Notes on the meta-analysis: HIIT and postprandial glucose/insulin (Khalafi et al., Diabetes Research and Clinical Practice, 2022)

Overview

• Topic: The impact of high-intensity interval training (HIIT) on postprandial glucose (PPG) and postprandial insulin (PPI).
• Type of study: Systematic review and meta-analysis.
• Authors and affiliation highlights: Khalafi et al. from departments of physical education, exercise physiology, and nutrition/dietetics (multinational collaboration).
• Objective: Compare HIIT to non-exercise control (CON) and to moderate-intensity continuous training (MICT) regarding glucose and insulin dynamics after a meal, assessed via oral glucose tolerance tests (OGTT).
• Population: Adults with normal glucose regulation and those with impaired glucose regulation (including prediabetes and type 2 diabetes risk states).
• Primary outcomes: Glucose AUC and insulin AUC derived from OGTT following HIIT vs CON or HIIT vs MICT.
• Key results at a glance:
  • HIIT vs CON: glucose AUC ↓ by a small, significant amount; insulin AUC ↓ similarly.
  • HIIT vs MICT: no significant difference in glucose AUC or insulin AUC.
• Take-home message: HIIT and MICT both reduce postprandial glycemia and insulinemia, with moderate-duration HIIT (≥ 8 weeks) showing clearer benefits, especially in those with impaired glucose status. HIIT can offer time-efficient benefits similar to longer-duration MICT.

Key definitions and concepts

• Postprandial glucose (PPG): blood glucose levels after a meal; an independent risk factor for cardiovascular disease and diabetes complications.
• Postprandial insulin (PPI): insulin response after meal intake.
• Postprandial glucose/insulin AUC: area under the curve for OGTT-derived glucose or insulin responses over a set time (commonly 120 or 180 minutes after a 75 g glucose load).
• HIIT (high-intensity interval training): repeated bouts at 80–100% HR_peak with recovery periods. SIT (sprint interval training): all-out or supramaximal efforts (>100% of maximal work rate/VO2max).
• LI-HIIT (long-interval HIIT): HIIT with longer intense bouts (e.g., several minutes).
• MICT (moderate-intensity continuous training): continuous exercise at moderate intensity.
• Outcome measurement tool: OGTT with a 75 g glucose load, typically 120–180 minutes long, with glucose AUC and insulin AUC calculated.
• Baseline glucose status subgroups:
  • Normal glucose: fasting glucose < 5.6 mmol/L and/or 2-h post-load glucose < 7.8 mmol/L.
  • Impaired glucose: fasting glucose ≥ 5.6 mmol/L or 2-h post-load glucose ≥ 7.8 mmol/L.

Methods (how the review was done)

• Protocol and registration:
  • Followed PRISMA guidelines; registered with PROSPERO (ID: CRD42021254582).
• Search strategy and data sources:
  • Databases: PubMed, Scopus, Web of Science.
  • Inclusion window up to October 2021.
  • Key terms covered HIIT varieties and glucose tolerance/OGTT terms; language restricted to English; human studies.
• Study selection criteria:
  • Randomized trials with a control or MICT arm.
  • Reports of glucose and/or insulin AUC via OGTT at baseline and post-intervention or change scores.
  • Intervention duration ≥ 2 weeks.
  • Excluded: non-English, non-peer-reviewed, pregnant populations, HIIT plus resistance training, cross-over designs were included in one study, etc.
• Data extraction and handling:
  • Independent extraction by two reviewers; cross-checked with a third party.
  • Parameters collected: study design, sample size, participant characteristics, intervention details (mode, intensity, frequency, duration), OGTT meal details, glucose/insulin AUC means and SDs (or SEs/CIs converted to SDs).
  • For two-arm HIIT trials or mixed-gender splits, multiple HIIT arms were included.
  • iAUC vs AUC: when both were reported, AUC data were used for meta-analyses.
• Quality assessment:
  • PEDro scale used (9 applicable items; blinding not feasible for exercise trials).
  • Score range: 0–9; higher is better quality.
  • Sensitivity analyses performed by removing studies one-by-one to assess influence on results.
• Statistical analysis:
  • Software: Comprehensive Meta-Analysis (CMA) Version 2.0.
  • Analyses:
  • HIIT vs CON on glucose AUC and insulin AUC.
  • HIIT vs MICT on glucose AUC and insulin AUC.
  • Effect size: Standardized Mean Difference (SMD) with 95% CI.
  • Model choice: random-effects model for heterogeneity; fixed-effects considered when heterogeneity was low.
  • Heterogeneity assessment: I^2 statistic.
  • Publication bias: Funnel plots and Egger’s test (significant bias set at p < 0.1).
  • Subgroup analyses predefined by:
  • Baseline glucose status (normal vs impaired).
  • Intervention duration (short < 8 weeks vs moderate ≥ 8 weeks).
  • HIIT type (LI-HIIT vs SIT).
  • Additional comparisons: independent t-tests for weekly duration comparisons between HIIT and MICT arms.

Included studies and participants

• Number of studies: 25 randomized trials.
• Total participants: 870.
• Gender composition:
  • 15 studies included both sexes.
  • 6 studies included males only.
  • 4 studies included females only.
• Age range and BMI:
  • Ages ranged from as young as 15.5 years to about 66 years.
  • BMI ranged from roughly 22.6 to 37.8 kg/m^2.
• Health status: mixture of healthy, overweight/obese, metabolic syndrome, prediabetes, or type 2 diabetes risk states.

Intervention characteristics (what the HIIT protocols looked like)

• Intervention duration:
  • Range: 2 to 16 weeks.
  • Most studies used 2, 8, or 12 weeks; a few used longer durations.
• Frequency:
  • 2 to 6 sessions per week; 3 per week most common.
• HIIT protocols used:
  • Repeated HIIT bouts, often cycling, running, walking, or swimming.
  • Common HIIT formats included alternating 3–4 minutes at high intensity and recovery or short all-out bouts (e.g., 6–30 seconds).
  • Intensities based on: %HRpeak, MAV, HRpeak, HRmax, RPE, or W_peak.
• MICT protocols:
  • 30–60 minutes at low to moderate intensity.
• Control conditions:
  • No exercise, maintain usual daily activities, or receive standard care.
• Dietary controls:
  • A few studies included dietary guidance that was matched across HIIT and control groups; otherwise diet not a major differentiator.
• OGTT details for outcomes:
  • Most used a 75 g glucose load with a 120-minute post-load window to assess glucose and insulin AUC.
  • Four studies used a 180-minute OGTT (instead of 120 minutes).
• Outcome focus:
  • OGTT-derived glucose AUC and insulin AUC as primary outcomes.

Meta-analysis results (HIIT vs CON)

• Glucose AUC:
  • Overall effect: HIIT reduces glucose AUC compared with CON.
  • Quantitative result: SMD = -0.37,
    \ 95\% \ CI = [-0.60, -0.13],\ p = 0.002
  • Heterogeneity: low (I^2 = 14.81%, p = 0.28).
  • Subgroup: baseline glucose level
  • Impaired glucose at baseline: SMD = -0.72,
    \ 95\% \ CI = [-1.37, -0.07],\ p = 0.03; I^2 = 59.05%, p = 0.04
  • Normal glucose at baseline: SMD = -0.22,
    \ 95\% \ CI = [-0.51, 0.05],\ p = 0.11; I^2 = 0.00%, p = 0.93
  • Subgroup: intervention duration
  • Moderate-duration (≥ 8 weeks): SMD = -0.41,
    \ 95\% \ CI = [-0.75, -0.08],\ p = 0.01; I^2 = 33.16%, p = 0.13
  • Short-duration (< 8 weeks): SMD = -0.31,
    \ 95\% \ CI = [-0.78, 0.14],\ p = 0.18; I^2 = 0.00%, p = 0.70
  • Subgroup: HIIT type
  • SIT: SMD = -0.56,
    \ 95\% \ CI = [-1.03, -0.10],\ p = 0.01; I^2 = 0.00%, p = 0.91
  • LI-HIIT: SMD = -0.33,
    \ 95\% \ CI = [-0.67, 0.02],\ p = 0.06; I^2 = 38.09%, p = 0.10
• Insulin AUC:
  • Overall effect: HIIT reduces insulin AUC vs CON.
  • Quantitative result: SMD = -0.36,
    \ 95\% \ CI = [-0.68, -0.04],\ p = 0.02
  • Heterogeneity: not significant (I^2 = 0.00%, p = 0.72).
  • Note: Subgroup analyses for insulin AUC by baseline glucose, duration, or HIIT type were not conducted due to a small number of studies.

Meta-analysis results (HIIT vs MICT)

• Glucose AUC:
  • No significant difference: SMD = -0.02,
    \ 95\% \ CI = [-0.20, 0.14],\ p = 0.76
  • Heterogeneity: none (I^2 = 0.00%, p = 0.99).
  • Subgroup analyses by baseline glucose, duration, or HIIT type did not reveal significant effects.
• Insulin AUC:
  • No significant difference: SMD = -0.08,
    \ 95\% \ CI = [-0.29, 0.12],\ p = 0.43
  • Heterogeneity: none (I^2 = 0.00%, p = 0.76).
  • Subgroup analyses not performed due to limited numbers.
• Overall interpretation for HIIT vs MICT:
  • HIIT and MICT yield similar improvements in postprandial glycemic markers when compared directly, with no clear advantage of HIIT over MICT for these outcomes in the analyzed data.

Quality, bias, and sensitivity analyses

• Study quality (PEDro):
  • Scores ranged from 4 to 9 out of 9; common limitations linked to lack of blinding (not feasible in exercise interventions).
• Publication bias:
  • Visual funnel plots suggested publication bias for glucose and insulin AUC in HIIT vs CON analyses.
  • Egger’s test: glucose AUC (p = 0.11) and insulin AUC (p = 0.98) did not jointly confirm bias; for HIIT vs MICT, funnel plots suggested no strong bias.
• Sensitivity analyses:
  • Excluding individual studies did not alter significance or direction of main results when more than 10 arms were present.

Subgroup and mechanistic insights

• Baseline glucose status effects:
  • Greater improvements in impaired glucose vs normal glucose, especially for glucose AUC in HIIT vs CON (consistent with reduced postprandial excursions in those with prediabetes/diabetes risk).
• Duration effects:
  • Moderate-duration HIIT (≥ 8 weeks) more consistently reduced glucose AUC than short-duration HIIT.
• HIIT modality effects:
  • SIT showed a clear reduction in glucose AUC; LI-HIIT showed a near-significant reduction with some heterogeneity across studies.
• Mechanistic considerations discussed in narrative synthesis:
  • Potential improvements in insulin secretion and action, hepatic glucose production, and whole-body insulin sensitivity with HIIT.
  • HIIT may improve pancreatic β-cell function, reduce hepatic gluconeogenesis, and enhance glucose disposal via skeletal muscle adaptations.
  • Reductions in postprandial glycemia may be mediated by improved insulin resistance, better β-cell signaling, and pancreatic–tissue coordination.
• Cross-study context:
  • Findings align with broader evidence that exercise improves postprandial glycemic control, sometimes with less impact on fasting glucose, and that postprandial markers are important predictors of cardiometabolic risk.

Practical implications and interpretation for practice

• Time efficiency:
  • HIIT achieved comparable postprandial benefits to longer-duration MICT with substantially less weekly exercise time (HIIT weekly duration ≈ 46% less than MICT in many arms).
  • Could be a practical option to improve postprandial glycemia and insulin responses in at-risk populations.
• Target populations:
  • Greatest likelihood of benefit appears in individuals with impaired glucose status rather than those with normal glucose regulation.
• Program design considerations:
  • Moderate-duration HIIT (≥ 8 weeks) may be more effective for PPG reduction.
  • Both LI-HIIT and SIT appear effective for lowering glucose AUC; however, energy expenditure, total work, and other protocol specifics differ and require more controlled comparisons.
• Matching total work/energy expenditure between HIIT and MICT is important in future trials to isolate modality effects from dose effects.

Limitations of the evidence

• Sample sizes: Several included trials had small samples; overall meta-analysis could be underpowered for some comparisons.
• Heterogeneity: HIIT protocols varied across studies (different modes, intervals, intensities, rest periods, and total work); while statistical heterogeneity was generally low to moderate, clinical heterogeneity is notable.
• Few studies directly contrasted HIIT vs MICT with energy-expenditure-matched designs; most HIIT arms used shorter total weekly time than MICT arms, which complicates direct dose-response interpretation.
• OGTT protocols varied: most used 120-minute tests; four used 180-minute OGTTs, which could affect AUC comparability.
• Cross-over design in one study and mixed-arm designs required careful handling to avoid unit-of-analysis errors.
• Generalizability: Many studies recruited overweight/obese or metabolic syndrome populations; less data in healthy young adults or older adults with normal glucose status.

Conclusions (synthesized)

• HIIT is an effective modality to reduce postprandial glucose and insulin AUC when compared with no-exercise control, particularly in individuals with impaired glucose tolerance and with longer (moderate-duration) intervention programs.
• When compared to MICT, HIIT does not show superior advantages for postprandial glycemic markers; HIIT and MICT yield similar reductions in PPG and PPI AUC.
• From a time-management perspective, HIIT offers a more time-efficient route to achieving similar postprandial glycemic benefits as longer-duration MICT, which may support adherence and real-world implementation.
• Future research needs:
  • Directly energy-expenditure matched HIIT vs MICT trials.
  • Larger samples and longer follow-ups to examine sustained postprandial glycemic benefits.
  • More precise analyses of SIT vs LI-HIIT and mechanistic pathways (β-cell function, hepatic metabolism, insulin signaling).

Appendix: supplementary notes

• The protocol and data availability: supplementary materials include additional tables and figures; data can be requested from the corresponding author or accessed via PROSPERO (ID: CRD42021254582).
• Figures referenced in the study include forest plots for HIIT vs CON and HIIT vs MICT effects on glucose AUC and insulin AUC (not reproduced here but described in text).