REF-1400 · Salvesen et al., 2012 · British Journal of Sports Medicine · Paper · PubMed

Where does the 90%-of-max heart-rate limit in pregnancy come from?

From 6 Olympic athletes on a treadmill at 23-29 weeks: the baby's heart slowed only above about 90%, then recovered on stopping. An observation, not a tested limit.

Main caveat: Read from the abstract only. The number of women above 90%, the number and duration of bradycardia episodes, how maximal heart rate was set and the stage…
limited, mixed evidenceMeasured on elite to world-class athletes

Salvesen, Hem and Sundgot-Borgen studied fetal wellbeing and uteroplacental blood flow during strenuous treadmill running in six pregnant Olympic-level endurance athletes, aged 28-37 and training 15-22 h a week before the pregnancy.

Each woman was tested once at 23-29 weeks and ran three to five submaximal workloads at about 60-90% of maximal oxygen consumption, with Doppler ultrasound of the uterine and umbilical arteries before, during and after exercise. Mean uterine artery volume blood flow fell to 60-80% of its initial value after warming up and stayed at 40-75% during exercise.

The fetal heart rate stayed within the normal range of 110-160 bpm as long as the woman exercised below 90% of her maximal heart rate. Fetal bradycardia and a high umbilical artery pulsatility index occurred when she exercised above 90% of maximal heart rate and mean uterine artery blood flow was below 50% of its initial value. Both normalised quickly after she stopped.

The authors conclude that exercise above 90% of maximal maternal heart rate in pregnant elite athletes "may compromise fetal wellbeing".

Our reading: this is the study the advice to stay under 90% of maximum heart rate in pregnancy is usually traced to (the review in REF-1396 pairs it with one other small study), and it is an observation in very few women during one session of three to five treadmill workloads: the 90% line marks where bradycardia appeared together with a halving of uterine flow, not a threshold that was tested, and the transient changes were not followed to birth.

A randomised crossover in 15 women found no fetal bradycardia before or immediately after 1-min intervals that peaked near 96% of maximum heart rate (REF-1401).

Protocol

Who: n = 6 pregnant Olympic-level athletes in endurance events, aged 28-37 years, training 15-22 h per week before the pregnancy.
When: one test each at 23-29 weeks of pregnancy.
Exercise: treadmill running, three to five submaximal workloads at approximately 60-90% of maximal oxygen consumption (stage durations and speeds are not in the abstract).
Measures: Doppler ultrasound of the uterine arteries (volume blood flow) and umbilical artery (pulsatility index) and the fetal heart rate, before, during and after exercise; maternal heart rate expressed as a percentage of maximal heart rate (how the maximum was established is not in the abstract).
Not in the abstract: how many women exceeded 90% of maximal heart rate, how many bradycardia episodes occurred and how long they lasted, how quickly values normalised, and any birth outcomes.
Funding and interests: not available from the abstract or the PubMed record; Crossref lists no funder. No device maker, federation or data platform is named.

Key Results

Condition during the sessionUterine artery volume blood flow (vs initial value)Fetal heart rate and umbilical artery
After warming up60-80%—
During exercise (about 60-90% VO2max)40-75%—
Mother below 90% of maximal heart rate—fetal heart rate within 110-160 bpm (normal range)
Mother above 90% of maximal heart rate and uterine flow under 50% of initial<50%fetal bradycardia and a high pulsatility index
After stopping—fetal heart rate and pulsatility index normalised quickly
What the 90% line is: the maternal heart rate above which bradycardia was observed in this sample, coinciding with uterine flow below half its initial value; the abstract gives no number of women or episodes behind it. The review in REF-1396 describes the study as seven athletes, of whom only two reached more than 90% of maximal heart rate; the abstract says six athletes.
What was not measured: any outcome at birth, any later session, any woman who was not an elite endurance athlete.

Mechanisms

Measured: uterine artery blood flow fell with exercise, to 40-75% of its initial value, and the fetal heart rate stayed normal until the mother exceeded 90% of maximal heart rate with uterine flow under half its initial value, when fetal bradycardia and a raised umbilical artery pulsatility index appeared; both recovered quickly once she stopped.
Authors' interpretation: exercise above 90% of maximal maternal heart rate in pregnant elite athletes may compromise fetal wellbeing.
Inference: the bradycardia went with a large fall in uterine blood flow, so the 90% heart-rate figure may stand in for how much blood is diverted from the uterus at that effort, which will differ between women, fitness levels and modes of exercise; the study cannot separate the two. Its long-term meaning is unknown: the review in REF-1396 states that it is not clear what the long-term impact of this transient bradycardia is.

Practical Application

Where does the 90%-of-max heart-rate limit in pregnancy come from? From 6 Olympic athletes on a treadmill at 23-29 weeks: the baby's heart slowed only above about 90%, then recovered on stopping. An observation, not a tested limit.
What it observed, exactly (B-, six elite women, one session each, from the abstract): three to five submaximal treadmill workloads at about 60-90% of VO2max (stage durations not in the abstract); fetal heart rate normal below 90% of the mother's maximal heart rate; fetal bradycardia above 90% when uterine flow was below half its initial value; quick recovery after stopping.It did not test whether staying under 90% prevents anything, or whether crossing it harms the baby.
Not a law: the number rests on very few women (how many crossed 90% is not in the abstract; a later review says two of seven).One-minute reps peaking near 96% of maximum caused no fetal bradycardia before or immediately after the session in 15 women (REF-1401); a meta-analysis counted three episodes after high-intensity exercise in two studies with 7 women, all resolved within 10 min (REF-1395), most likely these six athletes plus one case report, so not independent of this study; and vigorous work, mostly below 90%, into the third trimester did not change birth weight in a meta-analysis of low-risk pregnancies (REF-1396), whose authors nevertheless advise avoiding 90% or more until it is shown to be safe.Not from this paper; the base's own practice: no heart-rate number is given as a rule; perceived effort and the talk test set the ceiling, long near-maximal efforts are eased late in pregnancy, and any limit comes from the obstetric team.
Why heart rate is a weak governor here (inference): the signal that went with bradycardia was uterine blood flow, which a watch does not measure; maximal heart rate itself is hard to establish in pregnancy (the authors of REF-1396 recommend perceived exertion for this reason).
What it is good for as a source: where the 90% figure comes from and what it was, so the assistant can say it plainly. What it does not show: a safe or unsafe heart rate for anyone; any effect at birth; responses in recreational athletes, in intervals, in cycling or in early or late pregnancy.

Limitations

Read from the abstract only. The number of women above 90%, the number and duration of bradycardia episodes, how maximal heart rate was set and the stage protocol could not be checked; a later review's description (seven athletes, two above 90%) differs from the abstract (six).
Six women, one session each. No repeat tests, no control condition, no follow-up to birth.
Acute markers. Fetal heart rate and umbilical artery pulsatility index during and just after exercise; their meaning for the baby's later health is not known from this study.
One population and one mode: Olympic-level endurance athletes running three to five submaximal treadmill workloads in mid-pregnancy.
Funding and interests: not available from the abstract or the PubMed record; Crossref lists no funder.
How this entry was checked and changed
Why this level

(from the abstract) — the rubric places an elite case series at B; held one step lower because six women were each tested once, the outcomes are acute fetal markers (fetal heart rate, umbilical artery pulsatility index), not birth outcomes, and the abstract does not say how many women exceeded 90% of maximal heart rate, how maximal heart rate was established or how long the bradycardia lasted; a later review describes the study as seven athletes, two of whom exceeded 90% (REF-1396), which the abstract cannot confirm. Read from the abstract (closed access; Europe PMC holds no full text and OpenAlex lists no open copy)