Lisboa – Portugal SEASONAL MORTALITY PATTERNS DUE TO DISEASES OF THE CIRCULATORY SYSTEM IN PORTUGAL

exposure to cold weather has negative consequences on human health. studies have been showing that the seasonality of mortality has an evident peak during winter months in european countries. however, the highest increases in mortality are registered in countries with mild winters. according to several studies winter deaths seem to be associated with low socioeconomic conditions. the main aims of this study are to identify the trends of mortality due to diseases of the circulatory system and excess winter deaths in portugal and to assess the geographical pattern of seasonal mortality. in the 20 years under analyses, mortality due to diseases of the circulatory system increased 38% during winter months when compared to the non-winter period. important regional disparities were found, the excess winter death index ranged from 21% to 48%, the central regions tend to have better results. our results indicate that although circulatory mortality is significantly decreasing in portugal, the vulnerability to seasonal cold weather remains as an important public health issue. these findings suggest that the exposure to cold weather is an important determinant of cardiovascular diseases that is still neglected in portugal.


INTRODUCTION
in portugal, mortality due to diseases of the circulatory system is decreasing at an important pace since the last decades in result of significant improvements in healthcare and life conditions [santana, 2014].however, still remains as one of the main causes of morbidity and death.in most european countries, mortality has an important seasonal variation, characterized by a peak during the winter months and a decrease in rest of the year [healy, 2003], this pattern is even more significant when the analysis is focus on diseases of the circulatory or respiratory system [eurowinter Group, 1997;Rau, 2006].cardiovascular mortality is strongly related with weather conditions, being the cause of death with higher increases due to temperature variations.the increases are recorded both during high and low temperatures, although the number of excess winter deaths (ewd) is higher than the heat related mortality.exposure to lower temperatures has important biological effects.within certain thresholds, the human body can maintain thermal comfort through appropriate thermoregulatory responses so that physical and mental activities can be pursued without any detriment to health [xu et al., 2012].however, when the human body is exposed to temperatures below the biological optimum the circulatory system becomes under stress, causing an increase in blood pressure, blood viscosity and fibrinogen concentration [Keatinge, 2002;maheswaran et al., 2004;stocks et al., 2004].
healy [2003], mentions the "paradox of excess winter mortality", as the higher mortality rates are generally found countries with milder winter where, all else equal, there should be less potential for cold strain and cold related mortality.this pattern suggest that excess winter mortality should not be seen as an environmental inevitability [davie et al., 2007] thus, excess winter mortality should be seen as an important public health issue that demands proper measures and policies [monteiro et al., 2012]. in some countries, such as scotland and united Kingdom, ewd are seen as a public health priority [Gemmell, 2001], and specific actions to mitigate it are supported.Reduction of fuel poverty, monitoring and subsidizing the vulnerable homes are some examples of the national initiatives developed at a national level.unlike other countries, in portugal seasonal mortality and morbidity is underrated as a severe threat to public health, there is no systematic quantification or measures to tackle the vulnerability to cold weather.this paper aims to: i) characterize the evolution of overall and seasonal mortality due to diseases of the circulatory system in portugal (since 1990 until 2009); and ii) identify the geographical pattern of excess winter mortality (in the ten years period 2000-09).

DATA AND METHODS
this study uses standardized deaths Rates due to diseases of circulatory system, obtained from the european health for all database (http://data.euro.who.int/hfadb/), and monthly deaths through diseases of the circulatory system (icd 10: i00-i99), obtained from statistics portugal (www.ine.pt).data was collected from 1990 until 2009.deaths by month were standardized to 30-days with an adjustment for leap years, and analysed by Nut iii.
the number of ewd and the ewd index were calculated according to Johnson and Griffi [2003] in order to quantify the winter mortality burden.
the number of ewd is calculated by comparing the number of deaths in winter months (december to march) with the average number in non-winter months (the previous august to November and the following april to July): once the number of ewd was found, the ewd rate was calculated in portugal (from 1990 to 2009) and each Nut iii (ten years period 2000-09) using the resident population available in statistics portugal.
the seasonal impact of winter was assessed through the ewd index.it indicates whether there are higher than expected deaths in the winter compared to the rest of the year: and their 95% confidence intervals: winter deaths Non-winter deaths 1 1 1,96 winterdeaths Non-winter deaths cl 2 1.
the joinpoint regression model was used to identify significant changes cardiovascular mortality, ewd rate and ewd index.this method detects statistically significant changes in trend, and the annual percent of change is computed by each of those trends by means of generalised linear models assuming a poisson distribution [Fernandez et al., 2001].the joinpoint analyses were performed using the "Joinpoint" software from the us National cancer institute (available at http://surveillance.cancer.gov/joinpoint/).

RESULTS
portugal is located by the atlantic ocean (Fig. By analyzing the monthly distribution of the number of deaths due to diseases of the circulatory system an important seasonal pattern was identified.the winter months (from december to march) have the highest number of deaths.January is the month with higher proportion of deaths (11.5%) while september had the lowest 6.5% (Fig. 3). the ewd rate is significantly decreasing since 1990 with an annual percent change of 3.43% (p < 0.05), however this trend is not constant due to important inter-annual variations.the highest values were recorded between 1999 with 749 per 1.000.000inhabitants (Fig. 4 a).
the ewd index is not decreasing significantly; the evolution in the 20 years under analysis is marked by the inter-annual fluctuations.the highest value was recorded in 1999 when the number of deaths in winter months was 61.3% higher than the number of death recorded in the rest of the year (Fig. 4 B).

GEOGRAPHICAL PATTERN
winter mortality increases were recorded throughout the country, although, the seasonal burden varied remarkably across regions.the ewd rate through circulatory diseases varied from 249 to 742 deaths per million inhabitants, while the country average is 378 per million inhabitants, the worst results were in the inland and southern regions, while the best were recorded at the coastal areas (Fig. 5).
the ewd index country average is 40%, but important regional disparities were found (ranging from 21% to 48%), the central regions of the country tend to have better results, even though a clear pattern was not identified.during winter-moths the number of deaths was 40% higher than in non-winter months, two regions had significantly lower ewm index than the portuguese average and any region had significantly higher values.

DISCUSSION
Joinpoint analyses allowed the identification of significant decreasing trends, between 1990 and 2009, in the standardized death rates due to diseases of the circulatory system and ewd rate.despite these trends, the ewd index is not decreasing significantly.the geographical pattern shows important regional disparities, the ewd rate tend to be higher in the regions near the border and the ewd index is lower in the central regions, although is difficult to identify a clear pattern.
the improvements in healthcare and living conditions can explain the significant decrease in the standardized death rate due to diseases of the circulatory system.these diseases have been the major cause of death in europe for some decades and therefore several european countries have undertaken health policies to control this public health burden [levi et al., 2002].the gap between portugal and the other european countries is diminishing but still has higher mortality rate than the eu ' 15 average [santana, 2014].
there is an uneven distribution of mortality throughout the year in portugal, the mortality seasonal pattern shows a clear cluster during winter months, this trend is usually found in several european countries [Rau, 2006], and a B  with the improvements in socioeconomic and healthcare conditions it would be expected to assist a significant decrease in both ewd rate and index.although the decreasing trend observed in the standardized death rate was followed by the ewd rate there is no significant trend in the ewd index. in england (since 1991), the ewd index does not have a significant decreasing trend either [office for National statistics, 2013].

CONCLUSIONS
these findings highlight that although significant efforts have been developed to reduce the impacts of cardiovascular diseases in portugal, their effects on the disease seasonality are still far from being achieved.one possible explanation is that other epidemiology aspects of the cardiovascular diseases need to be addressed, such as the indoor environment and the reduction of cold exposure.
Nevertheless, the seasonal pattern in the mortality by circulatory system diseases is highly evident and it is still not properly addressed by health authorities in portugal.
1), according to Köppene and Geiger the southern regions are classified as a hot-summer mediterranean climate (csa) and the northern as warm-summer mediterranean climate (csb), both with mild winters [vasconcelos et al., 2011].TREND ANALYSISsince 1990, a decline in the standardized death rate due to diseases of the circulatory system has been observed.two points, where the rate changes significantly, were found and three periods with significant decreases (p < 0.05) were identified: 1991-03; 2003-06; 2006-09.thesecond period (2003-06) had the higher decreasing rate (annual percent change of 7.95%) (Fig.2).

Fig. 4 .Fig. 5 .
Fig. 4. Trends of Excess Winter Death rate (A) and index (B) due to diseases of circulatory system