Classification of small-for-gestational age births: weight-by-gestation standards of second birth conditional on the size of the first.

Percentiles of weight-by-gestational age were constructed for first and second births, based on linked sibship-data from the Medical Birth Registry of Norway. Standards were made for weight-by-gestational age of second births conditional on whether the first birth was small-for-gestational age (SGA) or large-for-gestational age (LGA). These standards were compared with the conventional, cross-sectional standard of all second births. The relevance of the conditional standards was assessed on the basis of perinatal mortality, using logistic regression analyses. When applying cross-sectional standards of second births, more than 30% of the births following a SGA first birth were classified as SGA, compared with only 1.7% following an LGA first births. The overall risk for a perinatal loss in second births following a SGA first birth was twice that among second births following a LGA first birth. When second births were themselves categorised as SGA or non-SGA using the cross-sectional standards, the mortality among the SGA second births was such that the risk was 4 to 5 times higher following LGA first births compared with SGA first births. When conditional standards were applied to define SGA among second births, the risk relation between the subgroups (defined by classification of first birth) corresponded to the observed overall risk pattern. An unconditional SGA classification conceals important differences between clinically distinct subgroups.