La mortalité différentielle aux âges adultes et avancés selon le groupe linguistique au Québec : une étude de suivi sur la période 1991-2011

The author explains that Quebec’s unique linguistic situation in North America leads us to question the effect of belonging to a linguistic group on observed inequalities in mortality. She notes that the new 1991 Canadian Census Health and Environment Cohorts (CanCHEC) database, which is derived from the pairing of the 1991 census, vital statistics and administrative files from 1991 to 2011, provides an unparalleled opportunity to study mortality differentials by language in Quebec with more accuracy. The main objective of the author's work is to compare the distributions of deaths by age and language most often spoken at home (i.e. French, English or another language). She begins by smoothing the age-specific mortality rate curves by sex, language and region using the P-splines approach. From the smoothed density functions, she derived four indicators of adult mortality and old-age mortality which are the life expectancy at age 25, at age 65, and the modal age (i.e. most common age) at death as well as the dispersion measure associated to the mode. Her results show that in Quebec, there are indeed significant differences between linguistic groups in adult mortality, except for Francophone and Anglophone women. However, she notes that the mortality gaps for the life expectancy at age 25 are less important than stated in the literature. The author found that although these observed differences in adult mortality persist among those 65 years of age and older, they do not carry on among those who reach the modal age at death, with the exception of Allophone men who retain their mortality advantage in very old age. In addition, the dispersion of deaths beyond the modal age is not significantly different between linguistic groups. She also found different levels of mortality between the Island of Montreal and the rest of Quebec, for Francophones and Allophones. The author points out that her results shed new light on the subject given the greater reliability of the linguistic variable from the 1991 Census in CanCHEC, and the improvements made in terms of data processing and analysis, which include the consideration of emigration and the Indigenous population, and region-based analyses.

Master's Thesis

Université de Montréal

Montréal

2019

xiv-104p.

fr