Limits...
Leptin levels are negatively correlated with 2-arachidonoylglycerol in the cerebrospinal fluid of patients with osteoarthritis.

Nicholson J, Azim S, Rebecchi MJ, Galbavy W, Feng T, Reinsel R, Rizwan S, Fowler CJ, Benveniste H, Kaczocha M - PLoS ONE (2015)

Bottom Line: However, little is known about the corresponding central signaling.Endocannabinoids were quantified by liquid chromatography - mass spectrometry.AEA and 2-AG levels in the serum and CSF did not correlate with either age or BMI.

View Article: PubMed Central - PubMed

Affiliation: Department of Orthopedic Surgery, Stony Brook University, Stony Brook, New York, United States of America.

ABSTRACT

Background: There is compelling evidence in humans that peripheral endocannabinoid signaling is disrupted in obesity. However, little is known about the corresponding central signaling. Here, we have investigated the relationship between gender, leptin, body mass index (BMI) and levels of the endocannabinoids anandamide (AEA) and 2-arachidonoylglycerol (2-AG) in the serum and cerebrospinal fluid (CSF) of primarily overweight to obese patients with osteoarthritis.

Methodology/principal findings: Patients (20 females, 15 males, age range 44-78 years, BMI range 24-42) undergoing total knee arthroplasty for end-stage osteoarthritis were recruited for the study. Endocannabinoids were quantified by liquid chromatography - mass spectrometry. AEA and 2-AG levels in the serum and CSF did not correlate with either age or BMI. However, 2-AG levels in the CSF, but not serum, correlated negatively with CSF leptin levels (Spearman's ρ -0.48, P=0.0076, n=30). No such correlations were observed for AEA and leptin.

Conclusions/significance: In the patient sample investigated, there is a negative association between 2-AG and leptin levels in the CSF. This is consistent with pre-clinical studies in animals, demonstrating that leptin controls the levels of hypothalamic endocannabinoids that regulate feeding behavior.

No MeSH data available.


Related in: MedlinePlus

Relationship between endocannabinoid and leptin levels in the serum (A, B) and CSF (C, D).Shown are the data for AEA (Panels A, C) and 2-AG (Panels B, D). The Spearman ρ values are provided in Table 2.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4383333&req=5

pone.0123132.g001: Relationship between endocannabinoid and leptin levels in the serum (A, B) and CSF (C, D).Shown are the data for AEA (Panels A, C) and 2-AG (Panels B, D). The Spearman ρ values are provided in Table 2.

Mentions: Bivariate correlations were determined between the outcome parameters used in the present study. Before undertaking such correlations, it is important to determine whether the data are normally distributed because a parametric Pearson correlation is not appropriate if this is not the case. We used the D'Agostino & Pearson omnibus normality test and found that the serum 2-AG values were not normally distributed (all other variables reported passed this test). Consequently, we used non-parametric Spearman’s correlations for the whole dataset. The correlation coefficients for the endocannabinoids and for leptin are shown in Table 2, and the scatterplots for the endocannabinoids against the corresponding leptin concentrations are shown in Fig 1. As expected from the literature [15], CSF and serum leptin concentrations were highly correlated (Spearman’s ρ = 0.83, n = 31, P<0.0001) and thus it is not surprising that the significant association of BMI with leptin was seen for both serum and CSF leptin. In their original study, Schwartz et al. [41] reported that the best correlation for CSF leptin as dependent variable was seen in a model incorporating gender, plasma leptin and log plasma leptin as independents. However, this is likely to result in multicollinearity (a potential source of misleading results when one or more of the predictor variables are highly correlated, i.e. not independent), and this was indeed seen in our dataset. A simpler model with gender and either unlogged serum or logged serum leptin (but not both) gave similar r2 values (0.69 and 0.71) with both parameters contributing significantly, and the lowest level of collinearity (variance inflation factor value 1.8; as a rule of thumb, a value ≥5 indicates multicollinearity) was seen with the unlogged values (data not shown).


Leptin levels are negatively correlated with 2-arachidonoylglycerol in the cerebrospinal fluid of patients with osteoarthritis.

Nicholson J, Azim S, Rebecchi MJ, Galbavy W, Feng T, Reinsel R, Rizwan S, Fowler CJ, Benveniste H, Kaczocha M - PLoS ONE (2015)

Relationship between endocannabinoid and leptin levels in the serum (A, B) and CSF (C, D).Shown are the data for AEA (Panels A, C) and 2-AG (Panels B, D). The Spearman ρ values are provided in Table 2.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4383333&req=5

pone.0123132.g001: Relationship between endocannabinoid and leptin levels in the serum (A, B) and CSF (C, D).Shown are the data for AEA (Panels A, C) and 2-AG (Panels B, D). The Spearman ρ values are provided in Table 2.
Mentions: Bivariate correlations were determined between the outcome parameters used in the present study. Before undertaking such correlations, it is important to determine whether the data are normally distributed because a parametric Pearson correlation is not appropriate if this is not the case. We used the D'Agostino & Pearson omnibus normality test and found that the serum 2-AG values were not normally distributed (all other variables reported passed this test). Consequently, we used non-parametric Spearman’s correlations for the whole dataset. The correlation coefficients for the endocannabinoids and for leptin are shown in Table 2, and the scatterplots for the endocannabinoids against the corresponding leptin concentrations are shown in Fig 1. As expected from the literature [15], CSF and serum leptin concentrations were highly correlated (Spearman’s ρ = 0.83, n = 31, P<0.0001) and thus it is not surprising that the significant association of BMI with leptin was seen for both serum and CSF leptin. In their original study, Schwartz et al. [41] reported that the best correlation for CSF leptin as dependent variable was seen in a model incorporating gender, plasma leptin and log plasma leptin as independents. However, this is likely to result in multicollinearity (a potential source of misleading results when one or more of the predictor variables are highly correlated, i.e. not independent), and this was indeed seen in our dataset. A simpler model with gender and either unlogged serum or logged serum leptin (but not both) gave similar r2 values (0.69 and 0.71) with both parameters contributing significantly, and the lowest level of collinearity (variance inflation factor value 1.8; as a rule of thumb, a value ≥5 indicates multicollinearity) was seen with the unlogged values (data not shown).

Bottom Line: However, little is known about the corresponding central signaling.Endocannabinoids were quantified by liquid chromatography - mass spectrometry.AEA and 2-AG levels in the serum and CSF did not correlate with either age or BMI.

View Article: PubMed Central - PubMed

Affiliation: Department of Orthopedic Surgery, Stony Brook University, Stony Brook, New York, United States of America.

ABSTRACT

Background: There is compelling evidence in humans that peripheral endocannabinoid signaling is disrupted in obesity. However, little is known about the corresponding central signaling. Here, we have investigated the relationship between gender, leptin, body mass index (BMI) and levels of the endocannabinoids anandamide (AEA) and 2-arachidonoylglycerol (2-AG) in the serum and cerebrospinal fluid (CSF) of primarily overweight to obese patients with osteoarthritis.

Methodology/principal findings: Patients (20 females, 15 males, age range 44-78 years, BMI range 24-42) undergoing total knee arthroplasty for end-stage osteoarthritis were recruited for the study. Endocannabinoids were quantified by liquid chromatography - mass spectrometry. AEA and 2-AG levels in the serum and CSF did not correlate with either age or BMI. However, 2-AG levels in the CSF, but not serum, correlated negatively with CSF leptin levels (Spearman's ρ -0.48, P=0.0076, n=30). No such correlations were observed for AEA and leptin.

Conclusions/significance: In the patient sample investigated, there is a negative association between 2-AG and leptin levels in the CSF. This is consistent with pre-clinical studies in animals, demonstrating that leptin controls the levels of hypothalamic endocannabinoids that regulate feeding behavior.

No MeSH data available.


Related in: MedlinePlus