Body Composition Estimated by Bioelectrical Impedance Analysis in Patients with Acute Coronary Syndrome.

Introduction

It is indisputable that obesity is a global epidemic. According to the latest data from the World Health Organization (WHO), the prevalence of obesity has tripled since 1975. In 2016, there were 650 million obese people worldwide, representing 13% of those over 18 years of age. Its definition is simple: a body mass index greater than 30. Moreover, it is a preventable condition.

In developing countries, its prevalence is steadily increasing. According to data from the Permanent Household Survey conducted in Argentina, the prevalence of obesity in 2013 was 20.8%; this represented an increase of 15.6% compared to the 2009 edition (prevalence of 18.0%) and 42.5% compared to the 2005 edition (prevalence of 14.6%). The obesity indicator was higher among men (22.9%) than among women (18.8%) and higher among older people (with a maximum of 29.6% in the 50 to 64 age group) compared to younger individuals (7.7%).

The health importance of obesity is mainly due to its relationship with cardiovascular disease. In a classic study analyzing 457,785 men and 588,369 women published in 1999, it was found that the total mortality risk due to obesity was 2.58 and 2.0 times higher, respectively. The relative risk of cardiovascular death in obese men was 2.9, with a 95% confidence interval (CI95%) of 2.37 to 3.56. Obesity increases the metabolism of free fatty acids, reduces insulin sensitivity, increases sympathetic activity, promotes inflammation, and causes a state of higher coagulability: all of these factors can contribute to the development of coronary disease.

Despite the well-known predisposition to cardiovascular disease caused by obesity, once established, its relationship with body weight is more complex. In 2002, a study analyzing the impact of obesity in patients undergoing percutaneous coronary intervention (PCI) was published. Gruber et al. analyzed data from 9,633 patients and observed that obese individuals were generally younger but had more cardiovascular risk factors (hypertension, diabetes, hypercholesterolemia, and smoking). Although angiographic success rates were similar in all groups, the presence of normal BMI was associated with higher cardiovascular mortality. For long-term death, BMI presented in multivariable analysis an odds ratio (OR) of 0.94, with a CI95% of 0.94 - 0.98, showing to be a protective factor. This effect was named the "obesity paradox," which could be formulated as follows: obese individuals have a higher risk of coronary disease, but obese coronary patients have a lower risk of mortality. This was confirmed in a more recent study in patients with ST-elevation acute coronary syndrome (STEMI), where after analyzing data from 50,149 patients, hospital mortality rates of 7.7% were observed for individuals with normal weight and 4.3%, 4.4%, and 6.1% for subjects with class I, II, and III obesity, respectively.

The first explanations for the obesity paradox revolved around a higher risk of bleeding due to increased anticoagulation and the possible presence of other non-cardiovascular diseases in low-weight patients. Later, attention was given to adiponectin, a mediator predominantly produced by adipose tissue, although it can also be synthesized in cardiac myocytes in response to angiotensin II. In rodents, adiponectin increases fatty acid oxidation in muscle, improving insulin sensitivity. Despite being produced in adipocytes, blood levels of adiponectin are inversely proportional to BMI, and hypoadiponectinemia in obese individuals is related to insulin resistance and higher levels of plasma C-reactive protein. Adiponectin levels below 4.0 mg/dl can double the risk of coronary disease.

Beyond possible explanations for the obesity paradox, it has been found that BMI does not always reflect fat mass, and body composition may not be represented by this simple index. In a study where both BMI and waist circumference were measured in patients with ACS, a trend was found towards a higher number of events in patients with lower BMI but higher waist circumference. In another study, waist circumference showed a higher correlation with myocardial infarction size than BMI.

There is limited evidence on the role of body composition in patients with ACS. In a retrospective study in China, the Clínica Universidad de Navarra formula was used to estimate body fat (BF) in patients with impaired renal function, where sex is considered 0 in men and 1 in women: BF = -44.988 + (0.503 × age) + (10.689 × sex) + (3.172 × BMI) - (0.026 × BMI^2) + (0.181 × BMI × sex) - (0.02 × BMI × age) - (0.005 × BMI^2 × sex) + (0.00021 × BMI^2× age). They found that in multivariable analysis, increased body fat and decreased lean mass were associated with a higher risk of death. They also found higher mortality in patients with higher BMI, but without a significant increase with intermediate BMI values. Another study, conducted in the United States, analyzed fat mass based on estimation by body folds in 570 patients with stable coronary disease and found that both reduced fat mass and lean mass were associated with higher mortality.

Adiposity and lean mass can be evaluated non-invasively, cheaply, and quickly using bioimpedance analysis (BIA). Total body composition estimation through total body bioimpedance can be carried out using the equation V = ρ x S^2/R, where V is the conduction volume representing total body water or fat-free mass, ρ is the specific resistivity of the conductor, S is height, and R is total body resistance, measured with four surface electrodes placed on one wrist and one ankle.

In Latin America, at least two BIA validation studies have been conducted to estimate body composition. In the study by Fjeld et al. in Peru, the correlation coefficient for cross-validation of formulas for calculating body composition was 0.96. In Argentina, in 2008, Rodríguez et al. conducted a study aimed at comparing body composition estimated by two simple anthropometric methods, BIA, and dual-energy X-ray absorptiometry (DXA) and studying the correlations between them in a pediatric population. They found good correlation between simple anthropometric methods (waist circumference) and bioimpedance and DXA, but the results were not interchangeable, even between BIA and DXA.

In cross-validation studies of formulas to estimate body composition from bioimpedance, high correlation coefficients and small standard errors of estimation were found, with acceptable precision. Due to this, it is considered that, despite their limitations, they can be extrapolated to different populations. However, it is important to note that although the estimates are validated and highly accurate for evaluating populations, they are less reliable for determining body composition in individual subjects.

Very few studies have measured body composition using BIA in patients with stable coronary disease. Puri et al. estimated the percentage of BF using this method in 477 individuals divided into three groups: normal, with established coronary disease, and at high risk of coronary disease. BF was higher in the at-risk group for coronary disease (30.7%), compared to 25.4% in patients with established coronary disease and 23.9% in controls. However, most strikingly, despite a high correlation between BF percentage and BMI (r = 0.8), 34% of men and 44% of women with normal BMI had increased BF. This suggests that BMI significantly underestimates obesity in certain patients.

In a cross-sectional study conducted on 161 patients undergoing coronary angiography for stable coronary disease, characteristics were compared between those with coronary lesions and those without. BMI and body composition estimated by BIA were taken into account. There were no differences in total weight and BMI, but BF was higher among those with coronary lesions, while lean mass was higher among those without lesions.

Only one study has used BIA in patients with ACS. It was found that visceral fat is a better risk indicator for the no-reflow phenomenon after PCI in STEMI patients compared to BMI and total BF.

Given the limited experience with BIA in patients with ACS and the effects that body composition may have on clinical outcomes and as a confounding factor with BMI, this study proposes to investigate this topic in greater depth.

Objectives

2.1 General To describe the body composition of patients with ACS estimated through BIA and evaluate its association with clinical outcomes.

2.2 Specific

Describe the body composition of patients with ACS in terms of lean mass, body fat, and fluid volume estimated through BIA.

Analyze the relationship between body composition parameters with age, sex, and BMI.

Determine if there is an association between a higher percentage of body fat and mortality during hospitalization and at one year.

Determine if there is an association between a lower percentage of lean mass and mortality during hospitalization and at one year.

Determine if a lower percentage of body water is associated with a higher incidence of renal failure during hospitalization.

Determine if a higher percentage of body water is associated with a higher incidence of heart failure during hospitalization.

Hypotheses

There is a difference in the percentage of body fat between those who die and those who survive during hospitalization and at one year.

The percentage of lean mass is lower among those who die and those who survive during hospitalization and at one year.

Patients with lower body water estimated by BIA have a higher risk of acute renal failure (ARF) during hospitalization.

Patients with higher body water estimated by BIA have a higher risk of heart failure during hospitalization.

Methodology

4.1 Study Design This is a prospective cohort study of patients with ACS.

4.2 Population The study population will include all adult patients with ACS.

4.3 Data Collection Patient data will be collected from the electronic medical record. BIA will be performed within the first 72 hours of admission.

4.4 Data Analysis Statistical analysis will be conducted using R software. Categorical variables will be compared using the chi-square test or Fisher's exact test as appropriate. Continuous variables will be analyzed using t-tests or Mann-Whitney U tests for independent samples, depending on their distribution. Multivariable analysis will be conducted using logistic regression.

Ethical Considerations The study protocol will be submitted to the Research Ethics Committee of the Hospital for approval before implementation. All participants will sign an informed consent form before inclusion in the study.