ABSTRACT
[Objective]To evaluate the terminal hair growth of different body positions among in pregnant women ,analyze the contribution of each area to the diagnosis to hirsutism ,to improve the mFG scoring system.[Method]Pregnant women aged 20~41 years,with normal range of total testosterone levels and non-hirsute were recruited at their prenatal care in Sun Yat-sen Memorial Hospital,Sun Yat-sen University. They were followed up before pregnancy,at the 5th and the 9th week,the 10th and the 14th week, the 15th and the 20th week,the 21th and the 24th week of gestation. Then 72 more cases of pregnant women were recruited and followed up at 15~24th week. At each time of followed up,their total testosterone(TT)levels was examined by liquid chromatography tandem mass spectrometry(LC/MS-MS)and terminal hair growth were assessed by mFG score. Significant difference procedure least(LSD) analysis of variance was used to compare the levels of testosterone and mFG score in different gestational weeks. Receiver operating characteristics(ROC)analysis and logistics analysis were conducted to evaluate the contributory strength of hair growth in each body position for the diagnosis of hirsutism. The scores in the body area which made a significant contribution to the total were summed up and termed the simplified mFG score(sFG score). Following,the sFG scores were subjected to ROC analysis to determine the thresh-old that would maximize both the sensitivity and specificity of the measure to accurately distinguish hirsute from non-hirsute patients.[Results]Among the forty three pregnant women who were followed-up from before pregnancy to 24th week,the mean±SD for TT was (1.09 ± 0.59)nmol/L before pregnancy,and(1.13 ± 0.40),(1.28 ± 0.38),(1.83 ± 0.63),(1.82 ± 0.52)nmol/L for 5~ 9th,10~14th,15~20th week,and the mFG score was 1.65 ± 0.60,2.30 ± 0.45,3.60 ± 0.68,4.20 ± 0.41 and 4.40 ± 0.77,respectively. The order of the facial and body sites presented with new terminal hair growth was upper abdomen,lower abdomen,lower back,up-per lip,thighs,upper back,chest,upper arm,and chin,in sequence. After analyzing 115 cases(including the 72 cases recruited later),ROC analysis showed that the diagnostic value of different sites for hirsutism(mFG≥5):upper lip>lower back>thighs>lower abdomen>upper arm>upper back>chest/upper abdomen. Four sites among them ,namely upper lip ,lower back ,thighs and lower abdomen had the best diagnostic value,and the AUC for ROC were 0.779,0.728,0.675,and 0.626,respectively. Both ROC and logistic analysis indicated that he most significant body areas in defining hirsutism(defined as an mFG score≥5)were the upper lip, lower back,thighs,and lower abdomen. Using a cut-off value of 3,the combination of the four areas has the best sensitivity and specificity in distinguishing hirsute from non-hirsute women. [Conclusion]The study suggested that the mFG score increased as pregnancy progressed before the 24 weeks of gestation. The subset of upper lip,lower back,thighs and lower abdomen may be a reli-able simplification of mFG system for the evaluation of excess hair growth. The cut-off value was of≥3.
ABSTRACT
Objective To compare the difference of serum adiponectin levels between polycystic ovary syndrome (PCOS) patients and age, boby mass index (BMI) and insulin-resistance index matched controls, and explore its influence factors.Methods Case-control study, involving 97 women with PCOS and 116 age, BMI, fasting plasma glucose and insulin levels, homeostasis model assessment-insulin resistance index (HOMA-IR) matched controls.Hormone profiles, and serum adiponectin levels were measured and compared.Hormone profiles and serum adiponectin levels were compared among the four PCOS phenotypes.Multiple regression analysis was used to evaluate the factors affecting serum adiponectin levels.Results (1) Serum adiponectin level was significantly lower in PCOS group [(21 ± 16) mg/L] than controls [(25± 13) mg/L, P=0.038], and the same result in stratified analysis on weight height ratio (WHR, ≥0.8 and ＜0.8).(2) There was statistical differences in testosterone among different four PCOS phenotypes (P=0.001), there were no statistical differences in FSH, LH, WHR and serum adiponectin levels among four PCOS phenotypes (P＞0.05).(3) WHR and PCOS status were independent determinants of serum adiponectin levels (P＜0.05).Conclusions Low serum adiponectin levels in the women with PCOS is correlated with PCOS per se, independent of insulin resistance and obese.This fact supports the further study of the effect of adiponection in the pathophysiology of PCOS and its log-term impact.