Affecting Variables Breastfeeding in Indonesian Urban Areas:

2017 IDHS Data Analysis

Onny Fatimah and Risni Julaeni Yuhan

Politeknik Statistika STIS, Indonesia

Keywords: Breast Milk, Breastfeeding, Urban, Binary Logistic Regression.

Abstract: Breast milk is an important source of nutrients for child growth and development. The data of BPS 201,

achievement of breastfeeding in urban areas of Indonesia (81.72 percent) is more critical than rural areas

(73.82 percent). This is exacerbated by the infant mortality rate (IMR) in urban areas which is higher than in

rural areas. Therefore, in this study, we examine the variables that affect breastfeeding in Indonesian urban

areas using the 2017 IDHS data using the binary logistic regression method. The variables studied included

mother's education, mother's working status, birth attendant, place of delivery, early initiation of

breastfeeding, pregnancy visits, and household economic status. The data showed that 94.7 percent of mothers

gave breast milk and 5.3 percent of mothers did not breastfeed their children. The result showed that several

variables that had a significant effect on breastfeeding for children under two years of age were the mother’s

working status, pregnancy visits, and early initiation of breastfeeding. Mothers tend to breastfeed is the mother

who does not work, carry out the IMD and women who undergo pregnancy visit of more than equal to four

times.

1 INTRODUCTION

According to Law Number 35 of 2014, a child is

someone who is not yet 18 years old, including

children who are still in the womb. Children are an

asset for a nation in the future. For this reason, the

government through Law Number 36 Year 2009

states that efforts to maintain infant and child health

must be aimed at preparing future generations who

are healthy, intelligent, and qualified and to reduce

infant and child mortality. In order to maintain the

health of infants and children, it is important to meet

the nutritional intake for children.

Breast milk is the main source of nutrition for

children. Breast milk contains white blood cells,

immunoglobulins, enzymes, hormones, specific

proteins, and other nutrients needed for the growth

and development of children (Law Number 3, 2010).

Breastfeeding is recommended for the first six

months without complementary foods. Then

breastfeeding is continued until the child is two years

old.

Based on the 2030 Sustainable Development

Goals (SDGs), there are several goals that are closely

related to breastfeeding. This goal is contained in the

second goal of ending hunger and the third goal of

good health and well-being. This goal can be

achieved by ending all forms of malnutrition and

reducing neonatal mortality. Every year, there are

more than 25,000 babies in Indonesia and 1.3 million

babies in the world can be saved from death by

breastfeeding.

Source: IDHS 2002-2017

Figure 1. Percentage of coverage of exclusive breastfeeding

in Indonesia in 2002-2017.

Figure 1 shows that the coverage of exclusive

breastfeeding from 2002 to 2017 has fluctuated. The

coverage of exclusive breastfeeding in Indonesia has

Fatimah, O. and Yuhan, R.

Affecting Variables Breastfeeding in Indonesian Urban Areas: 2017 IDHS Data Analysis.

DOI: 10.5220/0010759800003235

In Proceedings of the 3rd International Conference on Social Determinants of Health (ICSDH 2021), pages 189-194

ISBN: 978-989-758-542-5

189

not met the target set by the government, which is 80

percent. Meanwhile, the percentage of children aged

0-23 months who were still breastfed in 2017 was

77.4 percent. This figure has decreased from 2016 of

78.49 percent. The practice of breastfeeding that has

not met the target will cause problems in achieving

optimal growth and development as well as

malnutrition in children. Septikasari (2018) states that

the success of exclusive breastfeeding has an

influence on the incidence of malnutrition in children

aged 6-12 months.

Based on 2017 BPS data, the achievement of

exclusive breastfeeding in urban areas in Indonesia in

2017 was 54.77 percent. Meanwhile, the coverage of

exclusive breastfeeding in rural areas of Indonesia in

the same year was 57.22 percent. The same condition

also occurs in children aged 0-23 months who are still

breastfed in rural areas by 81.72 percent. This figure

is higher than urban areas with a percentage of 73.82

percent. The condition of breastfeeding in urban areas

in Indonesia is more critical than in rural areas. This

is also exacerbated by the infant mortality rate (IMR)

in urban areas which is higher than in rural areas

IMR in urban areas is 24 deaths per 1000 live

births. In rural areas, this figure is lower at 23 deaths

per 1000 live births. It is interesting to study further

with regard to the characteristics of urban areas that

affect breastfeeding.

In order to optimize breastfeeding for children, it

is necessary to know what factors are related to a

person's behavior related to health. According to

Lawrence Green (1980), human behavior related to

health is formed through several factors including the

following:

1. Predisposing factors which include

knowledge, attitudes, beliefs, beliefs, socio-

cultural values, and so on,

2. Enabling factors which include the physical

environment, available or the unavailability

of health facilities or facilities, such as health

centers, medicines, sterile tools and so on.

3. The driving factor which includes the

attitudes and behavior of health or other

staff, which is a reference group for

community behavior.

Predisposing factors are factors that become the

basis of a person's motivation or intention to do

something. Variables that are included in the

predisposing factors include education level and work

status. Enabling factors are factors that make it

possible to perform a behavior or action. Enabling

factors in this study used variables of economic

status, place of delivery, pregnancy visits and BMI

status. While the driving factor is a factor that

strengthens a person's behavior. The driving factor in

this study was a delivery assistant variable.

Research conducted by Astuti (2013) concluded

that mother's education, mother's occupation,

mother's knowledge, mother's attitude, role of

officers, media exposure, husband's role, parental role

had a significant effect on exclusive breastfeeding.

Hasiana (2016) in his research stated that there was a

significant effect between the implementation of IMD

and the success of exclusive breastfeeding.

Rahmawati (2012) concluded that maternal age,

employment status, baby birth order and support from

health workers had a significant effect on exclusive

breastfeeding. In a study conducted by Elsera (2015)

it was found that there was an effect between neonatal

visits and work with exclusive breastfeeding. In a

study conducted by Tampah-Naah and Kumi-

Kyereme (2013) stated that there is a relationship

between marital status, area and place of delivery

with the practice of exclusive breastfeeding.

Based on the description of the problem, this

study has the following objectives to examine the

description of the characteristics of breastfeeding in

urban areas of Indonesia, analyze the variables that

have a significant effect on breastfeeding in urban

areas of Indonesia, and determine the tendency of

each variable that has a significant effect on

breastfeeding. in children under two years in urban

areas of Indonesia.

2 METHODOLOGY

2.1 Data Collection Methods

This study used secondary data sourced

frommaterials raw data 2017 IDHS. Women of

childbearing age (15-49 years) living in urban areas

who have their last child under two years old are the

unit of analysis in this study. The implementation of

the 2017 IDHS covers 49,250 households with a total

of 1970 census blocks in urban and rural areas.

From the 2017 IDHS data, there are 49,627 WUS

respondents from all over Indonesia. For research

purposes, respondents who lived in rural areas were

excluded, leaving 26,425 WUS respondents. Then

another selection was made to select WUS who had

the last child under two years of age and clean up the

missing value on the dependent variable so that there

were 3467 WUS respondents left. Of the 3467

respondents who were selected, then re-

categorization was carried out.

The dependent variable used in this research is the

status of breastfeeding for children under two years

ICSDH 2021 - International Conference on Social Determinants of Health

190

old, namely breastfeeding and not breastfeeding.

While the independent variables include: education

level, work status, birth attendant, place of delivery,

pregnancy visits, early initiation of breastfeeding, and

economic status.

2.2 Analysis Method

The analytical method used in this research is

descriptive analysis and inferential analysis.

Descriptive analysis used in the form of graphs and

tables to see the general picture of breastfeeding in

urban areas. Meanwhile, inferential analysis uses

binary logistic regression analysis with the aim of

seeing the effect of the independent variable on the

dependent variable. The significance level used in

this study is 5 percent. The binary logistic regression

model used in this study is as follows:

𝑔



𝐷



𝛽





𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



𝛽





𝐷



(1)

Model suitability test was conducted to see if the

model used was suitable to explain breastfeeding

status. Testing using Hosmer and Lemeshow

Goodness of fit test.

Simultaneous test was used to see the effect of the

independent variables together on the variable of

breastfeeding status. Simultaneous test using G.

Partial test is used to determine the significance of

each independent variable on the dependent variable.

This test is carried out using thetest Wald.

The accuracy of the model in classifying the

observed values can be obtained from looking at the

classification table. The last step is to calculate the

odds ratio of each significant independent variable to

see the trend of breastfeeding between a category and

the reference category.

3 RESULT AND DISCUSSION

3.1 Descriptive Analysis

Based on the results of the study, it was found that

94.7 percent of mothers living in urban areas with

children under two years of age gave breast milk and

5.3 percent did not. This indicates that there are still

mothers in urban areas who do not breastfeed their

children.

Source: IDHS data (processed)

Figure 2. Percentage of women of childbearing age based

on breastfeeding status in urban Indonesia.

Based on table 1. Mothers with junior high-high

school education are more dominant than mothers

who do not attend school, elementary/equivalent

education, and education above high school. The

distribution of mothers who do not work is much

greater than that of working mothers. The percentage

of mothers is dominated by mothers with birth

attendants by health workers. Likewise with the

characteristics of the mother's place of delivery. The

percentage of mothers was dominated by mothers

who gave birth in health facilities, which was 93.6

percent. The distribution of mothers was dominated

by mothers who had four or more pregnancy visits

than mothers with less than four pregnancy visits. The

percentage of mothers was dominated by mothers

who did early initiation of breastfeeding than mothers

who did not do early initiation of breastfeeding.

Meanwhile, the percentage of mothers whose

economic status is rich is 58.6 percent, middle

economic status is 21.6 percent, and the economic

status is poor is 19.7 percent. These results indicate

that the number of mothers whose economic status is

rich is more than that of mothers with middle and poor

economic status.

94,7%

5,3%

Breastfeeding

Not

breastfeeding

Affecting Variables Breastfeeding in Indonesian Urban Areas: 2017 IDHS Data Analysis

191

Table 1. Percentage of women of childbearing age in urban

Indonesia based on their characteristics.

Characteristics Percentage

(1) (2)

Mother's education 100.0

No school 0.3

Elementary school 14.8

Middle school 63.1

High school 21.9

Status Workin

Mothe

100.0

Working 39.1

Not workin

60.9

Birth attendant 100.0

Assisted by health workers 97.4

Not assisted b

health workers 2.6

Place of deliver

100.0

Not in health facilities 6.4

At health facilities 93.6

Pregnancy visit 100.0

≥4 93.9

<4 6.1

Earl

initiation of breastfeedin

status 100.0

Early initiation of breastfeeding 65.4

No earl

initiation of breastfeedin

34.6

Economic status 100.0

Rich 58.6

Middle 21.6

Poo

19.7

Source: IDHS 2017 (processed)

3.2 Binary Logistics Regression

Analysis

3.2.1 Model Fit Testmodel Fit

Based on the results of thetest , the table shows the C

statistic value of 6.917 with a p-value of 0.546. P-

value is more than = 0.05 so the decision is to accept

which means that the model is fit. Thus, it can be

concluded that with a 95 percent confidence level the

model used is appropriate in explaining the status of

breastfeeding in children.

Table 2 The value of thetest statistic Hosmer and

Lemeshow.

Chi-s

e df Si

(

)

(

)

(

)

6.917 8 0.546

Source: IDHS 2017 (processed)

3.2.2 Simultaneous Test

Based on the results of the simultaneous test, the G

statistic value is 51,742 with a p-value of 0.000. As in

table 3, it can be seen that the P-value <α=0.05 so that

the decision obtained is to reject H

. Thus, the results

of the study indicate that there is at least one

independent variable that has a significant effect on

breastfeeding children in urban Indonesia.

Table 3. Omnibus test of model coefficient.

Omnibus

test

Chi-square df Sig.

(1) (2) (3)

51.742 10 0.000

Source: IDHS 2017 (processed)

3.2.3 Partial Test

The next stage is to test the significance of the

independent variable partially on the dependent

variable. Table 5 shows that the variables of

employment status, pregnancy visits and BMI status

were significant at the 5 percent level of significance.

This means that these variables partially affect

breastfeeding for children in urban Indonesia. Based

on the results of the partial test, the equation of the

binary logistic regression model is obtained as

follows:

𝑔





𝐷



 0,057  1,067𝐷



 0,964𝐷



 1,488𝐷



 0,544𝐷



 0,221𝐷



 0,035𝐷



 0,916𝐷



 0,645𝐷



 0,148𝐷



 0,169𝐷



(1)

There are four variables no significant effect on

breastfeeding in children. Mother's education level

has no effect on breastfeeding children. Fakhidah and

Palupi (2018) state that mothers with basic education

are no less advanced than mothers with higher

education in accessing information about

breastfeeding because they can use electronic media

or from health workers. The birth attendant had no

significant effect on breastfeeding the child. This

result is in line with Paschalia (2017) in her research,

which revealed that because birth attendants lacked

information about the practice of exclusive

breastfeeding, breastfeeding mothers did not

understand the benefits and advantages of exclusive

breastfeeding. In a study conducted by Sembiring

(2018), it was stated that after the delivery process,

the mother and baby were separated in different

rooms, the mother was in the patient room while the

baby was in the baby room so that the mother could

not give breast milk to her baby.

The ROC curve is used to obtain the cut value

which is then used to form a classification table.

Overall percentage value obtained by 83.1 percent.

This means that the model obtained as a whole is able

ICSDH 2021 - International Conference on Social Determinants of Health

192

to predict the breastfeeding status of children well by

83.1 percent. In addition, the model can classify 36.1

percent of mothers who do not give breast milk

properly and 85.6 percent of mothers who give breast

milk appropriately. The classification table is used as

a support for thetest Goodness of fit.

Comparison the value of the breastfeeding

variable for each change in the value or category of

the independent variable is seen using the odds ratio.

The trend ratio is calculated using the exponential

value of the regression coefficient.

1. Working

Status Mother's working status has a significant

effect on breastfeeding children. Mothers who do not

work have a positive effect on breastfeeding with a

coefficient of 0.544. This means that mothers who do

not work have a tendency to give breast milk by 1.723

times compared to mothers who work. This result is

in line with research conducted by Rahmawati (2010)

which states that mothers who do not work have the

opportunity to breastfeed their babies four times

compared to working mothers. Working mothers face

their own problems in providing breast milk for their

children. In short, the period of maternity or maternity

leave certainly disrupts breastfeeding efforts.

2. Early Initiation of Breastfeeding Status

Early initiation of breasfeeding status has a

significant effect on breastfeeding children. Mothers

who do IMD have a positive effect on breastfeeding

with a coefficient of 0.645. This means that mothers

who do IMD tend to give breast milk to their children

by 1.906 times compared to mothers who do not do

early initiation of breastfeeding. Babies who are

given the opportunity to breastfeed early by putting

the baby in skin-to-skin contact for at least one hour,

the result is twice as long as breastfeeding (Luluk and

Fitria, 20120). This result is also in accordance with

the descriptive which describes the percentage of

mothers who carry out early initiation of

breastfeeding giving more milk than mothers who do

not carry out early initiation of breasfeeding.

3. Pregnancy visits

Pregnancy visits have a significant effect on

breastfeeding children. Mothers who made pregnancy

visits four times or more had a positive effect on

breastfeeding with a coefficient of 0.916. This means

that mothers with four pregnancy visits or more tend

to breastfeed their children by 2,498 compared to

mothers with less than four pregnancy visits. These

results are in accordance with the research conducted

by Ogbo et al. (2019) stated that pregnancy visits

have a relationship with breastfeeding. In Chipo

(2019) explained that the knowledge and skills gained

during pregnancy visits can increase the mother's

confidence and intention to breastfeed her child. This

result is in accordance with the descriptive which

illustrates that the percentage of mothers who visited

pregnancy check-ups more than four times more gave

breast milk than mothers who checked their

pregnancy less than four times.

4 CONCLUSIONS AND

RECOMMENDATIONS

Based on the results and discussions that have been

described, the following conclusions can be drawn:

1. The status of breastfeeding for children under two

years old in urban areas in Indonesia has not yet

reached 100 percent. Mothers who breastfeed

their children the most with the characteristics:

high school education, not working, delivery

assisted by health workers, giving birth in health

facilities, having pregnancy visits four or more

times, carrying out early initiation of

breastfeeding, and rich economic status.

2. Variables that significantly affect breastfeeding

for children under two years old are: working

status of the mother, pregnancy visits, and early

initiation of breastfeeding.

3. The tendency of mothers who give breast milk to

children under two years old are mothers who do

not work, make pregnancy visits four or more

times, and mothers who carry out early initiation

of breastfeeding.

Suggestions that can be given by the author in

relation to the results of this study are:

1. For government or private agencies that have

female employees in breastfeeding conditions,

they are expected to provide adequate facilities for

breastfeeding such as providing lactation rooms

and so on.

2. The government and all parties can increase

educational activities in the form of counseling

and socialization about the importance of

maternal health during pregnancy so that at the

time of delivery they are ready to carry out IMD

and provide breast milk.

3. In this study, the variables used only looked at

aspects of the mother and household conditions.

In future research, it is possible to use variables

related to gender equality and variables from the

husband's aspect.

Affecting Variables Breastfeeding in Indonesian Urban Areas: 2017 IDHS Data Analysis

193

REFERENCES

Astuti, Isroni. (2013). Determinants of Exclusive

Breastfeeding in Breastfeeding Mothers. Journal of

Health Quality. Accessed on September 10, 2020 via

https://www.poltekkesjakarta1.ac.id/wp-

content/uploads/legacy/jurnal/document/41Jurnal_ISR

ONI.pdf

Central Bureau of Statistics. (2018). Maternal and Child

Health Profile. Jakarta: BPS.

BKKBN, BPS, Ministry of Health. (2018). Indonesia

Demographic and Health Survey 2017. Jakarta:

BKKBN, BPS, and Ministry of Health.

Hasiana, S., Ivone, J., & Wiharja, W. (2014). Effect of Early

Initiation of Breastfeeding on the Success of Exclusive

Breastfeeding. Accessed on 15 February 2021 via

https://repository.maranatha.edu/12604/

Hosmer, DW, & Lemeshow, S. (2000). Applied Logistic

Regression (2nd ed.). New York: John Wiley and Son.

Ministry of Health. (2014). Exclusive Breastfeeding

Situation and Analysis. Jakarta: Ministry of Health.

Ogbo. (2019). Regional prevalence and determinants of

exclusive breastfeeding in India. International

Breastfeeding Journal. Accessed on November 20,

2020 via https://doi.org/10.1186/s13006-019-0214-0

Government of Indonesia. (2009). Law of the Republic of

Indonesia Number 36 Year 2009 concerning Health.

Jakarta: Government of the Republic of Indonesia.

Indonesian government. (2014). Republic Law Number 35

of 2014 concerning Amendments to Law Number 23 of

2000 concerning Child Protection. Jakarta:

Government of the Republic of Indonesia.

Rahmawati, M. (2010). Factors influencing exclusive

breastfeeding for breastfeeding mothers in Pedalangan

Village, Banyumanik District, Semarang City. Kusuma

Husada Health Journal. Accessed on November 20,

2020 via

https://internationalbreastfeedingjournal.biomedcentral

.com/articles/10.1186/1746-4358-8-13?optIn=true

Tampah-Naah, A., & Kumi-Kyereme, A. (2013).

Determinants of exclusive breastfeeding among

mothers in Ghana: a cross-sectional study. Accessed

on November 20, 2020 via

https://internationalbreastfeedingjournal.biomedcentral

.com/articles/10.1186/1746-4358-8-13?optIn=true

ICSDH 2021 - International Conference on Social Determinants of Health

194