2025 Volume 72 Issue 3 Pages 307-317
The Japan Society for the Study of Obesity recommends a weight loss of 3% of body weight over a period of 3–6 months. However, the effects of rapid weight loss on the body composition have not yet been adequately studied. Therefore, we observed the changes in the body composition induced by rapid weight loss and its effects on the pathophysiological mechanisms involved in obesity. The subjects were people with obesity admitted to our institution. The goal was to achieve a 3–5% body weight loss in the subjects by combining a carbohydrate-controlled therapeutic diet of 25–30 kcal/day per kg target body weight, exercise therapy, and pharmacotherapy. The body composition was measured at admission and at discharge by the dual bioelectrical impedance analysis. After 2 weeks, the participants’ body weight decreased by 4.2%; the visceral fat area decreased by 16.7%, the subcutaneous fat area by 2.4%, and the lean area by 4.0%. The moderate weight loss, moderate energy restriction and adequate protein intake significantly reduced the visceral fat area while allowing the lean area to be preserved. Improvements were also noted in the peripheral white blood cell count and C-reactive protein level. However, no statistically significant changes in homeostasis model assessment for insulin resistance and the adiponectin level were noted. Regarding clinical parameters, improvements of the systolic and diastolic blood pressures, fasting plasma glucose, triglycerides, low-density lipoprotein cholesterol, and degree of microalbuminuria were observed. Short-term comprehensive treatment produced beneficial body composition changes, and improvements in the pathophysiological mechanisms involved in obesity.
The incidence of obesity is rising around the world [1]. Obesity is associated with various obesity-related health disorders, such as glucose intolerance, dyslipidemia, hypertension, coronary artery disease, fatty liver [2]. Therefore, development of effective management strategies for obesity is urgently needed.
Where fat accumulates in the body is important. Visceral fat accumulation leads to chronic inflammation and imbalances in adipokine secretion [3-5]. It causes insulin resistance which, in turn, leads to metabolic abnormalities. Previous reports have shown a positive correlation between increased visceral fat area (VFA) and the complication rate associated with cardiovascular risk factors such as hyperglycemia, dyslipidemia, and elevated blood pressure [6]. Therefore, evaluation of the VFA is important for assessing the various complications associated with obesity, including cardiovascular and other complications risks.
The Japan Society for the Study of Obesity recommends a weight loss of 3% of body weight over a period of 3–6 months [7]. This was based on evidence provided by a study that metabolic parameters, such as blood glucose, serum lipid levels and blood pressure, improved in subjects in whom a 3% weight loss was achieved over a 6-month period [8]. Although various guidelines recommend measurement of body weight as a simple indicator for evaluating obesity, body weight does not accurately assess whether the weight loss in a given subject is beneficial, not beneficial or even harmful as it includes not only visceral fat mass, but also subcutaneous fat mass and skeletal muscle mass. Since obesity-related health disorders are strongly associated with the accumulation of visceral fat [9-11], it is important to prioritize reduction of visceral fat in people being treated for obesity [12]. Preservation of the muscle mass while attempting to achieve body weight loss is also important for energy expenditure, energy storage, and avoiding sarcopenia [13]. Hence, the appropriate goal of weight loss is to achieve loss of visceral fat preferentially over loss of muscle mass.
Furthermore, the effects of rapid weight loss on the body composition and metabolic parameters have not yet been adequately studied. Most previous body composition research has covered monthly or yearly periods. This is due to the problem of radiation exposure since VFA is measured by CT, making it difficult to measure VFA over a short period of time.
The purpose of this study is to examine the changes in the body composition associated with rapid weight loss resulting from comprehensive treatment of obesity with a body weight loss goal of 3–5% in two weeks, using bioelectrical impedance analysis. Furthermore, we also attempted to evaluate the impact of rapid weight loss on the pathophysiology of obesity by assessing changes in clinical endpoints such as insulin resistance, inflammatory marker levels, and serum adiponectin levels.
This was a prospective single-center observational study. The subjects were consecutive people with obesity who were hospitalized at the Department of Diabetes and Metabolic Diseases, The University of Tokyo Hospital, between August 16 and November 15, 2021. People with obesity who were assessed by outpatient physicians as requiring weight loss through hospitalization were admitted to the hospital. The subjects were people with a body mass index (BMI) of 25 kg/m2 or more in the age range of 20–80 years. The exclusion criteria are shown in Fig. 1. People who were hospitalized for less than 7 days and people who could not comply with the protocol of this study due to their medical schedule were excluded. People with inflammatory diseases (rheumatic diseases and chronic infectious diseases) were excluded from examination of the changes in the inflammatory responses after hospitalization as compared with the values measured before. People with secondary obesity, post-laparoscopic sleeve gastrectomy, and unstable Basedow’s disease, all of which could have an influence on weight loss, were excluded. In addition, pregnant women, people with heart failure, severe edema, and/or severe abdominal obesity (waist circumference ≥140 cm, as such people cannot be fitted with a visceral fat measurement device) were also excluded. As the impedance method is affected by water content, it cannot be used in such cases. This study received ethical approval from the Research Ethics Committee of the University of Tokyo Hospital [Approval No. 3659]. Informed consent was obtained from all the participants.
Participants were hospitalized and treated by therapeutic modification of the diet, exercise, and medication, with the aim of achieving a 3–5% loss of body weight over a period of 2 weeks. The energy content of the diet was adjusted to 25–30 kcal/day per kg of the target body weight (TBW). In regard to the nutritional composition of the diet, we provided a diet containing 40–60% carbohydrates (hereafter, 40–60% carbohydrate-controlled diet). We first determined the content of carbohydrates in the diet, then increased the amount of protein to compensate for any decrease of the carbohydrate content and fixed the amounts of the other nutrients (Supplementary Table 1). The procedure adopted for determining the amount of carbohydrates in the diet is shown in Supplementary Table 2. For people with renal dysfunction, the energy level is established first, and then the protein level is determined subsequently in accordance with the guidelines for chronic kidney disease (CKD) [14]. For people with CKD G3 who also had sarcopenia or frailty, a target of 1.3 g/TBW/day is set as the upper limit for the daily protein intake. This recommendation was adopted in the present study. If the participant was severely obese or the weight loss stopped midway, the diet was modified to provide an energy level of 20 kcal/TBW/day or to a 40% carbohydrate-controlled diet. Exercise therapy was prescribed, including walking, ergometer, and resistance exercises. Participants who did not have any exercise restrictions were encouraged to walk a minimum of 10,000 steps/day and/or to use an ergometer for a minimum of 30 minutes/day. Resistance exercises such as twisting the body with a ball and squats were recommended at a total of 3 sets of 10 repetitions each, 2–3 times a week. For participants with exercise restrictions due to musculoskeletal disorders, etc., the target exercise levels were set on an individual basis by the attending physician. For participants with type 2 diabetes, the insulin secretion capacity was evaluated, and the medications were adjusted accordingly. The subjects also received antihypertensive, lipid-lowering therapies and antihyperuricemic drugs, as needed, according to the guidelines.
MeasurementsWe measured the VFA, subcutaneous fat area (SFA), and lean area (LA) at admission and at discharge in all the subjects. These body composition parameters were measured with DUALSCAN HDS-2000 (Omron Healthcare Co., Ltd, Japan), which utilizes a measurement principle based on the dual bioelectrical impedance analysis [15, 16]. VFA determined by this method has been shown to be strongly correlated with the VFA measured by CT [15, 17]. The participants’ height, weight, and waist circumference were measured at admission. The waist circumference was measured at the level of the umbilicus in accordance with the measurement standards of the Japan Society for the Study of Obesity [7]. The body weight and blood pressure of the participants were also recorded daily before breakfast. Blood and urine samples were collected in the early morning on the day of admission and the day of discharge. Blood and urine tests included measurement of the peripheral white blood cell (WBC) count, C-reactive protein (CRP), fasting plasma glucose (FPG), fasting plasma insulin, hemoglobin a1c (HbA1c), serum total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides (TG), aspartate aminotransferase (AST), alanine aminotransferase (ALT), and γ-glutamyl trans peptidase (γ-GTP), estimated glomerular filtration date (eGFR), and the urinary albumin-creatinine ratio (ACR). Serum adiponectin was measured at admission and at discharge in some of the participants. The homeostatic model assessment for insulin resistance (HOMA-IR) was calculated in participants who were not on insulin by multiplying the fasting blood glucose (mg/dL) and fasting plasma insulin (μIU/mL) and dividing the product by 405 [18]. Low-density lipoprotein cholesterol (cLDL cholesterol) was calculated using the Friedewald equation [19]. Non-HDL cholesterol was calculated as total cholesterol (mg/dL) minus HDL cholesterol (mg/dL). The fibrosis-4 index (FIB-4 Index) was calculated as follows: (age [in years] × AST [IU/L])/(platelet [109/L] × √ALT [IU/L]) [20].
Statistical MethodsStatistical analysis was performed using GraphPad Prism ver. 9.5.1 for Mac (GraphPad Software). Descriptive statistics are presented as the number of cases (proportion) for categorical variables or as median values (interquartile range [IQR]) for continuous variables. Because the sample size was small, the Wilcoxon signed rank test was used for paired comparisons, while the Mann-Whitney’s U test was used for unpaired comparisons. The Kruskal-Wallis test was used as the significance test for multigroup comparisons, with post-hoc analyses performed using Dunn’s multiple comparisons. Fisher’s exact probability test was used for the analysis of categorical variables.
Table 1 shows the characteristics of the 29 participants in this study. The subject’s body mass index (BMI) was 32.6 (27.6 to 36.4) kg/m2. The body composition parameters at admission were as follows: VFA, 145.9 (107.3 to 181.1) cm2; SFA, 307.6 (232.8 to 371.7) cm2; and LA, 291.3 (273.5 to 321.6) cm2. The percentage of participants with impaired glucose tolerance, type 1 diabetes, type 2 diabetes was 10.3%, 3.4%, 82.8%, respectively. The rates of use of insulin, sulfonylureas, and thiazolidinediones in the subjects were 13.8%, 3.4%, and 3.4%, respectively. On the other hand, the rates of use of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) and sodium-glucose cotransporter 2 inhibitors (SGLT2 inhibitors) were 20.7% and 41.4%, respectively. Blood tests showed mildly elevated CRP 0.25 (0.08 to 0.41) mg/dL in the subjects. The HbA1c was elevated to 8.0 (7.0 to 9.2)% and the HOMA-IR was 3.1 (1.7 to 4.0). The category distribution of eGFR is shown in Supplementary Table 3.
| Variables | Overall | Male | Female | p |
|---|---|---|---|---|
| n | 29 | 15 (51.7) | 14 (48.3) | — |
| Age (years) | 55 (49 to 67) | 57 (49 to 64) | 53 (46 to 68) | 0.628 |
| Weight (kg) | 84.0 (75.0 to 99.3) | 92.6 (78.4 to 113.0) | 79.3 (70.0 to 96.0) | 0.079 |
| BMI (kg/m2) | 32.6 (27.6 to 36.4) | 31.7 (28.1 to 36.3) | 32.6 (27.1 to 36.9) | 0.872 |
| Waist Circumference (cm) | 105 (98 to 115) | 105 (99 to 126) | 106 (95 to 113) | 0.381 |
| VFA (cm2) | 145.9 (107.3 to 181.1) | 164.7 (132.4 to 221.2) | 142.0 (103.1 to 156.9) | 0.172 |
| SFA (cm2) | 307.6 (232.8 to 371.7) | 326.1 (225.9 to 372.4) | 282.0 (235.9 to 368.9) | 0.652 |
| LA (cm2) | 291.3 (273.5 to 321.6) | 305.1 (276.0 to 355.7) | 281.5 (259.1 to 304.9) | 0.125 |
| Obesity-related health disorders | ||||
| Glucose intolerance | 28 (96.6) | 15 (100.0) | 13 (92.9) | 0.483 |
| Impaired glucose tolerance | 3 (10.3) | 2 (13.3) | 1 (7.1) | >0.999 |
| Type 1 diabetes | 1 (3.4) | 1 (6.7) | 0 (0) | >0.999 |
| Type 2 diabetes | 24 (82.8) | 12 (80.0) | 12 (85.7) | >0.999 |
| Hypertension | 19 (65.5) | 11 (73.3) | 8 (57.1) | 0.450 |
| Dyslipidemia | 26 (89.7) | 14 (93.3) | 12 (85.7) | 0.598 |
| Hyperuricemia | 5 (17.2) | 2 (13.3) | 3 (21.4) | 0.651 |
| Coronary artery disease | 1 (3.4) | 1 (6.7) | 0 (0) | >0.999 |
| Cerebral infarction | 2 (6.9) | 1 (6.7) | 1 (7.1) | >0.999 |
| MASLD | 26 (89.7) | 13 (86.7) | 13 (92.9) | >0.999 |
| MetALD | 3 (10.3) | 2 (13.3) | 1 (7.1) | >0.999 |
| Microalbuminuria | 14 (48.3) | 10 (66.7) | 4 (28.6) | 0.066 |
| Medications at admission | ||||
| Insulin | 4 (13.8) | 4 (26.7) | 0 (0) | 0.100 |
| Sulfonylureas | 1 (3.4) | 1 (6.7) | 0 (0) | >0.999 |
| Glinides | 0 (0) | 0 (0) | 0 (0) | — |
| GLP-1 RAs | 6 (20.7) | 1 (6.7) | 5 (35.7) | 0.080 |
| DPP-4 inhibitors | 11 (37.9) | 7 (46.7) | 4 (28.6) | 0.450 |
| Biguanides | 14 (48.3) | 7 (46.7) | 7 (50.0) | >0.999 |
| Thiazolidinediones | 1 (3.4) | 1 (6.7) | 0 (0) | >0.999 |
| SGLT2 inhibitors | 12 (41.4) | 6 (40.0) | 6 (42.9) | >0.999 |
| α-glucosidase inhibitors | 4 (13.8) | 3 (20.0) | 1 (7.1) | 0.598 |
| Antihypertensive drugs | 15 (51.7) | 9 (60.0) | 6 (42.9) | 0.466 |
| Lipid-lowering drugs | 14 (48.3) | 7 (46.7) | 7 (50.0) | >0.999 |
| Antihyperuricemic drugs | 1 (3.4) | 0 (0) | 1 (7.1) | 0.483 |
| sBP (mmHg) | 124 (110 to 130) | 122 (112 to 128) | 124 (108 to 133) | 0.855 |
| dBP (mmHg) | 70 (63 to 80) | 68 (62 to78) | 74 (64 to 83) | 0.236 |
| WBC (×103/μL) | 6.6 (5.1 to 8.3) | 6.5 (4.5 to 9.4) | 6.9 (5.8 to 8.0) | 0.872 |
| CRP (mg/dL) | 0.25 (0.08 to 0.41) | 0.26 (0.04 to 0.42) | 0.22 (0.13 to 0.39) | 0.822 |
| FPG (mg/dL) | 142 (99 to 161) | 142 (119 to 162) | 139 (98 to 157) | 0.739 |
| Fasting plasma insulin (μIU/mL) | 8 (7 to 13) (n = 25) | 9 (5 to 13) (n = 11) | 8 (7 to 14) (n = 14) | 0.818 |
| HOMA-IR | 3.1 (1.7 to 4.0) (n = 25) | 3.1 (1.4 to 4.1) (n = 11) | 2.9 (1.7 to 4.2) (n = 14) | 0.840 |
| HbA1c (%) | 8.0 (7.0 to 9.2) | 8.2 (7.5 to 11.0) | 7.7 (6.5 to 8.5) | 0.125 |
| Total cholesterol (mg/dL) | 193 (160 to 218) | 180 (159 to 211) | 196 (172 to 223) | 0.382 |
| HDL cholesterol (mg/dL) | 46.4 (36.9 to 53.1) | 37.5 (33.7 to 46.5) | 50.4 (42.5 to 60.5) | 0.003 |
| Triglycerides (mg/dL) | 153 (103 to 202) | 185 (82 to 233) | 144 (107 to 171) | 0.477 |
| cLDL cholesterol (mg/dL) | 109 (84 to 139) (n = 28) | 109 (79 to 140) (n = 15) | 108 (85 to 140) (n = 13) | 0.717 |
| Non-HDL cholesterol (mg/dL) | 145 (117 to 170) | 144 (115 to 174) | 146 (116 to 168) | 0.747 |
| Uric acid (mg/dL) | 5.4 (4.5 to 6.4) | 5.5 (4.4 to 6.7) | 5.3 (4.5 to 5.7) | 0.497 |
| AST (IU/L) | 31 (21 to 43) | 35 (28 to 47) | 24 (20 to 41) | 0.087 |
| ALT (IU/L) | 48 (22 to 67) | 53 (30 to 69) | 25 (20 to 63) | 0.285 |
| γ-GTP (IU/L) | 41 (27 to 89) | 41 (22 to 98) | 43 (27 to 73) | 0.723 |
| FIB-4 index | 1.26 (0.72 to 1.69) | 1.39 (1.18 to 2.05) | 1.25 (0.62 to 1.45) | 0.148 |
| eGFR (mL/min/1.73 m2) | 76.6 (64.9 to 89.0) | 76.6 (64.1 to 89.6) | 77.4 (64.9 to 89.3) | 0.974 |
| U-ACR (mg/gCre) | 23 (10 to 101) | 37 (12 to 343) | 19 (9 to 38) | 0.095 |
| Adiponectin (μg/mL) | 5.2 (3.2 to 7.8) (n = 12) | 4.6 (3.0 to 7.4) (n = 5) | 5.7 (3.0 to 9.7) (n = 7) | 0.561 |
Data are presented as n (%) or medians (interquartile range). Microalbuminuria was defined as a urine albumin level of ≥30 mg/gCre. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; cLDL cholesterol, calculated low density lipoprotein cholesterol; CRP, C-reactive protein; dBP, diastolic blood pressure; DPP-4 inhibitors, dipeptidyl peptidase-4 inhibitors; eGFR, estimated glomerular filtration rate; FIB-4 index, fibrosis-4 index; FPG, fasting plasma glucose; γ-GTP, gamma-glutamyl transpeptidase; GLP-1 RAs, glucagon-like peptide-1 receptor agonists; HbA1c, hemoglobin A1c; HDL cholesterol, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment for insulin resistance; LA, lean area; MASLD, metabolic dysfunction–associated steatotic liver disease; MetALD, metabolic and alcohol related/associated liver disease; sBP, systolic blood pressure; SFA, subcutaneous fat area; SGLT2 inhibitors, sodium-glucose cotransporter-2 inhibitors; U-ACR, urinary albumin-creatinine ratio; VFA, visceral fat area; WBC, peripheral white blood cell count.
We compared the changes in the body composition parameters and clinical parameters from admission to discharge (Table 2, Fig. 2). The median duration of hospitalization was 13 (11 to 14) days. The participants were prescribed a diet with an energy intake of 25.8 (24.2 to 27.5) kcal/TBW/day. The percentages of participants on 40%, 50%, and 60% carbohydrate-controlled diets were 6.9%, 79.3%, and 13.8%, respectively. The median protein intake was 1.2 (1.2 to 1.4) g/TBW/day. Of all the participants included in the study, 58.6% walked less than 5,000 steps/day, 20.7% walked between 5,000 and 9,999 steps/day, and 20.7% walked more than 10,000 steps/day; 44.8% of the participants performed ergometer exercises (Supplementary Table 4) and 17.2% of the participants engaged in resistance training. During the study period, the participants showed rather poor adherence with the prescribed exercise therapy, possibly attributable to the restrictions on activities imposed during the participants’ hospitalization during the COVID-19 pandemic.
| Variables | Overall (n = 29) | |||
|---|---|---|---|---|
| Hospital days (day) | 13 (11 to 14) | |||
| Diet therapy | ||||
| Energy intake (kcal/TBW/day) | 25.8 (24.2 to 27.5) | |||
| Carbohydrate-controlled diet (%) | ||||
| 40% | 2 (6.9) | |||
| 50% | 23 (79.3) | |||
| 60% | 4 (13.8) | |||
| Protein intake (g/TBW/day) | 1.2 (1.2 to 1.4) | |||
| Exercise therapy | ||||
| Walking (%) | ||||
| <5,000 steps/day | 17 (58.6) | |||
| 5,000–9,999 steps/day | 6 (20.7) | |||
| 10,000 steps/day ≤ | 6 (20.7) | |||
| Ergometer exercise (%) | 13 (44.8) | |||
| Resistance exercise (%) | 5 (17.2) | |||
| Medications | Admission | Discharge | p | |
| Insulin | 4 (13.8) | 3 (10.3) | >0.999 | |
| Sulfonylureas | 1 (3.4) | 0 (0) | >0.999 | |
| Glinides | 0 (0) | 0 (0) | — | |
| DPP-4 inhibitors | 11 (37.9) | 3 (10.3) | 0.030 | |
| GLP-1 RAs | 6 (20.7) | 17 (58.6) | 0.007 | |
| Biguanides | 14 (48.3) | 17 (58.6) | 0.599 | |
| Thiazolidinediones | 1 (3.4) | 0 (0) | >0.999 | |
| SGLT2 inhibitors | 12 (41.4) | 14 (48.3) | 0.792 | |
| α-glucosidase inhibitors | 4 (13.8) | 1 (3.4) | 0.353 | |
| Antihypertensive drugs | 15 (51.7) | 12 (41.4) | 0.599 | |
| Lipid-lowering drugs | 14 (48.3) | 16 (55.2) | 0.793 | |
| Antihyperuricemic drugs | 1 (3.4) | 3 (10.3) | 0.612 | |
| Variables | n | Admission | Discharge | Median differences (95% confidence interval) |
| Weight (kg) | 29 | 84.0 (75.0 to 99.3) | 81.3 (72.0 to 95.8) | –3.6 (–4.3 to –2.7) |
| VFA (cm2) | 29 | 145.9 (107.3 to 181.1) | 118.1 (94.7 to 166.4) | –19.0 (–38.7 to –8.1) |
| SFA (cm2) | 29 | 307.6 (232.8 to 371.7) | 293.0 (235.0 to 359.8) | –6.6 (–14.5 to 3.00) |
| LA (cm2) | 29 | 291.3 (273.5 to 321.6) | 279.6 (262.0 to 305.5) | –10.0 (–18.5 to –1.2) |
| sBP (mmHg) | 29 | 124 (110 to 130) | 110 (106 to 122) | –6 (–16 to 0) |
| dBP (mmHg) | 29 | 70 (63 to 80) | 66 (60 to 68) | –4 (–10 to 0) |
| WBC (×103/μL) | 29 | 6.6 (5.1 to 8.3) | 5.8 (4.7 to 7.1) | –0.6 (–1.3 to –0.2) |
| CRP (mg/dL) | 28 | 0.26 (0.08 to 0.41) | 0.11 (0.07 to 0.19) | –0.10 (–0.16 to 0) |
| FPG (mg/dL) | 28 | 138 (98 to 161) | 106 (90 to 119) | –27 (–46 to –9) |
| Fasting plasma insulin (μIU/mL) | 23 | 8 (7 to 13) | 9 (5 to 13) | –0.9 (–1.3 to 1.1) |
| HOMA-IR | 23 | 3.1 (1.5 to 4.1) | 2.2 (1.4 to 3.5) | –0.6 (–1.4 to 0.1) |
| Total cholesterol (mg/dL) | 28 | 193 (160 to 221) | 152 (128 to 184) | –31 (–46 to –21) |
| HDL cholesterol (mg/dL) | 29 | 46.4 (36.9 to 53.1) | 43.8 (34.3 to 51.2) | –2.6 (–4.9 to 1.5) |
| Triglycerides (mg/dL) | 29 | 153 (103 to 202) | 110 (80 to 171) | –41 (–52 to –14) |
| cLDL cholesterol (mg/dL) | 27 | 108.2 (83.9 to 138.6) | 84.2 (59.0 to 112.8) | –22.1 (–30.1 to –12.1) |
| Non-HDL cholesterol (mg/dL) | 28 | 145 (116 to 166) | 112 (84 to 132) | –31 (–39 to –19) |
| Uric acid (mg/dL) | 29 | 5.4 (4.5 to 6.4) | 6.2 (5.3 to 7.6) | 0.7 (0.4 to 1.3) |
| AST (IU/L) | 29 | 31 (21 to 43) | 37 (21 to 54) | 0 (–4 to 6) |
| ALT (IU/L) | 29 | 48 (22 to 67) | 58 (25 to 81) | 4 (2 to 12) |
| γ-GTP (IU/L) | 29 | 41 (27 to 89) | 36 (22 to 73) | –6 (–17 to –5) |
| eGFR (mL/min/1.73 m2) | 29 | 76.6 (64.9 to 89.0) | 77.0 (62.9 to 84.0) | –5.8 (–8.7 to –1.3) |
| U-ACR (mg/gCre) | 28 | 27 (10 to 106) | 11 (5 to 40) | –14 (–45 to –3) |
| Adiponectin (μg/mL) | 12 | 5.2 (3.2 to 7.8) | 4.8 (2.9 to 6.8) | –0.3 (–2.1 to 0.1) |
Data are presented as n (%) or medians (interquartile range). ALT, alanine aminotransferase; AST, aspartate aminotransferase; cLDL cholesterol, calculated low density lipoprotein cholesterol; CRP, C-reactive protein; dBP, diastolic blood pressure; DPP-4 inhibitors, dipeptidyl peptidase-4 inhibitors; eGFR, estimated glomerular filtration rate; FPG, fasting plasma glucose; γ-GTP, gamma-glutamyl transpeptidase; GLP-1 RAs, glucagon-like peptide-1 receptor agonists; HDL cholesterol, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment for insulin resistance; LA, lean area; sBP, systolic blood pressure; SFA, subcutaneous fat area; SGLT2 inhibitors, sodium-glucosecotransporter-2 inhibitors; TBW, target body weight; U-ACR, urinary albumin-creatinine ratio; VFA, visceral fat area; WBC, peripheral white blood cell count.
A: Change in body weight. B: Change in visceral fat area. C: Change in subcutaneous fat area. D: Change in lean area. E: Percent change in body composition. F: Changes in the VFA in subjects <65 years old (y) and ≥65 y. G: Changes in the SFA in subjects <65 y and ≥65 y. H: Changes in the LA in subjects <65 y and ≥65 y.
Graphs show the medians and interquartile ranges.
LA, lean area; SFA, subcutaneous fat area; VFA, visceral fat area.
Some participants were started on a GLP-1 RA during hospitalization (rate of GLP-1 RA use: 6 participants at admission [20.7%] vs. 17 participants at discharge [58.6%], p = 0.007). The details of the prescribed GLP-1 RAs at discharge were as follows: 15 participants (51.7%) were receiving semaglutide and 2 participants (6.9%) were receiving dulaglutide (Supplementary Table 5). Dipeptidyl peptidase-4 inhibitors (DPP-4 inhibitors) were discontinued in some participants during hospitalization, which was thought to be due to the start of GLP-1 RA administration in the participants.
Fig. 2 shows the differences in the body composition between admission and discharge. The body weight decreased significantly during hospitalization (median of differences, –3.6 kg; 95% CI [confidence interval], –4.3 to –2.7 kg). The VFA and LA also decreased significantly during hospitalization (VFA: median of differences, –19.0 cm2; 95% CI, –38.7 to –8.1 cm2; LA: median of differences, –10.0 cm2; 95% CI, –18.5 to –1.2 cm2). There was no significant difference in SFA between admission and discharge. Importantly, comparison of the percent changes in the parameters revealed that VFA by –16.7% (–25.9 to –4.0%), the SFA changed by –2.4% (–5.2 to 3.1%), and the LA by –4.0% (–7.8 to –0.2%), suggesting a greater percent decrease of the VFA as compared with that of the SFA and LA (Fig. 2E, Supplementary Table 6). On the other hand, comparison of the changes in the body composition between participants who were <65 years old (y.o) and ≥65 y.o revealed no differences in the percent decrease of the VFA or LA between the two groups (change of VFA: median of differences, 8.3 cm2; 95% CI, –17.8 to 25.2 cm2; change of LA: median of differences, –12.2 cm2; 95% CI, –21.9 to 10.0 cm2) (Supplementary Table 7).
We compared the clinical data at admission and at discharge (Table 2). The systolic blood pressure (sBP) and diastolic blood pressure (dBP) at discharge tended to be lower than the values at admission (sBP: median of differences, –6 mmHg; 95% CI, –16 to 0 mmHg; dBP: median of differences, –4 mmHg; 95% CI, –10 to 0 mmHg). The peripheral WBC count was significantly lower at the time of discharge, and the CRP tended to be lower (peripheral WBC count: median of differences, –0.6 × 103/μL; 95% CI, –1.3 × 103 to –0.2 × 103/μL; CRP: median of differences, –0.10 mg/dL; 95% CI, –0.16 to 0 mg/dL). The FPG was also significantly lower at discharge (median of differences, –27 mg/dL; 95% CI, –46 to –9 mg/dL). There was no statistically significant change in the HOMA-IR after hospitalization as compared with that measured before (median of differences, –0.6; 95% CI, –1.4 to 0.1). The changes in the total daily insulin dose (TDD) are shown in Supplementary Table 8. The TDD decreased in the study participants and some participants, except for those with type 1 diabetes and those who desired to become pregnant, could stop their insulin. The TG and cLDL cholesterol were significantly lower at discharge (TG: median of differences, –41 mg/dL; 95% CI, –52 to –14 mg/dL; cLDL cholesterol: median of differences, –22.1 mg/dL; 95% CI, –30.1 to –12.1 mg/dL). Significant increase of the serum uric acid was observed (median of differences, 0.7 mg/dL; 95% CI, 0.4 to 1.3 mg/dL). The eGFR decreased significantly (median of differences, –5.8 mL/min/1.73 m2; 95% CI, –8.7 to –1.3 mL/min/1.73 m2). The ACR was significantly lower at discharge (median of differences, –14 mg/dL; 95% CI, –45 to –3 mg/dL). No statistically significant change in the adiponectin level was observed between admission and discharge (median of differences, –0.3 μg/mL; 95% CI, –2.1 to 0.1 μg/mL).
In this study, we studied the effects of short-term comprehensive therapy administered with the goal of achieving weight loss over the short term on the body composition and pathophysiological mechanisms involved in obesity. Our hospitalized study subjects showed a 4.2% body weight loss after 2 weeks of inpatient treatment, with significant loss of the VFA and minimal loss of the muscle mass. Furthermore, we also observed improvements in the peripheral WBC count, CRP which led us to conclude that this change in the body composition was associated with improvements of the important pathophysiological parameters involved in the development of obesity.
Most importantly, a 4.2% loss of body weight was achieved after only 2 weeks of inpatient treatment in the subjects of our study, with preferential loss of the VFA over that of muscle mass. There have been few reports on the changes in the body composition associated with rapid weight loss. Our results suggest that it is possible to achieve beneficial changes of the body composition with intensive weight loss over a short period of time. It has been reported that visceral fat and subcutaneous fat differ in their adrenergic receptor sensitivity, with visceral fat expressing more lipolytic β1, β2, and β3 adrenergic receptors and less lipolytic inhibitory α2 adrenergic receptors than subcutaneous fat [21-23]. Therefore, visceral fat is thought to show more rapid lipolysis than subcutaneous fat. However, the methods to control changes in the body composition associated with weight loss have not yet been adequately studied. Previous systematic reviews have reported that the relative rate of reduction in visceral fat is associated with the rate of weight loss [24]. If the rate of weight loss is about 5%, the amount of visceral fat is likely to decrease more relative to the amount of subcutaneous fat. On the other hand, a larger rate of weight loss tends to result in a smaller rate of visceral fat loss and larger rate of subcutaneous fat loss. In our study, the average percent weight loss was 4.2%, suggesting that a preferential decrease of visceral fat over subcutaneous fat is observed with moderate weight loss rates. Another systematic review showed that a greater degree of reduction of visceral fat was obtained with exercise alone or combined exercise plus diet therapy than with diet therapy alone [25]. This is thought to be due to the activation of adrenergic signaling by exercise therapy, which facilitates the reduction of visceral fat.
Our study also showed that the rate of decrease in the skeletal muscle mass was much smaller than the rate of decrease of the VFA. Since there have been few studies on the changes in the body composition associated with rapid weight loss, there has been concern that significant weight loss within a short period of time could also result in a loss of muscle mass along with that of the fat mass. Therefore, it was very important that, by using an appropriate weight loss method, rapid weight loss was associated with a preferential reduction of visceral fat while maintaining muscle mass. We prescribed moderate energy restriction, adequate protein intake, and exercise therapy to prevent participants from losing muscle mass. According to a previous report, when weight loss from very low calorie diets (500 kcal/day) and low calorie diets (1,250 kcal/day) dietary interventions was compared, the rate of lean mass loss was smaller with low calorie diets, even though the rate of weight loss was the same [26]. In this study, the energy intake was set at 25–30 kcal/TBW/day (the actual daily intake was 1,200–1,800 kcal/day). We considered mild catabolism due to the mild energy restriction as the reason for the maintained muscle mass. We provided carbohydrate-controlled diets to achieve sufficient weight loss, but the proportion of dietary protein content was increased in proportion to the decrease in the dietary carbohydrate content. The participants’ protein intake was 1.2 (1.2 to 1.4) g/TBW/day. Appropriate protein intake is thought to have contributed to the maintenance of the muscle mass in the subjects. In addition, our weight loss program includes a combination of aerobic and resistance exercises. However, unfortunately, due to the impact of the restrictions on activities during hospitalization imposed owing to the prevailing COVID-19 pandemic, the adherence rate to the exercise therapy was low. If the adherence to exercise therapy had been higher, we may have observed even better maintenance of the muscle mass. On the other hand, some people with obesity are unable to exercise sufficiently due to musculoskeletal disorders or heart failure. The results of this study suggest that even in such people, it may be possible to prevent a decrease in the lean body mass if the subjects participated in this weight loss program involving consumption of a moderate amount of energy and adequate amount of protein.
Along with the change in the body composition, the TDD decreased in insulin users, and the serum TG, cLDL, and U-ACR levels improved, suggesting improvement also of markers of obesity-related diseases. On the other hand, the eGFR decreased. The decrease in eGFR is considered to be derived from dehydration and administration of SGLT2 inhibitors. The fact that the participants were restricted to drink only water and tea and were not allowed to drink soft drinks or other types of beverages was considered as the reason for the reduced fluid intake. The increase in the serum uric acid levels was also significant. We considered this increase of the serum uric acid as being attributable to dehydration and catabolism. Due to the energy restriction and increased energy expenditure, catabolism was accelerated and ketone bodies were produced. It is known that inhibition of tubular transport of ketone bodies reduces the urinary excretion of uric acid, thereby increasing the blood level of uric acid [27]. To achieve rapid weight loss, it is necessary to ensure adequate fluid intake and avoid extreme carbohydrate restriction.
We showed that short-term comprehensive therapy improved inflammation, suggesting that rapid weight loss was associated with improvement of the pathophysiology of obesity. No significant improvement of the HOMA-IR was observed, but we consider this as being attributable to the small sample size. On the other hand, we did not note any improvement of the adiponectin levels in the subjects. Serum adiponectin blood levels have been reported to improve with a 5–10% or more body weight loss [28, 29]. Also, adiponectin is secreted from subcutaneous fat, so that a decrease of the serum adiponectin levels may be observed with a decrease of the subcutaneous fat mass. Furthermore, it is possible that it takes longer to observe improvement of the serum adiponectin levels in association with weight loss. Further studies are needed to clarify the effects of changes of the body composition on the serum adiponectin levels.
In this study, the number of prescriptions for GLP-1 RAs increased significantly by the time of discharge. Of all the participants, 51.7% were prescribed semaglutide, which has the effect of reducing the visceral fat area, and it was thought that use of this drug also possibly contributed to the visceral fat reduction in the study participants.
This study had some limitations. Firstly, it was an observational study conducted at a single institution, on a relatively small number of subjects. Secondly, we only evaluated the muscle mass and not the muscle strength in this study. A subsequent evaluation of the muscle strength too is required. Thirdly, this study was conducted in 2021, and due to the prevailing COVID-19 pandemic, the activities of the hospitalized participants were rather restricted. As a result, the participants exhibited inadequate adherence with the exercise therapy, which limited our ability to reliably investigate the effects of exercise therapy on the body composition. Fourthly, the impact of the weight loss achieved through comprehensive treatment in this study on long-term outcomes is unknown. In a previous study, we showed that people admitted for obesity showed weight loss that persisted for a long time after discharge [30], but whether the changes in the body composition might also persist over the long term needs further investigation.
In summary, we observed that short-term comprehensive treatment preferentially reduced VFA in the subjects while allowing the muscle mass to be maintained (Graphical Abstract). This change in body composition was also accompanied by improvement of markers of inflammation, which are pathophysiological mechanisms underlying the development of obesity. This rapid weight loss might be a useful treatment strategy for obesity. In order to further investigate this treatment strategy, long-term observation with a large sample size will be required.
Y.K., T.A., Y.H., and N.K. conceived the study, defined the objectives, and analyzed and interpreted the data; Y.K., T.A., and N.K. primarily wrote the manuscript; M.S. and R.S. contributed to data collection; M.A. and H.I. advised us on the statistical analysis methods used in this study; and T.K. and T.Y. contributed to critical revision of the manuscript for important intellectual content. All authors contributed substantially to the conception or design of the work or to the acquisition, analysis, or interpretation of data for the work, drafted the work or revised it critically for important intellectual content, approved the final version for publication, and ensured that questions relating to the accuracy or completeness of any part of the work were properly investigated and resolved.
We would like to thank the staff of the Department of Diabetes and Metabolism of the University of Tokyo Hospital and our dietitian for their invaluable support. We would like to thank Tetsuya Sato (Omron Healthcare Co., Ltd, Kyoto, Japan) for his help in determining the body composition of the subjects.
This work was supported by a grant for TSBMI from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to T. Kadowaki). Omron Healthcare Co., Ltd provided us with the DUALSCAN HDS-2000 machine for this study.
The authors have no conflict of interests to declare. This study received ethical approval from the Research Ethics Committee of the University of Tokyo Hospital [Approval No. 3659] on January 30, 2012, and was conducted in accordance with the principles of the Declaration of Helsinki. The first author was approved as a researcher on June 23, 2021.
