European Journal of Clinical Investigation (2005) 35, 723–730 © 2005 Blackwell Publishing Ltd Blackwell Publishing, Ltd. Frailty predicts long-term mortality in elderly subjects with chronic heart failure F. Cacciatore * , P. Abete † , F. Mazzella † , L. Viati * , D. Della Morte † , D. D’Ambrosio † , G. Gargiulo † , G. Testa † , D. De Santis † , G. Galizia † , N. Ferrara *‡ and F. Rengo *† * ‘Salvatore Maugeri Foundation’, Institute of Care and Scientific Research, Telese (BN), † Department of Clinical Medicine, Cardiovascular and Immunological Sciences, Division of Genetics, University of Naples ‘Federico II’ and ‡ Department of Animal, Vegetable and Environmental Sciences, Division of Internal Medicine, University of Molise, Italy Abstract Background The elderly are characterized by a high prevalence of chronic heart failure (CHF) and frailty, which is a complex interaction of physical, psychological and social impairment. This study aimed to examine the predictive role of frailty on long-term mortality in elderly subjects with CHF. Materials and methods The study assessed long-term mortality after 12-year follow up in 120 subjects with CHF and 1139 subjects without CHF, selected in 1992, from a random sample of the elderly population in the Campania region of Italy. Frailty was assessed according to a ‘Frailty Staging System’. Results Subjects with CHF were prevalently female (60%) and older than 75 years (mean 75·9 ± 6·7); subjects without CHF were prevalently female (56·4%) and younger than 75 years (mean 74·0 ± 6·3). In subjects with and without CHF stratified into classes of frailty there was a statistically significant increase in age, comorbidity, disability and low social support, and a decrease in MMSE score. Moreover, death progressively increased more with frailty in subjects (70·0% to 94·4%, P < 0·03) than in those without (43·8.% to 88·3%, P < 0·0001) CHF.The Kaplan–Meier analysis shows that at 9 years the probability of survival progressively decreased as frailty increased (45·5% to 0%) in subjects with CHF and from 62·8% to 25·9% in subjects without CHF. The Cox regression analysis indicated that frailty is predictive of mortality in the multivariate model adjusted for several variables including sex and age in subjects with and without CHF. Moreover, the analysis showed that frailty is more predictive of mortality in elderly subjects with CHF when it was analyzed either as continuous (1·48 vs. 1·36) or as a dummy (3 vs. 1 = 1·62 vs. 1·24) variable. Conclusions Thus mortality among elderly subjects with or without CHF increases with frailty. Moreover, frailty is more predictive of long-term mortality in elderly subjects with than in those without CHF. Hence, frailty represents a new independent variable for predicting long-term mortality in elderly subjects with CHF. Keywords Elderly, frailty, heart failure, mortality. Eur J Clin Invest 2005; 35 (12): 723–730 ‘Salvatore Maugeri Foundation’, Institute of Care and Scientific Research, Telese (BN) (F. Cacciatore, L. Viati, N. Ferrara, F. Rengo); Department of Clinical Medicine, Cardiovascular and Immunological Sciences, Division of Geriatrics, University of Naples ‘Federico II’ (P. Abete, F. Mazzella, D. Della Morte, D. D’Ambrosio, G. Gargiulo, G. Testa, D. De Santis, G. Galizia, F. Rengo); Cattedra di Medicina Interna, Department of Animal, Vegetable and Environmental Sciences, University of Molise (N. Ferrara), Italy. Correspondence to: Pasquale Abete, MD, PhD, Department of Clinical Medicine, Cardiovascular and Immunological Sciences, Division of Geriatrics, University of Naples ‘Federico II’, Via S. Pansini, n°5, 80136 Napoli, Italy. Tel.: 0039 81 7462270; fax: 0039 81 7462270; e-mail: p.abete@unina.it Received 30 May 2005; accepted 26 September 2005