This study tests the hypothesis that size at birth is associated with diarrhoeal incidence from birth to 24 mo of age, and the hypothesis that diarrhoeal incidence from birth to 24 mo is associated with body size at age 24 mo. This is a longitudinal study of 1476 infants born in Lahore. Pakistan, in 1984-1987. Diarrhoeal incidence was enumerated monthly. A generalized estimating equations approach with Poisson regression showed that birthlength standard deviation score (SDS) and ponderal index at birth (rate ratio = 1.01 and 1.00; each p > 0.05) were not significantly associated with diarrhoeal incidence. Multiple linear regression showed that diarrhoeal incidence was significantly associated with weight SDS and body mass index at 24 mo (regression coefficient or beta = -0.58 and -1.02: each p < 0.05). but not with height SDS (beta = -0.10: p > 0.05). The associations were mainly due to a transient effect of diarrhoea in the period 18-24 mo.
Conclusion: Prevention of low birthweight is unlikely to have an impact on diarrhoeal incidence in infants. Control of diarrhoeal incidence may not improve the growth of infants in developing countries.