Delayed excess mortality after exposure to measles during the first six months of life.

In an urban area of Guinea-Bissau, 71 children exposed to measles before age 6 months had a mortality risk of 34% (95% confidence interval (Cl) 24-47) between 6 and 60 months of age. The mortality risk for the 205 other children of the same birth cohort who had not been exposed to or developed measles was 11% (95% Cl 9-15), a significant difference compared with exposed children. With a version of the Cox regression model, maternal education was found to be the only background factor with a significant effect on mortality. When background factors were controlled for, the mortality of children exposed to measles was significantly higher than that of controls in each of the age intervals 6-11, 12-23, and 24-35 months. For the large subgroup of children of mothers without any formal education, exposed children had 5.7 times (95% Cl 2.7-12.0) higher mortality than did the control children in the age interval 6-35 months. Diarrhea deaths were particularly common among exposed children. Of 22 children who had been exposed before age 6 months during a subsequent epidemic and had a blood test taken, there was a significantly higher mortality risk (27%) between 6 months and 5 years than in the 26 controls who had a blood test (0%). Children who had elevated antibody titers to measles after exposure had a particularly high mortality compared with controls. These results suggest that later childhood mortality may be related to infectious experiences during the first months of life. The possible long-term health consequences of exposure to measles virus should be considered when assessing the value of measles control programs.