Relationship of high leptin levels with an adverse lipid and insulin profile in 6–8 year-old children in Spain

Accepted Manuscript Relationship of high leptin levels with an adverse lipid and insulin profile in 6-8 yearold children in Spain A. Jois, P. Navarro, H. Ortega-Senovilla, T. Gavela-Pérez, L. Soriano-Guillén, Dr C. Garcés PII:

S0939-4753(15)00216-1

DOI:

10.1016/j.numecd.2015.09.005

Reference:

NUMECD 1487

To appear in:

Nutrition, Metabolism and Cardiovascular Diseases

Received Date: 20 March 2015 Revised Date:

31 July 2015

Accepted Date: 8 September 2015

Please cite this article as: Jois A, Navarro P, Ortega-Senovilla H, Gavela-Pérez T, Soriano-Guillén L, Garcés C, Relationship of high leptin levels with an adverse lipid and insulin profile in 6-8 yearold children in Spain, Nutrition, Metabolism and Cardiovascular Diseases (2015), doi: 10.1016/ j.numecd.2015.09.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Relationship of high leptin levels with an adverse lipid and insulin profile in 6-8 year-old children in Spain Jois Aa, 1, Navarro Pa, Ortega-Senovilla Hb, 2, Gavela-Pérez Tc, Soriano-Guillén Lc and Garcés Ca

RI PT

* Dedicated to Prof. Manuel de Oya, as the warmest homage to his memory

Lipid Laboratory, IIS-Fundación Jiménez Díaz, UAM, Madrid, Spain.

b

Servicio de Bioquímica-Investigación, Hospital Ramón y Cajal, Madrid, Spain

c

Department of Pediatrics, IIS-Fundación Jiménez Díaz, UAM, Madrid, Spain.

1

Present address: Faculty of Medicine, Nursing and Health Sciences, Monash University,

M AN U

SC

a

Melbourne, Australia. 2

Present address: Faculties of Pharmacy and Medicine, Universidad San Pablo-CEU, Madrid,

TE D

Spain

Word counts: text: 2450; abstract: 247; references: 31; figures: 2; tables: 4.

AC C

Dr Carmen Garcés

EP

Corresponding author:

Lipid Laboratory

IIS-Fundación Jiménez Díaz, UAM. Avda. Reyes Católicos, 2 28040 Madrid

Spain

[email protected] Telephone/ FAX number: +34-91-5432880

ACCEPTED MANUSCRIPT Abstract Background and aims: Leptin, an adipokine elevated in obesity, may be related to an adverse cardiovascular risk profile in childhood. However, evidence for this relationship in pre-

and insulin profiles in Spanish children.

RI PT

pubertal children is scarce. We aimed to analyze the relationship between leptin levels and lipid

Methods and results: Our population-based sample included 389 males and 369 females aged 6 to 8 years. Lipid levels were determined by standard methods, insulin by

SC

radioimmunoassay and leptin by sandwich ELISA.

Leptin levels were higher in girls (8.6 ng/ml) than boys (4.7 ng/ml) (p<0.001). Leptin

M AN U

increased from ages 6 to 8 in girls, but remained steady in boys. In both sexes, leptin increased significantly (p<0.001) across weight category from normal weight to obese. Children in the highest tertile of leptin concentration showed significantly (p<0.01) lower levels of HDLcholesterol (HDL-C) and apolipoprotein-AI (apo-AI) and significantly higher triglyceride (TG)

TE D

levels than children in lower tertiles. However, in linear regression analysis, after adjustment for body mass index (BMI), leptin only accounted for 1.5% of the variance of HDL-C in boys, and 2.6% of the variance of apo-AI in girls. Leptin was strongly and positively correlated with

EP

insulin and HOMA. Upon regression analysis, leptin contributed to over 20% of the variability in insulin and HOMA, independent of BMI.

AC C

Conclusion: Leptin levels show sex differences in pre-pubertal children. In this age group, leptin levels are strongly related to insulin, and affect lipid profile –namely HDL-C, apo-AI and TG– particularly when leptin levels are high.

Keywords: Leptin; HDL-Cholesterol; apo-AI; TG; insulin; 6-8 year-old children.

ACCEPTED MANUSCRIPT Introduction Childhood obesity is associated with an adverse cardiovascular risk profile, which contributes to an increased prevalence of cardiovascular and metabolic disease in both childhood and adulthood1.

RI PT

Obesity is characterized by excess adipose tissue, which is no longer solely considered an energy reservoir, but rather a complex endocrine organ with critical roles in energy

metabolism2. Its functions are mediated by adipose tissue-derived hormones known as

SC

adipokines3. In obesity, the secretion of these adipokines becomes dysregulated, which is

thought to contribute to adverse cardiovascular and metabolic risk profiles4. The strongest

M AN U

evidence linking cardiovascular risk and adipokines involves leptin5. It has been suggested that leptin could be involved in not only the consequences, but also in the aetiology of obesity, through the development of hyperleptinaemia and leptin resistance6. Leptin plays a key role in the regulation of body weight and energy expenditure, through

TE D

its effects on the hypothalamus7, as well as a number of peripheral actions. These include the direct stimulation of glucose metabolism, and improvement in insulin sensitivity8. In addition, leptin has a number of regulatory roles in the cardiovascular system4, 5. Consequently, an

EP

imbalance in leptin levels, such as hyperleptinaemia, can lead to adverse cardiovascular and metabolic outcomes.

AC C

Hyperleptinaemia is apparent in both childhood and adulthood obesity6 9, and may be implicated in the onset of insulin resistance. An association has been found between high leptin levels and insulin resistance10 11 12 13, although findings are conflicting14, and appear to differ depending on ethnicity10 and weight category11. In addition to its relationship with insulin, leptin has been positively associated with an adverse lipid profile in adults15 16 and children12 17 18 19 20. However, most studies in children have focused on specific populations, such as obese children, children with concomitant disease, and varying ethnicities. Furthermore,

ACCEPTED MANUSCRIPT most previous studies have had small sample sizes and included a broad age range. Thus, limited evidence is available in Caucasian pre-pubertal children regarding the possible association of leptin with obesity-related alterations independent of BMI. In our study, we investigated the relationship between leptin and lipid levels and insulin

AC C

EP

TE D

M AN U

SC

of Spanish children aged between 6 and 8 years.

RI PT

sensitivity depending on sex, by analyzing this association in a large, population-based cohort

ACCEPTED MANUSCRIPT Methods Study population: We included a population-based sample of 758 (389 boys and 369 girls) 6-8-year-old children. Subjects were participants in a cross-sectional study designed to analyze cardiovascular risk factors in Spanish schoolchildren. Children were selected by

RI PT

random cluster-sampling in schools, and stratified by sex and type of school (public versus private). Sampling was carried out in two stages: in the first stage, schools were selected from lists supplied by the Regional Educational Authorities; in the second, classrooms and pupils

SC

were selected. All children reported by parents to be suffering from metabolic, endocrine, liver or kidney disorders were excluded in order to minimise the possibility of the values of any of

M AN U

the variables of interest being altered. The study protocol complied with the Helsinki Declaration guidelines and Spanish legal provisions governing clinical research on humans, and was approved by the Clinical Research Ethics Committee of the Fundación Jiménez Díaz Hospital in Madrid. Parents were required to sign a written consent form allowing their children

Data collection

TE D

to participate.

Anthropometric variables: Measurements were taken with children wearing light clothing

EP

and barefoot. Weight was determined to the nearest 0.1 kg using a standardized digital scale, and height was measured to the nearest 0.1 cm using a portable stadiometer. Body mass index

AC C

(BMI, weight in kilograms divided by height in meters squared, kg/m2) was calculated from these parameters. We considered children to be overweight (OW) or obese if their BMI exceeded the age- and sex-specific cut-off points proposed for children by Cole et al. in a synthesis of international studies21. Biochemical data: Blood samples were obtained early in the morning after a 12-h fasting period by venipuncture into vacutainer tubes. Samples were kept on ice and sent to the laboratory for analysis. Once centrifuged, fractions were separated and frozen at –70ºC.

ACCEPTED MANUSCRIPT Cholesterol and triglycerides (TG) were measured enzymatically (Menarini Diagnostics, Firenze, Italy) with an RA-1000 Autoanalyzer (Technicon Ltd, Dublin, Ireland). High-density lipoprotein-cholesterol (HDL-C) was measured after precipitation of apo-B-containing lipoproteins with phosphotungstic acid and Mg (Roche Diagnostics, Madrid, Spain). Low-

RI PT

density lipoprotein-cholesterol (LDL-C) was calculated according to Friedewald’s formula. Plasma apolipoprotein A-I (apo-AI) and apolipoprotein B (apo-B) concentrations were

measured by immunonephelometry (Dade Berhing, Frankfurt, Germany). Serum leptin levels

SC

were determined by ELISA using a commercially available kit (Leptin CAN-L-4260, Diagnostics Biochem Canada Inc.). Serum insulin concentrations were measured by

M AN U

radioimmunoassay (RIA) using a commercial kit (BI-Insulin IRMA, Bio-Rad, France). Insulin resistance was estimated using the homeostasis model assessment for insulin resistance (HOMA; fasting insulin [µU/ml] x fasting glucose [mmol/l]/22.5). Free fatty acid (FFA) levels were measured by using the Wako commercial kit (Wako Industries, Osaka, Japan)

TE D

Statistical analysis: Statistical analyses were performed using the SPSS software package, version 21.0 (SPSS, Inc. Chicago, IL). The results are expressed as mean with the 95% confidence interval. Given their skewed distribution, leptin, TG, FFA, insulin and HOMA data

EP

were log transformed prior to statistical analysis. Differences in leptin levels by sex were assessed using a student t-test. Differences in variables by age, weight category and leptin tertile

AC C

in boys and girls were evaluated by one-way analysis of variance (ANOVA), and subsequent Post-hoc analysis. As leptin levels did not demonstrate normal distribution, as assayed by the one-sample Kolmogorov-Smirnov test, Spearman correlation analyses were performed to evaluate the relationships between leptin levels and anthropometric and biochemical variables. Partial correlations were used to analyze this association after adjusting for BMI. Regression analyses were performed with lipid and insulin factors as dependent variables and leptin as an independent variable.

ACCEPTED MANUSCRIPT Results Anthropometric and biochemical characteristics of the children studied are shown in table 1. Mean leptin levels in boys were 4.8 ng/ml (95% confidence interval: 4.1-5.5) with a median value of 2.25 ng/ml; girls’ mean leptin levels were 8.6 ng/ml (95% confidence interval: 7.6-9.5),

RI PT

with a median value of 5.24 ng/ml. As leptin levels were significantly higher in girls than in boys (p<0.001), all results are presented for girls and boys separately.

Figure 1 shows serum leptin levels in boys and girls according to age. In girls, leptin

SC

levels are higher with each year of increase in age from 6-8 years (p<0.001). The association between leptin levels and age remains significant after adjusting for BMI in a univariate

M AN U

analysis of variance. No significant variations in leptin levels across age categories were observed in boys.

Serum leptin levels are significantly higher in obese than in normal weight (NW) or overweight (OW) boys and girls, and significantly higher in OW than in NW children (table 2).

TE D

Leptin levels were found to have a significant strong positive correlation with BMI in both boys (0.629, p<0.001) and girls (0.663, p<0.001).

In Spearman correlation analysis (table 3), leptin levels showed significant negative

EP

correlations with HDL-C and apo-AI, and a positive correlation with TG levels in both boys and girls. Leptin levels were strongly positively correlated with insulin and HOMA, and negatively

AC C

correlated with FFA in both sexes. Partial correlation coefficients were computed adjusting for BMI (table 3). Following adjustment, leptin levels remained negatively correlated with apo-AI levels in boys only. However, associations between leptin and insulin and HOMA remained highly significant. The association between leptin and FFA also persisted following adjustment. Stepwise regression analyses were performed with HDL-C, apo-AI, TG, insulin, HOMA and FFA as dependent variables, after adjustment for BMI. Leptin was considered as an independent variable. Leptin was a predictive variable for HDL-C in boys, and for apo-AI in

ACCEPTED MANUSCRIPT girls, accounting for 1.5% and 2.6% of the variance respectively (table 4). In both genders, leptin explained up to 20% and 17% of the variance of insulin and HOMA respectively. Leptin accounted for 2.1% of the variance of FFA in boys and 5.3% in girls (table 4). The relationship between leptin and lipid levels was also analyzed by mean values by

RI PT

leptin tertile (figure 2). Children in the highest leptin tertile had significantly (p<0.01) higher TG levels, and significantly (p<0.01) lower apo-AI and HDL-C levels than the first or second

AC C

EP

TE D

M AN U

SC

tertiles (figure 2).

ACCEPTED MANUSCRIPT Discussion We described leptin levels in a large, population-based sample of 6-8 year-old children. Information regarding leptin levels in healthy Mediterranean children is scarce22. As such, our study may help to characterize the normal leptin range in a large, well-described cohort of

RI PT

pre-pubertal Caucasian children.

In agreement with previously reported findings23, sex-related differences in leptin levels were already apparent in our pre-pubertal children. Girls had higher mean leptin levels than

SC

boys, and their levels were higher from ages 6 to 8, while boys’ levels did not show any

significant difference between ages. This sexual dimorphism could be due to sex differences

M AN U

in body composition. It is known that the physical changes that occur during puberty are in part related to sex-specific hormonal profiles. Girls have significantly more body fat, which has been related to the increase of estradiol levels23. However, sex differences in body composition exist before puberty, and Garnett et al showed that hormones explained only 3-17% of the

TE D

variations in body composition in prepubertal children, suggesting that other factors are related with prepubertal sexual dimorphism24. As leptin is secreted by adipocytes in proportion to adipose tissue mass25, this may contribute to the higher levels reported in girls. Unfortunately,

EP

specific body composition data were not available in our population, and is the main limitation of our study. However, we found a strong association between leptin and anthropometric

AC C

variables, including BMI, as well as with the presence of overweight and obesity. The association of leptin levels with obesity and its correlation with BMI in childhood is well described in the literature 23 26. Our study revealed a significant association between leptin and features of an adverse cardiometabolic profile in children as young as 6 years old, a finding with important clinical implications. Higher leptin levels were associated with features of an adverse lipid (decreased HDL-C and apo-AI levels, and increased TG) and insulin profile.

ACCEPTED MANUSCRIPT To our knowledge, this is the first population-based study analyzing the relationship of leptin with lipid profile in a large, well-described cohort of pre-pubertal Caucasian children. In our study, the correlations between leptin with TG and HDL-C disappeared after adjusting for BMI, and its relationship with apo-AI remained significant in boys only. In regression

RI PT

analysis, leptin explained only a small percentage of variability in HDL-C levels in boys, and apo-AI in girls, after adjusting for BMI. In contrast, when analyzing lipid profile by tertile, children in the highest leptin tertile had significantly lower HDL-C and apo-A1, and higher

SC

TG levels compared to the lower two tertiles.

Previous studies in children of different ages and ethnicities have also reported a positive

M AN U

association between leptin and TG levels and a negative association with HDL-C12 17 18. However, BMI is also inversely associated with HDL-C and positively associated with TG levels. Whether leptin has an association with an adverse lipid profile independent of BMI remains unclear. The design of our study does not allow us to clarify this aspect, however a

TE D

study in mice has shown that low dose leptin administration decreased apo-AI mRNA and lowered HDL-C without any change in body weight27. Comparable to our results, past studies in children have shown a similar pattern, where

EP

significant findings on Spearman analysis have disappeared following adjustment for BMI 17-19. However, there is evidence to support our finding that upon regression analysis, after adjusting

AC C

for BMI, the relationship between leptin and TG levels persists17 18, suggesting an independent association.

A possible reason for the discrepancy in both our results and the current body of literature based on the type of analysis performed is that the relationship between leptin and aspects of lipid profile is non-linear. This is supported by our findings on tertile analysis, where children in the highest tertile showed an adverse lipid profile, while the lower two tertiles had relatively similar lipid profiles. Indeed, the Taipei Children Heart Study also reported that 12-16 year old

ACCEPTED MANUSCRIPT children in the highest quartile of leptin levels exhibited a worse lipid profile (low HDL-C and higher TG) than children in the lower quartiles17 18. Therefore, we propose that the association between leptin and TG, HDL-C and apo-AI may be present only in children with leptin resistance. The assumption that leptin could have

RI PT

an adverse relationship to lipid profile only in those with leptin resistance is supported by previous animal studies. In leptin sensitive rats, acute leptin administration reduced TG levels through fatty acid oxidation. However, in rats that were leptin resistant through a genetic

SC

defect, no changes occurred upon leptin administration, and TG levels were elevated28.

Our study also showed a strong positive association between leptin levels and insulin and

M AN U

HOMA index, independent of BMI. Leptin accounted for more than 20% of variability in both parameters. This suggests that leptin levels may already be contributing to insulin sensitivity in pre-pubertal children. Other studies in children and adolescents have demonstrated a significant association between leptin and insulin resistance, independently of gender and anthropometric

TE D

variables13. Studies in children of similar age to ours support our conclusion that the association of leptin with insulin sensitivity is already present before puberty12 13 19. The relationship between leptin and insulin appears to be bi-directional. Leptin receptors

EP

have been found in pancreatic tissue, and hyperleptinaemia has been found to modify glucose stimulated insulin secretion29. However, insulin has also been shown to have a direct regulatory

AC C

effect on leptin levels30. Current opinion is that the two hormones are closely connected, supported by the discovery that the two hormones use some of the same transduction and intracellular signaling pathways31. Despite these interesting findings, an inherent limitation of all cross-sectional studies is the inability to demonstrate causality. Therefore, further studies are needed to confirm the causal nature of these associations. In conclusion, our data have shown an association between high leptin levels and an

ACCEPTED MANUSCRIPT unfavorable cardiometabolic profile in children of pre-pubertal age, and confirmed the considerable strength of the relationship between leptin levels and insulin sensitivity. Rather than a simple linear association, there may be a leptin resistant specific relationship between

AC C

EP

TE D

M AN U

SC

RI PT

leptin and adverse lipid profile.

ACCEPTED MANUSCRIPT Acknowledgements The article is dedicated to the late Prof. Manuel de Oya. Prof. de Oya designed the Four Province Study and the ideas reflected in our work can be traced back to his. This work was supported by a grant from the Fondo de Investigación Sanitaria (FIS

RI PT

14/00344) and Biobank grant FEDER RD09/0076/00101. The contract of C Garcés is co-

financed by the Fondo de Investigación Sanitaria. Pilar Navarro is a fellow of the Conchita

AC C

EP

TE D

M AN U

SC

Rábago Foundation.

ACCEPTED MANUSCRIPT

1 Baker JL, Olsen LW, Sørensen TI. Childhood body-mass index and the risk of coronary heart disease in adulthood. N Engl J Med 2007; 357:2329–37. 2 Galic S, Oakhill JS, Steinberg GR. Adipose tissue as an endocrine organ. Mol Cell Endocrinol

RI PT

2010; 316:129–39.

3 Havel PJ. Control of energy homeostasis and insulin action by adipocyte hormones: leptin, acylation stimulating protein, and adiponectin. Curr Opin Lipidol 2002; 13:51-9.

SC

4 Mattu HS, Randeva HS. Role of adipokines in cardiovascular disease. J Endocrinol 2013; 216:T17-36.

Endocrinol 2004; 181:1–10.

M AN U

5 Ren J. Leptin and hyperleptinemia - from friend to foe for cardiovascular function. J

6 Zhang Y, Scarpace PJ. The role of leptin in leptin resistance and obesity. Physiol Behav 2006; 88:249-56.

TE D

7 Sahu A. Leptin signaling in the hypothalamus: emphasis on energy homeostasis and leptin resistance. Front Neuroendocrinol 2003; 24:225–53. 8 Kamohara S, Burcelin R, Halaas JL, Friedman JM, Charron MJ. Acute stimulation of glucose

EP

metabolism in mice by leptin treatment. Nature 1997; 389:374–7.

AC C

9 Klok MD, Jakobsdottir S, Drent ML. The role of leptin and ghrelin in the regulation of food intake and body weight in humans: a review. Obes Rev 2007; 8:21-34. 10 Mente A, Razak F, Blankenberg S, Vuksan V, Davis AD, Miller R, et al. Ethnic variation in adiponectin and leptin levels and their association with adiposity and insulin resistance. Diabetes Care 2010; 33:1629–34. 11 Galletti F, D'Elia L, De Palma D, Russo O, Barba G, Siani A, et al. Hyperleptinemia is associated with hypertension, systemic inflammation and insulin resistance in overweight but not in normal weight men. Nutr Metab Cardiovasc Dis. 2012; 22:300–6.

ACCEPTED MANUSCRIPT 12 Chu NF, Chang JB, Shieh SM. Plasma leptin, fatty acids, and tumor necrosis factorreceptor and insulin resistance in children. Obes Res 2003; 11:532–40. 13 Slinger JD, van Breda E, Keizer H, Rump P, Hornstra G, Kuipers H. Insulin resistance,

RI PT

physical fitness, body composition and leptin concentration in 7-8 year-old children. J Sci Med Sport 2008; 11:132–8.

14 Ceddia RB, Koistinen HA, Zierath JR, Sweeney G. Analysis of paradoxical observations

SC

on the association between leptin and insulin resistance. FASEB J 2002;16:1163–76.

15 Rainwater DL, Comuzzie AG, VandeBerg JL, Mahaney MC, Blangero J. Serum leptin levels

M AN U

are independently correlated with two measures of HDL. Atherosclerosis 1997; 132:237-43. 16 Haluzik M, Fiedler J, Nedvidkova J, Ceska R. Serum leptin levels in patients with hyperlipidemias. Nutrition 2000; 16:429-33.

17 Wu DM, Shen MH, Chu NF. Relationship between plasma leptin levels and lipid profiles

17:911–6.

TE D

among schoolchildren in Taiwan–the Taipei Children Heart Study. Eur J Epidemiol 2001;

18 Kavazarakis E, Moustaki M, Gourgiotis D, Drakatos A, Bossios A, Zeis PM, et al.

50:1091–4.

EP

Relation of serum leptin levels to lipid profile in healthy children. Metabolism 2001;

AC C

19 Valle M, Gascón F, Martos R, Bermudo F, Ceballos P, Suanes A. Relationship between high plasma leptin concentrations and metabolic syndrome in obese pre-pubertal children. Int J Obes Relat Metab Disord 2003; 27:13–8. 20 Demirel FF, Bideci AA, Cinaz PP, Camurdan MOM, Biberoğlu GG, Yesilkaya EE, et al. Serum leptin, oxidized low density lipoprotein and plasma asymmetric dimethyl arginine levels and their relationship with dyslipidaemia in adolescent girls with polycystic ovary syndrome. Clin Endocrinol (Oxf) 2007; 67:129–34.

ACCEPTED MANUSCRIPT 21 Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320:1240–3. 22 Venner AA, Doyle-Baker PK, Lyon ME, Fung TS. A meta-analysis of leptin reference

RI PT

ranges in the healthy paediatric prepubertal population. Ann Clin Biochem 2009; 46:65–72. 23 Blum WF, Englaro P, Hanitsch S, Juul A, Hertel NT, Müller J, et al. Plasma leptin levels in healthy children and adolescents: dependence on body mass index, body fat mass, gender,

SC

pubertal stage, and testosterone. J Clin Endocrinol Metab 1997; 82:2904–10.

24 Garnett SP, Högler W, Blades B, Baur LA, Peat J, Lee J, et al. Relation between hormones

72.

M AN U

and body composition, including bone, in prepubertal children. Am J Clin Nutr 2004; 80:966–

25 Maffei M, Halaas J, Ravussin E, Pratley RE, Lee GH, Zhang Y, et al. Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weight-reduced

TE D

subjects. Nat Med 1995; 1:1155–61.

26 Schoppen S, Riestra P, García-Anguita A, López-Simón L, Cano B, de Oya I, et al. Leptin and adiponectin levels in pubertal children: relationship with anthropometric variables and

EP

body composition. Clin Chem Lab Med 2010; 48:707–11. 27 Silver DL, Jiang XC, Tall AR. Increased High Density Lipoprotein (HDL), defective hepatic

AC C

catabolism of ApoA-I and ApoA-II, and decreased ApoA-I mRNA in ob/ob mice. Possible role of leptin in stimulation of HDL turnover. J Biol Chem. 1999; 274:4140-6 28 Shimabukuro M, Koyama K, Chen G, Wang MY, Trieu F, Lee Y, et al. Direct antidiabetic effect of leptin through triglyceride depletion of tissues. Proc Natl Acad Sci USA. 1997; 94:4637–41. 29 Emilsson V, Liu YL, Cawthorne MA, Morton NM, Davenport M. Expression of the functional leptin receptor mRNA in pancreatic islets and direct inhibitory action of leptin on insulin secretion. Diabetes 1997; 46:313–6.

ACCEPTED MANUSCRIPT 30 Wabitsch M, Jensen PB, Blum WF, Christoffersen CT, Englaro P, Heinze E, et al. Insulin and cortisol promote leptin production in cultured human fat cells. Diabetes 1996; 45:1435–8. 31 Niswender KD, Baskin DG, Schwartz MW. Insulin and its evolving partnership with leptin

AC C

EP

TE D

M AN U

SC

RI PT

in the hypothalamic control of energy homeostasis. Trends Endocrinol Metab 2004; 15:362–9.

ACCEPTED MANUSCRIPT Table 1. Anthropometric and biochemical characteristics (mean±SD) among Spanish school children. Girls n=369 7.1 ±0.6

0.61

BMI (Kg/m2)

16.8 ±2.4

17.0 ±2.4

0.25

Total cholesterol (mg/dl)

184.4 ±26.3

186.9 ± 27.9

0.20

Triglycerides (mg/dl)

71.1 ± 23.0

73.1 ± 22.9

0.06

HDL-C (mg/dl)

60.4 ± 13.5

58.7 ± 13.0

0.08

LDL-C (mg/dl)

109.7 ± 25.1

113.6 ± 26.3

0.04

Apo B (mg/dl) Glucose (mg/dl)

AC C

EP

FFA (mEq/l)

139.6 ± 19.1

136.9 ± 18.3

0.05

68.8 ± 13.6

71.5 ± 14.7

0.009

92.3 ± 8.8

89.2 ± 8.9

0.000

3.6 ± 2.5

3.7 ± 2.5

0.37

0.79 ± 0.61

0.82 ± 0.71

0.40

0.68 ± 0.27

0.72 ± 0.30

0.03

TE D

Insulin (µU/mL) HOMA

SC

M AN U

Apo AI (mg/dl)

p

RI PT

Age (years)

Boys n=389 7.2 ±0.6

ACCEPTED MANUSCRIPT Table 2. Leptin levels (ng/ml) (mean, 95% CI) by weight category in boys and girls Boys (n=389)

Girls (n=369)

2.6 (2.2-2.9)

NW (n=259 )

5.3 (4.5-6.0)

OW (n= 53)

8.5 (6.2-10.9)

OW (n= 77)

12.5 (10.6-14.4)

16.7 (12.2-21.3)

Obese(n= 33)

27.2 (22.8-31.6)

Obese (n= 30)

AC C

EP

TE D

M AN U

SC

NW: normal weight; OW: overweight.

RI PT

NW (n= 306)

ACCEPTED MANUSCRIPT

Spearman correlations

RI PT

Table 3: Spearman and partial correlations of leptin with lipid and insulin levels by sex Partial correlations (adjusted for BMI) Boys

Girls

0.040

-0.039

0.075

0.017

HDL-C

-0.111*

-0.137**

-0.058

0.002

Apo-AI

-0.111*

-0.121*

-0.113*

-0.012

TG

0.121*

0.143**

0.053

0.092

LDL-C

0.065

-0.015

0.098

0.002

Apo-B

0.082

-0.052

0.073

-0.052

Glucose

0.145**

0.073

0.034

0.017

Insulin

0.471***

0.447***

0.339***

0.316***

HOMA

0.459***

0.432***

0.325***

0.309***

-0.183**

-0.157*

-0.172**

FFA

TE D EP

AC C

TC

-0.157*

SC

Girls

M AN U

Boys

*p<0.05 **p<0.01 ***p<0.001; TC: total cholesterol; TG: triglycerides; HDL-C: high density lipoprotein-cholesterol; LDL-C: low density lipoprotein-cholesterol; FFA: free fatty acids.

ACCEPTED MANUSCRIPT

HDL-C

Apo-AI

TG

Insulin

B (SE)

B (SE)

B (SE)

B (SE)

NS

NS

1.5

-

-

NS

-6.311 (2.19)*

NS

-

2.6

-

R2 (%)

R2

EP

Leptin

20.5

TE D

Girls

FFA

B (SE)

B (SE)

0.280 (0.32)*** 0.291 (0.34)***

M AN U

Leptin -3.030 (1.42)*

HOMA

SC

Boys

RI PT

Table 4: Multiple linear regression models for HDL-C, apo-AI and TG as dependent variables after adjustment for BMI

20.2

-0.050 (0.02)* 2.1

0.266 (0.34)*** 0.264 (0.35)*** -0.084 (0.02)*** 17.2

16.2

5.3

AC C

SE: Standard error * p< 0.05 ** p< 0.01 *** p< 0.001. NS: not significant BMI: body mass index; TG: triglycerides; HDL-C: high density lipoprotein-cholesterol; FFA: free fatty acids

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Figure 1. Serum leptin levels (ng/ml) in 6-to-8 year-old boys and girls

ACCEPTED MANUSCRIPT Figure 2: Lipid levels (mg/dl) by leptin tertiles in boys and girls a) High density lipoprotein-cholesterol (HDL-C) by leptin tertile 65 60 Tertile 1 55

45

RI PT

Tertile 2

50

HDL-C girls

b) Apolipoprotein AI (Apo AI) by leptin tertile 150

130

M AN U

140

SC

Tertile 3 HDL-C boys

Tertile 1

120

Tertile 2

110 100

Tertile 3

Apo AI boys

Apo AI girls

70 60 50 40

EP

80

TE D

c) Triglycerides (TG) by leptin tertile

AC C

TG boys

Tertile 1 Tertile 2 Tertile 3

TG girls

Error bars: 95% confidence intervals. In boys: tertile 1 (0.25-1.42 ng/ml); tertile 2 (1.44-3.68 ng/ml); tertile 3 (3.69-40.89 ng/ml) In girls: tertile 1 (0.25-3.28 ng/ml); tertile 2 (3.29-8.90 ng/ml); tertile 3 (8.93-54.41 ng/ml)

ACCEPTED MANUSCRIPT •

The association of leptin with an adverse lipid profile in children, independent of BMI, remains unclear.

•

Our study showed that children in the highest tertile of leptin had an adverse lipid profile, while children in the lower tertiles had relatively similar lipid profiles.

•

The relationship between leptin levels and lipid variables may be non-linear, and

Our data contribute to explain discrepancies in the literature regarding the

EP

TE D

M AN U

SC

association between lipid profile and leptin levels.

AC C

•

RI PT

could be present only in children with leptin resistance.