Previous studies have demonstrated that intermittent fasting (IF) has garnered scientific attention and gained recognition for its beneficial effects on metabolic outcomes. However, the results are inconsistent. Accordingly, this systematic review and meta-analysis aimed to evaluate the effect of fasting on glycemic control, lipid profile, and inflammatory markers.
Methods
Databases such as PubMed, Embase, Cochrane, Scopus, and Web of Science were used to retrieve relevant studies published until September 2025. The quality of the included studies was evaluated using the Cochrane Risk-of-Bias 2 (RoB2) tool. Moreover, the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) approach was employed to evaluate the quality of evidence.
Results
A total of 10 studies, involving 701 individuals, were included in the current meta-analysis. The combined effect of various types of fasting significantly reduced fasting blood sugar (FBS) [standard mean difference (SMD) = −0.51; 95% confidence interval (CI): −0.81, −0.20; p = 0.001], insulin (SMD = −0.27; 95% CI: −0.52, −0.03; p = 0.027) and Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) (SMD = −0.39; 95% CI: −0.65, −0.12; p = 0.004), and HbA1c (SMD = −0.25; 95% CI: −0.49, −0.02; p = 0.034) levels. Moreover, the regimen successfully exerted its beneficial effect on low-density lipoprotein cholesterol (LDL-C) (SMD = −0.34; 95% CI: −0.53, −0.14; p = 0.001) and interleukin-6 (IL-6) (SMD = −0.30; 95% CI: −0.57, −0.03; p = 0.029) levels as well. The sensitivity analysis indicated that excluding any single study had no effect on the overall effect size (ES) for FBS, blood sugar (BS), HOMA-IR, LDL-C, and high-density lipoprotein cholesterol (HDL-C). Moreover, the results of the GRADE approach scored high quality of evidence for FBS, insulin, HOMA-IR, HbA1c, total cholesterol (TC), LDL-C, and IL-6, which suggests the robustness of the results. No evidence of publication bias was detected using Egger’s and Begg’s test (p > 0.05).
Conclusion
The findings suggest that intermittent fasting may have favorable effects on the metabolic panel, specifically, FBS, insulin, HOMA-IR, (HbA1c), LDL-C, and IL-6 levels.
intermittent fastingmetabolic syndromelipid profileinflammatorymeta-analysisThe author(s) declare that no financial support was received for the research and/or publication of this article.section-at-acceptanceNutrition and MetabolismIntroduction
Metabolic health is influenced by glycemic levels and lipid profiles, both of which are associated with several diseases such as obesity, metabolic syndrome, impaired glucose tolerance, insulin resistance, and dyslipidemia (1–3). Moreover, low-grade inflammation triggers metabolic abnormalities, which subsequently affect metabolic health (4, 5). The growing global prevalence of these health conditions has raised public health concerns that need rapid intervention. In addition to pharmacological strategies, there is a need for the development of non-pharmacological therapies as an adjunctive and complementary therapy to enhance the success of metabolic interventions.
In this regard, fasting approaches are known to exert beneficial effects on metabolic health by modulating metabolic responses. Several types of fasting regimens have been developed with distinct metabolic effects. Intermittent fasting (IF) is a general and broad term representing dietary patterns alternating between periods of eating and fasting (6). Time-restricted fasting (TRF) limits food intake to a specific time frame within a day, whereas alternate-day fasting (ADF) alternates between days of normal eating and days of fasting (6–8). Evidence shows that a fasting regimen contributes to decreased fasting and postprandial glucose levels as well as improved insulin sensitivity, thereby serving as a key determinant of glycemic control (9). Furthermore, fasting regimens have been shown to increase lipolysis and decrease lipid synthesis, thereby improving dyslipidemia (10). In addition, several anti-inflammatory effects have been linked to fasting, although the extent of these effects may vary depending on the type of IF.
Although numerous studies have been conducted to evaluate the efficacy of various fasting regimens on metabolic health and generally reported beneficial outcomes, some studies have found no significant effects. Nonetheless, there is no consensus about the metabolic benefits of fasting. Therefore, this study aimed to provide a firm conclusion regarding the impact of fasting regimen on metabolic health.
Method
The study was conducted following the Cochrane Handbook for Systematic Reviews of Interventions (11) and the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) (12) guidelines. In addition, the study protocol was approved by the International Prospective Register of Systematic Reviews (CRD420251142741).
Search strategy
The PubMed, Scopus, Embase, Web of Science, and Cochrane Library were used to search for the relevant studies. The search was conducted from its inception through September 2025. Additionally, the reference list of relevant studies has been screened through a manual search. The search strategy was constructed based on an appropriate combination of Medical Subject Headings (MeSH) terms and keywords (Supplementary Table S1). The language was restricted to English.
Inclusion and exclusion criteria
Inclusion criteria for this study were defined using Population, Intervention, Comparison, and Outcome (PICO) criteria; Population (P): adults aged >18 years old with metabolic syndrome, Intervention (I): all types of fasting patterns, Comparison (C): placebo or control group, Outcome (O): blood glucose, glycemic control [fasting plasma glucose (FPG), HbA1c, Homeostatic Model Assessment for Insulin Resistance (HOMA-IR)], insulin, lipid profile [total cholesterol (TC), low-density lipoprotein cholesterol (LDL-c), high-density lipoprotein cholesterol (HDL-c), and triglyceride (TG)] levels, inflammatory markers [C-reactive protein (CRP), IL-6, and tumor necrosis factor (TNF-α)]. Studies with no placebo control group, studies with other designs (observational, animal studies), studies with duplicate data, and studies that evaluated the efficacy of Islamic fasting, which is structurally different from other types of fasting and complicates the comparison, were excluded.
Data screening procedures and data extraction
All retrieved studies, duplicates, and references were managed using EndNote (version 9) to screen and remove any duplicates. Two independent researchers completed the screening process based on the title and abstract. Subsequently, the remaining studies were checked for their eligibility using full-texts. Any disagreements were resolved by consulting a third researcher. Similarly, the following data were extracted from the included studies: the name of first author, publication year, country, study design, gender, mean age, body mass index (BMI), and sample size of both intervention and control groups, study duration, health condition of study participants, type of fasting and control group, changes in main outcomes for both intervention and control groups [mean ± standard deviation (SD)].
Quality assessment and quality of evidence
The quality of included randomized controlled trials (RCTs) was evaluated using the RoB2 tool (13), which contains five domains: concealment of allocation, generation of random sequences, selective reporting, blinding of outcome assessment, and incomplete outcome data. Each of these domains was evaluated and scored as low, unclear, or high risk. In addition, the quality of evidence for all study outcomes was assessed using the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) guidelines (14). Accordingly, the quality of evidence was classified as high, moderate, and low for each outcome.
Statistical analysis
STATA Statistical Software version 14 (Stata Corp, College Station, TX, United States) was used to perform analyses using a random-effect model (15). The standard mean difference (SMD) and 95% confidence intervals (CIs) of changes for each outcome were used to express the overall effect size. The I2 index was used to present the heterogeneity of studies. In addition, a sensitivity analysis was conducted to evaluate the effect of individual studies on the overall effect size. Funnel plots and Begg’s and Egger’s tests were used to show any evidence of publication bias.
ResultsStudy selection and characteristics
The search retrieved 1,268 records, of which 627 records were duplicates and were removed. The remaining 641 RCTs were checked by title and abstract, resulting in the exclusion of 628 irrelevant RCTs. Subsequently, 13 studies were screened using full-texts, and 3 articles that had no control group (n = 2) and a review (n = 1) were excluded. Finally, 10 studies were included for the meta-analysis. The PRISMA flow diagram is presented in Figure 1.
PRISMA flow chart of selection studies.
Flowchart depicting the selection process for a meta-analysis. It starts with 1268 records identified through database searching. After removing duplicates, 641 records remain. 628 articles are excluded based on title and abstract. Thirteen full-text articles are evaluated for eligibility. Three articles are excluded because two lack a control group and one has another design. Ten studies are included in the meta-analysis. The process follows the identification, screening, eligibility, and inclusion steps.
The study characteristics of the 10 studies included are presented in Table 1. The total sample size of the included RCTs comprised 701 individuals, with an age range of 25–75 years. The included trials were published between 2017 and 2025. Both men and women were included in all studies. BMI values indicated that all study subjects were classified as overweight or obese. The intervention duration ranged from 1 to 16 weeks. The majority of the participants were diagnosed with metabolic syndrome (MetS) or exhibited MetS components. Moreover, the type of intervention varied between studies, including TRF, ADF, and ICR.
Basic characteristics of included RCTs.
Author, Year
Country
Study, Design
Gender
Mean age (Int, Cont)
BMI (Int, Cont)
N (Int, Cont)
Duration (week)
Health condition
Intervention type
Control group
Main outcome (Int)
Main outcome (Cont)
Sun et al. (2025) (22)
China
Parallel-arm, RCT
M/F
49.346.5
2931.1
60 (30, 30)
12
MASLD
(Intermittent calorie restriction) Two successive days of fasting and 5 days of recovery per week. On fasting days: Consuming only a fixed amount of plant-based meal replacement, which provided 497.6 kcal/day
(Time-restricted eating) Restriction of daily food intake to a 9 h window (between 8:00 a.m. and 5:00 p.m.), without any limitation on the types of food and beverages consumed
Usual care for impaired glucose
FBS (−3.03 ± 1.94)HbA1c (0.09 ± 0.08)
FBS (−1.79 ± 1.75)HbA1c (0.15 ± 0.08)
Cramer et al. (2022) (18)
Germany
SB, multicenter, parallel, RCT
M/F
58.6 ± 10.860.8 ± 10.8
33.7 ± 4.5
145 (73, 72)
10
MetS
(Intermittent calorie restriction) 2 calorie-restricted vegan days (max 1,200 kcal/day), followed by a 5-day modified fasting intervention (intake of 300–350 kcal/day, obtained from vegetable juices and vegetable broth)
(Time-restricted eating) The 8-h TRE group was instructed to consume all calories from 8 a.m. to 4 p.m. each day and fast from 4 p.m. to 8 a.m., or to consume all calories from 12 p.m. to 8 p.m. each day and fast from 8 p.m. to 12 p.m. (16-h fast). During the 8-h meal eating windows, they could eat ad libitum without any restriction on the quantities and types of food
(Modified Alternate-Day Fasting) A very low-calorie diet (75% energy restriction) during the 3 fast days (Saturday, Monday, Wednesday) and then ate a diet that provided 100% of their energy needs on each feed day (Sunday, Tuesday, Thursday)
The impact of the fasting approach on glycemic control
The pooled analysis of five studies encompassing 393 individuals evaluating the effect of fasting regimen on fasting blood sugar (FBS) indicated a significant reduction (SMD = −0.51; 95% CI: −0.81, −0.20, p = 0.001; I2 = 53.3%, p = 0.073) (Figure 2A). The subgroup analysis revealed that the fasting approach has a significant reducing effect on older adults (>50 years old) with a higher BMI (>30 kg/m2) and in short-term treatment (p < 0.05). In addition, it has demonstrated that most of this beneficial effect is related to TRF treatment rather than other fasting methods (p < 0.05) (Table 2).
(A) Forest plot of intermittent fasting vs. control on FBS level. (B) Forest plot of intermittent fasting vs. control on the BS level. (C) Forest plot of intermittent fasting vs. control on insulin level. (D) Forest plot of intermittent fasting vs. control on HOMA-IR level. (E) Forest plot of intermittent fasting vs. control on HbA1c level.
Forest plot comparing two sets of studies, labeled A and B. Study A includes five studies with effect sizes ranging from -1.08 to -0.11. The overall effect is -0.51 with a confidence interval of -0.81 to -0.20. Study B includes five studies with effect sizes from -0.75 to 1.21. The overall effect is -0.06 with a confidence interval of -0.66 to 0.55. Each study shows a horizontal line representing confidence intervals and a diamond for overall effect size. Both plots note weights are from random effects analysis. Two forest plots display meta-analysis results for various studies. Plot C shows studies with standardized mean differences (SMD) ranging from -0.84 to 0.11, with an overall SMD of -0.27. Plot D presents SMDs from -0.97 to 0.04, with an overall SMD of -0.39. Both plots include confidence intervals, study weights, and durations in weeks. The plots feature diamonds indicating overall effect sizes and lines for confidence intervals. Forest plot depicting results from multiple studies, measuring standardized mean differences (SMD) with 95% confidence intervals for various interventions. Studies listed include Sun et al, Manoogian et al, Suthutvoravut et al, Cramer et al, He et al, Kunduraci et al, and Li et al. The overall effect size is marked as a diamond shape, suggesting a pooled estimate value of -0.25. The plot indicates heterogeneity with an I-squared value of 46.5% and a p-value of 0.082. Weights for each study, based on random effects analysis, and study durations in weeks, are provided.
Pooled estimate effects of fasting regimen on the metabolic markers across different subgroups.
Group
Number of comparisons
ES (95% CI)
p-value
I2 (%)
p-heterogeneity
FBS
Age
<50 years
3
−0.49 (−1.03, 0.05)
0.077
75.7%
0.016
>50 years
2
−0.54 (−0.86, −0.21)
0.001
0.0%
0.661
BMI
<30
2
−0.23 (−0.53, 0.07)
0.133
0%
0.567
>30
2
−0.76 (−1.33, −0.18)
0.010
69.6%
0.070
NR
1
−0.65 (−1.24, −0.05)
0.033
–
–
Duration
<8
2
−0.89 (−1.31, −0.47)
<0.001
14.0%
0.281
>8
3
−0.33 (−0.57, −0.09)
0.006
0.0%
0.497
Type of fasting
TRF
3
−0.43 (−0.68, −0.19)
<0.001
0.0%
0.582
ADF
1
−1.06 (−1.58, −0.57)
<0.001
–
–
ICR
1
−0.11 (−0.62, 0.39)
0.663
–
–
Sample size
<100
3
−0.61 (−1.19, −0.03)
0.038
71.4%
0.030
>100
2
−0.39 (−0.66, −0.12)
0.004
0.0%
0.484
BS
Age
<50 years
3
−0.17 (−0.63, 0.30)
0.484
55.1%
0.108
>50 years
2
0.20 (−1.72, 2.13)
0.838
95.4%
<0.001
BMI
<30
2
0.0 (−0.57, 0.58)
0.988
50.6%
0.155
>30
2
−0.66 (−0.94, −0.38)
<0.001
0.0%
0.350
NR
1
1.21 (0.46, 1.97)
0.002
–
–
Duration
<8
2
0.43 (−1.07, 1.94)
0.574
89.2%
0.002
>8
3
−0.34 (−0.93, 0.25)
0.263
81.3%
0.005
Insulin
Age
<50 years
5
−0.10 (−0.31, 0.12)
0.371
0.0%
0.786
>50 years
3
−0.52 (−0.96, −0.09)
0.019
65.5%
0.055
BMI
<30
3
−0.12 (−0.39, 0.15)
0.381
0.0%
0.677
>30
4
−032 (−0.73, 0.10)
0.137
75.2%
0.007
NR
1
−0.46 (−1.16, 0.25)
0.202
–
–
Duration
<8
3
−0.31 (−0.64, 0.03)
0.072
0.0%
0.845
>8
5
−0.25 (−0.60, 0.11)
0.175
73.5%
0.005
Type of fasting
TRF
2
−0.12 (−0.39, 0.14)
0.368
0.0%
0.461
ADF
1
−0.21 (−0.69, 0.26)
0.374
–
–
ICR
5
−0.36 (−0.75, 0.03)
0.073
65.4%
0.021
Sample size
<100
5
−0.17 (−0.41, 0.07)
0.168
0.0%
0.693
>100
3
−0.37 (−0.87, 0.13)
0.147
82.3%
0.004
HOMA-IR
Age
<50
5
−0.37 (−0.73, −0.02)
0.039
61.4%
0.035
>50
3
−0.41 (−0.86, 0.04)
0.076
68.1%
0.043
BMI
<30
3
−0.29 (−0.83, 0.24)
0.284
70.4%
0.034
>30
4
−0.53 (−0.81, −0.24)
<0.001
46.5%
0.132
NR
1
0.0(−0.69, 0.69)
1.00
–
–
Duration
<8
3
−0.42 (−0.93, 0.09)
0.107
52.7%
0.121
>8
5
−0.38 (−0.72, −0.03)
0.031
70.8%
0.008
Type of fasting
TRF
2
−0.11 (−0.41, 0.19)
0.463
20.0%
0.264
ADF
1
−0.31 (−0.78, 0.17)
0.205
–
–
ICR
5
−0.55 (−0.90, −0.20)
0.002
55.3%
0.063
Sample size
<100
5
−0.42 (−0.75, −0.09)
0.013
43.4%
0.132
>100
3
−0.34 (−0.83, 0.14)
0.166
81.4%
0.005
HbA1c
Age
<50
3
0.01 (−0.24, 0.27)
0.909
0.0%
0.786
>50
4
−0.49 (−0.71, −0.27)
<0.001
0.0%
0.502
BMI
<30
2
0.04 (−0.26, 0.34)
0.779
0.0%
0.704
>30
3
−0.41 (−0.67, −0.15)
0.002
26.6%
0.256
NR
2
−0.36 (−1.03, 0.30)
0.283
52.7%
0.146
Duration
<8
2
−0.36 (−1.03, 0.30)
0.283
52.7%
0.146
>8
5
−0.22 (−0.50, 0.05)
0.104
54.6%
0.066
Type of fasting
TRF
3
−0.36 (−0.77, 0.05)
0.082
61.7%
0.074
ICR
4
−0.17 (−0.50, 0.17)
0.324
47.5%
0.126
Sample size
<100
4
−0.14 (−0.48, 0.20)
0.425
32.1%
0.220
>100
3
−0.35 (−0.68, −0.02)
0.040
60.9%
0.078
However, the combined effect of studies which have evaluated the effects of fasting regimen on BS illustrated a non-significant effect (SMD = −0.06; 95% CI: −0.66, 0.55, p = 0.849; I2 = 85.4%, p < 0.001) (Figure 2B). Similarly, subgroup analysis showed that individuals with higher BMI had reduced BS levels following a fasting regimen (p < 0.05) (Table 2).
In total, eight studies including 628 individuals assessed the effect of fasting on insulin levels. It has been demonstrated that a fasting regimen could have a significant and beneficial reducing effect on insulin levels (SMD = −0.27; 95% CI: −0.52, −0.03; p = 0.027) with moderate heterogeneity (I2 = 54.6.0%, p = 0.031) (Figure 2C). The subgroup analysis revealed that older adults (>50 years old) may benefit more from the fasting approach in terms of insulin levels (p < 0.05) (Table 2).
Overall, eight trials with a total of 628 adults indicated that a fasting regimen could reduce HOMA-IR levels significantly (SMD = −0.39; 95% CI: −0.65, −0.12, p = 0.004; I2 = 61.0%, p = 0.012) (Figure 2D). The subgroup analysis showed that the fasting regimen has a more favorable effect on younger adults (<50 years old) and obese individuals (BMI > 30 kg/m2) (p < 0.05). In addition, long-term fasting (>8 weeks) was associated with a significant reduction in HOMA-IR levels as well (p < 0.05). Moreover, the subgroup analysis based on the type of fasting method revealed a significant reduction effect for the ICR methodology (p < 0.05) (Table 2).
In addition, the effect of fasting on HbA1c levels was explored from 7 studies with 566 participants. Random-effects model indicated that fasting significantly reduced HbA1c levels (SMD = −0.25; 95% CI: −0.49; −0.02, p = 0.034; I2 = 46.5%, p = 0.082) (Figure 2E). The results of the subgroup analysis revealed that older adults (>50 years old) with higher BMI (>30 kg/m2) demonstrated significantly greater improvements than other subgroups (Table 2).
The impact of the fasting approach on lipid profile
Seven studies, including 520 individuals, investigated the effect of fasting regimen on TC levels. Accordingly, it has been shown that fasting did not significantly affect the TC level (SMD = 0.13; 95% CI: −0.07, 0.33, p = 0.212; I2 = 20.9%, p = 0.270) (Figure 3A). Fasting intervention showed similar effects in terms of TG (SMD = −0.17; 95% CI: −0.38, 0.03, p = 0.097; I2 = 36.5%, p = 0.138) (Figure 3B) and HDL-C (SMD = 0.07; 95% CI: −0.28, 0.42, p = 0.690; I2 = 78.2%, p < 0.001) (Figure 3C) levels. However, the pooled effect of seven studies (518 participants) showed a reduced effect on LDL-C levels following fasting (SMD = −0.34; 95% CI: −0.53, −0.14, p = 0.001; I2 = 19.3%, p = 0.282) (Figure 3D). In addition, heterogeneity for TC, TG, and LDL-C levels was below 50%, and therefore, no subgroup analysis was performed. However, the subgroup analysis carried out for HDL-C failed to identify the source of heterogeneity in the HDL-C values (Table 2).
(A) Forest plot of intermittent fasting vs. control on TC level. (B) Forest plot of intermittent fasting vs. control on TG level. (C) Forest plot of intermittent fasting vs. control on HDL-C level. (D) Forest plot of intermittent fasting vs. control on LDL-C level.
Two forest plots titled A and B compare study results using standardized mean differences (SMD) with 95% confidence intervals. Both plots list authors, SMD values, weights, and durations in weeks. Plot A shows an overall SMD of 0.13 with I-squared at 20.9 percent, and plot B an overall SMD of 0.17 with I-squared at 36.5 percent, both using random effects analysis. Forest plots labeled C and D depict the standardized mean differences (SMD) with 95% confidence intervals for various studies, including Sun et al and Cramer et al. Each line represents a study's effect size and confidence interval. The diamond represents the overall effect size. Plot C shows high heterogeneity (I-squared = 78.2 percent), while plot D indicates lower heterogeneity (I-squared = 19.3 percent). Each study’s weight and duration in weeks are provided. These plots illustrate data from a random effects analysis.The impact of the fasting approach on inflammatory markers
The present study also attempted to evaluate the effect of fasting on inflammatory markers (CRP, IL-6, and TNF-α). The meta-analysis showed that fasting regimen has significant ameliorative effects on the IL-6 levels (SMD = −0.30; 95% CI: −0.57, −0.03, p = 0.029; I2 = 0.0%, p = 0.0512) (Figure 4). However, it had no significant effect on CRP and TNF-α levels (p > 0.05).
Forest plot of intermittent fasting vs. control on inflammatory markers.
Forest plot displaying effect sizes for studies on CRP, IL-6, and TNF-a. Each study is associated with a diamond and horizontal line, representing the effect size and confidence interval. Subtotals and heterogeneity statistics are provided.Sensitivity analyses and publication bias
Sensitivity analyses employed a leave-one-out approach to exclude an individual study and evaluate the overall effect on the results. It was shown that excluding each individual study has no effect on the overall results of FBS, BS, and HOMA-IR. In contrast, the studies by Manoogian et al. (16), Suthutvoravut et al. (17), and Cramer et al. (18) have the potential to alter the levels of HbA1c to a non-significant form. Additionally, sensitivity analyses showed that the exclusion of Manoogian et al. (16), Cramer et al. (18), Guo et al. (19), and Li et al. (20) altered the pooled results for insulin levels. In addition, sensitivity analyses of lipid markers demonstrated that excluding the study by He et al. (21) could affect the overall results of TC. Similarly, excluding Sun et al.(22), Manoogian et al. (16) were able to significantly alter the effect of fasting on TG levels. Nonetheless, the leave-one-out approach had no significant effect on LDL-C and HDL-C outcomes.
Publication bias for included studies was assessed using Egger’s and Begg’s tests. No evidence of publication bias was observed for any glycemic-related outcomes (Egger’s and Begg’s test results: FBS, 0.551 and 0.806; BS, 0.087 and 0.086; insulin, 0.539 and 0.711; HOMA-IR, 0.994 and 0.902; and HbA1c, 0.510 and 0.764). Similarly, no publication bias was detected for lipid markers (TC, 0.667 and 0.230; LDL-C, 0.942 and 0.230; HDL-C, 0.718 and 0.266; and TG, 0.954 and 1.0). Visual inspection of the funnel plots also indicated no publication bias across all study outcomes as presented in Supplementary Figure S1.
Quality assessment and GRADE approach
Quality assessment of the included studies using the RoB2 tool classified three studies as having a low risk of bias, while four studies had some concerns. All of the included studies had a low risk of bias for the following domains: randomization process, selection of the reported result, and missing outcome data. The details of the quality assessment based on the domains are presented in Table 3.
Results of risk of bias assessment for included RCTs in the present study.
Author, Year
Randomization process
Deviation from intended interventions
Selection of the reported result
Measurement of the outcome
Missing outcome data
General risk of bias
Sun et al. (2025) (22)
Low
High
Low
Low
Low
High
Manoogian et al. (2024) (16)
Low
Unclear
Low
Unclear
Low
Some concern
Suthutvoravut et al. (2023) (17)
Low
Low
Low
Low
Low
Low
Cramer et al. (2022) (18)
Low
Unclear
Low
Low
Low
Some concern
Sun et al. (2025) (22)
Low
Unclear
Low
Low
Low
Some concern
Guo et al. (2021) (19)
Low
High
Low
High
Low
High
Razavi et al. (2021) (35)
Low
Low
Low
Low
Low
Low
Kunduraci and Ozbek (2020) (33)
Low
Unclear
Low
Unclear
Low
Some concern
Parvaresh et al. (2019) (34)
Low
Low
Low
Low
Low
Low
Li et al. (2017) (20)
Low
High
Low
Low
Low
High
Each study was assessed for risk of bias using the ROB2 tool. Each domain was rated as “high risk” if it contained methodological issues that may have affected the results, “low risk” if the flaw was deemed inconsequential, and “some concern” if information was insufficient to determine.
The GRADE quality of evidence for the effect of fasting on FBS, insulin, HOMA-IR, HbA1c, TC, LDL-C, and IL-6 levels was assessed as high, suggesting the robustness of the results. However, the effect of fasting was considered moderate for TG and TNF-α values (Table 4).
Summary of findings and quality of evidence assessment using the GRADE approach.
Outcomes
No of patients (meta-analysis)
SMD (95% CI)
Risk of bias
Inconsistency
Indirectness
Imprecision
Publication bias
Quality of evidence
BS (mg/dL)
341 (5)
−0.06 (−0.66, 0.55)
Not serious
Serious
Not serious
Serious
Not serious
Low
FBS (mg/dL)
393 (5)
−0.51 (−0.81, −0.20)
Not serious
Not serious
Not serious
Not serious
Not serious
High
Insulin (mU/L)
628 (8)
−0.27 (−0.52, −0.03)
Not serious
Not serious
Not serious
Not serious
Not serious
High
HOMA-IR
628 (8)
−0.39 (−0.65, −0.12)
Not serious
Not serious
Not serious
Not serious
Not serious
High
HbA1c
566 (7)
−0.25 (−0.49, −0.02)
Not serious
Not serious
Not serious
Not serious
Not serious
High
TC (mg/dL)
520 (7)
0.13 (−0.07, 0.33)
Not serious
Not serious
Not serious
Serious
Not serious
High
LDL-C (mg/dL)
261 (7)
−0.34 (−0.53, −0.14)
Not serious
Not serious
Not serious
Not serious
Not serious
High
HDL-C (mg/dL)
628 (8)
0.07 (−0.28, 0.42)
Not serious
Serious
Not serious
Serious
Not serious
Low
TG (mg/dL)
628 (8)
−0.17 (−0.38, 0.03)
Not serious
Not serious
Not serious
Serious
Not serious
Moderate
CRP
322 (3)
−0.35 (−0.94, 0.24)
Not serious
Serious
Not serious
Serious
Not serious
Low
IL-6
214 (2)
−0.30 (−0.57, −0.03)
Not serious
Not serious
Not serious
Not serious
Not serious
High
TNF-α
69 (1)
−0.23 (−0.71, 0.24)
Not serious
Not serious
Not serious
Serious
Not serious
Moderate
Discussion
This systematic review and meta-analysis of 10 RCTs provides a valuable insight regarding the impact of IF on metabolic health. It has been demonstrated that IF is associated with an improvement in glycemic control and insulin sensitivity, as evidenced by a reduction in FBS, insulin, HOMA-IR, and HbA1c levels. These favorable effects seem to be mediated through several probable mechanisms. Previous studies indicate that IF inhibits gluconeogenesis in hepatocytes, thereby reducing HOMA-IR levels and improving overall insulin sensitivity (23). In addition, IF has been reported to promote pancreatic islet neogenesis, which may enhance the β-cell function and improve glycemic regulation (23, 24). It is worth noting that postabsorptive glucose levels are dependent on daily intake and dietary fluctuations and appear to be less influenced by IF intervention. Similarly, the higher sensitivity and responsiveness of postprandial glucose to short-term interventions cannot be ignored. In contrast, FBS, after a night of fasting, and HbA1c levels are more stable and provide a precise judgment in this regard (25). Another probable explanation that can unveil these findings is attributed to the way that fasting exerts its effect. The fasting regimen predominantly influences improvements in hepatic glucose production and insulin sensitivity, rather than postprandial glucose levels. In addition, subgroup analysis indicated that fasting contributes to a more significant improvement in FBS, BS, HOMA-IR, and HbA1c levels in obese individuals (BMI > 30 kg/m2). These findings are promising for individuals with obesity who are at risk of MetS and diabetes progression. It could be a beneficial strategy for the management of metabolic health in high-risk populations.
Additionally, subgroup analyses demonstrated that fasting intervention resulted in a greater reduction in FBS, HOMA-IR, and HbA1c levels among participants aged >50 years. Some probable mechanisms may explain this age-specific finding. Older adults tended to show higher baseline FBS, insulin, and HbA1c values, and the aging process is accompanied by insulin resistance. Likewise, they may respond more efficiently to fasting intervention.
In addition, fasting was accompanied by a significant reduction in LDL-C levels as well. This literature review has illustrated that fasting suppresses sterol regulatory element-binding protein 2 (SREBP-2), thereby inhibiting the activation of 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) synthase and ultimately reducing the synthesis of TC (26–28). Moreover, IF induces proliferator-activated receptor alpha (PPARα), which participates in the activation of the JMJD3-SIRT1-PPAR α complex (26, 29). Through histone modification, this pathway promotes β-oxidation of fatty acids (26). Although improvements in TC, TG, and HDL-C levels might be expected with IF, these lipid markers are influenced by multiple factors such as genetic variability, dietary composition, and lifestyle habits, which may mask the overall pooled effect in meta-analyses (30).
Furthermore, IF was associated with an improvement in IL-6 levels, suggesting an anti-inflammatory effect of IF (31, 32). However, there are only a few number of studies that evaluate other inflammatory markers, which limits the statistical power to assess their actual effect.
From a clinical perspective, the effect sizes (ESs) of the glycemic markers were all below the minimal clinically important difference (MCID), indicating that IF may have only a limited clinical effect, despite achieving statistical significance. The pooled ESs for HbA1c (SMD: −0.25) are below the MCID (MCID: 0.3–0.5), suggesting that although the reduction in HbA1c levels was statistically significant, it is unlikely to be clinically meaningful too. Similarly, the effect sizes of other evaluated markers were below the MCID. These findings highlight the need for further studies to determine whether IF can serve as an effective adjunctive strategy to improve glycemic control in high-risk populations. In addition, the majority of the included RCTs were of short duration (≤16 weeks), which limits the ability to assess the long-term efficacy.
Overall, the quality assessment of the included studies revealed a low publication bias for the majority of the included studies. In addition, the certainty of the evidence, as assessed using the GRADE method, showed a higher level of certainty for the HOMA-IR outcome, suggesting that these findings are more reliable and generalizable. Additionally, HbA1c, insulin, TC, TG, LD-C, and HDL-C levels received moderate certainty of evidence, reflecting a reasonable but less robust level of confidence. The current study also had some limitations that need to be addressed. The type of fasting regimen may affect the observed outcomes.
Conclusion
In conclusion, the current systematic review and meta-analysis demonstrate that fasting regimens may help improve glycemic control by significantly reducing the FBS, HbA1c, and HOMA-IR levels. Similarly, IF was also successful in reducing the LDL-C levels.
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The authors declare that no Gen AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2025.1664811/full#supplementary-material
ReferencesReinMBen-YacovOGodnevaAShiloSZmoraNKolobkovD. Effects of personalized diets by prediction of glycemic responses on glycemic control and metabolic health in newly diagnosed T2DM: a randomized dietary intervention pilot trial. BMC Med. (2022) 20:56. doi: 10.1186/s12916-022-02254-y, 35135549KazeADSanthanamPMusaniSKAhimaREchouffo-TcheuguiJB. Metabolic dyslipidemia and cardiovascular outcomes in type 2 diabetes mellitus: findings from the look AHEAD study. J Am Heart Assoc. (2021) 10:e016947. doi: 10.1161/JAHA.120.016947, 33728932MusazadehVRoshanravanNMohammadizadehMKavyaniZDehghanPMosharkeshE. Curcumin as a novel approach in improving lipid profile: an umbrella meta-analysis. Nutr Metab Cardiovasc Dis. (2022) 32:2493–504. doi: 10.1016/j.numecd.2022.07.021, 36058763LasselinJCapuronL. Chronic low-grade inflammation in metabolic disorders: relevance for behavioral symptoms. Neuroimmunomodulation. (2014) 21:95–101. doi: 10.1159/000356535, 24557041Dabbaghi VarnousfaderaniSMusazadehVGhalichiFKavyaniZRazmjoueiSFaghfouriAH. Alleviating effects of coenzyme Q10 supplements on biomarkers of inflammation and oxidative stress: results from an umbrella meta-analysis. Front Pharmacol. (2023) 14:1191290. doi: 10.3389/fphar.2023.1191290, 37614320MattsonMPLongoVDHarvieM. Impact of intermittent fasting on health and disease processes. Ageing Res Rev. (2017) 39:46–58. doi: 10.1016/j.arr.2016.10.005, 27810402BahadoranZMirmiranPKashfiKGhasemiA. Effects of time-restricted feeding (TRF)-model of intermittent fasting on adipose organ: a narrative review. Eat Weight Disord. (2024) 29:1–22. doi: 10.1007/s40519-024-01709-w, 39719521ParkJSeoYGPaekYJSongHJParkKHNohHM. Effect of alternate-day fasting on obesity and cardiometabolic risk: a systematic review and meta-analysis. Metabolism. (2020) 111:154336. doi: 10.1016/j.metabol.2020.154336, 32777443AntoniRJohnstonKLCollinsALRobertsonMD. Effects of intermittent fasting on glucose and lipid metabolism. Proc Nutr Soc. (2017) 76:361–8. doi: 10.1017/S0029665116002986, 28091348SantosHOMacedoRC. Impact of intermittent fasting on the lipid profile: assessment associated with diet and weight loss. Clin Nutr. (2018) 24:14–21. doi: 10.1016/j.clnesp.2018.01.002, 29576352MacaskillP.GatsonisC.DeeksJ.HarbordR.TakwoingiY.Cochrane handbook for systematic reviews of diagnostic test accuracy. 2010, Version.MoherD. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. (2015) 4:1. doi: 10.1186/2046-4053-4-1, 25554246SterneJASavovićJPageMJElbersRGBlencoweNSBoutronI. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. (2019) 366:l4898. doi: 10.1136/bmj.l4898GuyattGHOxmanADVistGEKunzRFalck-YtterYAlonso-CoelloP. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. (2008) 336:924–6. doi: 10.1136/bmj.39489.470347.AD, 18436948BorensteinMHedgesLVHigginsJPTRothsteinHR. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. (2010) 1:97–111. doi: 10.1002/jrsm.12, 26061376ManoogianENWilkinsonMJO’NealMLaingKNguyenJVanD. Time-restricted eating in adults with metabolic syndrome: a randomized controlled trial. Ann Intern Med. (2024) 177:1462–70. doi: 10.7326/M24-0859, 39348690SuthutvoravutUAnothaisintaweeTBoonmanuntSPramyothinSSiriyothinSAttiaJ. Efficacy of time-restricted eating and behavioral economic intervention in reducing fasting plasma glucose, hba1c, and cardiometabolic risk factors in patients with impaired fasting glucose: a randomized controlled trial. Nutrients. (2023) 15:4233. doi: 10.3390/nu15194233, 37836517CramerHHohmannCLaucheRChoiKE(A)SchneiderNSteckhanN. Effects of fasting and lifestyle modification in patients with metabolic syndrome: a randomized controlled trial. J Clin Med. (2022) 11:4751. doi: 10.3390/jcm11164751, 36012990GuoYLuoSYeYYinSFanJXiaM. Intermittent fasting improves cardiometabolic risk factors and alters gut microbiota in metabolic syndrome patients. J Clin Endocrinol Metabol. (2021) 106:64–79. doi: 10.1210/clinem/dgaa644, 33017844LiCSadraieBSteckhanNKesslerCStangeRJeitlerM. Effects of a one-week fasting therapy in patients with type-2 diabetes mellitus and metabolic syndrome–a randomized controlled explorative study. Exp Clin Endocrinol Diabetes. (2017) 125:618–24. doi: 10.1055/s-0043-101700, 28407662HeMWangJLiangQLiMGuoHWangY. Time-restricted eating with or without low-carbohydrate diet reduces visceral fat and improves metabolic syndrome: a randomized trial. Cell Rep Med. (2022) 3:100777. doi: 10.1016/j.xcrm.2022.100777, 36220069SunXLiFYanHChangXYaoXYangX. Intermittent compared with continuous calorie restriction for treatment of metabolic dysfunction-associated steatotic liver disease: a randomized clinical trial. Am J Clin Nutr. (2025) 121:158–66. doi: 10.1016/j.ajcnut.2024.10.012, 39447676AlbostaMBakkeJ. Intermittent fasting: is there a role in the treatment of diabetes? A review of the literature and guide for primary care physicians. Clin Diabetes Endocrinol. (2021) 7:3. doi: 10.1186/s40842-020-00116-1, 33531076DiNicolantonioJJMcCartyM. Autophagy-induced degradation of Notch1, achieved through intermittent fasting, may promote beta cell neogenesis: implications for reversal of type 2 diabetes. Open Heart. (2019) 6:e001028. doi: 10.1136/openhrt-2019-001028, 31218007LiuYMeiHXueLChengCWuYZouC. Testing the carbohydrate-insulin model: short-term metabolic responses to consumption of meals with varying glycemic index in healthy adults. Cell Metab. (2025) 37:606–615.e3. doi: 10.1016/j.cmet.2025.01.015, 40043690MarkoDMConnMOSchertzerJD. Intermittent fasting influences immunity and metabolism. Trends Endocrinol Metab. (2024) 35:821–33. doi: 10.1016/j.tem.2024.04.014, 38719726MérianJGhezaliLTrenteseauxCDuparcTBeuzelinDBouguetochV. Intermittent fasting resolves dyslipidemia and atherogenesis in apolipoprotein E-deficient mice in a diet-dependent manner, irrespective of sex. Cells. (2023) 12:533. doi: 10.3390/cells12040533, 36831200MusazadehVZarezadehMGhalichiFKalajahiFHGhoreishiZ. Vitamin D supplementation positively affects anthropometric indices: evidence obtained from an umbrella meta-analysis. Front Nutr. (2022) 9:980749. doi: 10.3389/fnut.2022.980749, 36159504GallageSAliABarragan AvilaJESeymenNRamadoriPJoerkeV. Beneficial effects of intermittent fasting in NASH and subsequent HCC development are executed by concerted PPARα and PCK1 action in hepatocytes. bioRxiv. (2023). doi: 10.1101/2023.10.23.562885WangYWuR. The effect of fasting on human metabolism and psychological health. Dis Markers. (2022) 2022:5653739. doi: 10.1155/2022/5653739AksungarFBTopkayaAEAkyildizM. Interleukin-6, C-reactive protein and biochemical parameters during prolonged intermittent fasting. Ann Nutr Metab. (2007) 51:88–95. doi: 10.1159/000100954, 17374948Al-IslamMEFarisEKacimiSAl-KurdRAAFararjehMABustanjiYK. Intermittent fasting during Ramadan attenuates proinflammatory cytokines and immune cells in healthy subjects. Nutr Res. (2012) 32:947–55. doi: 10.1016/j.nutres.2012.06.021, 23244540KunduraciYEOzbekH. Does the energy restriction intermittent fasting diet alleviate metabolic syndrome biomarkers? A randomized controlled trial. Nutr. (2020) 12:3213., 17374948ParvareshARazaviRAbbasiBYaghooblooKHassanzadehAMohammadifardN. Modified alternate-day fasting vs. calorie restriction in the treatment of patients with metabolic syndrome: A randomized clinical trial. Complementary Therapies in Medicine (2019) 47:102187., 23244540RazaviRParvareshAAbbasiBYaghooblooKHassanzadehAMohammadifardN. The alternate-day fasting diet is a more effective approach than a calorie restriction diet on weight loss and hs-CRP levels. Int. J Vit. Nutr. Res. (2020)., 23244540
Edited by: Stefan Kabisch, Charité University Medicine Berlin, Germany
Reviewed by: Yvelise Ferro, Magna Græcia University, Italy
Ehsan Hejazi, Shahid Beheshti University of Medical Sciences, Iran