- Email: [email protected]
Different thermic effects of leptin in adolescent females with varying body fat content
Clinical Nutrition 29 (2010) 639e645
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original Article
Different thermic effects of leptin in adolescent females with varying body fat content Verena K. Haas a, b, *, Kevin J. Gaskin c, Michael R. Kohn a, Simon D. Clarke d, Manfred J. Müller e a
Department of Adolescent Medicine, The Children’s Hospital at Westmead, University of Sydney, Australia Franz-Volhard Clinical Research Center, ECRC, Charité University Hospital, Berlin, Germany c The James Fairfax Institute of Pediatric Nutrition, The Children’s Hospital at Westmead, University of Sydney, Australia d The Centre of Research into Adolescent’s Health, Westmead Hospital, University of Sydney, Australia e Institut für Humanernährung und Lebensmittelkunde der Christian-Albrechts-Universität zu Kiel, Germany b
a r t i c l e i n f o
s u m m a r y
Article history: Received 19 October 2009 Accepted 31 March 2010
Background & aims: Investigating the effect of leptin on energy expenditure in undernutrition might lead to a better understanding of the role of leptin in regulating body weight in humans. Methods: 73 underweight female adolescents with anorexia nervosa (AN) were compared with 23 healthy normal weight (nwC), and 9 overweight girls (OW); 37 AN were followed during 7 months of weight recovery. Resting energy expenditure (REE, by indirect calorimetry), body composition (fat mass, FM; lean tissue mass, LTM; by Dual-Energy X-Ray Absorptiometry) and plasma hormones of leptin and 3,5,30 -Triiodothyronine (T3) were measured. Results: In underweight, leptin, T3 and REE adjusted for lean tissue mass (REELTM) were decreased; in OW, FM and leptin were increased at unchanged T3 and REELTM. There was a significant positive relation between FM and leptin at low and normal (AN, r2 ¼ 0.26; nwC, r2 ¼ 0.51, p < 0.001), but not at high adiposity. Leptin and REELTM were positively associated in underweight (r2 ¼ 0.14, p ¼ 0.001) but not in normal or overweight subjects. T3 was linearly related to REELTM over the whole range of adiposity (r2 ¼ 0.42, p < 0.001). With weight gain in AN (5.0 3.5 kg) the relationship between leptin and REELTM changed toward the conditions seen in normal weight controls. Conclusions: At low adiposity the interrelated fall of leptin and REE reflect an adaptive mechanism to preserve body weight. High leptin production associated with excessive adiposity was without effect on metabolic adaptation. Ó 2010 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Resting energy expenditure Leptin T3 Body composition Underweight Anorexia nervosa
1. Introduction Since its discovery in 1994,1 the role of leptin in the regulation of body weight by affecting energy homeostasis has been extensively discussed.1e16 Many authors agree that leptin functions as an adipocyte-derived signal to convey information about energy status to the brain, which then synchronizes a range of physiological responses, such as the adjustment of REE, to altering energy scenarios. In mouse experiments it has been shown that leptin injection increased whole-body energy expenditure and directly stimulated thermogenesis in skeletal muscle.3,6,11 However when
* Corresponding author. Dr. oec. troph. Verena Haas, Charité University Hospital, Franz-Volhard Clinical Research Center, ECRC, Lindenberger Weg 80, 13125 Berlin, Germany. Tel.: þ49 30 450 540 324; fax: þ49 30 450 540 920. E-mail addresses: [email protected] (V.K. Haas), [email protected] (K.J. Gaskin), [email protected] (M.R. Kohn), [email protected]. au (S.D. Clarke), [email protected] (M.J. Müller).
compared with animal data a thermic effect of leptin remains equivocal in humans. Cross-sectional studies have not been able to show an association between plasma leptin concentrations and REE in normal weight and obese subjects.9,13,15 Although plasma leptin concentrations decreased with weight loss and fasting in lean and obese individuals, the magnitude of change in leptin did not correlate with the reduction of REE in a linear relationship in overweight and obese subjects.2,7,10,14,16 Finally there was no substantial impairment of REE in a child with congenital leptin deficiency, and weight loss following leptin treatment in this child was largely attributed not to changes in REE but to decreased food intake.4 These contradictory findings make it difficult to designate leptin as a key target in obesity treatment, and a better understanding of the role of leptin in the regulation of body weight is required. Adipocytokines have been studied with increasing intensity as a result of the emergence of obesity as a serious public health problem. However in his recent review,8 Leibel argues that from an evolutionary perspective, the need to suppress food intake in response to adiposity has not been
0261-5614/$ e see front matter Ó 2010 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. doi:10.1016/j.clnu.2010.03.013
640
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
particularly important, and therefore the physiological “defence” of body fat is weaker against upward deflections than against those bringing body fat below a threshold for minimum allowable FM. Prentice et al. argue along the same lines, namely that attempts to understand leptin’s function through the studies of overnourished people that now dominate the literature may be misleading, and that further investigations are required to elucidate its possible modulating role in the state of energy deficit.12 While data about the role of leptin in adaptive thermogenesis is still contradictory, it has long been accepted that thyroid hormone is an important determinant of overall energy expenditure and the basal metabolic rate.17 Thyroid hormone concentrations have been described to be related to weight status, with their concentrations being decreased in underweight, and increased in overweight conditions.18 The endocrinology of the leptin axis linked with energy metabolism might be different and not be proportional over a wide range of its concentrations. Considering there was no relation between REE and leptin in normal weight or obese people, high levels of leptin might be not responded to. However low concentrations could lead to reduced REE in an effort to conserve and restore body fat stores,8 therefore the focus of leptin research should also be on underweight or “underfat” populations. Anorexia nervosa (AN) is an illness characterized by low body weight and FM, and reduced leptin, REE as well as T3.5 In AN as well as in healthy controls, T3 and REE adjusted for FFM were linearly related.19 During shortterm weight gain over a period of 12 weeks FM, leptin and REE increased in AN and leptin secretion was associated with weight gain.5 The aim of the current study was to investigate the effect of leptin on REE in a study cohort of a broad range of FM, and to compare the relation between leptin and REE with that of T3. Therefore body composition and metabolic and endocrine function was assessed in anorexic as well as healthy adolescent females. In addition a subgroup of AN patients was followed longitudinally over 7 months to assess whether body composition, REE and endocrine dysfunction normalized with nutritional intervention.
2. Subjects and methods 2.1. Study populations A group of 73 adolescent female patients with anorexia nervosa (AN; mean age 14.9 1.6 years; range 11e19) was recruited from two inpatient hospital units associated with the Departments of Adolescent Medicine, The Children’s Hospital at Westmead and Westmead Hospital in Sydney/Australia. All patients were amenorrheic (12 with primary amenorrhea), diagnosed with AN according to DSM-IV (American Psychiatric Organization, 199420), and admitted to an inpatient Eating Disorder Program. Baseline testing was done after medical stabilization. Mean Tanner stage as assessed by the treating physician was 4. Forty AN returned for follow-up approximately 7 months after baseline testing and 37 of these had gained weight. Three AN who had lost weight between discharge and follow-up were excluded from further analysis. The 37 follow-up AN had had an average length of hospital stay of 43 23 days. Average therapeutically induced weight gain during hospital stay was 4 2 kg, with the rest of the weight gain occurring after discharge. A healthy sex- and age matched control group was recruited from a local high school (n ¼ 32; age 14.6 0.6 years). Median Tanner stage in the control group as self-reported by the girls with the use of questionnaires21 was 4. Three of the girls in the control group had not reached menarche at the time of testing, the remaining 29 girls were tested in the follicular phase of their menstrual cycle (i.e. day 1e10 of their menstrual cycle). There was no follow-up testing for the control group.
2.2. Body composition measurements For each individual all measurements were performed on the same morning, after an overnight fast and voiding. Height (0.1 cm) was measured with a stadiometer (Holtain Ltd, Crymych, United Kingdom) and body weight (0.1 kg) in light clothing using electronic scales (AND FW-150K, Tokyo, Japan). BMI was calculated as body weight/height2 (in kg/m2). Whole-body composition (FM, LTM expressed in kg and as a percentage of body weight) was assessed by Dual-Energy X-Ray Absorptiometry (DXA) using a Lunar Prodigy whole-body scanner (Lunar Corp, Madison, WI, USA) in conjunction with EnCORE Software Version 8.10. 2.3. Hormone assays Blood samples were drawn between 07:00 and 09:00 am after the subjects had fasted overnight. Samples for analysis of thyroid hormones were sent to the hospital laboratory for immediate analysis. 3,5,30 -Triiodothyronine (free T3) concentrations were quantified using a microparticle Enzyme Immuno Assay (MEIA, AxSYM, U.S.A). Samples for leptin analysis were spun for separation into plasma and serum, and kept frozen at 80 C, until batch analysis at the end of the study. Plasma leptin concentrations were measured by Immunoassay (Lincoplex Kit, Linco Research, St. Charles, MO, U.S.A). 2.4. Metabolic measurements Metabolic testing took place in the early morning after 8e12 h of fasting, in a thermoneutral room. REE and Respiratory quotient (RQ) were assessed by indirect calorimetry (Deltatrac TMII, Datex Instrumentarium Corp., Helsinki/Finland). Before each measurement, the machine was calibrated with room air and a 95%O2/5% CO2 gas mixture. Participants were taken to the metabolic room by wheelchair, then lying down, and a canopy connected to the metabolic cart was placed over the head. Oxygen consumption and carbon dioxide production were monitored for 30 min at 1-min intervals. Results taken in the first 10 min were considered as adaptation phase and discarded. REE and RQ were calculated according to Weir et al.22 REE was adjusted for LTM using a regression analysis (¼ REELTM). REE was also predicted according to Schofield23 and Müller et al.24 2.5. Statistical analysis Statistical analysis was carried out by using SPSS (version 16.0; SPSS Inc. Cary. NC). The data shown in Tables 1e3 are presented as means SDs at a significance level of P < 0.05. Normal distribution was tested with the KolmogoroveSmirnov-Test. Between-group comparison was carried out with a one-way ANOVA and a 2-tailed post-hoc Dunnett t-test, comparing both AN and OW with normal weight controls. When AN at baseline were compared with AN at follow-up, a paired samples t-test was used. To account for the effect of interindividual differences in LTM on variance of REE (¼ REELTM), REE was adjusted in a linear regression analysis with REE as dependent and LTM as independent variable. To calculate individual REELTM, individual regression residual was subtracted from group mean REE. To describe the relationship between leptin and either FM or REELTM, Pearson correlation coefficients were calculated using SPSS as well as best fitting mathematical models using Microsoft Excel Office Standard Edition 2003. Stepwise liner regression analysis with REELTM as dependent, and group affiliation (AN, nwC or OW) and leptin or T3 was carried out using SPSS (cases were excluded listwise; p of F-value for inclusion < 0.05, for exclusion > 0.1).
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
641
Table 1 Body composition, metabolic and endocrine function of AN, nwC and OW.
Age (y) Height (cm) Weight (kg) BMI (kg/m2)
AN baseline (n ¼ 73)
nwC (n ¼ 23)
OW (n ¼ 9)
14.9 1.6 163 8 43.0 6.0 16.2 1.4
14.6 0.7 161 8 51.3 9.3 19.6 2.4
14.6 0.3 165 7 77.9 11.8 28.5 2.8
(11e19) (128e176) (22.7e59.0)*** (13.1e19.7)***
(13e16) (142e174) (33.9e71.6) (15.9e23.6)
(14e15) (150e173) (55.0e90.4)*** (24.4e33.2)***
Data are presented as means SD (range); AN, patients with anorexia nervosa; nwC, normal weight controls; OW, overweight subjects; BMI, Body Mass Index. ***, p ¼ 0.000 when comparing AN and OW with nwC using one-way ANOVA and post-hoc Dunnett-t-test.
2.6. Ethics The Ethics Committees at the Children’s Hospital at Westmead and Westmead Hospital approved the study protocol. Prior to participating, all girls and their parents gave informed written consent. Assessment and analysis of data was carried out according to privacy laws. 3. Results 3.1. Descriptive presentation of general characteristics, body composition, and metabolic and endocrine function Within the control group, 9 girls were classified as overweight (OW) according to a BMI cut-off of >23.9 for girls aged 14.5 years as established by Cole et al.25 Table 1 presents the general characteristics of AN at baseline, and of nwC and OW. There was no significant difference in age or height across the study groups, but weight and BMI were significantly lower in AN, and significantly higher in OW when compared with nwC. Body composition assessed with DXA, as well as metabolic and endocrine data of all 73 AN at baseline, and of nwC and OW are presented in Table 2. When compared with nwC, AN had less body fat and unchanged LTM which at decreased weight explained an increased percentage of LTM. REE, REELTM, leptin, and T3 were significantly lower in AN than in nwC at increased RQ. In OW body fat, leptin, LTM, and REE when compared with nwC were increased but there was no difference in RQ, REELTM or T3. 3.2. Association between leptin, fat mass and metabolic rate When combining all subjects to represent the entire range of adiposity present in this study, the best mathematical model to describe the relation between percentage body fat and leptin was
an exponential curve (Fig. 1A). When analyzing the 3 study groups separately, there was a significant and positive relationship between absolute FM and leptin in both AN and nwC (Fig. 1B). The slope of the regression line was steeper in nwC when compared with AN, suggesting that an increase in 1 kg of body fat was associated with a higher increase in leptin in nwC compared to AN. In AN, the variance around the regression line was lower than in nwC (Fig. 1B, AN: SEE 1.37 vs. nwC: 2.90). In OW variability in leptin concentrations was high and the association between FM and leptin was lost (Fig. 1A and B). Fig. 2 shows the associations between REELTM and leptin (Fig. 2A) and T3 (Fig. 2B) in AN, nwC and OW. There was a positive association between leptin and REELTM in AN (r2 ¼ 0.14, p ¼ 0.001) but not in nwC, OW, or nwC þ OW combined. Fig. 2A shows that when data from all subjects were combined, the best fitting mathematical model to describe the relation between leptin and REELTM, was a logarithmic curve (logarithmic function, r2 ¼ 0.43 versus linear function, r2 ¼ 0.33). Contrary to the findings shown in Fig. 1, the slope of the curve is steep at low leptin and flat at normal leptin and suggests that the effect of leptin on REELTM differs at varying leptin concentrations. In a stepwise linear regression with REELTM as dependent, and leptin and group affiliation (AN, nwC, OW) as independent variables, both group affiliation and leptin were significant predictors and together explained 50% of the variance in REELTM (Table 4). When data of all three groups were combined, T3 was positively associated with both leptin and FM (p ¼ 0.001). In a second regression model with REELTM as dependent, and T3 and group affiliation as independent variables, both group affiliation and T3 were significant predictors and together explained 53% of the variance in REELTM (Table 4). Fig. 2B shows that when data from all subjects were combined, there was a significant and positive relationship between T3 and REELTM. However when compared with leptin, the mathematical model was linear, suggesting that the effect of T3 on REELTM was constant over the entire T3 concentration spectrum.
Table 2 Body composition, metabolic and endocrine function of AN, nwC and OW. AN baseline (n ¼ 73) Body composition FMDXA (kg) FMDXA (%) LTMDXA (kg) LTMDXA (%) Metabolism REE (kcal/d) REEpred Schofield (kcal/d) REEpred Müller (kcal/d) RQ REELTM (kcal/d) Hormones Leptin (mg/L) T3 (pmol/L)
6.3 15.0 34.7 81.2
2.8 5.7 4.4 5.9
(1.4e12.9)*** (4.2e26.3)*** (20.5e43.5) (70.4e94.3)***
1037 181 1316 83 1380 101 0.92 0.10 1054 173
(579e1457)*** (986e1499)** (991e1554) (0.70e1.20)*** (613e1515)***
2.0 1.6 (0.1e8.6)*** 3.1 0.8 (1.9e4.9)***
nwC(n ¼ 23) 13.5 26.8 35.2 69.5
5.6 6.9 4.4 6.7
1373 166 1380 109 1428 128 0.82 0.05 1379 128
(5.3e26.3) (14.1e40.0) (25.1e43.6) (55.8e81.2) (1105e1745) (1145e1609) (1151e1660) (0.70e0.93) (1141e1726)
6.6 4.1 (1.7e18.1) 4.2 0.6 (3.0e5.1)
OW (n ¼ 9) 33.6 44.8 40.8 52.7
7.3 4.5 5.2 4.3
1598 216 1620 109 1629 135 0.75 0.02 1458 124
(20.7e40.9)*** (36.9e51.2)*** (31.6e46.6)** (46.1e60.2)*** (1284e1890)** (1145e1609)*** (1339e1757)*** (0.71e0.77) (1228e1599)
17.5 7.6 (5.8e31.9)*** 4.0 0.8 (2.6e5.3)
Data are presented as means SD (range); AN, patients with anorexia nervosa; nwC, normal weight controls; OW, overweight subjects; FM, Fat Mass; DXA, Dual-Energy X-ray Absorptiometry; LTM, Lean Tissue Mass; REE, Resting Energy Expenditure; pred., predicted; RQ, Respiratory Quotient, REELTM, REE adjusted for LTM; *, p < 0.05; ***, p < 0.001 when comparing AN and OW with nwC using one-way ANOVA and post-hoc Dunnett-t-test.
642
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
Table 3 General characteristics, body composition, and metabolic and endocrine function of AN patients on admission and after weight gain. AN baseline (n ¼ 37)
AN follow-up (n ¼ 37)
Age (y) Weight (kg) Height (cm) BMI (kg/m2)
14.6 1.5 42.7 5.9 163 9 16.0 1.3
(11e17) (22.7e50.7) (128e176) (13.4e18.4)
15.1 1.5 47.7 7.1 164 9 17.7 1.8
(11e18)*** (25.5e62.2)*** (129e177)** (14.7e21.2)***
Body composition FMDXA (kg) FMDXA (%) LTMDXA (kg) LTMDXA (%)
6.0 14.5 34.4 81.1
(1.4e11.3) (4.8e25.3) (20.5e42.9) (70.4e91.8)
10.4 22.5 34.8 73.6
(3.6e18.7)*** (9.0e34.6)*** (19.4e43.9) (61.3e85.3)***
Metabolism REE (kcal/d) RQ REELTM (kcal/d) Hormones Leptin (mg/L) T3 (pmol/L)
2.7 5.6 4.4 6.2
4.1 6.8 4.9 6.5
995 173 (579e1301) 0.93 0.10 (0.74e1.20) 1020 168 (613e1406)
1158 188 (747e1582)*** 0.89 0.07 (0.76e1.03)* 1172 178 (851e1556)***
1.6 0.9 (0.1e3.9) 3.1 0.8 (2.0e4.9)
4.0 3.1 (0.8e10.3)*** 3.4 0.7 (2.2e4.7)*
Data are presented as means SD (range); AN, patients with anorexia nervosa; FM, Fat Mass; DXA, Dual-Energy X-ray Absorptiometry; LTM, Lean Tissue Mass; REE, Resting Energy Expenditure; RQ, Respiratory Quotient, REELTM, REE adjusted for LTM; *, p < 0.05; **, p < 0.01; ***, p < 0.001; AN baseline vs. AN at follow-up: paired samples t-test.
3.3. Longitudinal observations in an during weight gain In Table 3, the change of nutritional status as well as metabolic and endocrine function in AN after medium-term weight gain of 7 months is presented. After 7 months of weight gain AN had significantly increased their body weight by 5.0 3.5 (range: 0.4e14.4) kg, and BMI by 1.8 1.3 units (range: 0.1e4.7 kg/m2). During weight gain in AN, FM but not LTM increased significantly when considered in absolute terms. When described as a percentage of body weight, FM increased, but LTM decreased during weight gain in AN. REE as well as REELTM, leptin and T3 also increased, and RQ decreased significantly during weight recovery in AN. However at follow-up, REE as well as REELTM were still significantly lower than in nwC (p ¼ 0.000, data not shown). Fig. 3 shows the relationship between leptin and REELTM in AN before and after 7 months of weight gain. The association between leptin and REELTM before and after weight gain in AN was in accordance with cross-sectional data: in AN after weight gain the slope of the regression line was less steep when compared with AN before weight gain.
A
35
4. Discussion Leptin is a fat cell-derived hormone with multiple functions, including a role in the adaptation of an organism to semistarvation.26 The present study demonstrated an exponential association between leptin and fat mass with a strong and positive relation at low and normal, and a loss of this relation at high degrees of adiposity. The effect of leptin on REELTM was not constant over its whole concentration range and was best described with a logarithmic curve. By including anorexic subjects with drastically reduced body fat we were able to show a significant positive association between leptin and REELTM at low and a lack thereof at normal or high leptin concentration as seen in healthy normal weight and overweight girls. T3 was positively associated with both leptin and fat mass. Opposed to the different thermic effects of leptin with varying adiposity, there was a linear and positive association between T3 and REELTM over the whole range of adiposity. Weight gain of AN patients resulted in a shifting of the relationship between leptin and REELTM in accordance with the cross-sectional data. 4.1. Association between leptin and fat mass Many studies have described rising leptin concentrations with increasing adiposity, suggesting a linear relationship. However some studies investigating subjects with a wide BMI range have supported an exponential model to describe the relationship between leptin and percent body fat.27,28 In addition to the nonlinear relation between fat mass and leptin Luke et al. have described an increasing variance and loss of relation with increasing adiposity.27 Our findings examining a cohort including a group of patients with anorexia nervosa as well as subjects with normal, and a small group with excessive fat mass, are consistent with these results (Fig. 1A), showing an exponential relationship between fat mass and leptin. While the biological plausibility of the resulting mathematical exponential curve is unclear, the data suggest that leptin dynamics are non-linear and differ with varying amounts of body fat. Leptin appeared to reflect the body’s energy stores in a sensitive dose response manner at low and normal body fat. In contrast in the girls with excessive body fat, there was a wide variation in leptin levels and the association between leptin and FM was lost. Prentice et al. point out that when evolution has pressed a single molecular species such as leptin into a number of different functions in several organs, there must be mechanisms that permit
B
30
30
all subjects, exponential y = 0.45e 0,0847x R2 = 0.74***
25
35
nwC y = 0.53x - 0.49
25 leptin (µg/L)
leptin (µg/L)
R2 = 0.51*** 20 all subjects, linear y = 0.42x - 4,09 R2 = 0.68***
15
20 AN y = 0.29x + 0.23
15
10
10
5
5
0
0
R2 = 0.26***
AN nw C
0
10
20
30
40
50
60
OW
0
Body fat (DXA, %)
10
20
30
Body fat (DXA, kg) Fig. 1. Relation between body fat and leptin.
40
50
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
A 2000
B 2000
AN, linear y = 41x + 972 R2 = 0.14**
1800
1800
1600 1400
1400 all subjects, linear y = 24x + 1056 R2 = 0.33***
1000 800
REE LTM (kcal/d)
REE LTM (kcal/d)
all subjects, linear y = 165x + 596 R2 = 0.42***
1600
1200
all subjects, logarythm ic y = 140Ln(x) + 1031 R2 = 0.43***
600
643
1200 1000 800 600
400
400
200
200
AN nw C OW
0
0 0
5
10
15
20
25
30
0,0
35
1,0
2,0
leptin (µg/L)
3,0
4,0
5,0
6,0
T 3 (pmol/L)
Fig. 2. Relationship between leptin and energy expenditure.
4.2. The role of leptin and T3 in adaptive thermogenesis Leptin is thought to regulate body weight through hypothalamic effects on satiety and energy expenditure.29 However human data has rarely supported a correlation between fasting leptin concentrations and REE. Most of these studies mainly include normal weight, overweight and obese subjects. Even though leptin may be particularly important in signalling the transition between energy surplus and deficit,12 the thermogenic effect of leptin on REE has rarely been studied in underweight populations. Studying the role of leptin in modulating adaptive thermogenesis specifically in undernourished people could lead to a better understanding and help to resolve some of the existing controversies. The significant effect of group affiliation (being anorexic, normal weight, or overweight) on the variance in REELTM shown in our regression models (Table 4) points towards metabolic adaptation to different nutritional states beyond the effect produced by changes in lean tissue mass. In these same regression models, both leptin and T3 were independent predictors of REELTM and are thus likely to play a role in adjusting energy homeostasis. In malnourished patients with anorexia nervosa (AN) plasma leptin is reduced5,30e32 and a positive association between leptin and REE has been documented earlier.5,32 Our data showing a steep Table 4 Multiple linear regression analysis of resting energy expenditure adjusted for lean tissue mass (REELTM) with group affiliation (anorexic, normal weight, overweight) and plasma hormone concentrations (model 1: leptin; model 2: T3). Independent variables Model 1 Group Leptin Model 2 Group T3
b-coefficient (T-value)
P
and positive association between leptin and REELTM in undernourished subjects (Fig. 2A) are in accordance with these previous findings and suggest that low leptin levels in malnourished AN patients are interrelated with low energy expenditure and could reflect an adaptive mechanism to preserve body weight as others have alluded to.3,8,12 However after including normal- and overweight subjects, our present findings show that the effect of leptin on REELTM was not constant over its entire concentration spectrum and suggest a loss of association within the normal or increased range of FM (Fig. 2A). These different thermic effects of leptin warrant further discussion. Underweight is characterized by low FM, LTM and REE. Lean tissue mass is a major determinant of REE and part of the fall in REE with weight loss is thus explained by LTM losses.33,24 However by reanalyzing data from the Minnesota experiment of semistarvation, Dulloo et al. have suggested that in addition to the relation between the depletion in LTM and the fall in REE, there is an additional feedback control system linking the state of depletion of FM to mechanisms suppressing thermogenesis.34
2000 1800
AN follow -up y = 27x + 1062 R2 = 0.23**
1600 1400 REE LTM (kcal)
variable sensitivity across the physiologic range of circulating leptin.12 From a teleological point it can be argued that the major role of leptin is to signal energy deficit rather than surfeit, and marginal nutrition or undernutrition were the states that have mostly driven metabolic selection processes during the past millennia.12 The role of leptin in regulating energy metabolism might therefore be most obvious in subjects with low body fat.
1200 1000 800 600 AN basline y = 57x + 923 R2 = 0.10 (p=0.06)
400 200
AN baseline
2
Adjusted r ¼ 0.501 0.489 (6.219) 0.339 (4.315) Adjusted r2 ¼ 0.533 0.425 (5.340) 0.414 (5.199)
0.000 0.000
AN follow -up
0 0
5
10
15
20
leptin ( µ g/L) 0.000 0.000
Fig. 3. Relationship between leptin and energy expenditure in AN patients on admission and after weight gain.
644
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
In accordance with this hypothesis it has been shown that FM is an independent but relatively small predictor of REE,35e38 which is most likely explained by the endocrine activity of adipose tissue. In the current study a decrease in leptin (or body fat e we are not able to distinguish between these two parameters with our data) was associated with a moderate drop in REELTM in control subjects but with a large drop in REELTM in anorexic subjects. It can therefore be hypothesized that starvation-induced falls in leptin or FM cause metabolic adaptation to starvation by severely downregulating REE only in subjects falling short of a critically low threshold range of FM or leptin. Fig. 2A shows that nwC with low leptin concentrations (i.e., below 5 mg/L) were located close to the steep regression line described for AN. However AN subjects with relatively high leptin concentrations (i.e., above 5 mg/L) were located close to the flatter regression line described for healthy controls. Therefore the relationship between leptin and REELTM did not seem strictly dependent on subject group (anorexic patient or healthy control), but rather linked with the amount of circulating leptin or FM per se. In contrast to underweight patients with low FM, subjects with excessive body fat demonstrated a 3-fold increase in serum leptin but a plateauing of REE (Fig. 2A). This unresponsiveness of thermogenesis to high leptin in the overweight or obese could be the equivalent to a state that has been described as “leptin resistance” by several investigators. Whether this specific metabolic situation reflects a saturation of leptin receptors or end organ resistance, alterations either in the transport of leptin across the blood brain barrier and/or cellular leptin receptor signalling or whether there are other underlying reasons, is still unknown and needs further study. The sample size in the OW group was small and this limits our findings with respect to the lack of adaptive thermogenesis in the overweight state. To confirm our hypothesis, and to pinpoint the threshold for the loss of linearity between fat mass and leptin, further studies with higher subject numbers are required. While leptin has been suggested as a promising link between the regulation of energy balance and thyroid hormones,39 the exact nature of this relationship is still not clear. Our study shows that firstly, there was a positive association between leptin and T3. Second, when compared with leptin, plasma T3 concentrations had a linear and positive association with REELTM over their whole concentration range (Fig. 2B). These findings are in accordance with previous data in healthy subjects as well as in patients with AN19,39 and in line with the assumptions of Flier et al.40 suggesting that a decrease in FM and concomitantly leptin concentrations sensed by the brain could lead to a lowering of REE brought about by decreasing serum T3 concentrations. 4.3. Observations in anorexic patients during weight gain After weight gain, the relationship between leptin and REELTM changed in direction toward the conditions seen in healthy and normal weight controls (Fig. 3). Once acute starvation of AN was over, thermic adaptation to the new and positive energy balance was achieved. Starvation metabolism and the depression of REE were slowly released, which has been shown before in AN during and following weight gain.5,32 While in the current and also in our previous longitudinal study5 metabolic function had not completely normalized in AN after weight gain, this can be explained by the fact that at follow-up testing, these AN patients had still been underweight. Another study examining weight-recovered AN patients however suggests a full reversibility of this adaptive mechanism.32 4.4. Implications for clinical nutrition practice As thermogenesis directly determines the energy needs, knowledge about adaptive mechanisms or the lack thereof is
crucial when deciding on nutritional strategies in clinical practice. More specifically, when relying on commonly used prediction equations, energy needs will be considerably overestimated in undernourished and metabolically compromised subjects such as patients with anorexia nervosa (Table 2). The discrepancy between measured and predicted energy expenditure originates in the inability of the algorithms to account for endocrine mechanisms leading to adaptive thermogenesis as described in the current study. 4.5. Study limitations There were several limitations in our study. First of all the sample size in the overweight group was small and therefore our findings require further confirmation with higher subject numbers. Second, blood sampling was not carried out on the first day after the AN patients were admitted to hospital as recommended previously,26 but within the first week of admission. This was due to the fact that before testing, patients needed to be medically stabilized, and patient and parental consent had to be obtained. The extent of refeeding before blood sampling influences leptin levels in AN. However mean leptin concentrations assessed in the current study are well within the low range provided by earlier studies on AN patients who had been tested shortly after hospital admission.26 5. Conclusion After examining a cohort with a wide range of body fat, our findings suggest that leptin closely reflected fat mass at low and normal but not at high adiposity. The association between leptin and resting energy expenditure also varied with the degree of adiposity. The sharp decrease of both leptin and REE at low adiposity suggests that low leptin concentrations could be involved in reducing thermogenesis in order to conserve energy in underweight. On the contrary an increase in fat mass or leptin above normal range associated with high adiposity was without effect in terms of metabolic adaptation. Conflict of interest We have no conflict of interest. Authorship statement Study design and critical review of manuscript: Kevin Gaskin, Manfred James Müller. Practical performance: Verena Haas, Michael Kohn, Simon Clarke. Data analysis and preparation of manuscript: Verena Haas, Manfred James Müller. Acknowledgements The present study was supported by grants from the Centre of Research into Adolescent’s Health, Westmead Hospital, Sydney, Australia; the James Fairfax Institute of Pediatric Nutrition, The Children’s Hospital at Westmead, Sydney, Australia; and the DAAD, German Academic Exchange Service, Bonn, Germany. We thank Madeleine Thompson and July Briody for assistance with the body composition measurements. VKH collected and analyzed the data and wrote the manuscript. MK and SDC collected the data and provided advice. KJG and MJM designed the study and provided advice. None of the authors had a conflict of interest.
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
References 1. Zhang YY, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homolog. Nature 1994; 372:425e32. 2. Ahima RS. Role of leptin in the neuroendocrine response to fasting. Nature 1996;382:250e2. 3. Dulloo AG, Stock MJ, Solinas G, Boss O, Montani JP, Seydoux J. Leptin directly stimulates thermogenesis in skeletal muscle. Febs Lett 2002;515:109e13. 4. Farooqi IS, Jebb SA, Langmack G, Lawrence E, Cheetham CH, Prentice AM, et al. Effects of recombinant leptin therapy in a child with congenital leptin deficiency. N Engl J Med 1999;341:879e84. 5. Haas V, Onur S, Paul T, Nutzinger DO, Bosy-Westphal A, Hauer M, et al. Leptin and body weight regulation in patients with anorexia nervosa before and during weight recovery. Am J Clin Nutr 2005;81:889e96. 6. Hwa JJ, Ghibaudi L, Compton D, Fawzi AB, Strader CD. Intracerebroventricular injection of leptin increases thermogenesis and mobilizes fat metabolism in ob/ ob mice. Horm Metab Res 1996;28:659e63. 7. Klein S, Horowitz JF, Landt M, Goodrick SJ, Mohamed-Ali V, Coppack SW. Leptin production during early starvation in lean and obese women. Am J Physiol Endocrinol Metab 2000;278:E280eE284. 8. Leibel RL. Molecular physiology of weight regulation in mice and humans. Int J Obes 2008;32(Suppl. 7):S98eS108. 9. Martin LJ, Jones PJ, Considine RV, Su W, Boyd NF, Caro JF. Serum leptin levels and energy expenditure in normal weight women. Can J Physiol Pharmacol 1998;76:237e41. 10. Ostlund RE, Yang JW, Klein S, Gingerich R. Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab 1996;81:3909e13. 11. Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D, Boone T, et al. Effects of the obese gene-Product on body-weight regulation in Ob/Ob mice. Science 1995;269:540e3. 12. Prentice AM, Moore SE, Collinson AC, O’Connell MA. Leptin and undernutrition. Nutr Rev 2002;60:S56eS67. 13. Roberts SB, Nicholson M, Staten Dallal GE, Sawaya AL, Heyman MB, Fuss P, et al. Relationship between circulating leptin and energy expenditure in adult men and women aged 18 years to 81 years. Obes Res 1997;5:459e63. 14. Rosenbaum M, Nicolson M, Hirsch J, Murphy E, Chu F, Leibel RL. Effects of weight change on plasma leptin concentrations and energy expenditure. J Clin Endocrinol Metab 1997;82:3647e54. 15. Soares MJ, Piers LS, O’Dea K, Collier GR. Plasma leptin concentrations, basal metabolic rates and respiratory quotients in young and older adults. Int J Obes Relat Metab Disord 2000;24:1592e9. 16. Weigle DS, Duell PB, Connor WE, Steiner RA, Soules MR, Kuijper JL. Effect of fasting, refeeding, and dietary fat restriction on plasma leptin levels. J Clin Endocrinol Metab 1997;82:561e5. 17. Kim B. Thyroid hormone as a determinant of energy expenditure and the basal metabolic rate. Thyroid 2008;18(2):141e4. 18. Reinehr T, Isa A, de Sousa G, Dieffenbach R, Andler W. Thyroid hormones and their relation to weight status. Horm Res 2008;70:51e7. 19. Onur S, Haas V, Bosy-Westphal A, Hauer M, Paul T, Nutzinger D, et al. L-triiodothyronine is a major determinant of resting energy expenditure in underweight patients with anorexia nervosa and during weight gain. Eur J Endocrinol 2005;152:179e84. 20. Diagnostic and statistical manual of mental disorders (DSM-IV) 1994. Washington, DC: American Psychiatric Association.; 2000.
645
21. Duke PM, Litt IF, Gross RT. Adolescents self-assessment of sexual-maturation. Pediatrics 1980;66:918e20. 22. Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949;109:1e9. 23. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr 1985;39c(1s):5e42. 24. Müller MJ, Bosy-Westphal A, Klaus S, Kreymann G, Lührmann PM, NeuhäuserBerthold M, et al. World Health Organization equations have shortcomings for predicting resting energy expenditure in persons from a modern, affluent population: generation of a new reference standard from a retrospective analysis of a German database of resting energy expenditure. Am J Clin Nutr 2004;80:1379e90. 25. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320:1240e3. 26. Hebebrand J, Muller TD, Holtkamp K, Herpertz-Dahlmann B. The role of leptin in anorexia nervosa: clinical implications. Mol Psychiatry 2007;12:23e35. 27. Luke AH, Rotimi CN, Cooper RS, Long AE, Forrester TE, Wilks R, et al. Leptin and body composition of Nigerians, Jamaicans, and US blacks. Am J Clin Nutr 1998; 67:391e6. 28. Kennedy A, Gettys TW, Watson P, Wallace P, Ganaway E, Pan Q, et al. The metabolic significance of leptin in humans: gender-based differences in relationship to adiposity, insulin sensitivity, and energy expenditure. J Clin Endocrinol Metab 1997;82:1293e300. 29. Clement K, Vaisse C, Lahlou N, Cabrol S, Pelloux V, Cassuto D, et al. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature 1998;392:398e401. 30. Grinspoon S, Gulick T, Askari H, Landt M, Lee K, Anderson E, et al. Serum leptin levels in women with anorexia nervosa. J Clin Endocrinol Metab 1996;81: 3861e3. 31. Hebebrand J, Blum WF, Barth N, Coners H, Englaro P, Juul A, et al. Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration. Mol Psychiatry 1997;2:330e4. 32. Polito A, Fabbri A, Ferro-Luzzi A, Cuzzolaro M, Censi L, Ciarapica D, et al. Basal metabolic rate in anorexia nervosa: relation to body composition and leptin concentrations. Am J Clin Nutr 2000;71:1495e502. 33. Keyes A, Brozek J, Henschel A, Mickelsen O, Taylor HL. The biology of human starvation. Minneapolis: University of Minnesota Press; 1950. 34. Dulloo AG, Jacquet J. Adaptive reduction in basal metabolic rate in response to food deprivation in humans: a role for feedback signals from fat stores. Am J Clin Nutr 1998;68:599e606. 35. Johnstone AM, Murison SD, Duncan JS, Rance KA, Speakman JR. Factors influencing variation in basal metabolic rate include fat-free mass, fat mass, age, and circulating thyroxine but not sex, circulating leptin, or triiodothyronine. Am J Clin Nutr 2005;82:941e8. 36. Blaak EE, Hul G, Verdich C, Stich V, Martinez JA, Petersen M, et al. Impaired fatinduced thermogenesis in obese subjects: the NUGENOB study. Obesity 2007;15:653e63. 37. Elia M. Energy expenditure in the whole body. In: Kinney JM, Tucker HN, editors. Energy metabolism: tissue determinants and cellular corollaries. New York: Raven Press; 1992. p. 19e59. 38. Nelson KM, Weinsier RL, Long CL, Schutz Y. Prediction of resting energy expenditure from fat-free mass and fat mass. Am J Clin Nutr 1992;56:848e56. 39. Flier JS, Harris M, Hollenberg AN. Leptin, nutrition, and the thyroid: the why, the wherefore, and the wiring. J Clin Invest 2000;105:859e61. 40. Reinehr T. Obesity and thyroid function. Mol Cell Endocrinol 2010;316:165e71.
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original Article
Different thermic effects of leptin in adolescent females with varying body fat content Verena K. Haas a, b, *, Kevin J. Gaskin c, Michael R. Kohn a, Simon D. Clarke d, Manfred J. Müller e a
Department of Adolescent Medicine, The Children’s Hospital at Westmead, University of Sydney, Australia Franz-Volhard Clinical Research Center, ECRC, Charité University Hospital, Berlin, Germany c The James Fairfax Institute of Pediatric Nutrition, The Children’s Hospital at Westmead, University of Sydney, Australia d The Centre of Research into Adolescent’s Health, Westmead Hospital, University of Sydney, Australia e Institut für Humanernährung und Lebensmittelkunde der Christian-Albrechts-Universität zu Kiel, Germany b
a r t i c l e i n f o
s u m m a r y
Article history: Received 19 October 2009 Accepted 31 March 2010
Background & aims: Investigating the effect of leptin on energy expenditure in undernutrition might lead to a better understanding of the role of leptin in regulating body weight in humans. Methods: 73 underweight female adolescents with anorexia nervosa (AN) were compared with 23 healthy normal weight (nwC), and 9 overweight girls (OW); 37 AN were followed during 7 months of weight recovery. Resting energy expenditure (REE, by indirect calorimetry), body composition (fat mass, FM; lean tissue mass, LTM; by Dual-Energy X-Ray Absorptiometry) and plasma hormones of leptin and 3,5,30 -Triiodothyronine (T3) were measured. Results: In underweight, leptin, T3 and REE adjusted for lean tissue mass (REELTM) were decreased; in OW, FM and leptin were increased at unchanged T3 and REELTM. There was a significant positive relation between FM and leptin at low and normal (AN, r2 ¼ 0.26; nwC, r2 ¼ 0.51, p < 0.001), but not at high adiposity. Leptin and REELTM were positively associated in underweight (r2 ¼ 0.14, p ¼ 0.001) but not in normal or overweight subjects. T3 was linearly related to REELTM over the whole range of adiposity (r2 ¼ 0.42, p < 0.001). With weight gain in AN (5.0 3.5 kg) the relationship between leptin and REELTM changed toward the conditions seen in normal weight controls. Conclusions: At low adiposity the interrelated fall of leptin and REE reflect an adaptive mechanism to preserve body weight. High leptin production associated with excessive adiposity was without effect on metabolic adaptation. Ó 2010 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Resting energy expenditure Leptin T3 Body composition Underweight Anorexia nervosa
1. Introduction Since its discovery in 1994,1 the role of leptin in the regulation of body weight by affecting energy homeostasis has been extensively discussed.1e16 Many authors agree that leptin functions as an adipocyte-derived signal to convey information about energy status to the brain, which then synchronizes a range of physiological responses, such as the adjustment of REE, to altering energy scenarios. In mouse experiments it has been shown that leptin injection increased whole-body energy expenditure and directly stimulated thermogenesis in skeletal muscle.3,6,11 However when
* Corresponding author. Dr. oec. troph. Verena Haas, Charité University Hospital, Franz-Volhard Clinical Research Center, ECRC, Lindenberger Weg 80, 13125 Berlin, Germany. Tel.: þ49 30 450 540 324; fax: þ49 30 450 540 920. E-mail addresses: [email protected] (V.K. Haas), [email protected] (K.J. Gaskin), [email protected] (M.R. Kohn), [email protected]. au (S.D. Clarke), [email protected] (M.J. Müller).
compared with animal data a thermic effect of leptin remains equivocal in humans. Cross-sectional studies have not been able to show an association between plasma leptin concentrations and REE in normal weight and obese subjects.9,13,15 Although plasma leptin concentrations decreased with weight loss and fasting in lean and obese individuals, the magnitude of change in leptin did not correlate with the reduction of REE in a linear relationship in overweight and obese subjects.2,7,10,14,16 Finally there was no substantial impairment of REE in a child with congenital leptin deficiency, and weight loss following leptin treatment in this child was largely attributed not to changes in REE but to decreased food intake.4 These contradictory findings make it difficult to designate leptin as a key target in obesity treatment, and a better understanding of the role of leptin in the regulation of body weight is required. Adipocytokines have been studied with increasing intensity as a result of the emergence of obesity as a serious public health problem. However in his recent review,8 Leibel argues that from an evolutionary perspective, the need to suppress food intake in response to adiposity has not been
0261-5614/$ e see front matter Ó 2010 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. doi:10.1016/j.clnu.2010.03.013
640
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
particularly important, and therefore the physiological “defence” of body fat is weaker against upward deflections than against those bringing body fat below a threshold for minimum allowable FM. Prentice et al. argue along the same lines, namely that attempts to understand leptin’s function through the studies of overnourished people that now dominate the literature may be misleading, and that further investigations are required to elucidate its possible modulating role in the state of energy deficit.12 While data about the role of leptin in adaptive thermogenesis is still contradictory, it has long been accepted that thyroid hormone is an important determinant of overall energy expenditure and the basal metabolic rate.17 Thyroid hormone concentrations have been described to be related to weight status, with their concentrations being decreased in underweight, and increased in overweight conditions.18 The endocrinology of the leptin axis linked with energy metabolism might be different and not be proportional over a wide range of its concentrations. Considering there was no relation between REE and leptin in normal weight or obese people, high levels of leptin might be not responded to. However low concentrations could lead to reduced REE in an effort to conserve and restore body fat stores,8 therefore the focus of leptin research should also be on underweight or “underfat” populations. Anorexia nervosa (AN) is an illness characterized by low body weight and FM, and reduced leptin, REE as well as T3.5 In AN as well as in healthy controls, T3 and REE adjusted for FFM were linearly related.19 During shortterm weight gain over a period of 12 weeks FM, leptin and REE increased in AN and leptin secretion was associated with weight gain.5 The aim of the current study was to investigate the effect of leptin on REE in a study cohort of a broad range of FM, and to compare the relation between leptin and REE with that of T3. Therefore body composition and metabolic and endocrine function was assessed in anorexic as well as healthy adolescent females. In addition a subgroup of AN patients was followed longitudinally over 7 months to assess whether body composition, REE and endocrine dysfunction normalized with nutritional intervention.
2. Subjects and methods 2.1. Study populations A group of 73 adolescent female patients with anorexia nervosa (AN; mean age 14.9 1.6 years; range 11e19) was recruited from two inpatient hospital units associated with the Departments of Adolescent Medicine, The Children’s Hospital at Westmead and Westmead Hospital in Sydney/Australia. All patients were amenorrheic (12 with primary amenorrhea), diagnosed with AN according to DSM-IV (American Psychiatric Organization, 199420), and admitted to an inpatient Eating Disorder Program. Baseline testing was done after medical stabilization. Mean Tanner stage as assessed by the treating physician was 4. Forty AN returned for follow-up approximately 7 months after baseline testing and 37 of these had gained weight. Three AN who had lost weight between discharge and follow-up were excluded from further analysis. The 37 follow-up AN had had an average length of hospital stay of 43 23 days. Average therapeutically induced weight gain during hospital stay was 4 2 kg, with the rest of the weight gain occurring after discharge. A healthy sex- and age matched control group was recruited from a local high school (n ¼ 32; age 14.6 0.6 years). Median Tanner stage in the control group as self-reported by the girls with the use of questionnaires21 was 4. Three of the girls in the control group had not reached menarche at the time of testing, the remaining 29 girls were tested in the follicular phase of their menstrual cycle (i.e. day 1e10 of their menstrual cycle). There was no follow-up testing for the control group.
2.2. Body composition measurements For each individual all measurements were performed on the same morning, after an overnight fast and voiding. Height (0.1 cm) was measured with a stadiometer (Holtain Ltd, Crymych, United Kingdom) and body weight (0.1 kg) in light clothing using electronic scales (AND FW-150K, Tokyo, Japan). BMI was calculated as body weight/height2 (in kg/m2). Whole-body composition (FM, LTM expressed in kg and as a percentage of body weight) was assessed by Dual-Energy X-Ray Absorptiometry (DXA) using a Lunar Prodigy whole-body scanner (Lunar Corp, Madison, WI, USA) in conjunction with EnCORE Software Version 8.10. 2.3. Hormone assays Blood samples were drawn between 07:00 and 09:00 am after the subjects had fasted overnight. Samples for analysis of thyroid hormones were sent to the hospital laboratory for immediate analysis. 3,5,30 -Triiodothyronine (free T3) concentrations were quantified using a microparticle Enzyme Immuno Assay (MEIA, AxSYM, U.S.A). Samples for leptin analysis were spun for separation into plasma and serum, and kept frozen at 80 C, until batch analysis at the end of the study. Plasma leptin concentrations were measured by Immunoassay (Lincoplex Kit, Linco Research, St. Charles, MO, U.S.A). 2.4. Metabolic measurements Metabolic testing took place in the early morning after 8e12 h of fasting, in a thermoneutral room. REE and Respiratory quotient (RQ) were assessed by indirect calorimetry (Deltatrac TMII, Datex Instrumentarium Corp., Helsinki/Finland). Before each measurement, the machine was calibrated with room air and a 95%O2/5% CO2 gas mixture. Participants were taken to the metabolic room by wheelchair, then lying down, and a canopy connected to the metabolic cart was placed over the head. Oxygen consumption and carbon dioxide production were monitored for 30 min at 1-min intervals. Results taken in the first 10 min were considered as adaptation phase and discarded. REE and RQ were calculated according to Weir et al.22 REE was adjusted for LTM using a regression analysis (¼ REELTM). REE was also predicted according to Schofield23 and Müller et al.24 2.5. Statistical analysis Statistical analysis was carried out by using SPSS (version 16.0; SPSS Inc. Cary. NC). The data shown in Tables 1e3 are presented as means SDs at a significance level of P < 0.05. Normal distribution was tested with the KolmogoroveSmirnov-Test. Between-group comparison was carried out with a one-way ANOVA and a 2-tailed post-hoc Dunnett t-test, comparing both AN and OW with normal weight controls. When AN at baseline were compared with AN at follow-up, a paired samples t-test was used. To account for the effect of interindividual differences in LTM on variance of REE (¼ REELTM), REE was adjusted in a linear regression analysis with REE as dependent and LTM as independent variable. To calculate individual REELTM, individual regression residual was subtracted from group mean REE. To describe the relationship between leptin and either FM or REELTM, Pearson correlation coefficients were calculated using SPSS as well as best fitting mathematical models using Microsoft Excel Office Standard Edition 2003. Stepwise liner regression analysis with REELTM as dependent, and group affiliation (AN, nwC or OW) and leptin or T3 was carried out using SPSS (cases were excluded listwise; p of F-value for inclusion < 0.05, for exclusion > 0.1).
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
641
Table 1 Body composition, metabolic and endocrine function of AN, nwC and OW.
Age (y) Height (cm) Weight (kg) BMI (kg/m2)
AN baseline (n ¼ 73)
nwC (n ¼ 23)
OW (n ¼ 9)
14.9 1.6 163 8 43.0 6.0 16.2 1.4
14.6 0.7 161 8 51.3 9.3 19.6 2.4
14.6 0.3 165 7 77.9 11.8 28.5 2.8
(11e19) (128e176) (22.7e59.0)*** (13.1e19.7)***
(13e16) (142e174) (33.9e71.6) (15.9e23.6)
(14e15) (150e173) (55.0e90.4)*** (24.4e33.2)***
Data are presented as means SD (range); AN, patients with anorexia nervosa; nwC, normal weight controls; OW, overweight subjects; BMI, Body Mass Index. ***, p ¼ 0.000 when comparing AN and OW with nwC using one-way ANOVA and post-hoc Dunnett-t-test.
2.6. Ethics The Ethics Committees at the Children’s Hospital at Westmead and Westmead Hospital approved the study protocol. Prior to participating, all girls and their parents gave informed written consent. Assessment and analysis of data was carried out according to privacy laws. 3. Results 3.1. Descriptive presentation of general characteristics, body composition, and metabolic and endocrine function Within the control group, 9 girls were classified as overweight (OW) according to a BMI cut-off of >23.9 for girls aged 14.5 years as established by Cole et al.25 Table 1 presents the general characteristics of AN at baseline, and of nwC and OW. There was no significant difference in age or height across the study groups, but weight and BMI were significantly lower in AN, and significantly higher in OW when compared with nwC. Body composition assessed with DXA, as well as metabolic and endocrine data of all 73 AN at baseline, and of nwC and OW are presented in Table 2. When compared with nwC, AN had less body fat and unchanged LTM which at decreased weight explained an increased percentage of LTM. REE, REELTM, leptin, and T3 were significantly lower in AN than in nwC at increased RQ. In OW body fat, leptin, LTM, and REE when compared with nwC were increased but there was no difference in RQ, REELTM or T3. 3.2. Association between leptin, fat mass and metabolic rate When combining all subjects to represent the entire range of adiposity present in this study, the best mathematical model to describe the relation between percentage body fat and leptin was
an exponential curve (Fig. 1A). When analyzing the 3 study groups separately, there was a significant and positive relationship between absolute FM and leptin in both AN and nwC (Fig. 1B). The slope of the regression line was steeper in nwC when compared with AN, suggesting that an increase in 1 kg of body fat was associated with a higher increase in leptin in nwC compared to AN. In AN, the variance around the regression line was lower than in nwC (Fig. 1B, AN: SEE 1.37 vs. nwC: 2.90). In OW variability in leptin concentrations was high and the association between FM and leptin was lost (Fig. 1A and B). Fig. 2 shows the associations between REELTM and leptin (Fig. 2A) and T3 (Fig. 2B) in AN, nwC and OW. There was a positive association between leptin and REELTM in AN (r2 ¼ 0.14, p ¼ 0.001) but not in nwC, OW, or nwC þ OW combined. Fig. 2A shows that when data from all subjects were combined, the best fitting mathematical model to describe the relation between leptin and REELTM, was a logarithmic curve (logarithmic function, r2 ¼ 0.43 versus linear function, r2 ¼ 0.33). Contrary to the findings shown in Fig. 1, the slope of the curve is steep at low leptin and flat at normal leptin and suggests that the effect of leptin on REELTM differs at varying leptin concentrations. In a stepwise linear regression with REELTM as dependent, and leptin and group affiliation (AN, nwC, OW) as independent variables, both group affiliation and leptin were significant predictors and together explained 50% of the variance in REELTM (Table 4). When data of all three groups were combined, T3 was positively associated with both leptin and FM (p ¼ 0.001). In a second regression model with REELTM as dependent, and T3 and group affiliation as independent variables, both group affiliation and T3 were significant predictors and together explained 53% of the variance in REELTM (Table 4). Fig. 2B shows that when data from all subjects were combined, there was a significant and positive relationship between T3 and REELTM. However when compared with leptin, the mathematical model was linear, suggesting that the effect of T3 on REELTM was constant over the entire T3 concentration spectrum.
Table 2 Body composition, metabolic and endocrine function of AN, nwC and OW. AN baseline (n ¼ 73) Body composition FMDXA (kg) FMDXA (%) LTMDXA (kg) LTMDXA (%) Metabolism REE (kcal/d) REEpred Schofield (kcal/d) REEpred Müller (kcal/d) RQ REELTM (kcal/d) Hormones Leptin (mg/L) T3 (pmol/L)
6.3 15.0 34.7 81.2
2.8 5.7 4.4 5.9
(1.4e12.9)*** (4.2e26.3)*** (20.5e43.5) (70.4e94.3)***
1037 181 1316 83 1380 101 0.92 0.10 1054 173
(579e1457)*** (986e1499)** (991e1554) (0.70e1.20)*** (613e1515)***
2.0 1.6 (0.1e8.6)*** 3.1 0.8 (1.9e4.9)***
nwC(n ¼ 23) 13.5 26.8 35.2 69.5
5.6 6.9 4.4 6.7
1373 166 1380 109 1428 128 0.82 0.05 1379 128
(5.3e26.3) (14.1e40.0) (25.1e43.6) (55.8e81.2) (1105e1745) (1145e1609) (1151e1660) (0.70e0.93) (1141e1726)
6.6 4.1 (1.7e18.1) 4.2 0.6 (3.0e5.1)
OW (n ¼ 9) 33.6 44.8 40.8 52.7
7.3 4.5 5.2 4.3
1598 216 1620 109 1629 135 0.75 0.02 1458 124
(20.7e40.9)*** (36.9e51.2)*** (31.6e46.6)** (46.1e60.2)*** (1284e1890)** (1145e1609)*** (1339e1757)*** (0.71e0.77) (1228e1599)
17.5 7.6 (5.8e31.9)*** 4.0 0.8 (2.6e5.3)
Data are presented as means SD (range); AN, patients with anorexia nervosa; nwC, normal weight controls; OW, overweight subjects; FM, Fat Mass; DXA, Dual-Energy X-ray Absorptiometry; LTM, Lean Tissue Mass; REE, Resting Energy Expenditure; pred., predicted; RQ, Respiratory Quotient, REELTM, REE adjusted for LTM; *, p < 0.05; ***, p < 0.001 when comparing AN and OW with nwC using one-way ANOVA and post-hoc Dunnett-t-test.
642
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
Table 3 General characteristics, body composition, and metabolic and endocrine function of AN patients on admission and after weight gain. AN baseline (n ¼ 37)
AN follow-up (n ¼ 37)
Age (y) Weight (kg) Height (cm) BMI (kg/m2)
14.6 1.5 42.7 5.9 163 9 16.0 1.3
(11e17) (22.7e50.7) (128e176) (13.4e18.4)
15.1 1.5 47.7 7.1 164 9 17.7 1.8
(11e18)*** (25.5e62.2)*** (129e177)** (14.7e21.2)***
Body composition FMDXA (kg) FMDXA (%) LTMDXA (kg) LTMDXA (%)
6.0 14.5 34.4 81.1
(1.4e11.3) (4.8e25.3) (20.5e42.9) (70.4e91.8)
10.4 22.5 34.8 73.6
(3.6e18.7)*** (9.0e34.6)*** (19.4e43.9) (61.3e85.3)***
Metabolism REE (kcal/d) RQ REELTM (kcal/d) Hormones Leptin (mg/L) T3 (pmol/L)
2.7 5.6 4.4 6.2
4.1 6.8 4.9 6.5
995 173 (579e1301) 0.93 0.10 (0.74e1.20) 1020 168 (613e1406)
1158 188 (747e1582)*** 0.89 0.07 (0.76e1.03)* 1172 178 (851e1556)***
1.6 0.9 (0.1e3.9) 3.1 0.8 (2.0e4.9)
4.0 3.1 (0.8e10.3)*** 3.4 0.7 (2.2e4.7)*
Data are presented as means SD (range); AN, patients with anorexia nervosa; FM, Fat Mass; DXA, Dual-Energy X-ray Absorptiometry; LTM, Lean Tissue Mass; REE, Resting Energy Expenditure; RQ, Respiratory Quotient, REELTM, REE adjusted for LTM; *, p < 0.05; **, p < 0.01; ***, p < 0.001; AN baseline vs. AN at follow-up: paired samples t-test.
3.3. Longitudinal observations in an during weight gain In Table 3, the change of nutritional status as well as metabolic and endocrine function in AN after medium-term weight gain of 7 months is presented. After 7 months of weight gain AN had significantly increased their body weight by 5.0 3.5 (range: 0.4e14.4) kg, and BMI by 1.8 1.3 units (range: 0.1e4.7 kg/m2). During weight gain in AN, FM but not LTM increased significantly when considered in absolute terms. When described as a percentage of body weight, FM increased, but LTM decreased during weight gain in AN. REE as well as REELTM, leptin and T3 also increased, and RQ decreased significantly during weight recovery in AN. However at follow-up, REE as well as REELTM were still significantly lower than in nwC (p ¼ 0.000, data not shown). Fig. 3 shows the relationship between leptin and REELTM in AN before and after 7 months of weight gain. The association between leptin and REELTM before and after weight gain in AN was in accordance with cross-sectional data: in AN after weight gain the slope of the regression line was less steep when compared with AN before weight gain.
A
35
4. Discussion Leptin is a fat cell-derived hormone with multiple functions, including a role in the adaptation of an organism to semistarvation.26 The present study demonstrated an exponential association between leptin and fat mass with a strong and positive relation at low and normal, and a loss of this relation at high degrees of adiposity. The effect of leptin on REELTM was not constant over its whole concentration range and was best described with a logarithmic curve. By including anorexic subjects with drastically reduced body fat we were able to show a significant positive association between leptin and REELTM at low and a lack thereof at normal or high leptin concentration as seen in healthy normal weight and overweight girls. T3 was positively associated with both leptin and fat mass. Opposed to the different thermic effects of leptin with varying adiposity, there was a linear and positive association between T3 and REELTM over the whole range of adiposity. Weight gain of AN patients resulted in a shifting of the relationship between leptin and REELTM in accordance with the cross-sectional data. 4.1. Association between leptin and fat mass Many studies have described rising leptin concentrations with increasing adiposity, suggesting a linear relationship. However some studies investigating subjects with a wide BMI range have supported an exponential model to describe the relationship between leptin and percent body fat.27,28 In addition to the nonlinear relation between fat mass and leptin Luke et al. have described an increasing variance and loss of relation with increasing adiposity.27 Our findings examining a cohort including a group of patients with anorexia nervosa as well as subjects with normal, and a small group with excessive fat mass, are consistent with these results (Fig. 1A), showing an exponential relationship between fat mass and leptin. While the biological plausibility of the resulting mathematical exponential curve is unclear, the data suggest that leptin dynamics are non-linear and differ with varying amounts of body fat. Leptin appeared to reflect the body’s energy stores in a sensitive dose response manner at low and normal body fat. In contrast in the girls with excessive body fat, there was a wide variation in leptin levels and the association between leptin and FM was lost. Prentice et al. point out that when evolution has pressed a single molecular species such as leptin into a number of different functions in several organs, there must be mechanisms that permit
B
30
30
all subjects, exponential y = 0.45e 0,0847x R2 = 0.74***
25
35
nwC y = 0.53x - 0.49
25 leptin (µg/L)
leptin (µg/L)
R2 = 0.51*** 20 all subjects, linear y = 0.42x - 4,09 R2 = 0.68***
15
20 AN y = 0.29x + 0.23
15
10
10
5
5
0
0
R2 = 0.26***
AN nw C
0
10
20
30
40
50
60
OW
0
Body fat (DXA, %)
10
20
30
Body fat (DXA, kg) Fig. 1. Relation between body fat and leptin.
40
50
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
A 2000
B 2000
AN, linear y = 41x + 972 R2 = 0.14**
1800
1800
1600 1400
1400 all subjects, linear y = 24x + 1056 R2 = 0.33***
1000 800
REE LTM (kcal/d)
REE LTM (kcal/d)
all subjects, linear y = 165x + 596 R2 = 0.42***
1600
1200
all subjects, logarythm ic y = 140Ln(x) + 1031 R2 = 0.43***
600
643
1200 1000 800 600
400
400
200
200
AN nw C OW
0
0 0
5
10
15
20
25
30
0,0
35
1,0
2,0
leptin (µg/L)
3,0
4,0
5,0
6,0
T 3 (pmol/L)
Fig. 2. Relationship between leptin and energy expenditure.
4.2. The role of leptin and T3 in adaptive thermogenesis Leptin is thought to regulate body weight through hypothalamic effects on satiety and energy expenditure.29 However human data has rarely supported a correlation between fasting leptin concentrations and REE. Most of these studies mainly include normal weight, overweight and obese subjects. Even though leptin may be particularly important in signalling the transition between energy surplus and deficit,12 the thermogenic effect of leptin on REE has rarely been studied in underweight populations. Studying the role of leptin in modulating adaptive thermogenesis specifically in undernourished people could lead to a better understanding and help to resolve some of the existing controversies. The significant effect of group affiliation (being anorexic, normal weight, or overweight) on the variance in REELTM shown in our regression models (Table 4) points towards metabolic adaptation to different nutritional states beyond the effect produced by changes in lean tissue mass. In these same regression models, both leptin and T3 were independent predictors of REELTM and are thus likely to play a role in adjusting energy homeostasis. In malnourished patients with anorexia nervosa (AN) plasma leptin is reduced5,30e32 and a positive association between leptin and REE has been documented earlier.5,32 Our data showing a steep Table 4 Multiple linear regression analysis of resting energy expenditure adjusted for lean tissue mass (REELTM) with group affiliation (anorexic, normal weight, overweight) and plasma hormone concentrations (model 1: leptin; model 2: T3). Independent variables Model 1 Group Leptin Model 2 Group T3
b-coefficient (T-value)
P
and positive association between leptin and REELTM in undernourished subjects (Fig. 2A) are in accordance with these previous findings and suggest that low leptin levels in malnourished AN patients are interrelated with low energy expenditure and could reflect an adaptive mechanism to preserve body weight as others have alluded to.3,8,12 However after including normal- and overweight subjects, our present findings show that the effect of leptin on REELTM was not constant over its entire concentration spectrum and suggest a loss of association within the normal or increased range of FM (Fig. 2A). These different thermic effects of leptin warrant further discussion. Underweight is characterized by low FM, LTM and REE. Lean tissue mass is a major determinant of REE and part of the fall in REE with weight loss is thus explained by LTM losses.33,24 However by reanalyzing data from the Minnesota experiment of semistarvation, Dulloo et al. have suggested that in addition to the relation between the depletion in LTM and the fall in REE, there is an additional feedback control system linking the state of depletion of FM to mechanisms suppressing thermogenesis.34
2000 1800
AN follow -up y = 27x + 1062 R2 = 0.23**
1600 1400 REE LTM (kcal)
variable sensitivity across the physiologic range of circulating leptin.12 From a teleological point it can be argued that the major role of leptin is to signal energy deficit rather than surfeit, and marginal nutrition or undernutrition were the states that have mostly driven metabolic selection processes during the past millennia.12 The role of leptin in regulating energy metabolism might therefore be most obvious in subjects with low body fat.
1200 1000 800 600 AN basline y = 57x + 923 R2 = 0.10 (p=0.06)
400 200
AN baseline
2
Adjusted r ¼ 0.501 0.489 (6.219) 0.339 (4.315) Adjusted r2 ¼ 0.533 0.425 (5.340) 0.414 (5.199)
0.000 0.000
AN follow -up
0 0
5
10
15
20
leptin ( µ g/L) 0.000 0.000
Fig. 3. Relationship between leptin and energy expenditure in AN patients on admission and after weight gain.
644
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
In accordance with this hypothesis it has been shown that FM is an independent but relatively small predictor of REE,35e38 which is most likely explained by the endocrine activity of adipose tissue. In the current study a decrease in leptin (or body fat e we are not able to distinguish between these two parameters with our data) was associated with a moderate drop in REELTM in control subjects but with a large drop in REELTM in anorexic subjects. It can therefore be hypothesized that starvation-induced falls in leptin or FM cause metabolic adaptation to starvation by severely downregulating REE only in subjects falling short of a critically low threshold range of FM or leptin. Fig. 2A shows that nwC with low leptin concentrations (i.e., below 5 mg/L) were located close to the steep regression line described for AN. However AN subjects with relatively high leptin concentrations (i.e., above 5 mg/L) were located close to the flatter regression line described for healthy controls. Therefore the relationship between leptin and REELTM did not seem strictly dependent on subject group (anorexic patient or healthy control), but rather linked with the amount of circulating leptin or FM per se. In contrast to underweight patients with low FM, subjects with excessive body fat demonstrated a 3-fold increase in serum leptin but a plateauing of REE (Fig. 2A). This unresponsiveness of thermogenesis to high leptin in the overweight or obese could be the equivalent to a state that has been described as “leptin resistance” by several investigators. Whether this specific metabolic situation reflects a saturation of leptin receptors or end organ resistance, alterations either in the transport of leptin across the blood brain barrier and/or cellular leptin receptor signalling or whether there are other underlying reasons, is still unknown and needs further study. The sample size in the OW group was small and this limits our findings with respect to the lack of adaptive thermogenesis in the overweight state. To confirm our hypothesis, and to pinpoint the threshold for the loss of linearity between fat mass and leptin, further studies with higher subject numbers are required. While leptin has been suggested as a promising link between the regulation of energy balance and thyroid hormones,39 the exact nature of this relationship is still not clear. Our study shows that firstly, there was a positive association between leptin and T3. Second, when compared with leptin, plasma T3 concentrations had a linear and positive association with REELTM over their whole concentration range (Fig. 2B). These findings are in accordance with previous data in healthy subjects as well as in patients with AN19,39 and in line with the assumptions of Flier et al.40 suggesting that a decrease in FM and concomitantly leptin concentrations sensed by the brain could lead to a lowering of REE brought about by decreasing serum T3 concentrations. 4.3. Observations in anorexic patients during weight gain After weight gain, the relationship between leptin and REELTM changed in direction toward the conditions seen in healthy and normal weight controls (Fig. 3). Once acute starvation of AN was over, thermic adaptation to the new and positive energy balance was achieved. Starvation metabolism and the depression of REE were slowly released, which has been shown before in AN during and following weight gain.5,32 While in the current and also in our previous longitudinal study5 metabolic function had not completely normalized in AN after weight gain, this can be explained by the fact that at follow-up testing, these AN patients had still been underweight. Another study examining weight-recovered AN patients however suggests a full reversibility of this adaptive mechanism.32 4.4. Implications for clinical nutrition practice As thermogenesis directly determines the energy needs, knowledge about adaptive mechanisms or the lack thereof is
crucial when deciding on nutritional strategies in clinical practice. More specifically, when relying on commonly used prediction equations, energy needs will be considerably overestimated in undernourished and metabolically compromised subjects such as patients with anorexia nervosa (Table 2). The discrepancy between measured and predicted energy expenditure originates in the inability of the algorithms to account for endocrine mechanisms leading to adaptive thermogenesis as described in the current study. 4.5. Study limitations There were several limitations in our study. First of all the sample size in the overweight group was small and therefore our findings require further confirmation with higher subject numbers. Second, blood sampling was not carried out on the first day after the AN patients were admitted to hospital as recommended previously,26 but within the first week of admission. This was due to the fact that before testing, patients needed to be medically stabilized, and patient and parental consent had to be obtained. The extent of refeeding before blood sampling influences leptin levels in AN. However mean leptin concentrations assessed in the current study are well within the low range provided by earlier studies on AN patients who had been tested shortly after hospital admission.26 5. Conclusion After examining a cohort with a wide range of body fat, our findings suggest that leptin closely reflected fat mass at low and normal but not at high adiposity. The association between leptin and resting energy expenditure also varied with the degree of adiposity. The sharp decrease of both leptin and REE at low adiposity suggests that low leptin concentrations could be involved in reducing thermogenesis in order to conserve energy in underweight. On the contrary an increase in fat mass or leptin above normal range associated with high adiposity was without effect in terms of metabolic adaptation. Conflict of interest We have no conflict of interest. Authorship statement Study design and critical review of manuscript: Kevin Gaskin, Manfred James Müller. Practical performance: Verena Haas, Michael Kohn, Simon Clarke. Data analysis and preparation of manuscript: Verena Haas, Manfred James Müller. Acknowledgements The present study was supported by grants from the Centre of Research into Adolescent’s Health, Westmead Hospital, Sydney, Australia; the James Fairfax Institute of Pediatric Nutrition, The Children’s Hospital at Westmead, Sydney, Australia; and the DAAD, German Academic Exchange Service, Bonn, Germany. We thank Madeleine Thompson and July Briody for assistance with the body composition measurements. VKH collected and analyzed the data and wrote the manuscript. MK and SDC collected the data and provided advice. KJG and MJM designed the study and provided advice. None of the authors had a conflict of interest.
V.K. Haas et al. / Clinical Nutrition 29 (2010) 639e645
References 1. Zhang YY, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homolog. Nature 1994; 372:425e32. 2. Ahima RS. Role of leptin in the neuroendocrine response to fasting. Nature 1996;382:250e2. 3. Dulloo AG, Stock MJ, Solinas G, Boss O, Montani JP, Seydoux J. Leptin directly stimulates thermogenesis in skeletal muscle. Febs Lett 2002;515:109e13. 4. Farooqi IS, Jebb SA, Langmack G, Lawrence E, Cheetham CH, Prentice AM, et al. Effects of recombinant leptin therapy in a child with congenital leptin deficiency. N Engl J Med 1999;341:879e84. 5. Haas V, Onur S, Paul T, Nutzinger DO, Bosy-Westphal A, Hauer M, et al. Leptin and body weight regulation in patients with anorexia nervosa before and during weight recovery. Am J Clin Nutr 2005;81:889e96. 6. Hwa JJ, Ghibaudi L, Compton D, Fawzi AB, Strader CD. Intracerebroventricular injection of leptin increases thermogenesis and mobilizes fat metabolism in ob/ ob mice. Horm Metab Res 1996;28:659e63. 7. Klein S, Horowitz JF, Landt M, Goodrick SJ, Mohamed-Ali V, Coppack SW. Leptin production during early starvation in lean and obese women. Am J Physiol Endocrinol Metab 2000;278:E280eE284. 8. Leibel RL. Molecular physiology of weight regulation in mice and humans. Int J Obes 2008;32(Suppl. 7):S98eS108. 9. Martin LJ, Jones PJ, Considine RV, Su W, Boyd NF, Caro JF. Serum leptin levels and energy expenditure in normal weight women. Can J Physiol Pharmacol 1998;76:237e41. 10. Ostlund RE, Yang JW, Klein S, Gingerich R. Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab 1996;81:3909e13. 11. Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D, Boone T, et al. Effects of the obese gene-Product on body-weight regulation in Ob/Ob mice. Science 1995;269:540e3. 12. Prentice AM, Moore SE, Collinson AC, O’Connell MA. Leptin and undernutrition. Nutr Rev 2002;60:S56eS67. 13. Roberts SB, Nicholson M, Staten Dallal GE, Sawaya AL, Heyman MB, Fuss P, et al. Relationship between circulating leptin and energy expenditure in adult men and women aged 18 years to 81 years. Obes Res 1997;5:459e63. 14. Rosenbaum M, Nicolson M, Hirsch J, Murphy E, Chu F, Leibel RL. Effects of weight change on plasma leptin concentrations and energy expenditure. J Clin Endocrinol Metab 1997;82:3647e54. 15. Soares MJ, Piers LS, O’Dea K, Collier GR. Plasma leptin concentrations, basal metabolic rates and respiratory quotients in young and older adults. Int J Obes Relat Metab Disord 2000;24:1592e9. 16. Weigle DS, Duell PB, Connor WE, Steiner RA, Soules MR, Kuijper JL. Effect of fasting, refeeding, and dietary fat restriction on plasma leptin levels. J Clin Endocrinol Metab 1997;82:561e5. 17. Kim B. Thyroid hormone as a determinant of energy expenditure and the basal metabolic rate. Thyroid 2008;18(2):141e4. 18. Reinehr T, Isa A, de Sousa G, Dieffenbach R, Andler W. Thyroid hormones and their relation to weight status. Horm Res 2008;70:51e7. 19. Onur S, Haas V, Bosy-Westphal A, Hauer M, Paul T, Nutzinger D, et al. L-triiodothyronine is a major determinant of resting energy expenditure in underweight patients with anorexia nervosa and during weight gain. Eur J Endocrinol 2005;152:179e84. 20. Diagnostic and statistical manual of mental disorders (DSM-IV) 1994. Washington, DC: American Psychiatric Association.; 2000.
645
21. Duke PM, Litt IF, Gross RT. Adolescents self-assessment of sexual-maturation. Pediatrics 1980;66:918e20. 22. Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949;109:1e9. 23. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr 1985;39c(1s):5e42. 24. Müller MJ, Bosy-Westphal A, Klaus S, Kreymann G, Lührmann PM, NeuhäuserBerthold M, et al. World Health Organization equations have shortcomings for predicting resting energy expenditure in persons from a modern, affluent population: generation of a new reference standard from a retrospective analysis of a German database of resting energy expenditure. Am J Clin Nutr 2004;80:1379e90. 25. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320:1240e3. 26. Hebebrand J, Muller TD, Holtkamp K, Herpertz-Dahlmann B. The role of leptin in anorexia nervosa: clinical implications. Mol Psychiatry 2007;12:23e35. 27. Luke AH, Rotimi CN, Cooper RS, Long AE, Forrester TE, Wilks R, et al. Leptin and body composition of Nigerians, Jamaicans, and US blacks. Am J Clin Nutr 1998; 67:391e6. 28. Kennedy A, Gettys TW, Watson P, Wallace P, Ganaway E, Pan Q, et al. The metabolic significance of leptin in humans: gender-based differences in relationship to adiposity, insulin sensitivity, and energy expenditure. J Clin Endocrinol Metab 1997;82:1293e300. 29. Clement K, Vaisse C, Lahlou N, Cabrol S, Pelloux V, Cassuto D, et al. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature 1998;392:398e401. 30. Grinspoon S, Gulick T, Askari H, Landt M, Lee K, Anderson E, et al. Serum leptin levels in women with anorexia nervosa. J Clin Endocrinol Metab 1996;81: 3861e3. 31. Hebebrand J, Blum WF, Barth N, Coners H, Englaro P, Juul A, et al. Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration. Mol Psychiatry 1997;2:330e4. 32. Polito A, Fabbri A, Ferro-Luzzi A, Cuzzolaro M, Censi L, Ciarapica D, et al. Basal metabolic rate in anorexia nervosa: relation to body composition and leptin concentrations. Am J Clin Nutr 2000;71:1495e502. 33. Keyes A, Brozek J, Henschel A, Mickelsen O, Taylor HL. The biology of human starvation. Minneapolis: University of Minnesota Press; 1950. 34. Dulloo AG, Jacquet J. Adaptive reduction in basal metabolic rate in response to food deprivation in humans: a role for feedback signals from fat stores. Am J Clin Nutr 1998;68:599e606. 35. Johnstone AM, Murison SD, Duncan JS, Rance KA, Speakman JR. Factors influencing variation in basal metabolic rate include fat-free mass, fat mass, age, and circulating thyroxine but not sex, circulating leptin, or triiodothyronine. Am J Clin Nutr 2005;82:941e8. 36. Blaak EE, Hul G, Verdich C, Stich V, Martinez JA, Petersen M, et al. Impaired fatinduced thermogenesis in obese subjects: the NUGENOB study. Obesity 2007;15:653e63. 37. Elia M. Energy expenditure in the whole body. In: Kinney JM, Tucker HN, editors. Energy metabolism: tissue determinants and cellular corollaries. New York: Raven Press; 1992. p. 19e59. 38. Nelson KM, Weinsier RL, Long CL, Schutz Y. Prediction of resting energy expenditure from fat-free mass and fat mass. Am J Clin Nutr 1992;56:848e56. 39. Flier JS, Harris M, Hollenberg AN. Leptin, nutrition, and the thyroid: the why, the wherefore, and the wiring. J Clin Invest 2000;105:859e61. 40. Reinehr T. Obesity and thyroid function. Mol Cell Endocrinol 2010;316:165e71.