Introduction Studies indicate connections between childhood conditions and health in old ages. Early conditions of life may be related to individual factors of development. Thus, rural areas can harbour states of child development different from those in the urban ambience.

Objective To evaluate the impact of rural origin on mortality of elders living in a urban region.

Methods Data are from SABE: a longitudinal survey in São Paulo—2000/2006. The explanatory variable “origin”, was obtained from the question: “Have you lived in the countryside for more than 5 years before the age 15?” Control variables are sex, age, education, income; having had malaria, pneumonia or typhoid fever before the age of 15. Outcome was “death”: people alive in 2000 (n=2143) and confirmed dead before or during the second round in 2006 (n=649).

Results Rao-Scott tests showed differences according to the origin of the elders for all variables except age and typhoid fever. A Poisson regression was applied to evaluate the associations between “origin” and “death” in the presence of all other variables, controlling for time of exposure. Incidence Rate Ratio (IRR) for rural origin was 1.54 (p=0.03). That is, having lived in the countryside before the age of 15 increased the death rate by 54%. Sex, age and income were also significant, with IRR of 1.79; 3.57 and 1.69.

Conclusions These results demonstrate that inequities such as rural-urban inequality, even occurring in the young ages, can perpetuate differences through a person's life, including an increased chance of dying when older.