Low birth weight (LBW) remains a public health problem throughout the world. Globally, about 20 million births (18%) were reported with weights less than 2,500g, of which 96% were in developing countries. The worldwide incidence of LBW varied between 8%-20%, depending on the level of development of the countries ¹ with about 85% of all neonatal deaths being associated with LBW ^2,3. The most affected region is South Asia (22-25%), and the least affected region is the Western Pacific (5-6%) ^4,5. Approximately 40% of all LBW births in the world are registered in India. Since in many developing countries, babies are not weighed at birth, there is an estimated global underreporting of LBW of 58% ⁶.

Regarding low birth weight in the Americas, the countries that report the highest incidences are Honduras (15.0%), Dominican Republic (14.0%), Guatemala (12.4%), Ecuador (12.1%), El Salvador (9.4%), Colombia with (9.0%), Panama (7.8%), Peru (7.5%), Argentina (7.2%), Costa Rica (6.8%), Chile (6.2%), Paraguay (6.2%), Mexico (5.7% ) and Cuba (5.3%). In South America, the incidence ranges between 6.2% and 12.0% and in North America, the incidence is less than 8.3% ^7,8.

In Colombia, during 2016, 48,340 births were reported, of which 292 (0.6%) died due to some complication related to the LBW ⁹. Between the years of 2004-2015, LBW has exhibited an irregular behaviour and in some municipalities, it has decreased slightly ⁵.

The proportion of children with LBW in the Valle del Cauca department of Colombia was 6.5% ¹⁰, while the reported incidence in Cali was 9%, with Commune 14 having the highest prevalence of LBW (8.1%). During 2016, the Hospital Universitario del Valle (HUV) registered 90 cases of LBW, which represented 23.4% of all births ¹¹.

Several risk factors associated with low birth weight have been identified. Those related to the mother include periodontitis, depression, alcohol consumption and smoking, preterm birth, birth interval, multiple births, primiparity, prenatal care, parents' socioeconomic level, maternal age, maternal occupation, maternal educational level and maternal marital status ¹².

The INS has identified five types of risk associated with low birth weight: sociodemographic (maternal age, low socioeconomic status), medical history (hypertension, previous births, kidney, cardiorespiratory and autoimmune diseases, maternal malnutrition, primigravida), risks during pregnancy (gestational hypertension, anemia, vaginal bleeding, gestational diabetes, urinary tract infection, gestational syphilis, TORCH), inadequate prenatal care (late check-ups, few check-ups, lack of micronutrient supplementation) and environmental risks (excessive work, stress, anxiety, depression, consumption of tobacco, alcohol, coffee and drugs) ¹⁰.

Other factors linked to LBW are adverse social conditions including unemployment, lack of food security, extreme poverty and barriers to accessing health services. Anemia in the early stages of pregnancy is the main biological determinant for LBW, which in turn explains why lack of prenatal care is one of the main risk factors related to deficiencies in health care ¹³. With regard to the main barriers to health services in Colombia which are related to LBW, these are geographical, economic, related to the complexities of the national health system ^14,15, and also include inaccessibility, limitations in transportation, health, social, cultural, regulatory, administrative and political barriers ^16-22.

Given that many of the determinants associated with LBW are structural, each region or administrative unit needs to identify and evaluate the specific regional social, cultural, political and economic determinants that are associated with LBW, in order to improve the processes of quality and safety of care of health to the vulnerable population, for LBW, ADD and perinatal mortality, which are the main public health events that are indicators of social development of the regions ^16-23. This research aimed to determine the frequency and determinants of LBW in a high-complexity hospital in Cali, Colombia, using data from 2017.

A descriptive observational study with an analytical scope (cases and controls) was carried out. All the records of births in the HUV were analyzed, in single-term newborns of both sexes in a high-complexity hospital in Cali, Colombia. The WHO parameters were used to classify LBW ²⁴.

53 cases of LBW and 853 normal-weight neonates were identified in the HUV newborn registries for the last quarter of 2017. A case of LBW was defined as “any patient who was born in the HUV within the study period, weighing between 1,800 and 2,499 g”. A control was defined as “any patient who was born in the HUV in the same period with a weight greater than 2,500 g”. Records of patients with extreme low weight were excluded, as they were generally related to unfavorable clinical conditions of the mother or the pregnancy, including pre-eclampsia, severe urinary tract infection and/or congenital malformations, amongst others.

During the period October 1st to December 31st, 2017, 1,131 live births were registered in the HUV. 179 (11%) records for preterm pregnancy (less than or equal to 36.6 weeks of gestation) and 46 (3.0%) records for extreme low birth weight were excluded. Information from 906 records of patients who met the selection criteria was analyzed, of which 53 (6.0%) were classified as cases of LBW with 853 children with weights considered normal being classified as controls (Figure 1).

Figure 1

The median birth weight was 3,181.9 g (IR +/- 4,32.7) in the study population and in the 853 controls, it was 3,233.5 g (IR +/- 390.8). In the LBW group, the median birth weight was 2,351.6 g (IR +/- 96.7). The predominant sex in the LBW cases was female, at 53% (28/53), in contrast to the control group in which it was male, at 61% (411/953). Regarding ethnicity, the most frequent ethnicity in both groups was “mestizo” (64%), followed by Afro-Colombians (19; 36%). Regarding insurance, 91% of the cases belonged to the subsidized regime in the LBW cases and 90% in the controls (Table 1).

Table 1

Caracteristic	Description	n	LBW (n=53)	Controls (n=853)	OR	IC (95%)	p
Sex	Female	424	28	396	1.29	0.74-2.25	0.364
	Male	482	25	457
Gestation time (weeks)	37-38	348	45	303	10.21	4.75-21.9	0.000
	39-41	558	8	550
Primigestation	Primigestante	382	29	353	1.71	0.98-2.99	0.056
	More than one pregnancy	524	24	500
Type of delivery	Cesarean	200	18	182	1.90	1.05-3.43	0.032
	Vaginal Delivery	706	35	671
Ethnicity	Mestizo	588	34	554	0.96	0.54-1.72	0.906
	Others	318	19	299
Young age group	Adolescent	176	13	163	1.37	0.72-2.63	0.333
	Not Adolescent	730	40	690
Older age group	40 or more	26	2	24	1.36	0.31-5.90	0.685
	Minor 40	880	51	829
Mother's education	elementary or less	106	8	98	1.37	0.63-2.99	0.428
	More that elementary	800	45	755
Relationship status	With partner	672	39	633	0.96	0.52-1.82	0.920
	Without partner	234	14	220
Assurance	Subsidized Regime	814	48	766	1.09	0.42-2.81	0.858
	No Subsidized Regime	92	5	87
Social status	low	509	27	482	0.79	0.46-1.40	0.428
Other	397	26	371

* The most statistically relevant variables were gestation time, primigestation and type of delivery.

Regarding the factors related to LBW in the bivariate analysis, which showed an association with LBW, these included the mother's gestation time (OR: 10.21; 95% CI: 4.75-1.90; p= 0.000 ) and type of delivery (OR: 1.90; 95% CI: 1.05 - 3.43 p= 0.032), with first-time pregnancy being at the limit of significance (OR: 1.71; 95% CI: 0.98-2.99 p= 0.056) (Table 1 ).

Logistic regression was performed on the statistically significant variables which showed that the variables that explain LBW in this study were gestation time and first-time births (Table 2). The variable “type of delivery”, that had a significant association in the bivariate analysis, had a diluted association in the multivariate analysis, showing itself as confounding. After adjusting for first-time pregnancy, newborns with gestations between 37-38 weeks presented a positive association with an OR of 10.2 (95% CI: 4.7-21.9). Similarly, newborns with LBW from first-time deliveries presented an adjusted OR of 1.9 (95% CI: 1.1-3.5) compared with newborns of mothers with a history of previous deliveries. The association between LBW and cesarean deliveries presented adjusted OR 1.69 (95% CI: 0.92-3.13), however, their confidence intervals include the null value, so the association was not significant.

Table 2

Caracteristic	Description	n	LBW (n=53)	Controls (n=853)	OR	IC 95%	p	OR Adjusted	IC 95%	p
Gestation time (weeks)	37-38	348	45	303	10.2	4.7-21.9	0.000	10.2	4.7-21.9	0.000
	39-41	558	8	550
Primigestation	Primigestants	382	29	353	1.7	1.0-3.0	0.056	1.9	1.1-3.52	0.022
	More than one pregnancy	524	24	500
Type of delivery	Cesarean	200	18	182	1.9	1.0-3.4	0.032	1.7	0.9-3.1	0.092
Vaginal delivery	706	35	671

In general terms, the adjusted ORs provided by logistic regression show that the exposure variables with raw ORs that were significant maintained the direction and strength of the association, indicating that they behave as true determinants that explain low birth weight in this population. Those variables with significance greater than 0.05 were excluded in the multivariate analysis which suggests that these did not behave as confounding factors (Table 2).

This study evaluated the determinants of LBW in patients born at the HUV, evaluating clinical and demographic information available in the medical records of 906 patients who met the selection criteria. The findings of this study demonstrate that in the HUV, LBW is associated with the length of gestation and the history of previous pregnancies, regardless of the mother's age.

Regarding the gestation time, it was found that LBW was more frequent in mothers who had gestations between 37-38 weeks compared to mothers with gestations between 29-41 weeks. These findings agree with those reported in Venezuelans where a positive association was found between gestational age of 37-38 weeks and LBW (27). Additionally, in 2013, 502 births were evaluated in a high-complexity private hospital in Cali, Colombia, where a significant dependence was found between LBW and gestational age (p: 0001) ²⁸. This association between LBW and gestational age is explained by fetal maturity, which depends on gestation time and therefore it is expected that the longer the gestation period, the greater the weight in the newborns. Generally, gestations between 37-38 weeks lead to newborns with weights greater than 2,500g, demonstrating that there are other factors that interact with gestational age, including the social status of the mother ²⁹.

In this research, this association can be explained by the type of population that accesses the HUV, with patients normally living in low socioeconomic strata, coming from deprived areas of Cali and rural areas of the areas of influence. Although other studies have shown significance with other variables related to gestational immaturity, including access to health services, low adherence to prenatal care and food safety ³⁰, this information was not available in the medical records used for this study.

Regarding maternal age, this research did not show an association between LBW and extreme age of the pregnant woman. The association was also not statistically significant in the multivariate analysis, which would indicate that in this study the age of the pregnant woman is affected by other variables. Associations have been found between LBW and extreme ages ³¹, where a greater association was reported between adolescents and those over 35 years of age at the time of delivery, the authors arguing that in the youngest mothers this was due to immaturity and in those older than 35 years this was due to ageing. This was not found during our study, perhaps due to the characteristics of the pregnant women treated at the HUV, where the poor clinical and sociodemographic conditions are more important for the first pregnancy, probably due to the care provided during pregnancy and adherence to prenatal care as a protective factor in future pregnancies, regardless of the age at which gestation occurs.

Regarding the type of delivery, our findings agree with two studies ^12,32. Cesarean delivery showed a significant association before the multivariate analysis, confirming a high risk of LBW in Cesarean delivery. Cesarean delivery was also performed in the presence of clinical, psychological, or social conditions that were not favorable for a vaginal delivery and so we consider Cesarean delivery is a possible response to the gestation process, rather than a risk factor for LBW.

Regarding the limitations of the study, the short period of analysis (which included only three months) could be regarded a limitation: the HUV statistical records do not, however, report seasonality in births or in LBW in previous years and so this limitation can be discarded. Only three months of 2017 were included in the study because the HUV was undergoing administrative restructuring allowing it to expand its coverage and installed capacity to provide a better service to the entire city, which has been reflected in the better quality of services it has provided since mid-2017.

The results of this study showed that LBW can be explained by biological variables including sociodemographic background: specifically, first-time pregnancy and gestation time represented risks for LBW in our population.

This suggests the need to reinforce sexual, nutritional and reproductive education programs at the public health level, in order to prepare women to become pregnant only in favorable biological and psychological conditions.

We have considered the possibility of carrying out prospective studies delving into sociodemographic variables and the perception of support for pregnant women by health promoters and health providers, in addition to replicating this study in hospitals that treat patients with private health insurance and which cater to patients with more favorable sociodemographic situations. This would allow us to compare our findings with the findings from these other hospitals and to improve the quality of life of the poorer population by focusing efforts on installing improvements from the point of view of public health.