LIMB REDUCTION DEFECTS IN SOUTH-AMERICA

Objective Limb reduction defects were considered as possible indicator s of environmental teratogenesis; it was suggested that also invasive prenatal procedures could increase the risk for limb reduction defects , The purpose of this work is to give a baseline frequency for limb re duction defects, using data from a population not exposed to prenatal diagnosis procedures. Design Using data collected in the period 1967 t o 1992 within the frame of the Latin American Collaborative Study of C ongenital Malformations which clinically examined 2,917,074 newborn in fants, a total of 1715 with limb reduction defects were found. All cas es were classified and analysed in 25 categories. Geographic differenc es in recorded rates were tested by chi(2) for homogeneity. Secular tr ends were analysed using chi(2) test for linear trends. Results The ov erall birth prevalence rate of limb reduction defects among liveborn i nfants was 4.91 (per 10,000 births) (3.05 for isolated and 1.85 for as sociated cases). For stillbirths, the total prevalence was 26.73/10,00 0 (5.53 for isolated and 21.20 for associated cases), The inclusion of the brachydactylies increased those figures to 5.55/10,000 (3.39 for isolated and 2.16 for associated cases), and 27.42, respectively, (5.5 3 for isolated and 21.89 for associated cases). When isolated and asso ciated cases were considered together, a geographic heterogeneity was found in pre-axial limb reduction defects; there was also some heterog eneity for amputations. A maternal age effect was found for the isolat ed hypoplasias. Standardising by maternal age, the overall prevalence of limb reduction in liveborn infants was 5.66 per 10,000 (95% CI = 5. 38-5.93). An increasing trend was suggested by the isolated form of di stal amputations which involved hands, feet, or digits. Conclusions Ou r data suggest that clustering limb reduction defects in wide groups a s transverse and longitudinal may lead to heterogeneous entities. When a possible association is suspected, it would be preferable to presen t and analyse data in the most discriminant form available. Due to the maternal age effect, it would be advisable to standardise the rates o f transversal limb reduction defects by this variable.