MPIDR Working Paper

Differences in lifespan by month of birth for the United States: the impact of early life events and conditions on late life mortality

Abstract

We find significant differences in the mean age at death by month of birth on the basis of 15 million US death certificates for the years 1989 to 1997: Those born in fall live about 0.44 of a year longer than those born in spring. The difference depends on race, region of birth, marital status, and education: The differences are largest for the less educated, for those who have never been married and for blacks, and the differences are more marked in the South than in the North. They are only slightly larger for males than for females. For blacks, the shape of the month-of-birth pattern is significantly different from that of whites. We present evidence that this difference is due to whether one has an urban or a rural place of birth. We find a significant month-of-birth pattern for all major causes of death including cardiovascular disease, malignant neoplasms, in particular lung cancer, and other natural diseases like chronic obstructive lung disease, or infectious disease. We reject the hypotheses that the differences in life span by month of birth are caused by seasonal differences in daylight or by seasonal differences in temperature. Our results are consistent with the explanation that seasonal differences in nutrition of the mother during pregnancy and seasonal differences in the exposure to infectious disease early in life lead to the differences in lifespan by month of birth.