Nipple shields and milk yields: an update

In a previous post I questioned the received wisdom that using nipple shields will have a negative impact on your ability to nurse your baby (see nipple shields: always a bad thing?). The post challenged one of the main criticisms made of shields — that they slow milk transfer and may therefore mean your baby is inadequately nourished — on the grounds that most of the studies demonstrating this were conducted a long time ago, and more recent research indicates that this problem does not exist for modern shields1.

The researchers who conducted the study in question concluded that nipple shields do not affect the amount of milk babies consumed in a feed by weighing before them before and after they nursed: when a mother was using a shield, the amount by which her baby had increased in weight at the end of the feed was roughly the same as when she was breastfeeding directly. Whilst this result looks positive for nipple shields, ‘test-weighing’ babies in this way is not without its critics, so one could argue that concluding nipple shields have no effect on milk consumption based solely on this evidence is a little premature.

Since writing the post, a follow-up study has been published, this time looking at the relationship between shield use and infant weight gain over a much longer period2. 54 mothers who used a nipple shield provided by a nurse or lactation consultant in the period just after the birth of their babies were recruited for the study, and completed interviews when their babies were 2 weeks, 1 month and 2 months old. Over time, the proportion of mothers using shields diminished (at 2 weeks,  69% of the mothers were still doing it, at 1 month 48%, and at 2 months 33%) and at each stage the responses of women who were still using the shields were compared with those who weren’t.

The main aim of the study was to determine whether nipple shields had a negative impact on weight gain — if babies whose mothers were still using shields grew more slowly than those whose mothers had stopped, then this could be taken as an indication that medium to long term use of shields was causing a real problem. Happily, there was no difference between the groups: whether a mother used a shield made no difference to her baby’s pattern of weight gain.

There were a few complaints about nipple shields: 8 women thought they caused nipple soreness; 2 found them messy; 2 found them inconvenient and 3 had problems with them falling off. In spite of this, 90% of the mothers in the study said that using the shield was a positive experience, and 67% felt it helped prevent them from giving up breastfeeding.

If you are a mother who relies on a nipple shield to breastfeed, these results make reassuring reading. Although shields appeared to cause difficulties for a few women, these were generally minor, and crucially they concerned practical issues, not the health of their babies. As most women felt that shields helped to prolong the period that they were able to breastfeed, this study ultimately supports the view they could be an important intervention for mothers who are having problems, rather than something that will make them worse.

  1. J Obstet Gynecol Neonatal Nurs. 2006 Mar-Apr;35(2):265-72.
  2. J Clin Nurs. 2009 Nov;18(21):2949-55.

Breastfeeding and thrush: what are the treatment options?

medicineThe symptoms associated with thrush (a candida yeast infection) in breastfeeding women vary. Some have red or shiny nipples, fissures or flaking skin, others exhibit no visible signs at all1. There is one symptom, however, that is reported almost universally: pain. It is usually described ‘in the strongest terms, with words such as “agonizing” or “excruciating” often being used’, and can occur throughout a feed and continue for sometime afterwards2. Faced with this level of discomfort, it is not surprising that many mothers diagnosed with thrush feel unable to continue breastfeeding3,4.

This shouldn’t be the case, of course; fungal infections are highly treatable, so contracting one need not automatically spell the end of breastfeeding. Unfortunately, getting medication for this type of thrush isn’t always straightforward, particularly if you are unlucky enough to have an unsympathetic doctor (see breastfeeding and thrush: it’s complicated). Health professionals who are reluctant to diagnose thrush, are naturally hesitant to prescribe for it, so some mothers may be left suffering unnecessarily, or given the wrong medication, such as antibiotics, which may actually make the condition worse2.

An additional problem is that the evidence base for treating breastfeeding yeast infections is sorely lacking. Although there are many antifungal drugs available, there have yet to be any controlled clinical trials examining their effectiveness for treating a candida infection of the breast. The difficulty confirming a diagnosis may be one reason for the lack of trials: milk and skin cultures aren’t always reliable, so it could be difficult to know whether a treatment failed because the drug was ineffective, or because the symptoms weren’t actually caused by candida. Alternatively, it may be because this research simply isn’t viewed as a priority: if an infection is caused by candida, then it should clear up eventually providing enough antifungals are thrown at it5. This ignores the possibility that the unique environment breasts are subject to during nursing may impact on the effectiveness of a treatment, but at present it’s the most ‘scientific’ approach we have.

Case studies and anecdotal information sources (which generally consist of health professionals giving opinions based on their clinical experience) also offer theories about the best way of treating thrush. A traditional medication for nipple thrush recommended on numerous breastfeeding websites (albeit mostly in articles written by the same person) is gentian violet, a purple ointment that can be applied to both nipples and babies’ mouths. Although there are no controlled clinical trials supporting its use for nipple thrush, it is known to be an effective antifungal, and there is anecdotal evidence it can be helpful for breastfeeding women2. Gentian violet does have its drawbacks, however. A minor issue is that it is messy (it is used as a dye); a more serious concern is that it can cause skin irritation6 and may be carcinogenic7, and as such it is not available for this purpose in the UK.

Due to the lack of clinical research in this area, there is no definitive list of drugs that are suitable for treating thrush in breastfeeding women, so theoretically any medication that is antifungal and unlikely to cause problems for a nursing baby could be prescribed. Common topical treatments include miconazole and clotrimazole, which are usually supplied in creams or powders. These are not advised for internal use, so it is generally recommended that they are removed before breastfeeding, although this is due to the ingredients in the base of the medication, rather than the antifungals themselves (both can be used to treat oral thrush if they are supplied in the appropriate preparation). Nystatin can also be applied topically to the nipples and is usually supplied in a formulation that does not need to be washed off. It is often used for treating thrush in babies’ mouths, and as such is often the first medication suggested for treating mothers too5.

Anecdotal evidence suggests that topical treatments aren’t always effective, however, and thrush can be treated more reliably with fluconazole5,8 (a clinical study suggests that this is the case for oral thrush too9.) Unlike creams and ointments that only treat the skin where they are applied, fluconazole has the additional advantage of being taken internally, providing a blanket assault on candida wherever the site of infection.

Although organizations like the Breastfeeding Network and the NHS suggest that fluconazole is a good treatment for mothers who appear to be suffering from persistent thrush, it can pass into milk in small amounts, and is not currently licensed for use when breastfeeding. As it can be safely given to newborns, this is not a great concern. The fact that the breastfeeding box isn’t officially ticked can make doctors cautious, however, leaving some women without potentially important medication. The opinion that fluconazole shouldn’t be given to breastfeeding women unless it is part of a controlled clinical trial can also be found in the scientific literature10, and while this idea is reasonable in theory, until someone actually gets round to running the trial, it isn’t so great for those women experiencing ‘intolerable and incredible pain’ that may be due to thrush11.

At present, there isn’t any concrete evidence that fluconazole (or indeed, any medication) is suitable for treating yeast infections in breastfeeding women, because there haven’t been any controlled clinical trials looking at its use in this situation. Until such trials are conducted, there are a couple of options. One is to deny women antifungal medication on the grounds that there is no proof it works. An alternative – supported by numerous breastfeeding organizations – is to advise women with suspected thrush to take fluconazole orally, to treat nipples topically and ensure their babies’ mouths are treated too, in the hope of alleviating symptoms. Proper research in this area would naturally be a great step forward, but until it occurs (if it ever does), it seems only fair to offer women suffering very painful symptoms at least a chance at respite, particularly if it enables them to continue breastfeeding.

  1. J Hum Lact. 2004 Aug;20(3):288-95.
  2. Aust N Z J Obstet Gynaecol. 1991 Nov;31(4):378-80.
  3. J Obstet Gynecol Neonatal Nurs. 2005 Jan-Feb;34(1):37-45.
  4. Clin Pediatr (Phila). 2001 Sep;40(9):503-6.
  5. J Hum Lact. 1999 Dec;15(4):281-8.
  6. J Hum Lact. 1990 Dec;6(4):178-80.
  7. Fundam Appl Toxicol. 1985 Oct;5(5):902-12.
  8. J Hum Lact. 2002 May;18(2):168-71.
  9. Pediatr Infect Dis J. 2002 Dec;21(12):1165-7.
  10. Breast. 2002 Feb;11(1):88-90
  11. J Hum Lact. 1997 Dec;13(4):307-11.

Breastfeeding and thrush: the difficulty getting a diagnosis

stack-of-petri-dishesAlthough many lactation specialists agree that breasts are susceptible to thrush (the candida fungus) when a woman is nursing, others feel that the yeast infection is ‘overdiagnosed and overtreated1.’ A quick perusal of the Mumsnet discussion forum shows that some health professionals are prepared to go a step further, and deny it is a problem at all (see breastfeeding and thrush: it’s complicated).

The scepticism that thrush can interfere with breastfeeding exists because at present there is no absolute proof that candida is the root cause of breastfeeding pain. A mother can become colonized with candida but remain infection-free, or can display many of the symptoms commonly associated with thrush (such as red, shiny, flaky or burning nipples) but test negative for it in the laboratory2.

This doesn’t mean that thrush isn’t a problem however: there are several controlled clinical studies that point to an association – not a perfect one, but a significant one nonetheless – between the presence of candida and certain symptoms suffered by breastfeeding mothers.

Evidence that thrush can affect breastfeeding mothers

Support for this relationship was provided by a study that assessed whether mothers who tested positive for candida shortly after they had given birth went on to develop the symptoms of thrush2,3. Nipple swabs and milk samples that were taken from 100 women at their two week post-birth check underwent microbiologic culturing to test for candida, and women completed two interviews to determine whether they had symptoms associated with thrush: one at the time the swabs were taken, and another 7 weeks later.

Although there was no association between colonization and symptoms initially, 20 of the 23 women colonized by candida went on to develop at least one of the symptoms of thrush (burning nipples, stabbing breast pain, shiny and/or red nipples). The percentage of cases (breasts rather than mothers, as it’s possible to be affected on only one side) with a positive or negative candida result that developed symptoms are listed below.

Nipple culture results (22 positive cases, 178 negative cases):

  • 95% of positive cases and 49% of negative cases reported sore nipples
  • 100% of positive cases and 33% of negative cases reported burning nipples
  • 82% of positive cases and 18% of negative cases reported non-stabbing breast pain
  • 91% of positive cases and 24% of negative cases reported stabbing breast pain
  • 45% of positive cases and 11% of negative cases reported shiny nipples
  • 36% of positive cases and 13% of negative cases reported flaky nipples

Milk culture results (32 positive cases, 168 negative cases):

  • 78% of positive cases and 50% of negative cases reported sore nipples
  • 81% of positive cases and 33% of negative cases reported burning nipples
  • 72% of positive cases and 16% of negative cases reported non-stabbing breast pain
  • 81% of positive cases and 21% of negative cases reported stabbing breast pain
  • 41% of positive cases and 10% of negative cases reported shiny nipples
  • 38% of positive cases and 11% of negative cases reported flaky nipples

Every symptom occurred at a significantly higher rate in the cases where candida had been detected. Each symptom was also reported in several negative cases, however, indicating either that the mothers had become colonized by candida after the initial test, or that the symptoms were caused by something else.

Another study investigating the link between the results of milk cultures and breastfeeding pain also reported that candida was found in a significantly greater proportion of women who were suffering from sharp, shooting breast pain: 6/20 (30%) of the women with pain tested positive for it, but only 6/78 (5%) of the no pain group4. This study also tested for levels of common bacteria, and found they did not differ significantly between the two groups.

Evidence that symptoms associated with thrush may be caused by something else

There are other experiments, however, which indicate that bacteria – rather than fungi – may be the culprit in some occurrences of breastfeeding pain. A study comparing culture results of 20 women with deep, stabbing pain, 20 with superficial nipple pain and 20 controls with no pain found an association between candida and superficial pain, and bacteria and deep pain: candida was found on the nipples of 1 of the control group, 12 of the superficial pain group and 10 of the deep pain group; and in the milk of 1 of the control group, 10 of the superficial pain group and 5 of the deep pain group5. By contrast, pathogenic bacteria were detected on the nipples of 4 of the control group, 11 of the superficial pain group and 19 of the deep pain group, and in the milk of 6 of the control group, 8 of the superficial pain group and 14 of the deep pain group.

Support for the idea that thrush does not underlie all breastfeeding pain is also provided by a recent study that compared levels of candida in the milk of 18 healthy breastfeeding mothers and 16 with thrush symptoms (sore, inflamed or traumatized nipples, intense stabbing or burning pain and painful breastfeeding that had not received any other diagnosis)6. The investigators were very careful to avoid contamination of the sample: nipples were thoroughly washed, and breasts were pumped for 12 minutes before any milk was collected. No evidence of candida was found in any of the samples, leading the authors to suggest that it does not infect milk ducts, and is unlikely to be the cause of the symptoms associated with thrush. They also go a step further, and suggest that positive nipple cultures cannot be trusted, as they may be contaminated by the baby’s saliva, which often contains candida. They conclude that, ‘candida infection on the mother’s skin remains a possibility, but it seems unlikely.’

This interpretation of the results seems fairly extreme. In previous studies, candida has been found to colonize milk (but not nipples)2 so the possibility that thrush may infect milk ducts cannot be ruled out. It is also not really appropriate to draw conclusions from the data about candida infection of the nipple, or the role of bacteria in breast pain, as neither of these things was actually examined. There is some evidence that bacteria may underlie certain instances of breast pain – it has been found at higher concentrations than candida in the milk of women suffering from deep pain5 – but this result conflicts with another reporting higher levels of candida in women suffering from a similar complaint4.

It is not only the role of bacteria that is hazy, however: despite the association between candida and symptoms reported in some studies, this link is far from concrete. Not every woman with the symptoms of thrush tests positive for candida, and some who are colonized by it never report symptoms. If you also consider that cultures themselves can be easily contaminated and may therefore be unreliable, the picture gets even more complicated.

Treating the symptoms of thrush

What does all this mean for breastfeeding women who appear to be suffering from thrush? Some researchers take the view that as the relationship isn’t sufficiently proven, women should not be treated with antifungal medication, or should receive it only as part of a controlled trial7. Others take a more pragmatic view, and suggest that such medication can actually have a useful diagnostic value: if the symptoms clear up after using an antifungal, then this ‘confirms’* that the symptoms were caused by candida8. This is far from an ideal solution (using such medications unnecessarily is expensive and increases the chances of fungi developing resistance to them), but it may be preferable to leaving women to suffer with very painful symptoms when there is chance they could be cured. Women who exhibit the symptoms of thrush when breastfeeding are at a very high risk of giving up3, so if there’s a medication that could stop this from happening, it may not be a bad idea to use it.

  1. Breastfeed Med. 2009 Jun;4(2):55.
  2. J Hum Lact. 2004 Aug;20(3):288-95.
  3. J Obstet Gynecol Neonatal Nurs. 2005 Jan-Feb;34(1):37-45.
  4. Am J Obstet Gynecol. 2007 Oct;197(4):424.e1-4.
  5. Gynecol Obstet Invest. 1998 Aug;46(2):73-4.
  6. Breastfeed Med. 2009 Jun;4(2):57-61
  7. Breast. 2002 Feb;11(1):88-90
  8. J Hum Lact. 1999 Dec;15(4):281-8.

*Of course, this doesn’t really confirm that a mother was suffering from thrush, as the alleviation of her symptoms may have been coincidental.

Breastfeeding and thrush: it’s complicated

woman-doing-mathI sometimes feel as if I’ve experienced pretty much every breastfeeding problem going: sore/bleeding/fissured nipples; mastitis (although fortunately only the early stages); not enough milk; too much milk; and a very temperamental let-down reflex. One of the only things I haven’t suffered from is thrush – a fungal or yeast (candida) infection that allegedly causes excruciating nipple and breast pain. I use the word ‘allegedly’ simply because it is not always easy to determine whether the symptoms associated with thrush are definitely caused by a fungal infection, or whether they are in fact down to something else. I am not a thrush doubter – I think there is sufficient evidence to justify taking it very seriously, as does the NHS. Many health professionals who do not specialize in breastfeeding are yet to be convinced, however, as the Mumsnet discussions below testify:

(These are just a drop in the ocean – you will find all manner of breastfeeding ignorance from health professionals on these noticeboards. The GP who suggested a mother might pass mastitis on to her baby deserves a special mention.)

It isn’t just those outside the field who disagree about thrush: within the scientific literature there are conflicting results and opinions, as well as holes in clinical knowledge because the relevant research simply hasn’t been conducted. I’m currently wading through the published work in this area, and am finding it’s actually quite difficult to get to the bottom of the relationship between yeast infections and breastfeeding problems. As it’s important to try to make sense of it though, I’ll be writing several posts on it over the next few weeks, starting with one that attempts to address the controversy that still surrounds the diagnosis.

Should ‘moist wound healing’ principles be applied to cracked nipples?

pot-of-vaselineMaintaining a slightly damp, rather than dry, environment under a dressing has been shown in many studies to help wounds heal faster. Moisture prevents a scab from forming, which allows new epithelium cells to move across the surface of the wound more quickly, and reduces the time it takes it to close1.

As the technique appears to improve the healing rates of a range of skin lesions2, it has been suggested that its benefits may extend to cracked nipples3, and there are now several off-the-shelf creams and dressings that claim to maintain a moist environment, and thus help the healing process. The inference is not, at first glance, unreasonable – if moisture helps skin to regenerate more quickly, then it may well help injured nipples to recover too. It is important to remember, however, that nipples are subject to a unique set of circumstances when a mother is breastfeeding, so it is also quite possible that the application of the technique might not be appropriate. Is there any scientific evidence that applying moist wound healing principles can aid the recovery of cracked or injured nipples?

In a study comparing dressings made from hydrogel with lanolin (both classified by the Breastfeeding Network as having ‘moist wound healing’ properties)4, 106 mothers were allocated at random to one of two groups: the first was given hydrogel dressings to use between feeds; the second was given lanolin cream. Mothers started using the treatments within 24 hours of giving birth, and their effectiveness was assessed via telephone interviews 3, 4 or 5, 7, 10 and 12 days later. The researchers found mothers reported significantly lower pain scores (a difference of just under 1 point, on a scale of 1-5) in the hydrogel group on days 10 and 12.

There were a couple of problems with this study (actually, there were several, but I’ll limit it to the major ones). Firstly, there was no baseline group of mothers not applying anything to their nipples, so it is not possible to say whether either treatment was better than simply leaving nipples alone. A second issue is that the people conducting the interview were aware, when they spoke to the participants, which treatment they were receiving. The researchers claim that to have conducted the study blind would have been ‘impossible’, an unsubstantiated and somewhat odd statement, as it would appear to be completely possible to interview a mother over the phone without knowing what she had on her nipples. Knowledge of the treatment group in this type of study is a problem if there is any chance that the researcher may have a bias towards a particular treatment, as they may subconsciously influence the patients’ responses. It may be worth mentioning at this point that the research was funded by Tyco, the manufacturers of the dressings.

These criticisms are, however, a digression. The main thing to note about this research is that it did not test whether the dressings actually helped wounds to heal. Although moist wound healing is touted in the introduction as the ‘science bit’ justifying the use of the dressings, it is not mentioned anywhere in the procedure or the results.

A hospital-funded study comparing hydrogel and lanolin – this time documenting the impact that the treatments had on bleeding and cracked nipples – did not find the dressings to be quite so effective5. Researchers who were blind to the treatment group rated nipples as healing significantly better when women used lanolin with breast shells, rather than hydrogel dressings. Self-reported measures of pain were also significantly lower in the lanolin group. A final point worth mentioning is that the study was halted early, due to a third of the 21 women in the hydrogel dressing group developing an infection.

Although this study compared two treatments, it is again compromised by the lack of a proper control group. We can see that lanolin appears to result in improved healing and lower pain scores when it is compared with hydrogel, but we still do not have any evidence that moist wound healing techniques are useful for treating injured nipples when breastfeeding – to ascertain this requires a control group where mothers keep their nipples dry.

So far, there appear to be only two studies that have looked at this issue. An experiment published in 1995 examined whether using a polyethylene adhesive dressing had any effect on the development of nipple redness, fissures and pain6. 50 mothers took part in the study, using a dressing on one nipple, and leaving the other untreated. Although the mothers reported less pain when feeding with the treated nipples, the researchers caution that this may simply have been because of the ‘Hawthorne Effect’ (the mere fact there is an intervention is enough to cause an improvement.) The dressings made no difference to the development or healing of fissures or redness, as reported by observers blind to the treatment group. 16% of the participants dropped out due to finding the dressings uncomfortable, and 66% said they found it uncomfortable to remove them – something they had to do before every feed.

A more recent study looked specifically at the effect of lanolin on the healing of nipple fissures7. 225 women, all with fissures, were randomly allocated to one of three groups: in the first group mothers applied lanolin 3 times a day; in the second they applied breast milk after each feed; in the third they applied nothing. The appearance of their nipples was assessed 3, 5, 7 and 10 days after starting the treatment by researchers who did not know which group the mothers were assigned to. There was no significant difference in healing time between the breast milk and no-treatment groups. The nipples of the women using lanolin, however, took significantly longer to heal (45% of this group took longer than 7 days, as opposed to 32% of the milk group, and 25% of the no-treatment group).

There is another area of research, which does not examine the use of moist wound healing directly, but is still relevant to the debate. Broken skin makes nipples vulnerable to infection8, which may mean there is an additional problem with keeping nipples damp, rather than dry: organisms like thrush are known to thrive in warm, moist environments9.

Given the possible risk of infection, and the evidence that maintaining a moist environment around cracked nipples may potentially delay the healing process, it seems to unwise to recommend the application of the products described above to mothers with cracked or fissured nipples. In spite of this, they continue to be promoted by both commercial companies and health professionals on scientific grounds. Until evidence that genuinely supports its use is found, presenting moist wound healing to mothers as a clinically tested treatment is at best misguided, and at worst dishonest.

  1. Nature. 1962 Jan 20;193:293-4.
  2. Br J Nurs. 2008 Aug 14-Sep 10;17(15):S4, S6, S8 passim.
  3. J Hum Lact. 1997 Dec;13(4):313-8.
  4. J Obstet Gynecol Neonatal Nurs. 2003 Jul-Aug;32(4):486-94.
  5. Arch Pediatr Adolesc Med. 1998 Nov;152(11):1077-82.
  6. Nurs Res. 1995 Nov-Dec;44(6):347-51
  7. Saudi Med J. 2005 Aug;26(8):1231-4
  8. J Hum Lact. 1991 Dec;7(4):177-81.
  9. Hum Lact. 1999 Dec;15(4):281-8.
Follow

Get every new post delivered to your Inbox.

Join 25 other followers