We read with interest the article by Wilson and colleagues [1]. We would like to congratulate the authors on a well-conducted, randomised, controlled study that supports the growing body of evidence that epidural analgesia is not an independent factor associated with initiation or duration of breastfeeding.

We conducted a retrospective review of our maternity database to determine factors affecting mothers’ intention to breastfeed at the time of discharge from hospital. There were 13 714 deliveries during the 27-month review period (July 2007 to October 2009). Breastfeeding intention at discharge was recorded in 13 320 mothers, of whom 5652 intended to breastfeed their babies or use breastfeeds with some form of complement feeds. Seven thousand, six hundred and sixty-eight intended to use artificial feeds solely. Statistical analysis was performed with MINITAB 15.1 Statistical Software. Factors affecting intention to breastfeed at discharge were analysed using binary logistic regression. Odds ratios (OR) with 95% CI were calculated for each variable. A p value of < 0.05 was considered significant. We defined a non-ethnic group as mothers belonging to White-British backgrounds and ethnic as African/Caribbean-Black, Chinese, Indian, Pakistani, non-British White and Asians of unspecified origin.

Our results showed that epidural uptake is not an independent factor affecting intention to breastfeed following discharge from hospital after childbirth (p = 0.319), supporting the work of Wilson et al. Our analysis also included smoking, a variable not considered in their paper. Apart from ethnic group (OR 5.06 (95% CI 3.94–6.44), p < 0.001), non-smoking (OR 3.11 (95% CI 2.47–3.92), p < 0.001) was strongly associated with intention to breastfeed. The multiplicative effect of the model suggests that a non-smoker ethnic woman is estimated to have 5.06 × 3.11 = 15.73 times the odds of intending to breastfeed compared to a woman who is non-ethnic and a smoker. Advancing age was associated with an increased intention to breastfeed (OR 1.12 (95% CI 1.10–1.14), p < 0.001). Factors that had a negative impact on intention to breastfeed were multiparity (OR 0.57 (95% CI 0.47–0.70), p < 0.001), caesarean delivery (OR 0.75 (95% CI 0.59–0.96), p = 0.023) and opioid use during labour (OR 0.82 (95% CI 0.69–0.97), p = 0.022). Obesity (p = 0.992) and gestation at delivery (p = 0.729) showed no effect.

Our retrospective review of over 13 000 women showed no link between epidural use and intention to breastfeed. Instead, smoking and not belonging to an ethnic group were the main factors associated with not intending to breastfeed.

No external funding and no competing interests declared.


  1. Top of page
  2. Reference