Relationship Between Frailty and Risk of Falls Among Hospitalised Older People with Cardiac Conditions: An Observational Cohort Study

1. Introduction

Frailty syndrome, cardiovascular diseases (CVDs), and falls are three of the most prevalent health issues among older adults [1], with significantly different outcomes between sexes [2,3,4].

The ageing process begins at 60 years of age, a stage marked by declining physical condition and an increased risk of disease. Globally, the number of people aged 60 and above reached 1 billion in 2019 and is projected to rise to 2.5 billion by 2050 [5].

CVDs are one of the most common conditions worldwide during this period, directly linked to morbidity, disability, and mortality [6]. The risk of CVD is particularly elevated among women due to the physiological changes associated with menopause [7, 8].

In later life, an imbalance in homeostasis can lead to the emergence of frailty syndrome, first defined by Linda Fried et al. as a physical phenotype in 2001. Various tools for identifying frailty exist in the literature, with Fried’s five-criteria scale being among the most widely used [9].

Frailty has been associated with adverse outcomes such as increased risk of hospitalisation, falls, and mortality, all of which lead to higher healthcare costs [10].

Ongoing research into defining frailty and developing tools to address its multifactorial nature reveals varying prevalence rates depending on the study setting. Doody et al. (2023) reported a frailty prevalence of 10.7% among community-dwelling older adults, compared to 47.4% among hospitalised older patients [1]. Despite having a longer life expectancy, women are more frequently affected by frailty [2], with a prevalence of 9.6% compared to 5.4% in men [11].

Xu, Ou, and Li provided evidence that CVD, frailty, and hypertension increase the risk of falls [12]. Similarly, a study in the United Kingdom found that pre-frail and frail patients had a 47% and 15% higher risk of developing CVD, respectively, compared to non-frail patients [13]. Other researchers have linked CVD to an elevated risk of mortality (relative risk [RR]=1.81; 95% CI:1.67–1.97) among fall-related incidents, with age being a contributing factor [14,15].

Falls compromise personal safety and are more common among older adults [16], with a higher prevalence among women [17]. In hospital settings, falls are the most frequent adverse event, with an incidence of 3–5 falls per 1,000 patient-days and an estimated annual occurrence of 700,000 to 1 million cases [18, 19].

Preventing falls has been a key area of study. Multifactorial programmes such as 6-PACK and Best Practice Spotlight Organisations (BPSO®) include guidelines like the Prevention of Falls and Fall Injuries Best Practice Guideline [20, 21]. A common recommendation across these programmes is to identify patients at risk of falling so as to implement targeted interventions [22, 23].

The interplay between CVD, frailty, and falls poses a significant challenge to therapeutic actions for these patients. These challenges are further influenced by sex differences and the traditional threshold for ageing set at 65 years, even though physiological changes begin at 60 years. Addressing these factors could have a profound impact on population health and the economic resources of healthcare systems. For these reasons, this study aimed to analyse the relationship between frailty, risk of falls, and length of hospital stay in hospitalised older adults with cardiac conditions.

2. Materials and Methods

2.1. Study Design, Setting, and Sampling

An observational cohort study was conducted in a tertiary hospital within the public network of the Spanish National Health System. Recruitment took place between March 2022 and September 2024, involving patients admitted to a conventional cardiology hospitalisation unit. Individuals aged ≥60 years who were conscious and oriented were included. The study is reported according to STROBE reporting guidelines for observational research.

The study participants were selected using a non-probabilistic convenience sampling method in the cardiology unit.

2.2. Instruments

2.2.1. Fall Risk Assessment

Fall risk screening was conducted using the BPSO® programme, introduced in Spain in 2010 through the Carlos III Health Institute under the Ministry of Science, Innovation, and Universities. The programme was implemented in the study hospital in 2018, with the Prevention of Falls and Fall Injuries guideline being adopted in 2019. Fall risk was evaluated using the J.H. Downton scale, which assesses five dimensions: previous falls, medication, sensory deficits, mental state, and gait [24]. The J.H. Downton scale has a sensitivity of 0.58 and specificity of 0.63. Scores were interpreted as follows: 0 points = no risk, 1–2 points = low/moderate risk, and ≥3 points = high risk [25].

2.2.2. Frailty Identification

Patients classified as frail were identified using Fried’s five criteria [9]:

1. Unintentional weight loss >4.5 kg or >5.0% (in the past year).

2. Generalised exhaustion (low energy and endurance measured by the CES-D depression scale [26]).

3. Weakness (handgrip strength adjusted for sex and body mass index using a hand dynamometer).

4. Walking speed (time to cover 4.57 m adjusted for sex and height).

5. Weekly physical activity (Minnesota Leisure Time Activity Questionnaire [MLTAQ] stratified by sex; men: 383 kcal/week, women: 270 kcal/week [27]).

Patients were classed based on the number of criteria they met:

• Frail: three or more criteria.
• Pre-frail: one or two criteria.
• Non-frail: no criteria.

2.2.3. Functional Dependency

Functional dependency was assessed using the Barthel Index, which has a Cronbach’s alpha of 0.70 [28]. Scores were categorised as:

• 0-20: total dependence
• 21-60: severe dependence
• 61-90: moderate dependence
• 91-99: minimal dependence
• 100: independence

2.3. Description of Variables

• Sociodemographic variables: sex (biological attribute); age.
• Dependent variable: fall risk (yes/no); score on the J.H. Downton scale.
• Independent variables:

1. Frailty status.

2. Clinical variables: main diagnosis (coronary heart disease; infectious endocarditis; heart failure; arrhythmias; heart transplant; and valvular diseases); diabetes mellitus (yes/no); readmission to the cardiology unit (yes/no); presence of dyspnoea at admission (yes/no); presence of chest pain at admission (yes/no); systolic blood pressure (SBP; mmHg); diastolic blood pressure (DBP; mmHg); heart rate (HR; beats per minute); oxygen saturation (SpO₂; %); and blood glucose (mg/dL).

3. Anthropometric variables: height (cm); weight (kg); waist circumference (cm); body mass index (BMI; kg/m²: overweight (≥25 kg/m²), obesity (≥30 kg/m²)).

4. Length of stay: number of hospitalisation days.

2.4. Procedure

Data collection was conducted within the first three days of hospital admission. Patients meeting the inclusion criteria were informed about the study, and after obtaining their informed consent, the following data were extracted from their electronic health records: sex, age, primary diagnosis, diabetes mellitus status, readmission to the cardiology unit, presence of dyspnoea, SBP, DBP, HR, SpO₂, and blood glucose levels. At admission, the degree of dependency was systematically assessed using the Barthel index, and fall risk was screened using the J.H. Downton scale, with scores of 1 or higher considered positive.

Anthropometric measures such as height, weight, and BMI were recorded, and Fried’s scale was applied to identify frail patients. The length of stay in days was documented for all participants upon hospital discharge.

2.5. Ethical Considerations

The anonymity and confidentiality of participant data were safeguarded using Research Electronic Data Capture (REDCap) software, in compliance with the European Regulation 2016/679 of the European Parliament and of the Council of 27 April 2016 and Spanish Organic Law 3/2018, of 5 December, on the Protection of Personal Data and the Guarantee of Digital Rights. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the relevant Valladolid Ethics Committee for Research with medicinal products (ECRmp) involving humans under the reference code PI-20-1612 on 23 January 2020.

2.5. Statistical Analysis

Data normality was assessed using the Kolmogorov-Smirnov test, kurtosis, and skewness. Quantitative variables were presented as means and standard deviations (SD), while categorical variables were expressed as frequency distributions. Quantitative variables were compared using Student’s t-test, categorical variables with Pearson’s chi-squared test, and comparisons among three groups using the Kruskal-Wallis test. Homogeneity of variance was evaluated using Levene’s test.

Univariate binary logistic regression was employed to identify factors associated with frailty and fall risk, with adjusted odds ratios (OR) calculated for the included variables. A confidence interval of 95% and a statistical significance level of p<0.05 were applied.

Statistical analyses were performed using IBM® SPSS® Statistics v.26 software (SPSS, Inc., Chicago, IL, USA).

3. Results

3.1. Descriptive Analysis

During the study period, 144 patients agreed to participate (33.30% women, n=48; 66.70% men, n=96) with a mean age of 73.08 years (SD=7.95). The most prevalent diagnosis was coronary heart disease (61.11%). Diabetes mellitus was present in 36.11% of patients, while 35.42% reported chest pain at admission, and 30.56% presented with dyspnoea.

Regarding frailty, 33.30% of patients were classed as frail. Fall risk was identified in 97.22% of participants, with 36.81% of the total sample considered at high or very high risk (Table 1).

The mean Downton score indicated a low/moderate fall risk (2.24 points; SD=1.25). The Barthel Index showed a mean score of 95.76 points (SD=9.28), reflecting a minimal level of dependency in most patients. The mean BMI fell within the overweight category ($\overline{\mathrm{X}}$=27.69 kg/m²; SD=4.35), and the mean waist circumference exceeded 100 cm. The mean overall hospital stay was 8.66 days (SD=5.56), while patients at high fall risk experienced a longer mean stay of 9.71 days (SD=7.39) (Table 2).

After analysing the data by sex, the mean age was similar between men and women, as was the Downton score. Women were proportionally frailer and had a lower mean score on the Barthel index (93.54 points; SD=12.80). There were no differences in mean BMI (men: $\overline{\mathrm{X}}$=27.71 kg/m²; SD=4.09; women: $\overline{\mathrm{X}}$=27.67 kg/m²; SD=4.89). However, the mean waist circumference was larger in men than in women. The mean length of hospital stay was longer among men ($\overline{\mathrm{X}}$=9.40 days; SD=6.13) compared to women ($\overline{\mathrm{X}}$=7.11 days; SD=3.78). When analysed by frailty status, frail men had the longest mean hospital stays (Table 2 and Table 3).

The most prevalent diagnosis was coronary heart disease. The proportion of patients with diabetes was higher among men (39.58%). Fall risk was greater among frail and pre-frail women compared to other groups (37.50% and 43.80%, respectively) (Table 4).

When analysing frailty distribution within the high/very high fall risk group, 45.30% of the patients were classified as frail. Women in this group exhibited a higher percentage of pre-frailty compared to men. This group of patients had a mean hospital stay of 12.06 days (SD=9.05) (Table 5).

According to the J.H. Downton scale, the most frequently used category of medications among patients was other medications (27.10%), followed by non-diuretic antihypertensives taken in combination with other drugs (16.00%). Notably, 99.30% of the patients were taking at least one medication during the assessment (Table 6).

3.2. Comparison Between Frailty Status and Demographic and Clinical Variables

A comparative analysis between categorical variables and frailty status revealed statistically significant differences in gait (assessed within the Downton scale) (χ2=25.14; df=10; p<0.00); social risk (χ2=8.88; df=2; p<0.01), and gender (χ2= 5.99; df=2; p<0.05). However, no significant differences were found with risk of falls (χ2=0.62; df=2; p=0.73).

Regarding quantitative variables, significant differences were observed between non-frail and frail patients in diastolic blood pressure (t=2.19; df=92; p=0.03); waist circumference (t=-2.22; df=91; p=0.02); length of hospital stay (t=-2.74; df=92; p=0.00), and height (t=2.96; df=92; p<0.00).

Among women, a significant association was observed between frailty status (frail and non-frail) and diastolic blood pressure (t=-2.06; df=25; p=0.05). In men, waist circumference was significantly associated with frailty status (frail vs. non-frail) (t=-2.61; df=65; p=0.01).

In the analysis of the three frailty groups (frail, pre-frail, and non-frail), statistically significant differences were found in the following variables: Downton score (H=7.73; df=2; p=0.02); Barthel index (H=25.68; df=2; p=0.00); length of hospital stay (H=6.97; df=2; p=0.03); waist circumference (H=6.78; df=2; p=0.03); age (H=19.75; df=2; p=0.00), and height (H=6.75; df=2; p=0.03) (Table 7).

Among women, a significant association was observed between frailty status and the Barthel index in the Pre-frail–Frail group (H=7.71; p=0.04). For men, a significant association was found in the Frail–Non-frail group (H=14.00; p<0.00).

3.3. Logistic Regression Analysis

No significant association was found between risk of falls and frailty status (OR: 0.959; 95% CI: 0.058-15.752; p=0.977). However, waist circumference was significantly associated with risk of falls (OR: 1.063; 95% CI: 1.003-1.125; p=0.038).

Frailty status was directly related to Downton scores, age, Barthel Index, oxygen saturation (SpO₂), diastolic blood pressure, and waist circumference. No statistically significant relationship was observed with sex (Table 8).

When examining the relationship between frailty and specific items from the Downton scale, the use of diuretics, sensory deficits in the extremities, and gait were significantly associated with frailty (Table 9).

A statistically significant relationship was observed between coronary heart disease and frailty status (Table 10).

4. Discussion

Scientific cardiology societies should prioritise including frail patients with cardiovascular disease (CVD) in their research projects in order to optimise their clinical management, particularly considering the existing link between frailty and risk of falls [29, 30].

Our study identified a relationship between frailty status and scores on the J.H. Downton scale, with frail hospitalised cardiac patients scoring higher and thus facing a greater risk of falls. However, no significant association was found between frailty status and binary fall risk (Yes/No), as determined by the same scale. This finding aligns with a study conducted on a rural Ecuadorian population aged ≥60 years (OR: 0.63; 95% CI: 0.29–1.36; p= 0.237), but contrasts with Yang et al.’s meta-analysis (RR: 1.48; 95% CI: 1.27–1.73; p<0.05) [14, 31]. These differences could be attributed to the use of varied tools for assessing both frailty and fall risk.

The limited predictive value of the J.H. Downton scale has been noted by other researchers, who have emphasised the critical role of nurses’ clinical judgement. In hospital settings, the Downton scale may overidentify fall risk, as the threshold begins at just 1 point. Despite its widespread use, pairing it with another tool that captures the multifactorial nature of fall risk is recommended [25, 32, 33].

Although no statistically significant relationship was found between frailty and fall risk by sex, differences were observed in related factors such as blood pressure and waist circumference. Previous studies have linked frailty in women to higher mortality (hazard ratio [HR]: 0.65; 95% CI: 0.58–0.74; p<0.001) [34].

Larger waist circumference in this study was associated with increased fall risk, consistent with Bu et al.’s findings (OR: 0.98; 95% CI: 0.96–1.00; p= 0.017) [35]. Polypharmacy is also a well-documented fall risk factor in CVD patients. In this study, frail patients were more likely to use diuretics than their non-frail counterparts. Diuretics are among the drugs most commonly linked to fall risk in this population, as highlighted by the American Heart Association [29].

The prevalence of frailty in our study was lower than reported in other non-community settings but higher than in community settings [1, 36]. Differences in frailty assessment tools, population groups, and identification criteria continue to hinder the development of standardised protocols [37].

Among the CVD diagnoses included in the study, frailty was only significantly associated with coronary heart disease. This contrasts with findings from Wang, Hu, and Wu, who found no overall relationship between frailty and CVD (RR: 1.00; 95% CI: 0.92–1.09; p=0.95) [38]. Liperoti et al. estimated that nearly one-fifth of patients with ischaemic heart disease were frail but did not establish a clear link [39]. While frailty is often associated with heart failure [40], this was not statistically significant in our study. Diastolic blood pressure was inversely related to frailty, suggesting that frail patients tended to have lower diastolic pressures. This finding aligns with previous studies [41], although Vetrano et al. reported a relationship between frailty and hypertension (pooled OR: 1.33; 95% CI: 0.94–1.89; I2=79.2%) [42]. Age also emerged as a significant factor, with increasing age linked to higher frailty risk, consistent with prior studies ([OR: 1.15; 95% CI: 1.09–1.21; p<0.001]; [OR: 3.910; 95% CI: 2.021–5.402; p<0.05]) [31, 41].

Frail patients also experienced longer hospital stays, with men staying the longest. When combined with fall risk, hospital stays extended to a mean of 12.06 days, likely increasing institutional costs [10].

4.1. Limitations

The main limitation of this study was the sample size, which was selected through convenience sampling. A larger sample would have allowed for greater generalisability of results. However, the study objectives were successfully addressed. Future studies should examine factors influencing the relationship between frailty and fall risk for each clinical diagnosis and include a wider range of hospital settings.

Future lines of research

Larger, more diverse samples are essential to generalise findings and uncover the complex relationship between frailty, risk of falls, and cardiovascular conditions. Developing multifactorial models that combine clinical and demographic variables could transform personalised care for these patients.

5. Conclusions

Frailty was directly associated with Downton scores in hospitalised older adults patients with cardiac conditions. In this study, 33.30% of patients were classed as frail, and 97.22% were at risk of falls. Women at high fall risk were proportionally frailer than men in the same situation. Frail patients also had longer hospital stays than their non-frail counterparts.

As the older population grows worldwide, the number of individuals with CVD who are frail and at risk of falls is expected to increase. Identifying shared characteristics within this population can help standardise assessment tools and improve clinical management, leading to better outcomes for patients and institutions alike. Combining frailty and fall risk scales in hospitalised older adults with cardiac conditions can refine prognostic accuracy, address comorbidities, and reduce hospital stays. Tailoring interventions to the differences between men and women is equally important for effective management.