Page 1 of 13

Journal for Studies in Management and Planning

Available at

http://internationaljournalofresearch.org/index.php/JSMaP

e-ISSN: 2395-0463

Volume 01 Issue 02

March 2015

Available online: http://internationaljournalofresearch.org/ P a g e | 90

Determinants of Low Birth Weight Neonates: A Case

Study of Tamale Metropolis in Ghana

R. Puurbalanta 1, A. O. Adebanji 2

1University for Development Studies, Faculty of Mathematical Sciences, Department of

Statistics, Navrongo Campus, Ghana

E-mail: pubarichard@yahoo.com

Tel: +0233244546726

2 Kwame Nkrumah University of Science and Technology, Faculty of Physical Sciences, College

of Science, Department of Mathematics, Kumasi, Ghana

E-mail: tinuadebaji@yahoo.com

Tel: +0233241860372

ABSTRACT

Low Birth Weight (LBW), a birth weight less than 2.5kg, is an important public health problem

because LBW infants are at greater risk of mortality and morbidity in early infancy (WHO,

2004; UNICEF, 2004). The rate of LBW in the Northern Region consistently ranks high among

the ten regions in Ghana, and Tamale metropolis has the highest percentage of LBW births

among the twenty districts in the Northern Region, and this is a major concern for health care

providers given the high cost of caring for LBW infants. In this study, logistic regression model

was used to identify the determining variables in predicting LBW babies in the metropolis. The

model was based on the birth records of 500 mothers of singleton neonates resident in the

Tamale metropolitan area of the Northern Region of Ghana from November 2010 to January

2011. The significant model coefficients were Gestation (p-value = 0.0008), Household size (p- value = 0.0160), Maternal food intake (p-value = 0.0002), Maternal health (p-value = 0.0000),

Passive smoking (p-value = 0.0003) and Type of fuel used for cooking (p-value = 0.0418). A test

of predictive ability of the model showed correct classifications of 93% for normal birth weight

infants and 76.8% for LBW infants. The likelihood ratio and Nagelkerk R2

tests showed positive

correlation between the predictors and LBW. Using the Hosmer and Lemeshow test of goodness

of fit, a p-value 0.206 was obtained and thus the null hypothesis that the model fits the data well

could not be rejected.

Keywords

Birth Weight, Logistic, Predictors, Maternal, neonate

INTRODUCTION

The presumption that women of childbearing age should not possess any special skills in order to

give birth to normal and healthy babies may be untrue, especially for women in Tamale

metropolis in the Northern Region of Ghana. Available records show that the metropolis

consistently records higher rates of LBW infants in the region (Northern Region Reproductive

Page 2 of 13

Journal for Studies in Management and Planning

Available at

http://internationaljournalofresearch.org/index.php/JSMaP

e-ISSN: 2395-0463

Volume 01 Issue 02

March 2015

Available online: http://internationaljournalofresearch.org/ P a g e | 91

and Child Health (NRRCH) annual reports, 2007, 2008 and 2009, unpublished).

In informal conversations, many health experts in the region agreed with reports by UNICEF and

WHO (2004) and UNICEF (2005) that high rate of LBW is a major problem in developing

countries yet the issue and factors influencing it has not received the much-needed attention. The

result being that some expectant mothers undermine foetal welfare and thus many births result in

LBW babies.

Unavailability of adequate information on the causes of LBW to the general public, especially

expectant mothers has compounded the situation. In this evidence-based era, this is a real

concern, as no evidence will be available for policy decision-making. This underscores the need

to identify and highlight the main risk factors that cause LBW, with the ultimate aim of

eliminating it.

LBW prevalence is a major public health issue since it is considered the single most important

predictor of infant mortality, particularly in the first month of life (Blanc et al, 2005), and is a

significant factor in many adverse child health and development outcomes (WHO, 2004).

Risk factors include maternal age, alcohol abuse, smoking, lack of pre-natal care, gestational age,

and maternal ill health (Allen et al, 2001; Steyn et al, 2006; Knopik, 2009; Salmasi et al, 2010).

In this study, logistic regression model was used to investigate the determining factors of LBW

neonates and postulating a predictive model of the likelihood of a pregnancy resulting in a LBW

neonate in the metropolis.

MATERIALS AND METHODS

Study Design and Data Sources

The study was a cross-sectional design. Data included all singleton births over a period of three

months (November 2010 to January 2011). Five hospitals randomly selected in the metropolis

served as study sites. They are the Choggu Clinic, Tamale Central Hospital, Tamale West

Hospital, The SDA Hospital and Fulera Maternity Home. The above community-based hospitals

provide both antenatal care and counseling services to expectant mothers. The locations of these

hospitals in the metropolis portray respondents of variable socio-economic backgrounds.

After obtaining the consent of the women to partake in the study, a structured questionnaire was

used to elicit data pertaining to mother’s socio-economic and reproductive-obstetrical

information. Maternal and newborn medical history such as number of antenatal visits,

gestational age, maternal health, maternal weight gain during pregnancy, neonate’s sex and birth

weight recorded in the folders of mothers at the medical facility during the course of pregnancy

were all captured by the questionnaire.

Sampling Techniques

Page 3 of 13

Journal for Studies in Management and Planning

Available at

http://internationaljournalofresearch.org/index.php/JSMaP

e-ISSN: 2395-0463

Volume 01 Issue 02

March 2015

Available online: http://internationaljournalofresearch.org/ P a g e | 92

In order to meet the data requirement for this study, all singleton births to mothers who had

medical history recorded at the medical facility during the course of their pregnancy were

sampled. Mothers had to be ordinarily resident in the metropolis to be eligible for the study.

Study Area

Tamale metropolis

Study Population

All mothers who delivered between November 2010 and January 2011 in the metropolis.

Determination of Sample Size

Using a guideline provided by Hosmer et al (2000) that the minimum number of cases per

independent variable is 10, with a preferred ratio of 20 to 1, a sample of 500 mothers were

selected for the study.

Study Variables

We define our outcome variable as a binary response corresponding to the risk of a neonate

being born LBW. That is,

1, if neonate is LBW

0, otherwise

Explanatory variables include maternal age, marital status, parity, cigarettes smoking by the

mother during pregnancy, maternal educational attainment, family income, gestational age,

passive smoking, type of fuel used by the mother for cooking, type of residence, number of

antenatal visits, maternal health, maternal food intake, household size, maternal weight gain

during pregnancy, employment status, occupation, history of previous LBW, birth spacing and

new-born sex.

Logistic regression model was used to identify risk factors for LBW by estimating the odds

ratios (OR) and their 95% confidence interval (CI). A multivariable analysis was conducted to

control for confounders. This analysis was carried out by the simultaneous method, where all

variables were entered at once. A 0.05 level of significance was used and the data processed and

analysed using Epi info3.4.1 software.

Logistic Regression Model for Binary Data

When we have a binary response variable , and a vector of associated covariates , the

preferred mathematical model to deal with the complex and generally nonlinear

interrelationships among the many variables is the binary logistic regression model (Agresti,

2007). Logistic regression combines the independent variables to estimate the probability that an

event of interest will occur, that is, a subject will be a member of one of the groups defined by

the dichotomous dependent variable.