Table 2. Grading of Recommendations Assessment, Development, and Evaluation (GRADE) Summary of Findings and Certainty of Evidence for Meal Timing for Anthropometric and Metabolic Measures.
| Outcomes | Follow-up | No. of participants (studies) | Certainty of the evidence (GRADE)a | Mean difference (95% CI) |
|---|---|---|---|---|
| Time-restricted eating | ||||
| Weight, kg | Follow-up: range 12 to 52 wk | 1527 (17 RCTs) | Lowb,c | –1.37 (–1.99 to –0.75) |
| Relative weight loss, % | Follow-up: range 12 to 52 wk | 770 (8 RCTs) | Lowb,c | −1.82 (−2.81 to −0.83 ) |
| BMId | Follow-up: range 12 to 52 wk | 993 (14 RCTs) | Lowb,c | –0.44 (–0.67 to –0.2) |
| Lean mass, kg | Follow-up: range 12 to 52 wk | 858 (11 RCTs) | Moderateb | –0.42 (–0.7 to –0.15) |
| Waist circumference, cm | Follow-up: range 14 to 52 wk | 530 (7 RCTs) | Lowb | –1.96 (–3.24 to –0.68) |
| HbA1c, % | Follow-up: range 12 to 52 wk | 1022 (13 RCTs) | Lowb,c | –0.08 (–0.15 to –0.01) |
| Fasting plasma glucose, mg/dL | Follow-up: range 12 to 52 wk | 1161 (14 RCTs) | Moderateb | –1.15 (–1.77 to –0.53) |
| LDL, mg/dL | Follow-up: range 12 to 52 wk | 1151 (13 RCTs) | Lowb,d | –1.51 (–1.3 to 4.32) |
| Systolic blood pressure, mmHg | Follow-up: range 12 to 52 wk | 1065 (13 RCTs) | Lowb,c,d | –0.54 (–2.42 to 1.33) |
| Diastolic blood pressure, mmHg | Follow-up: range 12 to 52 wk | 1065 (13 RCTs) | Lowb,c,d | –1.14 (–2.41 to 0.14) |
| Energy intake, kcal/d | Follow-up: range 12 to 52 wk | 843 (10 RCTs) | Lowb,c | –164 (–242.21 to –84.85) |
| Meal frequency, lower frequency vs higher frequency | ||||
| Weight, kg | Follow-up: range 12 to 52 wk | 396 (5 RCTs) | Very lowe,f | –1.84 (–3.55 to –0.13) |
| BMId | Follow-up: range 12 to 26 wk | 369 (5 RCTs) | Very lowe,f | –0.65 (–1.09 to– 0.21) |
| Lean mass, kg | Follow-up: range 13 to 26 wk | 91 (2 RCTs) | Very lowe,f | 1.35 (–0.18 to 2.88) |
| Waist circumference, cm | Follow-up: range 12 to 13 wk | 228 (3 RCTs) | Very lowe,f,g | –0.83 (–4.34 to 2.68) |
| HbA1c, % | Follow-up: range 12 to 12 wk | 310 (4 RCTs) | Very lowe,f,g | –0.14 (–0.39 to 0.11) |
| Fasting glucose, mg/dL | Follow-up: range 12 to 52 wk | 490 (6 RCTs) | Very lowe,f,g | –5.4 (–17.22 to 6.42) |
| LDL, mg/dL | Follow-up: range 12 to 52 wk | 458 (5 RCTs) | Very lowe,f,g | 4.27 (–3.34 to 11.87) |
| Systolic blood pressure, mmHg | Follow-up: mean 52 wk | 140 (1 RCT) | Very lowe,g | 0.7 (–3.28 to 4.68) |
| Diastolic blood pressure, mmHg | Follow-up: mean 52 wk | 140 (1 RCT) | Very lowe,g | –0.1 (–3.45 to 3.25) |
| Energy intake, kcal/d | Follow-up: range 12 to 26 wk | 159 (2 RCTs) | Very lowf,g | –0.64 (–208.34 to 207.07) |
| Calorie distribution, distribution of calories earlier vs later in the biological day | ||||
| Weight, kg | Follow-up: range 12 to 12 wk | 272 (4 RCTs) | Lowe | –1.75 (–2.37 to –1.13) |
| BMId | Follow-up: range 12 to 12 wk | 272 (4 RCTs) | Very lowe,f | –1.06 (–1.82 to –0.3) |
| Waist circumference, cm | Follow-up: range 12 to 12 wk | 272 (4 RCTs) | Very lowc,e | –1.77 (–2.89 to –0.65) |
| HbA1c, % | Follow-up: range 12 to 12 wk | 162 (2 RCTs) | Very lowb,g | –0.01% (–0.06 to 0.04) |
| Fasting glucose, mg/dL | Follow-up: range 12 to 12 wk | 272 (4 RCTs) | Very lowe,f,g | –3.06 (–6.73 to 0.6) |
| LDL, mg/dL | Follow-up: range 12 to 12 wk | 272 (4 RCTs) | Very lowe,f,g | –3.95 (–11.67 to 3.77) |
| Systolic blood pressure, mmHg | Follow-up: range 12 to 12 wk | 110 (2 RCTs) | Very lowe,h | –4.96 (–8.54 to –1.38) |
| Diastolic blood pressure, mmHg | Follow-up: range 12 to 12 wk | 110 (2 RCTs) | Very lowc,e,g | –4.64 (–10.79 to 1.51) |
| Energy intake, kcal/d | Follow-up: range 12 to 12 wk | 80 (1 RCT) | Very lowb,c,g | –51 (–96.6 to –5.4) |
Abbreviations: BMI, body mass index; HbA1c, glycated hemoglobin; LDL, low-density lipoprotein cholesterol; RCT, randomized clinical trial.
SI conversion factors: To convert fasting glucose to mmol/L, multiply by 0.0555. To convert LDL-cholesterol to mmol/L, multiply by 0.0259.
GRADE Working Group grades of evidence: high certainty: high confidence that the true effect lies close to that of the estimate of the effect; moderate certainty: moderate confidence that the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different; low certainty: the true effect may be substantially different from the estimate of the effect; very low certainty: the true effect is likely to be substantially different from the estimate of effect. RCTs were downgraded from an initial high rating if a serious flaw was present in any of the following domains: risk of bias (eg, large proportion of information from studies at high risk of bias that is sufficient to affect the interpretation of results), inconsistency (ie, substantial unexplained heterogeneity I2>75%), indirectness (ie, major limitations of the generalizability of the results), imprecision (ie, 95% CIs overlap with minimally important difference for benefits or harms), and publication bias (or small study effect, where >25% of participants were from small studies with <100 participants).
Risk of bias was assessed as serious due to many trials with concerns primarily related to blinding and missing data; these studies were rated down by 1 level for risk of bias.
Inconsistency was assessed as serious due to dissimilarities in point estimates, lack of overlap in CIs, and statistical evidence of heterogeneity; these studies were rated down by 1 level for inconsistency.
Calculated as weight in kilograms divided by height in meters squared.
Risk of bias was assessed as very serious due to many trials with concerns primarily related to blinding and missing data; these studies were rated down by 2 levels for risk of bias.
Inconsistency was assessed as very serious due to dissimilarities in point estimates, lack of overlap in CIs, and statistical evidence of heterogeneity; these studies were rated down by 2 levels for inconsistency.
Imprecision was assessed as very serious because the 95% CI included a point of no difference and failed to exclude important benefits; these studies were rated down by 2 levels for imprecision.
Imprecision was assessed as serious because the 95% CI included a point of no difference and failed to exclude important benefits; these studies were rated down by 1 level for imprecision.