Randomised controlled trial of pelvic floor muscle training during pregnancy.

Abstract

To examine a possible effect on labour of training the muscles of the pelvic floor during pregnancy.
Randomised controlled trial.
Trondheim University Hospital and three outpatient physiotherapy clinics in a primary care setting.
301 healthy nulliparous women randomly allocated to a training group (148) or a control group (153).
A structured training programme with exercises for the pelvic floor muscles between the 20th and 36th week of pregnancy.
Duration of the second stage of labour and number of deliveries lasting longer than 60 minutes of active pushing among women with spontaneous start of labour after 37 weeks of pregnancy with a singleton fetus in cephalic position.
Women randomised to pelvic floor muscle training had a lower rate of prolonged second stage labour (24%, 95% confidence interval 16% to 33%; 22 out of 105 women were at risk (undelivered) at 60 minutes in the survival analysis) than women allocated to no training (38% (37/109), 28% to 47%). The duration of the second stage was not significantly shorter (40 minutes v 45 minutes, P = 0. 06).
A structured training programme for the pelvic floor muscles is associated with fewer cases of active pushing in the second stage of labour lasting longer than 60 minutes.

Full text

Page 1

Primary care
Randomised controlled trial of pelvic floor muscle
training during pregnancy
Kjell Å Salvesen, Siv Mørkved
Abstract
Objectives To examine a possible effect on labour of
training the muscles of the pelvic floor during
pregnancy.
Design Randomised controlled trial.
Setting Trondheim University Hospital and three
outpatient physiotherapy clinics in a primary care
setting.
Participants 301 healthy nulliparous women
randomly allocated to a training group (148) or a
control group (153).
Intervention A structured training programme with
exercises for the pelvic floor muscles between the
20th and 36th week of pregnancy.
Main outcome measures Duration of the second
stage of labour and number of deliveries lasting
longer than 60 minutes of active pushing among
women with spontaneous start of labour after 37
weeks of pregnancy with a singleton fetus in cephalic
position.
Results Women randomised to pelvic floor muscle
training had a lower rate of prolonged second stage
labour (24%, 95% confidence interval 16% to 33%; 22
out of 105 women were at risk (undelivered) at 60
minutes in the survival analysis) than women
allocated to no training (38% (37/109), 28% to 47%).
The duration of the second stage was not significantly
shorter (40 minutes v 45 minutes, P = 0. 06).
Conclusions A structured training programme for
the pelvic floor muscles is associated with fewer cases
of active pushing in the second stage of labour lasting
longer than 60 minutes.
Introduction
Training the pelvic floor muscles during pregnancy
can prevent urinary incontinence,1 2and pregnant
women are encouraged to do exercises for these mus-
cles. A myth prevails among birth attendants that
strong pelvic floor muscles (for example, as a result of
horse riding) may obstruct labour.3However, training
of the pelvic floor muscles may produce strong and
well controlled muscles that will facilitate labour.4A
prospective study with 86 women found no effect of
such training on the course of delivery,5but possible
effects of pelvic floor muscle training on labour have
been sparsely scientifically documented.
The primary aim of this trial was to assess if train-
ing the muscles of the pelvic floor during pregnancy
could prevent urinary incontinence. Women in the
study group had stronger pelvic floor muscles and
reported less urinary incontinence after the training
period.2This report deals with secondary outcomes of
the trial. We wanted to study any effect of pelvic floor
muscle training on labour.
Methods
The population and methods are described elsewhere.2
One group of women (n=148) trained with a
physiotherapist for 60 minutes once per week for a
period of 12 weeks between the 20th and 36th week of
pregnancy. In addition, the women were encouraged to
perform eight to 12 intensive contractions of the pelvic
floor muscle at home twice a day. Adherence to the
training programme was 81% (120 women).2Women in
the control group (n=153) were not discouraged from
doing pelvic floor muscle exercises on their own. Figure
1 shows the flow of participants through the trial.
We reviewed hospital records two to three years
after delivery and recorded mode of delivery, epidural
analgesia or oxytocin augmentation during labour,
episiotomy,perineal tears,and neonatal outcomes.The
reviewer (KÅS) was not involved in training the women
and was blinded to group allocation while recording
and plotting the data. The midwife in charge of labour
judged the need for oxytocin augmentation or
episiotomy. The obstetrician on call decided the need
for operative delivery. Birth attendants were unaware
of women’s group status.
We recorded the lengths of the first and second
stages of labour from partograms. We defined the
onset of labour as the beginning of the active phase of
the first stage of labour,6or from the time of admission
if the cervix was dilated more than 3 cm on arrival.Our
definition of the second stage of labour was “active
pushing time.” We recorded the lengths of the first and
second stages of labour in minutes. Most clinics have
rules that limit the duration of the second stage.6In
Norway, the recommended “second stage rule” is one
hour, with active pushing after complete dilatation of
the cervix.7In this study we defined prolonged second
stage as active pushing for longer than 60 minutes.
This article was posted on bmj.com on 14 July 2004: http://bmj.com/cgi/
doi/10.1136/bmj.38163.724306.3A
Department of
Obstetrics and
Gynaecology,
Trondheim
University Hospital
St. Olav, N-7006
Trondheim, Norway
Kjell Salvesen
professor in obstetrics
and gynaecology
Department of
Community
Medicine and
General Practice,
Norwegian
University of
Science and
Technology, N-7489
Trondheim, Norway
Siv Mørkved
physiotherapist
Correspondence to:
K Å Salvesen
pepes@medisin.
ntnu.no
BMJ 2004;329:378–80
378
BMJ VOLUME 329 14 AUGUST 2004 bmj.com

Page 2

We undertook our analysis by intention to treat. It
was restricted to 111 women in the training group and
113 women in the control group (fig 1). They had
spontaneous start of labour after 37 weeks of
pregnancy with a singleton fetus in cephalic position.
We used a Kaplan-Meier survival analysis to test for
differences between groups in proportions of women
with prolonged second stage and the duration of
labour. We censored operative deliveries and deliveries
with a prolonged second stage. We also performed a
Cox regression analysis with possible confounding
variables and appropriate statistical tests for categori-
cal and normally distributed variables.We considered P
values < 0.05 significant.
Results
Women randomised to pelvic floor muscle training
had a lower rate of prolonged second stage labour than
women allocated to no training (table). The number
needed to treat (NNT) to benefit was 8. In a survival
analysis the difference between groups was 24% (95%
confidence interval 16% to 33%) for the training group
versus 38% (28% to 47%) for the control group.
The duration of the second stage of labour was not
statistically different between groups (40 min v 45 min,
P=0.06). Figure 2 shows the time to delivery from the
start of active pushing up to 60 minutes. The infants in
the training group were slightly younger and smaller
(table).ACoxregressionanalysiswithgestationallength,
birth weight, and head circumference as possible
confounders did not change the estimates materially.
Fewer women had breech presentations (fig 1; 1 v
9, P = 0.01). The rates of operative delivery for
prolonged second stage did not differ between the two
groups (table 1). Fewer women had episiotomies (51%
v 64%, odds ratio 0.59, 0.35 to 1.00; NNT 7), but we
found no other significant differences in outcomes
related to labour.Apgar scores and umbilical artery pH
did not differ between groups (data not shown).
Discussion
Pelvic floor muscle training during pregnancy results
in improved muscle control and strong flexible
muscles. The effect may be on the central nervous sys-
tem and the muscles, and training seems to facilitate
rather than obstruct labour.
Possible limitation of the study
Since this report deals with secondary outcomes and
the differences reached borderline significance, the
results should be viewed with caution. However, this
was a randomised controlled trial with blinding
technique, few withdrawals, and high adherence to the
training protocol.
The second stage of labour begins when cervical
dilatation is complete and ends with fetal expulsion.6We
defined the second stage as “active pushing time.” This
definition of the second stage is suitable for clinical
Signed informed consent for participation (n=342)
Excluded before randomisation (n=41)
Did not meet the inclusion criteria (n=31)
high risk pregnancy (n=10), geographical reasons (n=17), not nulliparous (n=4)
Withdrew/refused to participate (n=5)
Not able to meet for assessment (n=5)
Randomised (n=301)
Allocated to intervention (n=148)
Received standard intervention as allocated
(n=120)
Did not receive standard intervention as
allocated (n=28)
Allocated to control group (n=153)
Received intervention as allocated (n=153)
Lost to follow up before labour (n=5)
Lost to follow up, delivery data (n=5)
Lost to follow up before labour (n=7)
Lost to follow up, delivery data (n=0)
Fetus in cephalic position with spontaneous
start of labour after 37 weeks (n=111)
Fetus in cephalic position with spontaneous
start of labour after 37 weeks (n=113)
Breech presentation after 37 weeks (n=1)Breech presentation after 37 weeks (n=9)
Twin pregnancy (n=1)
Preterm delivery (n=6)
Planned caesarean section (n=2)
Induced labour (n=17)
Twin pregnancy (n=3)
Preterm delivery (n=8)
Planned caesarean section (n=2)
Induced labour (n=11)
Fig 1 Flow of participants through the trial
Mean outcome variables (with standard deviations) among 224 women with a singleton fetus in cephalic position with spontaneous
start of labour after 37 weeks of pregnancy. Values are numbers (percentages) of women unless otherwise indicated
Study group (n=111)
260 (195 to 325)
5 (5)
40 (33 to 47)
22 (21)
15 (15)
56 (51)
7 (6)
47 (42)
69 (63)
Control group (n=113)
259 (215 to 303)
3 (3)
45 (38 to 52)
37 (34)
19 (17)
72 (64)
9 (8)
49 (43)
74 (66)
P value
0.44*
0.46†
0.06*
0.03†
0.57†
0.05†
0.64†
0.88†
0.67†
Median duration of first stage of labour in minutes (95% CI)
Acute caesarean section in the first stage
Duration of second stage of labour in minutes (95% CI)
Deliveries with prolonged second stage (≥60 minutes)
Vaginal operative delivery in the second stage
Episiotomy
Third or fourth degree tears
Epidural analgesia
Oxytocin augmentation
Characteristics of infants at birth:
Gestational age in days
Weight (g)
Length (cm)
Head circumference (cm)
280 (8.6)
3514 (391)
50.4 (1.9)
35.3 (1.4)
283 (7.1)
3622 (431)
50.8 (1.7)
35.9 (1.6)
0.04‡
0.05‡
0.11‡
0.002‡
*Log rank test.
†?2test or Fisher’s exact test.
‡t test.
Primary care
379
BMJ VOLUME 329 14 AUGUST 2004bmj.com

Page 3

research,since the start of active pushing is easily identi-
fied from partograms, and it is related to a second stage
rule.7The risk for bias should be small since the reviewer
of the partograms was blinded to group status.
Chance finding
The difference in breech presentations should be
interpreted as a possible chance finding. The women
trained in different positions, but there is insufficient
evidence from well controlled trials to support the use
of postural management of breech presentations.8
Role of body mass index or exercise
We found no differences in body mass index or self
reported regular physical exercise after the training
period (data not shown). This argues for an effect of
increased strength and better control of pelvic floor
muscles rather than a general effect of physical training
during pregnancy. New trials from other populations
are needed.
Contributors: KÅS and SM were involved in designing and con-
ducting the study, analysing the data, and writing the report.
Kari Bø contributed to the design of the study. The physiothera-
pists Hildegunn Børsting, Trude Hoff Leirvik, Bente Olsen,
Monica U Tøndel, and Bjørg Vada led the group training
sessions. Pål Romundstad gave statistical advice. KÅS is the
guarantor.
Funding: Norwegian Fund for Postgraduate Training in
Physiotherapyand Norwegian
Association.
Competing interests: None declared.
Ethical approval: Regional medical ethics committee.
Women’sPublicHealth
1Reilly ETC, Freeman RM, Waterfield MR, Waterfield AE, Steggles P,
Pedlar F. Prevention of postpartum stress incontinence in primigravidae
with increased bladder neck mobility: a randomised controlled trial of
antenatal pelvic floor exercises. Br J Obstet Gynaecol 2002;109:68-76.
Mørkved S, Bø K, Schei B, Salvesen KÅ. Pelvic floor muscle training dur-
ingpregnancyto preventurinary
randomized controlled trial. Obstet Gynecol 2003;101:313-9.
UK Midwifery Archives. Hannah, midwife. Does horse-riding affect the
pelvic floor? www.radmid.demon.co.uk/pelvicfloor.htm (accessed 25 Jun
2004).
Baby Centre. How dopelvic
www.babycentre.co.uk/refcap/536339 (accessed 25 Jun 2004).
Nielsen CA, Sigsgaard I, Olsen M, Tolstrup M, Danneskiold-Samsoee B,
Bock JE. Trainability of the pelvic floor. A prospective study during preg-
nancy and after delivery. Acta Obstet Gynecol Scand 1988;67:437-40.
Cunningham FG, MacDonald PC, Gant NF, Leveno KJ, Gilstrap LC,
Hankins GDV, Williams obstetrics 20th edition. Stamford, CT: Appleton
and Lange, 1997.
Bergsjø P, Maltau JM, Molne K, Nesheim B-I. Obstetrikk [in Norwegian].
Oslo: Universitetsforlaget, 1987.
Hofmeyr GJ, Kulier R. Cephalic version by postural management for
breech presentation Cochrane Database Syst Rev 2000;(3):CD000051.
(Accepted 25 May 2004)
2
incontinence: A single-blind
3
4floor exerciseshelp me?
5
6
7
8
doi 10.1136/bmj.38163.724306.3A
Time to delivery (minutes)
% of women still in the second stage of labour
0 10 2030 405060
0
0.2
0.4
0.6
0.8
1.0
Control
Training
Fig 2 Survival plot of the duration of the second stage of labour for
women in the training group (n=105) and the control group (n=109).
Discrepancies in numbers are due to some missing data in each
group (six in the training group and four in the control group).
Operative deliveries for fetal distress (n=9) and slow progress (n=3)
during the first hour, and all deliveries lasting longer than 60
minutes (n=59) were censored. Log rank test, P=0.06 for comparison
of the two survival plots in the Kaplan-Meier analysis
What is already known about this topic
Exercising the pelvic floor muscles prevents
urinary incontinence in about one in six women
during pregnancy and in one in eight women after
pregnancy
It increases the strength of the pelvic floor muscles
What this study adds
Intensive training of the pelvic floor muscles
during pregnancy seems to facilitate rather than to
obstruct labour.
It could prevent a prolonged second stage in one
in eight women
For the pleasure of their company
It was a normal surgery, nothing exciting, the usual complaints.
Patients came and went. On looking back, however, I find that
three patients stand out.
One patient gave me a hug, another told me that I was the best
doctor in the world, and the third played a song for me with great
care and feeling. Such encounters are not common; in fact, they
are unique. The hug was given before I had time to anticipate it;
the compliment was well meant without any hidden agenda; and
the song—well, that was quite a treat on a grey winter morning.
It so happened that all three men had Down’s syndrome, and
any concern about having a too familiar relationship with their
doctor wouldn’t cross their minds. What they also had in
common was their total dedication to and enjoyment of the
moment, of an encounter with me. I know they will bring the
same quality of human awareness and directness to the next
person they meet, and the next doctor will be the best ever. These
three men, with ages from 22 to 60, are all healthy, have a good
lifestyle, don’t have many worries, and take pleasure in family and
social life. In short, they are a delight to meet.
On reflection, I feel troubled. These are exactly the people whom
we medics try with all our skills to prevent being born. What is so
awful, so dreadful about their destiny that it is not worth living?
The likelihood that one of them will commit a crime or
become a drug addict is quite slim. All three are in need of
guidance, which will cost society money, but is such money less
well spent than on the military or the new parliament building in
Scotland? And I even refuse to believe they give their family so
much heartache, after the initial shock at birth.
I am in the unusual situation of seeing patients with Down’s
syndrome quite regularly in my surgery, as more than 3% of our
practice population have this condition. And I can tell you, I don’t
mind a bit.
Marga Hogenboom general practitioner,Camphill Medical Practice,
Aberdeen
Primary care
380
BMJ VOLUME 32914 AUGUST 2004 bmj.com