Frailty commonly coexists with heart failure and although both have been associated with neurohormonal dysregulation, inflammation, catabolism, and skeletal muscle dysfunction, there are still no defined biomarkers to assess frailty, especially from the perspective of populations with cardiovascular diseases. This is a cross-sectional study with 106 outpatients with heart failure, aged ≥60 years, which aimed to assess frailty through a physical (frailty phenotype) and multidimensional (Tilburg Frailty Indicator) approach and to analyze its association with inflammatory and humoral biomarkers (high sensitivity C-reactive protein [hs-CRP], interleukin 6, tumor necrosis factor-α, insulin-like growth factor-1, and total testosterone), clinical characteristics, and functional capacity. In univariate analysis, hs-CRP was associated with frailty in both phenotype and Tilburg Frailty Indicator assessment (PR = 1.005, 95% confidence interval [CI] 1.001 to 1.009, p = 0.027 and PR = 1.015, 95% CI 1.006 to 1.024, p = 0.001, respectively), which remained significant in the final multivariate model in the frailty assessment by the phenotype (PR = 1.004, 95% CI 1.001 to 1.008, p = 0.025). There was no statistically significant difference between the groups for other biomarkers analyzed. Frailty was also associated with worse functional capacity, nonoptimized pharmacological treatment and a greater number of drugs in use, age, female gender, and a greater number of comorbidities. In conclusion, frailty is associated with higher levels of hs-CRP, which can indicate it is a promising frailty biomarker.
Introduction
Because of aging and the increasingly complex nature of the comorbidities that affect patients with heart failure (HF), frailty has emerged as a significant research area in HF. , Frailty is a biological syndrome, defined as a state of greater vulnerability to endogenous and exogenous stress factors, resulting from the decrease in physiological reserves and the dysfunction and dysregulation of multiple systems, which interfere with homeostasis and the stress response. , Symptoms and aspects related to HF considerably overlap the manifestations of frailty and both syndromes have been associated with neurohormonal dysregulation, inflammation, catabolism, and skeletal muscle disorders. , The bidirectional routes between both syndromes, however, remain not fully understood, and biomarkers can help to better determine this relation. Thus, this study aimed to assess frailty in a physical approach (frailty phenotype) and multidimensional (Tilburg Frailty Indicator [TFI]) in patients with HF and to analyze its association with inflammatory and humoral biomarkers.
Methods
This is a cross-sectional study, with subjects aged ≥60 years, diagnosed with HF confirmed by 2-dimensional echocardiography, of all functional classes of the New York Heart Association (NYHA), screened and recruited consecutively in outpatient care from a tertiary hospital in southern Brazil.
Subjects with the following conditions were excluded: (1) serum creatinine ≥2.0 mg/100 ml; (2) previous heart transplantation (3) decompensated HF; (4) symptomatic rheumatoid arthritis or other inflammatory condition; (5) acute myocardial infarction, stroke or surgery <3 months before the participation in the research; (6) peripheral congestion and/or edema; (7) history of unstable angina; (8) malignant tumors (active or in remission) <5 years; (9) acute infection; (10) previous diagnosis of neurodegenerative disease that prevented from carrying out the questionnaires reliably (such as dementia or Alzheimer); (11) unfeasibility of performing functional tests (wheelchair users, amputees, with motor sequels or disabilities); (12) using anti-inflammatory drugs; and (13) contraindications in performing the electrical bioimpedance analysis with pacemakers and implantable cardioverter-defibrillator, metallic prosthesis and body mass index (BMI) >39 kg/m². These subjects were excluded once this is a study focused on a project that evaluated quantity/muscle quality by electrical bioimpedance analysis.
Sociodemographic and clinical information, including variables used to calculate the MAGGIC (The Meta-Analysis Global Group in Chronic Heart Failure) mortality risk score, were collected from electronic medical records and checked during the research consultation. For the left ventricular ejection fraction (LVEF) data (2-dimensional echocardiography), patients with LVEF ≥50% were considered as having preserved LVEF. Anthropometrics were made to calculate BMI (weight [kilograms] divided by height [meters] squared), and functional capacity was assessed by the 6-minute walk test according to a standardized protocol. Of note, <300 m of total distance was characterized as low functional capacity.
Frailty was assessed using the frailty phenotype proposed by Fried et al and defined by the following criteria: (1) weight loss defined as an unintentional body weight loss ≥4.5 kg in the last year ; (2) exhaustion assessed by items 7 and 20 of the Center for Epidemiological Studies Depression Scale , (answers “always” or “frequently” to either of the 2 questions were scored as a criterion) ; (3) walk time defined as the time it would take the participant to walk the distance of 4.6 m at the usual speed. Values below the cut-off point, stratified by gender and height, scored for the criterion ; (4) physical activity assessed using the short version of the International Physical Activity Questionnaire, from which the weekly expenditure of energy in kilocalories was derived from the self-report of the activities performed. , Values below the cut-off point, stratified by gender, scored for the criterion ; (5) grip strength measured using the Jamar mechanical dynamometer (Sammons Preston Rolyan, Bolingbrook, Illinois), properly calibrated, according to protocols. , The test was repeated 3 times using the dominant hand, and we used the highest value of the 3 measurements. Values below the cut-off point, stratified by gender and quartiles of BMI, were scored for the criterion. In the end, subjects who scored for 3 or more criteria were considered frail, those who scored for 1 or 2 criteria were characterized as pre-frail and those who did not score for any of the criteria were classified as not frail.
Also, frailty was assessed by the TFI, using the validated Brazilian version, composed of 15 self-reported questions, divided into 3 domains (physical, psychological and social domains). For each item, scores equal to 0 or 1 were assigned, according to the question. The cut-off point for frailty was defined as ≥5 points.
Blood samples were collected from the antecubital vein and analyzed using standardized hospital protocols: high sensitivity C-reactive protein (hs-CRP) by immunoturbidimetry analysis, insulin-like growth factor-1 (IGF-1) by chemiluminescence and total testosterone by electrochemiluminescence by competition. For the interleukin 6 (IL-6) and tumor necrosis factor-α (TNF-α) analyses, blood samples were centrifuged at 4°C, 2,500 revolutions per minute, for 15 minutes to extract the serum which was stored at −80°C. For the analysis, we used ProcartaPlex 2-plex Human Custom High-Sensitivity and ProcartaPlex Simplex Human Custom High-Sensitivity immunoassay kits (Thermo Fisher Scientific [Vienna, Austria)] catalog number PPXS-02-MXPRKP3 and EPXS010-10213-901, respectively), according to the manufacturer’s instructions. For the present analysis, the samples went through a single defrosting process and the remaining unused were discarded.
This study was registered and approved by the Ethics and Research Committee of the institution (protocol number 2018-0683) and was conducted following the principles of the Declaration of Helsinki having all the subjects sign the informed consent form.
The sample size was calculated using the WINPEPI (PEPI-for-Windows, programs for epidemiologists) program version 11.43 and based on literature. , A minimum total of 106 patients was obtained considering 40% of prevalence of frailty in patients with HF, a significance level of 5%, power of 80% between groups regarding biomarkers, and the largest sample number in biomarkers.
Categorical variables were described using absolute and relative frequency, and associations were tested using Pearson’s chi-square test or Fisher’s exact test. In the case of statistical significance, the adjusted residual analysis was used to locate the associations. Continuous variables were described by means and standard deviation or median and 25th and 75th percentiles, as appropriate (Shapiro-Wilk test), and compared through the student’s t test, analysis of variance (post-Tukey Hoc), Mann-Whitney, or Kruskal-Wallis tests. To test the degree of relation between frailty scores and quantitative variables of interest, Spearman’s correlation coefficient was used.
To control the confounding factors, we used Poisson regression analysis with a robust estimator (not frail vs pre-frail/frail in the frailty assessment by the phenotype, and not frail vs frail in the frailty assessment by TFI). Variables with p value <0.1 in the univariate analysis were included in the multivariate model. The Kappa coefficient was used to evaluate the agreement between the methods for frailty assessment.
The level of significance adopted was 5% (p ≤0.05) and the analyses were performed using the IBM SPSS Statistics for Windows, version 21.0 (IBM Corp, Armonk, New York).
Results
Demographic, clinical characteristics and their association with the frail state are described in Table 1 . Concerning frailty according to the phenotype, 28% were classified as frail ( Figure 1 ), showing a higher frailty prevalence in women than men (31% and 27%, respectively). Through the TFI, 47% were classified as frail ( Figure 1 ), also showing a higher frailty prevalence in women to men (71% and 35%, respectively). There was no agreement between the methods for assessing frailty (Kappa = 0.007; p = 0.910).
| Variables | Total (n = 106) | Frailty Phenotype (n=106) | Tilburg Frailty Indicator (n = 102)* | |||||
|---|---|---|---|---|---|---|---|---|
| Not Frail (n = 12) | Pre-Frail (n = 64) | Frail (n = 30) | p | Not Frail (n = 54) | Frail (n = 48) | p | ||
| Age (years) | 68 (63,0-74,0) | 66,0 (62,25-71,25) | 67,0 (63,25-73,0) | 70,0 (63,0-75,0) | 0,264 | 67 (63,0-72,0) | 69,50 (64,0-75,0) | 0,058 |
| Female | 35 (33%) | 5 (42%) | 19 (30%) | 11 (37%) | 0,635 | 10 (18%) | 24 (50%)† | 0,001 |
| Male | 71 (67%) | 7 (58%) | 45 (70%) | 19 (63%) | 44 (82%)† | 24 (50%) | ||
| Whites | 81 (76%) | 7 (58%) | 47 (73%) | 27 (90%) | 0,062 | 39 (72%) | 39 (81%) | 0,283 |
| Not Whites | 25 (24%) | 5 (42%) | 17 (27%) | 3 (10%) | 15 (28%) | 9 (19%) | ||
| Years of education | 0,562 | 0,083 | ||||||
| 0-3 | 55 (55%) | 7 (64%) | 33 (54%) | 15 (54%) | 27 (53%) | 26 (58%) | ||
| 4-8 | 17 (17%) | 2 (18%) | 8 (13%) | 7 (25%) | 6 (12%) | 11 (24%) | ||
| ≥9 | 28 (28%) | 2 (18%) | 20 (33%) | 6 (21%) | 18 (35%) | 8 (18%) | ||
| BMI (kg/m 2 ) | 27,10±4,47 | 28,33±3,87 | 27,31±4,48 | 26,15±4,62 | 0,304 | 26,70±4,18 | 27,45±4,90 | 0,405 |
| Etiology of HF | 0,909 | 0,561 | ||||||
| Ischemic | 30 (28%) | 4 (33%) | 18 (28%) | 8 (27%) | 13 (24%) | 14 (29%) | ||
| Nonischemic | 76 (72%) | 8 (67%) | 46 (72%) | 22 (73%) | 41 (76%) | 34 (71%) | ||
| LVEF (%) | 34,56±11,87 | 32,83±11,49 | 34,62±12,59 | 34,47±10,69 | 0,155 | 30,0 (23,0-42,0) | 35,0 (27,0-42,50) | 0,184 |
| LVEF | 0,883 | 1,000 | ||||||
| HFrEF | 96 (91%) | 11 (91%) | 57 (89%) | 28 (93%) | 49 (91%) | 44 (92%) | ||
| HFpEF | 10 (9%) | 1 (8%) | 7 (11%) | 2 (7%) | 5 (9%) | 4 (8%) | ||
| Functional class | 0,003 | 0,303 | ||||||
| NYHA I/II | 80 (75%) | 9 (75%) | 55 (86%) † | 16 (53%) | 43 (80%) | 34 (71%) | ||
| NYHA III/IV | 26 (25%) | 3 (25%) | 9 (14%) | 14 (47%) † | 11 (20%) | 14 (29%) | ||
| Functional capacity | ||||||||
| Distance (m) | 365,50 (302,53-428,40) | 404,29±54,71 ‡ | 379,76±82,37 ‡ | 284,59±82,99 § | <0,001 | 387,38±71,34 | 315,63±98,52 | <0,001 |
| Normal | 81 (77%) | 12 (100%) | 54 (84%)† | 15 (52%) | <0,001 | 48 (90%) † | 29 (62%) | 0,001 |
| Low | 24 (23%) | 0 (0%) | 10 (16%) | 14 (48%) † | 6 (11%) | 18 (38%) † | ||
| HF time | 0,300 | 0,100 | ||||||
| <18 months | 35 (37%) | 4 (40%) | 24 (42%) | 7 (25%) | 21 (45%) | 13 (28%) | ||
| ≥18 months | 60 (63%) | 6 (60%) | 33 (58%) | 21 (75%) | 26 (55%) | 33 (72%) | ||
| Mortality risk score MAGGIC | 2 | |||||||
| Score total | 21,0 (16,0-26,0) | 21,20±4,07 ‡,§ | 20,32±5,65 § | 23,82±6,62 ‡ | 0,037 | 19 (16,0-25,0) | 22,0 (16,0-27,0) | 0,248 |
| 1-year mortality | 11,10 (7,0-17,50) | 12,20 (7,52-14,70) | 9,30 (7,0-16,75) | 13,40 (8,62-19,10) | 0,096 | 9,30 (7,0-16,0) | 12,20 (7,0-19,10) | 0,248 |
| 3-y mortality | 26,90 (17,50-39,70) | 29,20 (18,70-34,20) | 22,70 (17,50-38,30) | 31,60 (21,35-42,70) | 0,096 | 22,70 (17,50-36,90) | 29,20 (17,50-42,70) | 0,248 |
| No. of comorbidities | 2,0 (1,0-3,0) | 1,50 (1,0-3,0) | 2,0 (1,0-3,0) | 2,0 (1,0-3,25) | 0,826 | 2,0 (1,0-3,0) | 2 (1,0-3,75) | 0,036 |
| Comorbidities | ||||||||
| DM | 39 (37%) | 5 (42%) | 24 (37%) | 10 (33%) | 0,865 | 13 (24%) | 25 (52%) † | 0,003 |
| SAH | 73 (69%) | 9 (75%) | 45 (70%) | 19 (63%) | 0,704 | 36 (67%) | 35 (73%) | 0,493 |
| Previous stroke or AMI | 23 (22%) | 2 (17%) | 13 (20%) | 8 (27%) | 0,709 | 9 (17%) | 13 (27%) | 0,202 |
| No. of drugs in use | 8,0 (6,0-9,0) | 6,50 (6,0-8,0) | 7,5 (6,0-9,0) | 8,0 (6,0-10,0) | 0,279 | 6,96±1,98 | 8,58±2,65 | 0,001 |
| ACEI/ARB | 96 (91%) | 11 (92%) | 60 (94%) | 25 (83%) | 0,241 | 48 (89%) | 44 (92%) | 0,638 |
| β-blocker | 100 (94%) | 12 (100%) | 59 (92%) | 29 (97%) | 0,835 | 50 (93%) | 46 (96%) | 0,681 |
| Diuretics | 99 (93%) | 12 (100%) | 58 (91%) | 29 (97%) | 0,499 | 51 (94%) | 45 (94%) | 1,000 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
