Abstract
Elevated serum leptin levels following rapid therapeutically induced weight gain in anorexia nervosa (AN) patients are discussed as a potential biomarker for renewed weight loss as a result of leptin-related suppression of appetite and increased energy expenditure. This study aims to analyze the predictive value of leptin levels at discharge as well as the average rate of weight gain during inpatient or day patient treatment for body weight at 1-year follow-up. 121 patients were recruited from the longitudinal Anorexia Nervosa Day patient versus Inpatient (ANDI) trial. Serum leptin levels were analyzed at referral and discharge. A multiple linear regression analysis to predict age-adjusted body mass index (BMI-SDS) at 1-year follow-up was performed. Leptin levels, the average rate of weight gain, premorbid BMI-SDS, BMI-SDS at referral, age and illness duration were included as independent variables. Neither leptin levels at discharge nor rate of weight gain significantly predicted BMI-SDS at 1-year follow-up explaining only 1.8 and 0.4 % of the variance, respectively. According to our results, leptin levels at discharge and average rate of weight gain did not exhibit any value in predicting weight at 1-year follow-up in our longitudinal observation study of adolescent patients with AN. Thus, research should focus on other potential factors to predict weight at follow-up. As elevated leptin levels and average rate of weight gain did not pose a risk for reduced weight, we found no evidence for the beneficial effect of slow refeeding in patients with acute AN.
Similar content being viewed by others
Introduction
Lower levels of leptin are found in patients with acute anorexia nervosa (AN) compared to healthy subjects. Similar to healthy subjects, serum or plasma levels are correlated with reduced fat mass and low body mass index (BMI) in patients with AN [1–5]. Leptin also has a strong dynamic component that causes it to rise or fall disproportionately after energy intake or short-term fasting, respectively; accordingly, during acute weight gain or fasting, leptin levels increase or decrease more strongly than would be expected based on a change in fat mass alone (for reviews, see [6, 7]).
Because of this dynamic effect, leptin levels can increase substantially during weight gain in patients with AN, sometimes peaking at supra-physiological levels in comparison with BMI-matched healthy controls [7–11]. Rapid weight gain and increased leptin levels at the end of inpatient treatment have thus been considered a potential risk factor for renewed weight loss [6, 10, 11]. We identified two studies that investigated the predictive value of leptin levels. Haas et al. [12] studied 19 adult patients with AN (mean age 27 ± 7 years, mean BMI 15.2 ± 1.5 kg/m2), those patients with a higher increase in leptin levels during the first 6 weeks of inpatient refeeding showed reduced weight gain during weeks 6–12 of continued inpatient treatment compared with those with a slower increase in leptin levels. Holtkamp et al. [13] showed that high leptin levels at discharge predicted weight loss [a reduction in age-adjusted BMI-standard deviation scores (SDS)] 2 months and 1 year after discharge in 18 adolescent patients with AN (mean age 14.1 ± 1.2 years, mean BMI 14.4 ± 1.2 kg/m2). This study used a linear regression model that also included BMI-SDS at discharge and delta BMI-SDS between referral and discharge in the prediction [13].
However, elevated leptin levels have been shown to decline after discharge within a matter of weeks [6, 7]. Thus, the question arises whether a short-term peak in leptin surrounding discharge is indeed a predictive factor in a long-term effect on patients’ weight following discharge. A small initial intervention study found that measuring leptin levels during treatment with the aim of optimizing energy intake during refeeding to prevent high leptin levels did not result in better weight gain [14].
Accordingly, the aim of this study was to analyze the predictive value of leptin levels at discharge for BMI-SDS 1 year after referral while controlling for premorbid, referral and discharge BMI-SDS as well as age at referral, illness duration prior to referral and the average rate of weight gain between referral and discharge in a large sample of adolescent patients with AN.
Methods
Study sample
The participants were adolescent females aged 11–18 years with their first referral for AN recruited from the multicenter Anorexia Nervosa Day Patient versus Inpatient (ANDI) trial [15]. The weight threshold for inclusion in the study was a BMI below the 10th age-percentile. Serum blood samples and complete follow-up data were available for 121 of all 172 participating patients. Blood samples were drawn after an overnight fast in the morning and within the same week as anthropometric measurements. The leptin levels in blood samples at referral and discharge were analyzed. Serum concentrations of leptin were measured by a commercially available enzyme-linked immunosorbent assay at the Institute for Clinical Chemistry and Pathobiochemistry, Otto-von-Guericke-University Magdeburg, according to the manufacturer’s instructions (BioVendor; Czech Republic) with an intra-assay coefficient of variation (CV) of 5.9 %, inter-assay CV of 5.6 % and a lower limit of detection of 0.2 ng/ml.
This study was approved by the Ethics Committee of the University Hospital of the Technical University of Aachen and by all participating centers and was conducted in accordance with the Declaration of Helsinki. The participants and their legal guardians provided informed written consent after receiving a complete description of the study.
In brief, the subjects received a three-week stepped care program of stabilizing inpatient treatment and then, on a randomized basis, were assigned to treatment groups receiving continued inpatient care (treatment as usual) or day patient care. For all patients, a minimum weight gain of 300–500 g/week was intended. The subjects were discharged after a mean duration of treatment of 16 ± 6.7 weeks and with a mean weight gain of 8.9 ± 3.7 kg (0.58 ± 0.24 kg/week), with continued outpatient treatment according to the usual German practice until 12 months follow-up after referral [15]. For the purpose of this study, we analyzed the data sets of patients in both treatment groups jointly (see Table 1 for the sample characteristics). The data indicated that 21 of the 121 patients had to be readmitted after an average of 22.7 weeks after discharge (SD: 13.4 weeks, min 2.1 weeks, max 48 weeks, interquartile range 22.9 weeks). 22 patients (18.2 %) were of the binge-purging subtype and 99 (81.8 %) were of the restrictive subtype.
Statistical analysis
Variables of interest
BMI-SDS for premorbid weight (BMI-SDSpremorbid), weight at referral (BMI-SDSreferral), discharge (BMI-SDSdischarge) and 1-year follow-up (BMI-SDS1 year) were calculated using the Kromeyer-Hauschild normative data set of German youth [16]. All anthropometric variables (age, height and weight) were measured at each time point, except for premorbid height and weight. Premorbid BMI was approximated by measuring height at referral and by the patient’s recalled premorbid weight, in accordance with previous research [17]. Delta BMI-SDSpremorbid-discharge (therapeutic weight gain) was calculated as the difference between premorbid and discharge BMI-SDS, and delta BMI-SDSreferral-discharge (distance to premorbid weight) was computed as the difference between discharge and referral BMI-SDS, equivalent to the weight gain in BMI-SDS during treatment. The onset of AN was defined as the approximate time point at which the weight loss first occurred, the time point at which no further weight gain occurred despite an age-appropriate growth spurt, or the first occurrence of secondary amenorrhea, whichever came first. The duration of illness was calculated as the time in weeks between the onset of AN and the time of referral. The average rate of weight gain during treatment measured in BMI-SDS was calculated by dividing delta BMI-SDSreferral-discharge by the number of weeks in treatment. For descriptive purposes only, we also calculated the aforementioned measure of the average rate of weight gain in kg/week.
Correlations of serum leptin levels at discharge with weight-related measures
Pearson’s correlations between serum leptin at discharge and the weight-related measures available at discharge (BMI-SDSpremorbid, BMI-SDSreferral, BMI-SDSdischarge, delta BMI-SDSpremorbid-discharge, delta BMI-SDSreferral-discharge and average rate of weight gain) were calculated. Additionally, Pearson’s correlations of both serum leptin at discharge and the average rate of weight gain during treatment with BMI-SDS at 1-year follow-up were calculated. Logarithmic leptin levels (LogLeptin) were used because raw leptin levels have been shown to follow a skewed (approximately exponential) distribution in prior studies [6, 9, 10].
Linear regression
To determine the predictive power of LogLeptindischarge for BMI-SDS1 year independent of other potentially predictive clinical parameters, we performed a multiple linear regression. The dependent variable was BMI-SDS1 year; the independent variables were BMI-SDSpremorbid, BMI-SDSreferral, BMI-SDSdischarge, average rate of weight gain, age and illness duration and LogLeptindischarge. To determine the amount of total variance explained by LogLeptindischarge, and the average rate of weight gain, we performed two more multiple repression analyses by omitting LogLeptindischarge and average rate of weight gain, respectively, and then calculated the differences in total explained variances.
Note that it is not possible to include BMI-SDS differences such as delta BMI-SDSreferral-discharge (therapeutic weight gain) simultaneously with BMI-SDSdischarge and BMI-SDSreferral in the regression since these are completely colinear (e.g., delta BMI-SDSreferral-discharge is the difference between BMI-SDSdischarge and BMI-SDSreferral). However, upon inclusion of BMI-SDSdischarge in the regression, BMI-SDSreferral explained the same amount of (joint) variance as delta BMI-SDSreferral-discharge. The same holds true for delta BMI-SDSpremorbid-discharge (distance to premorbid weight), which cannot be combined with BMI-SDSdischarge and BMI-SDSpremorbid.
Additional analyses
Additional sensitivity analyses were performed to assess the effects of interaction terms, the omission of particular variables, and the non-linearity of LogLeptin. In addition, among the 121 patients with complete BMI-SDS values at 1-year follow-up, 21 had to be readmitted prior to this date. As readmission was necessary primarily because of low body weight and might have led to a renewed treatment-induced increase in BMI-SDS before the 1-year follow-up point, we recalculated the above linear regression models using an adapted BMI-SDS outcome. In relapsed patients, we replaced BMI-SDS at 1 year with the readmission BMI-SDS to consider a worst-case scenario [15]. Furthermore, we recalculated the above linear regression entirely without the 21 patients who had to be readmitted. SPSS 20 was used for all statistical analyses [18]. Lastly, we analyzed, whether the change of Leptin between discharge and referral (Delta_LogLeptindischarge-referral) explained more variance than LogLeptindischarge alone and whether inclusion of the AN DSM-IV subtypes, restricting and binge eating/purging subtypes, increased explanatory power and validity of the resulting estimates.
Results
Correlations of LogLeptindischarge with weight-related measures
LogLeptin at discharge was negatively correlated with BMI-SDSpremorbid and with delta BMI-SDSpremorbid-discharge (see Table 2). The positive correlation between LogLeptindischarge and delta BMI-SDSreferral-discharge showed a trend towards significance. However, the negative correlation between LogLeptindischarge and BMI-SDS1 year and between average rate of weight gain and BMI-SDS1 year did not reach significance. We found no significant association between LogLeptindischarge and average rate of weight gain.
Linear regression
LogLeptindischarge was not linearly correlated with BMI-SDS1 year (see Table 2). When using all available BMI-SDS values, age and illness duration including LogLeptindischarge and average rate of weight gain to predict BMI-SDS at 1 year, we found that 32.4 % of the variance could be explained (see Table 3). In this multiple regression analysis, BMI-SDSdischarge was identified as an independently significant factor. LogLeptindischarge and average rate of weight gain were not significant. Separately omitting LogLeptindischarge and average rate of weight gain reduced the explained variance in BMI-SDS1 year by only 1.8 and 0.4 %, respectively.
Additional analysis
Independently including non-linear correlation terms for LogLeptindischarge or interactions with age, BMI-SDSdischarge or average rate of weight gain did not add more than 2.3 % of the explained variance to the above model (all p values were non-significant), excluding non-linear effects.
Omitting BMI-SDSpremorbid resulted in LogLeptindischarge becoming a significant predictor of BMI-SDS1 year (β = −0.177, p = 0.028).
Repeating the above analysis with the alternative endpoint combining BMI-SDS at relapse (for those who were readmitted) and BMI-SDS at 1 year (for those who were not readmitted) yielded results similar to those reported above (see supplementary Table 1). The same holds true when the analysis is repeated without the readmitted patients (see supplementary Table 2). Including Delta_LogLeptindischarge-referral instead of LogLeptindischarge alone did not explain more variance (see supplementary Table 3). Inclusion of the AN subtypes did not significantly increase explanatory power or validity of the resulting estimates (R 2 = 0.326 vs R 2 = 0.324, LogLeptinreferral explained 1.9 vs 1.8 % of the variance).
Discussion
We found no significant association between leptin levels at discharge and body weight at 1-year follow-up. Leptin levels at discharge and the average rate of weight gain during therapy were not significantly correlated. Moreover, the average rate of weight gain did not predict body weight at 1-year follow-up. When analyzing the independent predictive power of leptin levels at discharge for body weight at 1-year follow-up, we found that leptin accounted for 1.8 % of the additional explained variance compared with current and past body weights as the sole predictors. Average weight gain rate accounted for only 0.4 % of the additional explained variance.
The leptin results of the current study appear to point in a different direction than previous reports that found a predictive role of increased leptin during the first 6 weeks of treatment in ensuing slower weight gain during weeks 6–12 [12], potentially because of hunger suppression and increased metabolism. In addition, Holtkamp et al. [13] showed a significant predictive effect of increased leptin at discharge for lower BMI-SDS at 2 months and 1-year follow-up when they included BMI-SDS at discharge and 2 months, weight gain and leptin at referral in the regression. However, neither study included premorbid weight in the models which could be the reason for this difference in findings. Note that when excluding premorbid BMI-SDS from our descriptive model, leptin levels at discharge had a significant but minor influence in predicting BMI-SDS at 1 year. Both premorbid weight and delta BMI-SDSpremorbid-discharge were significantly correlated with serum leptin at discharge in our study, a greater delta was associated with lower leptin values.
This fits well with previous results. Föcker et al. have shown that premorbid BMI and BMI measurements at referral and discharge predicted future weight development in AN in this same sample [19]. Thus, BMI tracking appears to be an underlying factor predicting BMI-SDS at 1 year [20]. Leptin at discharge appears to be an indicator of the delta remaining between BMI-SDS at discharge and premorbid BMI-SDS but is not independently predictive of ensuing weight changes.
The therapeutic impact
High leptin levels at discharge do not predict low weight at 1-year follow-up in a clinically relevant manner; consistent in all regression models used in this study, leptin levels at discharge had only a minor influence on follow-up BMI-SDS. This result should encourage further research on available variables at discharge to predict weight at follow-up. Additionally, the average rate of weight gain of each patient within the range used in this trial was not associated with patients’ weight at 1-year follow-up. As this is not an intervention study systematically varying weight gain, no causal inferences can be made. However, our large observation study showed no indication for a disadvantage for patients with rapid refeeding or an advantage for patients with slow refeeding as hypothesized by Holtkamp et al. [13]. This complements the findings of recent studies that actively favor more rapid weight gain in intervention trials [21–24]. Future studies should include a broader weight gain range above 1000 g per week to provide further information about the appropriate weight gain rate in an attempt to limit the amount of time spent in the state of semi-starvation, with its harmful somatic consequences for bone density and brain volume, for example [25, 26]. There is currently an active debate on how quickly the process of refeeding patients with AN should proceed [27–29]. Several studies have shown a reduction in the length of hospital treatment with rapid refeeding, leading to weight restoration of up to 1700 g per week [21–24] with a reduction in health care costs and potential benefits for rapid reintegration into the social environment [15]. The first systematic studies on refeeding syndrome, a potentially harmful complication of rapid realimentation, appear to indicate that this is not a common problem [24, 30]. An initial intervention trial with an average of 1350 g weight gain per week showed no negative consequences and lower readmission rates [31].
Limitations
This study used outcome data for only 121 of the 172 patients in the ANDI-trial in comparing day patient treatment after 3 weeks of inpatient stabilization with inpatient treatment. However, comparing non-participants to participants did not result in significant differences. As weight gain is not necessarily a linear process, more detailed studies on the course of weight gain and its predictive power are needed, in addition to randomized longitudinal intervention studies in which the target weight gain per week is varied to distinguish correlation from causation. This need is particularly relevant because our variable for the average rate of weight gain cannot separate patients with a high initial weight gain from those who experience a more substantial weight gain in a later stage of treatment. We did not measure physical activity levels in our patients which are known to be associated with leptin levels in AN [32], so we cannot make a statement about their potential influence. Furthermore, we did not measure fat mass in patients, which is closely related to leptin values. Both could have led to a higher amount of total explained variance in the prediction of BMI-SDS [12] and potentially different partial effects. However, adding these two variables might have lowered the additional predictive power of leptin even further because of the high degree of shared variance of leptin with physical activity and body fat percentage. In relation to our results, the inclusion of physical activity or fat mass in our statistical regression model would thus be unlikely to change the clinically relevant finding that leptin at discharge does not independently predict BMI-SDS at 1 year.
Conclusion
In summary, leptin levels at discharge and average rate of weight gain were not predictive of weight at 1-year follow-up after referral. Other predictors are thus needed to estimate weight development after discharge. Knowledge of the appropriate predictors would facilitate the use of further required treatment strategies to prevent renewed weight loss. Furthermore, future research should identify the most appropriate weekly weight gain for inpatient or day patient treatment.
References
Föcker M, Timmesfeld N, Scherag S et al (2011) Screening for anorexia nervosa via measurement of serum leptin levels. J Neural Transm 118:571–578. doi:10.1007/s00702-010-0551-z
Grinspoon S, Gulick T, Askari H et al (1996) Serum leptin levels in women with anorexia nervosa. J Clin Endocrinol Metab 81:3861–3863. doi:10.1210/jcem.81.11.8923829
Köpp W, Blum WF, von Prittwitz S et al (1997) Low leptin levels predict amenorrhea in underweight and eating disordered females. Mol Psychiatry 2:335–340
Mathiak K, Gowin W, Hebebrand J et al (1999) Serum leptin levels, body fat deposition, and weight in females with anorexia or bulimia nervosa. Horm Metab Res Horm Stoffwechselforschung Horm Métabolisme 31:274–277. doi:10.1055/s-2007-978732
Monteleone P, Fabrazzo M, Tortorella A et al (2002) Opposite modifications in circulating leptin and soluble leptin receptor across the eating disorder spectrum. Mol Psychiatry 7:641–646. doi:10.1038/sj.mp.4001043
Hebebrand J, Muller TD, Holtkamp K, Herpertz-Dahlmann B (2007) The role of leptin in anorexia nervosa: clinical implications. Mol Psychiatry 12:23–35. doi:10.1038/sj.mp.4001909
Monteleone P, Maj M (2013) Dysfunctions of leptin, ghrelin, BDNF and endocannabinoids in eating disorders: beyond the homeostatic control of food intake. Psychoneuroendocrinology 38:312–330. doi:10.1016/j.psyneuen.2012.10.021
Chan JL, Mantzoros CS (2005) Role of leptin in energy-deprivation states: normal human physiology and clinical implications for hypothalamic amenorrhoea and anorexia nervosa. Lancet 366:74–85. doi:10.1016/S0140-6736(05)66830-4
Hebebrand J, Blum WF, Barth N et al (1997) Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration. MolPsychiatry 2:330–334
Holtkamp K, Hebebrand J, Mika C et al (2003) The effect of therapeutically induced weight gain on plasma leptin levels in patients with anorexia nervosa. J Psychiatr Res 37:165–169
Mantzoros C, Flier JS, Lesem MD et al (1997) Cerebrospinal fluid leptin in anorexia nervosa: correlation with nutritional status and potential role in resistance to weight gain. J Clin Endocrinol Metab 82:1845–1851. doi:10.1210/jcem.82.6.4006
Haas V, Onur S, Paul T et al (2005) Leptin and body weight regulation in patients with anorexia nervosa before and during weight recovery. Am J Clin Nutr 81:889–896
Holtkamp K, Hebebrand J, Mika C et al (2004) High serum leptin levels subsequent to weight gain predict renewed weight loss in patients with anorexia nervosa. Psychoneuroendocrinology 29:791–797
Lob S, Pickel J, Bidlingmaier M et al (2003) Serum leptin monitoring in anorectic patients during refeeding therapy. Exp Clin Endocrinol Diabetes Off J Ger Soc Endocrinol Ger Diabetes Assoc 111:278–282. doi:10.1055/s-2003-41286
Herpertz-Dahlmann B, Schwarte R, Krei M et al (2014) Day-patient treatment after short inpatient care versus continued inpatient treatment in adolescents with anorexia nervosa (ANDI): a multicentre, randomised, open-label, non-inferiority trial. Lancet 383:1222–1229. doi:10.1016/S0140-6736(13)62411-3
Kromeyer-Hauschild K, Wabitsch M, Kunze D et al (2001) Perzentile für den Body-mass-Index für das Kindes- und Jugendalter unter Heranziehung verschiedener deutscher Stichproben. Monatsschr Kinderheilkd 8:807–818
Coners H, Remschmidt H, Hebebrand J (1999) The relationship between premorbid body weight, weight loss, and weight at referral in adolescent patients with anorexia nervosa. Int J Eat Disord 26:171–178
IBM Corp (2011) IBM SPSS statistics for Windows, version 20.0. IBM Corp., Armonk, New York
Föcker M, Bühren K, Timmesfeld N et al (2014) The relationship between premorbid body weight and weight at referral, at discharge and at 1-year follow-up in anorexia nervosa. Eur Child Adolesc Psychiatry. doi:10.1007/s00787-014-0605-0
Hebebrand J (2015) Identification of determinants of referral and follow-up body mass index of adolescent patients with anorexia nervosa: evidence for the role of premorbid body weight. Eur Child Adolesc Psychiatry 24:471–475. doi:10.1007/s00787-015-0711-7
Agostino H, Erdstein J, Di Meglio G (2013) Shifting paradigms: continuous nasogastric feeding with high caloric intakes in anorexia nervosa. J Adolesc Health Off Publ Soc Adolesc Med 53:590–594. doi:10.1016/j.jadohealth.2013.06.005
Garber AK, Mauldin K, Michihata N et al (2013) Higher calorie diets increase rate of weight gain and shorten hospital stay in hospitalized adolescents with anorexia nervosa. J Adolesc Health Off Publ Soc Adolesc Med 53:579–584. doi:10.1016/j.jadohealth.2013.07.014
Golden NH, Keane-Miller C, Sainani KL, Kapphahn CJ (2013) Higher caloric intake in hospitalized adolescents with anorexia nervosa is associated with reduced length of stay and no increased rate of refeeding syndrome. J Adolesc Health Off Publ Soc Adolesc Med 53:573–578. doi:10.1016/j.jadohealth.2013.05.014
Leclerc A, Turrini T, Sherwood K, Katzman DK (2013) Evaluation of a nutrition rehabilitation protocol in hospitalized adolescents with restrictive eating disorders. J Adolesc Health Off Publ Soc Adolesc Med 53:585–589. doi:10.1016/j.jadohealth.2013.06.001
Misra M, Katzman D, Miller KK, et al (2011) Physiologic estrogen replacement increases bone density in adolescent girls with anorexia nervosa. J Bone Miner Res 26(10):2430–2438. doi:10.1002/jbmr.447
Seitz J, Bühren K, von Polier GG (2014) Morphological changes in the brain of acutely ill and weight-recovered patients with anorexia nervosa. A meta-analysis and qualitative review. Z Für Kinder- Jugendpsychiatrie Psychother 42:7–17. doi:10.1024/1422-4917/a000265
Katzman DK (2012) Refeeding hospitalized adolescents with anorexia nervosa: is “start low, advance slow” urban legend or evidence based? J Adolesc Health Off Publ Soc Adolesc Med 50:1–2. doi:10.1016/j.jadohealth.2011.10.003
Le Grange D (2013) Examining refeeding protocols for adolescents with anorexia nervosa (again): challenges to current practices. J Adolesc Health Off Publ Soc Adolesc Med 53:555–556. doi:10.1016/j.jadohealth.2013.08.015
Marzola E, Nasser JA, Hashim SA et al (2013) Nutritional rehabilitation in anorexia nervosa: review of the literature and implications for treatment. BMC Psychiatry 13:290. doi:10.1186/1471-244X-13-290
Rocks T, Pelly F, Wilkinson P (2014) Nutrition therapy during initiation of refeeding in underweight children and adolescent inpatients with anorexia nervosa: a systematic review of the evidence. J Acad Nutr Diet 114:897–907. doi:10.1016/j.jand.2013.11.022
Rigaud DJ, Brayer V, Roblot A et al (2011) Efficacy of tube feeding in binge-eating/vomiting patients: a 2-month randomized trial with 1-year follow-up. JPEN J Parenter Enteral Nutr 35:356–364. doi:10.1177/0148607110382422
Holtkamp K, Herpertz-Dahlmann B, Hebebrand K et al (2006) Physical activity and restlessness correlate with leptin levels in patients with adolescent anorexia nervosa. Biol Psychiatry 60:311–313
Acknowledgments
The authors would like to thank all patients and their parents for their participation. Furthermore, they would like to thank Wolfgang Scharke for his support in analyzing the data.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Ethical standards
This study was approved by the Ethics Committee of the University Hospital of the Technical University of Aachen and by all participating centers and was conducted in accordance with the Declaration of Helsinki. The participants and their legal guardians provided informed written consent after receiving a complete description of the study.
Financial statements and conflicts of interest
The ANDI-trial was supported by the German Federal Ministry of Education and Research (BMBF No. 01GV0602, ISRCTN67783402, DRKS00000101). None of the authors have conflicting interests to declare.
Additional information
Johannes Hebebrand and Manuel Föcker contributed equally to this work.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Seitz, J., Bühren, K., Biemann, R. et al. Leptin levels in patients with anorexia nervosa following day/inpatient treatment do not predict weight 1 year post-referral. Eur Child Adolesc Psychiatry 25, 1019–1025 (2016). https://doi.org/10.1007/s00787-016-0819-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00787-016-0819-4