Not a fact anymore

Prescribed bed rest prevents preterm birth and other common pregnancy complications.

What we know now

Routine activity restriction has not been shown to prevent preterm birth and can cause deconditioning, bone loss, psychosocial harm, financial disruption, and increased risk of blood clots.

Why it changed

Rest seemed mechanically and physiologically plausible and became common despite limited trial evidence. Reviews found no convincing benefit, while research documented substantial harms and professional societies recommended against routine restriction.

Status
Overturned
Category
Medicine
Accepted for
≈145 years
Accepted approximately
19th century–late 20th century
Changed approximately
1990s–2010s

Bed rest sounded like a conservative way to protect a pregnancy. If uterine activity, cervical change, bleeding, or physical strain seemed capable of triggering complications, reducing movement appeared to remove a plausible stressor.

The practice spread far more widely than the evidence supporting it. Women at risk of preterm birth or other complications were sometimes told to spend weeks largely confined to bed.

Trials and systematic reviews failed to show that routine activity restriction prevents preterm birth. Meanwhile, the harms became easier to measure: muscle deconditioning, bone loss, increased risk of venous thromboembolism, anxiety, social isolation, lost income, and disruption of family life.

Professional guidance consequently moved away from routine bed rest for common obstetric indications.

That does not mean every pregnant patient should ignore all restrictions. Temporary limits may be appropriate for a specific procedure, symptom, hospitalization, or individualized medical reason.

The overturned claim is the broad preventive prescription: prolonged inactivity has not been shown to protect pregnancies as a general strategy, and it is not biologically neutral.

Evidence

Sources and what they establish

Current evidence

Historical context

Related entries

Or go somewhere else: