Time-restricted eating with or without low-carbohydrate diet reduces visceral fat and improves metabolic syndrome: A randomized trial

Metabolic syndrome (MetS) combines abdominal obesity, elevated blood pressure and fasting blood glucose, and atherogenic dyslipidemia and markedly increases risks of type 2 diabetes and cardiovascular disease. In China, abdominal and general obesity have risen sharply alongside carbohydrate-rich diets and late eating patterns. Lifestyle interventions are first-line therapy, but adherence is challenging. Low-carbohydrate diets (LCD; <130 g/day) can promote fat loss and improve insulin sensitivity, while time-restricted eating (TRE), typically an 8-hour eating window with a 16-hour fast, aligns feeding-fasting cycles with circadian rhythms and may improve metabolic health without calorie counting. Whether an 8-h TRE can match or surpass LCD for reducing body weight and visceral fat in adults with MetS, and whether combining LCD with TRE yields additional benefit, had not been directly tested. This randomized trial compared LCD, 8-h TRE, and their combination on body weight, abdominal fat depots, and cardiometabolic risk in adults with MetS in Shaanxi, China, allowing participants to choose early (8 a.m.–4 p.m.) or late (12 p.m.–8 p.m.) TRE windows.

Prior work shows LCDs often induce faster weight loss, greater fat loss, and maintenance of lean mass compared with low-fat diets, with improved insulin sensitivity but variable effects on glycemia and lipids. Very-low-carbohydrate ketogenic diets can produce substantial weight loss but raise concerns about long-term safety and adherence. TRE regimens (8–10 h windows) have demonstrated modest weight loss (~2–4%) and improvements in cardiometabolic risk factors in individuals with obesity or MetS, possibly via circadian alignment and metabolic switching during fasting. Early TRE (eTRE) may improve insulin sensitivity and blood pressure even without weight loss, while late TRE (lTRE) has shown mixed weight outcomes. Effects of TRE and LCD on lipid profiles vary across studies; some LCDs increase LDL cholesterol, while TRE effects on triglycerides (TG) and HDL cholesterol (HDL-c) are inconsistent. Visceral fat is more strongly linked than subcutaneous fat to insulin resistance, dyslipidemia, blood pressure, and inflammation, highlighting the importance of targeting VFA in MetS.

Design: Randomized, open-label, single-center, three-arm clinical trial comparing LCD, 8-h TRE, and LCD+TRE combination over 3 months, following a 2-week baseline weight-stabilization period. Registered at ClinicalTrials.gov (NCT04475822). Conducted in Xi’an, China (Sept 2020–Jan 2021) with IRB approval; informed consent obtained. Participants: Adults 18–65 years with MetS (≥3 criteria: elevated waist circumference, TG, reduced HDL-c, elevated blood pressure, elevated fasting blood glucose). Stable weight for 3 months prior; medication doses (if any) unchanged during intervention. Exclusions included pregnancy, night-shift work, major diseases, special diets, weight-loss surgery, or drugs affecting appetite. Randomization and groups: 1:1:1 to LCD (A), TRE (B), or LCD+TRE (C). Block randomization via computer-generated list by an independent allocator. Interventions:

• LCD: <130 g/day carbohydrates (<26% energy), with provided food/menu guidance.
• TRE: 8-h daily eating window chosen by participant—either early (8 a.m.–4 p.m.) or late (12 p.m.–8 p.m.); ad libitum within window; 16-h fast allowing water/zero-calorie unsweetened beverages.
• Combination: LCD within the 8-h TRE window (participant-chosen early or late). Participants were asked to maintain habitual physical activity (verified by IPAQ and step counts) and not to count calories. Data collection and compliance: Internet hospital app used for guidance and optional dietary logs; biweekly Food Frequency Questionnaire assessed adherence days, eating window, and estimated macronutrient intake by a dietician. Willingness to continue was asked at 3 months among completers. Outcomes:
• Primary: Changes in body weight and abdominal fat area (visceral fat area, VFA; subcutaneous fat area, SFA) from baseline to 3 months.
• Secondary: Body composition (fat and muscle mass), waist/hip circumferences and WHR, glycemic control (FBG, HbA1c, insulin, C-peptide, HOMA-IR, HOMA-IS, QUICKI), plasma lipids (total cholesterol, LDL-c, HDL-c, TG, TG/HDL-c), uric acid (UA), systolic/diastolic blood pressure (SBP/DBP), adverse events, adherence, and willingness to continue. Measurements: Weight monthly (digital scale); height at screening; VFA/SFA by bioelectrical impedance (HDS-2000) at baseline and 3 months; body composition by DSM-BIA (InBody H20) at baseline and 3 months. Fasting morning blood at baseline and 3 months for HbA1c (TOSOH HLC-723G8), FBG, UA, total cholesterol, TG, HDL-c, LDL-c (HITACHI Labospect 008AS), insulin and C-peptide by magnetic bead immunoassay (Millipore HMHEMAG-34K). Blood pressure measured in triplicate (Omron HBP-9020). Indices: HOMA-IR = fasting insulin×glucose/405; HOMA-IS = 1/HOMA-IR; QUICKI = 1/[log(insulin)+log(glucose)]. Adverse events assessed biweekly. Statistical analysis: Intention-to-treat analyses. Linear mixed model for repeated weight measures; multiple imputation (MCMC) for other missing data. Normality testing; results as mean±SEM or median (IQR) as appropriate. Within-group changes by paired t-test/Wilcoxon; between-group pairwise comparisons at 3 months by t-test/Mann-Whitney U. Correlations (Pearson/Spearman) between abdominal fat areas and metabolic risk factors. Power calculations targeted detecting a 5% weight difference (A vs C) with >80% power, accounting for 20% dropout. Flow: 290 screened; 169 randomized (LCD n=56, TRE n=57, Combo n=56); 162 received intervention; 47 (LCD), 44 (TRE), 44 (Combo) completed. In TRE, 38 chose eTRE and 17 lTRE; in Combo, 32 eTRE and 20 lTRE.

• All three interventions reduced body weight vs baseline; the combination produced the greatest loss:
  • Weight change at 3 months: LCD −2.2 ± 0.3 kg; TRE −3.4 ± 0.4 kg; Combination −5.0 ± 0.4 kg. Pairwise: Combination > LCD (p<0.001) and > TRE (p=0.004); TRE > LCD (p=0.013).
  • Sustained weight reduction observed from month 1 in eTRE and lTRE subgroups.
• Subcutaneous fat area (SFA) decreased similarly in all groups (≈ −23 to −24 cm²), but visceral fat area (VFA) decreased only with TRE and Combination:
  • VFA change: LCD +6 ± 5 cm² (NS); TRE −13 ± 5 cm² (p=0.008); Combination −10 ± 3 cm² (p=0.006). TRE and Combination reduced VFA vs LCD (p=0.009 and p=0.016, respectively).
  • WHR decreased more with TRE (−0.04 ± 0.01) vs LCD (−0.01 ± 0.01, p=0.023) and Combination (−0.01 ± 0.01, p=0.033).
• Glycemic and insulin metrics:
  • FBG decreased with TRE (−0.18 mmol/L, p=0.024) and Combination (−0.21 mmol/L, p=0.048), not with LCD.
  • HbA1c decreased only with Combination (−0.1%, p=0.021).
  • Fasting insulin decreased in all groups; insulin sensitivity improved in all (HOMA-IR↓, HOMA-IS↑, QUICKI↑). Combination showed greater improvements vs LCD: UA (−51 ± 13 vs −17 ± 11 μmol/L, p=0.039), HOMA-IR (−2.16 vs −1.15, p=0.049), HOMA-IS (+0.10 vs +0.03, p=0.042), QUICKI (+0.02 vs +0.01, p=0.004).
• Lipids:
  • TG decreased and TG/HDL-c ratio improved with TRE and Combination; not with LCD. Differences: Combination vs LCD for TG (−0.51 vs −0.15 mmol/L, p=0.011) and TG/HDL-c (−0.59 vs −0.02, p=0.003).
  • HDL-c increased only in Combination (+0.09 mmol/L, p<0.001).
  • LDL-c increased with LCD (+0.28 mmol/L, p=0.042) and Combination (+0.30 mmol/L, p=0.026); TRE no significant change.
• Blood pressure: SBP unchanged; DBP decreased significantly only in Combination (−5 mmHg, p=0.005); between-group DBP differences not significant.
• Correlations: VFA (not SFA) correlated with HOMA-IS, UA, TG/HDL-c, SBP, and DBP.
• Adherence and feasibility: TRE had more adherence days (65.9 ± 3.0) than LCD (55.5 ± 3.5; p=0.024). Willingness to continue was high in LCD and TRE (98%) but lower in Combination (82%, p=0.010 vs LCD; p=0.014 vs TRE). Adverse events were mild and infrequent, with no serious events; rates similar across groups.
• eTRE vs lTRE (exploratory): eTRE reduced VFA and SFA; lTRE did not reach significance, but differences between eTRE and lTRE were not significant; metabolic improvements were generally comparable between windows.

This randomized trial directly compared LCD, 8-h TRE, and their combination in adults with metabolic syndrome. All interventions achieved weight loss and reductions in subcutaneous fat, but only TRE—with or without LCD—reduced visceral fat and improved multiple cardiometabolic markers, including fasting glucose, uric acid, triglycerides, and TG/HDL-c ratio. The combination regimen yielded the largest weight loss and improved diastolic blood pressure, along with superior improvements in insulin sensitivity indices compared with LCD alone, though LDL-c rose with LCD and the combination. Visceral fat reductions aligned with improvements in glycemic indices, lipids (TG/HDL-c), and blood pressure, underscoring the centrality of VFA as a cardiometabolic risk driver in MetS. TRE’s benefits without mandated caloric restriction and with higher adherence suggest it may be a more practical first-line dietary strategy for reducing visceral adiposity and cardiometabolic risk than LCD, while combining TRE with LCD can amplify weight loss and some metabolic benefits at the cost of lower willingness to continue and increased LDL-c. These findings support integrating TRE, with or without moderated carbohydrate intake, to address abdominal obesity and metabolic dysfunction in MetS without altering physical activity.

Over 3 months, LCD, 8-h TRE, and their combination all reduced body weight and subcutaneous fat in adults with metabolic syndrome. TRE, with or without LCD, additionally reduced visceral fat and improved fasting glucose, uric acid, triglycerides, and TG/HDL-c ratio, with higher adherence than LCD. The combined LCD+TRE intervention produced the greatest weight loss and improvements in insulin sensitivity and diastolic blood pressure but increased LDL-c and had lower willingness to continue. Collectively, an 8-h TRE regimen—without or with LCD—emerges as an effective strategy to reduce visceral obesity and cardiometabolic risk in MetS. Future research should include longer-term, multicenter trials across diverse populations, randomized comparisons of early versus late TRE, inclusion of control groups, objective dietary intake monitoring, and mechanistic studies to disentangle weight-loss–dependent versus independent effects on visceral fat and metabolic health.

• Self-reported dietary intake and adherence introduce recall bias; dietary logs were optional and may be incomplete. Meal-eating window estimates relied on 2-week recall of meal times and did not capture snacks or day-to-day variability, likely underestimating true eating windows.
• Except for the combination group, LCD and TRE alone did not achieve clinically significant ≥5% weight loss over 3 months; longer trials are needed to assess sustained weight loss and health benefits.
• Early vs late TRE comparisons were exploratory (non-randomized subgroup choice, small sample sizes).
• No non-intervention control group was included.
• Higher adherence days in TRE vs LCD may confound comparisons. Generalizability is limited to Chinese adults in Shaanxi; validation in other races/ethnicities is needed.