Skip to main content

Advertisement

SpringerLink
Log in

Polycystic ovary syndrome: clinical and laboratory variables related to new phenotypes using machine-learning models

  • Original Article
  • Published:
Journal of Endocrinological Investigation Aims and scope Submit manuscript

Abstract

Purpose

Polycystic Ovary Syndrome (PCOS) is the most frequent endocrinopathy in women of reproductive age. Machine learning (ML) is the area of artificial intelligence with a focus on predictive computing algorithms. We aimed to define the most relevant clinical and laboratory variables related to PCOS diagnosis, and to stratify patients into different phenotypic groups (clusters) using ML algorithms.

Methods

Variables from a database comparing 72 patients with PCOS and 73 healthy women were included. The BorutaShap method, followed by the Random Forest algorithm, was applied to prediction and clustering of PCOS.

Results

Among the 58 variables investigated, the algorithm selected in decreasing order of importance: lipid accumulation product (LAP); abdominal circumference; thrombin activatable fibrinolysis inhibitor (TAFI) levels; body mass index (BMI); C-reactive protein (CRP), high-density lipoprotein cholesterol (HDL-c), follicle-stimulating hormone (FSH) and insulin levels; HOMA-IR value; age; prolactin, 17-OH progesterone and triglycerides levels; and family history of diabetes mellitus in first-degree relative as the variables associated to PCOS diagnosis. The combined use of these variables by the algorithm showed an accuracy of 86% and area under the ROC curve of 97%. Next, PCOS patients were gathered into two clusters in the first, the patients had higher BMI, abdominal circumference, LAP and HOMA-IR index, as well as CRP and insulin levels compared to the other cluster.

Conclusion

The developed algorithm could be applied to select more important clinical and biochemical variables related to PCOS and to classify into phenotypically different clusters. These results could guide more personalized and effective approaches to the treatment of PCOS.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Availability of data and material

The datasets generated during the current study are not publicly available but are available from the corresponding author on reasonable request.

Code availability

Python programming language and scikit-learn Random Forest implementation (https://scikit-learn.org/stable/).

References

  1. Meier R (2018) Polycystic ovary syndrome. Nurs Clin North Am 53(3):407–420

    Article  PubMed  Google Scholar 

  2. Azziz R (2018) Polycystic ovary syndrome. Obstet Gynecol 132(2):321–336

    Article  PubMed  Google Scholar 

  3. Bozdag G, Mumusoglu S, Zengin D, Karabulut E, Yildiz BO (2016) The prevalence and phenotypic features of polycystic ovary syndrome: a systematic review and meta-analysis. Hum Reprod 31:2841–2855

    Article  PubMed  Google Scholar 

  4. Patel S (2018) Polycystic ovary syndrome (PCOS), an inflammatory, systemic, lifestyle endocrinopathy. J Steroid Biochem Mol Biol 182:27–36

    Article  CAS  PubMed  Google Scholar 

  5. Oh J, Lee J, Lee H, Oh Y, Sung Y, Chung H (2009) Serum C-reactive protein levels in normal-weight polycystic ovary syndrome. Korean J Intern Med 24(4):350–355

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Hilali N, Vural M, Camuzcuoglu H, Camuzcuoglu A, Nurten A (2013) Increased prolidase activity and oxidative stress in PCOS. Clin Endocrinol (Oxf) 79(1):105–110

    Article  CAS  Google Scholar 

  7. - National Institutes of Health (2012) Evidence-based methodology workshop on polycystic ovary syndrome. December 3–5. Executive summary. Final report. https://prevention.nih.gov/docs/programs/pcos/FinalReport.pdf. Accessed 22 May 2021

  8. March W, Moore V, Willson K, Phillips D, Norman R, Davies M (2010) The prevalence of polycystic ovary syndrome in a community sample assessed under contrasting diagnostic criteria. Hum Reprod 25(2):544–551

    Article  PubMed  Google Scholar 

  9. Sóter M, Ferreira C, Sales M, Candido A, Reis FM, Milagres K, Ronda C, Silva I, Sousa M, Gomes K (2015) Peripheral blood-derived cytokine gene polymorphisms and metabolic profile in women with polycystic ovary syndrome. Cytokine 76(2):227–235

    Article  PubMed  Google Scholar 

  10. Tosatti J, Sóter M, Ferreira C, Silva I, Cândido A, Sousa M, Reis FM, Gomes K (2020) The hallmark of pro- and anti-inflammatory cytokine ratios in women with polycystic ovary syndrome. Cytokine 134:155187

    Article  CAS  PubMed  Google Scholar 

  11. Carvalho L, Ferreira C, Sóter M, Sales M, Rodrigues K, Martins S, Candido A, Reis FM, Silva I, Campos F, Gomes K (2017) Microparticles: inflammatory and haemostatic biomarkers in polycystic ovary syndrome. Mol Cell Endocrinol 443:155–162

    Article  CAS  PubMed  Google Scholar 

  12. Carvalho L, Ferreira C, Oliveira D, Rodrigues K, Duarte R, Teixeira M, Xavier L, Candido A, Reis F, Silva I, Campos F, Gomes K (2017) Haptoglobin levels, but not Hp1-Hp2 polymorphism, are associated with polycystic ovary syndrome. J Assist Reprod Genet 34(12):1691–1698

    Article  PubMed  PubMed Central  Google Scholar 

  13. Xavier L, Sóter M, Sales M, Oliveira D, Reis H, Candido A, Reis FM, Silva I, Gomes K, Ferreira C (2018) Evaluation of PCSK9 levels and its genetic polymorphisms in women with polycystic ovary syndrome. Gene 644:129–136

    Article  CAS  PubMed  Google Scholar 

  14. Carvalho L, Ferreira C, Candido A, Reis FM, Sóter M, Sales M, Silva I, Nunes F, Gomes K (2017) Metformin reduces total microparticles and microparticles-expressing tissue factor in women with polycystic ovary syndrome. Arch Gynecol Obstet 296(4):617–621

    Article  CAS  PubMed  Google Scholar 

  15. Sales M, Sóter M, Candido A, Fernandes A, Oliveira F, Ferreira A, Sousa M, Ferreira C, Gomes K (2013) Correlation between plasminogen activator inhibitor-1 (PAI-1) promoter 4G/5G polymorphism and metabolic/proinflammatory factors in polycystic ovary syndrome. Gynecol Endocrinol 29(10):936–939

    Article  CAS  PubMed  Google Scholar 

  16. Xavier L, Gontijo N, Rodrigues K, Cândido A, Reis F, Sousa M, Silveira J, Oliveira F, Ferreira C, Gomes K (2019) Polymorphisms in vitamin D receptor gene, but not vitamin D levels, are associated with polycystic ovary syndrome in Brazilian women. J Gynecol Endocrinol 35(2):146–149

    Article  CAS  Google Scholar 

  17. Reis G, Gontijo N, Rodrigues K, Alves M, Ferreira C, Gomes K (2017) Vitamin D receptor polymorphisms and the polycystic ovary syndrome: a systematic review. J Obstet Gynaecol Res 43(3):436–446

    Article  PubMed  Google Scholar 

  18. Alves M, de Souza I, Ferreira C, Cândido AL, Bizzi M, OliveiraReisGomes FFK (2020) Galectin-3 is a potential biomarker to insulin resistance and obesity in women with polycystic ovary syndrome. Gynecol Endocrinol 36(9):760–763

    Article  CAS  PubMed  Google Scholar 

  19. Oliveira F, Mamede M, Bizzi M, Rocha A, Ferreira C, Gomes K, Cândido AL, Reis F (2019) Brown adipose tissue activity is reduced in women with polycystic ovary syndrome. Eur J Endocrinol 181(5):473–480

    Article  CAS  PubMed  Google Scholar 

  20. Rahul C (2015) Machine learning in medicine. Circulation 132(20):1920–1930

    Article  Google Scholar 

  21. Saber H, Somai M, Rajah G, Scalzo F, Liebeskind D (2019) Predictive analytics and machine learning in stroke and neurovascular medicine. Neurol Res 41(8):681–690

    Article  PubMed  Google Scholar 

  22. Handelman G, Kok H, Chandra R, Razavi A, Lee M, Asadi H (2018) eDoctor: machine learning and the future of medicine. J Intern Med 284(6):603–619

    Article  CAS  PubMed  Google Scholar 

  23. Breiman L (2001) Random forests. Mach Learn 45(1):5–32

    Article  Google Scholar 

  24. Lundberg M, Erion G, Chen H, DeGrave A, Prutkin J, Nair B, Katz R, Himmelfarb J, Bansal N, Lee S (2020) From local explanations to global understanding with explainable AI for trees. Nat Mach Intell 2(1):56–67

    Article  PubMed  PubMed Central  Google Scholar 

  25. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group (2004) The Rotterdam ESHRE/ASRM revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 81:19–25

    Article  Google Scholar 

  26. Santos R (2001) III Diretrizes Brasileiras Sobre Dislipidemias e Diretriz de Prevenção da Aterosclerose do Departamento de Aterosclerose da Sociedade Brasileira de Cardiologia. Arq Bras Cardiol 77(3):1–25

    Article  PubMed  Google Scholar 

  27. Tang Q, Xueqin L, Song P, Xu L (2015) Optimal cut-off values for the homeostasis model assessment of insulin resistance (HOMA-IR) and pre-diabetes screening: Developments in research and prospects for the future. Drug Discov Ther 9(6):380–385

    Article  CAS  PubMed  Google Scholar 

  28. Lwow F, Jedrzejuk D, Milewicz A, Szmigiero L (2016) Lipid accumulation product (LAP) as a criterion for the identification of the healthy obesity phenotype in postmenopausal women. Exper Gerontol 82:81–87

    Article  CAS  Google Scholar 

  29. Keany E (2021) BorutaShap 1.0.15 2020. https://pypi.org/project/BorutaShap/. Accessed 26 May 2021

  30. Kursa M, Rudnicki W (2010) Feature selection with the Boruta package. J Stat Softw 36(11):1–13

    Article  Google Scholar 

  31. Bock H (2007) Clustering methods: a history of k-means algorithms. Selected contributions in data analysis and classification. Springer, Berlin, pp 161–172

    Google Scholar 

  32. Teede H, Misso M, Costello M, Dokras A, Laven J, Moran L, Piltonen T, Norman R, International PCOS Network (2018) Recommendations from the international evidence-based guideline for the assessment and management of polycystic ovary syndrome. Hum Reprod 33(9):1602–1618

    Article  PubMed  PubMed Central  Google Scholar 

  33. Pall M, Azziz R, Beires J, Pignatelli D (2010) The phenotype of hirsute women: a comparison of polycystic ovary syndrome and 21-hydroxylase-deficient nonclassic adrenal hyperplasia. Fertil Steril 94(2):684–689

    Article  CAS  PubMed  Google Scholar 

  34. Saadia Z (2020) Follicle stimulating hormone (LH: FSH) ratio in polycystic ovary syndrome (PCOS) obese vs non- obese women. Med Arch 74(4):289–293

    Article  PubMed  PubMed Central  Google Scholar 

  35. Speiser P, Knochenhauer E, Dewailly D, Fruzzetti F, Marcondes J, Azziz R (2000) A multicenter study of women with nonclassical congenital adrenal hyperplasia: relationship between genotype and phenotype. Mol Genet Metab 71:527–534

    Article  CAS  PubMed  Google Scholar 

  36. Qiu L, Liu J, Hei Q (2015) Association between two polymorphisms of follicle stimulating hormone receptor gene and susceptibility to polycystic ovary syndrome: a meta-analysis. Chin Med Sci J 30(1):44–50

    Article  PubMed  Google Scholar 

  37. Deniz R, Yavuzkir S, Ugur K, Ustebay D, Baykus Y, Ustebay S, Aydin S (2021) Subfatin and asprosin, two new metabolic players of polycystic ovary syndrome. J Obstet Gynaecol 41(2):279–284

    Article  CAS  PubMed  Google Scholar 

  38. Glintborg D, Altinok M, Mumm H, Buch K, Ravn P, Andersen M (2014) Prolactin is associated with metabolic risk and cortisol in 1007 women with polycystic ovary syndrome. Hum Reprod 29:1773–1779

    Article  CAS  PubMed  Google Scholar 

  39. Yang H, Di J, Pan J, Yu R, Teng Y, Cai Z, Deng X (2020) The Association between prolactin and metabolic parameters in pcos women: a retrospective analysis. Front Endocrinol (Lausanne) 11:263

    Article  Google Scholar 

  40. Corona G, Wu F, Rastrelli G, Lee D, Forti G, O’Connor D, O’Neill T, Pendleton N, Bartfai G, Boonen S, Casanueva F, Finn J, Huhtaniemi I, Kula K, Punab M, Vanderschueren D, Rutter M, Maggi M, EMAS Study Group (2014) Low prolactin is associated with sexual dysfunction and psychological or metabolic disturbances in middle-aged and elderly men: the European male aging study (EMAS). J Sex Med 11(1):240–253

    Article  CAS  PubMed  Google Scholar 

  41. Wagner R, Heni M, Linder K, Ketterer C, Peter A, Bohm A, Hatziagelaki E, Stefan N, Staiger H, Häring H, Fritsche A (2014) Age-dependent association of serum prolactin with glycaemia and insulin sensitivity in humans. Acta Diabetol 51(1):71–78

    Article  CAS  PubMed  Google Scholar 

  42. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group (2004) Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 19(1):41–47

    Article  Google Scholar 

  43. Cupisti S, Haeberle L, Schell C, Richter H, Schulze C, Hildebrandt T, Oppelt P, Beckmann M, Dittrich R, Mueller A (2011) The different phenotypes of polycystic ovary syndrome: no advantages for identifying women with aggravated insulin resistance or impaired lipids. Exp Clin Endocrinol Diabetes 119:502–508

    Article  CAS  PubMed  Google Scholar 

  44. Mehrabian F, Khani B, Kelishadi R, Kermani N (2011) The prevalence of metabolic syndrome and insulin resistance according to the phenotypic subgroups of polycystic ovary syndrome in a representative sample of Iranian females. J Res Med Sci 16:763–769

    CAS  PubMed  PubMed Central  Google Scholar 

  45. Shroff R, Syrop C, Davis W, Van Voorhis B, Dokras A (2007) Risk of metabolic complications in the new PCOS phenotypes based on the Rotterdam criteria. Fertil Steril 88:1389–1395

    Article  CAS  PubMed  Google Scholar 

  46. Lizneva D, Suturina L, Walker W, Brakta S, Gavrilova-Jordan L, Azziz R (2016) Criteria, prevalence, and phenotypes of polycystic ovary syndrome. Fertil Steril 106(1):6–15

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

FMR, AAV and KBG are grateful to CNPq for the research fellowship.

Funding

The grants from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq).

Author information

Author notes

  1. A. A. Veloso and K. B. Gomes contributed equally to this study.

Authors and Affiliations

  1. Departamento das Ciências da Computação, Instituto de Ciências Exatas, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil

    I. S. Silva & A. A. Veloso

  2. Colégio Técnico, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil

    C. N. Ferreira

  3. Departamento de Genética, Ecologia e Evolução, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil

    L. B. X. Costa, L. M. L. Carvalho, M. F. Sales & K. B. Gomes

  4. Departamento de Análises Clínicas e Toxicológicas, Faculdade de Farmácia, Universidade Federal de Minas Gerais, Av. Antônio Carlos, 6627, Pampulha, Belo Horizonte, MG, 31270-901, Brasil

    M. O. Sóter, J. de C. Albuquerque & K. B. Gomes

  5. Departamento de Clínica Médica, Faculdade de Medicina, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil

    A. L. Candido

  6. Departamento de Ginecologia e Obstetrícia, Faculdade de Medicina, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil

    F. M. Reis

Authors
  1. I. S. Silva
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. C. N. Ferreira
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. L. B. X. Costa
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. M. O. Sóter
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. L. M. L. Carvalho
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. J. de C. Albuquerque
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. M. F. Sales
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. A. L. Candido
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. F. M. Reis
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. A. A. Veloso
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. K. B. Gomes
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization: ISS, AAV, and KBG. Data curation: CNF, LBXC, MOS, LMLC, JA, MFS, ALC, and FMR. Formal analysis: ISS. Funding acquisition: KBG. Investigation: ISS, AAV, FMR, and KBG. Methodology: ISS and AAV. Project administration: KBG. Resources: KBG. Software: ISS and AAV. Supervision: AAV and KBG. Validation: ISS. Visualization: ISS. Roles/writing—original draft: ISS and KBG. Writing—review and editing: ISS, FMR, and KBG.

Corresponding author

Correspondence to K. B. Gomes.

Ethics declarations

Conflict of interest

The authors declare that they have no conflicts of interest or financial/personal relationships that could have appeared to influence the work reported in this paper.

Ethical approval

The Ethics Committee (COEP) of the Federal University of Minas Gerais (UFMG) approved the study (CAAE 0379.0.203.000-11). We certify that the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki.

Informed consent

Written informed consent was obtained from all participants.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 Figure 1: Example of decision tree. (TIF 244 KB)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Silva, I.S., Ferreira, C.N., Costa, L.B.X. et al. Polycystic ovary syndrome: clinical and laboratory variables related to new phenotypes using machine-learning models. J Endocrinol Invest 45, 497–505 (2022). https://doi.org/10.1007/s40618-021-01672-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40618-021-01672-8

Keywords

Search

Navigation