This study is a parallel-group randomised trial with a one-year follow-up (Figure 1). The Human Ethics Committee of La Trobe University Human has approved the trail and all participants will give written informed consent. Participants will be randomly allocated to either a "usual care" control group or the "multifaceted podiatry" intervention group. Permuted block randomisation will be undertaken using an interactive voice response telephone service provided by the National Health and Medical Research Council Clinical Trials Centre at the University of Sydney. Assessors will be blinded to group allocation but due to the nature of the trial, the participants will not be blinded to group allocation.

Community dwelling men and women aged 65 years and over will be recruited by a mail-out letter from two databases: (i) people who are currently accessing podiatry services at the La Trobe University Health Sciences Clinic, Bundoora, Victoria, Australia and (ii) people who are currently accessing podiatry services at the Bundoora Extended Care Centre, Victoria, Australia as well as from advertisements placed in seniors newspapers and websites. Respondents will be initially screened by telephone to ensure they are able to walk household distances without the use of a walking aid and are able to read and speak basic English. Individuals who meet the initial screening criteria will then be invited to be assessed for eligibility.

To be included in the study, participants must meet the following inclusion criteria:

(i) an elevated risk of falling, defined as either a history of a fall in the previous 12 months, a score of > 1 on the short Physiological Profile Assessment (PPA) 5 or performance on the alternate stepping test (the time taken to alternately place each foot on a 19 cm high step eight times) of > 10 seconds 38;

(ii) self-reported disabling foot pain (defined as people who have had foot pain lasting for at least a day within the last month and a positive response to at least one item on the Manchester Foot Pain and Disability Index 39);

(iii) cognitively intact (defined as a score of ≥ 7 on the Short Portable Mental Status Questionnaire 40).

Participants will be excluded if they have Parkinson's disease (or other neurodegenerative disorders) or lower limb amputation (including partial foot amputation).

The control group will be asked to continue whatever podiatry treatment they currently receive for the 12 months of the study. All participants will be offered basic podiatry treatment free of charge in the La Trobe University Health Sciences clinic for the 12 months of the study regardless of whether they are current patients of the clinic. This will typically include toenail maintenance and scalpel debridement of hyperkeratotic lesions (corns and calluses). This intervention is consistent with the usual ongoing "maintenance" care that is provided to older people attending public sector podiatry services (such as community health centres, outpatient podiatry clinics and Department of Veterans' Affairs subsidised private podiatry).

The intervention group will also be asked to continue whatever podiatry treatment they currently receive for the 12 months of the study and all participants will be offered basic podiatry treatment free of charge in the La Trobe University Health Sciences clinic for the 12 months of the study regardless of whether they are current patients of the clinic. In addition, they will receive a multifaceted podiatry intervention consisting of:

(i) Footwear advice and provision: participants' outdoor footwear will be assessed using a footwear assessment form for which the component variables have been shown to have good intra-rater reliability (intra-rater kappa 0.62 – 1.0) 41. Participants will be deemed to have inappropriate footwear if the heel height is greater than 4.5 cm or the shoe has any two of the following; no fixation, no heel counter or the heel counter can be depressed to greater than 45°, the tread pattern of the sole is fully worn or manufactured with a smooth sole, or the shoe heel width is narrower than the participant's heel width by greater than or equal to 20%. Participants with inappropriate footwear will be counselled regarding the specific hazardous footwear feature/s identified, and will be provided with a handout on what constitutes a safe shoe. They will then be given the contact details of an extra-depth and medical grade footwear retailer and will be asked to purchase a more appropriate pair of shoes. Their purchase of footwear will be assisted by the provision of an AUD$100 voucher.

(ii) Foot orthoses: Prefabricated insoles (Formthotics™, Foot Science International Ltd, Christchurch, New Zealand) manufactured from a thermoformable cross-linked closed cell polyethylene foam will be shaped to fit the participant's foot (Figure 2). The orthoses will then be appropriately customised using 3 mm thick PPT urethane to redistribute pressure away from plantar lesions (Figure 3).

(iii) Home-based exercise program: participants will be provided with instructions and practical demonstrations at the baseline assessment to allow them to undertake a 30 minute home-based exercise program aimed at stretching and strengthening the muscles of the foot and ankle. Strength will be assessed at baseline and appropriate intensity prescribed. A summary of the individual exercises is provided in Table 1. All necessary equipment to undertake the exercise program, including an Archxerciser™ (Elginex Corporation, Lombard, Illonois, USA) (Figure 4) will be provided. Participants will also be given an illustrated explanatory booklet of each exercise, a DVD demonstrating the exercises and an exercise diary to be returned monthly to the researchers. The program will be performed three times per week for six months, and participants will be contacted at one, four, twelve and twenty weeks by telephone to promote adherence to the program.

(iv) Falls prevention education: the booklet "Don't fall for it. Falls can be prevented!" subsidised by the Commonwealth Department of Health and Ageing will be provided. This booklet provides a general overview of risk factors for falls and outlines strategies to prevent falls, including regular eye checks, medication assessment, household hazard assessment and general exercise guidelines.

Demographic data will be collected at baseline and will include age, gender, height, weight, country of birth, education, source of income, falls history, eye problems, health conditions, medications, walking aids, use of community services, foot problems, foot posture and use of podiatry services. The Active Australia survey questionnaire will be used to obtain a baseline measurement of the physical activity levels of the participants 42.

All participants will be followed for 12 months following baseline assessment to record the number of falls. Falls, defined as "an unexpected event in which the participant comes to rest on the ground, floor, or lower level" 43 will be monitored using monthly mail-out calendars. When a fall occurs, specific details about fall injuries will be obtained through structured telephone interviews. If falls calendars are not returned at the end of each month, research staff will contact the participants by telephone to obtain the missing data. Consistent with the recommendations of the Prevention of Falls Network Europe group, three falls outcomes will be used as the primary outcomes: the number of fallers, the number of multiple fallers and the falls rate. As a safety measure, the time to first fall will be recorded as a secondary measure to reflect adverse events from the intervention 43.

All participants will be assessed at baseline and at six months by an assessor blinded to group allocation. The secondary outcome measures include:

(i) Foot pain: the Manchester Foot Pain and Disability Index (MFPDI) will be used to assess foot pain 39. The MFPDI consists of 19 statements prefaced by the phrase "Because of pain in my feet", formalised under three constructs: functional limitation (10 items), pain intensity (five items), and personal appearance (two items), with three possible answers: "none of the time" (score = 0), "some days" (score = 1), and "most days/every day" (score = 2). The last two items are concerned with difficulties in performing work or leisure activities, which are omitted if the respondent is of retirement age. The total score (range: 0 to 34) will be used to measure the degree of improvement in foot pain. The MFPDI has been shown to be a suitable tool for assessing foot pain in the older population 44, has high construct validity and has been shown to be sensitive to improvement following a self-management intervention 45.

(ii) Physiological falls risk: changes in physiological falls risk will be assessed using the Physiological Profile Assessment (PPA) 5. Based on the performance of five physiological domains (vision, proprioception, strength, reaction time and balance), the PPA computes a falls risk score (standardised score) for each individual; this measure has a 75% predictive accuracy for falls in older people 46. The five assessment items (Figure 5) are a test of vision (edge contrast sensitivity using the Melbourne Edge test), peripheral sensation (lower limb proprioception), lower limb strength (knee extension strength), reaction time using a finger press as the response, and body sway (sway when standing on a medium density foam rubber mat). Details regarding each of these tests are provided elsewhere 5. Several studies have reported significant improvements in PPA scores following exercise interventions 4748.

(iii) Fear of falling: Fear of falling will be measured using the 7-item short Falls Efficacy Scale-International (FES-I) 49. The FES-I uses a Likert scale to score the participant's level of concern regarding the possibility of falling when performing certain activities of daily living (e.g. taking a bath/shower or climbing up or down stairs). There are four responses which are: "not at all concerned" (score = 1), "somewhat concerned" (score = 2), "fairly concerned" (score = 3) and "very concerned" (score = 4). The total score ranges from 7 (no concern about falling) to 28 (severe concern about falling). The FES-I has been shown to have excellent validity 50 and reliability in the older people (Cronbach's alpha 0.92, ICC = 0.83) 49.

(iv) Generic health related quality of life: Mental and physical health status will be measured with the Mental (MCS-12) and Physical (PCS-12) Component Summary scores of the short form Health Survey (SF-12). The SF-12 was developed and validated as a shorter alternative to the SF-36 51. To calculate PCS-12 and MCS-12 scores, SF-12 items are scored and normalized via a standardised algorithm. PCS-12 and MCS-12 scores range from 0 to 100, with higher scores indicating better functioning. The MCS-12 and PCS-12 were designed to have a mean score of 50 and a standard deviation of 10 in a representative sample of the US population 52. Test-retest correlations of 0.89 and 0.76 were observed for the 12-item PCS-12 and the MCS-12, respectively 51.

(v) Foot and ankle strength: Maximal isometric muscle strength of foot and ankle muscles (ankle dorsiflexion, plantarflexion, inversion and eversion, hallux plantarflexion and lesser toe plantarflexion) will be assessed as the average of three trials using a hand held dynamometer (Citec, CIT Technics, Haren, The Netherlands). Using hand-held dynamometry for testing foot and ankle strength has been shown to have good reliability (ICC 0.88 to 0.95) 53. Plantarflexion strength of the toes will also be tested using the paper grip test, where the participant is seated with their knee and ankle at 90°, and instructed to use their toe muscles to push down on a 1 mm piece of card (e.g. a business card) while the examiner stabilises their ankle and attempts to slide the card away from the toes (Figure 6A). An inability to hold the card on any one of three trials is recorded as a fail. Using receiver operating characteristics (ROC) curves, this test has been shown to have a positive predictive value of detecting weakness of 95% for the hallux and 90% for the lesser toes 54. Maximal isometric muscle strength for ankle dorsiflexion will also be measured as the maximum of three trials with the participant sitting in a high chair with their foot secured to a footplate attached to a spring gauge (Figure 6B). This test has been shown to have good reliability (ICC 0.88) 5. Testing will be conducted on the self-reported dominant side identified by the response to the question "Which foot would you use to kick a ball with?".

(vi) Foot and ankle range of motion: First metatarsophalangeal joint dorsiflexion range of motion will be measured using a goniometer as the maximum angle at which the hallux cannot be passively moved into further extension in a non-weightbearing position (Figure 7A). The reliability of this test (ICC 0.95) has been reported previously 55. Ankle dorsiflexion flexibility will be recorded using a modified version of the lunge test which has been shown to have high reliability (ICC = 0.88) 56. Participants will be asked to take a comfortable step forward and squat as low as possible, keeping their trunk upright, without lifting the heel of their backfoot from the ground. The degree of motion will be recorded using a digital inclinometer placed on the mid-point of the anterior tibial border (Figure 7B). The test will be conducted with the knee extended and then with the knee flexed 5758. To assess ankle inversion and eversion, participants will be seated with the lower leg unsupported. Landmarks will be made on the participants at the midpoint between the malleoli on the anterior aspect of the ankle, the midline on the anterior aspect of the lower leg using the crest of the tibia as a reference point, and the longitudinal midline on the anterior surface of the second metatarsal. A flexible universal goniometer will be aligned along the landmarks and the participants will move their ankle from a self-selected neutral position actively to the end of range of inversion and eversion (Figure 7C). High intra-observer reliability has been reported for these tests (ICC 0.82 to 0.96) 59. Testing will be conducted on the self-reported dominant side.

(vii) Balance and functional ability: Balance will be assessed through four tests. Postural sway will be measured as previously described in the section regarding the PPA (Figure 8A). The reliability (ICC 0.57 to 0.68) of this test has been reported previously 60. Lateral stability will be measured using a sway meter that measures displacement of the body at the waist with testing performed with the participant standing with one foot in front of the other. Participants will choose which foot to place in the forward position and will be instructed to stand as still as possible for 30 seconds. Lateral displacement of the body will then be recorded 61. Leaning balance will be measured using the maximum balance range test and coordinated stability tests 62. For the maximum balance test, (Figure 8B) the sway meter is attached around the waist with the rod extending anteriorly. Participants are asked to lean forward then backwards from the ankles without moving their feet, as far as possible, i.e. to the point where they can just retain their balance. Maximal anterior-posterior displacement is measured over three trials. Using similar apparatus, the coordinated stability test (Figure 8C) requires the participant to bend and rotate at the hips without moving their feet to move the sway meter pen around a convoluted track marked on a piece of paper attached to the top of an adjustable height table. A total error score is calculated by summing the number of occasions the pen strays outside the track 62. Good reliability has been reported for both the maximum balance test (ICC = 0.74) and the coordinated stability test (ICC = 0.83) 62 and both tests are predictive of falls63. These tests will be performed twice, once with the participant barefoot and again with their preferred outdoor footwear. Lateral stability, maximum balance range, coordinated stability and walking speed will be corrected for height prior to analysis 11.

Functional ability will be evaluated using the sit to stand test (time taken to rise from a 43 cm high chair five times without using the arms) (Figure 8D), alternate stepping test (the time taken to alternately place each foot on a 19 cm high step eight times) (Figure 8E), and walking speed over 6 metres (Figure 8F). These tests have been shown to be reliable in previous studies of older people (sit to stand (ICC = 0.89) 64, alternate stepping test (reliability coefficient of 0.98) 65 and 6 m walk (reliability coefficient of 0.93) 66 and have demonstrated good sensitivity and specificity in identifying multiple fallers 38.

The sample size for the study has been determined a priori using an appropriate sample size formula 67. A sample size of 143 participants in each group provides 80% power to detect a 30% reduction in the intervention group in the primary outcome of the percentage of fallers. This reduction in falls was based on the following: (i) a falling rate of 60% in the control group (as has been observed in older people with foot pain in a previous risk factor study by Menz et al 12), and (ii) a falling rate of 42% in the intervention group (which equates to a 30% reduction in the number of fallers, recently demonstrated in a similar multifaceted study by Clemson et al 68). We chose an alpha level of 0.05 and allowed for a drop-out rate of 15%. To allow for unforeseen circumstances we will aim to recruit 300 participants (i.e. 150 per group).

Statistical analysis will be undertaken using SPSS version 14.0 (SPSS Corp, Chicago, Ill, USA) and STATA 8 (Stata Corp, College Station, Tex., USA) statistical software. All analyses will be conducted on an intention-to-treat principle using all randomised participants 69. Missing data will be replaced with the last score carried forward. Demographic characteristics and baseline data will be summarised by descriptive statistics. Consistent with the recommendations of the Prevention of Falls Network Europe group 43, three falls outcomes will be used as the primary outcomes: the number of fallers, the number of multiple fallers and the falls rate. The number of fallers and multiple fallers (two or more falls) will be compared by calculating relative risks. The number of falls and falls rate per person per year in the two groups will then be compared using negative binomial regression models. This approach takes into account all falls and adjusts for varying duration of follow-up 7071. The continuously-scored secondary outcome measures at baseline and the six month follow-up appointments will be compared using analysis of covariance with baseline scores and intervention group entered as independent variables 7273.