Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Dr. William Avison

2nd Supervisor

Dr. Alain Gagnon

Joint Supervisor


Previous research has shown larger than expected numbers of deaths at the age of 28 during the 1918 Spanish influenza pandemic in Canada. To analyze whether this was related to the Russian influenza pandemic that occurred 28 years previously in 1890, the Western, McMaster, Montreal Influenza Pandemic (WMMIP) database was created. It utilizes the death records of 3,316 individuals who died in Ontario between the ages of 23 and 35 from September to December, 1918, and who were also born in Ontario. These were linked to birth records, the 1901 and 1911 Canadian censuses, marriage records, and attestation papers.

A reconstructed date of birth was created for each individual to analyze date of potential exposure in 1890. Those who were in utero in 1890 died in greater numbers than would be expected and those in the first trimester of gestation had an unusual sex-ratio at death. Of the various hypotheses proposed to account for the high young adult mortality, these data most closely support that of antigenic imprinting. There is cautious support for the fetal growth restrictions hypothesis, but these data do not support the scarring mechanism. Further, these data do not support the hypothesized relationship between tuberculosis infection and influenza mortality.

More individuals left agricultural homes of origin among the decedents than among the Ontario population in general. There were proportionally more French Canadians, more catholic individuals, and more people from Eastern Ontario. The decedents also came from larger families than were found in the general population, although this may be an artifact of the records linkage process. This research shows that the mortality pattern in Ontario during the pandemic was similar to what it was prior to the epidemic: mortality continued along the fault lines in society and did not equalize risk in a “democratic” manner.

The extant records are appropriate for historical demographic analyses and the strength and weaknesses for each are detailed. As expected, individuals from the north, who were aboriginal, from smaller families, or in transient occupations were harder to link.