Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 
  • Users Online: 201
  • Home
  • Print this page
  • Email this page
Cover page of the Journal of Health Sciences


 
 Table of Contents  
ORIGINAL ARTICLE
Year : 2022  |  Volume : 15  |  Issue : 2  |  Page : 121-125

Obesity and erectile dysfunction among type 2 diabetes mellitus patients: Applications of binary, ordinal, and polytomous logistic regression models


1 Department of Community Medicine, PSG Institute of Medical Sciences and Research, Coimbatore, Tamil Nadu, India
2 Department of Biostatistics, St. Thomas College, Palai, Kerala, India
3 Department of Endocrinology, PSG Institute of Medical Sciences and Research, Coimbatore, Tamil Nadu, India

Date of Submission29-Nov-2021
Date of Acceptance01-Feb-2022
Date of Web Publication24-May-2022

Correspondence Address:
Dr Senthil Kumar Rajasekaran
Department of Endocrinology, PSG Institute of Medical Sciences and Research, Coimbatore - 641 004, Tamil Nadu
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/kleuhsj.kleuhsj_266_21

Rights and Permissions
  Abstract 


PURPOSE: The purpose of this article is to provide a synthesized review of the models for analyzing data with ordinal response as the outcome variable and to evaluate their usefulness in examining the association of obesity with erectile dysfunction (ED) among type 2 diabetes mellitus patients.
MATERIALS AND METHODS: A total of 204 married men aged 20–60 years with a diagnosis of type 2 diabetes at the outpatient unit of the Department of Endocrinology at PSG Hospitals during May and June 2019 were studied. We examined the association between obesity and ED using binary, ordinal, and polytomous regression models.
RESULTS: Logistic regression model revealed that patients with diabetes who are obese, have a higher odds of ED (odds ratio = 2.508) after adjusting the effect of age and physical activity. In proportional odds model, obesity as compared with normal, independently increased the odds of ED (odds ratio = 2.264) across three severity categories. In the polytomous regression model, it was observed that mild ED has an odds ratio of 3.136 with respect to normal erectile function. It was observed that a cutoff value of waist circumference more than 92.71 cm would be suitable for the classification of ED in diabetic patients.
CONCLUSION: The study considered the ordinal nature of the outcome variable investigated and supported the association of obesity and ED among type 2 diabetes. In addition, we have also suggested an optimum cutoff value to waist circumference among diabetic patients with high ED.

Keywords: Erectile dysfunction, obesity, type 2 diabetes


How to cite this article:
Mathew AC, Benny B, Sunny D, Subramaniam S, Rajasekaran SK. Obesity and erectile dysfunction among type 2 diabetes mellitus patients: Applications of binary, ordinal, and polytomous logistic regression models. Indian J Health Sci Biomed Res 2022;15:121-5

How to cite this URL:
Mathew AC, Benny B, Sunny D, Subramaniam S, Rajasekaran SK. Obesity and erectile dysfunction among type 2 diabetes mellitus patients: Applications of binary, ordinal, and polytomous logistic regression models. Indian J Health Sci Biomed Res [serial online] 2022 [cited 2022 Jun 27];15:121-5. Available from: https://www.ijournalhs.org/text.asp?2022/15/2/121/345826




  Introduction Top


Diabetes mellitus is a chronic metabolic disorder associated with significant morbidity and mortality. In 2014, globally, 8.5% of adults aged above 18 years had diabetes WHO. Erectile dysfunction (ED) is a chronic condition in men. It is defined as a disability of getting or keeping a firm erection while having sexual intercourse.[1] Diabetes mellitus is one of the common comorbidities of ED. Many studies have reported that obesity is an independent risk factor for ED in men and have shown that obesity may increase the risk of ED by 30%–90% as compared with normal-weight participants.[2],[3] On the other hand, it was also observed that participants with ED tend to be heavier and with a greater waist than participants without ED and also were more likely to be hypertensive.[4] However, the association of obesity on ED among type 2 diabetes is unclear.

There are studies which have investigated the correlates of ED. Although the ED was measured on an ordinal scale, in most of these studies, these scales were often dichotomized and analyzed using standard logistic regression analysis. Although valid this approach loses information by collapsing some categories of the ordinal scale. In a recent article, Mathew et al. used proportional odds ordinal logistic regression models and continuation ratio models in examining the association of physical inactivity with ED among type 2 diabetic patients.[5] The polytomous logistic model has also been found useful in estimating the odds ratio when the outcome variable is in nominal scale and was used for multiple comparisons. The purpose of this article is twofold: first, to provide a synthesized review of the models for analyzing data with ordinal response as the outcome variable, and second, to evaluate their usefulness in examining the association of obesity with ED among type 2 diabetes mellitus patients.


  Materials and Methods Top


Study population and participants

This was a cross-sectional study among married men aged 20–60 years with a diagnosis of type 2 diabetes mellitus which was conducted at the Department of Endocrinology, PSG Hospitals, Coimbatore, during May and June 2019. All the participants were invited to a confidential interview during their hospital visit and the investigator collected responses from them. Erectile functioning scores were determined using the International Index of Erectile Function (IIEF).[6] IIEF contains 15 questions which is a validated, multidimensional, and self-administered investigation that has been found useful in the clinical assessment of ED and treatment outcomes in clinical trials. A preliminary report on the sociodemographic and clinical correlates of ED among type 2 diabetes has been published previously.[7]

Sample size calculation

With an expected prevalence of ED among type 2 diabetes mellitus patients as 40% and an allowable error of prevalence as 20% with 95% confidence interval and 25% nonresponse, the appropriate sample size was 204 to get reliable results for the study.

Inclusion criteria

All the confirmed type 2 diabetic patients attended the outpatient unit of the Department of Endocrinology at PSG Hospitals during May and June 2019 and those who were free from liver diseases, renal diseases, dialyzes, and patients without a history of coronary events in the previous 4 weeks were included in the study.

Ethical committee approval

The ethical committee approval for the study was obtained from the Institutional Human Ethics Committee PSG Institute of Medical Sciences and Research, Coimbatore, PSG/IHEC/2021/Renew/056.

Outcome variable

In the study, the outcome variable was erectile functioning scores which were obtained from the six questions (Q1, Q2, Q3, Q4, Q5, and Q15) of IIEF. A score of 0–5 was awarded to each of these six questions (maximum score is 30) that examine the erectile functioning status. The erectile functioning scores were divided into two categories. ≤14 as ED and >14 as normal. In the ordinal scale, ED was divided into three categories. >21 as normal, 12–21 as mild, and <12 as moderate/severe.[7]

Explanatory variable

Obesity was assessed based on the body mass index (BMI) scores and waist circumference of each patient. BMI scores were categorized as <30 kg/m2 as nonobese and ≥30 kg/m2 as obese.[8]

Data management and statistical analysis

Data management and statistical analysis were done using IBM SPSS Statistics for Windows, version 24.0 (IBM Corp., Armok, N.Y. USA).

Statistical models

Binary logistic regression

We have conducted binary logistic regression model to investigate the association of obesity on ED adjusting the effects of age and physical activity. Let y be the set of binary outcome variables and X be the set of explanatory variables. By taking the case of one explanatory variable X with one binary outcome variable y, the logistic model predicts the logit of y from X which represents a natural logarithm of odds of y.[9],[10] The logistic model is given as,



Receiver operating characteristic curve

The receiver operating characteristic (ROC) curve was then used to identify a suitable cutoff value for waist circumference in its relation to ED.[9],[10],[11]

Ordinal logistic (proportional odds) regression model

Further, we have used the ordinal logistic regression (proportional odds) model for the association of obesity with ED.

Proportional odds model

This model assumes that an explanatory variable exerts the same effect on each cumulative logit regardless of the cutoff g. Consider a proportional odds model with an outcome (D) with (G) levels (D = 0, 1,…, G-1) and one independent variable (X1). The probability that the disease outcome is in a category greater than or equal to G is given as,



Test of parallel lines

This is one of the assumptions to be satisfied for the use of proportional odds regression model and we have tested this assumption.[9],[10],[11],[12]

Polytomous regression model

Multiple comparisons of the association of obesity on the outcome variable have been done with polytomous regression model.[9],[10],[11],[12],[13]

In the study, there were three outcome categories and one predictor variable (X1), the polytomous model requires two regression expressions as given below,





These models were fitted to the data using IBM SPSS (24.0). P < 0.05 was considered statistically significant.


  Results Top


Of the 204 participants, 139 had normal erectile function, 35 had mild ED, and 30 had moderate/severe ED. Among these 139 participants, 127 (91.4%) were nonobese and the remaining 12 (8.6%) were obese. Among those who had mild ED, 27 (77.1%) were nonobese and 8 (22.9%) were obese. Similarly, in the moderate/severe category, 25 (83.3%) were nonobese and 5 (16.7%) were obese [Table 1].
Table 1: Prevalence of erectile dysfunction as related to obesity among type 2 diabetics

Click here to view


In logistic regression analysis, it was observed that the odds of ED among obese patients were 2.646 times higher than the nonobese (P < 0.05). After adjusting the effects of age and physical activity, the odds of ED among obese patients were 2.508 times greater than the nonobese (P < 0.05). The details are illustrated in [Table 2].
Table 2: Effect of obesity on erectile dysfunction among type 2 diabetics adjusting the effect of covariates in the association of obesity and erectile dysfunction

Click here to view


In proportional odds model, obesity as compared with normal, independently increased the odds of ED (odds ratio = 2.264, P < 0.05) across three severity categories. Test of parallel line analysis indicated that proportional odds assumption was satisfied (P = 0.235). The details are presented in [Table 3].
Table 3: Maximum likelihood estimate and model fitting information in proportional odds model in the association of obesity and erectile dysfunction

Click here to view


In the polytomous regression model, it was observed that mild ED has an odds ratio of 3.136 with respect to normal erectile function (P < 0.05). However, we could not find any significance for moderate/severe ED when compared it with normal erectile function (P = 0.193). The details are illustrated in [Table 4].
Table 4: Maximum likelihood estimate and model fitting information in polytomous regression model

Click here to view


The Area under the ROC curve obtained was 0.583. It was observed that a cutoff value of more than 92.71 cm would be suitable for the classification of ED in diabetic patients [Figure 1]. This has a sensitivity of 63% and specificity of 50% [Table 5].
Figure 1: Receiver operating characteristic curve of waist circumference in the association of erectile dysfunction

Click here to view
Table 5: Coordinates of the receiver operating characteristic curve

Click here to view



  Discussion Top


The clinical assessment of obese patients must include a detailed physical examinations and investigations to look out for associated cardiovascular, respiratory, endocrine, and metabolic comorbidities. However, despite their high prevalence, reproductive function abnormalities are often undertreated in the obese population compared to the cardiometabolic comorbidities. Obesity is associated with a constellation of endocrine disturbances and among them, reproductive functions abnormalities need more recognition as relevant clinical problems, particularly in those having diabetes.[14] In our study, we observed a positive association between obesity and ED among type 2 diabetes. Our findings support the hypothesis that low testosterone and high estradiol plasma levels are inversely related to BMI.[15] Many studies reported that metabolic complications associated with obesity, type 2 diabetes, in particular, affect gonadal function, and testosterone levels.[16],[17] In a study from Slovakia, 73% of men older than 40 years with abdominal obesity had some degree of ED. It also reported that those having obesity had three times greater risk for sexual dysfunction than the general population.[15] In our study, we observed a similar association of obesity and ED in type 2 diabetic patients. Different mechanisms have been suggested to explain how insulin resistance, type 2 diabetes, and hypogonadism are interconnected in males. Many studies have identified direct biological relationships between ED and obesity. Endothelial dysfunction reduces the production of nitric oxide which results in ED.[2] Decreased secretion of gonadotrophin-releasing hormone due to leptin resistance reduces the level of testosterone which causes ED.[15],[16],[17],[18] Obesity appears to damage the inner lining of blood vessels (endothelium) and this resists proper blood flow to the penis causing a difficulty in erection.[19] Aging men with obesity and the metabolic syndrome have a significant decrease in total serum testosterone level which leads to ED.[20]

In our study, in the multiple comparisons, we observed that obesity was statistically significant for those with mild ED compared with normal erectile function, whereas for the comparison of moderate/severe ED to normal erectile function, we could not find a statistical significance although the direction of association was positive. This may be due to the fact that those patients with diabetes and severe ED may have more duration of diabetes which may reduce their body weight, and thereby reduce their obesity level. In our study, those with an elevated waist circumference more than 92.71 cm were a high-risk group for ED among type 2 diabetics. Many studies have reported that an elevated waist circumference is the expression of increased visceral adipose tissue which can result in an enhancement in-situ of aromatase activity. This enzyme is highly expressed in adipose tissue and is thought to be responsible to the increased conversion of circulating testosterone to 17 β-estradiol in men with obesity which favors the development of secondary hypogonadism.[15]


  Conclusion Top


We observed a statistically significant association between obesity and ED among type 2 diabetes mellitus patients.

Our study is a cross-sectional study; hence, the cause-effect relationship cannot be established which is a major limitation. However, the study has several strengths. We have used a validated questionnaire to measure ED and also looked at the ordinal nature of the outcome variable in investigating the association with obesity.

What this study adds?

With the advanced statistical methods, this study considered the ordinal nature of the outcome variable investigated and supported the association of obesity and ED among type 2 diabetes. It also suggested an optimum cutoff value to waist circumference among diabetic patients with high ED.

Acknowledgments

The authors are thankful to Dr. Tadury Madhukar Subbarao, Principal, PSG Institute of Medical Sciences and Research for permitting us to do this study. We express our gratitude to Dr. Sudha Ramalingam, Director of Research and Innovation, HOD, Department of Community Medicine, PSG Institute of Medical Sciences and Research for permitting and providing necessary support and facilities for doing this study. We also express our gratitude to Dr. S.L Ravishankar, Professor, Department of Community Medicine, PSG Institute of Medical Sciences and Research for his valuable suggestions in improving the quality of the manuscript at various stages. We are thankful to Mr. Arjun V Sajeev and Mr. Sarath Mohan for their help in data collection.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
Hassan M, Naji M, Shams H, Sami O, Al-kuraishy H, Al-Gareeb A. Erectile dysfunction and type 2 diabetes mellitus: A new twist. Int J Nutr Pharmacol Neurol Dis 2020;10:43-9.  Back to cited text no. 1
    
2.
Skrypnik D, Bogdański P, Musialik K. Obesity – Significant risk factor for erectile dysfunction in men. Pol Merkur Lekarski 2014;36:137-41.  Back to cited text no. 2
    
3.
Tamler R. Diabetes, obesity, and erectile dysfunction. Gend Med 2009;6 Suppl 1:4-16.  Back to cited text no. 3
    
4.
Esposito K, Giugliano F, Ciotola M, De Sio M, D'Armiento M, Giugliano D. Obesity and sexual dysfunction, male and female. Int J Impot Res 2008;20:358-65.  Back to cited text no. 4
    
5.
Mathew AC, Siby E, Tom A, Kumar RS. Applications of proportional odds ordinal logistic regression models and continuation ratio models in examining the association of physical inactivity with erectile dysfunction among type 2 diabetic patients. Phys Act Nutr 2021;25:30-4.  Back to cited text no. 5
    
6.
Rosen RC, Riley A, Wagner G, Osterloh IH, Kirkpatrick J, Mishra A. The international index of erectile function (IIEF): A multidimensional scale for assessment of erectile dysfunction. Urology 1997;49:822-30.  Back to cited text no. 6
    
7.
Senthil KR, Arjun VS, Anandhu KJ, Sarath M, Darshan M, Yunsheng M, et al. Socio-demographic and clinical correlates of erectile dysfunction among men with type 2 diabetes mellitus: A cross sectional study in South India. Indian J Diabetes Endocrinol 2020;2:17-23.  Back to cited text no. 7
    
8.
Shiferaw WS, Akalu TY, Aynalem YA. Prevalence of erectile dysfunction in patients with diabetes mellitus and its association with body mass index and glycated hemoglobin in Africa: A systematic review and meta-analysis. Int J Endocrinol 2020;2020:5148370.  Back to cited text no. 8
    
9.
Kleinbaum DG, Klein M. Logistic Regression: A Self-Learning Text. New York: Springer-Verlag; 2010. p. 17.  Back to cited text no. 9
    
10.
Abdulqader Q. Applying the binary logistic regression analysis on the medical data. Sci J Univ Zakho 2017;5:330-4.  Back to cited text no. 10
    
11.
Habibzadeh F, Habibzadeh P, Yadollahie M. On determining the most appropriate test cut-off value: The case of tests with continuous results. Biochem Med (Zagreb) 2016;26:297-307.  Back to cited text no. 11
    
12.
Bender R, Grouven U. Ordinal logistic regression in medical research. J R Coll Physicians Lond 1997;31:546-51.  Back to cited text no. 12
    
13.
Etowa J, Hannan J, Etowa EB, Babatunde S, Phillips JC. Determinants of infant feeding practices among Black mothers living with HIV: A multinomial logistic regression analysis. BMC Public Health 2021;21:663.  Back to cited text no. 13
    
14.
Kalra S, Kapoor N, Bhattacharya S, Aydin H, Coetzee A. Barocrinology: The endocrinology of obesity from bench to bedside. Med Sci (Basel) 2020;8:51.  Back to cited text no. 14
    
15.
Moon KH, Park SY, Kim YW. Obesity and erectile dysfunction: From bench to clinical implication. World J Mens Health 2019;37:138-47.  Back to cited text no. 15
    
16.
Cheung KK, Luk AO, So WY, Ma RC, Kong AP, Chow FC, et al. Testosterone level in men with type 2 diabetes mellitus and related metabolic effects: A review of current evidence. J Diabetes Investig 2015;6:112-23.  Back to cited text no. 16
    
17.
Diaz-Arjonilla M, Schwarcz M, Swerdloff RS, Wang C. Obesity, low testosterone levels and erectile dysfunction. Int J Impot Res 2009;21:89-98.  Back to cited text no. 17
    
18.
Habous M, Tealab DA, Ali M, Raheem A, Ralph D, Binsaleh S. Obesity is an independent risk factor for low serum testosterone in adult males. J Mens Health 2015;11:30-4.  Back to cited text no. 18
    
19.
Barton M, Baretella O, Meyer MR. Obesity and risk of vascular disease: Importance of endothelium-dependent vasoconstriction. Br J Pharmacol 2012;165:591-602.  Back to cited text no. 19
    
20.
Wang C, Jackson G, Jones TH, Matsumoto AM, Nehra A, Perelman MA, et al. Low testosterone associated with obesity and the metabolic syndrome contributes to sexual dysfunction and cardiovascular disease risk in men with type 2 diabetes. Diabetes Care 2011;34:1669-75.  Back to cited text no. 20
    


    Figures

  [Figure 1]
 
 
    Tables

  [Table 1], [Table 2], [Table 3], [Table 4], [Table 5]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
   Abstract
  Introduction
   Materials and Me...
  Results
  Discussion
  Conclusion
   References
   Article Figures
   Article Tables

 Article Access Statistics
    Viewed178    
    Printed10    
    Emailed0    
    PDF Downloaded58    
    Comments [Add]    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]