Effectiveness of a home-based physical exercise intervention in patients with hip fragility fractures: a randomized controlled trial

Abstract

We aimed to test the effectiveness of an invidually tailored home-based exercise intervention compared with usual care for independence, functional capacity and quality of life in patients with hip fragility fractures. In this single-blinded randomized-controlled trial (NCT04934358), patients with surgically treated fragility hip fractures were randomized to receive a 3-month exercise intervention or usual care. The exercise group performed a predetermined number of steps per day with resting pauses and a number of sit-to-stand exercises at home. The outcomes collected at baseline, at the end of the program and at follow up (6-months) included: the Barthel Index (primary outcome), quality of life, cognition and depression levels, 6-minute walking test, and 5-time sit-to-stand test. A total of 35 patients were randomized: 16 to the exercise group and 19 to the control group (mean age 80 ± 10 years, men n = 9). At the end of the program, patients in the exercise group presented significantly greater improvement in all primary and secondary outcomes than those in the control group did (Barthel Index variations: 12 ± 4 vs. 7 ± 5; p = 0.046), although both groups exhibited significant improvements from baseline for almost all the outcomes. Compared with usual care, a simple, individually tailored home-based exercise program favored the recovery of independence, quality of life, and functional capacity after hip fragility fractures.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-55826-w.

Introduction

Noncommunicable diseases and chronic conditions associated with population aging increase the demand for care and services and necessitate the economic sustainability of health care systems¹.

One example of this complex scenario is fragility fractures, which occur worldwide at a frequency of 70 events per minute². Fragility fractures are defined as fractures resulting from low-energy trauma, typically a fall from standing height or less, and are strongly associated with underlying skeletal fragility. These fractures predominantly affect individuals aged ≥ 50 years and are more common in women, largely because of the higher incidence of low bone mineral density (BMD) and osteoporosis in this population^3–6. It is estimated that one in three women and one in five men over the age of 50 will experience an osteoporotic fracture during their lifetime^7–9. Osteoporosis, which affects approximately 500 million men and women worldwide^1,10, is characterized by reduced bone mass and deterioration of bonemicroarchitecture, resulting from an imbalance between bone resorption and bone formation. Its development is influenced by aging, hormonal changes, and several secondary causes, including chronic alcohol consumption, inflammatory rheumatic diseases, and endocrine, hepatic, or renal disorders^11,12. Fragility fractures most commonly involve the hip, vertebrae, distal radius, and proximal humerus¹². Although the association between osteoporosis and fragility fractures is well established in the scientific literature, substantial gaps persist in postfracture diagnosis, risk stratification, and the initiation of appropriate secondary prevention strategies¹³. These fractures are accompanied by physical and psychological complications, such as pain, loss of mobility¹⁴, and reduced quality of life, autonomy, and independence^15,16. Fear of falling, anxiety and depression^17–20 are often reported. Finally, fractures are also associated with increased hospitalization morbidity and mortality^21,22. The annual cost of osteoporosis fractures to healthcare systems worldwide is $400 billion, accounting for approximately 3% of healthcare costs¹². In the European Union, these costs amount to 37.4 billion euros, representing approximately 3% of total healthcare costs.

The International Osteoporosis Foundation implemented the “Capture the Fracture” program internationally to stimulate the detection of patients with osteoporotic fractures and avoid a second fracture. Treatment strategies include nonpharmacological approaches and osteoprotective measures. Several osteoporosis societies recognize fracture liaison services as the most efficient organizational model for enhancing assessment and treatment following osteoporotic fractures and as cost-effective for preventing secondary fractures^23,24. The imminent fracture risk, which is highest in the immediate postfracture interval²⁵, calls for prompt interventions to optimize the benefits of preventive and rehabilitation treatments. The use of behavior change interventions has increased in primary care. In this context, printed educational materials are advised to support brief interventions, even though no outcome improvement has been clearly demonstrated²⁶. Thus, simple behavioral patterns in everyday life can significantly impact health.

Pharmacological interventions can improve both bone strength and exercise, which, however, may provide better muscle control, balance, and coordination with reduced fall risk and increased muscle strength and BMD^27,28. Compared with conventional care, interventions aimed at improving mobility after hip fracture may lead to clinically meaningful improvements in mobility and walking speed in hospital and posthospital settings²⁹. However, considering the difficulty of providing adequate postrehabilitation treatment, home-based exercise has the potential to be beneficial and effective³⁰. Indeed, there is an urgent need to identify effective home-based exercise programs possibly powered by fall prevention^31,32, given that only a few clinical trials have focused on motivation and exercise strategies^26,33,34. Previous studies have successfully tested the feasibility and effectiveness of home-based low-intensity progressive exercise programs to improve the functional capacity of severely dependent populations (peripheral arterial disease, dialysis, etc.)^35–38. The development of this type of low-cost program has also been associated with persistent mobility and a reduction in hospitalizations and mortality in frail patients, including patients undergoing dialysis^39–41.

Specifically, we aimed to study whether in patients with fragility fractures, a progressive exercise program adjusted to individual exercise capacity combined with motivational interventions can improve functional independence and functional outcomes at a 3-month follow-up compared with a group of patients receiving usual care.

Methods

This prospective single-center single-blinded randomized controlled trial is reported according to the CONSORT, CERT and TIDIER guidelines. The study was conducted at Hospital Universitario Reina Sofia, Cordoba, Spain, between September 2022 and September 2024. The Research Ethics Committee of the province of Cordoba approved the project. The study was performed in accordance with relevant European Union regulations and in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants.

The trial was registered before its start on Clinicaltrials.gov with the following ID: NCT04934358 (registration date: 14/06/2021). Further details can be found in the published study protocol⁴².

Eligibility criteria

Male and female participants aged greater than 50 years who met the following inclusion criteria were invited to participate:

• i)Suffering a fragility⁴³ hip fracture (including cervical and trochanteric fractures), classified in the 10th version of the International Classification of Diseases (ICD-10-ES) with codes S72–S72.26X;
• ii)Surgical repair via arthroplasty or internal fixation (ICD-10-ES: 0SU, 0SR) within one week, with postoperative care following standard protocols, including early mobilization and weight-bearing as tolerated unless contraindicated surgeon;
• iii)Independent or mildly dependent on activities of daily living (Barthel index > 60) and living at home before the fracture.

Patients were excluded from the study if the following criteria were meet:

• i)Low cognitive level (scores below 23 points on the Lobo Mini-cognitive test);
• ii)Pathological fractures due to skeletal alterations other than osteoporosis;
• iii)History of more than one fall (other than the cause of hospital admission) in the previous year;
• iv)Uncorrected anemia (hemoglobin concentration 10 mg/L), severe cardiorespiratory disease (e.g., unstable angina; cardiac failure; heart failure).
• v)Conditions that do not allow safe exercise at home (e.g., severe cardiorespiratory disease; unstable angina, severe heart failure, major lower limb amputation, etc.)

Randomization

Eligible participants were randomized between the two groups (exercise or control) through a dynamic⁴⁴ computerized randomization stratification approach in which an external administrator was not involved in the trial. The randomization scheme was organized into a block of at least 4 people, and a 1:1 allocation ratio was used, with the physical activity capacity acting as a unique stratum.

Exercise group

Patients randomized to the exercise group performed home-based walking training tailored to their baseline physical capacity, as measured by the 6-minute walk test (Table 1). Participants were enrolled and randomized during hospitalization before discharge, following standard orthopedic and geriatric criteria. The home-based intervention started immediately after discharge when patients were stable and returned home. In addition to the optimization of medical therapy, a progressive home-based training program based on a number of daily steps and lower limb strength training was prescribed. Patients were asked to add a number of steps per day, excluding usual daily activities, and to perform a fixed number of sit-to-stand movements. Walking should be performed at least 5 days a week, and the resting time between each sit-to-stand series should be at least 1 min. Training progressed as reported in Table 1. A fitness tracker was provided to each patient to be connected to patients’ or caregivers’ smartphones to monitor the daily execution of the exercise program. At the end of the three-month exercise program, patients were advised to maintain an active lifestyle without any specific exercise prescription.

Table 1.

Proposed interventions according to patients’ baseline exercise capacity.

Exercise capacity	6MWD < 200 m		6MWD ≥ 200 < 400 m		6MWD ≥ 400 m
Week	Walking (steps per day)	Sit and stand (reps/set)	Walking (steps per day)	Sit and stand (reps/set)	Walking (steps per day)	Sit and stand (reps/set)
1	1000	3 × 2	2000	5 × 3	4000	4 × 4
2	1000	3 × 2	2000	5 × 3	4000	4 × 4
3	1250	3 × 3	2000	5 × 3	5000	5 × 4
4	1250	3 × 3	2500	4 × 4	5000	5 × 4
5	1500	4 × 3	2500	4 × 4	6000	5 × 4
6	1500	4 × 3	2500	4 × 4	6000	5 × 5
7	1750	5 × 3	3000	5 × 4	7000	5 × 5
8	1750	5 × 3	3000	5 × 4	7000	6 × 5
9	2000	4 × 4	4000	5 × 4	8000	6 × 5
10	2000	4 × 4	4000	5 × 5	8000	7 × 5
11	2250	5 × 4	4500	5 × 5	9000	7 × 5
12	2250	5 × 4	4500	5 × 5	9000	8 × 5
13	2500	6 × 4	5000	5 × 5	10,000	8 × 5

Abbreviations: 6MWD: 6-minute walking distance; reps/set: repetitions per set.

Control group

Patients enrolled in this group after discharge from the hospital will receive optimized medical therapy and recommendations to maintain an active lifestyle. In addition, if deemed necessary for the patient’s condition, additional multidisciplinary rehabilitation programs could be provided in different care settings.

Patients in both groups received a follow-up daily log to record the physical activity carried out and/or possible adverse events that may have occurred.

Outcomes

Outcome data were collected at baseline (T0), at the end of the program for the intervention group (T3) and at the 6-month follow-up (T6) by blinded assessors.

The primary outcomes of the study were functional independence and activities of daily living, measured through the Spanish version of the Barthel Index (BI)⁴⁵. The BI is a scale ranging from 0 to 100 and assesses a patient’s independence in terms of feeding, bathing, grooming, dressing, bowel and bladder control, transfer, mobility and climbing stairs. A higher value corresponds to greater independence.

The secondary outcomes were as follows:

• Number of hospital readmissions within 6 months of recruitment, along with reasons, dates and lengths of stay. The data will be gathered from the regional hospital database and eventually censored at the date of death.

• Cognitive capacity was assessed via the Lobo Cognitive Mini-examination⁴⁶. This scale is a cognitive test used to detect possible dementia in geriatric patients and monitor its progression. It has a maximum score of 35 points and covers five cognitive areas: spatial–temporal orientation, immediate memory, concentration and calculation, memory, and language.

• Physical activity levels were measured through the Spanish reduced version of the Minnesota Leisure Time Physical Activity Questionnaire (VREM)⁴⁷. This questionnaire consists of 6 questions, provides information on energy expenditure during leisure time and allows individuals to be classified into activity categories.

• Depression was measured via the Yesavage Depression Scale⁴⁸. This is a 15-items questionnaire used to screen for depression among people over 65 years of age.

• Health-related quality of life was assessed through the EQ-5D-5L questionnaire⁴⁹, which assesses a person’s health through questions on mobility, self-care, activities of daily living, pain and anxiety/depression, as well as a numerical scale of general health.

• Exercise capacity was measured via the 6-minute walking test⁵⁰. Patients were asked to walk back and forth on a 10-meter corridor, aiming to cover the longest distance possible, which was recorded as the 6-minute walking distance (6MWD). Participants were permitted to pause if necessary because of fatigue or symptoms and to resume as soon as possible. The 6MWT was conducted under typical ambulatory conditions, including assistive devices when necessary.

• Lower limb strength was measured by the 5-time sit-to-stand test (5STS). Patients whose arms were folded across the chest had to complete 5 movements of raising from the standard-height chair and sitting back to the starting position in as little time as possible (sit-to-stand exercises)⁵¹.

• Fear of falling was measured through the Falls Efficacy Scale, a 10-item self-report measure assessing confidence in performing everyday activities without falling, with higher scores indicating greater fear of falling or lower falls-related self-efficacy⁵².

• An ad hoc questionnaire consisting of 22 items was used to assess knowledge about osteoporosis, with a focus on dietary, lifestyle, and physiological factors. The topics covered included the definition of osteoporosis, risk factors, and the roles of calcium and vitamin D. Lifestyle choices, hormonal influences, and related health conditions were examined. The scoring system categorizes knowledge levels as high (17 to 22), medium (13 to 16), or low (0 to 12).

• Bone mineral density in measured as the amount of mineral matter per square centimeter of bone tissue via dual-energy X-ray absorptiometry (DEXA). The t-score serves as a statistical measure to compare a patient’s bone mineral density with the average of a healthy young adult population. A t-score less than − 1 indicates lower than normal bone density.

• Blood samples at T0 and T3 were analyzed for specific biological indicators, including calcium and vitamin D levels. Standard laboratory methods were employed for collection and measurement of these outcomes.

Sample size

In the absence of previous trials published with similar interventions and considering both the low rate of enrollment due to the strict inclusion criteria and the low rate of acceptance by patients, as previously reported^34,42, the sample size for a feasibility pilot superiority trial was calculated not only on a potential acceptance basis⁴² but also on the primary outcome. Hypothesizing, as previously reported³⁴, a mean baseline improvement in the Barthel Index of 0.6 for the control group and 2.0 for the intervention group, a total of 32 patients will be sufficient to confirm the superiority of an exercise intervention with respect to usual care, with 90% and a significance level of 0.05.

Statistical analysis

Standard methods of RCTs analysis were employed. The data distribution was verified through the Shapiro‒Wilk test. Baseline comparisons between the two groups were assessed by a chi-squared test for categorical variables and the Mann‒Whitney test or Student’s t test for continuous variables.

Comparisons between the two groups were performed with Student’s t test or the Mann–Whitney U test as appropriate, and between-group mean changes from baseline to the end of the study were compared. Repeated-measures analyses of variance (RM-ANOVAs) with group allocation as a between-subject variable and a Huynh-Feldt source of variation were employed to assess all the outcomes, including the follow-up time.

Within-group variations were assessed via paired samples t test or the Wilcoxon signed-rank test.

An intention-to-treat approach was used, with missing values replaced via the multiple imputation procedure. A sensitivity per-protocol analysis was also conducted. To evaluate the clinical effect size, Cohen’s d was calculated for the primary and main secondary outcomes. A p value < 0.05 was considered significant. The data were analyzed via SPSS 21.0 (IBM, Armonk, NY, USA) and MedCalc statistical software (version 23.1.1; MedCalc Software bvba, Ostend, Belgium).

Results

A total of 524 patients were screened for eligibility, 35 of whom met the inclusion criteria and were randomized into two groups (Fig. 1). At baseline, the two groups were similar in terms of their anthropometrics, fracture type and characteristics, comorbidities and all the outcome measures (Tables 2 and 3). No systematic differences in discharge timing between groups were observed. After the surgical intervention, all patients were treated with vitamin D supplementation as recommended by the guidelines.

Fig. 1.

Study flow diagram.

Table 2.

Baseline comparison of the two groups for demographics and clinical characteristics.

	Exercise (n = 16)	Control (n = 19)	p
Age, years	80 ± 9	80 ± 11	0.90
Males, n (%)	5 (31)	4 (21)	0.50
Current smoking, n(%)	1 (6)	1 (5)	0.78
Hypertension, n(%)	11 (67)	10 (53)	0.34
Hyperlipidemia, n(%)	4 (25)	7 (37)	0.46
Diabetes, n(%)	7 (44)	7 (37)	0.68
Chronic Kidney Disease, n(%)	1 (6)	3 (16)	0.38
Lives alone, n (%)	1 (6)	4 (21)	0.22
Fracture type, n(%)
Subcapital	10 (63)	9 (47)	0.35
Subtrocanteric	0 (0)	2 (10)
Peritrocanteric	6 (37)	8 (43)
Dominant leg	9 (60)	12 (63)	0.85

Table 3.

Baseline comparison of the two groups for variables under study.

	Exercise (n = 16)	Control (n = 19)	p
Barthel Index	71 ± 7	66 ± 11	0.11
EQ-5D-5 L VAS	62 ± 10	64 ± 7	0.41
VREM	124 ± 105	106 ± 72	0.54
Cognition level	9.4 ± 5.7	9.4 ± 5.4	0.98
Yesavage depression scale	3.7 ± 1.9	3.4 ± 1.2	0.53
Falls Efficacy Scale	41 ± 8	46 ± 9	0.08
5-time Sit-to-Stand test (s)	40.7 ± 4.6	38.1 ± 6.3	0.19
6-minute walking distance (m)	193 ± 44	193 ± 22	0.96
Osteoporosis score	1.3 ± 0.6	1.6 ± 0.8	0.30
Bone mineral density (t-score)	-1.96 ± 0.52	-2.23 ± 0.45	0.13
Vitamin D concentration	26 ± 11	31 ± 12	0.26
Serum calcium concentration	56 ± 40	72 ± 46	0.30

Abbreviations: EQ-5D-5 L VAS, Score on the Visual Analog Scale of the EQ-5D-5 L questionnaire; VREM: Spanish reduced version of the Minnesota Leisure Time Physical Activity Questionnaire.

All the participants received the allocated intervention. A sample of patients dropped out of the trial for intercurrent diseases or missed the follow-up testing sessions. No variations in pharmacological therapy were reported during the trial.

Adherence to the intervention

All patients assigned to the exercise group performed the assigned walking and strength training program. A total of 8 out of 12 patients completed more than 60% of the prescribed sessions, with the remaining four exhibiting lower adherence. In this group, 3 people received concomitant physiotherapy treatments, but all of the patients missed the 3-month testing session.

In the control group, 5 people underwent physiotherapy treatment. Only two people in this group returned the completed diary, reporting an average of 15 min of walking per day, three times a week.

None of the patients randomized to the exercise group reported any falls or other adverse effects related to the training sessions.

Primary outcome

At the end of the exercise intervention, both groups significantly improved the BI (p < 0.001), with a significant between-group difference in favor of the exercise group (p = 0.046). The median improvement in the exercise group (+ 12 points) surpassed the minimal clinically important difference (MCID) reported for the BI, demonstrating that the observed change was not only statistically significant but also clinically meaningful. A large effect size was recorded in favor of the exercise group (Cohen’s d = 1.21).

At the 6-month follow-up, both groups maintained a significantly greater BI value with respect to baseline but failed to reach a statistically significant difference in favor of the exercise group (p = 0.087), the improvement in exercise group (+ 13 points) remained above the MCID threshold, and the effect size was large (Cohen’s d = 0.88). RM-ANOVA confirmed significant between-subject effects in favor of the exercise group (p = 0.004) but missed statistical significance for the group per time interaction (p = 0.082). The data are reported in Table 4 and Supplementary Table I.

Table 4.

Primary and secondary outcome measures in the two groups according to intention-to-treat analysis.−.

	Exercise group (n = 16)					Control group (n = 19)					Between groups p
	T0	T3	T6	ΔT3 − T0	ΔT6 − T0	T0	T3	T6	ΔT3 − T0	ΔT6 − T0	ΔT3 − T0	ΔT6 − T0
Barthel Index	71 (67 − 75)	83** (79 − 87)	84** (80 − 89)	12 (9 − 15)	13 (8 − 18)	66 (61 − 71)	73** (68 − 77)	75** (71 − 79)	7 (2 − 12)	9 (5 − 12)	0.046	0.087
EQ−5D−5L VAS	62 (56 − 67)	74** (67 − 82)	81** † (73 − 89)	13 (6 − 20)	19 (12 − 27)	64 (61 − 67)	67 (63 − 72)	67 (62 − 71)	3 (− 3 − 10)	3 (− 4 − 10)	0.012	< 0.001
VREM	124 (68 − 180)	402** (301 − 502)	572**†† (452 − 692)	277 (148 − 407)	448 (283 − 612)	106 (71 − 140)	250** (169 − 330)	400**† (311 − 488)	144 (66 − 222)	294 (192 − 396)	0.020	0.036
Cognition level	9.4 (6.5 − 12.2)	15.5** (13.1 − 17.8)	18.6**†† (16.4 − 20.8)	6.1 (2.8 − 9.4)	9.3 (5.7 − 12.8)	9.4 (6.7 − 12.2)	12.1** (9.3 − 14.9)	13.8**† (11.2 − 16.3)	2.7 (1.3 − 4)	4.4 (2.2 − 6.5)	0.010	0.003
Yesavage depression scale	3.8 (2.7 − 4.8)	2.6** (1.7 − 3.4)	1.4**† (0.7 − 2.1)	−1.2 (− 1.8− −0.6)	−2.4 (− 3.8− −1)	3.4 (2.8 − 4)	3.1 (2.4 − 3.7)	3.1 (2.4 − 3.8)	−0.4 (− 0.9 − 0.2)	−0.3 (− 0.9 − 0.3)	0.012	< 0.001
Falls Efficacy Scale	40.6 (36.4 − 44.9)	31.8** (28.2 − 35.3)	25.4**† (21.8 − 28.9)	−8.9 (− 13− −4.7)	−15.3 (− 21.4− −9.1)	46.0 (41.5 − 50.5)	44.2 (40.6 − 47.8)	42.4 (38.4 − 46.3)	−1.8 (− 5.9 − 2.3)	−3.6 (− 8.4 − 1.2)	0.003	< 0.001
5 − time Sit-to-Stand test (s)	40.7 (38.3 − 43.1)	25.8** (22.0 − 29.6)	21.2**†† (16.8 − 25.7)	−14.9 (− 19.1− −10.7)	−19.4 (− 24.3− −14.6)	38.2 (35.1 − 41.2)	31.9** (29.0 − 34.8)	28.6**†† (25.5 − 31.8)	−6.3 (− 9.4− −3.1)	−9.5 (− 13.7− −5.4)	< 0.001	< 0.001
6 − minute walking distance (m)	193 (170 − 217)	371** (329 − 413)	477**†† (420 − 533)	178 (127 − 228)	284 (214 − 353)	193 (182 − 203)	256** (230 − 283)	291**†† (257 − 324)	64 (36 − 91)	98 (61 − 135)	< 0.001	< 0.001
Osteoporosis questionnaire score	1.3 (1 − 1.6)	2.3** (1.9 − 2.7)	2.6** (2.4 − 2.9)	1 (0.5 − 1.5)	1.3 (0.8 − 1.8)	1.6 (1.2 − 2.0)	1.9** (1.5 − 2.4)	2.1** (1.6 − 2.6)	0.4 (0.1 − 0.7)	0.6 (0.1 − 1)	0.008	0.004
Bone mineral density (t−score)	−1.96 (− 2.24− −1.68)	−1.86* (− 2.14− −1.58)		0.11 (0.02 − 0.19)		−2.22 (− 2.43− −2.01)	−2.24 (− 2.44− −2.04)		−0.02 (− 0.06 − 0.01)		0.004
Vitamin D concentration (ng/ml)	26 (21 − 32)	35 (23 − 48)		9 (− 5 − 23)		30 (24 − 36)	29 (24 − 34)		−1 (− 6 − 3)		0.119
Calcium concentration (mg/dl)	56 (35 − 78)	82 (42 − 122)		26 (− 18 − 70)		71 (50 − 93)	75 (54 − 96)		4 (− 7 − 14)		0.271

Legend: data are expressed as mean (95% confidence interval).

* Indicates p value < 0.05 respect to T0; ** Indicates p value < 0.01 respect to T0; † Indicates p value < 0.05 respect to T3; †† Indicates p value < 0.01 respect to T3.

Abbreviations: EQ-5D-5 L VAS, Score on the Visual Analog Scale of the EQ-5D-5 L questionnaire; VREM: Spanish reduced version of the Minnesota Leisure Time Physical Activity Questionnaire.

Secondary outcomes

A total of 5 hospitalizations occurred in the enrolled population, with 4 patients hospitalized in the control group and one in the exercise group (p = 0.22).

At T3 in the exercise group, all secondary outcome measures significantly improved with respect to baseline, whereas in the control group, significant within-group differences were observed for VREM, cognitive capacity, the 6MWD, the 5STS and the osteoporosis score (Table 4).

In the between-group comparisons, a significant difference was noted for all the secondary outcomes in favor of the exercise group, with a large effect size observed for all the variables, particularly Cohen’s d values of 2.69 for the 6MWD; 2.11 for the 5STS; 1.7 for the FES; and 1.33 for the Yesavage scale.

At the 6-month follow-up, in the exercise group, significant improvements were recorded for all the outcomes except for the osteoporosis questionnaire score with respect to both T0 and T3. In the control group, significant differences between T6 and both T3 and T0 were noted for VREM, Lobo, 5STS and the 6MWD.

In the between-group comparisons, for all the secondary outcomes, significant differences were observed in favor of the exercise group (Table 4). RM-ANOVA confirmed a significant group per time interaction in favor of the exercise group for the VAS scale of the EQ-5D-5L questionnaire (p < 0.001), VREM (p = 0.028), cognitive capacity (p = 0.002), Yesavage (p < 0.001), FES (p < 0.001), 5STS (p < 0.001) and 6MWD (p < 0.001) scores.

Bone health biomarkers

At T3 with respect to baseline, bone mineral density significantly improved in the exercise group but not in the control group (p = 0.004). Although statistically significant, the increase should be interpreted as the least significant change (LSC) in DXA, as short-term variations may approach the densitometric precision error.

No significant variations were observed in the circulating vitamin D levels or calcium concentrations (Table 4).

Sensitivity analyses

The per-protocol analyses (Supplementary Tables A and B) confirmed the results obtained by the ITT, highlighting the same significant differences for the primary and secondary outcomes at both T3 and T6.

Discussion

In this randomized controlled trial including older adults with surgically treated fragility hip fractures, a motivational and progressive physical exercise intervention improved their degree of independence, mobility, quality of life and psychological status compared with a group of patients receiving usual care.

The effects of exercise in patients following hip fracture have been previously reported. After early experiences of daily exercise carried out within the subjects’ home environments^32,53, twenty years ago, a randomized trial demonstrated that a 6-month rehabilitation program including progressive resistance training conducted in an outpatient setting improved physical function, mobility and quality of life in this population⁵⁴.

The benefits of extended exercise programs on functional capacity after hip fracture were subsequently confirmed in a meta-analysis⁵⁵. More recently, a large randomized trial in patients after former hip fracture rehabilitation aimed to determine the efficacy of a 6-month home exercise program with minimal supervision by a physical therapist and performed independently at home⁵⁶. The program required minimal visits, and no patient travel or instrumentation, was safe and had good adherence (70% of prescribed training sessions). The study revealed a modest improvement in physical function⁵⁶ but contributed to the introduction of an important theme into the discussion. Indeed, alternative interventions to those usually based on close supervision and frequent visits, which are usually limited by staff availability and patient adherence, need to be identified to improve the function of patients with fractures.

In recent years, several trials and meta-analyses have been published, showing small-to-moderate positive effects at the end of intervention on mobility, activities of daily living, lower limb muscle strength and balance, which partly persist at follow-up after at least 1 year post fracture^30,56,57. Finally, the Cochrane review, which included 40 randomized controlled trials in hospital and posthospital settings (4059 participants, average age 80 years, 80% women), reported that, compared with conventional care, interventions aimed at improving mobility after hip fracture, particularly when multiple components are involved, may be associated with a clinically meaningful effect on mobility and walking speed²⁹.

The present study highlights the first scientific experience of non-supervised trial⁵⁴ focusing on the possibility of conducting a home-based program that enhances the effects of rehabilitation and stabilizes a better level of function of the person, reducing most barriers to exercise (e.g. the need for transportation, supervision, and costs). The study stems from a long history of successful administration of structured exercise at home in individuals with mobility limitations and reduced functional reserve^35–41 through programs based on the principles of a sustainable semipersonalized dose and of specificity and progressivity of load, empowered by experience derived from implementation in a real clinical environment. Although no formal economic evaluation was conducted, the intervention was designed as a low-resource, home-based program requiring minimal equipment and limited supervision, avoiding transportation and facility-based costs. Future studies should formally assess cost-effectiveness and scalability in real-world settings.

First, a significant variation was observed in the primary outcome, or the level of functional independence assessed by the BI, which showed an almost 2-fold greater increase in the exercise group with respect to its natural improvement after usual care in the control group. To the best of our knowledge, only a few studies have employed the BI as an outcome in this population, resulting in improvements³⁴, but the BI is considered a valid measure of independence⁵⁸, despite being limited by a possible ceiling effect in high-functioning people.

Second, interesting results were obtained in terms of HR-QoL, cognitive levels and depression with significantly greater variations in the exercise group. Although the most recent meta-analyses failed to demonstrate a significant effect of posthospital rehabilitation mobility strategies on HR-QoL^29,58, we observed a significant improvement in the exercise group only, which was greater than the minimal clinically important difference (set at 7.5) for a deconditioned population⁵⁹. This positive variation was confirmed and was accompanied by a significant reduction in depression levels in the exercise group only, both in the short term and in the long term. Interestingly, depression was found to be associated with an increased rate of bone loss at the hip in a cohort of older men⁶⁰, making the results observed in our trial important to for adding an additional protective factor against the risk of another fracture in this population. In addition, improvements in cognitive function are worthy of discussion, given that previous authors reported that people who suffer from fragility fractures presented a greater degree of cognitive impairment than their peers did⁶⁰ and that cognition can be positively improved by rehabilitation programs^61,62. Interestingly, in our trial, the cognitive level increased in both groups, but there was twofold greater variation in the intervention group following the simple home-based exercise program.

Third, significant variations in several parameters were observed, starting with a very important reduction in the fear of falling measured by the FES. Our findings confirm all the evidence already published in the literature, where prefrail and frail people and people with fragility fractures benefit from different forms of rehabilitation or exercise programs^63–65. This perceived reduced risk of falling may be attributable to increased lower limb strength, as observed through the 5STS test.

Indeed, the exercise program was combined by including not only walking sessions but also repeated sit-to-stand exercises, which were reported as determinants for increasing muscle strength and reducing lower body muscular demands in physically frail elderly individuals⁶⁶. Both groups significantly improved this outcome, and patients in the exercise group had significantly greater variation, as each sit-to-stand movement was lowered by approximately 3 s, which was consistent with values previously reported in women with frail vertebral fractures⁶⁷. Finally, variations in mobility measured through the 6MWD were collected, with values consistent with those reported in the most recent Cochrane review²⁹. Despite the fact that “naturally driven” recovery was also observed in the control group, which, after 6 months, increased the mean walking speed by approximately 1.0 km/h, twofold greater variations occurred in the exercise group, with a very large effect size, suggesting safer conditions, as a low 6MWD has been associated with fall risk and mortality in several populations^68–70.

Interestingly, the osteoporosis score and BMD improved after the 3-month program. Physical exercise improves and maintains bone mineral density, a crucial factor in assessing and managing patients with fragility fractures^71–77. It also decreases the likelihood of developing an osteoporotic fracture or being the victim of a fatal fall, leading to a fragility fracture^{71–76,78–80}.

Exercise also sends plasticity and remodeling signals to osteocytes in elderly individuals via intermittent dynamic stimuli²⁸ or, in rats, via short bouts of loading bouts separated by recovery periods⁸¹, such as those applied in the study. Furthermore, the safety of exercise has been reported^29,82 and tested positively in the present study through a patient-centered model, which facilitates comprehensive exercise routines at home, assessing the level of disability and adjusting exercise prescriptions accordingly together with motivational support. Indeed, one of the key aspects of the success of the present study was the feasibility of the training program for each patient. Considering that the population at entry walked slowly, covering approximately 200 m in the 6-minute test, the program added only 12–15 min of daily walking while respecting this pace in the initial phase so as not to evoke fatigue and possibly including rest breaks.

This study has several limitations. First, the small sample size, owing to the pilot design and strict criteria, limits the statistical power and generalizability, although the population is representative of stable post hip fracture patients. Despite nearly 20% attrition, losses were balanced and mostly due to clinical conditions rather than intervention effects. Analyses included intention-to-treat with imputation and per-protocol methods, but the results are exploratory and need confirmation in larger multicenter trials. Second, the majority of participants reported problems in the use of the fitness tracker that they received, and patients and caregivers reported that the daily number of steps was measured in other ways (e.g., with a smartphone application already present or by counting it). Third, surgical techniques and weight-bearing progression may influence early recovery; however, fracture type and baseline functions were similar between groups, and randomization likely reduced bias from surgical differences. Finally, the amount of eventual additional physical activity of the control group was not measured, even though the number of people who performed additional physiotherapy treatments was recorded.

In conclusion, while supervised exercise is generally regarded as ideal and more effective for adults than unsupervised programs are⁸³, the need to develop home exercise interventions that are effective for physical function and good adherence through the reduction of known barriers should be seriously considered. Our study revealed that the intervention was acceptable and tolerable, with all participants in the exercise group engaging, with more than 60% adherence in most participants, and no adverse exercise-related events or falls reported. These interventions that address health outcomes, such as those reported in this randomized trial, may improve early physical rehabilitation interventions, emphasizing a strategic approach to mitigate the challenges related to hip fragility fractures.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

RSR, NL, FM, PJLS were responsible for the conception and design of the study and data collection; MRC, AMA, RJE, EMG, PSL, PJA, CRC collected the data; RSR, GP, NL, FM were involved in the processing and statistical analysis of data; RSR, NL, FM, PJLS were involved in the drafting of the manuscript; and all authors contributed to the interpretation of the data for the work and revising it critically for important intellectual content. All the authors finally approved the manuscript. RSR and PJLS were responsible for obtaining project funding and takes responsibility for the integrity of the work as a whole. All authors have read and agreed to the published version of the manuscript.

Funding

This research received a specific grant from the Ministry of Health, Government of Andalusia (PIGE-0040-2020). The funders were not involved in the study design, data collection, data analysis, or preparation of this manuscript.

Data availability

The dataset analysed during the current study is available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

The study protocol was approved by the Research Ethics Committee of the province of Cordoba, Spain. Written informed consent was obtained from all participants.