Mortality due to cardiovascular diseases in the Americas by region, 2000-2009

ABSTRACT CONTEXT AND OBJECTIVE: Cardiovascular diseases are the leading cause of death worldwide. The aim here was to evaluate trends in mortality due to cardiovascular diseases in three different regions of the Americas. DESIGN AND SETTING: This was a time series study in which mortality data from three different regions in the Americas from 2000 to the latest year available were analyzed. METHODS: The source of data was the Mortality Information System of the Pan-American Health Organization (PAHO). Data from 27 countries were included. Joinpoint regression analysis was used to analyze trends. RESULTS: During the study period, the age-adjusted mortality rates for men were higher than those of females in all regions. North America (NA) showed lower rates than Latin America countries (LAC) and the Non-Latin Caribbean (NLC). Premature deaths (30-69 years old) accounted for 22.8% of all deaths in NA, 38.0% in LAC and 41.8% in NLC. The trend analysis also showed a significant decline in the three regions. NA accumulated the largest decline. The average annual percentage change (AAPC) and 95% confidence interval was -3.9% [-4.2; -3.7] in NA; -1.8% [-2.2; -1.5] in LAC; and -1.8% [-2.7; -0.9] in NLC. CONCLUSION: Different mortality rates and reductions were observed among the three regions.


INTRODUCTION
Cardiovascular diseases (CVDs) are the leading cause of death worldwide, accounting for 30% of the total number of annual deaths globally. 1 They comprise major disorders of the heart and blood vessels, which include heart coronary disease and stroke.
Nowadays, this group of diseases is not considered to be only a problem of developed countries, since estimates for 2010 showed that low and middle-income countries were more affected, since they accounted for 80% of these deaths. 1 Globally, ageing and unhealthy behavioral changes like tobacco smoking, physical inactivity and unhealthy diets, have become important contributors to the increased prevalence of intermediate risk factors such as obesity, dyslipidemia, raised blood pressure and raised blood sugar. Age is an unavoidable risk factor, but avoidance of premature mortality should be a public health concern; the percentage of premature deaths ranges from 4% in high income countries to 42% in lowincome countries. 2 This group of diseases is important to public health not only because of their magnitude but also because of the possibility of intervention. Prevention strategies at population level to reduce circulatory system mortality rates can be classified as high-risk and community-based. 3 The first of these relates to identifying individuals who are at high risk, through screening, and referring them for treatment. The second of these comprises implementation of programs at population level aimed at modifying one or more scientifically well established risk factors such as use of tobacco, physical inactivity, etc. As a result of implementation of such programs and policies, declines in mortality due to CVDs have been documented, mainly in developed countries but also in some Latin America countries. [4][5][6][7][8] Socioeconomic indicators have been associated with differences in mortality, incidence and prevalence of cardiovascular risk factors, 1,7-8 and considerable disparities among countries in the Americas regarding CVD death rates have also been found. 8

OBJECTIVE
The objective of this study was to evaluate trends in mortality due to CVD in three different regions of the Americas.

Case definition and source of data
Data were obtained from the Mortality Information System (updated in August 2012), which comprises mortality databases provided to the Pan-American Health Organization (PAHO) by the countries in the Americas. To ensure comparability among countries with different qualities of information, we used the data corrected for under-reporting and ill-defined deaths, in accordance with published methodology. 9 This study included all the deaths for which the underlying cause of mortality was coded in Chapter IX, Diseases of the Circulatory System (I00-I99), of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10).
Premature mortality was defined as deaths that occurred among people aged 30-69 years old. For this definition, the life expectancy in the countries of the region of the Americas 10 and a publication from the World Health Organization (WHO) 11 were taken into account.
The criterion for including countries in this study was that they needed to have a time series available for the study period. The countries were distributed geographically into Age-adjusted rates were calculated using estimates from the United Nations Population Division 12 and the WHO Standard Population 13 (direct standardization). This study analyzed anonymous secondary data on mortality and therefore no ethics approval was required.

Data analysis
This was a time-series study. We began with a descriptive analysis on 16,940,728 deaths due to CVD registered in the system, focusing on sex, age, sub-region of residence and year of occurrence. To explain the trend patterns, a subsequent analysis was performed using the Joinpoint Regression Program, version 4.0.1 (January 2013), from the National Cancer Institute (http://surveillance.cancer.gov/joinpoint/). The dependent variable was the age-adjusted rates and the independent variable was the year. This software identified points at which the slope of the linear trend changed significantly, 5 by using the age-adjusted rates from 2000 to LAY to fit a regression line to the natural logarithm of the rates. It calculated the average annual percentage change (AAPC), which is a summary measurement of the trend over a fixed time period, taking into account the trend transitions. 14 The AAPC was considered to be significant when the slope was different from "zero" at alpha = 0.05. The significance tests used a Monte Carlo permutation method.

RESULTS
During the study period, in all the countries included in the study, females accounted for 51.0% and males for 49.0% of all CVD deaths. Although the average number of deaths was greater among females, the age-adjusted rates were not. In all regions, the rate among males was higher than the rate among female. Although the average number of deaths was greater in North America, the age-adjusted rates were not. Both in 2000 and LAY, the North American selected countries presented lower age-adjusted rates than the Latin American selected countries and Non-Latin Caribbean selected countries. Using the mortality rates for the latest year available, the risk of dying due to a CVD presented by a person who lived in the Non-Latin Caribbean region was 1.9 times the risk presented by a person who lived in the North American region.

Trend analysis
The results from the trend analysis (Table 1 and Figures 1, 2 and 3) showed that CVD mortality has declined in the three regions of the Americas. In North America, the overall age-   The results from the joinpoint analysis (

DISCUSSION
Our findings showed that mortality due to CVD has been decreasing in the three regions of the Americas, which is consistent with previous studies. 4   deaths, its age-adjusted rates were the highest. The trend curves for males and females followed the same pattern as shown by the total population curve (Figures 1, 2 and 3).
other treatments) and approximately half to changes in risk factors (reductions in total cholesterol, blood pressure, smoking and physical inactivity). 15 For some Latin America countries for which published trend series studies are available, the decline took place recently, or there was no decline or even an increase. [4][5]8 These countries have probably implemented community-based prevention programs more recently.
In addition, populations that live in countries with more disadvantaged incomes might face greater difficulties in obtaining access to treatment, compared with those who live in North American countries. Some countries in the Americas, such as Argentina, Brazil, Cuba and Canada, have public universal health coverage. Despite this, in Brazil, a study found that premature mortality due to circulatory system diseases was 2.6 times higher in poor areas in Porto Alegre than in rich areas. 16 In São Paulo, the decline in the risk of death due to heart disease was faster for people living in the wealthiest areas and slower for people living in lower-resourced neighborhoods. 17 In the United States, which has a different health system and the percentage health insurance coverage is around 83%, a study showed racial disparities in use of coronary artery bypass grafting and percutaneous transluminal coronary angioplasty among elderly women and men, which was probably due to access to care and financial barriers. 18 Regarding risk factors, a lot of work has to be done in the region. In the USA, around 90% of the people consume sodium at levels above the recommended guidelines. 19  was around US$ 314 billion, and the indirect costs due to disability and deaths was around US$ 161 billion, which raised the total cost of these diseases to US$ 475 billion. 23 The main limitations of the present study are the following: 1) although the rates were corrected for under-registration of deaths and ill-defined causes, it is possible that these problems might affect the rates in some countries, especially those with lower resources; and 2) some lower-resourced countries have made efforts to improve the quality of data, including decreasing the numbers of ill-defined deaths, which might lead to an increase in circulatory system disease rates.
Since disease is a great contributor to poverty, the Americas should strive to bridge the gap in treatment and preventive control of CVD among its regions. Gold-standard prevention requires complementary clinical and community approaches and monitoring information regarding mortality, morbidity and prevalence of risk factors. Eliminating health disparities should be a goal in relation to CVD mortality reduction in the region.

CONCLUSION
Mortality rates due to CVD have been decreasing since 2000 in the North American, Latin American and Non-Latin Caribbean regions. Disparities in risk, premature mortality and trends were seen across these three different regions.