📄 medicine

Divergent fasting lactate responses are associated with serum multi-omics remodelling after 14-week high-intensity interval training in type 2 diabetes

This study demonstrates that while high-intensity interval training improves glycaemic control in type 2 diabetes patients, divergent fasting lactate responses reveal distinct subgroups with unique serum multi-omics remodelling patterns involving lipid, redox, and immune pathways that are not fully captured by conventional metabolic markers.

Original authors: Jiuxiang Gao, Yuchen Wang, Hui Zuo, Xiaohui Liu, Tao Zhang, Ngai Choi, Chuanzhi Wang, Tianxiao Guo, Xinfei Wen, Weiliang Xia, Jingyan Tian, Kang Chen, Qingwei Lu, Liang Yu, Xiuqiang Wang, Kun Qian, Su

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jiuxiang Gao, Yuchen Wang, Hui Zuo, Xiaohui Liu, Tao Zhang, Ngai Choi, Chuanzhi Wang, Tianxiao Guo, Xinfei Wen, Weiliang Xia, Jingyan Tian, Kang Chen, Qingwei Lu, Liang Yu, Xiuqiang Wang, Kun Qian, Sulin Cheng

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Race with Different Finish Lines

Imagine a group of people with Type 2 diabetes (a condition where the body struggles to manage sugar) all running the same 14-week training program. The program is intense: high-intensity interval training (HIIT) on rowing machines, like sprinting and then resting, repeated over and over.

Usually, when we check if a workout program works, we look at the "big scoreboard": Did their blood sugar go down? Did they get fitter? In this study, everyone got better at these standard measures. Their blood sugar improved, and they could row harder.

However, the researchers decided to look at a different, less common scoreboard: Fasting Lactate.

Think of lactate not as "muscle poison" (an old myth), but as a fuel courier. When your muscles work, they send out lactate couriers to deliver energy to other parts of the body. The researchers wanted to see what happened to these couriers after 14 weeks of training.

The Surprise: Three Different Teams

When the researchers checked the "courier levels" (fasting lactate) before and after the training, they found that the group didn't react the same way. They split into three distinct teams:

The "Drop" Team (Lactate-Decreased): For these people, the number of lactate couriers in their blood went down significantly.
The "Steady" Team (Lactate-Maintained): For these people, the number of couriers stayed about the same.
The "Rise" Team (Lactate-Increased): Surprisingly, for these people, the number of lactate couriers in their blood went up.

This was a shock because everyone did the exact same exercise. It's like a class of students all studying the same way, but some get A's, some get B's, and some actually get worse grades on a specific test.

The Hidden Story: The "Rise" Team Had a Problem

Here is the most important part of the story. Even though the "Rise" Team got fitter and their blood sugar improved (just like the others), they had a hidden issue.

The "Drop" Team looked like the healthiest overall. Their insulin worked better, and their fat levels (triglycerides) were good.
The "Rise" Team, despite getting fitter, showed signs of metabolic stress. Their insulin resistance got worse, and their fat levels (triglycerides) went up.

The researchers suggest that for the "Rise" Team, the body was struggling to handle the fuel. Even though the exercise helped burn sugar, the body was having trouble managing fats and energy storage, leading to a buildup of these lactate couriers.

The Molecular Detective Work: Looking Under the Hood

To understand why these teams were different, the researchers acted like molecular detectives. They looked at the "gears and oil" inside the blood (proteins and metabolites).

They found that the difference wasn't just about lactate; it was about how the body handled fats and energy.

The "Drop" Team showed signs of smooth energy processing. Their bodies were good at handling fats and keeping their internal "machinery" (redox balance) clean.
The "Rise" Team showed signs of chaos in the lipid (fat) department. Their bodies were struggling with how to process fats, and their immune systems were sounding a bit of an alarm.

Think of it like two cars driving the same route. Both cars arrive at the destination (better fitness). But one car's engine is running cool and clean (the "Drop" Team), while the other car's engine is overheating and leaking oil (the "Rise" Team), even though it made it to the finish line.

The Baseline: Starting with a Head Start (or a Handicap)

Before the training even started, the researchers checked the people with diabetes against healthy people. They found that people with diabetes already had higher levels of lactate couriers in their blood, even when resting. This suggests that the "fuel delivery system" was already a bit jammed before they ever stepped on a rowing machine.

The Takeaway

This study tells us that exercise doesn't affect everyone's body chemistry in the same way, even if the results on the surface (like weight or blood sugar) look similar.

Fasting lactate acts like a "canary in the coal mine." It can reveal hidden metabolic stress that standard blood sugar tests might miss.
If someone's lactate goes up after exercise, it might be a signal that their body is struggling with fat and energy handling, even if their blood sugar looks good.
If someone's lactate goes down, it suggests their body is efficiently reorganizing its energy systems.

The researchers conclude that looking at lactate changes could help doctors understand who is truly getting a "deep" metabolic benefit from exercise and who might need extra help with their diet or medication to handle the fat and energy stress.

Note: The paper explicitly states these are "exploratory" findings. They are like a map of a new territory that needs more exploration, not a final rulebook for doctors to use immediately.

Technical Summary: Divergent Fasting Lactate Responses and Serum Multi-Omics Remodelling in Type 2 Diabetes After HIIT

Problem Statement
Type 2 diabetes mellitus (T2DM) is characterized by impaired glucose regulation, insulin resistance, and altered substrate metabolism. While exercise training is a cornerstone of management that improves glycaemic control and cardiorespiratory fitness, individual metabolic adaptation remains heterogeneous. Current literature recognizes lactate not merely as a metabolic waste product but as a carbon shuttle and signaling metabolite linking glycolytic flux, mitochondrial oxidation, and redox balance. However, gaps remain regarding whether fasting lactate responses to chronic exercise are uniform in T2DM, whether distinct lactate-response phenotypes exist, and how these phenotypes relate to broader serum metabolomic and proteomic remodelling. Furthermore, baseline differences in fasting lactate between T2DM patients and non-diabetic controls, potentially influenced by medication (e.g., metformin), require characterization to contextualize exercise-induced adaptations.

Methodology
The study employed a longitudinal design involving two cohorts:

Independent Baseline Validation Cohort: A cross-sectional comparison of 120 patients with T2DM and 232 non-T2DM controls to establish baseline fasting lactate levels and serum metabolomic profiles.
Exercise Intervention Cohort: A 14-week supervised rowing-based high-intensity interval training (HIIT) program involving 40 patients with T2DM (and 13 non-exercise controls). Participants underwent assessments at baseline (0 weeks), week 7, and week 14.

Key Procedures:

Exercise Protocol: Supervised rowing ergometer sessions (3x/week) with intensity prescribed based on maximal workload (40% low-intensity vs. 90% high-intensity intervals), progressing from weeks 1–7 to weeks 8–14.
Phenotyping: Participants were stratified into three groups based on the change in fasting lactate ( $\Delta$ Lactate) from week 0 to week 14: Lactate-Decreased (LD), Lactate-Maintained (LM), and Lactate-Increased (LI).
Multi-Omics Profiling:
- Metabolomics: Untargeted serum metabolomics via LC-MS (polar metabolites and lipids) using different workflows for baseline and longitudinal cohorts.
- Proteomics: Serum proteomics via Data-Independent Acquisition (DIA) mass spectrometry.
Clinical Metrics: Fasting glucose, insulin, HbA1c, HOMA-IR, continuous glucose monitoring (CGM) profiles, VO $_{2peak}$ , and body composition (DXA).
Statistical Analysis: Mixed-effects models for longitudinal data, Spearman correlations, and regression models adjusting for baseline lactate, age, sex, BMI, and drug class to identify molecular features associated with lactate response. Exploratory network analysis integrated clinical, metabolomic, and proteomic data.

Key Results

Baseline Distinctions: Patients with T2DM exhibited significantly higher fasting lactate (1.72 vs. 1.06 mmol/L) compared to non-T2DM controls, accompanied by impaired glucose-insulin homeostasis and exploratory metabolomic differences (e.g., elevated isomorellic acid, reduced brainesteroside C), though no features survived global FDR correction in the baseline cohort.
Heterogeneous Lactate Response: The 14-week HIIT intervention induced divergent fasting lactate trajectories:
- LD Group: Significant decrease ( $-1.08 \pm 1.20$ mmol/L).
- LM Group: Minimal change ( $+0.09 \pm 0.21$ mmol/L).
- LI Group: Significant increase ( $+1.15 \pm 0.60$ mmol/L).
- Baseline lactate was inversely associated with subsequent change ( $\rho = -0.63$ ).
Clinical Decoupling: While all exercise groups showed improved glycaemic control (HbA1c, CGM metrics) and VO $_{2peak}$ , the LI group displayed adverse trends in fasting insulin, HOMA-IR, and triglycerides compared to LD and LM groups. This suggests lactate response captures a metabolic dimension distinct from conventional glycaemic improvement.
Metabolomic Signatures: Divergent lactate responses were associated with specific serum metabolomic remodelling. The strongest adjusted candidate was PI(17:0/14:1(9Z)) (phosphatidylinositol). Other top features included phosphatidic acids (PG), phosphatidylcholines (PC), ubiquinone Q1, and acylcarnitines. These findings point to remodelling in phospholipid/sphingolipid pathways, fatty-acid handling, and redox metabolism.
Proteomic Signatures: Serum proteomics highlighted associations with lipid/adipokine signalling (e.g., ANGPTL8, CIDEB), immune/complement activity (e.g., FCN2), extracellular remodelling (e.g., NOTUM), and cellular stress/RNA regulation (e.g., DRG1).
Integrated Network: A lactate-centred network linked baseline lactate to insulin resistance (HOMA-IR) and, post-intervention, to triglycerides and age. The response layer connected lactate changes to glycerophospholipid remodelling, sphingolipid signalling, and systemic immune/lipid remodelling.

Significance and Claims
The paper claims that fasting lactate response serves as a heterogeneous metabolic adaptation phenotype in T2DM that is not fully captured by conventional glycaemic markers. Specifically:

Distinct Metabolic Axis: Fasting lactate response represents a distinct clinical axis from fasting glucose change, revealing metabolic heterogeneity even when glucose control improves uniformly.
Risk Stratification: The "Lactate-Increased" phenotype, despite improved fitness and glucose, is associated with adverse insulin resistance and triglyceride responses, suggesting a potential marker for residual cardiometabolic risk or substrate-handling stress.
Multi-Omics Context: The study provides exploratory evidence that divergent lactate responses are underpinned by systemic remodelling involving lipid handling (phospholipids, sphingolipids, acylcarnitines), redox balance, and immune/extracellular signalling, rather than isolated changes in lactate concentration.
Precision Medicine Potential: The authors suggest that fasting lactate may help identify patients with T2DM who exhibit adverse insulin and lipid responses to exercise, warranting targeted validation in larger cohorts to refine precision exercise medicine strategies.

The authors explicitly state that the metabolomic and proteomic findings are exploratory and hypothesis-generating, noting limitations such as modest sample size, lack of tissue-level validation (muscle/liver), and the absence of isotope-tracer flux quantification. They do not claim causal mediation or validated biomarker panels but rather a systems-level summary linking lactate dynamics to broader metabolic and proteomic remodelling.