Heart failure (HF) is a pandemic condition whose prevalence is steadily increasing as populations age.1 In Brazil, it is second only to childbirth as a cause of hospitalizations in adults and is the leading cardiovascular cause of hospitalization.2

The triad of HF symptoms (shortness of breath [SOB], fatigue and bilateral lower limb edema [LLE]) is found in more than 50% of patients, reaching 80% in some populations.3 These symptoms restrict patients’ daily life and impact their quality of life (QOL).3 A clinically based diagnosis of HF (clinical history, signs and symptoms) has low accuracy,4 and access to more accurate methods such as the B-type natriuretic peptide (BNP) blood test can be problematic in a primary care (PC) setting due to cost and limited availability of skilled professionals,4 resulting in delayed treatment. Early diagnosis and intervention can postpone the development of the disease,4 reverse left ventricular remodeling and improve QOL in several ways.5

This study aims to estimate the association of complaints of SOB, fatigue and LLE (typical symptoms of HF) with QOL in the population covered by the Family Doctor Program (FDP) in Niteroi, Rio de Janeiro, Brazil. Such an association may point to a possible alternative for prioritizing access to more accurate tests for the diagnosis of HF in patients with HF symptoms and low QOL.

This was a cross-sectional study, conducted within the CAMELIA study, which involved families covered by the FDP of Niteroi, Rio de Janeiro, Brazil. Randomly sampled individuals (index cases) and their spouses and children between 12 and 30 years of age were invited to participate. The index cases were classified as non-diabetic hypertensive, non-hypertensive diabetic, hypertensive diabetic or non-diabetic and non-hypertensive.

In this study, individuals were classified as positive for SOB, fatigue and LLE based on positive responses to direct questions in the self-administered questionnaire. For the QOL assessment, the Medical Outcomes Study 36-item Short Form Health Survey (SF-36) was used. The SF-36 is a self-administered general QOL questionnaire consisting of 36 items that evaluate the following dimensions: functional capacity, physical aspects, pain, general health, vitality, social and emotional aspects and mental health.6 For each QOL dimension, a median score was calculated and considered as the cutoff point. To assess depression, the Beck Depression Inventory short form (BDI-SF) consisting of 13 items,7 each with four possible answers, was used. Scoring was in a continuous and dichotomous form, with a cutoff of 9/10.

Of the 1098 CAMELIA study participants, 455 individuals (41.43%) were eligible for this study. Individuals over 30 years of age (664) were included, and the study excluded those with no available information for anthropometric measurements (10), hematocrit (25), glucose (68), and total cholesterol (57) and those who did not respond to the SF-36 (41) or to questions regarding fatigue (7), SOB and LLE (one person did not respond to either the SOB or the edema questions). Missing data were considered random and were due to difficulty in keeping the participants in the clinic on days when the wait was longer due to technical difficulties in processing blood and/or urine samples. Comparison between the two samples showed a similar demographic profile: 47.4% men among those excluded vs. 47.5% of those included, with a mean age of 48.9 and 47.7, respectively. Median BDI-SF score was also the same: 7 points. The median of five of the eight dimensions of SF-36 was similar in both groups. For physical functioning this was 85.0 points for those excluded and 90.0 points for those included, 75.3 and 100.0 for physical and 67.5 and 70.0 for general health, respectively. The prevalence of fatigue was 54.8% vs. 56.9%, SOB 30.5% vs. 22.6%, and LLE 6.3% and 16.9%, respectively. Between June 2006 and December 2007, 13 communities, selected by convenience, were visited and the following evaluations were performed: anthropometric assessment, medical consultation with medical history, family history and physical exam, and a self-administered questionnaire (with support from the researchers) on sociodemographic conditions, lifestyle and mood.

Patients who reported a previous diagnosis of hypertension or those who presented with systolic blood pressure (SBP)≥140mmHg or diastolic blood pressure (DBP)≥90mmHg8 were classified as hypertensive. Patients were classified as diabetic if they had a confirmed previous diagnosis or presented with a fasting glucose level≥126mg/dl.9

The following lipid values were considered elevated: total cholesterol (TC)≥200mg/dl, LDL cholesterol≥130mg/dl, triglycerides (TG)≥150mg/dl, and VLDL cholesterol≥30mg/dl.10 Patients were considered to be hyperuricemic when uric acid was ≥6mg/dl for females and ≥7mg/dl for males. Anemia was defined as hematocrit<40% for males and <36% for females. A body mass index (BMI) of ≥30kg/m2 was defined as obese.11

The majority of the 445 participants were between 40 and 49 years of age, female and mulatto, had only attended elementary school and had a per capita family income of up to R$200.00 (the latter two data are not shown). The majority stated the following about their lifestyle-associated cardiovascular risks: they never smoked, had not consumed alcohol regularly over the previous six months, did not use salt at the table and exercised for less than 150minutes per week (they were physically inactive). Only 10% had symptoms of depression. Regarding comorbidities related to cardiovascular risk, the majority were not classified as obese or diabetic but were hypertensive and dyslipidemic. Fatigue was the most prevalent symptom (56.9%), followed by SOB (22.6%) and LLE (16.9%).

We observed a negative association of breathlessness with race (mulatto) (data not shown) and creatinine and a positive association with smoking, depression (according to the BDI-SF), obesity, hypertension, use of antihypertensives, asthma, bronchitis, myocardial infarction, atrial fibrillation (AF), stroke, angina and HF. Fatigue was negatively associated with hemoglobin and hematocrit and positively with gender (female), depression (according to the BDI-SF), obesity, diabetes, dyslipidemia, atrial fibrillation and angina. A positive association of LLE with female gender and obesity was found. All these variables were entered in the models to control for possible confounding, with the exception of HF. The three symptoms were associated with each other, although the association between SOB and LLE was not statistically significant (Table 1).

There was a statistically significant crude association between SOB and fatigue with all SF-36 dimensions. The probability of scoring below median on all eight dimensions of the SF-36 was higher for those who reported the presence of these symptoms. The association of LLE with the vitality and general health dimensions did not reach statistical significance. In the adjusted analysis, SOB and fatigue were again associated with all of the SF-36 dimensions, with the exception of the emotional aspects dimension in the case of SOB. LLE was not significantly associated with any of the SF-36 dimensions in the adjusted analysis (Table 2).

The probability of an above-median QOL score decreased as the number of symptoms increased, independently of gender, age, race, smoking, depression, obesity, diabetes, dyslipidemia, hypertension, antihypertensive medication, asthma/bronchitis, previous myocardial infarction, atrial fibrillation, stroke, angina, creatinine, hemoglobin and hematocrit (Table 3).

In five of the eight dimensions (functional capacity, physical aspects, emotional aspects, mental health and pain), an increase in the number of symptoms was directly associated with an increased probability of a below-median score on the QOL questionnaire. For the vitality, social aspects and general health dimensions, the probability of a below-median score increased for up to two symptoms and decreased when a third symptom was present (data not shown). Combinations of the three symptoms did not show the same pattern of impact on QOL for all of the dimensions. On most dimensions, LLE and SOB had less impact on QOL than fatigue (Figure 1).

Figure 1.

Average rank of quality of life (SF-36) according to combinations of symptoms of heart failure. Niteroi Family Doctor Program, 2006–2008.

(0.53MB).

The prevalence of SOB was 22.6%, fatigue 56.9% and LLE 16.9%. These results are intermediate between those of general populations and those of patients with HF.

The associations with the dimensions of the SF-36 were statistically significant, independent of demographic and socioeconomic factors, metabolic characteristics and comorbidities, for SOB and fatigue but not for LLE.

Neither symptoms nor QOL have good predictive accuracy for a diagnosis of HF. The combination of symptoms (especially SOB and fatigue) with QOL has increased positive predictive value for an HF diagnosis and may be used as part of the triage process.

The authors have no conflict of interest to declare.