47

International Journal of Human and Health Sciences Vol. 05 No. 01 January’21

Original article :
Forecasting of Content Ca 125 Endometriosis Using Logistic Regression Model

Sardjana Atmadja1,  Gulam Gumilar2

Abstract
Objective : This study is to prove that there is a significant relationship between the 
absence of students participating in activities at school / on campus and the symptoms of 
primary dysmenorrhoea experienced during menstruation. Endometriosis is characterized 
as pain under the abdomen during menstruation. In addition, this study is also to obtain a 
profile of students and factors that influence primary dysmenorrhoea. A logistic regression 
model has been used to assess the main factors of dysmenorrhoea among these students. 
Methods : The study was conducted at the RSK Permata Hati Malang. A total of 123 
students were randomly selected in this study. The factors observed were menarche, 
menstruation, menstruation period and blood loss volume and CA 125 level. From the 
logistic regression model, it was found that there were three factors that influence the 
occurrence of dysmenorrhoea among students, namely menarche, menstruation period 
and menstrual blood volume. Results: The Hosmer and Lemeshow test showed that the 
measurement model of CA 125 levels in endometriosis was appropriate (Chi squar test 
value was 2.847 with p-value = 0.416). Instead of Press. (3) and Eq. (4), it was found that the 
contributors to dismenortea were menstrual length, menstrual discharge and the beginning 
of menarche. By looking at the odds ratio it is found that the risk of students experiencing 
dysmenorrhoea is (i) 2.5 times higher for those with longer menstrual periods (ii) 3.7 times 
higher for those who have menstrual expenditure which is a little and (iii) three times 
higher for those who have mined it for more than 13 years. Conclusion: Significant CA 
125 levels were obtained for students and students suffering from dysmenorrhoea. The 
study also found that the risk of getting dysmenorrhoea increased if students and students 
had menstrual periods longer than 35 days, menstrual expenditure levels were small and 
menarche was more than 13 years old.
Keywords: CA 125 level, logistic regression model; dysmenorrhoea and menarche.

Correspondence to: Sardjana Atmadja, Professor, Department of Obstetrics and Gynecology Syarif 
Hidayatullah State Islamic University Jakarta. Email : sardjana_spog@yahoo.com

1. Sardjana Atmadja, Professor, Department of Obstetrics and Gynecology Syarif Hidayatullah State 
Islamic University Jakarta

2. Gulam Gumilar, Assistant Professor, Department of Obstetrics and Gynecology Gajahmada University 
Yogyakarta.

International Journal of Human and Health Sciences Vol. 05 No. 01 January’21 Page : 47-49
DOI: http://dx.doi.org/10.31344/ijhhs.v5i1.232

Background
The use of logistic regression models has grown 
rapidly along with advances in medical science 
and technology (IPTEKDOK). The use of logistic 
regression in the field of clinical epidemiological 
inquiry has been used extensively in fields such 
as biomolkuler, ecology, clinical pharmacological 
investigations. Logistic regression analysis 
was first reserved by Cox (2010). The logistic 
regression model is a test of a linear linear model 
as introduced by Nelder and Wedderburn (2012). 
Hosmer and Lemeshow (2000) have discussed 
logistic regression models.

This paper discusses the use of logistic regression 
models to predict endometriosis among Malang 
students. Dysmenorrhoea is pain that is felt under 
the abdomen during menstruation. Dysmenorrhoea 
is a pain often faced by women. Endometriosis 
can cause students to not attend school / college. 
Sardjana (2018) has found that 20% of Malang 
students experience endometriosis and cannot go 
to school. Syamsul and colleagues (2017) report 
that 10% of career women experience serious pain 
caused by dysmenorrhoea and are not allowed to 
work. Alkaff (2016) reported that 52% of students 
in Yogyakarta were unable to do their daily 



International Journal of Human and Health Sciences Vol. 05 No. 01 January’21

48

activities well during menstruation.
Data
The study was conducted among students. A total 
of 123 students were randomly selected to answer 
the research questionnaire. Five non-leaning 
scales were selected for analysis.
Schedule 1. Change of study

Changes to 
Coded

Letters Kod

Dys experincing dismenorea or no
Dura 1 Menstrual period

Leng 1
Long time around 
menstruation

Menarke
The beginning of the 
menarche

Mens 1 Level of expenditure
Reg Menstruation is normal

 
Logistic Regression Model  
If π (x) = E (Y / x) is a conditional min if the 
distribution of logistics is used, then the logistic 
regression model is defined as follows:

π(x) =
















−+ ∑

=

n

i
i i

1
0exp1

1

χββ

(1)
where β0, β1, ..., βn are for non-leaning allowances 
and x1, x2, ..., x¬n are non-leaning change makers.
Logit incarnation of terhadap (x), provides logit 
for logit regression models for logistic regression 
models such as the following:

g(x) = In 
( )
( )









− x

x
λ

λ
1

  =  β0 + β1X1 + … + βnXn   (2)

Results and Analysis Step 2 provides pertinaian 
statistics as well as two chi-square tests conducted 
to determine whether there are differences between 
herds with dysmenorrhoea and non-dysminorrhea. 
From the steps it was found that the level of 
menstrual discharge and menstrual length gave a 
significant difference to dysmenorrhoea. Although 
menarche, the prevalence of menstruation and 
menstrual period is not meaningful. Using the 
backward logistic regression method using the 
exclusion criteria through the probability ratio 
test, the linear logistic regression model is given 
as follows:
π(x) = 

( ){ }321 077,1132,1916,0099,0exp1
1

xxx −+−−+

(3)
where x1 is the menstrual period, x2 is the rate of 
menstrual expenditure and x3 is the beginning of 
menarche.
Schedule 2. Menstrual cycle patterns for students

Characteristic
Dismenorea

n = 71

Not
Dismenorea

n = 52

P-value for 
the khi-

kuasadua 
test

Menarke
Min
Sisihan piawai
11-12 year
13 year
14-16 year

12,86
1,23
32
24
15

12,98
1,20
23
11
18

0,770

Normal menstruation
58 44 0,400

Not normal 13 7
Tempoh kitaran
<= 30 hari
> 30 hari

57
9

35
14

0,048

Periode menstruaation
<= 6 day
> 6 day

42
28

26
24

0,383

menstrual discharge 
levels
Lots
a little

48
22

46
6

0,010

Then, the logit for the logistic regression model is 
given by:

g(x) = In 
( )
( )









− x

x
λ

λ
1

 =  0,099 – 0,916x1 + 1,312x2  - 1,077x3 (4)
The Hosmer and Lemeshow tests show that the 
model fits well (the value of the second chi-power 
test is 2.847 with a p-value = 0.416). Instead 
of Press. (3) and Eq. (4), it was found that the 
contributors to dismenortea were menstrual length, 
menstrual discharge and the start of menarche. By 
looking at the odds ratio it is found that the risk 
of students experiencing dysmenorrhoea is (i) 2.5 
times higher for those who have high menstrual 
cycle periods, (ii) 3.7 times higher for those who 
has a small amount of menstrual expenditure 
and (iii) is three times higher for those who have 
menarcheenya more than 13 years.
Conclusion
From the research it was found that the prevalence 
of dysmenorrhoea among Malang students was 
58% and 20% reported being unable to attend 
college due to dysmenorrhoea. Press (4) provides 



49

International Journal of Human and Health Sciences Vol. 05 No. 01 January’21

the logit for a logistic linear regression model. 
From this model, it can be concluded that the 
three factors that influence dysmenorrhoea are 
the length of the menstrual period, the level of 
menstrual discharge and the start of menarche 
while the menstrual period and the prevalence of 
menstruation are not significant. This decision is 
the same as the research conducted by Andersch 
and Milsorm (in Ng and colleagues (2012)).
Other studies such as Harlow and Park (in Ng and 

colleagues (2012)) and Sundell and colleagues are 
Ng and colleagues (2012)] found dysmenorrhoea 
is influenced by several alterations including 
menarche and increased menstrual tempo. The 
study also found that the risk of experiencing 
dysmenorrhoea increases if students have a 
high menstrual cycle tempo, levels of menstrual 
expenditure are low and menarche is more than 13 
years old.

Reference:
1. Alkaf F ( 2016 ).Terapi Pilihan Endometriosis. 

Yogyakarta BP 19 : 9- 11
2. Cox, D.R. (2010). The Analysis Of Binary Data. 

London: Methuen and Co.
3. Hosmer, D.W. dan Lemeshow, S. (2010). Applied 

Logistic Regression, Second Edition. John Wiley & 
Sons.

4. Htut, Y., Amran, A. dan Shukri, Y.A (2017). A 
Prevalence Study Of Dysmenorrhoea In Students 
From Universiti Sains Malaysia, Perak Branch 
Campus, Tronoh. Medical journal Of Malaysia, 51: 
264-269.

5. Nelder, J.A. dan Weddeburn, R.W.M. 2012. 
Generalized Linear Models, Jour. Rpyal  Statistical 
Soc. A, 135: 370-384. 

6. Ng, T.R., Tan, T.C. dan Wansaicheong, G.K., (2012). 
A Prevalence Study Of Dysmenorrhoea In Female 
Residents Aged 15-54 Years In Clementi Town, 
Singapore, Annals Acad. Med. Singapore, 21(3): 323-
327.

7. Samsul  H ( 2017 ). Terapi alternative Endometriosis.
SBY,13 : 4- 5

8. Sardjana ( 2018 ) . Prevalensi Endometriosis pelajar 
dan mahasiswi di Malang.PIT 2000,22: 23- 27