IJCS | Volume 31, Nº2, March / April 2018

125 Soares et al. Cardiovascular Mortality X Gross Domestic Product Int J Cardiovasc Sci. 2018;31(2)123-132 Original Article in 2000 by the census, stratified into seven age groups (20‑29 years; 30-39 years; 40-49 years; 50-59 years; 60‑69 years; 70‑79 years; and 80 years or older) for each sex. Those rateswere denominated compensated and adjusted. The mortality rates and GDPpc were correlated by estimating the Pearson coefficients of correlation 18 in all combinations of time series allowed to determine the optimal annual lag, according to the availability of socioeconomic data, which could be 29 years maximally. The optimal annual lagwas that with the highest Pearson linear coefficient in all series combined. In addition, the regression slope coefficients were estimated between the dependent variable mortality (DCS, IHD, CBVD) and the independent variable (GDPpc), multiplied by 100 dollars, in series with optimal lag, according to the coefficient of linear correlation. The quantitative analyses were performed with the Excel-Microsoft 19 and STATA programs. 20 Results The optimal GDPpc time lags (Table 1) with the mortality from DCS group and with the mortality from CBVD subgroup were very close, with respective means of 20.4 and 20.3 years in the Rio de Janeiro State; however, that with the mortality from IHD subgroup was lower, with a mean of 18.1 years. Regarding the regions, the highest time lags were of GDPpc with DCS in the Southern-Central region (mean of 24.3 years), and the lowest, of GDPpc with IHD in the Northern region (mean of 11.5 years). The highest time lag of GDPpc in the municipalities, which was 29 years, the maximum limit allowed by the data available, occurred with DCS in the municipalities of São Pedro da Aldeia, Paraíba do Sul, Vassouras, Nilópolis, São João de Meriti and Niterói; with CBVD, in Cabo Frio, Nilópolis, São João de Meriti and Niterói; and with IHD, in Vassouras, Nilópolis and Niterói. Some municipalities showed no time lag between the variable ‘mortality rate’ and GDPpc, which occurred with DCS in Porciúncula, with CBVD in Silva Jardim, Miracema and Porciúncula, and with IHD in Saquarema and Sapucaia. The coefficients of correlation (Table 1) of GDPpc with DCS and CBVD were closer to the extreme value (-1.0), with means of -0.84 and -0.83, respectively; however, the coefficients of correlation of GDPpc with IHDwere closer to absence of correlation (0), withmean of -0.62. The most extreme of those coefficients was that withDCS inNiterói (-0.99). Only the municipalities of São Pedro da Aldeia and Cambuci showed positive coefficients of correlation of GDPpc with IHD, +0.49 and +0.20, respectively, but closer to the level of absence of correlation. The evolution of the GDPpc in the Rio de Janeiro State municipalities over the past six decades showed a GDPpc elevation with heterogeneous distribution of the mean GDPpc values between the regions and the municipalities (Figure 1). The highest GDPpc values over the years were found in the capital of the Rio de Janeiro State, in Niterói, and in some more industrialized municipalities of inner state, such as Resende and Barra Mansa; and, in the past decade, in the coastal municipalities of the Northern and Coastal Lowlands regions, which concentrate the oil industry. The death variations at every 100-dollar increment in GDPpc (Figure 2) were higher in the group of deaths from DCS, because that group includes the two subgroups, CBVDand IHD, showing an importantmortality reduction related toGDP elevation. Suchmortality reduction related to GDPpc elevation was very heterogeneous: there are municipalities where a 100-dollar increment in GDPpc correlated with a reduction by more than 60 deaths from DCS, such as in Cordeiro, a municipality of the Mountain region. However, in only two small municipalities, with less than 40,000 inhabitants aged 20 years or older in 2010, São Pedro da Aldeia and Cambuci, the GDPpc elevation correlated with a mild increase in the number of deaths from IHD. In addition, in four municipalities (Valença, Niterói, Rio de Janeiro andNova Friburgo), the 100-dollar increment in GDPpc correlated with a higher reduction in deaths from IHD than from CBVD, a pattern that is opposite to those of the other municipalities, where the GDPpc increment correlated with a higher reduction in deaths from CBVD. Discussion Reductions in the mortality rates fromDCS have been shown in the Rio de Janeiro State municipalities for the past three decades. 21 In addition, GDPpc elevations have been observed in allmunicipalities studied (Figure 1) since 1950. They reflect the improvement in the socioeconomic indicators occurring all over Brazil, where the following aspects have been observed: income increase; mortality rate decrease; life expectancy increase; fertility decrease; childmortality reduction; and educational level increase resulting from illiteracy reduction. In addition, the improvement in the indicators in Brazil is also associated with the great income concentration. 22-24