Background: Hyperprolactinemia is associated with obesity. Furthermore, in human adipose tissue cultured in vitro, prolactin (PRL) inhibited lipoprotein lipase (LPL) activity via functional PRL receptors.
Objective: Tostudy PRL and insulin ultradian rhythm and subcutaneous adipose tissue LPL mRNA and protein expressions in severely obese women before and after malabsorptive bariatric surgery.
Methods and Procedures: Seven severely obese, fertile women were studied twice, once before and the second time 1 year after bilio-pancreatic diversion (BPD), when the weight was stable for at least 3 months. Metabolizable energy intake and 24-h energy expenditure (EE) were measured. Fourier and PULSEFIT analyses were applied to 24-h hormonal time-series to study daily fluctuations and hormonal clearance. Insulin sensitivity was assessed by euglycemic-hyperinsulinemic clamp. Quantitative-competitive reverse transcriptase-PCR and western blot analysis were used to measure LPL gene expression.
Results: Spontaneous 24-h PRL secretion was significantly reduced after BPD (mean-daily release, 128.4 ± 28.1 μg/l vs. 67.2 ± 9.2 μg/l distribution volume (Vd/l·24 h), P = 0.02); insulin secretion also was significantly reduced (499.9 ± 204.0 μg/Vd/l·24 h vs. 85.6 ± 21.0 μg/Vd/l·24 h, P = 0.0001). Metabolizable energy/kgFFM did not change significantly after BPD. Twenty-four-hour EE, but not 24-h EE/FFM, was significantly decreased after BPD (P < 0.05). Insulin sensitivity significantly (P < 0.0001) increased after BPD from 21.41 ± 1.92 to 68.62 ± 5.03 μmol/kgFFM/min. LPL mRNA concentration (from 42.63 ± 4.21% to 19.00 ± 2.74% of cyclophilin mRNA, P = 0.001) as well as LPL protein level (from 8.94 ± 2.73 to 3.16 ± 1.05 as ratios of protein of interest vs. housekeeping protein, P = 0.038) significantly decreased after BPD. The major determinant of PRL secretion was insulin secretion, whereas the best predictors of LPL expression were insulin and PRL secretion rates.
Discussion: The restriction of lipid metabolizable energy rather than weight loss seems to be responsible for both reduction in PRL circulating levels and normalization of its secretion rhythm after bariatric surgery. Furthermore, the reduced adipose tissue LPL expression, being significantly correlated with the decrease in insulin and PRL, suggests a role of hyperinsulinemia and hyperprolactinemia in inducing and sustaining obesity.
Weight gain and elevated body weight are frequently associated with prolactinomas (1,2), and weight loss is recorded in 70% of all prolactinoma patients and, in particular, in 90% of male patients after normalization of prolactin (PRL) levels (1,3). PRL seems to play a relevant role in inducing weight gain during antipsycothic drug therapy, particularly in men (4,5,6). Hyperprolactinemia, induced by ectopic pituitary grafting, is capable of stimulating food intake and white fat deposition in female rats (7).
On the other hand, these features seem to be in contrast with a series of in vitro and in vivo studies, although most of them were performed during pregnancy and lactation.
Hyperprolactinemia, in fact, was found to reduce adipose tissue mass in mice (8,9). In human adipose tissue cultured in vitro, PRL inhibited lipoprotein lipase (LPL) activity (10) via functional PRL receptors. PRL-receptors knockout mice (11) show impaired development of both internal and subcutaneous adipose tissue as a consequence of the decreased number of adipocytes.
It is important to point out that the role of PRL during lactation is much different than that in pathological conditions such as obesity. In fact, during lactation PRL probably acts by diverting fuel substrates from the adipose tissue to the mammary gland. On this regard, Flint and Vernon (12) have demonstrated that, during short-term food restriction, the lactating rat is able to mobilize significant amounts of lipids stored in the adipose tissue in order to preserve the total output of triglycerides in the milk. In contrast, the role of hyperprolactinemia in other physiological and pathological conditions, such as obesity, is poorly understood.
The first evidence of an ultradian rhythm of PRL, showing a nocturnal rise occurring shortly after the onset of sleep, was made by Sassin et al. (13). Later, Veldhuis and Johnson (14) showed that PRL release in normal young men is characterized by significant circadian and ultradian periodicities and by a close temporal coupling with luteinizing hormone.
Van Cauter et al. (15) demonstrated in healthy men a bimodal profile of PRL secretion with a daytime as well as a nocturnal phase of increased PRL release, the diurnal phase being ∼50% of the magnitude of the nocturnal one. A shift of the normal nyctohemeral rhythm of PRL was observed in obese patients, but the normal pattern was restored after 12 days of fasting (16). Recently, Kok et al. (17) described a significant reduction in PRL secretion rate in obese women after the loss of 50% of the excess weight without reaching the levels found in lean, healthy subjects.
However, at least to our knowledge, no data are reported in the literature about the pulsatility of PRL nor any relationship between PRL secretion and adipose tissue mass or LPL expression in adipose tissue after weight loss in severely obese women. We hypothesize that a reduction in circulating levels of PRL together with the normalization of its pulsatility might play a role in weight loss by decreasing LPL expression in the adipocytes.
To this end, the ultradian rhythm of PRL was examined and LPL mRNA and protein expression were measured in the subcutaneous adipose tissue of 7 severely obese, fertile women before and 1 year after bariatric surgery.
Methods and Procedures
Seven fertile (every subject had at least 1 pregnancy with full-term fetuses) severely obese women (age 38 ± 5 years; BMI 52.8 ± 3.0 kg/m2; fat-free mass (FFM) 71.5 ± 7.9 kg and fat mass 70.1 ± 3.7 kg; waist 143.0 ± 6.6 cm) were included in the study. None had diabetes mellitus or any other endocrine or non-endocrine disease. The subjects were studied twice, once before and the second time 1 year after bilio-pancreatic diversion (BPD), when the weight was stable for at least 3 months (BMI 39.0 ± 2.2 kg/m2, P < 0.0001; FFM 59.5 ± 5.0 kg, P < 0.02; fat mass 44.4 ± 2.4 kg, P < 0.0001; waist 103.7 ± 5.6 cm, P < 0.001). At the time of the baseline study, all subjects were on an ad lib diet.
The subjects were studied on two separate days: one for the assessment of PRL rhythm during a standardized 24-h period, and one for performing the adipose tissue biopsy.
The study protocol was approved by the Institutional Review Board; the nature and purpose of the study were carefully explained to all subjects before they provided their written agreement to participate.
This essentially malabsorptive surgical procedure is reported in ref. 18.
During the two study sections (before and after BPD), body weight was measured to the nearest 0.1 kg using a beam scale, and height to the nearest 0.5 cm using a stadiometer (Holatin, Crosswell, Wales, UK).
Total body water measurement
Total body water was determined using 0.19 Bq of tritiated water in 5 ml of saline solution administered as an intravenous bolus injection (19). Blood samples were drawn before and 3 h after the injection. Radioactivity was determined in duplicate on 0.5 ml of plasma using a Beta-scintillation counter (model 1600TR; Canberra-Packard, Meriden, CT). Corrections were made (5%) for nonaqueous hydrogen exchange (20); water density at body temperature was assumed to be 0.99371 kg/l. Total body water (kg) was computed as 3H2O dilution space (l) × 0.95 × 0.99371. The within-subject coefficient of variation (CV) for this method was 1.5% (ref. 21). FFM in kilograms was obtained by dividing the total body water by 0.732 (ref. 22).
Euglycemic-hyperinsulinemic clamp procedure
Peripheral insulin sensitivity was evaluated by the euglycemic-hyperinsulinemic clamp procedure (23). After inserting a cannula in a dorsal hand vein for sampling arterialized venous blood, and another in the antecubital fossa of the contralateral arm for infusions, the subjects rested in the supine position for at least 1 h. They were placed with one hand warmed in a heated air box set at 60 °C to obtain arterialized blood samples. Whole-body glucose uptake ((M value) in μmol/kgFFM/min) was determined during a primed constant infusion of insulin (at the rate of 6 pmol/kg/min). The fasting plasma glucose concentration was maintained throughout the insulin infusion by means of a variable glucose infusion and blood glucose determinations every 5 min. Whole-body peripheral glucose utilization was calculated during the last 40-min period of the steady-state insulin infusion.
24-h Energy substrate output
Twenty-four-hour stool collections were carried out during the day spent in the calorimetric chamber.
Stool aliquots were homogenized and analyzed for nitrogen, carbohydrates, and lipid content using a Fenyr analyzer (PerCon Prüfgeräte, Hamburg, Germany). The s.d. of repeated (n = 3) measurements of 24-h fat loss in BPD patients was found to be <0.5 g/d.
Energy intake and composition were assessed by weighed food records kept over three 3-day periods (the 2 of the 3 days included the weekends), evenly distributed over 3 months preceding the study. The subjects were asked to record everything they ate or drank for a total of 9 days. For each recording period the dietitian had two individual meetings with the subject: the first to provide instructions and the second to check the food records. All the subjects were followed by the same dietitian.
The nutrient content of all food items was calculated using computerized tables (Food Processor II; Hesha Research, Salem, OR, modified according to the food tables of the Istituto Nazionale di Nutrizione, Rome, Italy). The energy content of the food was computed as follows: 4.3 kcal/g for protein, 4.2 kcal/g for starch (or starch equivalent), and 9.3 kcal/g for fat.
Metabolizable energy evaluation
Macronutrient intake was computed by 7-day diary recall method. The metabolizable energy intake was defined as the gross energy intake minus fecal losses.
Adipose tissue biopsies
Fat specimens were obtained using needle biopsy after local anesthesia with 1% lidocaine of the subcutaneous adipose tissue covering the vastus lateralis muscle. The specimens were immediately placed in liquid nitrogen and stored at −80 °C until analysis.
The subjects spent a day (starting at 8 am) in the respiratory chamber of the Metabolism Unit of the Catholic University School of Medicine in Rome. The characteristics of the device have been described previously (24). 24-h energy expenditure (EE) and nonprotein (np) Respiratory Quotients (RQ) were computed. npRQ was calculated based on measured 24-h nitrogen excretion (urinary urea + uric acid + creatinine + proteins).
During the study day, all subjects were assigned a diet with an energy content reflecting the diet habitually consumed by the subjects. The food given and returned was weighed to the nearest gram on precision scales (KS-01; Rowenta, Berlin, Germany). The nutrient content of all food items was calculated as detailed above.
At 4 pm, the subjects performed a physical exercise session on the motorized treadmill, walking for 30 min at a constant speed of 3 km/h up a 10% gradient.
Hourly blood samples were drawn from a central venous catheter brought outside the chamber through long plastic tubing for the measurement of hormone concentrations.
Assay of LPL mRNA levels
RNA extraction. White subcutaneous adipose tissue was weighed and immediately frozen in liquid nitrogen and stored at −80 °C. Total RNA was extracted from 50 mg of tissue using RNAgents Total RNA Isolation System (Promega, Southampton, UK). Tissue samples were homogenized in a guanidine-thiocyanate denaturing solution, extracted with phenol:chloroform:isoamyl alcohol and precipitated overnight in isopropanol. RNA pellets were washed in ethanol and resuspended in nuclease-free water. RNA concentrations were quantified spectrophotometrically at 260 nm using the GeneQuant pro RNA/DNA Calculator (Amersham Pharmacia Biotech, Cambridge, UK).
Quantitative competitive reverse transcriptase-PCR. Concentrations of mRNA of LPL were quantified by reverse transcriptase real-time PCR using a LightCycler (LightCycler II; Roche Diagnostics, Basel, Switzerland). This was performed using the multi-specific internal standard (25), kindly donated by Professor Hubert Vidal. First-strand cDNA synthesis was performed at 48 °C for 45 min and ended at 94 °C for 2 min. The cDNA products were amplified in the presence of Tfl DNA polymerase, annealing of primers at 60 °C for 1 min and extension at 68 °C for 2 min. Samples were subject to 27–30 cycles in a thermal cycler (Bio-Rad, Hercules, CA). In addition, cyclophilin mRNA was amplified in parallel tubes as an internal standard or “housekeeping” gene, using cyclophilin primers and under identical conditions. A Cy-5 5′-end-labeled sense primer was used in the cPCR to generate fluorescent PCR products that were analyzed with an automated fluorescence DNA sequencer (ALFexpress; Pharmacia, Uppsala, Sweden) in 4% denaturing polyacrylamide gels. The initial concentration of the target mRNA was determined at the competition equivalent point, as described in ref. 25. Cyclophilin mRNA was also measured using reverse transcriptase-cPCR, as a reference gene, and the levels of LPL mRNAs were expressed as percentage ratio referred to the expression of cyclophilin.
Western blot analysis
Biopsies were homogenized while still frozen using a Polytron Homogenizer (Brinkmann, Westbury, NY) in buffer consisting of 20 mmol/l Tris-HCl, 5 mmol/l EDTA, 10 mmol/l sodium pyrophosphate, 100 mmol/l sodium fluoride, 2 mmol/l sodium orthovanadate, 1 mmol/l phenylmethylsulfonyl fluoride, % Triton-X100, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 3 mmol/l benzamidine, on ice. The absolute protein concentration was assayed according to the Lowry protein assay using a Bio-Rad (Hercules, CA) kit. Approximately 50 μg of proteins were resolved on 10% sodium dodecyl sulfate-polyacrylamide gels and transferred to nitrocellulose membranes, as previously described (26). The membrane was subsequently probed overnight at 4 °C with a mouse monoclonal anti-LPL antibody raised against the full-length human protein, and developed with an anti-mouse IgG-horseradish peroxidase conjugate and enhanced chemiluminescence (ECL; Santa Cruz, CA). The membranes were stripped and re-probed with a rabbit polyclonal antibody against glyceraldehyde-3-phosphate dehydrogenase, as a loading control. Bands were quantified using a scanning densitometer (Bio-Rad) and expressed as ratios of protein of interest vs. housekeeping protein.
All reagents for electrophoresis were obtained from Bio-Rad (Richmond, CA). All other chemicals were purchased from Sigma (St. Louis, MO). The antibodies against LPL and glyceraldehyde-3-phosphate dehydrogenase were acquired from Abcam (Cambridge, MA). Horseradish peroxidase-linked goat anti-mouse antibody was obtained from Amersham (Piscataway, NJ).
Plasma insulin: microparticle enzyme immunoassay (Abbott, Pasadena, CA); sensitivity 1 μU/ml and an intra-assay CV 6.6%.
Plasma PRL: direct, two-site immunoradiometric assay without extraction using materials supplied by ICN Diagnostics (Costa Mesa, CA). Antiserum cross reactivity <0.01% for human chorionic gonadotropin, thyrotropic hormone, luteinizing hormone, and follicle-stimulating hormone. Intra- and inter-assay CVs were 4.2 and 4.5%, respectively. Minimal detectable concentration 2.5 ng/ml, equivalent to 108.8 pmol/l.
Gonadotrophin FSH (standard: second WHO-IRP 94/632) levels were measured using reagents obtained from Ciba-Corning (Medfield, MA) by chemiluminescent methods.
Sex hormone-binding globulin (SHBG) was measured using an immunoradiometric assay using plastic beads coated with monoclonal antibody against human SHBG and an I125-labeled soluble antibody against human SHBG. The assay sensitivity was 0.1 μg/dl. Intra- and inter-batch percentage CVs were 5.5 and 15.6, respectively.
Estradiol was measured using radioimmunoassay after organic extraction with hexane:ethyl acetate and LH20 column chromatography. Minimum reportable level was 0.5 ng/dl; intra- and inter-batch percentage CV was 7.2 and 16.3, respectively.
Ultradian variability analysis of hormones
Fourier analysis (27) was applied to the 24-h hormonal time-series to study fluctuations on selected time scales. Relative serum levels of PRL and insulin for each subject were first calculated using the formula:
The averaged patterns, before and after treatment, of these relative profiles were approximated using the finite Fourier series:
Each filtered data set was rescaled so that the 24-h average was set equal to 100% and data at each time point were defined as a percentage of the 24-h average. The fitting goodness was determined by means of the degree of freedom-adjusted coefficient of determination: A value of this latter parameter equal to 1 indicates a perfect agreement between the experimental data and the fitted curve.
Group averages for all seven subjects were then calculated for each of the two hormones. The mean 24-h level of plasma concentration and the total daily secretion, as area under the curve calculated by the trapezoidal rule, were computed for each patient.
MATLAB software-R2006b (MathWorks, Natick, MA) was used to carry out the calculations.
Hormonal clearance kinetics
PULSEFIT, a computerized algorithm (28), was used to measure the hormone clearance kinetics, i.e., the rate at which a secreted hormone decays in the circulation. At this regard, PULSEFIT assumes a first-order, linear kinetics, i.e., a decay according to a single negative exponential-time curve. The dimensional value of this parameter is the inverse of a time, in this case min−1.
Data are reported as mean ± s.d., unless otherwise specified. Wilcoxon test with Bonferroni-correction was performed to compare paired-data sets. Distribution of residuals was assessed, testing for normality and checking the linearity assumptions in the model by standard scatter plots. Linear correlations and multiple regression analyses were performed. Data analyses were performed using SPSS statistical software (SPSS, Chicago, IL). Two-sided P < 0.05 was considered significant.
Sex hormone changes
The rising levels of SHBG, follicle-stimulating hormone, and estradiol suggested an improved fertility status. After BPD, in fact, follicle-stimulating hormone increased from 4.94 ± 1.43 to 6.85 ± 1.08 mIU/ml (P < 0.05), estradiol from 60.7 ± 21.7 to 91.6 ± 33.2 pg/ml (P < 0.01) and SHBG from 23.4 ± 7.5 to 31.5 ± 11.2 nmol/l, P < 0.01.
Energy balance and EE
Energy intake did not differ significantly before and after BPD (Table 1); however, after bariatric operation the metabolizable energy became 60.86 ± 6.81% of energy intake, mainly as a consequence of the fecal fat loss.
Table 1. Energy intake, metabolizable energy, and energy expenditure
Notably, the metabolizable energy normalized by kgFFM was not statistically different before and after BPD.
As shown in Table 2, 24-h EE was significantly decreased after BPD (P < 0.05). However, when 24-h EE was normalized by FFM, the mean value after BPD did not differ significantly from that before BPD (see Table 2).
Table 2. PRL and insulin concentration parameters and secretion rates
npRQ was significantly higher after BPD during the daytime (0.88 ± 0.02 vs. 0.78 ± 0.05, P < 0.01) than during the night time (0.76 ± 0.03 vs. 0.72 ± 0.02, P < 0.05).
Insulin-mediated whole-body glucose uptake significantly (P < 0.0001) increased after BPD from 21.41 ± 1.92 μmol/kgFFM/min to 68.62 ± 5.03 μmol/kgFFM/min. No significant difference was observed in insulin plasma levels during the last 40 min of the clamp (506.57 ± 13.60 pmol/l before BPD vs. 460.28 ± 12.05 pmol/l after BPD).
Average PRL 24-h profiles before and after BPD are illustrated in Figure 1a. The best fit pattern reveals a nocturnal acrophase (midnight) of PRL secretion, which was shifted toward early morning (4 am) after BPD. In both clinical conditions, the fitting gave an and the same principal angular frequency, ω ∼ 0.227(rad/cycle) ∼ 27(hour/cycle). Moreover, the spectral amplitudes of the periodical components were unchanged after treatment.
Mean 24-h PRL concentration, mean peak-amplitude, peak-area, and inter-peak valley (nadir) were significantly lower, whereas peak-frequency was unaffected after weight loss (Table 2). Pulse analysis showed that mean 24-h PRL secretion was significantly reduced by weight loss (126.0 ± 18.6 μg/l before vs. 56.1 ± 14.5 μg/l distribution volume−1 × 24 h after weight loss, P = 0.04).
The mean insulin release level, as well as its total daily secretion, significantly decreased after treatment: 333.2 ± 121 μg/ml vs. 65 ± 24 μg/ml (P < 0.001) and 7,689 ± 3,158 μg/ml × day vs. 1,311 ± 331 μg/ml × day (P < 0.01), respectively. The insulin secretion and pulsatility data are reported in Table 2. The spectral composition of the averaged relative release pattern was quite similar in both clinical conditions. The best fitting was achieved with before treatment, and with after the treatment (Figure 1b). The period of the fundamental harmonic assumed was ∼ 23 and ∼ 22(hour/cycle), before and after BPD, respectively. The harmonics with period ∼ 4 and ∼ 5(hour/cycle) had a higher amplitude after treatment.
LPL mRNA levels significantly decreased after BPD (42.63 ± 4.21 to 19.00 ± 2.74% of cyclophilin mRNA, P = 0.001) in agreement with the decrease in LPL protein level (from 8.94 ± 2.73 to 3.16 ± 1.05, P = 0.038 expressed as ratios of protein of interest vs. housekeeping protein).
Single linear correlations are summarized in Table 3. In Table 4 are reported the data relative to a multiple regression analysis (R2 = 0.78) including fat mass, FFM, and waist circumference, as a measure of the abdominal fat deposition, insulin secretion, and insulin sensitivity as independent variables. The best predictor of PRL secretion turned out to be the insulin secretion rate (P = 0.003).
Table 3. Pearson's correlations
Table 4. Multiple regression analysis with PRL secretion as dependent variable
The major determinants of LPL expression in the adipose tissue were insulin secretion rate and PRL secretion rate.
The major findings of this study are that in severely obese women after bariatric surgery: (i) plasma levels of PRL consistently decrease during both day and night time; (ii) PRL peak shifts from late evening to early morning, peak time previously reported in the literature as peculiar of lean, healthy subjects; (iii) the decrease in PRL levels significantly correlates with the decrease in both insulin levels and LPL mRNA or protein expression in subcutaneous adipose tissue; (iv) insulin secretion is the major determinant of PRL secretion.
It has been recently shown that PRL release is enhanced in obese premenopausal women proportionally to their BMI and that it is associated with the size of the visceral fat area (29). However, PRL secretion rate in obese women, although significantly reduced after 50% loss of overweight (15% absolute weight loss), was blunted compared to lean controls (17). In this study we found that severely obese women, following major weight reduction as a consequence of malabsorptive bariatric surgery, had a significant lowering of plasma PRL levels during the whole day. All PRL concentration and secretion parameters became similar to the values reported in the literature for normal-weight, healthy women (15).
Although the weight loss achieved in our series was much higher (from BMI 33 to 28 kg/m2 in Kok et al. (17) and from BMI 53 to 39 kg/m2 in our study) than that reported in the study by Kok et al. (17), our patients remained frankly obese. It is relevant that while in the study by Kok et al. (17) the weight loss was obtained through a caloric restriction, in our series the subjects had a normal to high calorie intake, as previously reported (30), but, thanks to the lipid malabsorption induced by surgery, they lost weight. Although the absolute values of EE were significantly reduced after BPD, 24-h EE normalized by FFM did not differ significantly. These results agree with the data reported in the literature showing that FFM is the major determinant of resting metabolic rate, explaining 82% of its variance (31). In our series, insulin sensitivity was significantly increased after BPD. Accordingly, the metabolic capacity to oxidize glucose was higher after the operation as suggested by the higher values of npRQ.
Data in the literature seem to support the hypothesis that the restriction of lipid metabolizable energy intake rather than weight loss induces the reduction in the circulating levels of PRL as well as the normalization of its secretion rhythm. In fact, a significant weight loss achieved in obese women after 1 month of a very low-calorie diet was unable to change the circulating levels of PRL significantly (32). In contrast, a very low-calorie diet containing only 1.5 g of lipids was able to suppress PRL synthesis and secretion as measured in response to thyroid-releasing hormone stimulus (33).
The change in insulin secretion was the major determinant of the modification in PRL secretion rate, suggesting that hyperinsulinemia plays a major role in the hormonal alterations of severe obesity. In fact, in primary monolayer cultures of male rats' pituitary cells, insulin caused a significant stimulation of basal PRL secretion (34).
In our series, PRL secretion was significantly associated with both LPL gene and protein expression, suggesting a role of PRL in regulating adipose tissue metabolism in severe obesity.
LPL is produced in parenchymal cells, mainly adipocytes, and transported to the luminal surface of the capillary where it exerts its activity (35). Chylomicrons and very low-density lipoproteins activate LPL through C-II apoprotein. Free fatty acids produced by the action of LPL on these lipoproteins are, then, taken up by adipocytes through a concentration-dependent mechanism involving a transmembrane transport protein. Inside the adipocytes, free fatty acids are re-esterified to triglyceride and in part β-oxidized.
Insulin is the primary hormone acting by increasing lipogenesis. Both insulin and PRL (36) enhance the differentiation of fibroblasts and preadipocytes into mature adipocytes. Therefore, the reduction in PRL plasma levels is likely dependent on the lowering of insulin plasma concentration which, in turn, produces the reduction in LPL adipose tissue expression. On the other hand, the large increase in insulin sensitivity did not influence PRL secretion, whereas insulin secretion did.
In conclusion, lipid malabsorption but normal calorie intake rather than reduced calorie intake in severe obesity might be related to the net improvement in PRL secretion after weight loss, which became similar to the values reported in the literature for healthy women (15), even in the face of persistence of a frank obesity. Furthermore, the reduced expression of LPL in the adipose tissue observed after bariatric surgery, being significantly correlated with the decrease in insulin and PRL, suggests a role of hyperinsulinenima and hyperprolactinemia in inducing and maintaining obesity.