Physical frailty characteristics have a differential impact on symptoms as measured by the CAT score: an observational study

A cross-sectional design was used. Patients with stable COPD were prospectively recruited from an outpatient pulmonary service at Morales Meseguer General Universitary Hospital, Murcia, Spain. The diagnosis of COPD and the stage of disease were based on the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines [1]. Patients with COPD, aged 40–80 years, were eligible to participate when they displayed a post-bronchodilator ratio of forced expiratory volume in 1 s/forced vital capacity (FEV1/FVC) < 0.70. Patients with an unstable cardiac condition within 4 months of the start of the study, as well as those with cognitive deterioration or unable to walk were excluded. During a 1-year period, a consecutive sample of eligible patients was identified from patient health examinations. A pulmonary physician assessed the eligibility criteria for recruitment. The institutional review board of the hospital, the “Ethical Committee of Clinical Research of the General University Hospital”, approved the study protocol (approval number: EST-35/13) and informed consent was obtained from all study participants prior to study entry. The study sample included 137 patients.

Data were collected during face-to-face interviews and through hetero-administered questionnaires, from January to December 2015. Functional tests were performed to obtain clinical data, and other relevant clinical information was extracted from the patients’ clinical history.

Frailty was defined using the Fried phenotype model [14], which is well established and was validated in large epidemiological studies [16]. It consists of five features that reduce physiologic reserve and precipitate a state of vulnerability; unintentional weight loss, low physical activity, exhaustion, slow walking speed, and low grip strength. All cut-off points used were according to the original values of Fried [14]: unintentional weight loss equal or ~ greater than 4.6 kg of body weight in the last year assessed by collecting information from the medical record; low physical activity identified by the Spanish Short Version of the Minnesota Leisure Time Physical Activity Questionnaire (VREM) [17], in the lower quintile adjusted by sex; exhaustion identified by two questions from the CES-D scale [18], with a positive answer to one of the two questions; slow walking speed, assessed by 4-m gait-speed [19], lower than the 20th percentile and adjusted for sex and height; and low grip strength assessed by handgrip dynamometer [20], less than percentil 20, adjusted for sex and BMI; Patients meeting none of the Fried criteria were considered not-frail/robust, those meeting 1–2 criteria were considered pre-frail, and those with ≥3 criteria were considered frail.

For measurement of health status in patients with COPD, we used the CAT [21]. It is a very simple, short, and sensitive questionnaire used to evaluate and quantify the impact of the symptoms of COPD on patients’ health. It consists of eight items measuring different domains of respiratory health in patients with COPD (cough, phlegm, chest tightness, fatigue climbing stairs, household activities, trust/security to leave home, snoring and energy). The resulting score out of 40 indicates disease impact, with higher scores entailing worse health status.

A total of eleven variables were selected from literature research as covariates based on their potential association with either health status or CAT scores. These variables can be classified into two domains; demographic and health-related factors (clinical and psychological).

Demographic variables included sex and age (years), and were self-reported. Clinical variables included body mass index, smoking pack-years, dyspnea, exacerbations, comorbidities, percent of predicted forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC). They were obtained as follows: the level of dyspnea-associated activity was measured by the modified questionnaire hetero-administered and the British Medical Research Council (mMRC); spirometry was performed using standardized procedures indicated by the Spanish Society of Neumology and Thoracic Surgery; patients were measured and weighted to obtain the body mass index (kg/m²); and a thorough review of the patient’s clinical history, the smoking pack-years, and the number of moderate (use of corticosteroids and/or antibiotics) or serious (hospitalization) exacerbations in the previous year were considered. The psychological factors included depression and anxiety that were assessed with the Hospital Anxiety and Depression Scale (HADS) [22]. HADS is an assessment tool validated for cases of depression and anxiety in patient care consultations and in patients hospitalized with chronic diseases. The HADS consists of seven items for measuring anxiety (HADS-A) and seven items for measuring depression (HADS-D). It uses a Likert scale with four possible answers ranging from 0 to 3 points. Scores for each subscale range from 0 to 21; a score of 0–7 indicates the normal range, 8–10 is interpreted as a possible case of anxiety or depression, and 11–21 would indicate likely anxiety or depression.

We used descriptive statistics, including Pearson χ2or Fisher’s exact test, and one way analysis of variance or Kruskal-Wallis (if non-normally distributed data), to summarize the demographic, health-related and frailty characteristics of the whole sample and the groups of patients without and with frailty status (i.e. pre-frail or frail). P-values under 0.05 were considered as statistically significant, except for some intermediary regression models. In these models p < 0.10 was used, in order to reduce the exclusion of relevant variables according to the literature research. Next, univariate and multivariable linear regression models were used to determine the association between the CAT (as dependent variable) and the demographic, health-related and individual frailty characteristics. For these analyses, several models were constructed: model 1 was unadjusted and included each one of the independent variables; model 2 included those significant demographic and health-related variables (with p < 0.10) from model 1; model 3 included each individual frailty characteristic after adjusting for significant (p < 0.10) variables from model 2; the final model adjusted for all significant (p < 0.10) sociodemographic, health and frailty variables simultaneously. All multivariate models were produced using the enter method, except models 2 and final, for which the backward method was used. The association between CAT and frail status (i.e. pre-frail vs non-frail) was examined in a similar way, first using an unadjusted model and second after adjusting for significant (p < 0.10) variables from model 2.

Finally, we also used univariate logistic regression models (by enter method) to examine relationships between patients’ characteristics and those individual frailty characteristics more determinants of CAT.

Sample size calculation was based on the rule of thumb that 15 subjects per predictor are needed for a reliable equation in multiple regression models [23]. We recruited a minimum of 120 participants assuming a maximum of 8 predictors. All analyses were performed using the Statistical Package for the Social Sciences software program (SPSS version 19.0; IBM SPSS, Chicago, IL, USA).

The results of our study provide important aspects for clinical practice. On the one hand, because the presence of exhaustion in patients with COPD has a negative impact on the perception of their health status, it is necessary to explore it beyond other usual psychological aspects such as anxiety and depression. On the other hand, because psychological characteristics interact with physical symptoms and play an important role in exhaustion and how COPD patients experience and manage their disease [33], specific interventions must be conducted to encourage self-management and to provide disease education and health behaviors in order to reduce stress during situations that can result in psychological problems. These interventions can be performed as part of usual care provided by specialists or as part of pulmonary rehabilitation programs. The latter have also showed relevance on CAT score and other clinical outcomes [34]. Moreover they have the advantage that the approach is multidisciplinary (physician, physiotherapist, psychologist, nutritionist, etc.) and provides the proper context to simultaneously improve the physical and psychological condition of patients. Additionally, patients may benefit from contact with other patients with the same clinical condition [35].

It is important to recognize the limited scope of the present study and to consider future research directions. For example, further research could explore consequences over time of exhaustion, how this characteristic persists or remits over time, and eventual longitudinal associations between exhaustion and CAT scores.

To our knowledge, no previous studies have analyzed the relevance of individual frailty characteristics in patients with COPD. The inclusion of a wide variety of determinants of CAT in our models could be considered as one of the strengths of the study. Nevertheless, our study had several limitations. First, patients were excluded if they had unstable cardiac condition, cognitive deterioration or were unable to walk. This fact might have affected our findings, considering that some of the aforementionend patients could have been more frail than those included. Few patients were nonetheless excluded and, consequently, the possible bias should be low. Second, we include a wide and acceptable variability of determinants of CAT, but it may be speculated that some other factors not included in our models could even further improve its power (e.g. medication, participation in rehabilitation programs, etc) [9, 28, 36]. However, information on medication was not available and our hospital doesn’t have a respiratory rehabilitation unit. Finally, because of the small number of women in the cohort, caution should be taken when generalizing the results to women.