The prevalence of PCOS at An-Najah National University in age groups 18–24 years was found to be 7.3%. PCOS prevalence depends on the recruitment process of the study population and criteria used for its definition; using NIH criteria, two Iranian studies found PCOS prevalence to be 7.1% and 7%. The first was done in 2011 on a community sample of age group 18–45 years , and the second one was done in 2009 in Isfahan among females referred to the mandatory pre-marriage screening clinic of age 17–34 years . Shared common ethnic and sociodemographic factors between Iranian and Palestinian women might explains this similarity in PCOS prevalence in addition to using the same criteria of diagnosis in the study despite the differences in age group and sampling methods. Using same criteria for diagnosis in other parts of the world found a close range of PCOS prevalence; examples are Australia 8.7%, Spain, 6.5%, Greek Island of Lesbos 6.7%, the southeastern United States 4%, and Sweden 4.8% [5–9]. Clinical hirsutism in this study was found in 27% of participants (Table 2), about 70% of whom had idiopathic hirsutism; these women cannot be entirely excluded from the diagnosis of PCOS, because they may have been oligoovulatory, despite their reported regular episodes of vaginal bleeding.
It is expected that we will have double the prevalence of PCOS if we use Rotterdam criteria for diagnosis ; although this is a newer criteria for diagnosis, many authors argue that its use may overestimate the prevalence of PCOS. Disagreement has persisted regarding whether the two additional phenotypes (oligo-ovulation with PCO or hyperandrogenism with PCOS) in Rotterdam criteria actually do have classical PCOS, because PCO on ultrasound is a very common finding in normal population . Also in some studies they found that using NIH criteria predicts the metabolic risk more appropriately than Rotterdam criteria .
When we explored acne as possible risk factors for PCOS , 80%. of PCOS group were found to have acne, and having acne was found to increase risk to develop PCOS by eight times (Tables 3 and 4). This is consistent with findings of Turkish and American PCOS studies where women with PCOS had acne at a rate of 53% [23, 24].
Although we found frontal baldness in 20% of PCOS group, this was not statistically significant after adjustment for confounders (Table 4). Frontal baldness is usually found in disorders associated with higher levels of androgen such as androgen secreting tumors . This is supported by our finding of all PCOS group to have normal FT serum level (Table 5).
Smoking, history of hirsutism and history of irregular menses in mothers were found to be insignificantly related to the presence of PCOS (Tables 3 and 4). These finding are shared by some studies  but not by others ; differences in population characteristics, methodology and sample size might explain these finding differences. Of those who had menstrual abnormalities there was no statistical biochemical difference between those who met the definition of PCOS or not (Table 5). Although none of PCOS cases in our study had high serum FT , clinical hirsutism was significant in all cases using mF-G score (mean11.8). This is in agreement with Haung etal  who found that supra-normal levels of FT were present in 57.6% of PCOS patients diagnosed by NIH criteria. The main strength of this study is that being the first PCOS prevalence study in Palestine, and it touches very important women health issue. With the international criteria of diagnosis used in this study, very good results about prevalence and risk factors were drawn that further studies can be built on.