SEARCH

SEARCH BY CITATION

Keywords:

  • Anal sphincter tears;
  • obesity;
  • pregnancy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

Objective

To assess the risk for obstetric anal sphincter lacerations in relation to maternal obesity among primiparous women in Sweden.

Design

A population-based study.

Setting

Sweden.

Population

All women with vaginal delivery and singleton pregnancy in Sweden in the years 2003–2008 (n = 210 678).

Methods

The Medical Birth Registry, the National Board of Health and Welfare, was used to identify cases of rupture and body mass index (BMI) classes. The population was categorised into four classes with BMI of <25, 25 to <30, 30 to <35 and >35 kg/m2.

Main outcome measures

Odds ratios were estimated with 95% confidence intervals. In order to estimate the effect of BMI on obstetric anal sphincter lacerations, with possible confounders accounted for, uni- and multivariate logistic regressions were performed.

Results

In total, 8958 (4.25%) cases of anal sphincter lacerations (grade III–IV) occurred; increasing BMI showed a significant near-dose–response type of protective effect against grade III–IV lacerations when compared with women with BMI <25 kg/m2: BMI 25 to <30 kg/m2, 0.89; BMI 30 to <35 kg/m2, 0.84; BMI > 35 kg/m2, 0.70.

Conclusion

Overweight and obesity were associated with a decreased risk for obstetric anal sphincter lacerations.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

The incidence of obesity is increasing worldwide and, at present, about 12.6% of Swedish pregnant women are obese, compared with 20.2% of women of childbearing age in the UK[1] and 26.5% of US women of 20–39 years of age.[2] Recent findings lend support to a causal relation between overweight or obesity and the risk of adverse pregnancy outcomes.[3, 4] Pregnancy complications among obese women include an increased risk for macrosomia,[5, 6] gestational diabetes,[5, 7] pre-eclampsia,[8] gestational hypertension,[9] shoulder dystocia,[4] stillbirth[10] and malformations.[11, 12]

With regard to obstetric events, studies also suggest that rates of caesarean section and operative delivery are higher among obese women relative to normal weight women.[13] Whether or not maternal obesity at birth also increases the risk for intrapartum perineal and anal sphincter lacerations is, however, poorly investigated.

The risk for anal incontinence is increased among obese women[14] and, given that obstetric anal sphincter lacerations are the predominant risk factors for anal incontinence among premenopausal women, it is possible that obese women are more susceptible to intrapartum anal sphincter lacerations. Additional risk factors among obese women, such as physical inactivity, may further increase the risk of perineal lacerations among obese mothers.[15] In a study from Wales, the investigators found an overall increased risk for intrapartum complications, but not specifically third- and fourth-degree perineal lacerations, among women with a body mass index (BMI) over 30 kg/m2 living within a hospital catchment area.[4] There is, however, scarce population-based evidence to support clinical and patient guidelines on this subject. The aim of this nationwide register-based study was to assess the risk for obstetric anal sphincter lacerations in relation to maternal obesity among primiparous women in Sweden.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

Data sources

This was a nationwide population-based study designed to include all primiparous women giving vaginal birth between 1 January 2003 and December 2008, recorded in the Swedish Medical Birth Register. The register has recorded data on more than 98% of all deliveries in Sweden since 1973, and has been validated and described in detail previously.[16] Starting with the first antenatal visit, normally at 8–12 weeks of gestation, information on maternal demographic data, reproductive history, pregnancy and birth characteristics is collected prospectively by a midwife. The record also includes maternity weight and height from which the BMI (kg/m2) is calculated. The standardised records are identical throughout the country and are forwarded to the central birth register, where the information is computerised. Only primiparous women were included in the present study and women with multiple gestations and emergency or elective caesarean sections were excluded.

Outcome definition and ascertainment

Body mass index (BMI) was categorised into four classes according to the World Health Organization guidelines: BMI <25 kg/m2; BMI 25 to <30 kg/m2; BMI 30 to <35 kg/m2; BMI > 35 kg/m2.[17] In accordance with the national guidelines, a diagnosis of obstetric lacerations was performed by the responsible midwife, who inspects the vagina immediately after childbirth. The diagnosis and classification of perineal lacerations follow the International Classification of Diseases (ICD) 10th revision: grade I, labial tears (ICD O70.0); grade II, perineal and vaginal laceration, but no anal sphincter involvement (ICD O70.1); grade III, partial or complete anal sphincter laceration (ICD O70.2); grade IV, complete anal sphincter laceration including rectal mucosa (ICD O70.3). In cases in which the midwife suspects a grade III–IV anal sphincter laceration, the attending obstetrician is consulted to confirm or refute the diagnosis. Information on age at delivery, operative delivery, birthweight, head circumference and maternal diabetes mellitus was also derived from the register.

Statistical analysis

Proportions for grade I–IV lacerations are presented as absolute numbers and proportions with 95% confidence intervals (CIs) according to the BMI class. In order to estimate the effect of BMI on obstetric anal sphincter lacerations, with possible confounders accounted for, uni- and multivariate logistic regressions were performed. The univariate analysis used all the variables identified a priori as possible risk factors for obstetric anal sphincter lacerations, including BMI class, maternal age, instrumental delivery, birthweight, head circumference and diabetes mellitus. Variables showing a significant relation in the univariate analysis were included in the final multivariate model. Odds ratios (ORs) were estimated with 95% CIs. All statistical analyses were performed using SAS software (Cary, NC, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

A total of 210 678 primiparous deliveries were included in the database. Almost 8% of the study population had a BMI > 30 kg/m2 and were considered as obese. The distribution according to BMI class is presented in Table 1. The mean age of the study population was 27 years with only minor differences between the BMI classes.

Table 1. Distribution of individuals over the body mass index (BMI) groups
BMI classFrequency (n)Per centMean age (SD) (years)
  1. SD, standard deviation.

<25148 82670.6427.1 (4.8)
25 to <3045 05521.3927.4 (4.9)
30 to <3512 3525.8627.2 (5.0)
>3544452.1127.3 (5.0)

The prevalence of various obstetric lacerations in relation to BMI category is presented in Table 2. In total, 8958 (4.25%) cases of anal sphincter lacerations (grade III–IV) occurred. The proportions of women with an anal sphincter laceration were 4.25% among those with BMI <25 kg/m2, 4.38% among those with BMI 25 to <30 kg/m2, 4.03% among those with BMI 30 to <35 kg/m2 and 3.71% among those with BMI > 35 kg/m2. Sphincter lacerations which also involved the rectal mucosa were very few (0.17%). With regard to grade I and II lacerations, including lacerations of the clitoris, labia and vagina, there was an almost linear increase in the occurrence of injuries related to an increase in the BMI. In contrast, the occurrence of anal sphincter lacerations showed an overall inverse correlation with increasing BMI.

Table 2. Delivery damage by body mass index (BMI). Percentage of obstetric ruptures in different BMI groups
Delivery injuryPercentage (95% CI)
<25 kg/m2 (n = 148 826)25 to <30 kg/m2 (n = 45 055)30 to <35 kg/m2 (n = 12 352)>35 kg/m2 (n = 4445) Total ( n  = 210 678)
  1. CI, confidence interval.

Ruptures, clitoris (n = 65 554)31.7 (31.5–32.0)30.1 (29.7–30.6)28.7 (27.9–29.5)27.6 (26.3–28.9)31.1 (30.9–31.3)
Ruptures, vagina (n = 116 133)53.5 (53.2–53.7)58.3 (57.8–58.8)60.5 (59.7–61.4)63.9 (62.5–65.3)55.1 (54.9–55.3)
Ruptures, perineum (n = 79 659)37.0 (36.8–37.3)40.0 (39.5–40.4)39.5 (38.7–40.4)38.0 (36.5–39.4)37.8 (37.6–38.0)
Ruptures, sphincter (n = 8884)4.21 (4.11–4.31)4.35 (4.17–4.55)4.00 (3.66–4.36)3.64 (3.12–4.25)4.22 (4.13–4.30)
Ruptures, rectum (n = 354)0.16 (0.14–0.18)0.19 (0.16– 0.24)0.17 (0.11–0.26)0.18 (0.08–0.37)0.17 (0.15– 0.19)
Ruptures, cervix (n = 437)0.21 (0.19–0.23)0.22 (0.18–0.26)0.19 (0.12–0.28)0.18 (0.08–0.37)0.21 (0.19–0.23)
Grade I & II ruptures (n = 144 513)67.5 (67.3–67.7)70.8 (70.3–71.2)72.0 (71.2–72.8)73.5 (72.2–74.8)68.6 (68.4–68.8)
Grade III & IV ruptures (n = 8958)4.25 (4.14–4.35)4.38 (4.20–4.58)4.03 (3.70–4.40)3.71 (3.18–4.32)4.25 (4.17–4.34)

The univariate logistic regression analyses identified increasing BMI, age, instrumental delivery, birthweight and head circumference as risk factors for both grade I–II and grade III–IV perineal lacerations. Maternal diabetes was not found to have a significant effect on the risk for lacerations, regardless of severity, and was therefore not included in the final multivariate model. When the independent variables (age, instrumental delivery, fetal weight and head circumference) were included in a multivariate setting (Table 3), increasing BMI showed a near-dose–response type of protective effect against grade III–IV perineal lacerations (i.e. anal sphincter injury). At the same time, increasing BMI had a dose–response type of increase in risk for grade I–II lacerations. The greatest risk for anal sphincter laceration was observed among women having an instrumental delivery (OR, 2.92; 95% CI, 2.78–3.06) which, at the same time, decreased the risk for grade I–II lacerations (OR, 0.77; 95% CI, 0.75–0.79). In our study, we found that 15.3% of all women had an instrumental delivery, the occurrence of which was near-identical throughout the BMI classes:BMI <25 kg/m2, 15.1%; BMI 25 to <30 kg/m2, 16.4%; BMI 30 to <35 kg/m2, 14.3%; BMI > 35 kg/m2, 15.2%. The significant effect of head circumference which, in the univariate analysis, was associated with both grade I–II (OR, 1.08; 95% CI, 1.07–1.09) and grade III–IV (OR, 1.20; 95% CI, 1.19–1.22) perineal lacerations, attenuated to a nonsignificant level in the multivariate analysis (Table 3). Increasing birthweight increased the risk for anal sphincter lacerations, and a dichotomised birthweight of 4.5 kg or more (n = 4115) was associated with a more than doubled risk for grade III–IV perineal lacerations relative to a birthweight of <4.5 kg (ORadj, 2.40; 95% CI, 2.17–2.66).

Table 3. Univariate and multivariate logistic regression analyses for the association between body mass index (BMI) and perineal lacerations
 UnivariateMultivariate
OR95% CIPOR95% CIP
  1. CI, confidence interval; OR, odds ratio.

  2. Birthweight and head circumference were measured immediately after birth.

Grade I–II obstetric lacerations
BMI (kg/m2)1.031.03–1.03<0.0001   
<251.00  1  
25 to <301.171.14–1.19<0.00011.131.11–1.16<0.0001
30 to <351.241.19–1.29<0.00011.231.11–1.16<0.0001
>351.341.25–1.43<0.00011.321.23–1.41<0.0001
Age (years)1.071.07–1.08<0.00011.071.07–1.08<0.0001
Instrumental delivery (yes/no)0.890.87–0.92<0.00010.770.75–0.79<0.0001
Birthweight (kg)1.261.24–1.28<0.00011.081.05–1.11<0.0001
Head circumference (cm)1.081.07–1.09<0.00011.061.05–1.07<0.0001
Grade III–IV obstetric lacerations
BMI (kg/m2)1.000.99–1.000.68   
<251  1  
25 to <301.030.98–1.090.250.890.85–0.95<0.0001
30 to <350.950.86–1.040.250.840.76–0.920.0004
>350.870.74–1.010.080.700.59–0.82<0.0001
Age (years)1.061.05–1.06<0.00011.051.05–1.06<0.0001
Instrumental delivery (yes/no)3.421.05–1.06<0.00012.922.78–3.06<0.0001
Birthweight (kg)3.032.90–3.17<0.00012.862.70–3.02<0.0001
Head circumference (cm)1.201.19–1.22<0.00010.990.98–1.010.82

Discussion and conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

Obesity is a risk factor for adverse pregnancy outcomes and is associated with a number of well-known maternal complications during pregnancy.[4, 18] Obesity related maternal perineal complications during labour in specific are, however, poorly investigated. Previous studies have shown conflicting results regarding the influence of obesity on perineal lacerations. Edwards et al.[19] found an increasing risk for recurrent obstetric anal sphincter lacerations in obese women, whereas Buhimschi et al.[20] did not find any correlation, and Landy et al.[21] showed that women with a high BMI had a lower risk of sphincter lacerations at delivery.

In the present nationwide study, we found that maternal overweight and obesity decreased the risk for obstetric anal sphincter lacerations in a near-linear fashion. The inverse correlation between increasing BMI and the occurrence of obstetric anal sphincter lacerations was clear, but the highly significant P values were attributed to the very large sample size rather than large effect sizes. Because obese women are at increased risk for caesarean section as a result of obstetric complications, we restricted the study population to women with vaginal deliveries only. An increased caesarean section rate is likely to lead to increased blood loss,[22] wound infection[23] and thromboembolic events,[24] but decreases the risk for neonatal trauma, such as shoulder dystocia,[5] and perineal tears. Consequently, our estimates do not underestimate the occurrence of obstetric perineal and anal sphincter lacerations.

In our study, we found that the rate of third- and fourth-degree perineal tears decreased with increasing BMI, whereas the opposite was true for first- and second-degree perineal tears, which increased with increasing BMI. This inverse relationship with BMI was somewhat surprising because the study's assumption was that an increasing BMI would increase the risk for obstetric grade III–IV lacerations. We can only speculate as to the reasons for the seemingly protective effect of overweight and obesity on the pelvic floor. Oestrogen and oxytocin receptor efficacy in uterine smooth muscles are modulated by cholesterol which, in turn, is raised in the serum and myometrial membranes of obese women.[25, 26] This may provide a certain degree of protection against oxytocin overstimulation during the second stage of labour and thereby decrease the risk for excessive contractions and subsequent pelvic floor injury in obese women. On a similar note, the uterine smooth muscles have been shown to contract with less force and frequency, and to have less [Ca2+] flux, when comparing obese with normal weight women.[27] Choo et al.[28] reported that abdominal skin in morbidly obese women having bariatric surgery showed greater tensile strength than skin from normal weight patients having cosmetic procedures. Whether or not these results can be generalised to the pelvic floor connective tissues is uncertain, but provides a possible mechanism by which obese women experience less obstetric anal sphincter lacerations. Whether or not the maternal birth position influences the risk for anal sphincter laceration remains controversial. Gottvall et al.[29] showed that women who gave birth in the lithotomy or squatting position had a higher incidence of third- and fourth-degree anal sphincter tears. Overweight and obese women may be prone not to assume a lithotomy or squatting position during pushing, because it may restrict lung function in pregnant women with high BMI.[30]

The strengths of our study include the large study population, the nationwide design, the homogeneous classification of important confounders and outcome measures, and the categorisation of obesity according to international standards. The use of high-quality healthcare registers employing standardised records throughout the country and nonselective registration are major advantages with regard to data ascertainment. At the same time, we recognise that the classification of our primary outcome measure (prevalence of obstetric sphincter lacerations) has some limitations. First, the diagnosis was not verified by ultrasound, but relied on subjective clinical assessments only. However, anorectal ultrasound is also an investigator-dependent method and diagnosis may vary between investigators. Furthermore, ultrasound immediately after delivery can be difficult to interpret correctly because of oedema and stretching of the anal sphincter complex. Routines differ with regard to how midwives examine perineal lacerations after delivery. In some clinics, midwives routinely perform a per rectal examination immediately after birth, whereas others do not implement this as routine. There may also be regional variations with regard to the vigilance and willingness to report the occurrence of anal sphincter lacerations. As a consequence, local, regional and sometimes national under-reporting of anal sphincter injuries at delivery may be a limitation of register-based studies. It should also be noted that the obstetric classification of anal sphincter injuries is based on clinical morphology of the external anal sphincter, and that only grade IV perineal lacerations per definition include the internal anal sphincter. Thus, internal anal sphincter integrity in relation to maternal obesity at delivery has not been addressed in the present study.

Obese women are at increased risk for both anal and urinary incontinence,[14, 31-33] and it is tempting to conclude that the increase in risk can be attributed to an increased risk for pelvic floor trauma at childbirth. However, our data do not support this assumption, but rather indicate the opposite, i.e. that obesity decreases the risk for obstetric pelvic floor disruption. Thus, the reason for the increased prevalence of anal incontinence among obese women can probably be attributed to other factors, such as dietary habits, bowel dysfunction and co-occurring diseases. If the diagnosis of grade III–IV perineal lacerations is considered a proxy for pelvic floor trauma, one may postulate that the increased risk for stress urinary incontinence among obese and overweight women should be attributed to increased intra-abdominal pressure, resulting in urethral sphincter incompetence, rather than pelvic floor injury at delivery.

It is widely assumed that the first childbirth is associated with the greatest risk of pelvic floor trauma and that subsequent deliveries have a lesser impact on obstetric pelvic floor sequelae.[34] Because the present study was limited to primiparous women, we cannot say whether this is the case for overweight/obese women, although it is biologically plausible. In concurrence with most previous studies, we found that instrumental delivery (vacuum extraction) increased significantly the risk for anal sphincter lacerations.[34] The near-threefold increase in risk attributed to instrumental delivery also remained after adjustment in multivariate analysis and showed the strongest association with anal sphincter lacerations. Other factors, such as maternal age, diabetes and head circumference, showed no or limited associations with the occurrence of obstetric anal sphincter lacerations. Our analysis did not include information on fundal pressure during the final stages of labour (not registered) or episiotomy (not validated). Thus, we cannot account for the effects of these obstetric interventions which, to some extent, may have influenced our results. Macrosomia (having a baby of >4.5 kg) and increasing birthweight increased the risk for anal sphincter lacerations by the same magnitude as instrumental delivery, but did not change the overall inverse correlation between BMI and risk for obstetric sphincter lacerations in the adjusted analysis. As such, our data concur with previous studies suggesting that increasing birthweight and macrosomia enhance the risk for anal sphincter lacerations at delivery,[35, 36] although the longer term association with anal incontinence remains under debate.

In summary, this large-scale nationwide study provides robust epidemiological evidence that overweight and obesity decrease the risk for obstetric anal sphincter lacerations among primiparous women. As a consequence, decisions to perform elective caesarean section in obese women should not be based on concerns for increased risks of third- or fourth-degree perineal lacerations.

Contribution to authorship

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

DA was responsible for the study and the study design. ESL conducted the research, wrote the paper and was responsible, together with Dr Altman, for the study design. Both authors were responsible for and approved the manuscript.

Details of ethics approval

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References

The study was approved by the Stockholm Regional Board of Ethics at Karolinska Institute, Stockholm, Sweden, dnr: 2008/1427-31/3.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusion
  7. Acknowledgements
  8. Disclosure of interests
  9. Contribution to authorship
  10. Details of ethics approval
  11. Funding
  12. References