Freqüência de hipertensão arterial e fatores associados : Brasil , 2006 Frequency of arterial hypertension and associated factors : Brazil , 2006

MÉTODOS: Estudo baseado em dados do sistema de Vigilância de Fatores de Risco e Proteção para Doenças Crônicas por Inquérito Telefônico (VIGITEL), coletados em 2006 nas capitais brasileiras e Distrito Federal. Estimou-se a freqüência de hipertensão arterial sistêmica entre 54.369 adultos, estratifi cada por sexo, região geográfi ca, variáveis sociodemográfi cas e comportamentais e morbidades auto-referidas. Foram calculadas os odds ratios brutos de hipertensão e ajustados para variáveis do estudo.

Systemic arterial hypertension (SAH) affects approximately 25% of the world's population, and the forecast is that the disease's cases will increase by 60% in 2025. 7It is estimated that 62% of cerebrovascular disease and 49% of ischemic coronary artery disease can be attributed to suboptimal blood pressure, with a small variation between sexes. 19Besides the impact on the morbidity and mortality of populations, SAH is related to high socioeconomic costs. 1 Other cardiovascular risk factors are commonly associated with SAH, like obesity and glucose and lipids metabolism disorders.Others can be causally associated with the increase in blood pressure levels, such as inadequate diet, excess of salt, excessive alcohol consumption, physical inactivity, overweight and smoking.Based on this knowledge, lifestyle changes have been indicated in SAH prevention and treatment.
According to the World Health Organization (WHO), in the Americas sub-region, SAH is among the three main risk factors that concur to the total disease load. 19a In a household survey carried out in 15 capitals and the Federal District about risk factors and self-reported morbidity for non-communicable diseases (NCD), the frequency of SAH varied from 16% to 45%.b Methodological differences like non-representative sampling, distinct population groups, restricted geographic inclusion, criteria, quality of the diagnosis and different approaches to data analysis hamper the comparison between studies and, consequently, their utilization as a decision tool for public health.
Surveillance of risk factors and diseases enables to detect trends in time and geographic space and to plan preventive actions in public health.Self-reported morbidity, without the specifi c diagnostic parameter, represents an alternative to estimate the frequency of diseases in populations.
Telephone-based surveys have proved to be useful tools to monitor aspects related to the population's INTRODUCTION health, like the Behavioral Risk Factor Surveillance System, c in the United States, which monitors risk factors, diabetes and hypertension, among others.Analysis of data from another survey, the National Health and Nutrition Examination Survey III (1988-1991), d showed good sensitivity (71%) and specifi city (92%) in the identifi cation of risk factors and health problems, and it has been suggested that the frequency of SAH in the United States' population could be quantifi ed by that instrument. 17A similar experience was already reported in Brazil in 2004. 8e aim of this study was to analyze the frequency of self-reported systemic arterial hypertension and associated factors.

METHODS
Cross-sectional study based on data from the system of Telephone-based Surveillance of Risk and Protective Factors for Chronic Diseases (VIGITEL) carried out in 2006.The objective of the system is to monitor, on a regular basis, the frequency and distribution of risk and protective factors for NCD in the capitals of the 26 Brazilian states and in the Federal District.To perform this, computer-assisted, telephone-based interviews were conducted in probabilistic samples of the adult population living in households served by fi xed telephone lines of each city.e The sampling process of VIGITEL was based on the draw of 5,000 households with fi xed telephone line per city, followed by the draw of one dweller aged ≥ 18 years per household, who had accepted to be interviewed, until the minimum fi gure of 2,000 interviews per city was obtained.A total of 54,369 interviews were conducted.e The questionnaire administered by VIGITEL included questions about demographic and socioeconomic characteristics, dietary and physical activity pattern associated with NCD, weight and height, use of cigarettes and alcohol, self-rate of the health status and report of medical diagnosis of SAH, The analyzed data refer to SAH that was previously diagnosed by a doctor, considered as response variable and categorized as yes or no.The independent variables were divided into sociodemographic (age, skin color, level of schooling and marital status), behavioral (physical activity during leisure time, dietary habits and smoking) and associated diseases (obesity, DM, dyslipidemia and cardiovascular events).Physical activity during leisure time was considered present if it was of light or moderate intensity and practiced for ≥ 30 minutes, on at least fi ve days a week, or the practice of ≥ 20 minutes of vigorous physical activity on three or more days a week; walking was also considered, when practiced for ≥ 30 minutes, at least fi ve days a week.Smoking was stratifi ed in three categories (never smoked, smoker or ex-smoker); alcohol consumption, in two (yes for consumption higher than fi ve doses for men and four doses for women in at least one occasion in the last 30 days); and addition of salt to the prepared meal in three (no, sometimes, yes).
Body mass index (BMI = weight in kilograms divided by height in meters squared) was classifi ed according to WHO criteria (eutrophia if BMI < 25 kg/m², preobesity if BMI between 25 and 29.9 kg/m², and obesity for BMI ≥ 30 kg/m²). 18All individuals with BMI ≥ 25 kg/m² were considered with excessive weight.The variables DM, dyslipidemia and cardiovascular events (myocardial infarction or cerebrovascular accident) were analyzed as present or absent.
For continuous variables, means and standard deviations were calculated.Frequencies of SAH by point and 95% confi dence interval were calculated and stratifi ed by sex, geographic region and the variables mentioned above.As the VIGITEL sample is limited to adults with home telephone, expansion factors were used to estimate prevalence in the overall population.A weighting factor corresponding to the number of adults and the inverse of the number of telephone lines of the household was employed.Then, a factor was used to correct possible sociodemographic differences between the adult population that has telephone and the overall population of the municipality according to the 2000 census.Of the 54,369 interviewees, 46.1% were men.The mean age of the studied sample was 39.1 years (sd=16.0).Mixed-ethnicity individuals predominated (52.0%), followed by white (40.6%) and black (6.6%) individuals.The yellow and red skin colors were reported with frequencies of 0.6% and 0.1%, respectively, and were grouped into the category "others".The majority of the individuals (53%) reported having between zero and eight years of schooling (data not shown).
The frequency of self-reported SAH was 21.6% (CI 95% 20.9;22.4),adjusted for the population of the 27 cities.Women reported SAH more frequently than  1).These fi ndings were the basis for the adjustments in the analysis of the associations with SAH.
Table 2 shows SAH frequencies according to sex and behavioral characteristics.Among men, SAH was more frequent in individuals who did not practice physical activity during leisure time.A statistically signifi cant association was found between the presence of SAH and the practice of walking for both sexes.Higher SAH frequency was found among exsmokers and among the subjects who did not consume alcohol.Among individuals who added salt to the meals, the frequency of SAH was lower.
SAH frequency increased as BMI increased, occurring in approximately half of the obese women (Table 3).The prevalence ratio of SAH between obese and normal weight individuals, independently of the sex, was 3.18 (CI 95% 3.05;3.32)(data not shown).
As for association with other diseases, it was observed that the frequency of SAH was higher among individuals who reported diagnosis of DM and dyslipidemia.Similarly, cardiovascular events (myocardial infarction and stroke) were more frequent among individuals with self-reported SAH (Table 3).
Table 4 shows crude and adjusted OR of SAH.In the block of the behavioral variables, it was verifi ed that the OR of SAH was higher among individuals who practiced physical activity during leisure time, ex-smokers and men who reported excessive consumption of alcohol.On the other hand, this ratio was lower for those who added salt to prepared meals.It was found that there was association of SAH with the variables overweight, DM, dyslipidemia and cardiovascular events.Even after adjustment for other cardiovascular risk factors, like overweight, DM and dyslipidemia (model 3), SAH remained independently associated with the presence of cardiovascular events.

DISCUSSION
The present study estimated that the frequency of self-reported hypertension was 21.6%, being higher among women (24.4% versus 18.4%), lower in the North and Central-West regions and higher in the Southeast region.The frequency of hypertension increased with age, decreased with level of schooling, was higher among black and widowed individuals and lower among singles.The chance of hypertension, adjusted for confounding variables, was higher for individuals with overweight, diabetes, dyslipidemia and cardiovascular events.
The comparison between SAH frequencies based on self-reported information and those provided by studies with blood pressure measurement shows that the subjects have a high degree of knowledge of their hypertensive state, meaning that the initiatives of health organs and professionals concerning SAH detection have produced the desired result.However, knowledge of health status does not imply change in behavior and the present study does not inform the level of blood pressure control of these Brazilians.
The SAH frequency variation refl ects the potential of To be continued access to the healthcare services and to the diagnosis of the disease in these regions.The proportion of primary healthcare services per inhabitant is lower in the North and Central-West regions and higher in the Southeast region, a coinciding with the values of SAH frequency obtained in these regions of Brazil.
SAH has been pointed as being more frequent among men up to 50 years of age. 7b The present study confi rmed the increase in SAH frequency with age, with alarming percentages in younger age groups.Although cross-sectional studies do not allow to establish cause-effect relations, it is reasonable to suppose that this fi nding is due to the increasing role of the weight gain of the Brazilian population.c The association of SAH with the BMI categories corroborates this hypothesis.Obesity is considered a risk factor for SAH 5 and its genesis implies environmental factors, such as inadequate dietary habits and physical inactivity.
Information collected by VIGITEL enabled to investigate some associations between SAH and other variables, which may be of causal nature.Associations with skin color could represent a genetic predisposition, while the excessive consumption of certain foods or inactivity might refl ect previous habits that favor the increase in blood pressure.On the other hand, associations with DM and dyslipidemia might express a common physiopathological mechanism.Finally, fi nding a greater chance of cardiovascular events among hypertensive subjects would indicate predisposition to the main complication of SAH: thromboembolic phenomena in the atherosclerotic disease.
Results of the present study showed that SAH frequency was higher among black individuals, followed by those of mixed-ethnicity, and among subjects with low level of schooling.Stratifi cation according to skin color, similarly to the majority of the studies conducted in Brazil, presents some degree of inaccuracy, and it should be seen with caution.To eliminate the possible infl uence of the level of schooling on the frequency of SAH among African descendants, the estimates were adjusted, and the differences remained signifi cant.These results confi rmed fi ndings in other populations. 13 relation to physical activity during leisure time, smoking (ex-smokers) and salt consumption, the associations found in the present study were opposed to the ones that would be expected in longitudinal studies.This effect possibly derives from reverse causality, i.e., individuals diagnosed as hypertensive reported more frequently that they practice exercises, stopped smoking and reduced the amount of salt added to prepared meals.This fi nding may refl ect effective communication about the importance of improving life habits in order to minimize the cardiovascular risk.
As for alcohol consumption, the hypothesis of reverse causality remained for the female sex, but not for the men.In this sense, it is possible that women are more compliant than men regarding changes in life habits.
In the present study, SAH was associated with DM and dyslipidemia.These three diseases are among the main cardiovascular risk factors.Therefore, fi nding that hypertensive individuals presented higher frequency of myocardial infarction and cerebrovascular disease was expected, confi rming results of studies conducted in Brazil and in other countries. 2,6,9,15a Another criticism regarding the use of self-reported morbidity is related to the infl uence of the access to the medical services; therefore, it may be underestimated.However, the SAH frequency observed in the present study (21.6%) was close to those obtained in prevalence studies involving blood pressure measures 4,13 and was similar to the one observed in the 2003 national survey, which also used self-reported information.b On the other hand, it is also found in the literature a percentage of self-reported SAH that is higher than the one found in the present study, which can be attributed, at least in part, to the higher mean age of the studied population. 16other limitation of this study is the representativeness of the sample, which is restricted to individuals who own a fi xed telephone line.c However, this bias was minimized through the utilization of expansion factors to represent the overall adult population of the studied cities, according to the sociodemographic distribution of the 2000 census.
Another limitation was the utilization of self-reported weight and height for the BMI calculation; nevertheless, studies have confi rmed the validity of this kind of measure. 3,12 conclusion, 21.6% of the Brazilian population of the cities included in VIGITEL reported suffering from SAH. VIGITEL proved to be a useful tool in the monitoring of this disease and of its associated factors.The high frequencies of modifi able risk factors reveal potential intervention targets, aiming at the prevention and control of SAH.The continuity of data collection about SAH by the VIGITEL system will enable to assess trends and to provide subsidies to evaluate the impact of detection and intervention public policies on SAH.

Table 2 .
Standardized frequencies a of arterial hypertension, stratifi ed by sex according to behavioral characteristics.Brazil,  2006.(N=54,369) a Sociodemographic distribution of the sample of VIGITEL adjusted to that of the adult population of each city in the 2000 Demographic Census, considering the population weight of each city.aMinistériodaSaúde.Secretaria de Vigilância em Saúde.VIGITEL Brasil 2006.Vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico: estimativas sobre freqüência e distribuição sócio-demográfi ca de fatores de risco e proteção para doenças crônicas nas capitais dos 26 estados brasileiros e no Distrito Federal em 2006.Brasília; 2007 [cited 2008May 10].Available from: http://bvsms.saude.gov.br/bvs/publicacoes/relatorio_vigitel_2006_marco_2007.pdfmen (24.4% versus 18.4%), resulting in a prevalence ratio of 1.33.SAH frequency increased with age, decreased with level of schooling, was higher among black individuals and in widowed subjects and was lower among singles (Table

Table 3 .
Standardized frequencies a of arterial hypertension, stratifi ed by sex according to the presence of self-reported morbidities.Sociodemographic distribution of the sample of VIGITEL adjusted to that of the adult population of each city in the 2000 Demographic Census, considering the population weight of each city. a

Table 4 .
Crude and adjusted odds ratios of arterial hypertension, stratifi ed by sex, according to behavioral characteristics and associated diseases (N=54,369)