Received: 26 December 2017  |  Revised: 5 June 2018  |  Accepted: 6 August 2018  |  First published online: 27 August 2018 DOI: 10.1002/ijgo.12642 S P E C I A L A R T I C L E G y n e c o l o g y Why we need epidemiologic studies of polycystic ovary syndrome in Africa Ernest T. Maya1,2,* | Chris B. Guure1 | Richard M.K. Adanu1 | Bismark Sarfo1 |  Michael Ntumy2 | Evelyn Y. Bonney3 | Daria Lizneva4,5 | Walidah Walker4 |  Ricardo Azziz4,6,7,8 1School of Public Health, University of Ghana, Accra, Ghana Abstract 2School of Medicine & Dentistry, University of The primary objective of the Ghana Polycystic Ovary Syndrome Epidemiology and Phenotype Ghana, Accra, Ghana (Ghana-P EP) study will be to assess the relevance and phenotypic distribution of polycystic ovar- 3Noguchi Memorial Institute for Medical ian syndrome (PCOS) in a medically unbiased population of reproductive- aged women. In addi- Research, University of Ghana, Accra, Ghana tion, the study will also attempt to identify sociodemographic, environmental, and psychological 4Department of Obstetrics & factors that may play a role in the development of PCOS phenotype. The study aims to recruit 990 Gynecology, Augusta University, Augusta, randomly selected women aged 18–45 years living in Nsawam, the district capital of the Nsawam- GA, USA 5 Adoagyiri Municipality, in the Eastern region of Ghana. Participants will complete a questionnaire Division of Endocrinology, Diabetes and Bone Disease, Icahn School of Medicine at with the aid of trained personnel, undergo a physical examination, and undergo ultrasonography Mount Sinai, Icahn School of Medicine at and biochemical evaluations relevant to PCOS. It is anticipated that the study will provide the Mount Sinai, New York, NY, USA population prevalence and phenotypes, and distribution of PCOS. 6The State University of New York System Administration, Albany, NY, USA 7Department of Health Policy, Management & K E Y W O R D S : Ghana; PCOS; Phenotypes; Polycystic ovarian syndrome; Prevalence Behavior, School of Public Health, University at Albany, Albany, NY, USA 8Department of Obstetrics & Gynecology, Albany Medical College, Albany, NY, USA *Correspondence Ernest T. Maya, School of Public Health, University of Ghana, Legon, Accra, Ghana. Emails: emaya@ug.edu.gh; maya_ernest@yahoo.co.uk Polycystic ovarian syndrome (PCOS) is a common endocrine–meta- is also associated with an increased risk of endometrial and possibly bolic disorder. Although it is most clinically apparent among women ovarian carcinoma. Mood disturbances and psychosexual dysfunction of reproductive age, PCOS can be symptomatic in pre-a dolescent are more frequent in PCOS. Lastly, most women with PCOS, regardless and menopausal women, and potentially even in men.1 Typically, the of body mass, have variable degrees of chronic subacute inflammation syndrome is characterized by chronic oligo-a novulation, biochemical and insulin resistance, which are associated with an increased risk of and/or clinical hyperandrogenism, and polycystic ovarian morphol- type 2 diabetes mellitus, dyslipidemia, and vascular disorders including ogy (PCOM). In addition to hyperandrogenic dermatologic symptoms cerebrovascular incidents and possibly even cardiovascular disease.1,2 (acne, alopecia, and hirsutism), PCOS is associated with impaired As a syndrome, PCOS is defined by a collection of signs and symp- reproduction and obstetric outcomes.2 toms after the exclusion of related or mimicking disorders. In 1990, Women with PCOS seem to be more frequently obese than their the relatively strict US National Institutes of Health (NIH) criterion non- affected counterparts, although the degree of this association categorized two phenotypes of PCOS: phenotype A, comprising is weak when medically unbiased populations are studied.3,4 PCOS oligo-a novulation, hyperandrogenism, and PCOM; and phenotype B, Int J Gynecol Obstet 2018; 143: 251–254 wileyonlinelibrary.com/journal/ijgo © 2018 International Federation of  |  251 Gynecology and Obstetrics 252  |     Maya ET aL. comprising oligo- anovulation and hyperandrogenism.5 Subsequently, true prevalence and phenotype of the disorder in the region studied. the 2006 Androgen Excess and PCOS (AE- PCOS) Society diagnostic Second, they will allow investigators to assess the impact of differ- criterion included phenotype C, comprising hyperandrogenism and ences in race and/or ethnicity, environment, socioeconomics, and diet PCOM, in addition to A and B,6 while the 2003 Rotterdam criterion and nutrition on the development, complications, phenotype, and of the European Society for Human Reproduction and Embryology prevalence of PCOS. Third, they are essential for determining the rela- (ESHRE) and the American Society for Reproductive Medicine (ASRM) tionship between genotype and phenotype, potentially fostering an additionally included phenotype D (oligo-a novulation and PCOM).7,8 improved understanding of the molecular mechanisms underlying the Phenotypically, the Rotterdam 2003 and AE- PCOS Society 2006 defi- disorder. Fourth, well conducted epidemiologic studies may provide nitions are effectively expansions of the NIH 1990 criterion. In 2012, a clues to the evolutionary history of the disorder, which in turn could consensus workshop sponsored by the NIH recommended use of the potentially assist identification of the core elements of PCOS. Last, Rotterdam 2003 criterion to identify the broadest number of affected they will lead to a better assessment of the public health and economic individuals, coupled with the need to define the specific phenotype of implications of PCOS in a region, and facilitate the development of an affected individuals (i.e. as phenotypes A–D).9 Of note, the phenotypic informed, cogent, and effective public health and prevention policy. presentation of PCOS shows ethnic diversity.10 The earliest epidemiologic study to evaluate PCOS was carried There is strong evidence that the features of PCOS identified out in 1998 by Azziz and colleagues,18 who assessed reproductive- among clinically referred patients are more severe than the phe- aged women undergoing pre- employment physical examinations at notypes of PCOS identified in studies of medically unbiased (unse- a university in the southern United States. Based on the NIH 1990 lected) populations. For example, Ezeh et al.3 compared two cohorts criteria, the study observed a prevalence of PCOS of 4%, which varied from the same geographic area, the first consisting of individuals little by race (black or white).18 Subsequent studies have verified and with PCOS seeking medical care and the second comprising individ- expanded this assessment. To date, the reported prevalence of PCOS uals with PCOS identified during a routine pre- employment health varies from 5% to 20%, depending on factors such as which diagnostic assessment.3 They reported that the referred (medically biased) indi- criterion is used, how the study population is identified, the methods viduals more frequently showed the more complete PCOS pheno- used to define each phenotypic feature, and the completeness of the type (phenotype A), were more hirsute, had higher serum androgen phenotypic assessment and recruitment process of the populations. levels, and were more obese as compared with individuals identi- For example, a recent meta- analysis found that the reported preva- fied in the unselected cohort. A similar conclusion was reached in lence of PCOS based on the diagnostic criteria of the 1990 NIH, 2003 a study of Spanish women screened during blood donation.11 At Rotterdam, and 2006 AE- PCOS Society was 6% (95% confidence least in part, this bias is determined by the negative impact of PCOS interval [CI] 5%–8%; 18 trials), 10% (95% CI 8%–13%; 15 trials), and features, particularly obesity and hirsutism, on quality of life12–14 10% (95% CI 7%–13%; 10 trials), respectively.19 and ability to access medical care.3 A meta- analysis of studies on However, most if not all studies of PCOS prevalence are from pop- PCOS has provided further evidence of referral bias for the PCOS ulations in North America, Europe, the Middle East, southern Asia, phenotype.4 Consequently, to gain the most accurate understanding and Australia (Fig. 1).5 In fact, there are no significant data from South of the prevalence of PCOS, epidemiologic studies of unselected or America, Russia (i.e., northern Asia), the island countries of Oceania unbiased populations must be undertaken. (Melanesia, Micronesia, and Polynesia), or Africa. Also notable is the Polycystic ovarian syndrome seems to be a complex prehistoric paucity of studies among black women, with few exceptions.18, 20 genetic trait, possibly dating back 50 000 years or longer.15 Thus, An assessment of studies in Africa highlights the lack of research despite its clear reproductive disadvantages, the disorder has appar- among black women in Sub- Saharan Africa, where there have been no ently persisted for tens of thousands of years. Various potential bene- large- scale epidemiologic studies of PCOS and only a very few studies fits of PCOS among ancestral women have been considered to account of the PCOS phenotype.21–25 This is particularly alarming given the for this paradox, including metabolic, immune, and musculoskeletal need to improve women’s health in the region26 and the great need to benefits, patterns of child- rearing and mothering, reproductive longev- address health disparities globally.27 Furthermore, PCOS among black ity, and in utero or childhood survival advantages, which might have women may be associated with additional or more severe morbidities, had a significant role in our hunter–gatherer past.15 In reality, however, such as uterine leiomyomata, as compared with white women.28 The there is little evidence to indicate that the persistence of PCOS is due negative impact that PCOS has on fertility is particularly harmful for to direct positive selection. Instead, the evolution of PCOS has prob- African women in low socioeconomic settings. Many are affected by ably been driven by non-a daptive mechanisms, including genetic drift psychosocial effects such as anxiety and depression, stress, intimate and population balance.15–17 Among other approaches, insight into the partner violence, and divorce.29,30 origins of PCOS may emerge through the epidemiologic study of dif- It is clear that additional research is needed to address the gap in ferent populations around the globe, particularly populations whose studies of this highly prevalent and morbid disorder. To begin to address diet, fertility, and disease load resemble those of ancestral humans. this deficit, the Ghana Polycystic Ovary Syndrome Epidemiology and Overall, the available data suggest that well- controlled epidemi- Phenotype (Ghana- PEP) study is undertaking a large community- ologic studies of unselected (medically unbiased) populations are crit- based assessment of PCOS in the Nsawam municipality of Ghana. ically needed worldwide. First, such studies will help to define the The study has been approved by the Noguchi Memorial Institute for Maya ET aL.      |  253 F IGURE  1 Map showing the location of PCOS prevalence studies conducted worldwide through 2015. Red dots indicate individual study centers; green circles indicate regions of the world where the epidemiology of PCOS remains unknown or understudied. Medical Research, University Of Ghana, Institutional Review Board this information will help to document the public health and eco- (CPN 064/16- 17), and Ghana Health Service Ethics Review Committee nomic burden of PCOS in the region, and to foster the establishment (GHS- ERC-1 1-0 3-2 016). of sound public health policies to help to address this common and The primary endpoint of the Ghana- PEP study will be the prev- morbid syndrome. alence and phenotypic distribution of PCOS in a medically unbi- ased population of reproductive- aged women; secondary endpoints CONFLICTS OF INTEREST include the frequency of other common health disorders of women (overweightness and obesity, uterine leiomyomata, glucose intoler- RA is a consultant to Ansh Labs, Fractyl, Medtronics, and Longitude ance, and thyroid dysfunction, among others).31 The study will be Capital, and is on the advisory board of Global PET Imaging. RMKA is the first of its kind in Africa and will provide data not only for Ghana, the Editor of the International Journal of Gynecology and Obstetrics. but also for the broader Sub- Saharan African region. The study is The authors have no other conflicts of interest. a collaboration between researchers at the University of Ghana Schools of Public Health, Medicine and Dentistry, and Noguchi Memorial Institute for Medical Research, and researchers based in AUTHOR CONTRIBUTIONS the USA. ETM contributed to the design of the study and to writing the manu- The study plans to recruit and assess 990 randomly selected script. CG, BS, MN, EB, DL, and WW contributed to the design and women aged 18–45 years living in Nsawam, the district capital of planning of the study, and to writing the manuscript. RMKA con- the Nsawam–Adoagyiri municipality, in the eastern region of Ghana. tributed to the design and planning of the study, and to writing and The women will complete a questionnaire with the aid of trained revising the manuscript. RA contributed to designing, planning, and personnel, undergo a physical examination, and ultrasonography conducting the study, and to writing and revising the manuscript. All and biochemical evaluations. One of the strengths of the study is authors read and approved the final manuscript. that it will be community- based rather than clinic- based, thereby avoiding the bias that characterizes clinic- based studies. In addition to the primary and secondary endpoints noted above, the Ghana- REFERENCES PEP study will attempt to identify sociodemographic, environmental, 1. Azziz R, Carmina E, Chen Z, et al. Polycystic ovary syndrome. Nat Rev and psychologic factors that might be involved in the development Dis Prim. 2016;2:16057. of PCOS phenotype. 2. Zore T, Joshi NV, Lizneva D, Azziz R. Polycystic ovarian syndrome: Long- Understanding better the prevalence and phenotypes of PCOS term health consequences. Semin Reprod Med. 2017;35:271–281. in Sub- Saharan Africa will help to build local capacity in PCOS and 3. Ezeh U, Yildiz BO, Azziz R. Referral bias in defining the phenotype and prevalence of obesity in polycystic ovary syndrome. J Clin Endocrinol reproductive endocrine research. It is anticipated that these data Metab. 2013;98:1088–1096. will facilitate the development of locally tailored early screening 4. Lizneva D, Kirubakaran R, Mykhalchenko K, et al. Phenotypes and strategies and tools, which may lead to earlier diagnosis and man- body mass in women with polycystic ovary syndrome identified in agement of affected individuals and potentially a reduction in the referral versus unselected populations: Systematic review and meta- analysis. Fertil Steril. 2016;106:1510–1520.e2. long-t erm sequelae of PCOS in the region. The study will also facil- 5. Lizneva D, Suturina L, Walker W, Brakta S, Gavrilova-Jordan L, Azziz itate the development of education materials tailored to both local R. Criteria, prevalence, and phenotypes of polycystic ovary syndrome. healthcare providers and the population in general. Furthermore, Fertil Steril. 2016;106:6–15. 254  |     Maya ET aL. 6. Azziz R, Carmina E, Dewailly D, et al. Position statement: Criteria women of the Southeastern United States: A prospective study*. J for defining polycystic ovary syndrome as a predominantly hyper- Clin Endocrinol Metab. 1998;83:3078–3082. androgenic syndrome: An androgen excess society guideline. J Clin 19. Bozdag G, Mumusoglu S, Zengin D, Karabulut E, Yildiz BO. The prev- Endocrinol Metab. 2006;91:4237–4245. alence and phenotypic features of polycystic ovary syndrome: A sys- 7. PCOS Consensus Workshop Group TRE-SPCWG. Revised 2003 con- tematic review and meta- analysis. Hum Reprod. 2016;31:2841–2855. sensus on diagnostic criteria and long- term health risks related to 20. Azziz R, Woods KS, Reyna R, Key TJ, Knochenhauer ES, Yildiz BO. polycystic ovary syndrome. Fertil Steril. 2004;81:19–25. The prevalence and features of the polycystic ovary syndrome in an 8. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop unselected population. J Clin Endocrinol Metab. 2004;89:2745–2749. Group. Revised 2003 consensus on diagnostic criteria and long- 21. Wada I, Matson PL, Macnamee MC, Brinsden PRLB. High ovarian term health risks related to polycystic ovary syndrome (PCOS). Hum response in Yoruba African women during ovulation induction for Reprod. 2004;19:41–47. assisted conception. Hum Reprod. 1994;9:1077–1080. 9. National Institutes of Health. Evidence-based methodology work- 22. Pembe AB, Abeid MS. Polycystic ovaries and associated clinical and shop on polycystic ovary syndrome, December 3-5, 2012. Executive biochemical features among women with infertility in a tertiary hospi- summary. NIH. 2012. tal in Tanzania. Tanzan J Health Res. 2009;11:175–180. 10. Williamson K, Gunn AJ, Johnson N, Milsom SR. The impact of eth- 23. Amisi C, Mputu L, Mboloko E, Bieleli E, Pozzili P. Biological insulin nicity on the presentation of polycystic ovarian syndrome. Aust N Z J resistance in Congolese woman with polycystic ovary syndrome Obstet Gynaecol. 2001;41:202–206. (PCOS). Gynecol Obstet Fertil. 2013;41:707–710. 11. Luque-Ramírez M, Escobar-Morreale HF. Polycystic ovary syndrome 24. Ugwu GO, Iyoke CA, Onah H, Mba S. Prevalence, presentation, and as a paradigm for prehypertension, prediabetes, and preobesity. Curr management of polycystic ovary syndrome in Enugu, South East Hypertens Rep. 2014;16:500. Nigeria. Niger J Med. 2013;22:313–316. 12. Guyatt G, Weaver B, Cronin L, Dooley JA, Azziz R. Health- 25. Ogueh O, Zini M, Williams S, Ighere J. Polycystic ovary morphology related quality of life in women with polycystic ovary syndrome, in Nigerian women the prevalence of polycystic ovary morphology a self- administered questionnaire, was validated. J Clin Epidemiol. among women attending a new teaching hospital in Southern Nigeria. 2004;57:1279–1287. Afr J Reprod Heal. 2014;18:161–164. 13. Khomami MB, Tehrani FR, Hashemi S, Farahmand M, Azizi F. Of PCOS 26. Temmerman M, Khosla R, Laski Z, Say L, Women and Health Working symptoms, hirsutism has the most significant impact on the quality of Group for the Global Strategy for Women’s C and AH. Women’s life of Iranian women. PLoS ONE. 2015;10:1–10. health priorities and interventions. BMJ. 2015;351:h4147. 14. Panico A, Messina G, Lupoli GA, et al. Quality of life in overweight 27. World Health Organization. Summary Report. World Conference on (obese) and normal- weight women with polycystic ovary syndrome. Social Determinants of Health 19 -21 October. Rio de Janeiro, Brazil. Patient Prefer Adherence. 2017;11:423–429. 2011. 15. Fessler DMT, Natterson-Horowitz B, Azziz R. Evolutionary 28. Wise L, Palmer RJ, Stewart AA, Rosenberg L. Polycystic ovary syndrome determinants of polycystic ovary syndrome: Part 2. Fertil Steril. and risk of uterine leiomyomata. Fertil Steril. 2007;87:1108–1115. 2016;106:42–47. 29. Alhassan A, Ziblim AR, Muntaka S. A survey on depression among 16. Casarini L, Simoni M, Brigante G. Is polycystic ovary syndrome a sex- infertile women in Ghana. BMC Womens Health. 2014;14:42. ual conflict? A review. Reprod Biomed Online. 2016;32:350–361. 30. Dhont N, Van De Wijgert J, Coene G, Gasarabwe A, Temmerman M. 17. Casarini L, Brigante G. The polycystic ovary syndrome evolu- ‘Mama and papa nothing’: Living with infertility among an urban pop- tionary paradox: A genome- wide association studies- based, ulation in Kigali, Rwanda. Hum Reprod. 2011;26:623–629. in silico, evolutionary explanation. J Clin Endocrinol Metab. 31. Adanu R, Maya ET, Sarfo B, et al. The study design and methodol- 2014;99:E2412–E2420. ogy of the epidemiology & phenotype of PCOS in Ghana (Ghana-pep) 18. Knochenhauer ES, Key TJ, Waggoner W, Boots LR, Azziz R. Prevalence study. Presented at the 14th Annual Meeting of the Androgen Excess of the polycystic ovary syndrome in unselected black and white & PCOS Society, November 10–12. Lorne, Victoria, Australia. 2016.