Does Gynecologic Surgery Contribute to Low Back Problems in Later Life? An Analysis of the Women’s Health and Aging Study

Methods 

The WHAS is a longitudinal, population-based study of disabled women aged 65 or older living in the eastern region of the city and county of Baltimore, MD. Participants were eligible for the study if they reported difficulty with daily tasks from 2 or more functional domains (basic self-care, upper-extremity ability, mobility, and higher-functioning tasks). Severe cognitive impairment, defined by scores below 18 on the Mini-Mental Status Examination,25 was an exclusion criterion. In-home screening was completed on 4137 women; 1409 were eligible, and 1002 agreed to participate in the study.24

Data for this analysis were derived from the comprehensive baseline assessment including an interview about health and sociodemographic information, nurse’s physical examination, physician’s questionnaire, and physical performance testing.24 Diagnoses of disk disease and spinal stenosis were derived from disease ascertainment algorithms by using information from participant interviews, physician’s surveys, nurse’s examinations, medical records, and radiographs.24 These algorithms showed acceptable reliability.24 Disk disease and spinal stenosis in the thoracic or lumbar spine only were included in the present study to define the outcome measures of disk pathology and spinal stenosis. The presence and severity of LBP were assessed by 2 questions: “During the past year, have you had pain in your lower back on most days for at least one month?” and “Please rate the average pain in your back during the past month.” Participants were shown a card with an 11-point numeric pain rating scale, with 0 denoting no pain and 10 indicating excruciating pain (as bad as you can imagine). Participants were classified as having moderate or severe LBP if they answered yes to the first question and rated the severity of recent or current pain in the moderate or severe categories. LBP severity was classified into categories of mild pain (no LBP reported or LBP rating 0−3), moderate pain (4−6), and severe pain (7−10). Of the 1002 women, 597 reported no LBP and were combined with the 49 women reporting only mild pain to form the mild reference category compared with 162 women reporting moderate pain and 193 reporting severe LBP.

Surgical menopause was defined as present if a participant answered “surgery” when asked, “What was the reason your menstrual periods stopped?” Ovarian surgery was defined as a positive response to the question, “Did you ever have surgery on your ovaries?” The questionnaire did not query the participants for the type of hysterectomy performed. Body mass index (BMI) was calculated by using weight (in kilograms) divided by height (in square meters), both measured during the initial home physical assessments. Diagnoses of osteoporosis and hip and knee arthritis were determined from adjudication algorithms as described for the spine pathology variables.

Results 

The weighted prevalence of participant characteristics is shown in table 1. Of the 1002 women in this study, 42% reported LBP lasting 1 month or more in the prior year. Remarkably few women reported mild LBP (4.5%) compared with respondents reporting moderate (16.2%) and severe (20.7%) LBP in the prior month. Gynecologic surgery was common; nearly half (45.9%) reported a surgical menopause or ovarian surgery. Confirmed lumbar stenosis and disk disease were present in 2.8% and 7.8%, respectively.

Table 1. Weighted Prevalence of 1002 WHAS Participant Characteristics, 1992–1995

	Characteristic	N	Weighted Percentage
	Age 65–74y	388	44.2
	Age 75–84y	311	40.7
	Age ≥85y	303	15.1
	White	713	71.1
	Obesity (BMI ≥30kg/m2)	309	33.4
	OA hip	78	8.2
	OA knee	366	37.9
	Osteoporosis	155	15.9
	Back pain in last year	405	42.3
	No back pain in last year	597	57.7
	Mild LBP in last month⁎	49	4.5
	Moderate LBP in last month†	162	16.2
	Severe LBP in last month‡	193	20.7
	Surgical menopause	350	36.4
	Ovarian surgery	320	33.2
	Surgical menopause or ovarian surgery	448	45.9
	Lumbar stenosis§	28	2.8
	Disk pathology∥	78	7.8

⁎ LBP intensity 1–3 on 0–10 scale reported for last month. † LBP intensity 4–7 on 0–10 scale reported for last month. ‡ LBP intensity 8–10 on 0–10 scale reported for last month. § Physician-adjudicated data. ∥ Thoracic or lumbar levels.

Logistic regression analyses results estimating the association of LBP reported in the prior year and LBP severity reported in the last month with gynecologic surgery are shown in table 2. The strength of the association diminished after multivariate adjustment; however, women who reported a surgical menopause were nearly 60% more likely to have moderate LBP compared with women who did not have a history of gynecologic surgery (adjusted OR [AOR]=1.59; 95% CI, 1.1−2.3). Ovarian surgery was associated with moderate pain in the past month (AOR=1.39; 95% CI, 0.96−2.00) after multivariate adjustment, but this association was not statistically significant. A history of either form of gynecologic surgery was not significantly associated with severe LBP. Overall, the history of either surgical menopause or ovarian surgery (ie, any gynecologic surgery) was significantly associated with moderate LBP (AOR=1.44; 95% CI, 1.01–2.07).

Table 2. Unadjusted and Adjusted ORs for Any LBP in the Past Year⁎ and Severity of LBP† in the Past Month Related to History of Gynecologic Surgery

		Surgical Menopause		Ovarian Surgery		Either Surgery
	LBP Conditions	Unadjusted	Adjusted	Unadjusted	Adjusted	Unadjusted	Adjusted
	LBP last year	1.43(1.10–1.86)	1.20(0.90–1.59)	1.35(1.03–1.77)	1.21(0.91–1.61)	1.37(1.07–1.77)	1.22(0.93–1.59)
	Moderate LBP in last month	1.75(1.23–2.48)	1.59(1.10–2.30)	1.50(1.04–2.15)	1.39(0.96–2.01)	1.56(1.10–2.20)	1.44(1.01–2.07)
	Severe LBP in last month	1.40(1.00–1.95)	1.09(0.76–1.57)	1.34(0.95–1.88)	1.18(0.82–1.69)	1.37(0.99–1.89)	1.16(0.82–1.63)

NOTE. Values are ORs with 95% CIs. Adjustments for age, race, BMI, osteoporosis, and knee and hip OA.

⁎ From logistic regression. † From polychotomous regression.

Neither surgical menopause nor ovarian surgery was significantly associated with spine pathology manifested as lumbar stenosis or disk pathology in the lumbar and thoracic region (fig 1). In an additional analysis, obstetric history, derived from questions about living children, was not significantly associated with the outcomes of LBP in the last year, moderate or severe LBP in the past month, and spine pathology (data not shown).

View Large Image Download to PowerPoint

Fig 1. The likelihood (AOR with 95% CI) of having spine pathology (stenosis or disk disease) related to gynecologic surgery, either surgical menopause (SM) or ovarian surgery (OS). All models adjusted for age, race, BMI, osteoporosis, and knee and hip OA.

Discussion 

The results of this study suggest that gynecologic surgery resulting in menopause is associated with self-reported moderate severity LBP in late life among women. The relation was independent of other factors, available in the database, previously linked to LBP such as hip and knee arthritis, osteoporosis, and obesity.26, 27, 28 Prior gynecologic surgery was not associated with an increased risk for severe LBP in the last month or thoracolumbar spine pathology such as disk disease or stenosis.

Biomechanic models have proposed that stability from neuromuscular control mechanisms is important for protecting the lower spine from acute and chronic injury.9, 20, 30, 31 Within the biomechanic and rehabilitation literature, pelvic floor and abdominal muscle dysfunctions have been theorized as important precipitating factors leading to LBP.16, 21, 22, 32 Our large population-based study suggests a link between prior gynecologic surgery and subsequent LBP. Altered abdominal or pelvic floor muscle function after surgery could precipitate LBP. For the purposes of this investigation, we chose to evaluate gynecologic surgery that would not typically involve severing of muscle or nerve structures but rather stretching or ischemic injury during surgical retraction or compression. Such injuries may precipitate biomechanic dysfunction with reduced core stabilizing force production, as well as secondary muscle deconditioning from pain inhibition during the perioperative period. Spontaneous recovery of normal biomechanic muscle function after surgery may not be automatic. The observation that deep spine muscle atrophy was present long after symptom resolution in acute LBP patients, together with muscle fiber–type changes years after lumbar surgery, support the idea that spontaneous muscle recovery after surgical and nonsurgical muscle injury is not automatic.33, 34, 35 We are not aware of similar studies assessing recovery of biomechanic or neurophysiologic function after abdominal or gynecologic surgery.

Regardless of the potential mechanisms, our findings suggest that injury to, or deconditioning of, these muscle groups may occur during gynecologic surgery and predispose these women to core muscle dysfunction and LBP later in life. Other reports of LBP and muscle control loss after pelvic surgery and childbirth support this hypothesis. Biomechanic effects manifested initially as incontinence and weakened pelvic floor muscle contractions were reported in 25% of new mothers.36 An Australian survey37 of health problems after childbirth described persistent early back pain in half of the total (n=1193) respondents. Primiparous women reported more back pain in the first 8 weeks postpartum than multiparous women. However, multiparous women described continued LBP and urinary problems, whereas primiparous women described declining LBP and urinary dysfunction over time.37 Middle-aged women with urinary incontinence after Burch colposuspension surgery, involving abdominal incision to access the bladder and urethra, reported a higher incidence of LBP than women with incontinence who did not report a surgical history (66% vs 38%).38 A retrospective study of 1760 middle-aged women (ages 38−64y) reported an association between LBP and hysterectomy in the older group (50−64y) and an association between LBP and abortion, dilatation and curettage, and childbirth.39 Higher numbers of live births in the older group and abortions in the younger group were associated with LBP after controlling for other risk factors.

Rehabilitation focus on abdominal and pelvic muscles, potentially injured in gynecologic surgery, has shown some efficacy in some LBP populations. Specific activation of transversus abdominis in the low abdomen and the pelvic floor muscles showed long-term benefits in both pain intensity and the Oswestry disability instrument at 30-month follow-up compared with traditional physiotherapy (PT) approaches in spondylitic LBP subjects.40 Reduced pelvic pain was shown in postpartum women after pelvic girdle stabilization exercises focusing on the deep local muscles compared with controls given PT with strengthening exercises not focusing on the specific “core” muscles.41 Although not linking specific muscle injury etiologies to LBP conditions, the focus of these studies on the deep musculature rather than conventional PT interventions further confirms that specific muscle groups, at risk for altered function after gynecologic surgery, play an important role in LBP.

These outcome study observations together with our findings of the association between moderate severity LBP and surgical menopause support the belief that gynecologic and obstetric events might compromise protective abdominal and pelvic musculature with both immediate and long-term consequences. The treatment outcomes described previously, focusing on retraining of muscles deep in the abdomen and pelvis, support the concept that muscle function is important to understand both the etiology and treatment of LBP.

It is important to note that prior gynecologic surgery was not associated with severe LBP in our analyses. Generally, it is acknowledged that multiple factors, including psychosocial influences, may contribute to the presence of LBP.42 Severe LBP may reflect the presence of overwhelming ongoing pathology or more recent injury; thus, a remote injury to abdominal or pelvic musculature may play a reduced role. Additionally, the WHAS cohort represents a population of disabled women with multiple comorbidities. Thus, many factors that contribute to disability (a common manifestation of severe LBP) may reduce any risk that prior surgery could impart. A study involving a healthier cohort of older women would allow a clearer understanding of the relation between prior surgery and subsequent LBP.

The hypothesis that gynecologic surgery would increase the likelihood of the specific spine pathologies spinal stenosis and disk disease was not supported by the results of this analysis. Several limitations to the dataset could account for the lack of association. We may not have the power to detect the strength of the association using this relatively small dataset. We do not have information regarding the surgical type and time since surgery, and the diagnoses of stenosis and disk disease, done by record review and not full population radiographic screening, may not reflect the true incidence of these pathologies in older women.

Other potential limitations of our investigation exist. The dataset did not allow appropriate examination of the relation between obstetric events and the outcome variables of LBP and spine pathology. True parity of the participants was not available. The influence of the number of live births and birth methods as potential risk factors for LBP or spine pathology warrant further investigation. Surgical descriptors for surgical menopause and ovarian surgery were not recorded in the WHAS database, and the temporal relation between surgery and the onset of spine conditions has not been investigated. The potentially cumulative role of abdominal surgery such as cholecystectomy and other pelvic surgeries (eg, bladder surgery) was not examined; the physical examination data did not include surgical scar observations. We do not know whether abdominal surgeries in men would predispose them to back pain or spine pathology later in life.

Conclusions 

Despite the limitations previously mentioned, our findings suggest that a potential injury to the pelvic and lower abdominal musculature, surgical menopause in this case, puts women at increased risk for LBP later in life. Women reporting a surgical menopause described moderate LBP more often than women with no history of gynecologic surgery. Biomechanic and neurophysiologic research is posing mechanisms for how the deep pelvic and abdominal muscles provide lumbopelvic stability. There is growing rehabilitation focus for activating and strengthening these muscles not only for incontinence but for LBP treatment as well.22, 41, 43, 44, 45, 46 Rehabilitation of muscles potentially injured with gynecologic surgery is reasonable in hopes of avoiding the deleterious effects that may predispose women to LBP as they age. Our observation of the increased likelihood of having LBP years after gynecologic surgery suggests that preventive rehabilitation efforts after surgery should focus on restoring normal neurophysiologic and biomechanic function to the muscles of the lower abdomen and pelvis. Further investigation of the role of obstetric and gynecologic events on LBP and spine pathology is warranted to determine whether these “routine” life events may lead to hazardous biomechanic and neuromuscular changes in important protective muscle systems.

Supplier