Summary_Danilova (1136935), страница 4
Текст из файла (страница 4)
M., Shkolnikov V. M., Begun A. Z. Algorithm for decomposition of differences between aggregate demographicmeasures and its application to life expectancies, healthy life expectancies, parity-progression ratios and total fertility rates //Demographic Research. 2002. № 7(14). P. 499-522.compensate for each other, resulting in a stable inter-regional inequality in Russiaduring the recent years.To understand how the difference between the top and the bottom 15% groups ofregions looks not only when the groups are compared to one other but also compared toother countries, we decomposed the difference in life expectancy between Poland andthe two selected groups of regions.
We chose Poland because like Russia this countryhas experienced unfavorable mortality trends since the mid-1960s. However, after thefall of the Iron Curtain, Poland and Russia took different paths. Poland entered thecardiovascular revolution13, while Russia experienced crisis mortality fluctuations overthe subsequent 15 years.Life expectancy calculated for the 15% of the Russian population which lives inthe regions of the highest life expectancy was only 2.0 years lower in males and 2.5years lower in females compared to Poland in 2014.
The gap in life expectancy betweenRussian leading regions and Poland is mostly attributable to the excess mortality inyoung and middle ages. It is only due to higher mortality at age interval 25-49 yearsmale life expectancy, calculated for the top 15% of the Russian population, thatmortality is 1.7 years lower than that estimated for Poland. The difference in female lifeexpectancy between the top 15% group of the Russian population and Poland is mainlydue to excess mortality at the older ages in Russia, although excess mortality in youngages also contributes to this gap.
Meanwhile, mortality from neoplasms in the top 15%group of the Russian population is lower than in Poland for both in males and females.The gap in life expectancy between the bottom 15% group of the Russianpopulation and Poland is much larger: 11.7 and 7.1 years for males and femalesrespectively. At ages 25-79 years in males and 30-79 years in females, each 5-year ageinterval contributes at least 0.5 years to the difference in life expectancy between thisgroup of regions and Poland. At ages 30-49 years this contribution accounts for more13Fihel A., Pechholdová M.
Between ‘Pioneers’ of the Cardiovascular Revolution and Its ‘Late Followers’: Mortality Changes in theCzech Republic and Poland Since 1968 // European Journal of Population. 2017. Vol.33. №5. P. 651-678than 1 year for each 5-year age interval. In males, the contribution of external causesand circulatory diseases is nearly equal: 4.0 and 4.1 years respectively (calculated forages before 80).
The contribution of excess mortality from external causes is generallymanifested at young and middle ages, whereas at middle and senior ages it is fromcirculatory diseases. In females circulatory diseases contribute the most to the gap inlife expectancy between the bottom 15% group in Russia and Poland, 3.0 years, whileexternal causes contribute the other 1.4 years.Thus, the “lagging” regions in Russia are experiencing the double burden ofdisease today. The “crisis” component of mortality – primarily, premature mortalityfrom external causes and circulatory diseases – is very high in these regions and still isan important factor of low life expectancy.
However, mortality decline in the elderly –the process which researchers claim to be a sign of the beginning of the cardiovascularrevolution in Russia – has affected these regions only slightly.Spatial patterns of mortality distribution in RussiaTo determine whether there is any spatial dependence of mortality distributionwithin Russia, we calculated Moran’s indices of spatial correlation14. Moran’s indexshows whether the values of some indicator (in our case, mortality) are closer to oneother in the neighboring territories. In the study, several ways of neighborhooddetermination (neighborhood matrices) between regions were tested:-the matrix of common borders – the regions are defined as neighbors ifthey share a border;-the matrix of “five nearest neighbors” defined by the distance between thecentroids – region j is a neighbor for region i, if region j is among five closest regionsfor region i, where the closeness is determined by the geographical distance between thecentroids of regions;14Moran P.A.P.
Notes on Continuous Stochastic Phenomena // Biometrika. 1950. № 37 (1). P. 17–23.-the matrix of “five nearest neighbors” defined by the distance between themost populated cities – this matrix is similar to the previous one, but the closeness isdetermined according to the shortest distance by car between the most populated citiesof the regions;-the matrix of the inverse distances between the most populated cities. Thecloseness was also determined according to the shortest distance by car between themost populated cities of regions.
When the most populated cities of the two regions arelocated less than 1000 km by car from each other the regions are identified asneighbors. Unlike the previous matrices which were binary (the neighborhood eitherexists or not), in this matrix the weight of neighborhood for each pair of regions wasdefined depending on the distance between the two regions;-the matrix of migration flows between regions. The neighborhood betweenregions was defined based on the data of cross-regional migration during the period2009-2012.For all matrices, Moran’s indices for life expectancy of birth were positive duringthe entire period of observation. It means that regions neighboring the regions withlower life expectancy tend to have a lower life expectancy as well.
And in turn,neighbors of those regions with higher life expectancy are more likely to also havehigher values of life expectancy. During most of the observation period, the highestvalues of Moran’s index were found for the matrix of common boundaries. The othermatrices do not explain spatial variation in mortality within Russia so well. The lowestestimates of Moran’s index were found for the matrix of migration flows.The change of Moran’s index of spatial autocorrelation over time is negativelycorrelated with the change in life expectancy at the national level and positivelycorrelated with the change in inter-regional inequality. It means that the periods of lifeexpectancy increase in Russia which were followed by the decline in inter-regionaldisparities also resulted in the increasing patchiness of mortality map.For the 23 groups of causes of death mentioned above, the mapping of mortalityindicators was applied.
It allowed visual inspection of mortality levels from thesecauses across the Russian Federation. Mortality rates for years 2013-2015 werecalculated for four age intervals: 0-19 years, 20-39 years, 40-64 years, and 65-79 years.Also, for the analyzed mortality components (combinations of age groups and causes ofdeath) we compared cross-regional distributions (weighted by population size) betweenthe period 2013-2015 and 1989-1991.
This allowed an analysis of how the proportion ofpopulation inhabiting the territories with lower/higher levels of mortality has changed inthe last quarter of a century.The analysis performed showed that the mortality components are different bothby their time trends and by the changes in inter-regional inequality. The comparison ofdistributions between 1989-1991 and 2013-2015 showed that even though mortality in2013-2015 was lower, it was not the same for all causes of death.
The negative trendswere especially pronounced in mortality changes form HIV, digestive diseases and alsofor some neoplasms. The increase in mortality from these causes also resulted in asignificant rise in inter-regional inequality, i.e., regions were unequally affected byunfavorable mortality trends.For most mortality components analyzed (by ages and causes of death), mortalitylevels in 2013-2015 are lower than in 2013-2015. However, inter-regional mortalityinequality often remains higher than in the turn of the 1980s and 1990s. It means thatsuccesses in mortality decline also affected the regions unevenly.Looking at the maps, it can be concluded that spatial regularities of mortalitydistribution across the territory of Russia which were described in the 1970s and 1980sstill persist15.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
