How to cite this article: Heredia-Olivera K, Oscar Munares-García O. [Maternal factors associated with low birth weight]. Rev Med Inst Mex Seg Soc 2016 ;54(5):562-7.
Received: May 12th 2015
Accepted: July 13th2015
Karen Heredia-Olivera,a Oscar Munares-Garcíab
aFacultad de Medicina Humana “San Fernando”, Universidad Nacional Mayor de San Marcos, Lima, Perú
bSuperintendencia Nacional de Salud, Ministerio de Salud, Lima, Perú
Communication with: Karen Heredia-Olivera
Telephone: (051) 9995 83479
Background: In Peru, low birth weight is an indicator of risk of perinatal problems and childhood, the study objective was to determine maternal factors associated to low birth weight.
Methods: Case-control study in 123 low birth weight (cases) and 123 normal-weight newborns (controls) matched for date of birth and district. Associated maternal factors were investigated to low birth weight and compared between cases and controls. Chi square was applied, Odds ratio (OR) with confidence intervals at 95% (95% CI), binary logistic regression and ROC curve.
Results: Partnerships for maternal history of low birth weight (OR: 41.1; 95 % CI: 5.5-306.7) were found; prematurity (OR: 12.0; 95% CI: 1.5-94.3), history of eclampsia (OR: 5.8; 95% CI: 1.9-17.4), one to three prenatal controls (OR: 5.7; 95% CI: 2.6-12.3), multiple gestation (OR: 4.7; 95% CI: 1.3-17.0) and tobacco consumption (OR: 3.8; 95 % CI: 1.5-9.8), not being a teenager (OR: 0.3; 95% CI: 0.1-0.6), and not having a short birth interval (OR: 0.2; 95% CI: 0.1-0.7). In multivariate analysis, we pointed out that having from 1 to 3 prenatal visits, multiple pregnancy, being a teenager and a short birth interval are associated with low birth weight, the proposed model explained 18.6% of the event, the area under the curve was 72.9%, considering that adequately predicts (p ˂ 0.001).
Conclusions: There are maternal risk factors associated with low birth weight in pregnant women in urban areas of the coast as having from 1 to 3 prenatal visits, multiple gestation, being a teenager and having a period between pregnancies shorter than two years.
Keywords: Newborn infant; Premature infant; Low birth weight Infant
It is estimated that each year about 20.5 million children are born in the world with low birth weight, mostly in developing countries.1 Low birth weight has been defined by the World Health Organization (WHO) as the weight of a neonate below 2500 g during the first hour of life.2 Birth weight is an indicator that predicts the likelihood of perinatal survival.3 The rate of low birth weight is 15% for developing countries and 7% for developed countries;3,4 in Peru (2013) at the population level it is 7.3%,5 and in hospitals it was 8.24 per 100 live births.6 The annual prevalence of low birth weight in China was 6%, so more than a million cases occur in this country.7,8
Various maternal conditions during pregnancy are a predictor of risk of low birth weight, including antiretroviral therapy, being Black, anemia, urinary tract infection, preeclampsia, premature rupture of membranes, maternal weight gain below 8 kg, psychosocial factors, and smoking; other associated factors are elevated fetal glycosylated hemoglobin and prematurity, inadequate prenatal care, premature birth, primiparous mother, and less than 20 years old.
Several studies describe the association of low birth weight with some disorders during the first year of life, including delayed neurological development, cerebral hemorrhage, respiratory disorders, the risk of birth asphyxia, and other diseases that require prolonged hospitalization.5,6 The aim of the study is to determine maternal factors associated with low birth weight.
A retrospective, analytical case control study was done with deliveries diagnosed as newborns with low birth weight (cases), and a group of newborns between 2501-3999 g (controls), excluding macrosomic newborns and those with a syndrome, treated between 2010 and 2011 in the Instituto Nacional Materno Perinatal (Ministerio de Salud), in the capital of Peru. For the determination of the sample, the formula for comparative studies with a 95% confidence was applied, with a power of 80% and a 17.9% prevalence of factors in cases; assuming the highest prevalence for controls (50%), 121 cases were obtained. By seeking cases in the years included, 123 cases were obtained, so it was decided to take the total and match cases with controls by date of birth and district of origin, the total sample being 246, divided into 123 cases and 123 controls.
The study measured: maternal age, marital status, education level, number of pregnancies, birth interval or period between births (years), considering the birth interval short when it was less than 2 years, number of prenatal visits, type of delivery (vaginal or cesarean), as well as sex, weight (g), and gestational age (weeks) of the newborn. The study also measured as factors: maternal history of low birth weight, prematurity, eclampsia, multiple gestation, tobacco use, history of urinary tract infection, maternal anemia (Hb 11.0 g/dL), hypertension, preeclampsia, and eclampsia.
Being retrospective research, the documentary observation technique was applied, requesting the medical records of cases and controls, and placing all the information on a record sheet for subsequent digitization. The study was approved by the Ethics Committee of the Instituto Nacional Materno Perinatal (Letter No. 185-DG/No. 567-OEAIDE-INMP-13); the information was used anonymously, i.e. no maternal identification data were recorded for the study.
For the data analysis, frequency distribution and percentages were applied in qualitative variables, determining the mean and standard deviation for maternal age. Chi-squared and Fisher's Exact tests and the odds ratio (OR) with 95% confidence intervals were applied to determine risk factors in the bivariate analysis. To determine the model of risk factors, binary logistic regression was applied to calculate the adjusted OR (AOR) with 95% confidence intervals; the Hosmer-Lemeshow test was applied to determine the goodness of fit, considering the model appropriate when the p-value was < 0.05. The Cox-Snell R2 test was applied to indicate the predictive level of the model, and finally, to determine the discrimination power of the predictor variables, the ROC curve was applied to the estimated probabilities by applying logistic regression under the method of introducing its confidence intervals and statistical significance p < 0.05.
The average age of pregnant women was 26.6 ± 6.9 years. The highest percentage was between 26 and 42 years (52.6%). The most frequent marital status was cohabiting 42.3%; most have a secondary school education (71.5%), and 89.6% had between one and three pregnancies, 76% presented a birth interval between 1 and 4 years; 59.8% had between four to six prenatal visits, and only 21.1% presented more than 6 visits, and 56.9% of deliveries culminated in caesarean section (Table I).
|Table I Logistic regression model for factors associated with low birth weight|
|p||AOR||95 % CI|
|Lower L.||Upper L.|
|1 - 3 prenatal checkups||0.001||6.1||2.7||13.9|
|Short birth interval||0.017||0.2||< 0.1||0.7|
|Overall sensitivity: 66.3 s% Hosmer-Lemeshow test = 0.973; gl = 1; p = 0.914 Cox-Snell R2= 0.185|
Bivariate analysis found significant associations for maternal history of low birth weight: prematurity, history of eclampsia, one to three prenatal checkups, multiple gestation, and tobacco use. On the other hand, not being a teenager and not having a short birth interval were protective factors. No statistical significance was found for history of urinary tract infection, anemia, hypertension, male newborn, preeclampsia, and single marital status (Table II).
|Table II Factors associated with low birth weight. Instituto Nacional Materno Perinatal, Lima, Peru|
|Low birth weight||OR||95% CI||p1|
|(n = 123)||(n = 123)|
|Maternal history of low weight|
|History of eclampsia|
|1 - 3||30.9||7.3||5.7||2.6-12.3||˂ 0.001|
|History of urinary infection|
|History of anemia|
|History of hypertension|
|History of preeclampsia|
|Short birth interval|
|1 Level of statistical significance for Chi-squared or Fisher's Exact test|
Multivariate analysis found that between one and three prenatal visits, multiple gestation, being a teenager, and a short birth interval are associated with low birth weight. The proposed model had an overall sensitivity of 66.3%, and this explained 18.6% of events (Cox-Snell R2 = 0.185); the Hosmer-Lemeshow test indicated that the overall fit of the model was adequate (p = 0.914).
The ROC curve was used to determine the discriminating power of the predictor variables, which coincides with the probability of correctly distinguishing a case of low birth weight that is not, using the predictor variables, the worst case scenario being when the area is equal to 0.50. In our case, having between 1 to 3 prenatal visits, multiple pregnancies, a teen mother, and short birth interval less than 2 years, represented an area under the curve of 0.729 (95% CI: 0.667-0.791), finding that they adequately predict cases of low birth weight (p < 0. 001) (figure 1).
Figure 1 ROC curve for determining the sensitivity of the model of factors associated with low birth weight. Area = 0.729, Lower L.= 0. 667, Upper L. = 0.791, p < 0.001
In the present study we found association of low birth weight with four factors: having 1 to 3 prenatal visits, multiple gestation, a teen mother, and a short birth interval. Prenatal visits are an essential means of bringing health professionals into contact with mothers, and properly trained health personnel can promote preventive activities during gestation and encourage healthy diets.13 One of the points that best exemplifies prevention schemes is prenatal checkups; our study found 6 times the risk associated with having one to three prenatal checkups, another study found a relationship of almost twice the risk (OR: 1.9).6 An association was also found in Argentina, but when less than five prenatal checkups were done.9 This shows that preventive work in prenatal care is not being properly managed; in Colombia, Caceres Manrique10 argues that there are still barriers for pregnant women to attend prenatal checkups, it is assumed that the goal of checkups is to prepare women for motherhood and parenting, detecting risks on time, which will not be done due to multiple factors, including the lack of adherence to prenatal care; the same report indicates that this adherence should be expressed not only in visits but in compliance with the recommendations. Follow-up with the mother is flawed; a study in Cuba assessed the quality of prenatal care for pregnant women with low birth weight, and found that the level of knowledge about low birth weight in professionals who attend the antenatal clinic was inadequate (60% when the standard was 90%), and compliance with the flow chart of insufficient weight gain was 20%, when the standard was 90%.11 This is more difficult for those with few prenatal visits; one should study the reasons why pregnant women do not complete their prenatal checkups. A study in Peru found that providing incomplete services, not providing follow-up appointments, lack of coordination between services, lack of prenatal care, and little time for prenatal visits were associated with fewer prenatal visits.12
In Brazil, a study found that the increase in cases of low birth weight is due to the increased rate of multiple births and reduced rate of fetal death (500-999 g).4 The study found an association of nearly 6 times more for low birth weight (AOR: 5.7; 95% CI: 1.4-23.0). According Gallardo et al.,14 low weight is due to two main causes: birth before term, or the fetus presenting insufficient weight in relation to gestational age. In multiple gestations both events are presented due to the multiple gestation itself: fetuses rarely reach term and, therefore, the result of low weight is likely; another event leading to more multiple gestations is described by Kushner-Davalos,15 who maintains that a large number of people delay pregnancy due to sociocultural factors, preferring personal and professional development before having children, and therefore resort to assisted fertilization, making multiple gestations more likely.
Latin America remains the region with the highest number of teenage pregnancies, after sub-Saharan Africa. Countries with teen pregnancy above 13% are Argentina, Bolivia, Colombia, Ecuador, Guatemala, and the Dominican Republic, whose prevalence has increased in in recent years; the group of countries that have declined in prevalence include El Salvador, Honduras, Nicaragua, Panama, and Venezuela.
Countries with less than 13% of teen pregnancy that have lowered their prevalence are Brazil, Costa Rica, Haiti, Jamaica, Peru, and Uruguay; Mexico, which belongs to the group with less than 13% prevalence, is the only country where there has been an increase.16
A systematic review found that age under 20 years was associated with low birth weight in Latin American countries; the mechanisms that explain this event include young women with immaturity of the reproductive system and emotional immaturity.17 A case-control study in 380 adolescents in Argentina found a low birth weight ratio of 8.8%, compared to 8.4% in pregnant adult women.18 With regard to the factor of adolescence, an association of 0.3 was found (95% CI: 0.1-0.7); this means that not being an adolescent reduces the risk of low birth weight by 70%; other studies find this element as a risk factor.19
Castilla et al.20 argue that the worst results of low birth weight are more common in children of adolescents than those of adults. Morí Quispe et al.,21 said that mothers with underweight children are not prepared to handle situations that may threaten the lives of their babies, or situations with a high risk of sequelae threatening their autonomy, as these are more common in teenage mothers. Another event that would happen in pregnant teens is poor diet. Garces and Gómez22 point out that in malnourished pregnant women, and mainly in adolescents who have not completed development, there is an inadequate maternal-fetal exchange, as well as abnormal metabolism of proteins, lipids, carbohydrates and minerals in the mother that leads to the underutilization of nutrients by the fetus, affecting its development. In Argentina, Salcedo et al. 9 found that more than 60% of mothers who showed insufficient weight increase during pregnancy or deficient pre-pregnancy BMI, had children with poor nutritional status. Ariza et al.23 in Colombia, identified that teenage pregnancy occurred more frequently in disadvantaged social sectors, and that pregnant girls and young women recorded poor nutritional status, increasing the risk of low birth weight among other diseases. This group is also prone to unwanted pregnancy: in Peru the rate of unwanted pregnancies in pregnant women with infants with low birth weight is 30.2%.21
Another factor found was having a short birth interval (AOR: 0.2; 95% CI < 0.1-0.7), i.e., a period between pregnancies exceeding two years would reduce the risk of low birth weight by 80%. Allowing little recovery time for the uteruses of women with previous pregnancies would be the underlying cause, as subjecting the uterus to another pregnancy in less than two years’ time creates an unfavorable environment for new pregnancy. A study in Cuba found low birth weight 3 times more likely with a short birth interval (OR: 3.09), although the data were not significant.24 In Spain, a study found that 64% of pregnancies with birth intervals less than 24 months had preterm delivery.25
The present study could find no association of low weight with factors such as urinary tract infections, anemia, hypertension, preeclampsia, and female newborns; this is due to sample size, although the study took all the population at the Institute for two years.
Although it is an institution that annually attends more than 17,000 deliveries, the study also failed to find pregnant women with diabetes with a baby with low birth weight in that period. We would also like to indicate that because the population attending the Institute is mostly from low to middle-class districts, the population is usually young, so these events are likely to occur less frequently, which did not allow us to find a significant association.
We believe that these four factors: having 1 to 3 prenatal visits, multiple gestations, being a teenager, and having a short birth interval, are predictors of low birth weight and should be part, together, of the training, prevention, and monitoring process for pregnant women; these four factors that adequately predicted the event in the proposed model, have predictive probabilities as well.
For these reasons, the factors that can be handled by the health system would be those regarding: prenatal checkups, which must have clear objectives for each visit, especially in pregnant women with low weight; birth interval, because this event is a function of family planning among couples, which may not be functioning properly, perhaps because, once they have given birth, few pregnant women use a family planning method, and they do not go to their postpartum visits, thus postponing appropriate reproductive management.
We can conclude that there are maternal risk factors associated with low birth weight in pregnant women in urban areas of the coast such as: having 1 to 3 prenatal visits, multiple gestations, being a teenager, and having less than two years between pregnancies.
Conflict of interest statement: The authors have completed and submitted the form translated into Spanish for the declaration of potential conflicts of interest of the International Committee of Medical Journal Editors, and none were reported in relation to this article.