Multidimensional frailty evaluation in elderly outpatients with chronic heart failure: A prospective study

Chronic heart failure (CHF) is frequently linked to clinical frailty, a complex geriatric condition associated with negative outcomes. However, the full characterization of frailty is still debated. Fried and colleagues focused on a ‘physical phenotype’ of frailty based on the assessment of unintentional weight loss, muscle weakness, slow walking speed, low physical activity, and exhaustion,¹ while Rockwood and colleagues fostered a multidimensional concept involving, besides physical impairments, comorbidity, disability, mental, nutritional, and social components (known as the ‘Canadian frailty index’, see supplemental material).²

We recently validated, in the Italian language, a modified version of the Canadian frailty index (mFI).³ Briefly, this tool has been adapted from the Canadian frailty index⁴ and explores the four domains of frailty: physical, mental, nutritional and socioeconomic. The mFI differs from the Canadian frailty index for the nutritional domain, assessed with the mini nutritional assessment, and the socioeconomic domain, assessed with the social support score. Frailty was stratified in three different degrees: light (0.1–16), moderate (16.1–27.0) and severe (>27.0).

In the present report, outpatients older than 65 years, who underwent comprehensive geriatric assessment (CGA), were assessed for frailty using the mFI and stratified for the presence and absence of CHF and by frailty degree. Mortality, disability (defined as ≥1 basic activity of daily living (BADL) lost from the baseline) and hospitalizations were assessed at 6, 12, 18 and 24 months of follow up.

Out of the 1077 study participants, 12 were excluded because they did not present any degree of frailty while 158 were lost to follow up. Thus, the final number was 907 participants. Demographic and clinical parameters are shown in Table 1. Patients with CHF were 20.9% of the sample and when the sample was stratified by frailty degree, we observed a significant increase of CHF with increasing frailty score (from 13.4% to 25.1%, p < 0.05 per trend). Comorbidity (assessed by the cumulative illness rating scale) and disability (assessed by the BADL) were higher in patients with CHF and increased more in the presence than in the absence of CHF, with increasing frailty. In addition, the severity of the New York Heart Association (NYHA) class, the lower ejection fraction (≤45%), glomerular filtration rate, brain natriuretic peptide and use of drugs are negatively affected by increasing levels of frailty. Inflammatory biomarkers, such C- reactive protein, progressively increased with frailty more markedly in patients with CHF than in those without CHF. Finally, the frailty-related increase of mortality, disability, and hospitalization was more evident in patients with CHF than those without CHF. Cox analysis, corrected for age and sex, showed that for each mFI unit increase, the relative risk (RR) of mortality increases more in the presence of CHF than in the absence of CHF (Figure 1). Accordingly, disability and hospitalization were higher in the presence of CHF than in the absence of CHF (relative risk (RR) = 1.06, 95% confidence interval (CI) 1.02–1.09 versus 1.02, 95% CI 1.01–1.07 and 1.14, 95% CI 1.05–1.18 versus 1.03, 95% CI 1.01–1.05, respectively).

Figure 1.

Cox multivariate analysis on mortality stratified for the presence and absence of CHF and for frailty degree, quantified by the mFI.

CHF: chronic heart failure; CI: confidence interval; mFI: modified frailty index; RR: relative risk.

Open in new tabDownload slide

There is a growing interest in the role of frailty in cardiovascular patients, especially in those affected by CHF.⁵ Two main models of frailty have been advanced: the phenotypic model or deficits accumulation model (i.e. clinical frailty), and different instruments have been proposed and validated to measure frailty.⁶ For instance, clinical frailty was independently associated with mortality in elderly patients with non-ST elevated myocardial infarction.⁷ However, prognostic relevance of frailty in CHF in elderly patients is somewhat limited and difficult to interpret. It is conceivable that the real misinterpretation of frailty data stems from the type of frailty assessed. A recent meta-analysis reported that the prevalence of a frail-phenotype in CHF patients ranged from 19% to 77%, while multidimensional frailty ranged from 15% to 89%.⁸ This significant heterogeneity of the results implies a lack of specificity in the assessment. As underlined before, the most relevant problem is a reasonably accurate definition of frailty. The ‘frail phenotype’ has the advantage of being relatively easy to identify but also the disadvantage of identifying a frailty that favours only functional impairments (i.e. ‘physical’ frailty).¹ In addition, when considering ‘physical’ frailty in CHF patients, frailty might be often blurred by CHF-related signs and symptoms.^5,9 In contrast, a ‘multidimensional’ frailty involving not only physical and mental but also nutritional and psychosocial components could be more accurate for clinical frailty identification.¹⁰ In this context, mFI, which includes all domains of multidimensional assessment, seems to provide a good prognostic tool as compared with identification of a ‘frail phenotype’.³

Our study shows that clinical frailty, detected and quantified by mFI, is highly prevalent in elderly patients with CHF. More importantly, mortality, disability and hospitalization rates are significantly predicted by mFI more in the presence of CHF than in the absence of CHF. These results encourage identifying and, more importantly, quantifying the condition of ‘multidimensional’ frailty in the elderly with CHF. In this regard, the mFI seems to be an appropriate tool.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References