OBJECTIVE: Single-slice magnetic resonance imaging (MRI) and computed tomography (CT) are finding increasing use as methods to estimate visceral fat content in human studies. To assess the validity of this approach, we have compared single- and multi-slice MRI methods for the measurement of intra-abdominal adipose tissue content.
MEASUREMENTS: Multi-slice whole-body MR images and single-slices at the level of L2–L3 and L4–L5 were obtained from 59 healthy female volunteers chosen to cover a wide range of body size, and from 17 healthy female volunteers before and after a 6-month exercise intervention.
RESULTS: Taking the group as a whole, significant correlation between multi-slice and single-slices was observed (L2–L3: r=0.56; P<0.01; L4–L5: r=0.76; P<0.01). However, the ranking of subjects according to their intra-abdominal fat content was significantly altered by the choice of MRI strategy, especially using L2–L3 methodology. Whole-body (−23.8±20.0%; P<0.01) and single-slice (L2–L3: −31.03±35.0%; P<0.01; L4–L5: −22.1±37.24%; p<0.05) MRI methods also detected a significant decrease in intra-abdominal fat following the exercise protocol, although the latter techniques gave rise to increased spreading of the data from the mean. These results suggest that the use of single-slice imaging techniques can lead to inconsistencies in the estimation of intra-abdominal fat content, which in turn can have significant effects on data interpretation.
CONCLUSION: Single-slice MRI appears to be suitable for assessing changes in intra-abdominal fat content in interventional studies, especially in large cohort of subjects, where each subject can serve as its own control. However, for accurate determination of an individual's intra-abdominal fat content, and intersubject comparison, only multi-slice imaging will give precise results.
Accurate and reproducible quantitation of total and regional body fat is central to our understanding of the role of body adiposity in health and disease. Many indirect and direct methods have been developed and implemented in the last few decades to tackle this problem, including underwater weighing (UWW), bioelectrical impedance (BIA), dual-energy X-ray absorptiometry (DXA), anthropometry, computed tomography (CT) and magnetic resonance imaging (MRI).1,2 At present, it is generally accepted that CT and MRI are the in vivo methods of choice for directly assessing regional adipose tissue content in human subjects. However, a close scrutiny of the published literature shows that there is no clear consensus as to a standardised protocol for MRI or CT quantitation of visceral fat, with several approaches being used in a variety of studies.3,4
Measurement of body fat content by MRI has been fully validated using animal models and cadavers.5,6,7 From these and other studies, a number of MRI protocols have been developed, including total transaxial whole-body imaging (head-to-toes),8,9,10,11 multi-slice transaxial imaging of the entire abdominal area of the body,12,13,14,15 and single transaxial slice imaging at a predetermined area of the abdomen (L4–L5 or L2–L3).16,17,18,19 The former strategy is probably the most accurate MRI method for assessing total and regional adipose tissue content, avoiding inaccuracies that may arise from abnormal fat distribution. However, this method can be relatively expensive, requiring extensive analysis time. In practice, single-slice imaging at a predetermined area of the abdomen appears to be the most commonly used methodology in the literature.3
Single-slice approaches were partially developed to provide a simplified mode of estimating relative visceral fat content. Visceral adipose tissue area measurements at the level of L4–L5 have been shown to strongly correlate with the volume of the total visceral adipose tissue depot.20,21 Furthermore, Abate et al22 have shown that the single-slice approach has an even stronger predictive ability for the different body fat depots, if a slice at the level of L2–L3 is used. The main advantages of this method are its simplicity, reduced cost and analysis time. The disadvantages of the single-slice methods relate to inaccuracies in slice positioning and the fact that they assume visceral fat distribution within the abdominal cavity to be identical in all individuals.
In the present study, we compare single-slice MRI with multi-slice whole-body MRI to determine relative levels of visceral and abdominal adipose tissue content in a cohort of subjects with a large range of body adiposity. The results show that the choice of MRI protocol can significantly alter the apparent levels of adipose tissue content of a given subject, and may therefore have important implications for phenotypic measurements in genetic studies.
Subjects and methods
Written informed consent was obtained from all volunteers. Permission for this study was obtained from the Ethics Committee of the Hammersmith Hospital, London.
Two protocols were devised in order to determine the impact of undersampling on the quantitation of an individual's visceral adipose tissue content.
The first part of the study corresponds to a cross-sectional measurement of visceral adipose tissue in a large cohort of subjects by single-slice and whole-body MRI. A total of 59 premenopausal healthy female volunteers (mean age: 31.6±7.9 y; range: 18–46 y) covering a wide range of sizes (mean weight: 75.0±18.1 kg; range: 49.4–123 kg; mean BMI: 28.4±7.5 kg/m2; range: 17.7–51.9 kg/m2; mean total body adipose tissue content: 37.0±17.2 l; range: 13.8–86.3 l) participated in this study.
The second part of the study was designed to assess the ability of single-slice and whole-body methods to detect longitudinal changes in visceral fat content. A total of 17 nonobese, premenopausal healthy (previously nonexercising) female volunteers (mean age: 32.6±1.8 y; range: 25–45 y) were studied before and after 6 months of three times per week aerobic exercise. Each subject served as her own control for the purpose of this study. These subjects were part of a study investigating the effect of training on body fat content, the results of which were reported previously.23
All subjects were imaged in a prone position (to minimise respiratory motion) with their arms straight above the head in a Marconi 1.0 T HPQ system with a rapid T1 weighted spin-echo sequence (TR 36 ms, TE 14 ms). Images in the arms and legs were collected with two averages, and those in the torso with one average. Images were analysed using image segmentation software as previously described.11 Basically, the images were segmented and analysed using a software program that employed knowledge-based image processing to label pixels as fat and nonfat components. Each slice was manually reviewed using an interactive routine within the software program and unwanted voxels, such as those arising from fatty bowel content, deleted. The adipose tissue volumes (cm3) of each compartment were calculated by summing the relevant voxel counts and multiplying by the voxel dimensions in cubic centimetre3. The adipose tissue volume for the whole body was calculated by multiplying the adipose tissue volumes of each slice by the sum of the slice thickness (10 mm) and interslice distance. Note that this analysis provides a direct measure of the volume of adipose tissue rather than the quantity of triglyceride contained within the adipose tissue.11 Relative measurements of lipids can be ascertained from these if the values obtained by MRI are the relative density of adipose tissue (typically 0.9193 kg/l), although this makes the assumption that adipose tissue hydration is the same in all subjects and in all fat depots in the body.
Multi-slice whole-body imaging
Subjects were scanned from their fingertips to their toes by acquiring 10 mm thick transverse images with a 30 mm gap between slices in the arms and legs and a 10 mm gap in the torso.11 All images were acquired as single slices at the isocentre to avoid image distortion and the volunteers were moved through the magnet on a purpose-built platter.
From the MR images, particularly in subjects with a significant amount of intra-abdominal adipose tissue, it is not possible to differentiate between different depots of internal fat such as omental, mesenteric and retroperitoneal. There are generally no differences in the intensity of the different depots on the MR images and quite often there are no clear anatomical boundaries separating these different entities. For this reason, the total internal fat content for each subject was subdivided only into intra-abdominal and peripheral internal body fat. Total intra-abdominal adipose tissue content was obtained by quantifying voxels arising from adipose tissue in the slices from the femoral heads to the slice containing the top of the liver or the base of the lungs (T10). The mean number of slices thus covered for the volunteers in this study was 16.5±0.9, range 15–18 slices corresponding to a distance of 29–35 cm over the abdomen. All other internal fat was labelled as peripheral internal fat. Subcutaneous fat was similarly divided into ‘abdominal’ and peripheral fat.
Single transaxial slice imaging L2–L3
Using sagittal pilot images, transverse single slices were identified at the level of L2–L3 for each subject. This was then used to assess intra-abdominal and subcutaneous abdominal adipose tissue content.
Single transaxial slice imaging L4–L5
Again, using sagittal pilot images, transverse single slices were identified at the level of L4–L5 for each subject, which was then used to assess intra-abdominal and abdominal subcutaneous adipose tissue content.
All data are presented as mean±s.d. Differences before and after exercise were tested using the Student's paired t-test. Because of the relatively small number of subjects, and the possibility of a non-normal distribution, the data were also analysed using nonparametric statistics (Wilcoxon Signed Rank Test), with no difference in the results. Differences between the means of two groups of lean and obese volunteers were tested for using the Welch's t-test (not assuming equal variance), Pearson's product movement correlation coefficients (r) were used to assess the relation between variables. Repeated measures of ANOVA were performed for the single-slice data and the entire intra-abdominal fat depot to determine the influence of BMI and scanning regimen (n=59) and the exercise intervention (n=17) on the measurement of visceral adipose tissue. The level of significance was set at P<0.05.
Intra-abdominal and abdominal subcutaneous adipose tissue content for the whole cohort of volunteers (n=59), as measured by whole-body and single-slice (L4–L5 and L2–L3) MRI, are shown in Table 1. Intra-abdominal adipose tissue by whole-body MRI ranged from 0.47 to 8.04 l, with a mean value of 2.52±1.82 l. The content of abdominal subcutaneous adipose tissue for the same cohort of subjects was 8.89±5.49 l (range 2.02–24.63 l). The distribution of intra-abdominal adipose tissue measured by the whole-body MRI approach is shown in Figure 1a.
The relation between the single-slice methods and whole-body MRI measurement of intra-abdominal and subcutaneous abdominal adipose tissue are shown in Table 2. Taking the group as a whole, the correlation between the L4–L5 and whole-body MRI measurements of intra-abdominal fat was stronger (r=0.76; P<0.01) than that between L2–L3 and whole-body measurements (r=0.56; P<0.01). However, when the subjects were separated into two groups of lean (n=27; mean BMI=22.5±2.0 kg/m2, range 17.7–25.4 kg/m2) and overweight/obese individuals (n=32; mean BMI=33.3±6.9 kg/m2, range 25.7–51.9 kg/m2), some of these correlations were significantly altered. For lean subjects, the association between L2–L3 and whole-body measurements of intra-abdominal fat was found to be stronger than those observed for the group as a whole (r=0.88; P<0.01), but it was significantly weakened in overweight/obese subjects (r=0.33; P<0.05). The correlation between L4–L5 and whole-body measurements of intra-abdominal adipose tissue, in lean (r=0.67; P<0.01) and overweight (r=0.64; P<0.01) subjects, remained relatively unchanged. Similar results were observed for subcutaneous abdominal adiposity.
To further assess the potential effects of single-slice methods on fat quantitation, participants were ranked according to their relative intra-abdominal adipose tissue content. Again, taking whole-body MRI measurements as the standard, single-slice methods yielded significantly altered population distribution, with some subjects moving from the lower quartile to the median or upper quartile, Figures 1a–c. This reranking appeared to be independent of a subject's adiposity and equally pronounced for both L4–L5 and L2–L3. For example, subject ranked 32/59 by whole-body MRI (total intra-abdominal fat content 1.97 l), is reranked 58/59 by the single-slice L2–L3 measurement, that is, equivalent to 7.85 l of intra-abdominal fat. Likewise, the subject ranked 49/59 by the whole-body approach (4.22 l of intra-abdominal fat) was reranked 26/59 by the single-slice technique, equivalent to 1.62 l of intra-abdominal fat. A subset of these results are shown in Figure 2.
The ability of single-slice and whole-body MRI methods to detect longitudinal changes in intra-abdominal adipose tissue was also determined. Following 6 months exercise, a significant decrease in intra-abdominal adipose tissue was observed by both whole-body (−23.8±20.0%; P<0.01) and single-slice (L2–L3: −31.03±35.0%; P<0.01; L4–L5: −22.1±37.24%; P<0.05) MRI. However, there were some significant individual variations, with the single-slice techniques leading to an increase in data scattering, Figures 3a–c. Moreover, the magnitude of the changes appears to be related to the method used to assess adipose tissue content, with single-slice sampling at L2–L3 showing the greatest relative decrease in both intra-abdominal and subcutaneous abdominal adipose tissue content, compared with both whole-body MRI and a single slice at L4–L5, however using ANOVA, we found the interaction between scanning regimen and effect of exercise to be significant only for abdominal subcutaneous adipose tissue (Table 3).
Abdominal adiposity has become an important area of research, principally because of its association with insulin resistance. A number of imaging strategies have been adopted for its qualitative and quantitative assessment, ranging from whole-body to single-slice imaging. The former is generally accepted as the most accurate method, although strong correlation between single-slice and whole-body MRI of intra-abdominal adipose tissue has been previously reported.20,21,22 Single-slice imaging, MRI or CT, is increasingly used in human research and at present are being used as an adjunct to indirect methods such as UWW, BIA, DXA and anthropometry. In these instances, the indirect techniques are employed to measure the whole-body fat content, while single-slice imaging is used to provide an estimate of intra-abdominal fat content.16,17,18,19
Several groups have previously reported a significant correlation between single-slice and whole-body MRI visceral fat measurements. However, there appears to be some discrepancy as to whether single-slice at the L2–L3 or L4–L5 is the best predictor of total intra-abdominal adipose tissue.20,21,22 Abate et al22 found that the slice at the level of L2–L3 contained the highest amount of intra-peritoneal and retroperitoneal adipose tissue and was the best predictor of total intra-abdominal fat volume. However, Ross et al20 have reported almost identical correlation for the slice at the level of L4–L5 with intra-abdominal fat, and with the slices at 5, 10 and 15 cm above L4–L5. Similarly, Kvist et al21 found the slice at the level of L4–L5 to be the best predictor of intra-abdominal adipose tissue volume in both men and women. This apparent discrepancy in the published literature may reflect differences in the subject population. In the present study, we found that the correlation between single-slice and whole-body MRI, although significant, was highly dependent on the subject population. The findings in lean volunteers agree with those by Abate et al,22 while those arising from overweight/obese volunteers are in agreement with those reported by Ross et al20 and Kvist et al21, where L4–L5 appears as a better predictor of total intra-abdominal adipose tissue.
The reasons for the different relationship in lean and obese women are unclear and may relate to a number of factors including ethnicity, age, hormonal influences and/or genetic factors. In the present study, 59 healthy Caucasian female volunteers were studied with an average BMI of 22 kg/m2 in the lean subjects and an average BMI of 33 kg/m2 in the overweight/obese subjects. Ross studied men with a mean BMI of 28 kg/m2, Kvist studied both male and female (male BMI 31 kg/m2; female subjects BMI 26 kg/m2), and Abate studied men from a variety of ethnic groups, of which approximately 50% had NIDDM (BMI 28 kg/m2). Clearly, the discrepancy between the studies by Abate and those of Ross and Kvist may not be related to body size, but rather to ethnicity and disease state. Several groups have recently shown that BMI and body fat content and distribution are highly dependent on ethnicity.24,25 Thus, the predictive value of a single-slice MRI may not only depend on the subject adiposity, but also on factors such as ethnic background and genetics.
Physical constraints of the abdominal cavity on the intra-abdominal fat depot may also be a factor influencing the relation between a single-slice measure of intra-abdominal fat and the whole depot. During periods of weight gain, it is possible that initially intra-abdominal fat is deposited evenly throughout the abdominal cavity. As fat deposition continues to extremes of obesity, deposition of fat may no longer be possible at the level of L2–L3 because of physical constraints, such as the presence of the enlarged liver. Under those circumstances L4–L5 may become the main depository for fat, thereby accounting for its greater correlation with the total fat depot in obese subjects. Another possible source of variation may be associated with the constant movement of the small intestine/omentum.
Our results also reveal that despite the strong correlations, single-slice techniques showed a significant data spread postexercise intervention. The reasons for these differences are unclear, but may reflect the fact that single-slice images have intrinsically lower levels of signal compared to multi-slice images, making them more susceptible to changes in signal content because of potential misalignments during the follow-up scan. This in turn would lead to increased residual standard deviation, even though the mean value for a cohort of subjects may not be altered. In practical terms this means that, compared to multi-slice techniques, single-slice methods may be less likely to detect small changes in abdominal adiposity, or may require larger cohort of subjects to detect similar changes.
An interesting finding from our study is the fact that a greater reduction in subcutaneous adipose tissue was observed at L2–L3 compared with L4–L5, and even with whole-body data. The reason for this finding is unclear, but may reflect regional metabolic differences.
Another important observation arising from the present study is the fact that ranking of subjects according to their intra-abdominal fat content was significantly altered by single-slice approach, compared to multi-slice imaging. In general, single-slice imaging techniques rely on the assumption that although there are significant intersubject differences in abdominal fat content, their relative distribution is similar. In other words, even though subjects may have different absolute levels of abdominal adiposity, the relative proportions of fat at different levels within the depot are consistent between different subjects.
Although previous studies have shown significant regional variation in intra-abdominal fat within an individual, with L2–L3 level showing the highest intra-abdominal fat content, these studies did not report on interindividual variation in distribution of intra-abdominal fat within the abdominal cavity. We have shown in the present study that ranking of subjects, according to intra-abdominal fat content, was significantly altered if single-slice results were compared to whole-body imaging. This suggests that the intersubject variation in abdominal adiposity does not only relate to its absolute quantity but also to its distribution. If this is the case, as the results of the present study appear to suggest, intersubject comparison of intra-abdominal fat levels may be significantly compromised by the use of single-slice imaging. Interestingly, the reranking of subjects by single-slice method did not appear to be related to total body or abdominal adiposity, and may reflect regional differences in fat distribution arising from environmental (diet/exercise) and/or genetic factors. Clearly these findings require further investigation, as single-slice imaging techniques find increasing use in population studies, especially in genetic studies. Since the preparation of this paper, Greenfield et al26 have shown using CT the limitation of single-slice techniques for assessing intersubject comparison of abdominal fat.
In conclusion, assessment of intra-abdominal fat by MRI or CT can be significantly compromised by the imaging strategy chosen. Single-slice MRI appears to be suitable for assessing changes in intra-abdominal fat content in interventional studies, especially in large cohort of subjects, where each subject can serve as its own control. However, for accurate determination of an individual's intra-abdominal fat content, and inter-subject comparison, only multi-slice imaging will give precise results.
Fogelholm M, van Marken Lichtenbelt W . Comparison of body composition methods: a literature analysis. Eur J Clin Nutr 1997; 51: 495–503.
Deurenberg P, Yap M . The assessment of obesity: methods for measuring body fat and global prevalence of obesity. Baillieres Best Pract Res Clin Endocrinol Metab 1999;13: 1–11.
Smith SR, Zachwieja JJ . Visceral adipose tissue: a critical review of intervention strategies. Int J Obes Relat Metab Disord 1999; 23: 329–335.
Abate N, Garg A . Heterogeneity in adipose tissue metabolism: causes, implications and management of regional adiposity. Prog Lipid Res 1995; 34: 53–70.
Fowler PA, Fuller MF, Glasbey CA, Cameron GG, Foster MA . Validation of the in vivo measurement of adipose tissue by magnetic resonance imaging of lean and obese pigs. Am J Clin Nutr 1992; 56: 7–13.
Ross R, Leger L, Guardo R, De Guise J, Pike BG . Adipose tissue volume measured by magnetic resonance imaging and computerized tomography in rats. J Appl Physiol 1991; 70: 2164–2172.
Abate N, Burns D, Peshock RM, Garg A, Grundy SM . Estimation of adipose tissue mass by magnetic resonance imaging: validation against dissection in human cadavers. J Lipid Res 1994; 35: 1490–1496.
Fowler PA, Fuller MF, Glasbey CA, Foster MA, Cameron GG, McNeill G, Maughan RJ . Total and subcutaneous adipose tissue in women: the measurement of distribution and accurate prediction of quantity by using magnetic resonance imaging. Am J Clin Nutr 1991; 54: 18–25.
Ross R, Shaw KD, Martel Y, de Guise J, Avruch L . Adipose tissue distribution measured by magnetic resonance imaging in obese women. Am J Clin Nutr 1993; 57: 470–475.
Sohlstrom A, Wahlund LO, Forsum E . Adipose tissue distribution as assessed by magnetic resonance imaging and total body fat by magnetic resonance imaging, underwater weighing, and body-water dilution in healthy women. Am J Clin Nutr 1993; 58: 830–838.
Thomas EL, Saeed N, Hajnal JV, Brynes A, Goldstone AP, Frost G, Bell JD . Magnetic resonance imaging of total body fat. J Appl Physiol 1998; 85: 1778–1785.
Bonora E, Micciolo R, Ghiatas AA, Lancaster JL, Alyassin A, Muggeo M, DeFronzo RA . Is it possible to derive a reliable estimate of human visceral and subcutaneous abdominal adipose tissue from simple anthropometric measurements? Metabolism 1995; 44: 1617–1625.
Seidell JC, Bakker CJ, van der Kooy K . Imaging techniques for measuring adipose-tissue distribution—a comparison between computed tomography and 1.5-T magnetic resonance. Am J Clin Nutr 1990; 51: 953–957.
Gronemeyer SA, Steen RG, Kauffman WM, Reddick WE, Glass JO . Fast adipose tissue (FAT) assessment by MRI. Magn Reson Imaging 2000; 18: 815–818.
Abate N, Garg A, Peshock RM, Stray-Gundersen J, Adams-Huet B, Grundy SM . Relationship of generalized and regional adiposity to insulin sensitivity in men with NIDDM. Diabetes 1996; 45: 1684–1693.
Treuth MS, Hunter GR, Kekes-Szabo T, Weinsier RL, Goran MI, Berland L . Reduction in intra-abdominal adipose tissue after strength training in older women. J Appl Physiol 1995; 78: 1425–1431.
Gray DS, Fujioka K, Colletti PM, Kim H, Devine W, Cuyegkeng T, Pappas T . Magnetic-resonance imaging used for determining fat distribution in obesity and diabetes. Am J Clin Nutr 1991; 54: 623–627.
Marks SJ, Chin S, Strauss BJ . The metabolic effects of preferential reduction of visceral adipose tissue in abdominally obese men. Int J Obes Relat Metab Disord 1998; 22: 893–898.
Ronnemaa T, Koskenvuo M, Marniemi J, Koivunen T, Sajantila A, Rissanen A, Kaitsaari M, Bouchard C, Kaprio J . Glucose metabolism in identical twins discordant for obesity. The critical role of visceral fat. J Clin Endocrinol Metab 1997; 82: 383–387.
Ross R, Leger L, Morris D, de Guise J, Guardo R . Quantification of adipose tissue by MRI: relationship with anthropometric variables. J Appl Physiol 1992; 72: 787–795.
Kvist H, Chowdhury B, Grangard U, Tylen U, Sjostrom L . Total and visceral adipose-tissue volumes derived from measurements with computed tomography in adult men and women: predictive equations. Am J Clin Nutr 1988; 48: 1351–1361.
Abate N, Garg A, Coleman R, Grundy SM, Peshock RM . Prediction of total subcutaneous abdominal, intraperitoneal, and retroperitoneal adipose tissue masses in men by a single axial magnetic resonance imaging slice. Am J Clin Nutr 1997; 65: 403–408.
Thomas EL, Brynes AE, McCarthy J, Goldstone AP, Hajnal JV, Saeed N, Frost G, Bell JD . Preferential loss of visceral fat following aerobic exercise, measured by magnetic resonance imaging. Lipids 2000; 35: 769–776.
Prentice AM, Jebb SA . Beyond body mass index. Obes Rev 2001; 2: 141–147.
Raji A, Seely EW, Arky RA, Simonson DC . Body fat distribution and insulin resistance in healthy Asian Indians and Caucasians. J Clin Endocrinol Metab 2001; 86: 5366–5371.
Greenfield JR, Samaras K, Chisholm DJ, Campbell LV . Regional intra-subject variability in abdominal adiposity limits usefulness of computer tomography. Obes Res 2002; 10: 260–265.
Financial support from the Medical Research Council and Marconi Medical Systems is gratefully acknowledged. We thank Serena Counsell and Tracey Harrington for their help and Dr Kish Bhakoo for his useful discussions regarding this work. Finally, we thank all of the volunteers who took part in this study.
About this article
Cite this article
Thomas, E., Bell, J. Influence of undersampling on magnetic resonance imaging measurements of intra-abdominal adipose tissue. Int J Obes 27, 211–218 (2003). https://doi.org/10.1038/sj.ijo.802229
- intra-abdominal adipose tissue
- single slice
- whole body
<p>Relationship of Two-Hour Plasma Glucose and Abdominal Visceral Fat with Bone Mineral Density and Bone Mineral Content in Women with Different Glucose Metabolism Status</p>
Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy (2020)
Novel artificial neural network and linear regression based equation for estimating visceral adipose tissue volume
Clinical Nutrition (2020)
Body composition and its association with fatigue in the first 2 years after colorectal cancer diagnosis
Journal of Cancer Survivorship (2020)
European Journal of Clinical Nutrition (2020)
The “hyper-triglyceridemic waist phenotype” is a reliable marker for prediction of accumulation of abdominal visceral fat in Chinese adults
Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2020)