Magnitude and associated factors of low birth weight among term newborns delivered in Addis Ababa public hospitals, Ethiopia, 2021

Abstract

Only 14% of births had information on birth weight available at the time of birth in Ethiopia. Hence, previous studies underestimate the magnitude and associated factors of low birth weight (LBW). As a result, the goal of this study is to fill those gaps in the previous studies. An institution-based cross-sectional study was employed. Binary logistic regression was used to identify the associated factors. In this study, the magnitude of LBW was 13.06%. History of chronic medical illness (AOR = 3; 95% CI: (1.02, 9.17)), haemoglobin level during pregnancy (AOR = 0.23; 95% CI: (0.10, 0.50)), iron/folic acid supplementation (AOR = 0.27; 95% CI: (0.10, 0.72)) and extra meal during pregnancy (AOR = 3.2; 95% CI: (1.52, 7.00)) were significantly associated with LBW. The magnitude of LBW in this study was comparable to the Ethiopian Demographic and Health Survey (EDHS) report from 2016. It is better to intervene in those identified factors in order to reduce LBW.

Impact Statement

• What is already known on this subject? Low birth weight (LBW) accounts for 60–80% of all neonatal deaths each year. In developing countries like Ethiopia, LBW is a major public health concern. Almost half of the world's infants are not weighed at birth, a figure that is especially high in sub-Saharan Africa including Ethiopia.

• What do the results of this study add? Only 14% of births had information on birth weight available at the time of birth in Ethiopia. Hence, previous studies underestimate the magnitude and associated factors of LBW. To meet the Sustainable Development Goals (SDGs)-2030 targets for neonatal and child mortality, sufficient evidence on the magnitude of LBW and associated factors must be important in order to contribute to the development of timely interventions. A history of chronic medical illness, haemoglobin level, iron/folic acid supplementation and extra meal during pregnancy was associated with LBW.

• What are the implications of these findings for clinical practice and/or further research? The findings of this study will be useful in developing better health policies to prevent LBW as well as interventions that can target the identified factors.

Keywords:

• Magnitude
• associated factors
• low birth weight
• newborn
• Ethiopia

Background

The World Health Organization (WHO) defines low birth weight (LBW) as less than 2500 g at birth (5.5 lb). Low birth weight remains a major public health issue around the world, with a variety of both short- and long-term consequences (World Health Organization 2014). Low birth weight is estimated to account for 15–20% of all births worldwide, amounting to more than 20 million births per year. Globally, around 30 million LBW babies are born each year (23.4% of all births). Half of all LBW babies are born in South-Central Asia, where 27% are born weighing less than 2500 g, while LBW levels in sub-Saharan Africa are estimated to be 15% (Tessema et al. 2021). According to the EDHS 2016, the prevalence of LBW is estimated to be 13% in Ethiopia (CSA 2016b). In developing countries like Ethiopia, LBW is a major public health concern (Toru and Anmut 2020).

Low birth weight is the most important risk factor for adverse health outcomes, including common childhood diseases (Class et al. 2014). Evidence showed that the prevalence of LBW was 12% in Debre Tabor (Mekie and Taklual 2019), 10.3% in Dangla (Talie et al. 2019), 17.4% in Gondar (Zenebe et al. 2014) and 14.6% in the Central Zone of Tigray region (Gebremedhin et al. 2015). Low birth weight is widely considered the single most important predictor of infant mortality, particularly deaths occurring within the first month of life (Demelash et al. 2015).

Various studies revealed that place of residence, gravidity, the status of pregnancy, maternal age less than 20 years, haemoglobin level, maternal body mass index (BMI) less than 18.5, short pregnancy interval, drinking alcohol during pregnancy, no formal education and being unmarried were significant predictors of LBW (Gebremedhin et al. 2015; Aboye et al. 2018; Endalamaw et al. 2018; Mekie and Taklual 2019). Low birth weight accounts for 60–80% of all neonatal deaths each year (Eshete et al. 2019). Birth weight is an important indicator of an infant's health status and a major factor in determining the infant's physical, survival and mental growth. It also indicates the mother's past and present health status (Zenebe et al. 2014).

Between 2005 and 2016, neonatal mortality in Ethiopia fell by 10%, but it has since risen to 30 per 1000 live births in 2019 (CSA 2016a; Indicators 2019). Despite significant reductions in infant and under-5 mortalities in Ethiopia, the proportion of newborn deaths remains high (Maternal and Child Health Directorate Federal Ministry of Health 2015). As a result, the Ethiopian government recognised LBW as a serious public health issue and implemented a number of interventions. Some of the interventions to reduce LBW include focussed antenatal care (ANC), nutritional counselling during the prenatal and perinatal period, and institutional deliveries (Maternal and Child Health Directorate Federal Ministry of Health 2015). Furthermore, non-governmental organisations and professional associations, such as the Ethiopian Paediatric Society, have been working to address this issue (Fulton 2013; Health 2015). Even though much effort has been put into child health programs to combat the prevalence of LBW, associated factors and their contribution to neonatal death, more concern and commitments are needed. Data on the prevalence of LBW are required to assist countries in developing action plans and accountability measures, as well as to track progress (Doherty and Kinney 2019).

Accurate monitoring is difficult because nearly half of the world's infants are not weighed at birth, which is especially high in sub-Saharan Africa and Ethiopia, where 54% and 86% of infants are not weighed after birth, respectively. This issue is likely to underestimate the true magnitude of the problem (CSA 2016b; Lemlem et al. 2021). As a result, it is beneficial to conduct such research in urban settings like Addis Ababa, where approximately 83% of births are delivered in health facilities and babies are weighed at birth, as it is assumed to obtain more reliable and correct information than we get on average (CSA 2017). To meet the Sustainable Development Goals (SDGs)-2030 targets for neonatal and child mortality, sufficient evidence on the magnitude of LBW and associated factors must be generated in order to contribute to the development of timely interventions. The findings of this study will be useful in developing better health policies to prevent LBW as well as interventions that can target the identified factors.

Methods and materials

Study design and setting

An institution-based cross-sectional study was conducted among alive term newborns in selected public hospitals in Addis Ababa Ethiopia from January to June 2021. Addis Ababa, the capital city of Ethiopia and city has 11 sub-city administrations and 121 Woreda administrations. It is located at 9.02 latitudes and 38.75 longitudes and it is situated at an elevation of 2405 m above sea level. The city has around 100 public health centres, 13 public hospitals, 32 private hospitals and 1143 private clinics providing comprehensive health services (Federal Democratic Republic of Ethiopia Central Statistical Agency 2013, from 2014 to 2017). The study has been reported in line with the strengthening of the reporting of observational studies in epidemiology STROBE criteria (Von Elm et al. 2007).

Study population

The source population were all newborns delivered in public hospitals in Addis Ababa. Whereas all newborns delivered at selected public hospitals of Addis Ababa during the data collection period were the study population. Newborn babies whose mothers suffered from severe medical or surgical conditions at the time of data collection period, twin delivery, newborn delivered with congenital anomalies, and unknown last normal menstrual period with absent early ultrasound evidence were excluded from the study.

Sample size and sampling procedures

The sample size was calculated using a formula for estimating sample size for a single population proportion, assuming a confidence level of 95%, a margin of error of 4%, a magnitude of LBW 15.8% (Lake and Olana Fite 2019) and by adding 10% for non-response rate, the final sample size was 351. Simple random sampling was used to select four hospitals from among the 13 public hospitals in Addis Ababa. Based on the delivery caseload of each selected health public hospital (Tikur Anbessa Specialized Hospital, Gandhi Memorial Hospital, Zewditu Memorial Hospital and St Petros Specialized Hospital, which were 600, 700, 300 and 250, respectively), the calculated sample size (351) was proportionally allocated to these health facilities. Finally, data were systematically collected by skipping every fifth neonate until the required sample size was reached. The initial study participant was selected by lottery method. Then, each study participant was selected using a systematic random sampling technique.

Study variables

Dependent variable

Low birth weight.

Independent variables

Sociodemographic and socioeconomic factors. Age, residence, educational status, monthly income, occupation and marital status.

Maternal obstetrics related factors. Current pregnancy status, number of pregnancy, number of parity, history of LBW baby, history of abortion, history of still birth, history of ANC follows up and anaemia during pregnancy.

Labour and delivery related factors. Sex of neonate, gestational age at birth, and birth weight.

Nutritional status and behaviour factors. Iron/folic acid supplementation during the current pregnancy, extra meal during current pregnancy, nutrition counselling during current pregnancy, any substance use (khat, alcohol, cigarette smoking) during current pregnancy.

Data collection tools and quality control

A semi-structured questionnaire and data extraction checklist were prepared. To ensure consistency, the questionnaire was prepared in English and translated into the Amharic language, and then back to English. The training was provided for data collectors and supervisors on overall data collection procedures, interviewing techniques, and how to maintain the confidentiality of information obtained from respondents for one day. To ensure the tool's validity, it was pretested on 5% of the total sample size. Throughout the data collection period, the principal investigator and supervisors were conducted on-site supervision on a daily basis. The principal investigator and supervisors were checking and reviewing the collected data to ensure its completeness and consistency.

Data processing and analysis

The collected data were checked for completeness and consistency and entered into EPI data 3.1 software, and then, it was transferred to Statistical Package for the Social Sciences (SPSS) version 25 software (SPSS Inc., Chicago, IL) for the purpose of data cleaning, coding and analyses. Descriptive statics like frequencies with percentage and mean with standard deviation were done to describe the demographic and socio-economic characteristics of the study participants and to describe the outcome variable. Bivariate and multivariable logistic regression analyses were performed to test the association. The Hosmer and Lemeshow test was used to determine the goodness of fit. Variables with a 95% confidence interval and a p value less than .25 in the bivariate analysis were included in the multivariable logistic regression analysis to control for all potential confounding variables. Furthermore, even if the aforementioned parameters were not met, variables that were significant in previous studies and from a contextual point of view were included in the final model. Multi co-linearity was checked by co-linearity diagnostic statistics via variance inflation factors and tolerance test. Adjusted odds ratios with a 95% confidence interval were calculated and p value less than .05 was considered as statistically significant. Finally, data were presented using tables, graphs and texts.

Results

Sociodemographic and economic characteristics of study participants

Out of the 351 mothers, 337 participated in this study with a response rate of 96%. The mean age and standard deviation of study participants were 27.35 + 5.86 and 89.3% were age groups between 20 and 34 years. The majority (91.4%) of study participants (91.4%) were urban residents. About 145 (43%) had attained college and above level of education and 314 (93.2%) were married. According to the results of the household monthly income, 133 (39.47%) of study participants earned between 2100 and 4000 ETB per month (Table 1).

Table 1. Sociodemographic and economic characteristics of mothers who delivered at the public hospitals of Addis Ababa, Ethiopia, 2021 (n = 337).

Variable	Frequency	Percentage (%)
Age (years)
≤19	26	7.72 (%)
20–34	282	83.68 (%)
≥35	29	8.61 (%)
Educational status
No formal education	25	7.42 (%)
Primary school	56	16.62 (%)
Secondary school	111	32.94 (%)
College and above	145	43.03 (%)
Marital status
Single	18	5.30 (%)
Married	314	93.20 (%)
Widowed^a	5	1.50 (%)
Occupational status
House wife	130	38.58 (%)
Governmental employee	89	26.41 (%)
Merchant	72	21.36 (%)
Non-governmental employee	46	13.65 (%)
Residence
Urban	308	91.40 (%)
Rural	29	8.60 (%)
House hold monthly income (ETB)
≤2000	26	7.72 (%)
2100/4000	133	39.47 (%)
4100/6000	114	33.83 (%)
≥6100	64	18.99 (%)

^aWidowed: separated.

Obstetrics health care related, nutritional status and behavioural characteristics of study participants

Regarding the status of the current pregnancy of study participants, 304 (90.2%) of pregnancies were planned. More than half 186 (55.2%) of study participants were multigravida. Seventeen (5%) of respondents had a history of LBW and 41 (12.2%) of study participants had an abortion history. Three hundred and twenty-six (96.74%) of the respondents have ANC follow-up. Regarding medical status, 26 (7.72%) of study participants had chronic medical illnesses. Three hundred and three (89.91%) of study participants took iron and folic acid during their pregnancy, while 266 (78.9%) of respondents took extra meals during their pregnancy. About 314 (93.2%) of study participants received nutritional counselling during their pregnancy. Three hundred and twenty-five (96.44%) of study participants did not take any substance during their pregnancy (Table 2).

Table 2. Obstetrics health care related, nutritional status and behavioural characteristics of mothers who delivered at the public hospitals of Addis Ababa, Ethiopia, 2021 (n = 337).

Variable	Category	Frequency	Percentage (%)
Current pregnancy status	Planned	304	90.20 (%)
	Unplanned	33	9.80 (%)
Number of gravida	1	127	37.69 (%)
	2–4	172	51.04 (%)
	>5	38	11.28 (%)
History of low birth weight	Yes	17	5.00 (%)
	No	320	95.00 (%)
History of abortion	Yes	41	12.20 (%)
	No	296	87.80 (%)
Tetanus toxoid (TT) vaccination during pregnancy	Yes	328	97.30 (%)
	No	9	2.70 (%)
Chronic medical illness	Yes	25	7.42 (%)
	No	312	92.58 (%)
Haemoglobin result	<11 mg/dl	57	16.91 (%)
	≥11 mg/dl	280	83.09 (%)
Antenatal care follow up	Yes	326	96.74 (%)
	No	11	3.26 (%)
Iron/folic acid supplementation during the current pregnancy	Yes	303	89.91 (%)
	No	34	10.09 (%)
Extra meal during current pregnancy	Yes	266	78.90 (%)
	No	71	21.10 (%)
Nutrition counselling during current pregnancy	Yes	314	93.20 (%)
	No	23	6.80 (%)
Any substance user	Yes	12	3.56 (%)
	No	325	96.44 (%)

Magnitude of low birth weight

In this study, the magnitude of LBW was 13.06% (95% CI: (9.8, 17.11)). The mean and standard birth weight of the newborn was 2996 g (±517 g) (Figure 1).

Figure 1. Magnitude of low birth weight among newborn delivered in Addis Ababa public hospitals, Addis Ababa, Ethiopia, 2021.

Factors associated with low birth weight

In order to identify the associated factors of LBW, variables with a p value of less than .25 in bivariate analysis were included in the multivariable model. Respondents' age, history of stillbirth, previous history of LBW, history of chronic medical illness, haemoglobin level, history of ANC follow-up, sex of newborn, iron/folic acid supplementation during their pregnancy, and extra meal during their pregnancy were among the variables included in final model. After controlling confounding variables, history of chronic medical illness, haemoglobin level, iron/folic acid supplementation and extra meal during pregnancy were significantly associated with LBW. Women with a history of chronic medical illness were three times more likely to have an LBW than women without a history of medical illness (AOR = 3; 95% CI: (1.02, 9.17)). Low birth weight was 77% more likely in women with haemoglobin levels less than 11 g/dl than in pregnant women with haemoglobin levels greater than 11 g/dl (AOR = 0.23; 95% CI: (0.10, 0.50)). Those who did not take iron/folic acid supplementation during their pregnancy were 73% more likely to have an LBW than those who did take iron/folic acid supplementation (AOR = 0.27; 95% CI: (0.10, 0.72)). Pregnant women who did not consume extra food/meal during their pregnancy had 3.2 times higher odds ratio of having an LBW than those who did consume extra meals during their pregnancy (AOR = 3.2; 95% CI: (1.52, 7.00)) (Table 3).

Table 3. Factors associated with low birth weight among newborn delivered in Addis Ababa Public Hospitals, Addis Ababa, Ethiopia, 2021.

Variable	Category	Magnitude of LBW		COR (95% CI)	AOR (95% CI)
		Yes (%)	No (%)
Age (years)	≤19	6 (13.64)	20 (6.83)	1	1
	20–34	34 (77.27)	248 (84.64)	0.45 (0.17, 1.21)	0.47 (0.16, 1.42)
	≥35	4 (9.09)	25 (8.53)	0.53 (0.13, 2.15)	0.46 (0.92, 2.76)
History of still birth	Yes	6 (13.64)	19 (6.48)	1	1
	No	38 (86.36)	274 (93.52)	0.43 (0.16, 1.16)	0.6 (0.18,1.97)
Previous history of low birth weight	Yes	3 (6.82)	14 (4.78)	1.45 (0.40, 5.29)	1.6 (0.36, 7.15)
	No	41 (93.18)	279 (95.22)	1	1
Chronic medical illness	Yes	7 (15.91)	18 (16.14)	2.8 (1.13, 7.38)*	3 (1.02, 9.17)**
	No	37 (84.09)	275 (93.86)	1	1
Haemoglobin level	<11 g/dl	18 (40.91)	39 (13.31)	1	1
	≥11 g/dl	26 (59.09)	254 (86.69)	0.22 (0.11, 0.44)*	0.23 (0.10, 0.50)**
ANC follow up	Yes	42 (95.45)	284 (96.93)	0.66 (0.13, 3.18)	0.58 (0.10, 3.28)
	No	2 (4.55)	9 (3.07)	1	1
Sex of newborn	Male	24 (54.55)	117 (39.93)	1.8 (0.95, 3.41)	1.7 (0.85, 3.54)
	Female	20 (45.45)	176 (60.07)	1	1
Iron/folic acid for current pregnancy	Yes	32 (72.73)	271 (92.49)	0.21 (0.09, 0.47)*	0.27 (0.10,0.72)**
	No	12 (27.27)	22 (7.51)	1	1
Extra meal during current pregnancy	Yes	22 (50.0)	244 (83.28)	1	1
	No	22 (50.0)	49 (16.72)	4.9 (2.55, 9.69)*	3.2 (1.52, 7.00)**

1: reference.

*Statistically significant at bivariate analysis, **Statistically significant at p value <.05 with 95% CI in multivariable analysis.

Discussion

Low birth weight is a major public health concern because it increases the risk of developing non-communicable diseases later in life. LBW may be associated with immune system epigenetic modulation and cell maturation (Hayashi et al. 2020). Birth weight is an excellent summary measure of multifaceted public health issues such as long-term maternal malnutrition, ill health and inadequate prenatal health care (CSA 2016b). Hence, this study aimed to assess the magnitude and associated factors of LBW among newborns delivered in Addis Ababa public hospitals. Low birth weight was found to be 13.06% (95% CI: 9.8, 17.11) in this study, which was consistent with studies conducted in Adwa (10%) (Gebregzabiherher et al. 2017), the central zone of the Tigray region (14.6%) (Gebremedhin et al. 2015), Dangla (10.3%) (Talie et al. 2019), Butajira (12.5%) (Toru and Anmut 2020) and Ghana (12.9%) (Axame et al. 2022). However, this study's findings were higher than those of studies conducted in Australia (1.9%) (Herceg et al. 1994), Harar 23.3% (Abdurke Kure et al. 2021) and Addis Ababa (8.8%) (Mulatu et al. 2017), while this study finding was also lower than studies conducted in South Ethiopia (17.88%) (Wado et al. 2014) and Debre Markos (21.6%) (Alebel 2019). This discrepancy might be due to differences in study areas, study designs, inclusion and exclusion criteria, or the institutions' health service delivery systems. Another possible explanation could be the difference in health service utilisation and nutritional status of mothers during pregnancy.

The odds of having LBW newborns were found to be more than three times higher in mothers with a history of chronic medical illnesses than in mothers without a history of chronic medical illnesses. This finding was consistent with studies done in Mekelle (Gebremedhin et al. 2015; Desta et al. 2020). This could be due to increased risks of pregnancy complications. Women who took iron and folic acid supplements during pregnancy were found to have a protective effect against LBW in this study, which is supported by previous studies in Dilla (Mehare and Sharew 2020), Nepal (Khanal et al. 2014; Bhaskar et al. 2015) and China (Zhang et al. 2020). Folic acid's potential role in altering epigenetic processes that result in increased patterns of placental and foetal growth is one explanation that has been put forth. The modulation of placental growth and development by folic acid may also have an indirect influence on foetal growth (Timmermans et al. 2009; Yang et al. 2022).

Women with haemoglobin levels less than 11 g/dl had a higher risk of having an LBW baby. This is similar to studies conducted in Adwa (Gebregzabiherher et al. 2017), Debre Berhan (Asmare et al. 2018), Debre Tabor (Mekie and Taklual 2019), Harar (Abdurke Kure et al. 2021) and Pakistan (Khan et al. 2016). These may be due to anaemia's effect on oxygen-bearing capacity and its transportation tendency at the placental site, resulting in poor foetal growth and potential preterm delivery, which is a major cause of LBW. Another possible reason could be because of low haemoglobin levels during pregnancy impair the delivery of essential nutrients to the developing foetus, potentially compromising normal growth. This study showed that women who did not take extra meals during their pregnancy had a higher chance of having LBW neonates than mothers who did take extra meals. This finding was supported by studies done in the northwest part of Ethiopia (Talie et al. 2019), Ghana (Abubakari et al. 2019) and Pakistan (Jamshed et al. 2020). This could be due to a lack of food or a lack of awareness about the need for an extra meal during pregnancy that leads to undernutrition and pregnancy-related complication. Foetal growth is restricted as a result of undernutrition and pregnancy-related complications that also have an impact on the epigenetic process (Chiofalo et al. 2017).

The limitation of this study was that it was a cross-sectional study, so it does not show a cause and effect relationship, and it was unable to include mothers who delivered at home. Whereas the strength of this study was that direct measurement of newborn weight was used to eliminate recall bias.

Conclusions

The prevalence of LBW in this study was comparable to the EDHS report from 2016. Significant predictors of LBW were found to be a history of chronic medical illness, haemoglobin level, iron/folic acid supplementation and extra meal during pregnancy. Women with a history of medical illness, haemoglobin levels of 11 mg/dl, and who did not take iron/folic acid supplementation during pregnancy were more likely to have LBW newborns. Women who ate extra meals during their pregnancy, on the other hand, were less likely to have LBW newborns. Encouraging women to take iron with folic acid and extra meals during pregnancy is important for lowering the risk of LBW as well as maternal morbidity and mortality associated with anaemia.

Ethical approval

Ethical approval was obtained from the Addis Ababa Medical and Business College, School of Public Health Institutional Review Board (IRB), and official letters were sent to each irrespective public hospital. All necessary methods were used in accordance with the Guidelines of Institutional and Declaration of Helsinki.

Patient consent

Following an explanation of the research objectives to each study participant, informed written consent was obtained. The names and other identifiers of study participants were not recorded on the data collection tools to ensure confidentiality. During data collection, the possible COVID-19 prevention measures were implemented. All necessary methods were used in accordance with the Guidelines of Institutional and Declaration of Helsinki.

Author contributions

All authors contributed significantly to the study's conception, design, data collection, data analysis and interpretation of the findings. The authors also contributed to the manuscript's writing, reviewed the draft, and ultimately agreed on the journal in which the article should be published. All authors read and approved the final draft of the manuscript and agreed to accept responsibility for all of the manuscript's contents under any circumstances.

Acknowledgements

Our heartfelt gratitude goes to the Tikur Anbessa Comprehensive Specialized Hospital, Gandhi Memorial Hospital, Zewditu Memorial Hospital and St. Petros Hospital administration staff and health care professionals working in the hospital who provided support during data collection, data collectors and study participants.

Disclosure statement

All authors assert that they have no competing interests.

Data availability statement

The datasets generated and/or analysed during the current study are not publicly available due to preserving participant anonymity but are available from the corresponding author on reasonable request via [email protected]; [email protected].

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.