Annex: Maternal anthropometry: its predictive value for pregnancy outcome.

Introduction The Colombian study was undertaken in Cali, the third largest city in the country with a population of about 1 500 000. It has one of the best regionalized medical facilities in the country; most of our data were obtained from the tertiary-level hospital, a referral centre for most of the perinatal problems that appear in the urban area. In a sample of 5038 mothers with recorded pre-pregnancy weight and height during 1989, 1320 had information of weight gain during the pregnancy period. In addition to this sample from the high-risk hospital, a further sample from one of the six low-risk delivery hospitals was also studied for neonatal outcomes only - since by definition the later does not have maternal complications. A total of 527 mothers who had information on their pre-pregnancy weight and height during the year 1989 were chosen from this low-risk delivery group. It is worth noting that neither the mother's weight nor height appears as a risk indicator on the clinical record proposed by CLAP (Centro Latino (Americano de Perinatologia y Desarrollo Humana) (1) and routinely used in these hospitals; this can result in mothers with a low pre-pregnant weight and low height appearing in either the high-risk or low-risk delivery hospitals. Study objectives. The main objectives of the study were to investigate associations between maternal anthropometry (height, pre-pregnancy weight, and weight gain during pregnancy), and the presence of perinatal complications in the mother or neonate. Several categories of infant outcome were considered: full-term LBW infants (FT-LBW), all LBW infants, and pre-term deliveries, to see if it was possible to establish cut-off points for the anthropometric parameters in order to use them as risk indicators of the perinatal outcome. Methodology Since 1986 a number of workshops were held in all the hospitals in Cali to teach nurses and doctors to fill out the CLAP medical record in the follow-up of pregnant women and their newborns. For this study all medical records at the tertiary-level hospital were searched for the data on pre-pregnancy weight and mother's height. Of the 10 332 women attending, 5038 had the relevant information and these were selected for the study. In the low-risk delivery hospital, 527 out of 1976 deliveries in 1989 met the criteria for entry into the study. Length of pregnancy was determined by the date of LMP and by the Dubowitz score and if these disagreed by more than 2 weeks, the case was dropped from the study. Pre-pregnancy weight was based on patient recall. The definitions of study variables used here were the same as reported in the meta-analysis. Results and discussion Maternal anthropometry is an indicator of risk for pregnancy complications and neonatal problems when it is associated with other perinatal risk factors. Anthropometric data alone are not per se good risk indicators according to our findings. This hypothesis is very important from the public health point of view and will need to be further confirmed or rejected by prospective studies undertaken in communities where both high- and low-risk deliveries can be compared for maternal and newborn outcomes. As far as we know, no such study has been undertaken in developing countries where perinatal risk is high, especially in relation to IUGR. The percentage of mothers found to be of low pre-pregnancy weight (WTpp) was similar in the low-risk hospital (WTpp