Reference limits of five major reproductive hormones for some Ghanaian women of child bearing age
The need to provide a set of reference limits applicable to a laboratory arises when a new method is introduced or the values being used do not define the population, due to socioeconomic, genetic, cultural and environmental differences. This study aimed at establishing the laboratories own set of reference limits for the homones LH, FSH, PRL, E2, and Progesterone, to aid in diagnosis and treatment of reproductive endocrinopathies, and also, serve as the basis for epidemiological studies. From the two medical schools in the country, 85 randomly selected healthy females aged 17 to 45 years, (X=28.9± 6.8yr), with cycle length of 26 to 32 days (X=28.35±l .4dy), were choosen in the ratio of 1:2:6 from upper:middle:lower income earning groups. The response rate of the study was good (82%). Blood samples were taken on calculated days during the FP, MC and LP of their menstrual cycles. Serum levels of the hormones LH, FSH,E2 and progesterone were determined on the samples obtained from each phase of the cycle, while PRL levels were estimated on only the FP samples, using ELISA commercial kits from Biomerieux (France). The intra-assay precision for E2=2.3%, PRL=2.3%, LH=2.4%, P2.1% and FSH2.4%. The inter-assay precision for LFI=5.1%, FSH=5.4% PRL=5.0%, E2=6.0% and P9.0%. All the precisions fell within the recommended limits for the provision of Clinical Biochemical services for hormonal assays. Using the non-parametric binomial probability method, and the parametric method, 95% reference limits accomanied by 70% and 90% CI respectively, were established for all the hormones, during the various phases of the cycle. The percentiles obtained from both methods were then compared for statistical difference at p<0.00. Where the assumption of a Gaussian distribution held good, there was no difference between the reference limits obtained by either method. The inadequate sample size and the wide CI around the non-parametric percentiles reduced their precision and clinical usefulness. For this reason, the parametric estimates were being recommended for use. In comparison of the established reference limits for 95% of the population, to that of literature values established on Caucasians, it was realized that the average Ghanaian of child bearing age had relatively higher basal hormone levels than her Caucasian counterpart. The recommended 95% reference limits and 90% CI are: At P<0.05, weak correlations (all less than 0.5) existed in trends consistent with literature between age and PRL (r = -0.25), age and FP E2 (r = -0.22), PRL and MC LH (r = -0.25), FP E2 and MC E2 (r = 0.23), MC E2 and LP E2 (r =0.34). The established PRL limits were used as a tool for health screening for hyper-PRL among refferals to the laboratory in 22 months. A prevalence of 9.4% was found. Of 673 requests for PRL, 20.36% were hyper-PRL. Amenorrhoea was the most common presenting symptom (29.27%) with 19.8% having hyper-PRL, and they formed the majority of the hyper-PRL cases (28.47%). Galactorrhea which is used as a prominent sign of hyper-PRL was the least presented symptom (16.79%) among the referrals with requests for PRL determinations, and 23.9% of them had hyper-PRL. It was also found to be the least prevalent symptom in the hyper-PRL population of the study (19.77%). It can however not be conclusively said that the elevated PRL levels are all pathological, nor are an ovulation or defects in folliculogenesis the sole causes of the above symptoms. Comparing the various prevalence’s with literature, it was realized that hyper-PRL may not be a prominent causative agent of infertility in the population. Generally, the results of this study are unique to the University of Ghana medical school endocrinology laboratory, and will be appropriate to the Ghanaian woman of child bearing age who gets access to the laboratory.
A thesis submitted to the Board of Postgraduate Studies, Kwame Nkrumah University of Science and Technology, Kumasi, in partial fulfilment of the requirements for the award of the Degree of Master of Science in Chemical Pathology, 1993