Visceral fat mass is a strong predictor of circulating ghrelin levels in premenopausal women
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
European Journal of Endocrinology (2009) 160 375–379 ISSN 0804-4643
CLINICAL STUDY
Visceral fat mass is a strong predictor of circulating ghrelin
levels in premenopausal women
E Sondergaard, L C Gormsen, B Nellemann, E T Vestergaard, J S Christiansen and S Nielsen
Medical Department M (Endocrinology & Diabetes), Aarhus University Hospital, DK-8000 Aarhus C, Denmark
(Correspondence should be addressed to E Sondergaard; Email: esbens@dadlnet.dk)
Abstract
Objective: A well known inverse relationship exists between obesity and circulating ghrelin
concentrations. However, obesity is a heterogeneous disease entity and upper-body obesity (UBO) is
associated with more profound metabolic disturbances than lower-body obesity (LBO). We therefore
aimed to investigate the impact of body composition on circulating ghrelin levels in women spanning a
wide range of body composition phenotypes.
Subjects and methods: Ten (UBO; waist-to-hip ratio (WHR) O0.85, body mass index (BMI) O28 kg/m2),
ten LBO (WHR !0.80, BMI O28 kg/m2) and ten lean women (BMI!25 kg/m2) were studied. Total
ghrelin levels were measured under basal and hyperinsulinemic (0.6 mU/kg per min) conditions. Body
fat distribution was determined by dual X-ray absorptiometry in combination with computed
tomography at the L2-L3 level.
Results: As expected, an inverse correlation existed between basal ghrelin concentration and BMI
(rZK0.40, PZ0.03) and total fat mass (rZK0.39, PZ0.04). Visceral fat mass was a strong
predictor (rZK0.56, PZ0.003) of circulating ghrelin levels, even when adjusted for BMI (PZ0.02) or
body composition group (PZ0.04). The suppressive effect of insulin on ghrelin concentration was
significantly diminished in the UBO compared with the lean controls (PZ0.012) and a highly
significant inverse correlation existed with visceral fat mass (rZK0.52, PZ0.004).
Conclusions: Visceral fat mass is a strong predictor of basal ghrelin concentrations and also attenuates
the suppressive effect of insulin on ghrelin concentrations. These data provide further evidence that the
UBO phenotype is associated with more profound metabolic abnormalities than obesity per se.
European Journal of Endocrinology 160 375–379
Introduction study used dual X-ray absorptiometry (DXA) to measure
abdominal fat, a technique unsuited to differentiate
Ghrelin, an endogenous ligand for the GH secretagogue- between abdominal s.c. and visceral fat. In the latter
receptor, is involved in the regulation of appetite and study, magnetic resonance imaging (MRI) provided
stimulates food intake (1). Circulating ghrelin levels are estimates of visceral fat area rendering comparisons
inversely correlated with degree of obesity assessed by between the two studies difficult. To our knowledge,
body mass index (BMI) (2, 3). It is not clear which factor only one study (11) designed to assess the impact of
regulates this negative feedback mechanism, but various
body fat location on ghrelin levels have measured
hormones and metabolites that are altered in obesity
such as insulin (4) and free fatty acids (FFAs) (5) have visceral fat by combined DXA and computed
been suggested. However, obesity is a heterogeneous tomography (CT). This technique is considered by
disease entity, where upper-body obesity (UBO) is many the best to measure absolute visceral fat mass
characterized by a greater extent of hyperglycemia, (12). However, both men and women were included in
type 2 diabetes, and elevated FFAs (6, 7) than lower-body the study by Purnell et al. (11) and correct interpre-
obesity (LBO) (8). It is therefore possible that circulating tation of their results may therefore be difficult.
ghrelin levels may be influenced by body fat distribution. The aim of the present study was therefore to
Previous studies assessing the impact of body fat investigate the impact of body composition in healthy
distribution on ghrelin levels report conflicting results. premenopausal women on circulating levels of ghrelin.
A recent study in postmenopausal women reports an In these women, body fat distribution was thoroughly
inverse relationship with abdominally located fat (9), characterized by DXA, CT, and waist-to-hip ratio (WHR).
whereas others report a positive correlation with A second purpose was to assess the suppressive effect
visceral fat (10). Highlighting some of the difficulties of insulin on circulating ghrelin levels. It has previously
when comparing body composition studies, the former been demonstrated that insulin suppresses ghrelin
q 2009 European Society of Endocrinology DOI: 10.1530/EJE-08-0735
Online version via www.eje-online.org
Downloaded from Bioscientifica.com at 11/28/2020 10:09:26AM
via free access376 E Sondergaard and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2009) 160
levels (4) and that the suppression is blunted in obesity visceral fat mass were assessed using the CT measures
(13). But whether the inhibitory effect of insulin on of intra-abdominal and s.c. adipose tissue combined
ghrelin secretion is modified by body fat distribution with abdominal fat mass measured by DXA as
has to our knowledge not previously been investigated. previously described (12). Upper body s.c. fat was
estimated as the difference between upper body fat
(DXA) and visceral fat.
Methods
Thirty healthy, premenopausal women with varying Statistical analysis
body fat distribution (10 UBO (WHRO0.85, BMIO Data are expressed as the meanGS.D. Between groups,
28 kg/m2), 10 LBO (WHR!0.80, BMIO28 kg/m2), comparisons were done by one-way ANOVAwith Tukey’s
and 10 lean (BMI!25 kg/m2) were recruited in a study
test for multiple comparisons applied when appropriate.
designed to investigate triglyceride metabolism in
Univariate and multivariate regression analysis were
women with different obesity phenotypes (unpublished
performed to identify potential determinants of circulat-
observations). All women were studied in the luteal
ing ghrelin levels as well as determinants of the
phase. The two groups of obese women were matched
according to BMI to ensure comparability. All partici- suppressive effect of insulin on ghrelin levels. Corre-
pants were normotensive, non-smokers, used no lations were evaluated by Spearman’s rank correlation
medication except oral contraceptives, and had a coefficient. To compare the strength of correlations, the
normal blood count and chemistry panel documented approach of Choi was used (14). It involves regression
before participation. of the ranked dependent variable versus the difference
The study protocol was approved by the local Ethics in rank between the two independent variables. All
Committee (The Central Denmark Region Committees statistical analyses were performed using the SPSS 15.0
on Biomedical Research Ethics) and written informed software program. A level of P!0.05 was accepted as
consent was obtained from all participants. statistically significant.
Subjects were studied after an overnight fast. The
study day included a 5-h basal study period followed
by a 2 h hyperinsulinemic-euglycemic clamp (Insulin
(Actrapid; Novo Nordisk, Bagsvaerd, Denmark) 0.6 mU/ Results
kg per$min). During the clamp, plasma glucose concen-
trations were maintained at 5.0 mmol/l by infusion of Characteristics of the three study groups are presented
20% glucose solution. in Table 1. By design, both groups of obese subjects had
Serum ghrelin was determined by an in-house RIA comparable BMI’s and a significantly larger fat mass
(RIA). The assay measures immunoreactive levels of than the lean group. However, even though we had
ghrelin using 125I-labeled bioactive ghrelin as tracer anticipated that visceral fat mass would be greater in
and rabbit polyclonal antibodies raised against octa- the UBO group than in the LBO; the difference was not
noylated human ghrelin. The assay recognizes the statistically significant. The excess abdominal fat mass
COOH-terminal of ghrelin and determines acylated as of UBO subjects was therefore both s.c. and visceral.
well as des-acylated ghrelin. The intra-assay coefficient Both obese groups were insulin resistant in terms of
of variation is less than 3.9%. Values presented are greater levels of fasting insulin and impaired glucose
averaged from hourly measurements. infusion rate (GIR) during the hyperinsulinemic-eugly-
Total body fat, leg fat, fat percent, and fat-free mass cemic clamp. There were no differences in GIR between
were examined by DXA (QDR-2000). Upper-body and the obese groups (PZ0.54).
Table 1 Basal characteristics and ghrelin levels of the upper-body obese (UBO), the lower-body obese (LBO) and the lean groups.
UBO (nZ10) LBO (nZ10) Lean (nZ10) ANOVA (P value)
Age (years) 41.9G7.9 39.0G8.1 39.0G10.2 0.669
Weight (kg) 91.4G10.0* 86.3G6.9* 65.5G6.0 !0.001
BMI (kg/m2) 32.3G2.3* 30.4G1.9* 22.9G1.4 !0.001
Waist-to-hip ratio 0.91G0.05 0.77G0.04** 0.79G0.07** !0.001
Total fat mass (kg) 36.4G4.2* 35.7G3.8* 19.2G4.5 !0.001
Visceral fat (kg) 4.8G1.5* 4.0G1.5* 2.3G1.0 0.001
S.c. abdominal fat (kg) 13.5G2.6* 11.8G2.4* 5.3G1.9 !0.001
Fasting insulin (pmol/l) 60.3G16.2* 53.7G16.4* 23.4G10.0 !0.001
GIR (clamp; mg/kg per min) 3.9G1.6* 4.5G2.6 7.0G2.6 0.014
Ghrelin basal (mg/l) 0.60G0.16* 0.69G0.22 0.85G0.22 0.033
Ghrelin clamp (mg/l) 0.56G0.17 0.62G0.18 0.72G0.20 0.158
All values are meansGS.D. *P!0.05 compared with lean group, **P!0.05 compared with UBO group.
www.eje-online.org
Downloaded from Bioscientifica.com at 11/28/2020 10:09:26AM
via free accessEUROPEAN JOURNAL OF ENDOCRINOLOGY (2009) 160 Visceral fat mass and ghrelin levels 377
Basal ghrelin levels
The basal level of ghrelin was significantly lower in the
UBO (29%) and LBO (19%) group compared with the
lean group (ANOVA, PZ0.033), but there was no
significant difference between the UBO and the LBO
groups (Fig. 1).
As expected, an inverse correlation existed between
basal ghrelin concentration and BMI (rZK0.40,
PZ0.03), total fat mass (rZK0.39, PZ0.04), and
fasting insulin (rZK0.47, P!0.01). Additionally,
there was also a positive correlation between high Figure 2 Correlation between visceral fat mass and basal ghrelin
density lipoprotein (HDL) cholesterol and basal ghrelin concentrations. C, upper-body obese; B, lower-body obese; , lean.
(rZ0.51, P!0.01). Since abdominal s.c. fat has been
proposed as a determinant of ghrelin levels, we also controls (PZ0.012), whereas this was not the case in
tested whether this was the case in our study, but no the LBO group (PZ0.084; Dghrelin(BASAL-CLAMP) (mg/l):
significant correlation was found (PZ0.10). UBO: 0.03G0.03 vs LBO: 0.07G0.09 versus lean:
In line with our preliminary hypothesis, visceral fat 0.13G0.05, ANOVA PZ0.003; Fig. 3). The insulin
mass was a strong predictor (rZK0.56, PZ0.002) of stimulated ghrelin decrease did not significantly corre-
circulating ghrelin levels (Fig. 2). We used the approach late to BMI or total fat mass, whereas a highly
of Choi to determine if visceral fat mass was a better significant inverse correlation existed with visceral fat
predictor of basal ghrelin levels than BMI, but there mass (rZK0.52, PZ0.004; Fig. 4).
were no significant differences (PZ0.074).
Since visceral fat accumulation predisposes to insulin
resistance and elevated fasting insulin, a multivariate
regression analysis with basal ghrelin levels as depen- Discussion
dent variable and visceral fat mass and insulin as In the present study, we demonstrate that the visceral
independent variables was performed to elucidate the fat mass is a strong predictor of basal ghrelin
independent effect of visceral fat. The regression concentrations in a well characterized group of
analysis revealed that visceral fat was an independent premenopausal women. The inverse correlation
significant predictor of ghrelin levels even when between visceral fat mass and ghrelin levels persisted
adjusted for insulin (PZ0.03), GIR (PZ0.01), BMI even when adjusted for known predictors of ghrelin
(PZ0.02), body composition group (LBO, UBO or lean; levels such as BMI (2), insulin sensitivity (3), insulin
PZ0.04) or HDL cholesterol (PZ0.01), a known levels (3) or HDL cholesterol (11). In addition, we
ghrelin carrier molecule (15). observed that the suppressive effect of insulin on ghrelin
levels was blunted in UBO women compared with lean,
The suppressive effect of insulin on ghrelin whereas this was not the case in LBO women. Our
concentrations findings provide further evidence that upper-body
obesity impacts more profoundly on hormonal and
As anticipated, ghrelin was inversely correlated with metabolic parameters than LBO.
fasting insulin. The suppressive effect of insulin on Previous reports on the impact of body composition
ghrelin concentration (ghrelin decrease during the on ghrelin levels have been contradictory, which may
clamp (Dghrelin(BASAL-CLAMP))) was significantly atte- partly be attributable to methodological differences in
nuated in the UBO group when compared with the lean quantification of abdominal obesity. For instance,
Figure 3 Decrease of ghrelin under the hyperinsulinemic eugly-
Figure 1 Basal ghrelin levels in different body composition groups cemic clamp in different body composition groups (error bars are
(error bars are S.D). *PZ0.009 compared with lean group. S.D.) *PZ0.012 compared with the lean group.
www.eje-online.org
Downloaded from Bioscientifica.com at 11/28/2020 10:09:26AM
via free access378 E Sondergaard and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2009) 160
In addition to lowering total levels of ghrelin, obesity
also affects the normal oscillating pattern of the
hormone. Thus, the suppressive effect of a test meal
on ghrelin secretion (13) as well as the nocturnal
increase in ghrelin (7) is blunted in obese persons. In
this study, we demonstrate that not only the size but
also the location of excess fat depots impacts on the
ability of insulin to suppress ghrelin concentrations
under hyperinsulinemic-euglycemic clamp conditions.
To our knowledge, this loss of hormonal flexibility in
UBO subjects has not previously been reported.
Figure 4 Correlation between visceral fat mass and decrease of
ghrelin under hyperinsulinemic euglycemic clamp. C, upper-body However, since our obese volunteers were comparable
obese; B, lower-body obese; , lean. in terms of the absolute amount of visceral fat, this effect
cannot be attributed to visceral fat alone.
Some limitations to the study should be acknowl-
Lindeman et al. observed a positive relationship between edged. First, the measurement of visceral fat by both
visceral fat area and ghrelin levels in 11 obese women CT and DXA was laborious and prevented us from
(assessed by MRI) (10), whereas Katsuki et al. found a investigating a large cohort. That the LBO group did
negative correlation between visceral fat area (assessed not differ from the UBO or lean in terms of the
by CT) and ghrelin in 18 patients with type 2 diabetes suppressive effect of insulin on ghrelin levels may
(3). Both studies assessed visceral fat using a 2D therefore be a statistical type 2 error. Second, we used
technique that does not allow precise quantification of an in-house ghrelin assay that measures the total
fat mass. Other studies have assessed the impact of ghrelin as opposed to other assays specifically aimed at
abdominal fat on ghrelin levels by measuring waist determining the bioactive acylated ghrelin. Total
circumference (WC) (16), DXA (2) or by a combination ghrelin levels were assumed to serve as a marker of
of WC and DXA (9). Although these techniques are bioactive ghrelin because total and acylated ghrelin
useful when studying large cohorts and expose trial levels usually change in parallel (22). Whether upper-
subjects to minimal radiation, they do not accurately body obesity impacts the ratio between acylated and
assess the amount of visceral fat. To quantify visceral des-acylated ghrelin remains to be elucidated. Third,
fat mass it is necessary to use CT in combination with due to the cross-sectional nature of this investigation,
DXA. Using that technique, Purnell et al. found no the exact nature of the relationship between upper-
significant correlation between the visceral fat mass and body adiposity and ghrelin concentrations can
ghrelin levels. However, both men and women were obviously only be speculated upon. Whether selective
included in that study, which may have caused loss of visceral fat induces a more potent change in
interference by gender-related differences in ghrelin ghrelin concentrations than loss of s.c. fat could be an
levels and other hormones. interesting field for future studies.
Although ghrelin secreting cells have been demon- In summary, visceral fat mass is a strong predictor of
strated to exist in the mucosa lining the small intestine circulating ghrelin levels. In addition, the suppressive
(17) and therefore may be in close contact with visceral effect of insulin on ghrelin levels is attenuated in UBO
adipocytes, only w30% of total ghrelin is estimated to subjects but not in LBO subjects when compared with
be secreted from these cells (18). It is therefore more lean controls. These findings underscore the unfavor-
likely that the impact of visceral fat accumulation on able metabolic and hormonal profile of upper-body
ghrelin levels is indirect, i.e. via altered levels of obesity.
metabolites or hormones. Visceral adipocytes are more
lipolytically active than s.c. (19) thus releasing Declaration of interest
relatively larger quantities of FFAs into the circulation.
We have previously shown that elevated FFA concen- E Sondergaard, L C Gormsen, B Nellemann, E T Vestergaard, and
S Nielsen have no disclosures. J S Christiansen has served as a
trations suppress ghrelin levels (5). Since the mild consultant for Novo Nordisk, Pfizer and Ipsen. There is no conflict of
hyperglycemia observed in UBO subjects also elicits a interest that could be perceived as prejudicing the impartiality of the
hormonal response in the form of increased levels of research reported.
insulin, another known inhibitor of gastric ghrelin
secretion (20), we tested whether including insulin in Funding
the regression analysis would cancel out the effect of
visceral fat on ghrelin levels. That this was not the case The study was supported by The Danish Medical Research Council,
The Novo Nordisk Foundation, and The Danish Diabetes Foundation.
strengthens the notion that visceral adipocytes may This work was supported by grants from the Danish Medical Research
secrete ghrelin inhibiting substances. Potential candi- Council, the Novo Nordic Foundation, and the Danish Diabetes
dates include TNFa (21) and leptin (20). Foundation (to S Nielsen).
www.eje-online.org
Downloaded from Bioscientifica.com at 11/28/2020 10:09:26AM
via free accessEUROPEAN JOURNAL OF ENDOCRINOLOGY (2009) 160 Visceral fat mass and ghrelin levels 379
Acknowledgements 12 Jensen MD, Kanaley JA, Reed JE & Sheedy PF. Measurement of
abdominal and visceral fat with computed tomography and dual-
The authors acknowledge the excellent technical assistance of Ms energy X-ray absorptiometry. American Journal of Clinical Nutrition
Elin Carstensen, Ms Lene Ring, Ms Lone Svendsen, and Ms Susanne 1995 61 274–278.
Sørensen. 13 English PJ, Ghatei MA, Malik IA, Bloom SR & Wilding JPH. Food
fails to suppress ghrelin levels in obese humans. Journal of Clinical
Endocrinology and Metabolism 2002 87 2984.
14 Choi SC. Tests of equality of dependent correlation coefficients.
Biometrika 1977 64 645–647.
References 15 Beaumont NJ, Skinner VO, Tan TMM, Ramesh BS, Byrne DJ,
MacColl GS, Keen JN, Bouloux PM, Mikhailidis DP, Bruckdorfer KR,
1 Wren AM, Seal LJ, Cohen MA, Brynes AE, Frost GS, Murphy KG, Vanderpump MP & Srai KS. Ghrelin can bind to a species of
Dhillo WS, Ghatei MA & Bloom SR. Ghrelin enhances appetite and high density lipoprotein associated with paraoxonase. Journal of
increases food intake in humans. Journal of Clinical Endocrinology Biological Chemistry 2003 278 8877–8880.
and Metabolism 2001 86 5992. 16 Schutte AE, Huisman HW, Schutte R, van Rooyen JM, Malan L &
2 Tschop M, Weyer C, Tataranni PA, Devanarayan V, Ravussin E & Malan NT. Aging influences the level and functions of fasting
Heiman ML. Circulating ghrelin levels are decreased in human plasma ghrelin levels: the POWIRS-study. Regulatory Peptides 2007
obesity. Diabetes 2001 50 707–709. 139 65–71.
3 Katsuki A, Urakawa H, Gabazza EC, Murashima S, Nakatani K, 17 Krsek M, Rosicka M, Haluzik M, Svobodova J, Kotrlikova E,
Togashi K, Yano Y, Adachi Y & Sumida Y. Circulating levels of Justova V, Lacinova Z & Jarkovska Z. Plasma ghrelin levels in
active ghrelin is associated with abdominal adiposity, hyperinsu- patients with short bowel syndrome. Endocrine Research 2002 28
linemia and insulin resistance in patients with type 2 diabetes 27–33.
mellitus. European Journal of Endocrinology 2004 151 573–577. 18 Ariyasu H, Takaya K, Tagami T, Ogawa Y, Hosoda K, Akamizu T,
4 Flanagan DE, Evans ML, Monsod TP, Rife F, Heptulla RA, Suda M, Koh T, Natsui K, Toyooka S, Shirakami G, Usui T,
Tamborlane WV & Sherwin RS. The influence of insulin on Shimatsu A, Doi K, Hosoda H, Kojima M, Kangawa K & Nakao K.
circulating ghrelin. American Journal of Physiology. Endocrinology Stomach is a major source of circulating ghrelin, and feeding state
and Metabolism 2003 284 E313–E316. determines plasma ghrelin-like immunoreactivity levels in
5 Gormsen LC, Nielsen C, Gjedsted J, Gjedde S, Vestergaard ET, humans. Journal of Clinical Endocrinology and Metabolism 2001
Christiansen JS, Jorgensen JO & Moller N. Effects of free fatty acids, 86 4753–4758.
growth hormone and growth hormone receptor blockade on 19 Marin P, Andersson B, Ottosson M, Olbe L, Chowdhury B, Kvist H,
serum ghrelin levels in humans. Clinical Endocrinology 2007 66 Holm G, Sjostrom L & Bjorntorp P. The morphology and
641–645. metabolism of intraabdominal adipose tissue in men. Metabolism
6 Nielsen S & Jensen MD. Obesity and cardiovascular disease: is body 1992 41 1242–1248.
structure a factor? Current Opinion in Lipidology 1997 8 200–204. 20 Kamegai J, Tamura H, Shimizu T, Ishii S, Sugihara H & Oikawa S.
7 Yildiz BO, Suchard MA, Wong ML, McCann SM & Licinio J. Effects of insulin, leptin, and glucagon on ghrelin secretion
Alterations in the dynamics of circulating ghrelin, adiponectin, from isolated perfused rat stomach. Regulatory Peptides 2004 119
and leptin in human obesity. PNAS 2004 101 10434–10439. 77–81.
8 Jensen MD, Haymond MW, Rizza RA, Cryer PE & Miles JM. 21 St-Pierre DH, Bastard JP, Coderre L, Brochu M, Karelis AD,
Influence of body fat distribution on free fatty acid metabolism in Lavoie ME, Malita F, Fontaine J, Mignault D, Cianflone K,
obesity. Journal of Clinical Investigation 1989 83 1168–1173. Imbeault P, Doucet E & Rabasa-Lhoret R. Association of acylated
9 Ritland L, Alekel D, Matvienko O, Hanson K, Stewart J, Hanson L, ghrelin profiles with chronic inflammatory markers in overweight
Reddy M, Van Loan M & Genschel U. Centrally-located body fat is and obese postmenopausal women: a MONET study. European
related to appetitive hormones in healthy postmenopausal women. Journal of Endocrinology 2007 157 419–426.
European Journal of Endocrinology 2008 158 889–897. 22 Barazzoni R, Zanetti M, Ferreira C, Vinci P, Pirulli A, Mucci M,
10 Lindeman JH, Pijl H, Van Dielen FM, Lentjes EG, Van LC & Dore F, Fonda M, Ciocchi B, Cattin L & Guarnieri G. Relationships
Kooistra T. Ghrelin and the hyposomatotropism of obesity. Obesity between desacylated and acylated ghrelin and insulin sensitivity
Research 2002 10 1161–1166. in the metabolic syndrome. Journal of Clinical Endocrinology and
11 Purnell JQ, Weigle DS, Breen P & Cummings DE. Ghrelin levels Metabolism 2007 92 3935–3940.
correlate with insulin levels, insulin resistance, and high-density
lipoprotein cholesterol, but not with gender, menopausal status, or
cortisol levels in humans. Journal of Clinical Endocrinology and Received 8 December 2008
Metabolism 2003 88 5747–5752. Accepted 23 December 2008
www.eje-online.org
Downloaded from Bioscientifica.com at 11/28/2020 10:09:26AM
via free accessYou can also read
