Gynecological Surgery and Low Back Pain in Older Women

Abstract

Objective: To determine sacroiliac joint compliance characteristics and pelvic floor movements in older women relative to gynecological surgery history and back pain complaints. Design: Single-visit laboratory measurement. Setting: University clinical research center. Participants: Twenty-five women aged 65 years or older. Outcome Measures: Sacroiliac joint compliance measured by Doppler imaging of vibrations and ultrasound measures of pelvic floor motion during the active straight leg raise test. Results: Doppler imaging of vibrations demonstrated test reliability ranging from 0.701 to 0.898 for detecting vibration on the ilium and sacrum sides of the sacroiliac joint. The presence of low-back pain or prior gynecological surgery was not significantly associated with a difference in the compliance or laxity symmetry of the sacroiliac joints. No significant difference in pelvic floor movement was found during the active straight leg raise test between subject groups. All P values were ≥.4159. Conclusions: Prior gynecological surgery and low-back pain were not significantly associated with side-to-side differences in the compliance of the sacroiliac joints or in significant changes in pelvic floor movement during a loading maneuver in a group of older women.

Extensive low-back pain (LBP) research has modeled lumbopelvic stability derived from both global (multisegmental) muscles and more local or single-segment muscles.¹ A control system model integrated the active anatomic elements, passive anatomic elements, and the arthrokinetic control elements (sensory afferents; spinal, subcortical, and cortical neurons; efferent motor elements).^2,3 These models fostered extensive investigation into neuromuscular control and stability of the lumbopelvic junction with early observations of altered motor control in specific muscles in LBP populations^4–7 with rapid carryover into rehabilitation efforts focused on neuromuscular reeducation of the deep abdominal and pelvic floor muscles, part of the so-called core stability system, in LBP populations.^8,9

The abdominal wall muscles are postulated to contribute to lumbopelvic stability in part due to linkage with the thoracolumbar fascia system.¹⁰ Deep abdominal muscles were observed to have altered activation patterns in chronic LBP populations during specific load challenges compared with normal populations.^5,6,11,12 Further work described similar premovement activation implicated as postural contributions to spine stability, in the pelvic floor muscles, deep fibers of the multifidus muscle, and the respiratory diaphragm.^13–22 Altered neurophysiological control of pelvis and abdominal muscles using electromyographic, ultrasound, athletic performance, and postural measures has been linked to LBP conditions.^11,23–27 These observations of changes in muscle neurophysiological control in LBP populations, combined with observations of the stabilizing role of abdominal and pelvic muscles, suggest that pelvic and abdominal muscle functional integrity are important for some role in spine stability. The potential for a primary muscle injury leading to motor control alteration that is clinically silent after injury although resulting in new adverse tissue loads, perhaps cumulative in potential injury over time, is an intriguing model to link apparently minor events with later LBP conditions.^28,29

In parallel with the description of spine motor control and stability has been the enhanced understanding of lumbopelvic functional anatomy. Extensive anatomical and neurophysiological research has defined the biomechanical role provided by the pelvis linking the lower limb propulsion system and the spine.³⁰ The sacroiliac joint (SIJ) has been extensively studied relative to its ligament support and muscle-stabilizing system.^31–33 The SIJ contribution to postpartum LBP and nonradicular LBP has been described with biomechanical and injection response methods.^34–37 Asymmetric (left vs right) SIJ compliance, measured by Doppler imaging of vibration (DIV), has been correlated with peripartum LBP.^34,38 The DIV method was used to quantify the compliance of the SIJ by measuring the relative reduction in transmitted vibration energy applied by an oscillator motor, from ilium to sacrum, using real-time color Doppler images (Figure 1 ). DIV has been described in pelvis and cadaver models and human postpartum pain syndromes, and it was used to demonstrate the effects of pelvic belts and muscle activation on SIJ compliance.^34,38–47 The DIV method appears to offer an objective method to detect altered SIJ biomechanics relative to traditional physical examination methods primarily criticized for their lack of correlation with intra-articular diagnostic gold standard injections.^36,48–51 Specific quantification of the biomechanical integrity of the SIJ load transfer function influenced by joint architecture, ligament and fascia integrity, and muscle function is possible.

Figure 1.

Doppler imaging of vibration schematic demonstrating the piston transmitting vibration from the motor up through the anterior superior iliac spine. The ultrasound probe is shown straddling the sacroiliac joint at the posterior superior iliac spine level.

Pelvic floor dysfunction has been implicated with altered lumbopelvic motor control in LBP populations with ultrasound observations of greater decline in the pelvic floor during a standardized load challenge, the active straight leg raise (ASLR) test.^26,52–54 The ASLR (Figure 2 ) places a load (subject’s leg) across the pelvis, and abnormal responses have been linked to posterior pelvic pain in patients with SIJ dysfunction.^55,56 The normalization in pelvic floor descent in LBP patients with external pelvis compressive force applications during the ASLR supports the hypothesis that an altered motor control strategy (MCS) for load transfer across the pelvis creates a force imbalance that compromised the support system of the pelvic floor.²⁶ One hypothesis is that a lax or overly compliant SIJ requires greater muscle-stabilizing forces to transfer load across the lumbopelvic junction, creating excessive intra-abdominal pressure precipitating greater pelvic floor descent.⁵³

Figure 2.

Active straight leg raise schematic: intra-abdominal pressure influences pelvic pressure and pelvic floor movement during the leg lift.

The potential for a clinically silent injury precipitating SIJ dysfunction and LBP in later life due to altered abdominal and pelvis muscle function led us to study older women who had undergone gynecological surgery (GS). We speculated that women with a GS history would demonstrate altered MCS and SIJ biomechanics compared to women without a GS history. The observation that older women treated with hysterectomy reported a significantly higher degree of moderate severity of LBP in later life offers preliminary epidemiological support for our hypothesis.⁵⁷ GS might serve as a clinically silent marker for altered motor control of the lumbopelvic junction, possibly from surgically induced dysfunction of local muscles, leading to LBP later in life. The purpose of our study was to determine if older women with LBP and a history of GS demonstrated altered SIJ compliance using DIV and greater pelvic floor descent during the ASLR using real-time sonography to measure pelvic floor descent. Increased SIJ compliance would support the hypothesis that ineffective stabilizing MCS results in excessive stress, strain, and creep, as forces normally modulated by the MCS are applied to passive joint structures in excess. We hypothesized that women with a history of GS suffering from LBP would demonstrate the greatest degree of SIJ compliance asymmetry, similar to younger women with pregnancy-associated LBP shown previously.^38,40 In addition, it was hypothesized that women with a history of GS suffering from LBP would demonstrate a greater degree of pelvic floor descent during the ASLR, supporting our belief that prior GS would compromise the so-called core stability MCS of the lumbopelvic region, requiring greater large-muscle activation for low-load stability, leading to excessive abdominal pressure generation relative to pelvic floor muscle, ligament, and fascia support.

Materials and Methods

Subjects

Subjects were recruited to participate in a single data-gathering session through university medical center flyers, online journals circulated to geriatric populations, retirement facility lectures, and word of mouth. The Institutional Review Board of the university’s Office of Human Subjects Protection approved the protocol. Twenty-five women met the inclusion and exclusion criteria and participated in the study. Women aged 65 years or older were recruited and classified into one of the following medical and surgical history categories:

1. no history of LBP or GS,

2. history of LBP but no history of GS,

3. history of GS but no history of LBP, or

4. history of GS and history of LBP.

Exclusion criteria included any history of lumbar spine surgery to avoid surgically altered lumbopelvic biomechanics, neurological diseases that might affect underlying motor control function for spine stability, and spine compression fractures as a cause of LBP.

LBP was quantified using both the 100-mm visual analog scale (VAS) and the 11-point numerical rating scale. The impact of LBP on daily function was assessed using the Modified Oswestry Disability Questionnaire.⁵⁸ Clinical research center nursing staff recorded height and weight on the visit day to obtain a body mass index (BMI). The research team was blinded to prior history. Subjects were asked to not discuss their prior history or pain locations during the measurement protocol.

Procedures

SIJ compliance measure

The DIV measurement was performed by applying a sinusoidal vibration force to the prone subject’s anterior superior iliac spine (ASIS; Figure 1), as previously described.^40,46,53 The bed height was adjusted to ensure contact without excessive preload of the shaker motor floating platform. The subjects' feet were supported with pillows to ensure a relaxed prone position. The subjects were instructed to relax the muscles of the lower back, buttocks, abdomen, and pelvic floor to minimize the influence of muscle tone on SIJ compliance. The shaker motor (model LW126-13, Labworks, Inc, Costa Mesa, California) was programmed to provide a 200-Hz sinusoidal vibration with amplitude of approximately 0.05 mm. The motor’s driver amplifier was adjusted until the subject felt the vibration and the operator could palpate vibration in the posterior pelvis tissues. The 200-Hz frequency has been previously demonstrated to be safe.⁴³ The vibration signal of the posterior SIJ bones was quantified using real-time color Doppler imaging (HDI 5000 1997, software version 4252-0913-15-190-17; ATL Ultrasound, Inc, Bothell, Washington) of the ilium and sacrum. A linear probe (ATL Linear Array L7-4 38 mm) was positioned over the SIJ using the posterior superior iliac spine (PSIS) as the palpation landmark giving an image of the PSIS and sacrum (Figure 3 ). The scanner was configured in the peripheral vascular tissue specific preset using the venous category based on preliminary testing to optimize the DIV images. The image depth was adjusted to display both the PSIS and sacrum in the image. The system was converted to the power-imaging mode (PWR IMG), and the power region of interest column size was maximized and column position adjusted to include the entire image. The color gain was adjusted to show power intensity on both the PSIS and sacrum that was not present when the vibration was turned off (CPA range, 70%-90%). Once a stable power image was obtained with vibration applied, the image was saved using the research quality output module on the HDI 5000 system to an optical disc. Thus, the power Doppler signal data for the sacrum and the ilium were recorded simultaneously to compare the vibration intensity across the SIJ at one point in time. Three independent sequential measurements of each SIJ were performed with transducer repositioning each time. Left-sided measurements were performed first, and the subject turned around on the bed to perform right-sided measurements. One investigator (J.E.) performed all the measurements after training in the DIV method with previous authors⁴¹ and assembling the system and training for 3 months.

Figure 3.

Doppler image of vibration of a left sacroiliac joint (SIJ) showing posterior superior iliac spine and sacrum color proportional to vibration intensity. The ultrasound probe is positioned to overlap the sacrum medially and ilium laterally as shown in Figure 1; the dorsal SIJ ligament is seen superficially.

Pelvic floor measurements

Subjects were positioned supine and taught the ASLR by asking them to hold one foot 12 to 24 inches above the bed mattress and to demonstrate proper performance before testing. The ultrasound system was converted to the phased array probe (ATL P4-2 20 mm) and the abdominal imaging tissue-specific preset using the general category. The probe was positioned over the superior aspect of the pubic symphysis with the beam aimed caudally to capture the sharpest delineation of the inferior bladder floor and the image stored. The split-screen mode was used with the baseline bladder floor image saved on 1 screen. The subject was then asked to raise 1 leg, and the operator maintained the same inferior bladder floor image. The image was stored, and on-screen distance measurement was performed to determine the distance from the probe to the pelvic floor at the inferior bladder wall (Figure 4 ). The absolute values at rest and during the ASLR were recorded with 3 separate measures from each side obtained in serial fashion beginning with the right leg for all subjects. One investigator (J.E.) performed all measurements.

Figure 4.

Typical pelvis ultrasound demonstrating the change in pelvic floor position during the active straight leg raise.

Data Analysis

The optical disc images were transferred to a Windows operating system (Microsoft Inc, Bellevue, Washington) computer for analysis. For each DIV measure, the image was analyzed using the region of interest plug in for the QLAB software package (QLAB version 2.0, copyright 2003; Philips Corporation, Andover, Massachusetts). The images were accessed using a color suppression setting, enabling the operator to place the region of interest target (0.5 cm²) over the ultrasound image without motion color biasing placement of the target. The operator placed the target over the best bone image of the sacrum and ilium. The analysis was then done providing the average power signal (dB) of the region of interest based on the color intensity embedded in the image. A direct comparison between the vibratory motion of the sacrum and ilium was possible for the same moment in time, providing a differential measurement proportional to compliance of the SIJ. A highly compliant SIJ was predicted to have a greater difference in the power Doppler signal intensity between the ilium and sacrum. A stiffer or less compliant SIJ would demonstrate less difference or more similar power Doppler signals on both sides of the joint. The mean difference between the ilium and sacrum region of interest average power signal was calculated for the 3 measurements on each side. The side-to-side (left to right) difference was calculated by subtracting the mean right ilium to sacrum intensity difference from the mean left ilium to sacrum intensity difference.

Pelvic floor motion during the ASLR was calculated by subtracting the inferior bladder wall depth measurement (probe surface to bladder floor) during the ASLR from the resting measurement. Positive values indicate descent and negative values ascent of the pelvic floor during the load test of the ASLR. The 3 measurements were averaged for each side ASLR maneuver to provide a mean pelvic floor measurement for both the left and right ASLR.

Statistical Analysis

The baseline characteristics (age, BMI, pain scores, and the modified Oswestry Disability Questionnaire scores) were compared among the 4 groups using 2-way analysis of variance (ANOVA) models with factors for LBP and GS. Two-way ANOVA models were further used to test whether the 4 groups had significantly different compliance on either side, compliance asymmetry (right-left), or pelvic floor motion on either side. The consistency of the DIV measurement was evaluated using Cronbach α for the right and left ilium and sacrum measurements individually across the 3 measurements.

Results

DIV data were not obtained on 2 subjects due to data collection problems. These subjects were not included in the results analyses. Table 1 summarizes the characteristics of the groups relative to age, VAS pain score, modified Oswestry Disability Questionnaire rating, BMI, and GS history. Subjects reporting LBP (n = 7) described fairly modest degrees of LBP (mean VAS = 33.14, SD = 17.32 mm) and limited disability (mean Oswestry score = 14.71, SD = 6.58). There was no significant difference in the mean BMI between the pain-free and LBP groups, F(1, 21) = 3.61, P = .0714.

Table 1.

Baseline Characteristics Across Groups Based on Low-Back Pain (LBP) and Gynecological Surgery (GS) History, No Significant Differences Across Groups (P = .05)

	No GS		+GS
	No LBP (n = 8)	+LBP (n = 4)	No LBP (n = 8)	+LBP (n = 3)
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Age, y	72.4 (6.09)	73.3 (5.38)	73.6 (5.24)	68.7 (5.51)
Body mass index, kg/m2	26.7 (5.05)	27.6 (2.95)	25.6 (4.10)	33.1 (4.78)
Visual analog scale, mm	NA	33.3 (12.12)	NA	33.0 (26.06)
Modified Oswestry Disability Scale	NA	13.5 (6.61)	NA	16.3 (7.57)

The DIV measurement demonstrated excellent internal reliability for the ilium (Cronbach α = .852 [right], .701 [left]) and sacrum (Cronbach α = .898 [right], .810 [left]). Similarly, the ilium to sacrum DIV difference for the SIJ demonstrated reasonable internal reliability (Cronbach α = .757 [right], .705 [left]).

Table 2 shows measurements for each of the subgroups for the mean DIV power image data of the ilium and sacrum, the ilium to sacrum difference for each SIJ compliance measure, and the right to left SIJ compliance difference. There was no significant difference in the SIJ compliance measure right to left difference among the subgroups, F(3, 19) = 0.17, P = .9124. Similarly, there was no significant difference in the mean absolute SIJ compliance measurement for either SIJ across groups: right side, F(3, 19) = 0.25, P = .8613; left side, F(3, 19) = 0.99, P = .4159. The results of the pelvic floor movement during the ASLR measurement are shown in Table 2 for each of the subgroups. There was no significant difference in the amount of pelvic floor motion across groups during either the right- or left-sided ASLR: right side, F(3, 18) = 0.82, P = .4976; left side, F(3, 18) = 0.77, P = .5253). The ANOVA results are shown in Table 3 for both the SIJ compliance measurement and the pelvic floor measurement.

Table 2.

DIV Measurements for Each SIJ Ilium to Sacrum Difference (dB) and Pelvic Floor Motion (cm) During the ASLR by Groups

	No GS				+GS
	No LBP (n = 8)		+LBP (n = 4)		No LBP (n = 8)		+LBP (n = 3)
	Mean (SE)	95% CI	Mean (SE)	95% CI	Mean (SE)	95% CI	Mean (SE)	95% CI
R ilium to sacrum DIV  measure difference	4.27 (1.09)	(2.00 to 6.55)	3.46 (1.54)	(0.24 to 6.68)	4.43 (1.09)	(2.15 to 6.70)	2.87 (1.78)	(–0.85 to 6.59)
L ilium to sacrum DIV  measure difference	2.22 (0.75)	(0.65 to 3.79)	1.58 (1.06)	(–0.64 to 3.80)	3.44 (0.75)	(1.87 to 5.01)	1.60 (1.22)	(–0.96 to 4.16)
Pelvic floor motion  (right ASLR)	0.26 (0.13)	(–0.01 to 0.52)	0.19 (0.18)	(–0.18 to 0.56)	0.49 (0.13)	(0.21 to 0.77)	0.39 (0.21)	(–0.04 to 0.82)
Pelvic floor motion  (left ASLR)	0.32 (0.16)	(–0.02 to 0.66)	0.62 (0.23)	(0.14 to 1.10)	0.65 (0.17)	(0.29 to 1.01)	0.53 (0.26)	(–0.02 to 1.08)

Abbreviations: ASLR, active straight leg raise; CI, confidence interval; DIV, Doppler imaging of vibration; SE, standard error; SIJ, sacroiliac joint.

Table 3.

ANOVA Results for the DIV Measurement of Compliance Within Each SIJ and Left to Right Difference of the SIJ Compliance Measurement and the Pelvic Floor Motion During the ASLR by Side^a

	F Statistic	P Value
R ilium to sacrum DIV measure difference	0.25	.8613
L ilium to sacrum DIV measure difference	0.99	.4159
L to R SIJ DIV measure difference	0.17	.9124
Pelvic floor motion (right ASLR)	0.82	.4976
Pelvic floor motion (left ASLR)	0.77	.5253

Abbreviations: ANOVA, analysis of variance; ASLR, active straight leg raise; DIV, Doppler imaging of vibration; SIJ, sacroiliac joint.

^a Numerator and denominator degrees of freedom for all DIV tests were 3 and 19, respectively, and for all pelvic floor motion tests, 3 and 18, respectively.

Discussion

The results of this study indicate that older women with LBP or with a history of GS do not demonstrate greater SIJ compliance asymmetry relative to age-matched women with no LBP or history of GS. We hypothesized that prior GS could trigger an ineffective MCS, leading to increased SIJ stress and strain, precipitating more SIJ compliance asymmetry as reported in women with pregnancy-related LBP.^34,38,40 Women reporting both LBP and a history of GS (n = 3) did not demonstrate more asymmetry of SIJ laxity. Women with either LBP or history of GS did not demonstrate differences in right or left absolute SIJ compliance compared with their null counterparts. Thus, our hypothesis that prior GS or the presence of LBP in older women would be associated with more SIJ compliance asymmetry was not supported.

Differences in pelvic floor movement during the ASLR, interpreted to suggest altered lumbopelvic MCS with changes in the balance between intra-abdominal pressure and pelvic floor support mechanisms, were not observed in women with LBP or GS history. Thus, our hypothesis that altered MCS inherent in LBP or due to prior GS would be demonstrated by greater pelvic floor movement during a load challenge was not supported.

The DIV method employed in this study differed from initial reports in the analysis of SIJ compliance. Prior experiments measured the difference in color Doppler gain between the ilium and sacrum to provide a stiffness or, conversely, laxity measure.^{34,38,40,41,43–47} That method used serial changes in instrument settings to compare the threshold gain when the sacrum color Doppler signal was lost to when the ilium signal was lost. One limitation of that method is the moment-to-moment variation of the bone vibration from changing muscle tone and vibration plate contact from factors such as respiration. The ASIS contact and therefore intensity of the applied vibration force could potentially vary between the 2 measurement points for color Doppler thresholds of the sacrum and then ilium. We used a single-point-in-time measurement of the sacrum and ilium vibration analyzed to determine the relative difference in energy between the 2 vibrating bones. The internal consistency of the DIV measurements of the ilium and sacrum within subjects suggests our measurement on each SIJ structure and thus the difference between the 2 SIJ structures was derived with consistency.

Other factors limit the analysis of the results. The severity of LBP was small in women reporting current LBP in both the GS and no-GS groups (33.0 mm, 33.3 mm mean VAS scores), suggesting our groups lacked sufficient SIJ compliance asymmetry to precipitate LBP as described in peripartum women suffering LBP.^34,38,40 Those studies focused on the more specific pelvic girdle pain diagnosis rather than the nonspecific LBP our subjects described. We did not localize the region of our subjects' LBP to the lumbar spine or posterior pelvis. In addition, the small number in each subgroup limited the power to detect clinically meaningful differences.

Degenerative changes in the SIJ could reduce inherent compliance in older women with increased joint congruity relative to younger women. Thus, the population might have been too advanced in age to detect compliance asymmetry potentially present in younger years. It is possible that altered MCS for stability developed earlier in life leads to higher compressive loads over time, causing accelerated degenerative changes that reduce SIJ compliance. Histological analysis of older human SIJs demonstrated degenerative changes with incomplete ankylosis in a discussion of manual therapies for SIJ conditions.⁵⁹ Radiographic analysis demonstrated that degenerative changes progressed beyond age 40 years and varied in location within the SIJ studied.⁶⁰ The radiographic variation of degenerative changes with observations of marked variation in the viscoelastic properties of different SIJ regions suggests that the choice of the PSIS landmark to perform the DIV analysis might have missed more mobile sections of the SIJ and did not reflect overall SIJ compliance.⁶¹ Further refinement of the DIV method comparing measurements at various levels across the SIJ could enhance the understanding of this measurement in older subjects.

The pelvic floor motion assessment used an indirect measure of movement using the perceived bladder floor motion during the ASLR. Potential error exists in the method due to changes in probe angle and compression during the ASLR, creating a change in the perceived edge for measuring the movement. Altered MCS during the ASLR may not affect bladder floor motion in our cohort if the MCS included excessive pelvic floor muscle tone, best observed with electrophysiological methods, or increased stiffness of pelvic floor ligaments. As noted, the subjects with LBP had very little pain and disability and may not represent women with significant alterations in their MCS during the ASLR.

Conclusions

Older women reporting a history of GS and/or LBP did not demonstrate differences in absolute SIJ compliance or greater asymmetry of SIJ compliance (stiffness) compared with older women without LBP or a history of GS. Similarly, the cohort did not demonstrate a significant difference in pelvic floor motion during the ASLR challenge, suggesting they did not use excessive muscle activation to provide lumbopelvic stability during the load challenge that exceeded the pelvic floor active and passive support systems. We present a variation in the method for assessing SIJ compliance using DIV that compares induced vibration signals on both sides of the SIJ simultaneously that demonstrated reasonable internal consistency.