A randomized clinical trial demonstrating cell type specific effects of hyperlipidemia and hyperinsulinemia on pituitary function

Abstract

Introduction: Obesity is characterized by elevated lipids, insulin resistance and relative hypogonadotropic hypogonadism, reducing fertility and increasing risk of pregnancy complications and birth defects. We termed this phenotype 'Reprometabolic Syndrome' and showed that it can be recapitulated by acute infusions of lipid/insulin into healthy, normal weight, eumenorrheic women. Herein, we examined the broader impact of hyperlipidemia and euglycemic hyperinsulinemia on anterior pituitary trophic hormones and their targets.

Methods: Serum FSH, LH, TSH, growth hormone (GH), prolactin (PRL), thyroid hormones (free T4, total T3), cortisol, IGF-1, adiponectin, leptin and creatinine were measured in a secondary analysis of an interventional crossover study of 12 normal weight cycling women who underwent saline and heparin (control) infusion, or a euglycemic insulin infusion with heparin and Intralipid® (lipid/insulin), between days 2-5 in sequential menstrual cycles.

Results: In contrast to the decrease in gonadotropins, FSH and LH, infusion of lipid/insulin had no significant effects on other trophic hormones; TSH, PRL or GH. Thyroid hormones (fT4 and total T3), cortisol, IGF-1, adiponectin and creatinine also did not differ between saline or lipid/insulin infusion conditions. Leptin increased in response to lipid/insulin (p<0.02).

Conclusion: Acute hyperlipidemia and hyperinsulinemia exerted differential, cell type specific effects on the hypothalamic-pituitary-gonadal, adrenal and thyroid axes. Elucidation of mechanisms underlying the selective modulation of pituitary trophic hormones, in response to changes in diet and metabolism, may facilitate therapeutic intervention in obesity-related neuroendocrine and reproductive dysfunction.

Fig 1. Flow diagram showing process for participant screening, enrollment and randomization for study visit infusions.

The total participants analyzed was 12.

Fig 2. Basal and stimulated gonadotropin levels are decreased in response to acute insulin/lipid infusion.

FSH (A) and LH (B) were measured at 10-minute intervals as described [7, 8]. GnRH (75ng/Kg) was administered at 240 min. Values are means ± SEM; n = 12.

Fig 3. Creatinine levels are unaffected by lipid/insulin, confirming there is no hemodilution occurring.

Pooled levels of serum Creatinine were measured, as described in methods, approximately every hour; values are mean ± SEM; n = 12.

Fig 4. Lipid/insulin infusions may increase relative serum TSH levels but have no effect on thyroid hormones.

Hormones were measured as described in methods. (A) Pooled levels of serum TSH were measured every 30 minutes; values are mean ± SEM (n = 11). Pooled levels of (B) serum FT4 and (C) T3 measured at approximately 0, 30, 160, and 360-minutes; values are mean ± SEM (n = 11).

Fig 5. Prolactin and cortisol are unaffected by lipid/insulin infusions.

Pooled levels of (A) serum prolactin and (B) cortisol (n = 12) were measured at approximately 0, 60, 160, and 360-minutes, as described in methods; values are mean ± SEM.

Fig 6. Effect of lipid/insulin infusion on GH and IGF-1.

Pooled steady state levels of (A) serum GH and (B) serum IGF-1 were measured as described in methods; boxes encompass 25^th -75^th percentiles, whiskers are maximum and minimum values, and bar indicates median. Spaghetti plots indicate paired data points for each participant for the saline and lipid/insulin infusions. (n = 12).

Fig 7. Lipid/insulin infusion increases leptin and has no effect on adiponectin.

(A) Pooled levels of serum leptin and (B) adiponectin were measured as described in methods, during saline and lipid/insulin infusion visits; boxes encompass 25^th -75^th percentiles, whiskers are maximum and minimum values, and bar indicates median. Spaghetti plots indicate paired data points for each participant for the saline and lipid/insulin infusions. (n = 12).