Diabetes and Obesity-Related Risks for Pelvic Reconstructive Surgery in a Cohort of Swedish Twins

This nationwide twin cohort study was performed through cross-linkage of the Swedish Twin Register and the Swedish Inpatient Register. The Swedish Twin Register contains data on nearly all twins born in Sweden since 1886 (10). The present study included the female Swedish Twin Register cohort born from 1926 through 1958, for which we obtained information on age, diabetes type (i.e., type 1 versus type 2), BMI, and childbirths.

The cohort was subsequently linked to the Swedish Inpatient Register to obtain detailed information on pelvic floor surgery based on the Swedish Classification of Operations and Major Procedures. Subjects contributed to the study from 1973 until first occurrence of stress urinary incontinence and/or pelvic organ prolapse surgery, death, or the end of the observation period (31 December 2005). The study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational studies (11). The study was approved by the regional research ethics board at Karolinska Institutet.

Using the Swedish Classification of Operations and Major Procedures, stress urinary incontinence surgery was classified as any one of the following: Kelly/Kennedy plication, Burch colposuspensions, Stamey procedures, Marshal-Marchetti-Krantz procedures, Ingelman-Sundberg plasty, intravaginal slingplasty, bladder-neck and suburethral slings, and tension-free vaginal tapes (operation codes 6355, 6356, 6358, 7470, and 7471 for 1973–1996 and LEG00, LEG10, LEG20, and KDG10-40 for 1997–2005). Pelvic organ prolapse surgery was categorized as any one of the following: anterior or posterior repair, Manchester procedures, abdominal sacrocolpopexy, sacrospinous fixations, and enterocele obliteration and colpocleisis (operation codes 7120, 7121, 74607-464, 7466, 7469, and 7541 for 1973–1996 and LEF00, LEF03, LEF10-50, LEF53, and LEF96 for 1997–2005).

Logistic regression was used based on generalized estimating equations, which take into account the correlated (twin) structure of the data. Provided that there is a significant association between the exposure and outcome, using a twin-based study design allows further in-depth assessments of a genetic interaction. Multivariable analyses were adjusted for age, BMI, and childbirth (ever/never). Odds ratios (ORs) were estimated with 95% CIs. All statistical analyses were performed using SAS software (Cary, NC).

A total of 29,881 women were included in the study cohort, including 8,443 same-sex female twin pairs with known zygosity. Mean ± SD parity was 2.3 ± 0.97, BMI 21.2 ± 2.9 kg/m², and age 64.1 ± 9.2 years. The accumulated prevalence of pelvic floor surgery in the cohort was 5.6%: stress urinary incontinence surgery was performed in 555 women (1.9%), and pelvic organ prolapse surgery was performed in 1,099 women (3.7%).

We identified 3,376 complete monozygotic and 5,067 dizygotic same-sex female twin pairs from the cohort for correlated generalized estimating equations analysis (Table 1). For type 1 and 2 diabetes, no significant association was observed for stress urinary incontinence (OR 1.0 [95% CI 0.1–9.2] and 2.0 [1.0–4.0]). There were no cases of pelvic organ prolapse surgery in women with type 1 diabetes, and for type 2 diabetes, the risk estimate was nonsignificant (1.6 [1.0–2.7]). We therefore combined type 1 and 2 diabetes into a single variable for the regression analysis.

In the univariate (unadjusted) analysis, diabetes was associated with an increased risk of stress urinary incontinence (OR 1.9 [95% CI 1.1–3.4]). However, after adjustment, the association diminished to a nonsignificant level (1.6 [0.9–2.9]). For stress urinary incontinence, BMI >25 kg/m², age ≥60 years, and childbirth were independently associated with an increased risk for having incontinence surgery. Diabetes and BMI were not independent risk factors for pelvic organ prolapse surgery.

Women undergoing pelvic floor surgery are the ones most likely to have experienced severe symptoms (12), and a presumed causal association would be evident. Nonetheless, we found no independent association between type 1 or type 2 diabetes and surgically managed stress urinary incontinence. When considering both diabetes types as a single exposure, any diabetes was associated with an increased risk for stress urinary incontinence in a univariate setting, but the statistical significance of the association was lost when adjusting for established confounders. Similar to the results for stress urinary incontinence, we found no significant association between diabetes and pelvic organ prolapse surgery after adjusting for potential confounders. Thus, the often-promoted association between diabetes and pelvic floor disorders is confounded by environmental factors.

Genetic effects contribute to the occurrence of both stress urinary incontinence and pelvic organ prolapse (9). There were no consistent indications of a common genetic basis for diabetes and stress urinary incontinence or pelvic organ prolapse. Further exhaustive analyses on mono- or dizygotic twin similarity would therefore be futile.

Being overweight, but not obese, was a risk factor for stress urinary incontinence surgery. This paradoxical result is probably explained by a negative selection for surgery due to apprehension for increased complication rates and poor outcomes in obese subjects. Consistent evidence from observational studies suggests that obese women experience more severe stress urinary incontinence than women of normal weight (13,14). Thus, despite having more severe symptoms and comparable success rates after surgical treatment (15), obese women are less likely to undergo surgical treatment than women of normal weight.

The study was supported by a grant from the Nordic Urogynecological Association and a grant from The Regional Agreement on Medical Training and Clinical Research (ALF) between Stockholm County Council and Karolinska Institutet.