Table 1.

Key protocol details for FEMCO Study (LREC07/H0305/61).

Protocol title, REC number, REC committee	Regional thinning of the FEMoral neck COrtex in hip fracture; a case-control study LREC07/H0305/61 ARC17822 v3.6 Cambridge Research Ethics Committee 4
Objective	To evaluate a novel index of bone fragility (regional cortical thickness) using clinical quantitative computed tomography (QCT) scanning of the proximal femur
Study design	Case control study, convenience sampling for cases and controls
Setting	Multicentre, UK (Cambridge, Norwich, Torbay). Initiated in 2007
Participants	Eligibility- inclusion criteria. Cases Patients with first hip fracture (femoral neck or trochanteric) awaiting surgical fixation, not due for surgery within 4 h of consent, able to understand, ask questions and give witnessed consent (verbal or written), medically stabilised. Controls Patients with a recent admission to orthogeriatric unit following fall from standing height or less, not sustaining hip fracture. Exclusion criteria- Dementia/cognitive impairment (AMTS < 7/10 or MMSE < 19/30), unconsciousness, terminal illness, metastatic cancer, previous hip replacement (synthetic material at either hip), previous hip fracture, osteomyelitis, bone tumour, currently taking oral corticosteroids, prior hemiplegia, prior treatment with teriparatide or strontium ranelate
Matching criteria	Convenience sample of cases. Convenience sample of fallers, not matched beyond sex, minimum age and injurious fall.
Scan protocol	Patient Positioning for Hip QCT Supine on Siemens Somatom Sensation 16, 64 or GE Lightspeed 64 scanner. Mindways 5-compartment solid phantom positioned under the hips (calibrated to aqueous K2HPO4 density), or phantom-free (using ClinicQCT asynchronous calibration) if phantom previously calibrated on that machine. Acquisition parameters: scout view from iliac crest to lesser trochanters. 120 kV, Tube current target 160mAS with Siemens CARE dosing, GE target dosing up to 320mAS. Reconstruction: To capture both hips and the phantom in each reconstruction 1 mm slice thickness (0.5 mm increment) on Siemens or 1.25 mm (0.625 increment) on GE 64. DFOV 400 mm (512*512 pixel matrix) = pixel size 0.5859 mm. Siemens B20f convolution kernel, GE ‘bone’ kernel. CT DICOM format images. Fracture classification: a consultant radiologist (TDT) reconstructed CT images with a multiplanar reformat (MPR) and classified the fracture as subcapital, transcervical or trochanteric using the Muller AO classification, based on the anterior extent of the fracture line. Image processing: QCT PRO CTXA software (v5.1.3) - reconstruct 3D image, analyse each hip for vBMD, aBMD. Segmentation of contralateral hip in fracture patients (Stradwin v 4.0) or both hips in control patients followed by Bone Mapping.
Participants used for present analysis (n)	Study 1.1 Hip fracture cases 50 females (37 femoral neck, 13 trochanteric) Study 1.1 Hip fracture cases 20 males (14 femoral neck, 6 trochanteric) Study 1.2 Hip fracture cases 60 females (46 femoral neck, 14 trochanteric) Study 1.3 Frail fallers 50 females
Demographics	Fall description, site of impact, other injuries, admitted from home/institution, MUST category (malnutrition index), weight, last recorded height (either GP record or measured), FRAX questions, pre admission Barthel Index and Functional Ambulatory Category, bone active medications, EPOS hip questionnaire.
Bone density analysis	aBMD of the femoral neck and total hip region region using traditional ROIs specified in CTXA software (QCTpro v 5.1.3).