Insulin and Body Mass Index Decrease Serum Soluble Leptin Receptor Levels in Humans

Abstract

Purpose Test how serum soluble leptin receptor (sOb-R) is influenced by glucose, insulin, body fat, body mass index (BMI), food intake and physical activity. Methods We performed an epidemiological triangulation combining cross-sectional, interventional and Mendelian randomization study designs. In five independent clinical studies (n = 24-823), sOb-R was quantified in serum or plasma by commercial ELISA kits using monoclonal antibodies. We performed mixed models regression and two-sample Mendelian randomization. Results In pooled, cross-sectional data, levelling on study, sOb-R associated inversely with body mass index (BMI) (beta [95% CI] -0.19 [-0.21 to -0.17]), body fat (-0.12 [-0.14 to -0.10) and fasting C-peptide (-2.04 [-2.46 to -1.62]). sOb-R decreased in response to acute hyperinsulinaemia during euglycaemic glucose clamp in two independent clinical studies (-0.5 [-0.7 to -0.4] and -0.5 [-0.6 to -0.3]), and immediately increased in response to intensive exercise (0.18 [0.04 to 0.31]) and food intake (0.20 [0.06 to 0.34]). In two-sample Mendelian randomization, higher fasting insulin and higher BMI were causally linked to lower sOb-R levels (inverse variance weighted, -1.72 [-2.86 to -0.58], and -0.20 [-0.36 to -0.04], respectively). The relationship between hyperglycaemia and sOb-R were inconsistent in cross-sectional studies, non-significant in intervention studies, and two-sample Mendelian randomization suggested no causal effect of fasting glucose on sOb-R. Main conclusion Both BMI and insulin causally decreased serum sOb-R levels. Conversely, intensive exercise and food intake acutely increased sOb-R. Our results suggest that sOb-R is involved in short-term regulation of leptin signalling, either directly or indirectly, and that hyperinsulinaemia may reduce leptin signalling.

Metformin Affects Plasma Levels of Soluble Leptin Receptors in Type 2 Diabetes Mellitus and Metabolic Syndrome Patients

“…It has been found in this study that in T2DM patients, the correlation value of plasma levels of sOB-R with glycemic control (HbA1c) was statistically non-significant, moderately positive (r = 0.476, P-value = 0.117), whereas, in MetS patients, there was a weak positive correlation (r = 0.050, Pvalue = 0.864). These findings agreed with the results of Sommer et al, who reported that HbA1c was not associated with sOB-R in simple regression models ( β = 0.04, P-value = 0.70) [9].…”

“…A huge prospective study found that lower T2DM risk is linked to elevated plasma sOB-R levels [8]. Serum sOB-R may protect against future T2DM or function as a sign of protective characteristics, but its regulation is not yet known [9].…”

Insulin and body mass index decrease serum soluble leptin receptor levels in humans

“…Instrument strength in the 2-sample Mendelian randomization was calculated for each genetic instrument as (β/SE^2). Lists of independent genetic variants used as instruments for the exposures with effect size for both exposure and the outcome sOb-R are presented in Supplementary Tables S2 to S5 (26). Due to a lack of independent genetic instruments for sOb-R after pruning and harmonization, we were not able to test the effect of sOb-R on fasting glucose, fasting insulin, BMI, body fat, or T2D.…”

“…We calculated genetic correlation between sOb-R and its causally related phenotypes with the Complex-Traits Genetics Virtual Lab (37) using full GWAS summary statistics for sOb-R (29), fasting insulin (31), and BMI (33). To test whether causal effects of correlated exposures were independent of each other, we performed multivariable Mendelian randomization using the MendelianRandomization package (38) for R. Genetic variants used as instruments in the multivariable Mendelian randomization are presented in Supplementary Table S6 (26). To test instrument strength and validity in the multivariable Mendelian randomization, we used equation 12 and 13, respectively, from Sanderson et al (39), modified using the intercept from the genetic correlation analysis of BMI and insulin to estimate sampling covariances as suggested by Wu et al 2021 (40) (detailed description in the Supplementary file [26]).…”

“…Due to individual variation in plasma insulin in response to insulin infusion, we explored the relationship between actual changes in plasma insulin and sOb-R from fasting to physiological and supraphysiological hyperinsulinemia (Supplementary Fig. S2 [26]). At physiological infusion of insulin, we observed a negative relationship between Δinsulin and ΔsOb-R levels (rho = −0.38; P = .03), whereas the relationship leveled out at a supraphysiological insulin infusion (rho = 0.05; P = .78) (Supplementary Fig.…”