|
|
Metformin Therapy in Low Birth Weight Pubertal Girls« Back to Volume 23, Issue 1, March 2007 - Table of Contents Lourdes and colleague studied 22 girls with birth weight <–1.5 SDS for gestational age (a level of prenatal growth restraint known to be associated with early menarche and reduced adult stature when puberty starts at 8-9 years of age), with a history of breast development at age 8-9 and less than 1 year before the study start, and a height at time of enrollment of at least 1 SD above mid-parental height SD for chronologic age. Other entry criteria included a bone-age at least 1 year above chronologic age and central and progressive puberty as judged by gonadotropic responses to gonadotropic releasing hormone (GnRH), and pubertal dimensions of the genitalia on ultrasound exam. These girls were studied in order to better understand whether or not hyperinsulinemic insulin resistance could underpin a rapid transit through puberty and a subsequent loss of adult stature. The authors stated that multiple factors are involved in the timing of the onset of puberty, and interrelated neuroregulation of the hypothalamo-gonadotropic-leptinergic axes are thought to play a major role. For example, obese girls with insulin resistance and hyperleptinemia have an earlier onset and more rapid progression of puberty. Age of onset of puberty is inversely related to the rate of progression through puberty, just as less pubertal height gain is usually compensated by more pubertal growth. However, in girls with low birth weight followed by rapid catch-up growth in infancy, these mechanisms may fail to work, and thus these girls enter puberty at an early age (8-9 years) and are at risk for early menarche, as well as growth arrest and, therefore, adult short stature. In this open-label study, the 22 girls were randomized to remain untreated or to receive metformin (850 mg once a day) for a total of 36 months. Pubertal growth, body composition, uterine-ovarian size, and endocrine metabolic markers were assessed at 6 monthly intervals. All therapy was stopped in the treated group at 36 months, and an additional assessment was performed at 42 months after study start. Birth weight data was transformed into SDS for gestational age. Target height was defined as mid-parental height adjusted for female gender, and BMI was calculated for each girl and transformed into SDS for chronologic age. Ultrasounds were performed of the ovaries to calculate ovarian volume. Body composition was performed with dual energy x-ray absorptiometry. Assessments were made every 6 months for fasting serum insulin, sex hormone binding globulin, insulin-like growth factor (IGF)-I, HDL and LDL cholesterol, triglycerides. as well as renal and hepatic function. Serum leptin and IGF binding-protein (IGFBP)-1 were determined at 0 and 24 months. Fasting insulin resistance was estimated with fasting insulin and glucose levels by using HOMA. Results were expressed as mean + SEM. Baseline characteristics were similar in the untreated and metformin-treated girls. While not being obese (mean BMI <+2 SD) both groups had a high fat fraction for a given BMI and had high serum leptin concentration, on average >2-fold higher than the reference. Metformin treatment was associated with lower serum insulin, leptin, and IGF-I as well higher SHBG and IGFBP-3 levels; a less arthrogenic lipid profile and a leaner body composition. No differences were noted in bone mineral density or growth of the internal genitalia between the groups. Untreated and metformin-treated girls had a similar mean age at the start (8.7 vs 8.6 years) but menarche was significantly earlier in the girls who were untreated (2 years vs 3 years, P<0.01). Thus, metformin was also accompanied by a prolonged pubertal height gain (19.5 cm + 1.0 vs 16.0 cm + 1.0, P<0.05). In addition, metformin treatment was shown to increase the near adult height relative to the target height as inferred from the average height at 42 months. The final effect of metformin treatment might well be greater than that reported, since all the untreated girls had nearly stopped growing (height velocity <2 cm/year), whereas 4 of the 10 metformin-treated girls were still growing at a velocity >2 cm/year at 42 months. The authors stated that this prospective study shows that low birth weight girls with an early onset of breast development tend to pass through puberty rapidly and end up with an adult stature below their target height. Those treated with metformin for 36 months had prolongation of their pubertal growth, as well as a leaner body composition and lower levels of lipids. They speculate that the mechanism whereby metformin affects pubertal progression and growth may be related to reduction in body fat and less leptonemia, which could subsequently attenuate the activity of the hypothalamic GnRH pulse generator, enhance the gonadal feedback suppression on LH secretion, and decrease aromatase activity in the ovarian granulosa cells. Another possible mechanism would be the reduction in hyperinsulinism and concomitant rise in serum SHBG and IGFBP-1, thus altering unbound estrogen and IGF-I levels. Whether or not metformin could normalize the endocrine metabolic state, pubertal growth, and menarchal timing in other conditions of hyperinsulinemia and hyperleptinemia is unknown. Editor’s CommentThis study marks a very important contribution to the body of information regarding growth and development of low birth weight children and the relationship to insulin resistance of their final adult stature. It is interesting to note, as did the authors, that the girls in the study had early-normal, not precocious, puberty and thus were not candidates for GnRH agonist therapy. In addition, their stature for chronologic age was not at such a level for which growth hormone therapy would be considered. Yet, these girls clearly have a significant abnormality in the timing and rate of progression of puberty which causes them to fail to achieve their target heights. Metformin is clearly of benefit to these girls. Whether prolonged use of metformin would have been an even greater benefit remains to be shown. The girls in the current study had birth weight less than <–1.5 SDS for gestational age, putting at least some of them in a category that would not necessarily make them be considered small for gestational age and thus at risk for the numerous metabolic phenomena associated with that syndrome. The spectrum of size at birth and subsequent health and well-being remains to be elucidated. However, this study contributes to a better understanding of these complex processes. William L. Clarke, MD « Back to Volume 23, Issue 1, March 2007 - Table of Contents
|
Volume 24 Number 1
May 2008
www.GGHjournal.com
ISSN 1932-9032
