Can Preconception TRE and Exercise Prevent Gestational Diabetes?

Starting metabolic habits before pregnancy may be a better target for gestational diabetes prevention than waiting until pregnancy is well underway. That idea makes biological sense. It has not yet been proven.

The first randomized trial to test it found a lower 2-hour glucose level at 28 weeks of pregnancy among participants assigned to time-restricted eating and high-intensity interval training (HIIT) before conception. The difference was 0.48 mmol/L, but it did not reach statistical significance (p=0.08). More importantly, participants completed only 37% to 44% of the prescribed program during pregnancy, when the trial measured its primary outcome. The study tested an appealing strategy; it did not establish that the strategy prevents gestational diabetes.

A prevention experiment that begins before pregnancy

Gestational diabetes mellitus (GDM) is usually identified with an oral glucose tolerance test at 24 to 28 weeks of pregnancy. By then, pregnancy-related insulin resistance has been building for months. The question behind the BEFORE THE BEGINNING trial was whether changing diet and activity before conception could improve the starting metabolic position.

Researchers in Trondheim, Norway, conducted a single-center, parallel-group randomized controlled trial involving 111 pregnancies in people at elevated GDM risk because of a BMI of at least 25, prior GDM, or a family history of diabetes. They assigned participants to usual care or to a program that began before conception and continued through pregnancy:

• Time-restricted eating: an eating window of 10 hours or less per day, at least five days per week.
• HIIT: at least 100 weekly Physical Activity Intelligence points.

The primary endpoint was 2-hour glucose during an oral glucose tolerance test at gestational week 28. The intervention group had a 0.48 mmol/L lower value than the standard-care group, with p=0.08. That result points in a favorable direction, but a non-significant p-value means the trial did not rule out chance as an explanation. The publication does not provide a confidence interval for this estimate in the evidence summarized here.

This distinction matters. A directional result is not proof, and it is also not proof of no effect. The trial enrolled roughly 55 participants per group, a small sample for a prevention study with a variable glucose endpoint. Its observed difference may have been clinically interesting, but the study lacked the precision to make a firm claim.

The intervention dose fell apart at the decisive time

Adherence is the main practical problem. Participants followed the program reasonably well before conception, but adherence during pregnancy fell to 37% to 44%. An intention-to-treat analysis therefore compared standard care with an assigned intervention that many participants could not sustain at the prescribed dose.

Pregnancy is not a convenient time to add a demanding food schedule and HIIT target. Nausea, fatigue, changing work demands, childcare, and medical concerns all compete for space. That is not an excuse to ignore adherence; it is the result that a future trial must solve. A program that works only when people can follow it perfectly has not solved the real-world problem.

The trial also cannot separate the effects of its two components. It included no time-restricted-eating-only arm and no exercise-only arm. A lower glucose value, if real, could reflect HIIT, the eating schedule, both together, or participants making other health changes alongside the program.

Why the hypothesis remains plausible

Preconception metabolic health predicts later GDM risk. An automated machine-learning study using preconception data found that variables including BMI, fasting glucose, and lifestyle factors could predict GDM risk. Prediction does not prove that changing those variables will prevent GDM, but it identifies a sensible intervention window.

The same research group had previously found that a seven-week combined TRE and HIIT program in reproductive-aged women with overweight reduced HbA1c, body mass, fat mass, and visceral fat. Those are relevant metabolic targets, though they are not pregnancy outcomes and cannot substitute for a GDM-prevention result.

There is also broader support for combining exercise with intermittent fasting. A 2025 systematic review and meta-analysis in adults with overweight or obesity found that adding exercise to intermittent fasting improved body composition and cardiometabolic outcomes beyond intermittent fasting alone. The population was not pregnant, so the finding supports the program's rationale rather than its safety or effectiveness in pregnancy.

Early timing helps sometimes, not reliably

Other lifestyle trials show why timing alone cannot carry the whole argument. The BEFORE THE BEGINNING paper reports that RADIEL, which started a lifestyle intervention before 20 weeks of pregnancy in high-risk women, reduced GDM incidence by 39%. In contrast, UPBEAT and LIMIT, which began at 16 to 20 weeks, did not significantly reduce GDM incidence. PREPARE, a preconception weight-loss trial, reduced GDM diagnosis in early pregnancy but did not significantly reduce GDM at standard 24-to-28-week screening.

These trials differed in intervention content, populations, and outcomes, so they do not produce a clean ranking of start dates. They do show that starting early is plausible but insufficient. Sustained behavior, intervention intensity, baseline metabolic risk, and the way a trial defines GDM all matter.

A systematic review and meta-analysis on precision prevention found that responses to GDM-prevention lifestyle programs varied with characteristics such as prior GDM, insulin resistance, ethnicity, and BMI. That makes a mixed "high-risk" group scientifically useful but statistically messy: a treatment that helps one phenotype may appear weaker when averaged across people with different metabolic pathways to GDM.

Related evidence from a meta-analysis of prediabetes interventions found different responses among people with isolated impaired fasting glucose and those with isolated impaired glucose tolerance. These are not pregnancy trials, but they reinforce a restrained conclusion: metabolic phenotype can change how a lifestyle program performs.

TRE during pregnancy has a less encouraging record

A separate 2024 randomized feasibility trial from the same research group tested TRE alone in a similar high-risk pregnant population. Participants used a 10-hour eating window for five weeks during the second and third trimesters. The approach was feasible and acceptable, but researchers detected no cardiometabolic effect.

That study was short and designed for feasibility, so it cannot prove that TRE has no metabolic effect in pregnancy. It does weaken any assumption that an eating window by itself can offset rising gestational insulin resistance. In BEFORE THE BEGINNING, the exercise component may have mattered more, or the preconception phase may have needed to create durable changes that low pregnancy adherence could not maintain.

Another limitation is dietary content. BEFORE THE BEGINNING prescribed the timing of food but did not set calorie or diet-quality targets. An isocaloric intermittent-fasting meta-analysis found that intermittent fasting was not consistently superior to continuous calorie restriction. If some fasting benefits arise because people eat less overall, a time window without calorie guidance may produce modest and variable effects.

What the trial can and cannot tell someone planning pregnancy

The trial offers a reasonable research case for entering pregnancy with better fitness, activity habits, and metabolic health. It does not establish a specific TRE-and-HIIT prescription as a proven way to prevent GDM. Its single Norwegian center also enrolled a predominantly European cohort, limiting confidence that the result applies to populations with higher GDM burden, including South Asian, Middle Eastern, East Asian, and sub-Saharan African populations.

Neonatal outcomes offer no reason for extra optimism. The companion secondary-outcomes paper places its findings alongside broader reviews in which pregnancy lifestyle programs have shown minimal or no average effect on birth weight and neonatal outcomes. A preconception start has not yet shown that it changes this pattern.

Preconception TRE plus HIIT is a credible prevention hypothesis with one underpowered, poorly adherent trial behind it—not a validated GDM-prevention protocol.

For people with prior GDM, overweight, insulin resistance, or a strong family history of diabetes, preconception care is worth discussing with an obstetric clinician or diabetes specialist. A larger multicenter trial should test sustained behavioral support, dietary guidance alongside TRE, and prespecified risk phenotypes. Until then, the most accurate verdict is simple: the preconception window is promising enough to study and too uncertain to sell.