REF-1779 · Jin & Lee, 2026 · International Urogynecology Journal · Paper · PubMed

Pelvic-floor muscle training made the pelvic floor stronger in pregnancy (SMD 1.10, 95% CI 0.28-1.93) and after birth (SMD 1.74, 0.71-2.76) against usual care or no training, in a meta-analysis of 26 trials (I² 92-96%), but its effect on urinary symptoms, quality of life, sexual function and prolapse was less consistent (no pooled figures in the abstract); risk of bias not read

Main caveat: Heterogeneity of 92-96% in every strength pool; standardised differences hide what strength measure and units were used.
moderate evidenceMeasured on non-athletes and recreational peopleRead from the abstract only

Jin and Lee asked whether pelvic-floor muscle training helps women in pregnancy and after birth, whether adding feedback (for example biofeedback, which the abstract does not specify) adds anything, and whether timing matters.

They searched three databases for trials published from 2005 to 2025 that compared training, with or without feedback, against usual care or no intervention in pregnant or postpartum women; two reviewers selected trials and rated risk of bias with RoB 2.

Across 26 trials, training raised pelvic-floor muscle strength by a large standardised amount overall (SMD 1.39), in pregnancy (1.10) and after birth (1.74), with heterogeneity of 92-96% in every pool.

For urinary symptom severity, quality of life, sexual function and prolapse severity the abstract gives no pooled numbers and calls the evidence less consistent; programmes with feedback tended to show more reliable benefits.

The authors conclude that training improves strength and may reduce urinary symptoms, and that feedback helps women train correctly and stick with it rather than producing a larger physiological effect.

Participants

Pregnant or postpartum women in 26 trials; the split between pregnancy and postpartum trials, the number of women, their parity, continence status at entry and the countries are not in the abstract.

Protocol

Search: Ovid MEDLINE, Ovid Embase, Cochrane Library; studies published January 2005 to December 2025.
Inclusion: trials of pregnant or postpartum women comparing PFMT with or without feedback against usual care or no intervention; the abstract does not say whether all were randomised.
Risk of bias: RoB 2, by two independent reviewers; results not in the abstract.
Primary outcome: pelvic-floor muscle strength (measurement method not in the abstract). Secondary outcomes named: urinary severity and symptoms, quality of life, sexual function, pelvic organ prolapse severity.
Analysis: standardised mean differences with 95% CI; pooled separately for pregnancy and postpartum; how programmes with and without feedback were compared is not stated. Model type and registration not in the abstract.

Key Results

Outcome (PFMT vs usual care or no intervention)Effect (95% CI)I²
Pelvic-floor muscle strength, all trialsSMD 1.39 (0.74 to 2.05)95%
Strength, training in pregnancySMD 1.10 (0.28 to 1.93)92%
Strength, training after birthSMD 1.74 (0.71 to 2.76)96%
Urinary severity and symptoms, quality of life, sexual function, prolapse severity"less consistent"; no pooled values in the abstractnot given
With feedback vs withoutfeedback programmes "tended to show more reliable benefits"; no figuresnot given
The authors call the postpartum gain larger; the pregnancy and postpartum intervals overlap widely and no test of the difference is reported in the abstract.

Practical Application

Pelvic-floor training made the pelvic floor stronger, in pregnancy and after birth; its effect on urinary symptoms was less consistent in this review. Strength rose by large standardised amounts (SMD 1.10 in pregnancy, 1.74 after birth), with very high heterogeneity and risk of bias not reported in the abstract; for urinary symptoms the abstract gives no pooled figure.
How it bears on [SYNTH-95]: it supports the guide's reading and its rule not to promise that training stops leaking.Like REF-1709 (strength and endurance up, incontinence risk ratio not significant), it finds the pelvic floor trainable and the effect on symptoms uncertain; it does not report an odds or risk ratio for leaking, so it can neither confirm nor contest REF-1398's 37% lower odds.It mixes pregnancy and postpartum trials, and the abstract does not separate prevention in continent women from treatment of women who already leak.
Feedback may help women do the exercise correctly and keep doing it, the authors' reading; no adherence or accuracy figures are in the abstract, and they do not claim a larger physiological effect from feedback.
Inference, not tested here: for an athlete in pregnancy or after birth, pelvic-floor strength can be expected to improve with a programme; whether that translates into less leaking on runs or jumps was not measured.
Not treatment advice: the programme, any feedback device and what to expect from it are for a pelvic-health physiotherapist, with the obstetric provider, to decide.

Limitations

Heterogeneity of 92-96% in every strength pool; standardised differences hide what strength measure and units were used.
Risk-of-bias results, certainty, the number of women and the trial designs are not in the abstract; whether the trials overlap with REF-1709 or REF-1398 cannot be checked.
The secondary outcomes, which matter most to women (leaking, prolapse, sexual function), carry no pooled numbers in the abstract.
The feedback conclusion is qualitative ("tended to show"), and the claim about accuracy and adherence has no data in the abstract.
The pregnancy-versus-postpartum comparison is between groups of trials, not within the same women.
Written from the abstract and the PubMed record; full text not read; funding not stated in the abstract.
How this entry was checked and changed
Why this level

a meta-analysis of trials, risk of bias not read (abstract only): the abstract names RoB 2 but not what it found, gives no certainty rating and reports heterogeneity of 92-96%