TY - JOUR
T1 - Geographic variation in vascular mortality in Eurasia
T2 - Spatial autocorrelation analysis of mortality variables and risk factors
AU - Peter, I.
AU - Otremski, I.
AU - Livshits, G.
PY - 1996
Y1 - 1996
N2 - The geographic variation patterns of vascular mortality and their major risk factors from 68 samples in Eurasia are described in this work. The goodness-of-fit tests and analysis of variance indicate significant differences in incidence of mortality from the studied diseases, as well as in risk factors among the various geographic regions in four age groups. Correlation analysis points out the two general tendencies for the majority of studied traits: (a) significant positive association with latitude and (b) significant negative correlation with longitude. In turn, one-dimensional correlograms showed no specific geographic pattern at least up to 3000 km for all studied variables. However, at the large geographic scale a long-distance differentiation pattern was indicated for total serum cholesterol and body mass index; regional patches - for total death rate, mortality rate from cardiovascular and ischaemic heart diseases, systolic and diastolic blood pressure; and a local patches pattern was detected for mortality rate from stroke. Two-dimensional correlograms uncovered three distinct and significant patterns of variation: (a) a north-south trend for total mortality rate, for death rate from cardiovascular and ischaemic heart diseases, for diastolic and systolic blood pressure and for body mass index; (b) a northwest-southeast pattern for mortality rate from cerebrovascular disease; and (c) local patches for total serum cholesterol.
AB - The geographic variation patterns of vascular mortality and their major risk factors from 68 samples in Eurasia are described in this work. The goodness-of-fit tests and analysis of variance indicate significant differences in incidence of mortality from the studied diseases, as well as in risk factors among the various geographic regions in four age groups. Correlation analysis points out the two general tendencies for the majority of studied traits: (a) significant positive association with latitude and (b) significant negative correlation with longitude. In turn, one-dimensional correlograms showed no specific geographic pattern at least up to 3000 km for all studied variables. However, at the large geographic scale a long-distance differentiation pattern was indicated for total serum cholesterol and body mass index; regional patches - for total death rate, mortality rate from cardiovascular and ischaemic heart diseases, systolic and diastolic blood pressure; and a local patches pattern was detected for mortality rate from stroke. Two-dimensional correlograms uncovered three distinct and significant patterns of variation: (a) a north-south trend for total mortality rate, for death rate from cardiovascular and ischaemic heart diseases, for diastolic and systolic blood pressure and for body mass index; (b) a northwest-southeast pattern for mortality rate from cerebrovascular disease; and (c) local patches for total serum cholesterol.
UR - http://www.scopus.com/inward/record.url?scp=0030292097&partnerID=8YFLogxK
U2 - 10.1080/03014469600004692
DO - 10.1080/03014469600004692
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
C2 - 8933913
AN - SCOPUS:0030292097
SN - 0301-4460
VL - 23
SP - 471
EP - 490
JO - Annals of Human Biology
JF - Annals of Human Biology
IS - 6
ER -