STRUCTURAL EQUATION MODELLING ANOVUTORY INFERTILITY AMONG WOMEN WITH POLYCYTIC OVARIAN SYMDROME IN EDO STATE, NIGERIA.

Abstract

Infertility is the biological inability of an individual to contribute to conception, or to a female who cannot carry a pregnancy to full term. Infertility is one of the toughest challenges a couple can face which could lead to an irreparable damage. Many researches have been done on infertility among women in Nigeria, but none has explored structural equation model of anovulatory infertility among women with polycystic ovary syndrome in Nigeria. This study aimed to develop a structural equation model of the relationships between anovulatory infertility and some characteristics in women with polycystic ovarian disease. A total of 366 reproductive women aged 18-44 years responded to the questionnaire at the University of Benin Teaching Hospital (UBTH) and the Women's Health and Action Research Centre (WHARC), Benin, Edo State. Confirmatory factor analysis was performed on the hypothesized factors and structural equation modelling was fitted using AMOS software to know the relationship between women with anovulatory infertility and characteristics m women with polycystic ovarian syndrome. Analyses was perfom1ed at 5% significance level. The minimum sample discrepancy (CMIN) is given as 722.418 and after dividing it by the degrees of freedom, CMIN/DF which is given as 15.05. The RMSEA is 0.205 while the values of the GFI and AGFI are close to one. Furthermore, there exist a relationship between Estradiol and fasting glucose (r = 0.186, p = 0.001), follicle stimulating hormone and total testosterone (r = -0.172,p = 0.002), Progesterone and total testosterone (r = -0.182, p = 0.009), Estradiol and total testosterone (r = 0.333, p =< 0.001 ), Estradiol and fasting glucose (r = 0.186, p = 0.001), and some other hormonal, metabolic and bio-physical variables. Our results showed relationships between Hormonal, Metabolic and Bio-physical variable with polycystic ovarian syndrome. The SEM model had some fit indices significant but a non-significant chi square fit index which may be due to our sample size.