Socioeconomic Risk Factors for Adult Mortality in a Middle-Income Country: The Costa Rican Longitudinal Mortality Study 1984-2007

William H. Dow, University of California, Berkeley
Luis Rosero-Bixby, Universidad de Costa Rica

The Costa Rican Longitudinal Mortality Study (CR-LMS) is a unique new database that prospectively tracks 24-year mortality in a national sample of approximately 20,000 Costa Ricans who were aged 30 and over in the 1984 census. About 5,000 deaths had occurred by December 2008, mirroring death rates from national life tables. Numerous checks suggest that the death registry is extremely high quality; for example, the cohort that started over age 85 is now extinct, an indication that the death registry is complete. We estimate socioeconomic gradients in all-cause and cardiovascular adult mortality using hazard regressions, something rarely accomplished in a developing country population. The data indicate that SES gradients are weak throughout this 24-year period, even among the younger adults. College education is protective but rare, while completing primary or secondary education carries little mortality advantage. Adults in rural areas have lower mortality than in urban areas, an intriguing result.