Reproducibility and refinement of urinary LH in the screening of progressive puberty in girls

Abstract

Abstract Context First-voided urinary LH (FVU-LH) has been suggested as an alternative to gonadotropin-releasing hormone (GnRH) stimulation test for detection of precocious puberty. Objective To evaluate the reproducibility of FVU-LH, its correlation with basal and GnRH-stimulated gonadotropins, and its diagnostic value for differentiating progressive from non-progressive puberty. Design and participants Clinical and endocrine data were obtained from the medical records of 95 girls with suspected progressive puberty who underwent two consecutive FVU-LH tests. In 55 of these participants, GnRH stimulation test was performed close to the FVU-LH test. The reported cutoff levels of 5 IU/L and 1.16 IU/L for GnRH-stimulated LH and FVU-LH, respectively, were used as markers of progressive puberty, clinically defined as bone age advancement of >1 year and/or growth velocity SDS>2, in addition to thelarche. Results The two consecutive measurements of FVU-LH were highly correlated (r= 0.830; p<0.001). The higher of the two results was better correlated with basal gonadotropins and GnRH-stimulated LH. Furthermore, it aligned better with the clinical outcome of girls with early thelarche, which supports the approach of double tests of FVU-LH to distinguish progressive from non-progressive puberty. By comparison to GnRH-stimulated LH, the higher FVU-LH value had better sensitivity (68%), whereas peak LH had better specificity (91%) for the diagnosis of progressive puberty. Both tests had high positive predictive value and poor negative predictive value. Conclusions The higher value of paired FVU-LH tests can be used to screen girls with suspected progressive puberty and can reduce the need for GnRH stimulation test.