ABC | Volume 110, Nº1, January 2018

Original Article González-Rivas et al Dyslipidemias in Venezuela Arq Bras Cardiol. 2018; 110(1):30-35 a sub-national sample of Venezuela. Five municipalities from three regions were evaluated: Palavecino, in Lara State (urban), from the Western region; Ejido (Merida city), in Merida State (urban), and Rangel (Páramo area), in Merida State (rural), both from the Andes region; Catia La Mar, in Vargas state (urban), and Sucre, in the Capital District (urban), both from the Capital region. From 2006 to 2010, a total of 1,320 subjects aged 20 years or more, who had lived in their houses for at least six months, were selected by a two-stage random sampling. Three different geographic regions of the country – Andes, mountains at the south; Western, llanos in the middle; and the Capital District, coast at the north – were assessed. Each region was stratified by municipalities and one was randomly selected. A map and a census of each location were required to delimit the streets or blocks, and to select the households to visit in each municipality. After selecting the sector to be surveyed in each location, the visits to households started from number 1 onwards, skipping every two houses. Pregnant women and participants unable to stand up and/or communicate verbally were excluded. All participants signed the informed consent form for participation. The sample size was calculated to detect the prevalence of hypercholesterolemia (the lowest prevalent condition reported in Venezuela) in 5.7% 6 , with standard deviation of 1.55%, which allows to calculate the 95% confidence interval (95%CI). The minimal estimated number of subjects to be evaluated was 830. Overall, 1,320 subjects were evaluated (89.4% from the urban and 10.6% from the rural area). Clinical and biochemical data All subjects were evaluated in their households or in a nearby health center by a trained health team according to a standardized protocol. Each home was visited twice. In the first visit, the participants received information about the study and signed the written informed consent form. Demographic and clinical information was obtained using a standardized questionnaire. Weight was measured with as few clothes as possible, without shoes, using a calibrated scale. Height was measured using a metric tape on the wall. Waist circumference was measured with a metric tape at the iliac crest at the end of the expiration. Body mass index was calculated (BMI: weight[kg]/height[m] 2 ). In the second visit, blood samples were drawn after 12 hours of overnight fasting. Then, they were centrifuged for 15 minutes at 3000 rpm, within 30-40 minutes after collection, and transported with dry ice to the central laboratory, where they were properly stored at -40°C until analysis. Data from participants who were absent during the first visit were collected. Total cholesterol, 8 triglycerides, 9 LDL-c, and HDL-c 10 were determined by standard enzymatic colorimetric methods. Categorization of variables Dyslipidemia was defined according the National Cholesterol Education Program /Adult Treatment Panel III (NCEP/ATPIII) 11 , being categorized in 6 types. Of these, four were isolated dyslipidemias: Low HDL-c (hyperalphalipoproteinemia) < 40 mg/dL in men and < 50 mg/dL in women; high triglycerides: ≥ 150 mg/dL; hypercholesterolemia (≥ 240 mg/dL of total cholesterol); high LDL-c ≥ 160 mg/dL; and two were combined dyslipidemias: atherogenic dyslipidemia (triglycerides ≥ 150 mg/dL + low HDL-c) and mixed dyslipidemia (triglycerides ≥ 150 mg/dL + total cholesterol ≥ 240 mg/dL). Additionally, individuals were classified according to BMI as normal weight (BMI < 25 kg/m 2 ), overweight (BMI ≥ 25 kg/m 2 and < 30 kg/m 2 ), or obese (BMI ≥ 30 kg/m 2 ). 12 Abdominal obesity was established by waist circumference ≥ 94 cm in men and ≥ 90 cm in women. 13 Statistical analysis All calculations were performed using the SPSS 20 software (IBM corp. Released 2011. Armonk, NY: USA). It was verified that all variables had normal distribution using a normality test (Kolmogorov-Smirnov). All variables were continuous and data were presented as mean ± standard deviation (SD). Differences between mean values were assessed with the t-test. Proportions of subjects with dyslipidemia were presented as prevalence rates and 95% confidence intervals (CI). A Chi‑square test was applied to compare different frequencies by gender, nutritional status and abdominal obesity. P-value of < 0.05 was considered statistically significant. Results Characteristics of the subjects Two thirds of the study subjects were female. Men had higher triglycerides, waist circumference and lower HDL-c than women (Table 1). Age, BMI, total cholesterol and LDL-c were similar. Prevalence of dyslipidemia Low HDL-c was the most prevalent lipid change present in nearly seven of ten women, and in about four of ten men (p < 0.01), followed by high triglycerides that were present in half of the men and in one third of women (p < 0.01). Their combination, atherogenic dyslipidemia, was observed in 25.9% of subjects, followed in frequency by increasing LDL-c and total cholesterol levels (Table 2). Mixed dyslipidemia was observed in only 8.9% of the subjects, and was higher among men than in women. An increasing prevalence of all types of dyslipidemias was found when individuals were classified according to BMI and at the presence of abdominal obesity (Figure 1 and Figure 2). The prevalence of hypercholesterolemia, high LDL-c and mixed dyslipidemia were similar in overweight and obese subjects, but higher than those found in the normal weight group. Discussion The present study reports that the most prevalent lipid abnormality in our sub-national sample of adults in Venezuela is the low HDL-c (58.6%), followed by high triglycerides (38.7%), whereas the prevalence of hypercholesterolemia (22%) and its combination with hypertriglyceridemia (8.9%) were lower. Similar findings have been reported in earlier studies, both in Venezuela (Zulia state, Low HDL-c 65.3%, high triglycerides 32.3%), 7 and Mexico (Low-HDL 31