The CORDIAL (Cardiovascular Outcomes Registry in Dialysis Patients) study was designed to collect data on cardiovascular risk profile and carry out the follow-up of all CKD patients enrolled in a regular outpatient hemodialysis program in the city of Porto Alegre, Brazil. The present article describes the baseline phase data and reports a cross-sectional analysis of this population.

Eligibility criteria for enrollment included age 18 years or more, being in chronic outpatient dialysis for more than 30 days, and ability to provide informed consent for participation. The study complies with the Declaration of Helsinki and the protocol was approved by the Ethics Committee for Research of Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA) and by the review boards of the clinical centers included.

From August 2010 to March 2011, all CKD patients on hemodialysis in each of the fifteen dialysis centers of the city of Porto Alegre (Appendix 1) were considered for enrollment and 1215 (97.3% of all) matched the inclusion criteria and agreed to provide informed consent.

To collect the CORDIAL baseline data we used a standardized case report form adapted and extended to our needs from the REACH registry, a study performed in several countries including Brazil21. The form was comprehensively filled after visits to each center where patients, health care staff and electronic clinical records were accessed.

Weight, height and blood pressure measured before and after the dialysis session (mean of the last 3 sessions) were obtained from reviewing the patients' medical records. Prevalence of atherosclerotic cardiovascular disease, diabetes, and hypertension were established by reviewing all history and physical data in the dialysis clinic medical charts and other available medical documents. Furthermore, we considered as hypertensive any patient in use of antihypertensive drugs, or those with mean pre dialysis blood pressure in the stage 1 or up of the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) classification22.

The definition of diabetes included both type 1 and type 2, with or without oral hypoglycemic medication or insulin. Tobacco use was registered as past or present. In accordance with the Kidney Disease Outcomes Quality Initiative of the National Kidney Foundation (K/DOQI-NKF)23, dyslipidemia was defined as the presence of any of the following - total cholesterol > 200 mg/dL, low-density lipoprotein (LDL)-cholesterol > 100 mg/dL, high-density lipoprotein (HDL)-cholesterol < 40 mg/dL, triglycerides > 150 mg/dL, or the use of statin. Current physical activity was established by an inquiry asking about "any moderate activity (as light walking, bicycling, or gardening) or hard one (running, playing football - soccer, hard swimming) at least 30 min two times per week" and "how many minutes in a week".

Laboratory data were obtained from the medical chart as the most recent result in the six preceding months.

Data about gender were available for all CORDIAL participants, about age for 1213 (99.8%), and race, height and weight for 1206 (99.3%) each.

Information about hypertension status, cardiovascular disease, physical activity, smoking and lipid profile were available respectively for 1210 (99.6%), 1170 (96.3%), 1150 (94.7%), 1147 (94.4%), and 1121 (92.3%) of the individuals in the cohort.

Statistical analyses were carried out in STATA 9.0. Descriptive statistics were used to characterize the sample. Unadjusted and adjusted prevalence ratios were calculated separately for each dependent variable (hypertension, diabetes, dyslipidemia, tobacco use, sedentary lifestyle, and obesity) with independent variables (gender, age, race, time in dialysis, and diagnosis of cardiovascular disease). All tests were two-tailed and the significance level used was 0.05.

Hypertension was found in 87.5 % of our patients, a result higher than that found by Portolés et al. (the MAR study) in a multicenter cohort in Spain a decade ago (75.8 %)19, but similar to others, like 87.1 % in the study by Ohsawa et al. in Japan (KAREN Study, 2005)18, 86 % in a cohort of 2535 patients in the USA (Agarwal, 2003)25, and around 86 % in two studies in Spain9 ^, 12. The CHOICE study found a higher prevalence (96.0 % among 1041 individuals)14. Pre dialysis mean blood pressure in our population (about 147/82 mmHg) is similar to other studies (149/79 mmHg in CHOICE, and 147/80 mmHg in DMMS Wave2)10 ^, 14. Uncontrolled blood pressure in the pre dialysis measurements is 72 % in our cohort (Table 1), similar to the 70 % in the study of Agarwal et al25 and to the 69 % in the CHOICE study14. In CORDIAL, a multivariate adjusted analysis found that patients under 65 years-old presented a higher prevalence of hypertension.

Hypertension has been related to higher mortality in hemodialysis patients26 and the importance of treating high blood pressure in these individuals was highlighted in two recent meta-analysis suggesting a reduction in cardiovascular events and all-cause mortality in those treated with antihypertensive drugs27 ^, 28.

Among our patients, the prevalence of dyslipidemia was 84.7 %, a higher rate comparing to other studies like 50% in CHOICE14. The different criteria in each study preclude a more accurate comparison. Our mean LDL-cholesterol (94.3 mg/L) was similar to others9 ^, 12 ^, 14 ^, 18 ^, 19, with Kronnnenberg reporting a higher value (115 mg/dL)24.

The prevalence of elevated triglycerides was 45.2 % in our cohort, higher than the study by Collado et al12 and the CHOICE study (36.9 % over 200mg/dL)14. Mean values were 171.7 mg/dL in CORDIAL, higher than in KAREN and ANSWER, similar to the MAR study and Kronenberg et al24, and lower than CHOICE (199mg/dL)9 ^, 14 ^, 18 ^, 19 ^, 24. HDL-cholesterol under 40 mg/dL was more prevalent in our cohort (60.3 %) than in others9 ^, 14 ^, 18 ^, 19. An exception was the 66 % reported by Kronenberg et al24.

In CORDIAL, there was a higher adjusted prevalence of dyslipidemia in female patients, while Ohsawa et al. reported an opposite finding18. Although there is not a proven association between serum lipids and the development CVD outcomes in dialysis patients29, lipid measurements and treatment remains a cornerstone approach to prevent CVD even in CKD patients24.

Tobacco use is associated with a markedly increased risk of heart disease in dialysis patients8 ^, 16. Smoking (past or present) was described in 53.7 % of our patients, in accordance to others studies9 ^, 10 ^, 12 ^, 14 ^, 19. Thirteen percent of our patients were current smokers, similar to other authors9 ^, 14 ^, 19, except for the KAREN study, with 28.2%18. In CORDIAL, the adjusted prevalence of tobacco use was higher in males and in those with less than 65 years.

Sedentary lifestyle was identified in 73.1% of our patients. In the CHOICE study14, only 14 % reported physical activity to perspiration at least 3 times a week. Tentori et al. (2010)30 described results of 20,920 patients in the DOPPS study with a self-reported regular physical activity of 47.4 %. Comorbidities and physical limitations could explain the low adherence of haemodialysis patients to exercise programs31. Besides, older age could represent another limitation to physical activity - among our patients, adjusted prevalence of physical inactivity was higher in those over 65 years than in younger patients (Table 3).

Diabetes prevalence among dialysis patients is already high and increasing32. The 5-year survival rate among diabetic patients on hemodialysis is worse than in non-diabetics. A recent study in Brazil found 41.1 versus 62.7 %, respectively33. According to the 2009 United States Renal Data System (USRDS) database, only 30 % of the diabetic patients survived five years after beginning hemodialysis1.

In CORDIAL, diabetics were 35.8 % of the population, a lower rate than the one presented in CHOICE (54%)14, but similar to the 38.5% in the 2012 USRDS Annual Report34, with others ranging from 26 to 43 %9 ^, 10 ^, 18 ^- 20.

Obesity estimated by body mass index (BMI) was present in 12.4% of individuals in CORDIAL. The MAR study19 reported obesity in 14% of their patients, while in ANSWER and CHOICE the prevalence was higher (20 and 26%, respectively)9 ^, 14. Unlike the general population, overweight dialysis patients have better prognosis, supposedly due to a better nutritional status35. Recently, BMI has been considered a non-efficient tool in the assessment of excess body fat content and thereafter obesity36. Reverse epidemiology and inadequate assessment of fat accumulation could explain the "obesity paradox" in CKD37.

Age

In CKD patients on hemodialysis, older age confers a cardiovascular risk that parallels the relationship described in the general population38. There is a growing prevalence of elderly in the hemodialysis population all-over the world9 ^, 10 ^, 19. In the Brazilian dialysis census of 2000, those with age 60 or more were 26 %39, while in 2011 there were 31.5 % with age 65 or more40. In CORDIAL, also in 2011, we found a rate of 34.9 %.

Our data presents the usual drawbacks of cross-sectional studies limiting causal inferences. We had to rely on clinical laboratory measurements of 15 dialysis centers instead of using a core lab. We described similar rates observed in other studies, but our data cannot be generalized to other cities in our country. Our data, however, has the strength of having been acquired from a large population with few dropouts and double-checking structured individual interviews and clinical charts.