized strata, gender, country category, pre-fracture community-dwelling status, pre-fracture walking aid use, quality of surgical fixation, visit, and treatment-by-visit interaction and back-transformed using the exponential transformation.

Randomized participants who received at least one dose of study drug with available data at each time point.

During the timed-up-and-go test the clinician timed the participant while they stood up from a seated position in a chair, walked 3 meters, turned around, walked 3 meters back to the chair, and returned to a seated position. A TUG value of ≤ 10 seconds is considered normal for a healthy elderly person.

LSMs were based on a repeated measures model adjusting for treatment, randomized strata, gender, country category, pre-fracture community-dwelling status, pre-fracture walking aid use, quality of surgical fixation, visit, and treatment-by-visit interaction.

Missing TUG values for participants still on study were imputed using the last observation carried forward (LOCF) when possible. If no observation could be carried forward, the maximum TUG value observed among all participants at a given visit was used. TUG values obtained after unplanned revision surgery were replaced by carrying forward the last available observed or imputed value prior to unplanned revision surgery.

Randomized participants who received at least 1 dose of study drug. LOCF imputation was used when possible, as described above.

Time to radiographic healing is the time interval from the surgery date for the eligible hip fracture to the date of radiographic healing, defined as effacement of the fracture lines by newly formed bone along the cortices and within the trabecular bone on anteroposterior and lateral (or oblique) radiographs. Radiographic fracture healing was determined by a panel of independent reviewers blinded to treatment.

The cumulative incidence function (CIF) method was used to estimate the median time to radiographic healing and the confidence intervals. Unplanned revision surgery to promote healing was considered a competing risk in CIF estimate.

All randomized participants who received at least one dose of study drug.

The radiographic Union Scale for Hip (RUSH) is a semiquantitative scoring assessment to assess hip fracture healing after surgical repair. The RUSH has 4 key domains based on radiographic parameters used by orthopedic surgeons and radiologists in routine clinical practice including cortical bridging (4 to 12 points), cortical fracture line disappearance (4 to 12 points), trabecular consolidation (1 to 3 points), and trabecular index disappearance of fracture line (1 to 3 points). The score has a minimum of 10 points (definitely not healed) and a maximum of 30 points (definitely healed).

Randomized participants who received at least 1 dose of study drug, with available data at baseline and at each time point. Missing RUSH scores for participants who were still on study were imputed using the using last observation carried forward procedure when possible.

The Harris Hip Score is a clinician-based outcome that assesses pain, function, deformity, and range of motion. The pain domain measures pain severity and its effect on activities and need for pain medication. The function domain consists of daily activities and gait. Deformity takes into account hip flexion, adduction, internal rotation, and extremity length discrepancy. Range of motion measures hip flexion, abduction, external and internal rotation, and adduction. The score ranges form 0-100 (best possible outcome) covering pain (0-44 points), function (0-47 points), absence of deformity (4 points), and range of motion (5 points). LSMs were based on a repeated measures model fitted with the Harris hip score values at weeks 2, 6, 12, 16, 20, 24, 36, and 52 as the dependent variable and adjusted for treatment, randomized strata, gender, pre-fracture community-dwelling status, pre-fracture walking aid use, quality of surgical fixation, visit, and treatment-by-visit interaction.

Randomized participants who received at least 1 dose of study drug with available data at each time point.

Hip pain was assessed using a visual analog scale (VAS). Participants were asked to rate their pain as a result of the hip fracture on a 100 mm vertical scale with 0 indicating no pain at all and 100 indicating the worst pain they could imagine.

LSMs were based on a repeated measures model fitted with hip pain score values at weeks 2, 6, 12, 16, 20, 24, 36, and 52 as the dependent variable and adjusted for treatment, randomized strata, gender, pre-fracture community-dwelling status, pre-fracture walking aid use, quality of surgical fixation, visit, and treatment-by-visit interaction.

Randomized participants who received at least 1 dose of study drug with available data at each time point.