Abstract Data and Methods
1 Introduction
One of the first to describe differences in life span according to month of birth was Ellsworth Huntington. In his book "Season of Birth" [10] he presents data on longevity by month of birth based on genealogical memoirs of families from different regions of the United States. For all regions he found that people born in February or March live decidedly longer than those born in July or August.

A study of Japanese males born before 1890 [14] found that those born between May and July have a lower life expectancy after age 70 to 75 than those born in other months of the same year. Recent research on babies born in rural Gambia indicates that those born during the dry season suffer higher mortality later in life than those born during the wet season [15] .

Huntington’s findings stimulated a long line of research on season of birth and the likelihood of mental disorders, e.g. schizophrenia (for a review of 250 studies, see [12, 16, 19, 21] ), Alzheimer’s disease [22] , and autistic disorder [3] . Most of these studies found significant differences in the risk of developing the respective disease according to season of birth. Findings on cancer patients have also revealed significant seasonality in life span according to month of birth [17, 24] .

Up to now studies about the relationship between month of birth and longevity have used comparatively small data samples. For example, Huntington’s results are based on genealogical information for about 39,000 persons. Some of the data samples were confined to selected groups: Miura’s findings are based on graduates of the Tokyo medical school and on asylum inmates, each group numbering about 400 people.

In this study I use data for total populations to investigate the question of whether or not month of birth and life span are related. I employ two different data sources for two different countries: Austria and Denmark. The Austrian data are based on vital statistics. The exact date of birth and the exact date of death are known for each person who died between 1988 and 1996. Only deaths after age 50 are considered. For the years 1990 to 1997 the cause of death is also reported. The Danish data consist of the total population aged 50+ on 1 April 1968. These people were then followed until their death or until August 1998.

The result of this study suggests that life span and month of birth are in fact related. Having determined this, I then pay particular attention to the question of whether or not this relationship can be explained by factors that become active at the end of one’s life. I consider two factors: first, the impact of the seasonal distribution of mortality on life span and, second, the "birthday-effect". The latter refers to the alleged tendency of people to die shortly after their birthday. I show in this study that neither factor can explain the relationship between month of birth and life span. Furthermore, I find cause to question recent results that found a correlation between birthday and the time of death (e.g. [5, 23] ).

The article closes with a discussion of possible mechanisms that may explain the relationship between month of birth and longevity.

Abstract Data and Methods

logo70.gif (2450 bytes)

Longevity and Month of Birth:
Evidence from Austria and Denmark

Gabrielle Doblhammer
1999 - 2000 Max-Planck-Gesellschaft ISSN 1435-9871