Revision surgery (re-operation) was defined as any additional surgical intervention performed or recommended at the site of the index procedure, except those that were planned at the time of the index procedure.

Participants who were randomized and received at least 1 dose of study drug.

The signs of femoral neck fracture healing included disappearance of the fracture line on radiographs. If a participant had radiographic evidence of healing at the 12-month visit, that participant was considered to have radiographic evidence of healing.

Percentage was calculated as: (number of participants with radiographic evidence of healing / total number of participants analyzed) * 100.

Participants who were randomized and received at least 1 dose of study drug.

The worst pain numeric rating scale (NRS) was used to assess the impact of pain on a participant's life. NRS Item 3 assessed the worst musculoskeletal pain severity during the walking test. Pain was measured by an 11-point Likert scale. The following cut-points were used to categorize the NRS responses: 0 = no pain, 1 to 4 = mild pain, 5 to 6 = moderate pain, and 7 to 10 = severe pain. Participants with an NRS score of <7 were categorized as having no severe fracture-site pain with ambulation and no worsening of NRS scores >2 from baseline. Percentage was calculated as: (Number of participants with pain control during ambulation / total number of participants) * 100.

Participants who were randomized, received treatment, and had baseline and at least one nonmissing post-baseline measurement. Last observation carried forward (LOCF) values used.

The worst pain NRS was used to assess the impact of pain on a participant's life. Fracture-site pain severity was assessed for pain in the 24 hours preceding a visit. Pain was measured by an 11-point Likert scale. Participants with an NRS score of <7 in the 24 hours preceding a visit and no worsening of NRS >2 from baseline were categorized as having no severe fracture-site pain. Percentage was calculated as: (number of participants with pain control during 24 hours preceding a visit / total number of participants) * 100.

Participants who were randomized, received at least 1 dose of study drug and had baseline and at least one nonmissing post-baseline measurement for severe fracture-site pain in the last 24 hours. LOCF values used.

The worst pain NRS was used to assess the impact of pain on a participant's life. Fracture-site pain severity was assessed for pain on weight bearing. Pain was measured by an 11-point Likert scale. Participants with an NRS score of <7 during weight bearing and no worsening of NRS >2 from baseline were categorized as having no severe fracture-site pain. Percentage was calculated as: (number of participants with pain control during weight bearing / total number of participants) * 100.

Participants who were randomized, received at least 1 dose of study drug and had baseline and at least 1 nonmissing post-baseline measurement. LOCF values used.

Functional healing was defined as ability to walk with a gait speed ≥ 0.05 meters/second (m/s) with a change from baseline ≥ -0.1 m/s. The walking test involved having the participant walk a distance of 7 meters (m) at a self-selected, comfortable pace. A 4-m portion of the test was timed to determine the participant's gait speed in m/s.

Percentage was calculated as: (number of participants with functional evidence of healing / total number of participants analyzed) * 100.

Participants who were randomized, received at least 1 dose of study drug, and had either at least one nonmissing gait speed or non-ambulatory status. LOCF values used.

Ability to ambulate was defined as ambulatory with convalescent aid or without convalescent aid. Percentage was calculated as: (number of participants able to ambulate / number of total participants analyzed) * 100.

Participants who were randomized and received at least 1 dose of study drug and had at least 1 nonmissing post-baseline measurement. LOCF values used.

Prefracture ambulatory status was defined as either ambulatory with or without a walking aid. A participant was considered to have regained their prefracture ambulatory status if the participant's postsurgery ambulatory status was returned to or was improved from their pre-surgery ambulatory status. Percentage was calculated as = (number of participants who regained their ambulatory status / total number analyzed) * 100.

Participants who were randomized, received at least 1 dose of study drug, and had baseline and at least one nonmissing post-baseline measurement. LOCF values used.

The worst pain NRS was used to assess the impact of pain on a participant's life. Participants with an NRS score of <7 were categorized as having no severe fracture-site pain. Least Squares (LS) means was calculated using analysis of covariance (ANCOVA) and adjusted for baseline, treatment group, region, fracture type, and fixation type.

Participants who were randomized and received at least 1 dose of study drug and had baseline and at least 1 nonmissing post-baseline measurement.

The walking test involved having the participant walk a distance of 7 m at a self-selected, comfortable pace. A 4-m portion of the test was timed to determine the participant's gait speed in m/s. LS means was calculated using ANCOVA and adjusted for baseline, treatment group, region, fracture type, and fixation type.

Participants who were randomized, received at least 1 dose of study drug and had baseline and at least one nonmissing post-baseline measurement.

Time to revision surgery was defined as the time from initial hip fracture surgery to revision surgery, or recommendation for revision surgery if recommended but not performed. Time to revision surgery was censored at the date of the last contact.

Participants who were randomized, received at least 1 dose of study drug, and who did not have revision surgery or if they had revision surgery, it was adjudicated as not being related to the initial hip fracture surgery. Participants censored: Teriparatide = 51; placebo = 51.

SF-12 is a self-reported questionnaire covering a mental component score (MCS) and a physical component score (PCS), each scoring from a 0 to 100 (worst to best) scale. LS mean was calculated using ANCOVA and adjusted for baseline, treatment group, region, fracture type, fixation type, visit, and visit-by-treatment interaction.

Participants who were randomized, received at least 1 dose of study drug, were adjudicated as having the hip fracture in the neck of the femur and had baseline and at least 1 nonmissing post-baseline measurement.

WOMAC is: a self-reported questionnaire that consisted of 24 questions covering 3 health domains: Pain (5 items: during walking, using stairs, in bed, sitting or lying, and standing), Stiffness (2 items: after first waking and later in the day), and Physical Function. Each domain was scored by summing the individual items and transforming the scores into a 0 to 100 (best to worst) scale. LS mean was calculated using ANCOVA and adjusted for baseline, treatment group, region, fracture type, fixation type, visit, and visit-by-treatment interaction.

Participants who were randomized, received at least 1 dose of study drug, were adjudicated as having the hip fracture in the neck of the femur and had baseline and at least 1 nonmissing post-baseline measurement.

The EQ-5D is a 5-item, self-reported, generic, multidimensional, health-related, quality-of-life instrument with 5 items. Overall health state score was also self-reported using a visual analogue scale (VAS) marked on a scale scored from 0 (worse imaginable health state) to 100 (best imaginable health state). LS mean was calculated using ANCOVA and adjusted for baseline, treatment group, and region.

All randomized participants who were randomized, received at least 1 dose of study drug, were adjudicated as having the hip fracture in the neck of the femur, and had baseline and at least 1 nonmissing post-baseline measurement.