STRRIDE Finds Aerobic Volume Drives LP-IR Improvement

Adding resistance training to aerobic exercise did not improve LP-IR or the Diabetes Risk Index more than aerobic training alone in the STRRIDE trials. That result does not make lifting irrelevant to metabolic health. It does mean that if your target is the NMR-measured lipoprotein pattern captured by LP-IR, aerobic work appears to do the heavy lifting.

The finding cuts against the simplified version of the usual advice: do cardio and weights because more modalities must produce more metabolic benefit. Exercise science rarely rewards that kind of arithmetic. Different tests measure different biology, and the answer depends on what you measure.

What STRRIDE actually found

Ross et al.'s 2021 STRRIDE reanalysis applied two NMR-derived scores to stored blood samples from three randomized trials. LP-IR combines six lipoprotein measures: concentrations of large VLDL, small LDL, and large HDL particles, plus the sizes of VLDL, LDL, and HDL particles. The Diabetes Risk Index adds the branched-chain amino acids valine and leucine.

These are surrogate biomarkers, not diabetes diagnoses or cardiovascular events. Still, they aim at a metabolic signature associated with insulin resistance: high triglyceride-rich VLDL particles, smaller LDL particles, and unfavorable HDL patterns. LP-IR has also correlated with HOMA-IR and clamp-derived glucose disposal in validation work cited by the STRRIDE authors. Correlation does not prove that changing the score changes a person's clinical outcome.

In STRRIDE I, sedentary adults with dyslipidemia completed about six months of one of three aerobic programs. The largest LP-IR improvement occurred with high-volume vigorous exercise: 23 kilocalories per kilogram per week (KKW) at 65% to 80% of VO₂peak. Low-volume vigorous exercise—14 KKW at the same intensity—produced an intermediate response. Low-volume moderate exercise—14 KKW at 40% to 55% of VO₂peak—produced the smallest improvement.

The pattern points first to aerobic volume and then to intensity. At the same 14-KKW volume, vigorous training outperformed moderate training; the 23-KKW vigorous program did best. The published information available here does not provide the LP-IR effect sizes, confidence intervals, or between-group p-values, so the size and precision of those differences cannot be judged from numerical estimates.

The combined-training null result needs a careful reading

In the eight-month STRRIDE AT/RT randomized trial, dyslipidemic adults were assigned to aerobic training alone (14 KKW at 65% to 80% of VO₂peak), resistance training alone three days per week, or both programs together. Aerobic training improved LP-IR and the Diabetes Risk Index. Resistance training alone did not produce comparable improvement, and the combined program did not improve either score beyond aerobic exercise alone.

This is the provocative part: participants did both forms of training, yet the combined arm did not beat the aerobic arm on these two NMR scores. But a non-significant difference is not proof that the treatments are physiologically identical. The trial arms were relatively small—roughly 30 to 50 participants per arm by the reported trial scale—and the available results do not include confidence intervals for the aerobic-versus-combined comparison. A modest added effect could have escaped detection.

Nor did STRRIDE show that resistance exercise fails to improve insulin sensitivity. Earlier publications from the same AT/RT cohort found benefits on traditional insulin-action measures, including an intravenous-glucose-tolerance-test insulin sensitivity index and HOMA-IR. Muscle contraction can improve glucose uptake through pathways that a lipoprotein-based score may not fully capture. The sensible interpretation is narrower: resistance training did not add a measurable benefit for these specific lipoprotein-and-amino-acid risk scores under this trial's conditions.

Why LP-IR deserves attention—but not blind faith

LP-IR is more than a dressed-up cholesterol panel. It integrates particle concentrations and sizes measured by NMR spectroscopy, features that can track insulin-resistant dyslipoproteinemia. In a 2023 validation study in South Asian adults, Fosam et al. reported that LP-IR accurately identified insulin resistance in a population where HOMA-IR can perform poorly because insulin assays are not standardized. That supports LP-IR as a useful measurement, especially when conventional fasting-insulin calculations are shaky.

The score can also change with metabolic state. A 2024 prospective intervention study in adults with overweight or obesity found that LP-IR moved in favorable directions with weight loss and shifted away from that profile with weight maintenance. Separately, a 2024 Mediterranean-diet analysis linked greater adherence with more favorable NMR lipoprotein subclasses and glucose-metabolism markers in 196 older adults at high cardiovascular risk, assessed at baseline and one year later.

Those studies show that NMR lipoprotein measures respond to lifestyle and metabolic change. They do not establish that LP-IR should replace HOMA-IR, glucose tolerance testing, or clinical diabetes risk assessment. The decisive question is prospective prediction: whether LP-IR or the Diabetes Risk Index predicts future type 2 diabetes better than conventional measures in comparable people. The evidence summarized here does not resolve that head-to-head question.

Why this does not overturn the case for doing both

Combined training performs well when researchers measure conventional glucose and insulin outcomes. A 2024 systematic review and meta-analysis found that combined exercise training improved glucose-metabolism and inflammatory markers in sedentary adults, using outcomes such as fasting glucose, fasting insulin, and HOMA-IR. It does not answer STRRIDE's precise question because the available report does not provide a direct aerobic-only versus combined-training comparison using LP-IR.

Likewise, a 2022 randomized trial in adults with overweight or obesity found larger HOMA-IR-based improvements with concurrent aerobic-and-resistance training than with metformin. That comparison shows that combined training can work very well; it lacks an aerobic-only arm, so it cannot isolate what resistance exercise added beyond the aerobic component.

Other work reinforces that distinction. A 2025 bed-rest study assigned 22 adults aged 55 to 65 to 14 days of head-down bed rest with either passive movement or a combined aerobic, HIIT, and resistance protocol. The exercise program attenuated the bed-rest-related increase in insulin resistance and increased α-klotho. Because every exercise participant performed multiple modalities, the study cannot attribute the protection to resistance training, aerobic training, or their combination.

In a separate 2025 three-arm trial, older adults with obesity followed caloric restriction plus aerobic exercise, resistance exercise, or both. The study detected modality differences in insulin-secretion-related outcomes. It supports the broader point that exercise modes can affect insulin biology differently; it does not test LP-IR and therefore cannot overturn or confirm the STRRIDE result.

How to use this without turning a biomarker into a religion

If you monitor LP-IR or NMR lipoprotein profiling, protect your aerobic training volume. STRRIDE's strongest response came from 23 KKW of vigorous aerobic exercise, while 14 KKW produced less improvement and moderate intensity at that volume produced the least. The study supports aerobic volume as the main lever for this endpoint, with intensity adding something when volume stays constant.

Keep resistance training in the program unless you have a compelling reason not to. It can improve conventional insulin-resistance measures, preserve or build muscle, and supports bone and function—outcomes LP-IR does not measure. The practical conflict appears only when time is limited enough that adding lifting displaces meaningful aerobic work. For LP-IR specifically, STRRIDE suggests that trade may reduce the adaptation you want.

STRRIDE offers a biomarker-specific conclusion, not a verdict against resistance training: aerobic exercise drove the measurable improvement in LP-IR and the Diabetes Risk Index, while combined training did not exceed aerobic training alone.

That is a useful correction to exercise dogma, not a reason to choose sides in the cardio-versus-weights argument. Train both for broad health. If LP-IR is the scoreboard, make sure aerobic volume gets enough minutes on the field.