| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
JMA Mens, MD, is Researcher, Department of Rehabilitation Medicine, Faculty of Medicine, Erasmus University, Rotterdam, the Netherlands, and Head, Department of Spine Rehabilitation, Spine and Joint Centre, Westerlaan 10, 3016 CK, Rotterdam, the Netherlands (sjceco{at}wxs.nl). Address all correspondence to Dr Mens at the second address
CJ Snijders, PhD, is Professor of Medical Technology and Head, Department of Biomedical Physics and Technology, Faculty of Medicine and Allied Health Sciences, Erasmus University
HJ Stam, MD, PhD, is Professor of Medical Rehabilitation and Head, Department of Rehabilitation Medicine, Institute of Rehabilitation Medicine, University Hospital Rotterdam, Erasmus University
Submitted December 8, 1999;
Accepted May 28, 2000
 | Abstract |
|---|
Key Words: Low back pain • Physical therapy • Pregnancy • Pubic symphysis • Randomized clinical trial • Sacroiliac joint
| Introduction |
|---|
 TopAbstract Introduction MethodsResultsDiscussionConclusionAppendixReferences |
|---|
Many hypotheses on the pathogenesis of peripartum pelvic pain focus on decreased stability of the pelvic girdle.7–13 These hypotheses are based on the assumption that stability of the pelvic girdle is provided, in part, by the coarse texture of the sacroiliac (SI) cartilage surfaces, the complementary ridges and grooves, and the undulated shape ("form closure")7–9 and, in part, by compressive forces of muscles, ligaments, and the thoracolumbar fascia ("force closure").7–13 Muscles that generate a force perpendicular to the SI joints or that increase tension on the sacroiliac ligaments or thoracolumbar fascia could generate forces that stabilize the SI joint.8–13 The internal and external oblique abdominal muscles (which we refer to as the "anterior diagonal trunk muscle system") and the latissimus dorsi muscle, the transversospinal parts of the erector spinae muscle (especially the multifidus muscle), and the gluteus maximus muscle (which we refer to as the "posterior diagonal trunk muscle system") seem to be appropriate for this task.8–13
From this perspective, we believe that training of the diagonal trunk muscle systems will benefit people with peripartum pelvic pain, partly by increasing muscle force and endurance (the ability to function over a long period of time).8–13 We believe that many patients reduce peripartum pelvic pain with therapy that focuses on increased force production. It remains unclear, however, whether success is real and, if so, whether it is due to the increased stability resulting from the increased force of the diagonal trunk muscles, or to spontaneous recovery, placebo effects, or applied co-interventions. Exercises to achieve this goal of increased force production could exacerbate symptoms by the loading of the spinal and pelvic structures.14 The purpose of our study was to investigate whether the results of treatment of peripartum pelvic pain with graded exercises of the diagonal trunk muscle systems are better than the results without these exercises.
| Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAppendixReferences |
|---|
years, contacted the outpatient clinic of the Institute of Rehabilitation Medicine of the University Hospital Rotterdam in the Netherlands. A brochure with information about peripartum pelvic pain and a medical history questionnaire were mailed to all patients, and 891 questionnaires were returned. Eighty-four patients appeared to fulfill the selection criteria and were invited to visit the outpatient clinic and to participate in the trial. A physical examination was performed, routine blood tests (ie, sedimentation rate, white cell count, hemoglobin, alkaline phosphatase, calcium) and urine tests (ie, protein, glucose, sediment) were made, and radiographs of the lumbar spine and pelvis were made according to the procedure described by Chamberlain.15 In the second selection phase, 40 patients were excluded for various reasons (Figure), 11 of them because they were treated with exercises during the waiting period between the registration and the first examination.
|
Exclusion criteria were:
During a 2 -year period, 44 women with persistent pelvic pain after pregnancy (mean age=31.7 years, SD=3.2, range=23.6–37.5; mean period postpartum=4.1 months, SD=2.2, range=1.7–5.6) were included in the study. Subjects were randomly assigned to 1 of 3 groups: (1) a group that performed exercises to increase the force of the diagonal trunk muscle systems (experimental group), (2) a group that received training of the longitudinal trunk muscle systems (rectus abdo-minis muscle, longitudinal parts of the erector spinae muscle, and quadratus lumborum muscle) (control group 1), and (3) a group that was instructed to refrain from exercises (control group 2). No differences between the 3 groups were found for prognostic indicators, co-interventions during the study, and baseline values of outcome measures (Tab. 1).
|
Each subject received a 30-minute videotape in which explanations were given about the possible cause of peripartum pelvic pain, prognosis, and therapeutic possibilities. Furthermore, apart from ergonomic advice, information was given on how to behave if activities caused pain and how to use a pelvic belt (a nonelastic strap that gives support to the pelvic girdle). The last part of the videotape differed, depending on group assignment (Appendix). Videotape 1, which was given to subjects in the experimental group, gave instructions on how to train the diagonal trunk muscle systems. In videotape 2, which was given to subjects in control group 1, there was a demonstration of light exercises of the longitudinal trunk muscle systems, which we viewed as placebo exercises. Videotape 3, which was given to subjects in control group 2, illustrated how subjects should try to gradually increase the activities of daily living and to refrain from exercises. The same individuals described the exercises on each of the videotapes.
The exercises that we chose are based on the opinions of Kendall et al18 and, therefore, have not been subjected to research. The frequencies were based on the opinions prevalent in sports training19 rather than on the results of systematic research. In conformity with these opinions, heavy exercises to gain muscle force and endurance were performed 3 times a week. These exercises were partly isometric and partly nonisometric. Two series of exercises were performed, with a rest of 5 minutes between the series of exercises. The subjects had to try to gradually increase the amount of repetitions per series, and they were guided by their pain and fatigue. Light exercises designed to improve muscular awareness and recruitment were performed 3 times a day. To control and facilitate adherence, the subjects had to complete weekly visual analog scales (VASs) for pain and fatigue and send them to an administrative assistant by means of an addressed, prepaid envelope.
A designated form inquired about the subjects' frequency of training, their use of medication or a pelvic belt, whether their general health was disturbed (eg, by a cold or urinary tract infection), and whether they were working. Subjects were given the opportunity to ask questions. These questions were submitted to the principal investigator (JM) without the subjects being identified by the administrative assistant. The administrative assistant contacted the subjects by telephone if no forms were received or the forms were not completed appropriately, and to answer any questions. To check correctness of the exercise technique, the subjects were asked during the evaluation to demonstrate the way they trained after 8 weeks of intervention.
Assignment
After informed consent was obtained from the subjects, they were given a videotape in a sealed envelope. At home, they played the videotape in order to first learn to which group they had been assigned. Prior to the start of the trial, numbered, sealed envelopes containing a copy of 1 of the 3 different videotapes were prepared in random order.
Outcome Assessment
The outcome of treatment was assessed after conclusion of the 8-week intervention. To prevent the influence of fluctuations of complaints associated with the menstrual cycle, the day of the week, and the hour of the day, the second examination in the hospital was planned to occur exactly 8 weeks after the first examination on the same day of the week and at the same time. Because no agreed-on measures to evaluate treatment for peripartum pelvic pain exist, we decided to use scales for general health: a measure of pain, a measure of fatigue, and the Nottingham Health Profile (NHP).20 Moreover, we used a posterior pelvic pain provocation test (PPPP test)17 and a radiographic examination according to the procedure described by Chamberlain.15 Chamberlain described how mobility of the pelvic joints could be assessed by measuring the shift between the pubic bones when a person stepped with weight bearing while alternating between the left and right lower extremities. Berezin21 used the Chamberlain method to compare the mobility of the pelvic joints of women with and without pelvic girdle pain in the puerperium. The measured shift between the pubic bones was 5.9±3.3 mm in women with complaints and 1.9±2.2 mm in women without complaints (P=.0000). Because the validity of measurements obtained with the PPPP test and the radiographic examination as measures of effect is not known, we considered these measurements to be secondary.
Primary outcome measures.
The subjects scored their global impression of improvement on a 3-point Likert scale (1=worse, 2=unchanged, 3=improved). The mean severity of pain and fatigue were scored on a 100-mm horizontal VAS by asking subjects to rate their pain (or fatigue) in the morning (or in the evening), where 0 represented "no pain (or not tired) at all" and 100 represented "very severe pain (or extremely tired)."22,23 Because of the large variation in pain and fatigue between morning and evening, scores were obtained at both times of the day. The subjects were encouraged to complete the forms each week on the same day and at the same hour (preferably during the evening on the weekend). Reliability and validity for these measures have been examined and shown to be good.22,23
The 6 main outcome scales of the NHP were used to measure various aspects of perceived health: energy, pain, emotional reactions, sleep, social isolation, and physical mobility.20 The reliability and validity for this measure for assessing health-related quality of life have been examined and shown to be good.24
Secondary outcome measures.
Gluteal pain provoked by the PPPP test on the left and right sides was scored on a 2-point scale (yes or no). Radiographic examination was performed to assess mobility of the pubic symphysis during weight bearing while alternating between the right and left lower extremities.
Sample size calculations were based on a clinical success rate in the experimental group of at least 20% higher than in the control groups (outcome measure: global impression of improvement). The target sample was estimated at approximately 30 patients per group (
=.05, ß=.20). When about half of the subjects had been enrolled, an interim analysis was planned to investigate whether it was necessary to include 30 subjects per group before conclusions could be made. When 44 subjects were enrolled, this analysis took place and the study was terminated.
Blinding
It was impossible to keep subjects unaware of the kind of intervention they received. Before the subjects were randomly assigned to groups, they were told by the principal investigator that the approach to treatment for persistent peripartum pelvic pain involved the combination of use of a pelvic belt and ergonomic advice and that the study was initiated to answer the question of whether the addition of exercises is beneficial or harmful, or has no influence. In order not to influence the subjects, all the assessment forms had to be completed at home.
The examiner who determined the score of the PPPP test after 8 weeks of intervention was unaware of the subjects' group assignment. Before the examination, the subjects were asked not to inform the examiner about their treatment. The investigator was unaware of the subjects' group assignments during interpretation of the radiographic findings.
Data Analysis
SPSS statistical software* was used for data analysis. The 3 groups were checked when the study began for similarity of prognostic indicators and for initial values of outcome measures. All outcome measurements were analyzed as intention to treat (data obtained for dropouts were included in the results). Changes were calculated for each subject by subtracting the results obtained at the beginning of the study from those obtained after 8 weeks. Differences between measurements obtained at the beginning of the study and at conclusion of the study were analyzed using a one-sided analysis of variance. Categorical data were compared with the Kruskal-Wallis test (P<.05 was considered significant).
Subjects in the 2 exercise groups were encouraged to increase the number of repetitions. If a subject was unable to perform the exercises, she could decrease the amount or stop. Four subjects in the experimental group (25%) stopped the exercises due to increase of pain (2 in the sixth week, 1 in the seventh week, and 1 in the eighth week). One subject in control group 1 stopped the exercises due to increased pain in the eighth week. No subject was lost at conclusion of the study. Four subjects (2 in the experimental group and 2 in control group 1) refused to participate in the second examination because of exacerbated symptoms after the first examination; for these subjects, the results were based only on the primary outcome measurements. The 2 subjects in the experimental group who refused to participate in the second examination classified their result as "worse" and stopped the exercises before the end of the study.
All subjects considered the videotape comprehensive, and almost all of the subjects found the information sufficient (they had no further questions about the disease and how to cope with the problems). All subjects in the 2 exercise groups demonstrated that the way they had trained was the correct way. No subjects in control group 2 had performed any structured training.
| Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAppendixReferences |
|---|
Comparison of results at the end of the 8-week intervention showed no differences for the primary outcomes measures between the experimental group and both control groups (Tab. 2). The statement that global improvement in the experimental group was not 20% better than in the control groups could be made with a confidence level of 95% for control group 1 and with a confidence level of more than 99% for control group 2. With respect to change in the PPPP test scores on the right side, the experimental group scored better than the control groups (P<.05).
|
| Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAppendixReferences |
|---|
Our results show that 63.6% of the subjects improved during the program. We found no evidence that training of the diagonal muscle systems of the trunk was beneficial for patients with peripartum pelvic pain. There were minimal differences in results between the experimental group and both control groups. After 8 weeks of intervention, a difference was shown in only one item: the PPPP test on the right side improved more in the experimental group than in the control groups. The interpretation of this finding was hindered by the difference in baseline values of the groups for this test. The difference might be the result of a "regression to the mean."
A surprisingly large percentage of the experimental group (25%) had to cease training because of pain or fatigue. Many subjects in this group complained of increasing pain during the exercises; the majority attributed the pain to the exercise aimed at strengthening the hip extensors (ie, raising the lower extremity in prone position).
We conclude that training of the diagonal trunk muscle systems, without individualized coaching, as done in this study, is not more effective than low graded training of the longitudinal trunk muscle systems or no exercises. Training of the hip extensors in our subjects may have increased pain to such an extent that any benefit derived from increased stability of the pelvis was obscured. Vleeming and colleagues7,10–13 reported that tension of the gluteus maximus and hamstring muscles increases the tension of the ligaments and decreases the mobility of the SI joints. A decrease in the movements of the SI joints may be beneficial, but extra loading on the ligaments probably is not beneficial.
A literature search was made in MEDLINE for the period 1966 to 1998, in the Cochrane Controlled Trials Register,27 and in the proceedings of the 3 interdisciplinary congresses on low back pain.28–30 Two randomized clinical trials and 3 nonrandomized intervention studies on peripartum pelvic pain were found.26,31–34 In the first randomized trial,Östgaard et al26 investigated the preventive value of a back school education and training program during pregnancy. They concluded that the program could reduce short-term sick leave due to peripartum pelvic pain, provided that the instructions were individual based. Nilsson-Wikmar et al31 compared the effects of exercises given by a physical therapist with the effects of a program of home training and stretching and with the effects of a program without exercises; no differences were found. Noren et al34 studied the effects of an individual-based education and training program in patients who were pregnant and had peripartum pelvic pain. They found that days lost to sick leave were reduced in the intervention group compared with a group of women from another antenatal clinic who received no treatment. In a prospective nonrandomized trial, Dumas et al32 investigated the value of exercise classes in the prevention and treatment of peripartum pelvic pain. They found no effect on back pain during pregnancy and after childbirth. Mantle et al33 studied the effects of ergonomic advice on the development and course of back pain during pregnancy. The treated group in their study scored better than the control group. The results of our study and the literature search agree with the hypothesis that giving information about peripartum pelvic pain in combination with ergonomic advice is beneficial. Until now, however, the studied exercises have shown no additional value in the treatment of peripartum pelvic pain during pregnancy or during the first 6 months after childbirth.
At the beginning of the study, we hoped that the results would support the hypothesis that training of the diagonal muscle systems in patients with peripartum pelvic pain is better than other exercises or better than no exercises. If this would have been the case, the study would provide a rationale for physical therapy in this patient category. The results suggest that providing appropriate information to a patient with peripartum pelvic pain is useful. The results also suggest that training of the hip extensor muscles in a patient with peripartum pelvic pain may worsen the situation. Neither suggestion, however, was proven in our study. The most important consequence of our study is that the hypothesis about form closure and force closure, described by Vleeming and colleagues,7,10–13 needs to be revised. It is recommended that studies be done to examine the effect of training the diagonal trunk muscle systems without the hip extensors, eventually in combination with exercises, to strengthen the transverse abdominal muscle. Training of the transverse abdominal muscle has been recommended for lumbar segmental instability35,36 and as a stabilizer for the pelvic girdle.9,37
| Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAppendixReferences |
|---|
| Appendix |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAppendixReferences |
|---|
|
| Footnotes |
|---|
The study was approved by the Ethics Committee of the University Hospital Rotterdam.
* SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611. 
| References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAppendixReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
