Use of complementary and alternative medicines for menopausal symptoms in Australian women aged 40–65 years

For many of the 3.7 million women aged 40 to 65 years in Australia, who account for 14% of the population, menopause presents a health challenge. Use of oestrogen for the relief of menopausal vasomotor symptoms (VMS) (eg, hot flushes, night sweats) was common until 2002, when concerns about the safety of menopausal hormone therapy (MHT) arising from a Women’s Health Initiative study1 triggered a dramatic decline in the prescription of MHT. Concurrently, the use of complementary and alternative medicine (CAM) has grown considerably,2 with one Australian study reporting that most women presenting to a menopause clinic had used CAMs for their menopausal symptoms.3 However, there is insufficient evidence to support the use of CAMs that are most visibly promoted for managing VMS, such as black cohosh and phytoestrogens.4,5 Women at midlife are also likely to be taking other CAMs to manage other menopausal symptoms.

The current prevalence of the use of CAMs specifically for menopausal symptoms in Australia is not known. We aimed to investigate the prevalence of CAM use and consultations with CAM practitioners for menopausal symptoms by Australian women at midlife.

Methods

Study participants

Recruitment to this study has been reported in detail elsewhere.6 In brief, between October 2013 and March 2014, women aged 40–65 years listed on a large, representative, national, continually refreshed database based on the electoral roll were randomly invited by telephone to participate in a survey about the health of women at midlife. The only exclusion criterion was inability to complete a questionnaire in English. A study questionnaire with a reply-paid envelope was posted to each woman who gave verbal consent. Return of a completed questionnaire was accepted as written informed consent. The Monash University Human Research Ethics Committee approved the study. We used the STROBE guidelines to report the study findings.

Questionnaire

The questionnaire provided women with a specific list of CAMs that are commonly used for relieving menopausal symptoms and asked them to indicate which of these they had used in the past 4 weeks. For our analyses, we grouped these CAMS according to whether they are commonly used for VMS (eg, black cohosh, phytoestrogen products), sexual symptoms (eg, horny goat weed, maca) or other menopausal symptoms such as joint pain, sleep disturbance and mood (eg, fish or krill oil, glucosamine).

The women were also asked about their current use of systemic MHT, including prescription oestrogen, oestrogen plus progestogen (including the oral contraceptive pill and compounded hormones) and tibolone.

The questionnaire further provided women with a list of CAM practitioners and asked them to indicate which, if any, they had consulted for menopausal symptoms in the past 12 months. The practitioner list comprised: chiropractor, naturopath, acupuncturist, herbalist, Chinese medicine practitioner, homeopath, osteopath, spiritual healer, kinesiologist, aromatherapist, iridologist and Ayurveda therapist.

Assessment of menopausal status

Women were classified as premenopausal, perimenopausal or postmenopausal using published algorithms7 derived from the Stages of Reproductive Aging Workshop + 10 criteria.8 Classification was based on age, history of bilateral surgical menopause, menstrual bleeding, VMS and hysterectomy. Use of systemic hormonal contraception or MHT was also taken into account. We further divided the postmenopausal group into women aged 40 to < 55 years, 55 to < 60 years, and 60–65 years.

Assessment of menopausal symptoms

We assessed menopausal symptoms using the Menopause-Specific Quality of Life (MENQOL) questionnaire, a validated self-administered instrument comprising four symptom domains: vasomotor, physical, psychological and sexual.9 Women reported if they had experienced each symptom in the past 4 weeks (yes/no) and the extent to which they were bothered by it (ranging from 0 [no bother] to 6 [extreme bother]). Each individual question had a range of 1–8, with 1 being no symptoms. The symptoms were rated on a seven-point Likert scale, and women who ranked their degree of bother as more than the midpoint of the item (score, > 5–8) were considered to have moderate–severe symptoms.

Sample size

We calculated that our required sample size was 2000, based on the primary study outcome of moderate–severe VMS, with a 95% confidence interval of ± 2% around a percentage prevalence estimate of 30%. We purposefully sampled women so that the age distribution of our sample population mimicked the age distribution of Australian women aged 40–65 years in 2011.6

Statistical analyses

The characteristics of study participants, use of CAMs and consultations with CAM practitioners are presented as frequencies and percentages.

We performed three separate logistic regression analyses, where the outcome variables were CAM use for VMS, CAM use for other symptoms, and consultations with CAM practitioners. Too few women were using a CAM intended for sexual function to justify separate regression modelling for this outcome. Factors associated with CAM use for VMS and other symptoms in bivariate analysis at P < 0.05 were included in the final regression models. All tests were two-sided and the a level was 0.05 for all statistical tests. We performed all analyses using Stata version 12.0 (StataCorp).

Results

Of 5850 women who were contacted, 2911 agreed to participate, and 2023 subsequently returned completed questionnaires. Three completed questionnaires were from women outside the study age range, so the final sample consisted of 2020 women (response rate, 34.53%). Most women were postmenopausal (1109; 54.90%), resided in metropolitan areas and were born in Australia (Box 1).

Current use of MHT was reported by 12.18% of the women; 13.24% had undergone hysterectomy without oophorectomy, and 5.45% had surgical menopause (Box 1). Among women not using MHT, 52.06% reported having VMS, 86.69% had psychological symptoms and 97.40% had physical symptoms. Among women not using MHT or vaginal oestrogen, 57.18% reported sexual symptoms.

Of women reporting any symptoms in the MENQOL domains, the symptoms were reported as moderate–severe by 12.80% (221/1727) for VMS, 19.86% (332/1672) for sexual symptoms, 15.39% (266/1728) for physical symptoms, and 16.15% (279/1728) for psychological symptoms.

CAM use for menopausal symptoms

The prevalence of use of at least one CAM for any menopausal symptoms was 39.16% (Box 2). Of the 2020 women, 267 (13.22%) reported using CAMs for VMS, with the most frequent use reported by perimenopausal women (18.85%; 72/382) (Appendix 1). Phytoestrogens were the most commonly used CAM for VMS (6.29%), followed by evening primrose oil (3.91%), ginseng (1.73%) and black cohosh (1.49%). Nearly half the women who used MHT had also used a CAM (112/246; 45.53%), including 34 women (13.82%) who had used a CAM for VMS. Only 16 women (0.79%) used a CAM for sexual symptoms.

Of the 2020 women, 651 (32.23%) used at least one CAM for other symptoms, with the most frequent use by postmenopausal women aged 60–65 years (Appendix 1). Overall, fish or krill oil was most commonly used for other symptoms (24.36%), followed by glucosamine (12.13%) (Box 2). Of the 358 women aged 60–65 years, 128 (35.75%) reported taking a fish oil supplement and 61 (17.04%) were taking glucosamine.

Factors associated with CAM use for menopausal symptoms

Compared with premenopausal women, perimenopausal and early postmenopausal women were significantly more likely to use any CAM for VMS (Box 3).

The likelihood of CAM use for other symptoms was significantly greater for perimenopausal and all postmenopausal women than for premenopausal women. Women residing in non-metropolitan areas were less likely than metropolitan residents to use a CAM for other symptoms. Experiencing psychological or physical symptoms was not associated with the use of CAMs for other symptoms (Box 3).

Consultations with CAM practitioners

Overall, 8.33% of women (168/2017) reported having at least one consultation with a CAM practitioner for menopausal symptoms in the past 12 months (Appendix 2). The practitioners most frequently consulted were chiropractors (2.78%), naturopaths (2.68%) and acupuncturists (1.98%). The women most likely to have consulted a CAM practitioner were those experiencing VMS (adjusted odds ratio, 1.80; 95% CI, 1.20–2.71) and those who were educated beyond high school (adjusted odds ratio, 1.56; 95% CI, 1.07–2.29) (Appendix 3).

Discussion

In this large, representative, cross-sectional sample of Australian women aged 40–65 years, we found that 13.22% had used at least one CAM for VMS, similar to the proportion of women of this age who were using MHT. More than one in 10 women using MHT also used a CAM for VMS. We also found that a third of study participants reported the use of CAMs for other symptoms.

CAM use for VMS was highest in perimenopausal and early postmenopausal women, with the prevalence in these groups close to one in five. These women, compared with premenopausal and older postmenopausal women, also reported the most severe menopausal symptoms.6

Applying our findings to the 3.7 million Australian women aged 40–65 years would mean that 490 000 women had used CAMs for VMS in the past month, and 303 000 would have consulted a CAM practitioner for menopausal symptoms in the past year. In addition, 883 000 women would have taken fish or krill oil and 438 000 would have taken glucosamine in the past month.

Phytoestrogens, the most commonly used CAMs for VMS, have not been shown to be effective for treating VMS, either as food supplements or as concentrated tablets.5 Evening primrose oil, black cohosh and ginseng, the other commonly used CAMs for VMS, overall appear to be no more effective than placebo.10,11 Importantly, CAMs such as ginseng will have varying effects according to the species used and potency and bioavailability of the formulation.12,13 The effects are not always desirable, with the known side effects of ginseng including hypertension, diarrhoea, sleeplessness, mastalgia with diffuse mammary nodularity, skin eruptions and vaginal bleeding.12,13 Given the lack of evidence regarding benefit of CAMs for alleviating VMS, as well as the potential adverse effects and their high cost, the continuing use of these CAMs for this purpose cannot be supported.

We found that the prevalence of use of CAMs for other symptoms was high among older Australian women, with fish or krill oil and glucosamine the most commonly used. Fish oil is promoted for a wide range of conditions, most notably for preventing cardiovascular disease.14 Fish oil inhibits platelet aggregation and has been associated with a modest increase in bleeding time,15 but there has been controversy regarding its efficacy in preventing cardiovascular disease.16 Glucosamine, taken alone or in combination with other supplements such as chondroitin, is most commonly used for arthralgia. However, the United Kingdom’s National Institute for Health and Care Excellence does not recommend the use of glucosamine or chondroitin for the management of osteoarthritis.17 Recent studies linking glucosamine to ß-cell dysfunction and decreased insulin secretion18 raise concern, particularly in the context of high usage of this CAM by older people.

The higher prevalence of CAM use for symptoms other than VMS among women residing in metropolitan rather than non-metropolitan areas may reflect availability and income differentials between these regions. Others have found that women with poor self-reported health or chronic diseases are more likely than their healthier counterparts to take supplements.19 In our study, women with psychological and physical symptoms of menopause were not more likely to take a CAM. This could be because such a high proportion of women reported psychological and physical symptoms, and because of the widespread indiscriminate use of fish oil and glucosamine in the community.

There is evidence that CAM use is supported by health practitioners. A national survey found that about a third of Australian general practitioners self-identified as practising complementary therapy.20 It is a cause for concern that a sizeable proportion of Australian practitioners are recommending ineffective therapies.

We found that the proportion of women consulting a CAM practitioner for menopausal symptoms was threefold lower than that reported by menopause clinic respondents (25.2%) in Australia.3 This difference is likely because our study used a community-based, rather than clinic-based, sample. Our finding that chiropractors and naturopaths were the most commonly consulted CAM practitioners is consistent with a previous community-based survey.14

Our finding of a higher prevalence of CAM consultations among women educated beyond high school is similar to that in another study, which found that better educated women were more likely to consult CAM practitioners.21 This may be a function of greater uptake of private health insurance (PHI) among educated women. Higher educational achievement, as a proxy measure of higher income, has been found to be associated with greater uptake of PHI.22 Furthermore, people with PHI have been found to be significantly more likely than those without PHI to see CAM practitioners in Australia.23 Many PHI funds provide cover for consulting CAM practitioners, such as homeopaths and naturopaths. However, these rebates are not yet underpinned by credible evidence of clinical efficacy, cost-effectiveness, safety and quality.24

In contrast with consulting CAM practitioners, we found that use of CAM products was not associated with education level. This may be because of the lower cost burden of CAM products relative to consultations with CAM practitioners, and greater acceptance of use of CAM products in the community.14,25

This is the first Australian community-based study of use of CAMs and CAM practitioners specifically for menopausal symptoms. Previous studies have reported overall use of CAMs by Australian women at midlife, with prevalence ranging from 48.5% to 82.5%,2,3,14 higher than our finding of 39.16%. The higher prevalence in these studies is due to the inclusion of a broader range of CAM therapies, such as change in diet, exercise and use of vitamins. A recent Australian study reporting a higher rate of consultations with CAM practitioners was not menopause-specific and included practitioners such as massage therapists.26

An important strength of our study is the use of national random recruitment of a representative sample of women, rather than convenience sampling. Although we were unable to compare the demographics of respondents and non-respondents, comparison with Australian 2011 census data has shown that our sample is nationally representative of women in this age range in terms of basic demographic characteristics, including metropolitan location.27

A further strength of our study is that participants were provided with a comprehensive list of the CAMs specifically used for menopausal symptoms.

A potential limitation is that proficiency in English was required to complete the study questionnaire. This is unlikely to have affected the representativeness of our sample, as English literacy in Australian women exceeds 96%.6 Our response rate of 34.53% indicates that selection pressures were operative, but this rate is high compared with rates recently reported for similar studies.28

In conclusion, a significant proportion of Australian perimenopausal and early postmenopausal women are using CAMs that are known to be ineffective for managing VMS. Furthermore, the use of fish oil and glucosamine supplements is common in the community, yet the efficacy and long-term safety of these supplements are yet to be established. Health care providers need to actively guide women in the management of VMS and other menopausal symptoms.

1
Basic demographics of study participants (n = 2020)

Characteristic

No. (%)

Mean age in years (SD)

52.62 (6.81)

Age group (years)

40–49

835/2020 (41.34%)

50–59

827/2020 (40.94%)

60–65

358/2020 (17.72%)

Menopausal status

Premenopausal

529/2020 (26.19%)

Perimenopausal

382/2020 (18.91%)

Postmenopausal

Aged 40–< 55 years

307/2020 (15.20%)

Aged 55–< 60 years

444/2020 (21.98%)

Aged 60–65 years

358/2020 (17.72%)

Residential location

Metropolitan

1266/2019 (62.70%)

Non-metropolitan

753/2019 (37.30%)

Region of birth

Australia

1623/2018 (80.43%)

Europe

235/2018 (11.65%)

Other

160/2018 (7.93%)

Ethnicity

White

1896/2012 (94.23%)

Aboriginal or Torres Strait Islander

15/2012 (0.75%)

Other

101/2012 (5.02%)

Education

High school or less

832/2017 (41.25%)

Beyond high school

1185/2017 (58.75%)

Current relationship status

Married or long-term relationship

1358/2016 (67.36)

Not currently partnered

658/2016 (32.64%)

Paid work

742/2016 (36.81%)

Yes

1274/2016 (63.19%)

Have children

315/2017 (15.62%)

Yes

1702/2017 (84.38%)

Current use of systemic HT

246/2020 (12.18%)

Hysterectomy only

251/1896 (13.24%)

Surgical menopause

110/2020 (5.45%)

Menopausal symptoms*

Any vasomotor symptoms

899/1727 (52.06%)

Any sexual symptoms†

956/1672 (57.18%)

Any psychological symptoms

1498/1728 (86.69%)

Any physical symptoms

1683/1728 (97.40%)

HT = hormone therapy.
*Excluding women using systemic HT, hormonal contraception or non-hormonal therapy for menopausal symptoms.†Also excludes women using vaginal oestrogen.

2
Use of complementary and alternative medicines (CAMs) for menopausal symptoms in the past 4 weeks (n = 2020)

CAM use	No. (%)*

At least one CAM use	791 (39.16%)
CAMs for vasomotor symptoms
Phytoestrogens†	127 (6.29%)
Evening primrose oil	79 (3.91%)
Ginseng	35 (1.73%)
Black cohosh	30 (1.49%)
Licorice extract	25 (1.24%)
Chasteberry (Vitex agnus-castus)	15 (0.74%)
Sage	11 (0.54%)
Dong quai	5 (0.25%)
Hops (Humulus lupulus)	5 (0.25%)
CAMs for sexual symptoms
Macafem (Koech Corp) or maca	10 (0.50%)
Horny goat weed	6 (0.30%)
CAMs for other symptoms
Fish or krill oil	492 (24.36%)
Glucosamine	245 (12.13%)
Valerian	51 (2.52%)
Unknown compounds from a herbalist or naturopath	35 (1.73%)
Selenium	26 (1.29%)
St John’s wort	25 (1.24%)
Chlorophyll	24 (1.19%)
St Mary’s thistle	23 (1.14%)
Turmeric (Nalgesic Forte [Nutrition Care Asia])	18 (0.89%)
Grape seed extract	11 (0.54%)
Saw palmetto	3 (0.15%)
Zizyphus	2 (0.10%)
Tribulus	1 (0.05%)

*20% of women used more than one CAM.†Included all phytoestrogen products (flax seed, linseed oil, phytoestrogen, red clover and soy supplements, as well as Promensil [PharmaCare Laboratories], MenoEze [Naturopathica], PhytoLife [Blackmores Australia] and Harmony [Martin & Pleasance]).

3
Logistic regression for the outcome variables of using at least one complementary and alternative medicine (CAM) for vasomotor symptoms (VMS) and other symptoms of menopause

Characteristic

Use of at least one CAM for VMS

Use of at least one CAM for other symptoms

Crude odds ratio (95% CI)

Adjusted odds ratio (95% CI)

Menopausal status

Premenopausal

1.00

Perimenopausal

2.09 (1.42–3.06)

< 0.001

1.40 (1.03–1.89)

0.03

1.41 (1.05–1.91)

0.02

Postmenopausal

Aged 40–< 55 years

1.83 (1.21–2.76)

0.004

1.67 (1.22–2.29)

0.001

1.67 (1.22–2.29)

0.001

Aged 55–< 60 years

1.17 (0.78–1.75)

0.46

2.04 (1.54–2.70)

< 0.001

2.04 (1.54–2.71)

< 0.001

Aged 60–65 years

1.10 (0.71–1.70)

0.68

2.54 (1.90–3.41)

< 0.001

2.53 (1.88–3.40)

< 0.001

Residential location

Metropolitan

1.00

Non-metropolitan

1.08 (0.83–1.41)

0.55

0.82 (0.68–1.00)

0.05

0.80 (0.65–0.97)

0.02

Education

High school or less

1.00

Beyond high school

1.10 (0.84–1.42)

0.50

1.17 (0.97–1.41)

0.11

Current relationship status

Married or long-term relationship

1.00

Not currently partnered

1.16 (0.89–1.53)

0.27

1.02 (0.84–1.25)

0.83

Paid work

1.00

Yes

0.93 (0.72–1.22)

0.61

0.94 (0.77–1.14)

0.53

Have children

1.00

Yes

0.73 (0.53–1.02)

0.06

1.06 (0.82–1.38)

0.65

Hysterectomy only

1.00

Yes

0.97 (0.65–1.43)

0.86

0.94 (0.71–1.26)

0.70

Surgical menopause

1.00

Yes

0.64 (0.33–1.23)

0.19

1.61 (1.10–2.38)

0.02

1.24 (0.83–1.85)

0.3

Any psychological symptoms

1.00

Yes

1.20 (0.88–1.64)

0.24

Any physical symptoms

1.00

Yes

1.25 (0.64–2.45)

0.51

[Correspondence] Mobilising the Campaign to End Fistula

In 2004, the United Nations Population Fund (UNFPA) and its partners launched the Campaign to End Fistula, yet how much has changed for women’s health since its conception? Overall, improvements in women’s health would appear positive; as maternal and child mortality rates have halved in the past two decades, a successful foundation has been established for the Sustainable Development Goals (SDG).1 This is a remarkable achievement for many developing countries in view of the rise in population size, declining rates in income growth, and increasing political instability.

Medibank-Calvary contracts stand-off: what it means for doctors and patients

Medibank Private Ltd has announced that its contract with private hospitals operated by Calvary Health Care will end on 31 August this year.

Medibank cited a breakdown in negotiations over the rates Medibank pays for services provided by Calvary hospitals and new quality criteria for the failure to renew the contract, while Calvary has stated the new demands are financially rather than quality driven.

This will not affect the fees paid by Medibank to medical practitioners.

However, medical practitioners will need to ensure patients insured with Medibank seek information directly from Medibank about their out-of-pocket costs for accommodation, theatre and other items if they are admitted to a Calvary Health Care hospital.

Medibank has advised that it will continue to pay benefits for hospital expenses at the current contracted rates for any procedures pre-booked before 31 August, for a period of:

nine months if the pre-booking is obstetrics-related
six months if the pre-booking is to treat a chronic condition, e.g. chemotherapy or dialysis
two months for all other pre-booked admissions.

After this, Medibank will pay previously-contracted Calvary Health Care private hospital expenses at ‘second tier’ rates which are set at 85% of Medibank’s average contracted rate.

Medical practitioners can call Medibank’s dedicated ‘doctor hotline’ on 1300 130 460 for further information.

Georgia Morris

Non-invasive prenatal testing: the new era in reproductive medicine

Important advances in screening for genetic abnormalities have major ethical implications.

In Australia, it is routine practice to offer pregnant women screening for fetal chromosomal and structural abnormalities, and, if serious anomalies are diagnosed, the option of terminating the pregnancy. Screening for chromosomal abnormalities commenced in the 1970s and was initially restricted to women whose pregnancies were considered to be of increased risk because of an obstetric history of aneuploidy or advanced maternal age. Over the past five decades, prenatal chromosome screening has been expanded to encompass the entire obstetric population. However, the commonly used aneuploidy screening tests are dogged by high false-positive rates, typically 4%–5%.1 Confirmation of an increased screening risk for aneuploidy involves diagnostic tests such as amniocentesis or chorionic villus sampling, each associated with low but definite risks of pregnancy loss.

Many women are reluctant to proceed with a diagnostic test after a positive aneuploidy screening result,2 and the ability to assess fetal genetic material without directly sampling the amniotic fluid or placenta has long been a goal of prenatal diagnosis. The recent development of non-invasive prenatal testing (NIPT), a high-level screening test using cell-free fetal DNA, offers the opportunity to markedly reduce the requirement for invasive testing while potentially also increasing detection rates of chromosomal anomalies, in particular of trisomy 21. In addition, NIPT may be offered earlier in pregnancy than standard aneuploidy screening and diagnostic techniques.

Advances in DNA sequencing provide sensitivity and specificity

The clinical introduction of NIPT for prenatal diagnosis was initially made possible by the recognition that cell-free fetal DNA (cffDNA) is detectable in the maternal bloodstream. cffDNA is placental in origin, released as the result of trophoblastic apoptosis, and is reliably detectable in maternal blood after about 10 weeks’ gestation.3 Secondly, advances in DNA sequencing technology have led to the development of next-generation sequencing (NGS) techniques. DNA sequencing determines the exact order of nucleotides in a DNA fragment, and NGS enables millions of DNA fragments to be sequenced simultaneously. The use of massively parallel sequencing to precisely measure the relative abundance of sequences from each chromosome forms the basis of NIPT for aneuploidy screening.4,5 Further technical modifications have led to the introduction of more targeted approaches, the sequencing of selected regions of interest using single nucleotide polymorphism-based targeted assays, and, most recently, microarray-based cell-free DNA analysis. The focus of NIPT has progressively shifted from trisomy 21 to the detection of other trisomies and sex chromosomal aneuploidies. With the increase in the amount of sequencing data that can be obtained, the introduction of non-invasive testing for some chromosomal microdeletions and microduplications has become possible. These subchromosomal aberrations are much less common than the aneuploidies typically screened by existing tests, and these newer assays still require population studies to assess their performance characteristics.6

Although initially used in pregnancies at high risk for aneuploidy, recent data indicate that NIPT is also a robust screening test in lower-risk pregnancies.7 NIPT is now the most sensitive and specific screening test for the common trisomies, with detection rates greater than 99% for trisomy 21 and false-positive rates of less than 0.5%. The performance characteristics for trisomies 13 and 18 and the sex chromosome anomalies are lower than for trisomy 21, although the sensitivity is still typically greater than 90%.8

NIPT is currently available to Australian women from several international private companies on a user-pays basis, and a local provider has also recently offered the service. The price of testing varies between $420 and $900, and this is currently seen as too expensive for NIPT to be adopted as a publicly funded universal population screening strategy in Australia. Other, less expensive screening models have been proposed, and usually use maternal age and/or combined first trimester screening in a contingent screening model (typically using risk cut-offs of 1 in 1000–2000, capturing more than 95% of fetal trisomy 21 cases).9,10

Since its initial introduction in late 2012, NIPT has been embraced with vigour by some Australian women and health care providers;11 it is available to any woman who is able to pay for it, regardless of medical risk level.12 There are strong arguments, both ethical and financial, for making NIPT available through the public health system in the near future, at least for major chromosomal anomalies. The inequity of access to NIPT and its current lack of regulation is a concern to many health care professionals.

Caveats and limitations

NIPT is not without its limitations. Failure to obtain a result occurs in routine clinical practice in about 3%–4% of tests,11 usually due to a low cffDNA fraction, typically because the sample was collected too early in the pregnancy or because of maternal obesity. False-positive results have been associated with confined placental mosaicism, the death of a co-twin, maternal malignancy and maternal mosaicism. Detection rates appear to be lower and the chances of not obtaining a result are higher in twin than in singleton pregnancies.

Ethical questions, ever present and never fully resolved when discussing prenatal testing, will come more sharply into focus in Australia, as elsewhere, with the broader introduction of NIPT into obstetric practice. A woman and her partner have two options after trisomy 21 has been diagnosed: continuation or termination of the pregnancy. The option of termination is widely regarded in our society as part of the couple’s reproductive health rights. If the diagnosis is made earlier, and termination methods that are less stressful and safer for the woman and more acceptable to medical staff are available, there could be greater pressure to undergo testing (and termination, when abnormalities are detected) than is currently the case. What message does this then send to people with trisomy 21 in our community and their families? The same arguments would apply to other non-lethal chromosomal anomalies, such as Turner syndrome. Of even greater consequence would be the ability to discover the sex of the fetus at an early stage. Sex-based termination, widespread in some parts of Asia, is believed to be uncommon in Australia, but this situation could change were the sex known much earlier in pregnancy. The wider introduction of NIPT must be accompanied by appropriate increases in the provision of genetic counselling services for women and of education for health care providers.

The final goal of NIPT, although it may be some way off, will be sequencing of the entire fetal genome on a commercial basis, with the prospect of personalised prenatal diagnosis available to all, or at least to all who can afford it. Before this happens, it is important that we have widespread community understanding of all the questions involved, so that wise decisions can be made about how this testing should be made available to Australian women.12

Don’t panic about the pill – it’s safer than driving to work

A study published in The BMJ yesterday suggests a link between newer contraceptive pills and higher risk of serious blood clots. The finding is not new, but it may be cause for a different kind of concern.

During their fertile years, between three and five women of every 10,000 who are not pregnant and not taking the pill are likely to develop blood clots every year. The research published today found older contraceptive pills double this “background” risk of blood clots, and the newer pills have roughly doubled the risk again.

Several studies published over the past 20 years show very similar findings. What this research brings to the table are larger numbers of women and more careful attention to factors in their medical history that could potentially skew the results.

It’s likely the media will pounce on this story; there will be testimonies from women who have experienced blood clots while taking the pill and a plethora of personal injury lawyers spruiking their business. Women across the world will be scared into stopping their contraception until it all blows over. I know this because I’ve seen it before, and I think that’s what we should be concerned about.

Back and forth

We’ve known the pill increases a woman’s risk of blood clots and stroke since it was first marketed. By the 1990s, concern had been tempered by the fact that this risk was greatly reduced by pills containing only a quarter of the oestrogen compared to the 1960s. The development of several newer progestogens in the 1980s had also increased the range of pills available, making it more likely that most women could find a combination that suited them.

But then, in 1995, three studies published in The Lancet suggested pills containing these newer progestogens posed twice the risk of blood clots as the older ones, just as the study published today does. Frenzied media coverage of the finding led many women to simply discontinue their contraception.

As a result, 1995-96 saw a 9% increase in abortion rates in Britain as well as a 25% increase in births. And both pregnancy and birth hold significantly higher risks of blood clots than any contraceptive pill, with rates at least ten times higher.

Within a few years, the controversy settled down somewhat when a number of epidemiologists pointed out that doubling an extremely small risk has no significant public health impact. But then, between 2007 and 2014, it all started again when a number of studies reached conflicting conclusions about whether there were any real differences in clotting rates between various pills.

Every woman considering using the contraceptive pill should discuss the risks it poses to her health as well as available alternatives with her prescribing doctor.
Annabelle Shemer/Flickr, CC BY-NC-ND

Two very large studies that kept track of women from the time they started various pills showed no difference in blood-clotting risk between any of the pills the women were taking. But research like this is extremely expensive and only within the funding reach of either governments or pharmaceutical companies. In this case, it was the latter. Despite the fact that both studies were independently monitored, they were attacked as having commercial bias.

Four out of six of the remaining studies, which used various databases to look back from a blood clot diagnosis and capture the kind of pill the woman was taking, suggested the newer pills doubled risk. But the inevitable lack of “real-time” information in this kind of research also leaves it open to many potential biases.

The just-published BMJ study provides further evidence for research showing increased risk, and its publication will no doubt re-ignite the debate about the safety of newer contraceptive pills. It’s clearly time for an appraisal of the actual risks involved.

Being cautious

Even if we were to adopt the worst-case scenario from all the studies published to date, being on one of the older versions of the contraceptive pill increases the risk of blood clot from three and five per 10,000 women each year to somewhere between five and eight. Taking one of the newer ones raises it to between nine and 14.

So although a doubling of clotting risk sounds alarming, it actually translates to an additional four to six cases per 10,000 users of the newer pills a year.

It’s also important to recognise that only one in 100 women who have a blood clot will die from it. That risk of death could be cancelled out statistically by driving for two fewer hours each year. Put another way, the risk of a woman dying from a road accident is approximately 25 times that of death from a pill-related clot.

This is not to say we should be blasé about the risks posed by the contraceptive pill. It is above all a medication, which means some of its benefits may be compromised by – potentially serious – side effects. Every woman considering using the contraceptive pill should discuss the risks it poses to her health as well as available alternatives with her prescribing doctor.

This study adds to what is known about blood clot risk on various oral contraceptive pills, but it doesn’t claim to provide the definitive answer. I hope that, as we again debate the risks posed by the pill, we don’t lose sight of the fact that, for most women, the benefits of combined contraceptives far outweigh risks.

Terri Foran is Lecturer in the School of Women’s and Children’s Health at UNSW Australia.

This article was originally published on The Conversation.
Read the original article.

The global challenge of women’s health

Sierra Leone, a West African state of 6 million, saw 11 000 cases and over 3000 deaths during last year’s Ebola outbreak. A bitter civil war from 1991 to 2002, fuelled largely by fierce factions from neighbouring countries, led to 50 000 deaths and degradation of the country’s infrastructure and social fabric. Sierra Leone’s exports of diamonds and bauxite notwithstanding, the lack of a socially responsive polity and a largely agrarian population set the scene for the epidemic. Over 70% of its population live in extreme poverty.1

Sierra Leone also tops the 2013 chart when it comes to maternal deaths — 1100 per 100 000 live births.2 The comparable figure for Australia is six. UNICEF estimates that 88% of the women have been subject to genital mutilation.3

Improving maternal health

The Millennium Development Goals, promulgated by the United Nations in September 2000 and endorsed by 189 countries, sought to halve desperate poverty, defined as living on less than a dollar a day, by 2015. The metrics suggest that this goal has been achieved, and it is a remarkable tribute to international efforts. Among the eight goals, five concern health, and most have been achieved, including huge reductions in infant mortality.

Improving maternal health is one of the health-related goals that has proved harder to reach. Under Goal 5, countries committed to reducing maternal mortality by three-quarters between 1990 and 2015. Since 1990, maternal deaths worldwide have dropped by 45%.4

Maternal mortality — often due to blood loss and infection — has proved more resistant to efforts to substantially reduce it as a global health problem. It has been intractable in areas of poverty and social turmoil. There were 289 000 maternal deaths worldwide reported in 2013.4

The explanation for these disturbing figures has much to do with social attitudes and investment. When we encounter health disparities, the explanation is most often found outside the clinic, in society and politics. In preventing maternal death, strong investment in education for women is fundamental. Provision of the basic infrastructure necessary for safe childbirth comes next. But even more basic is a pathological view of women — that they are not a priority and that public resources should be invested elsewhere.

Broadening the focus

The World Health Organization draws our attention in 2015 to food security. Its importance is great for women’s health, before and during reproduction and throughout all adulthood, to reduce the risk of nutritional deficiencies, diabetes and heart disease.

When, in 2003–2004, my colleagues at Columbia University and I were examining cardiovascular disease in emerging economies, I was amazed to discover that it far outweighed obstetric and perinatal disorders, HIV and malaria as causes of death of women in the years of family formation and support. In seven out of nine developing countries that we studied, chronic diseases caused over 20% of deaths among women aged 15–34 years, while reproductive causes and HIV together accounted for about 10% of deaths.5 We questioned why the traditional conceptualisation of women’s health has more to do with disorders that impair their performance as reproductive machines than with the real threats to their wellbeing, including the precursors of cardiovascular catastrophe. Those who work on global programs to abate the scourge of diabetes make a major contribution to reducing deaths among women from cardiovascular disease.

Shaking stereotypic thinking

Even if our view of women’s health is restricted to an understanding of causes of death, it is clear we have a task to shake the stereotypic thinking and social relegation of women that foster a completely inadequate global response to their health needs.

There are tasks aplenty for those with advocacy in their blood at governmental, educational and individual levels. Heroic clinicians such as 91-year-old Dr Catherine Hamlin AC and her co-workers at the Addis Ababa Fistula Hospital, its five regional hospitals and the Hamlin College of Midwives set outstanding examples of other pathways.

Pharmacological management of low milk supply with domperidone: separating fact from fiction

To the Editor: After publication of our peer-reviewed perspective article on the safe and effective use of domperidone in the management of low milk supply,1 we are disappointed to have been informed that the Australian medicines handbook (AMH) has removed lactation stimulation as a recognised off-label indication for domperidone (as well as metoclopramide). It appears that this change could be a kneejerk response to recent regulatory warnings, despite these being highlighted as not being of relevance to most breastfeeding women.1,2 These changes carry with them the significant potential to create confusion among prescribers and the public.

We would like to re-emphasise the points we have made previously that further restrictions regarding the use and availability of domperidone do not appear to be warranted; they risk subjecting women to additional emotional trauma and are not in the interests of the immediate and long-term health of breastfeeding women and their babies.1,2 There is absolutely no evidence that domperidone places breastfeeding mothers at increased risk of severe cardiac events, whereas, in contrast, there is evidence of the significant benefits to be gained from the use of domperidone in supporting breastfeeding. Domperidone has been used widely in clinical practice for women experiencing low milk supply,3 with no reports of significant adverse effects.

We also re-emphasise our previous practice points regarding the importance of adequately screening women for factors that may place them at increased risk of cardiac arrhythmia, should this be of concern.

It is clear that domperidone is not a magic bullet for adequate milk production in breastfeeding women. Non-pharmacological management options are the mainstay of treatment, with medications only to be considered as a last resort, and always in addition to these management strategies.4 There is no easy fix when it comes to supporting women who are struggling with low milk supply, but when used appropriately, domperidone can be a very safe and effective treatment option.

Odds, risks and appropriate diagnosis of gestational diabetes: comment

In my opinion, McIntyre and colleagues have misunderstood the primary point I raised in my earlier article in the Journal.1 My concern was not the increased number of women diagnosed with gestational diabetes mellitus (GDM), but the accuracy of the diagnostic thresholds. Although I suggested that many women in the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study with one elevated blood glucose level may not be at risk, I also stated that additional women with two or more blood glucose levels marginally below the new thresholds will be at risk. The authors’ comment that an increased odds ratio has the same effect confirms that they misunderstood my concern.

My concerns about the combined primary end point were valid and I acknowledge the response provided — but is cord C-peptide > 90th centile a true adverse neonatal outcome?

I agree that thresholds should be determined by fully adjusted models.

However, no data have been provided to refute the assertion that many women with only one elevated blood glucose level may not be at risk. Each blood glucose threshold has been determined independently and not adjusted for normal blood glucose levels.2 The thresholds should accurately define patients at risk in the HAPO study,3 so the number of women diagnosed should approximate the total number of cases of GDM (since a woman can only have one case of GDM per pregnancy). Of 23 316 participants, 3746 (16.1%) were diagnosed with GDM. There were 1935 women with an elevated fasting blood glucose level, 1928 with a raised 1-hour blood glucose level, and 1309 with a raised 2-hour blood glucose level — 5172 cases of GDM (22.2%) in total, or 38% more cases than women diagnosed with GDM.2,4 This occurs because women with two or more elevated blood glucose levels are double- or triple-counted as diagnoses. Women with two or more elevated levels have the highest rate of adverse outcomes.4 These women contribute disproportionately to the total number of adverse events for each parameter.

There is an expectation that the lower limit of the confidence interval should exceed the diagnostic odds ratio threshold, given that the minimal odds ratio of groups of these women should be 1.75. This raised additional concerns1 that many women with only one elevated blood glucose level will belong to blood glucose level categories at reduced risk, when the overall odds ratio of these cohorts is equivalent to the risk threshold.4

This is not a new problem. It confronted the Framingham statisticians who stated “analyses that fail to examine risk factors in combinations usually greatly overestimate the population-attributable risks associated with individual risk factors”,5 which is exactly the situation for women with one elevated blood glucose level only. They also stated that “multivariable risk assessment also avoids overlooking high-risk CVD [cardiovascular disease] candidates with multiple marginal risk factors and avoids needlessly alarming persons with only 1 isolated risk factor”,5 which is also the situation for women with two or more blood glucose levels just below the diagnostic thresholds in contrast to those with a single marginally elevated blood glucose level. It is reasonable to question the recommended implementation of the current thresholds.2 Subgroup analysis of the 22 152 women with two normal blood glucose levels is a valid suggestion.

The HAPO study has a wealth of data.3 Clinicians require tools that accurately diagnose women at risk. The interaction of blood glucose parameters is similar to the interaction of cardiovascular risk factors; a single cholesterol threshold does not define absolute risk for all people. We have a great opportunity to develop a risk model that would more accurately identify women at risk, based on the three blood glucose levels, age, weight, ethnicity and other parameters. This would make the most of the data from this excellent study. It has been done for cardiovascular disease, why not for GDM?

Reduced breast milk feeding subsequent to cosmetic breast augmentation surgery

Breastfeeding is beneficial for infants and their mothers. It protects against diarrhoea, respiratory tract and other infant infections, atopic dermatitis, asthma, obesity, diabetes and cancer.1,2 Although exclusive breastfeeding achieves optimal infant growth and development, the World Health Organization recognises that providing some breast milk to the infant is better than none.3 For mothers, breastfeeding has a contraceptive effect, and reduces the risk of type 2 diabetes, breast cancer and ovarian cancer.1

Cosmetic breast augmentation is the most common plastic surgical procedure, and its use is rising dramatically. In Australia, this surgery increased by 150% between 2001 and 2011.4 In the United States, the estimated increase for this period was 45%, although this followed a 550% increase from 1992 to 2000.5 In the United Kingdom, rates increased by 200% from 2005 to 2013.6 In this article, cosmetic breast augmentation (or breast implants) refers to procedures that change the size, shape and texture of healthy breasts. This is distinct from reconstructive breast augmentation, such as following mastectomy.

Although most cosmetic breast surgery occurs among women of reproductive age, there has been little research into pregnancy outcomes, including breastfeeding. A systematic review of breastfeeding outcomes associated with cosmetic breast augmentation surgery identified only three small, observational studies.7 One study reported reduced rates of any breastfeeding among women with breast augmentation, while meta-analysis of all three studies suggested a reduced likelihood of exclusive breastfeeding (pooled rate ratio, 0.60; 95% CI, 0.40–0.90).7 The authors recommended that studies using larger cohorts and more representative study populations be used to explore the observed association.

To test the null hypothesis that augmentation has no effect on breast milk feeding, we conducted a population-based study to determine the effect of cosmetic breast augmentation: (i) on any breast milk feeding in a subsequent pregnancy; and (ii) on exclusive breast milk feeding among women who breast milk fed.

Methods

The study population was derived from the 391 979 women who gave birth in New South Wales from 1 January 2006 to 31 December 2011 (Box 1). As our intention was to examine the effect of cosmetic breast augmentation, women with breast cancer, mastectomy, breast reconstruction or other breast surgery before giving birth were excluded (n = 3831; Box 1 and Appendix 1). The remaining 388 148 women had 506 942 births. The first birth in the study period or the first birth after breast augmentation surgery was used in the primary analysis.

Data for the study were obtained from two linked population health datasets: the NSW Perinatal Data Collection (PDC; referred to as birth records) and the NSW Admitted Patient Data Collection (APDC; referred to as hospital records). The PDC is a statutory surveillance system of all births in NSW of at least 20 weeks’ gestation or at least 400 g birthweight. Information on maternal characteristics, pregnancy, labour, delivery, and infant outcomes are recorded by the attending midwife or doctor. The APDC is a census of all NSW inpatient hospital discharges from both public and private hospitals, and day procedure units, and includes demographic and episode-related data. Diagnoses and procedures are coded for each admission from the medical records according to the International Classification of Diseases, 10th revision, Australian modification (ICD-10-AM) and the Australian Classification of Health Interventions.8

Hospital records for individual women were linked cross-sectionally to birth records from 2006 to 2011 and longitudinally (from July 2000 to December 2011). Thus, the minimum lookback period for prior breast surgery ranged from 5.5 to 11.5 years. Record linkage was undertaken by the NSW Centre for Health Record Linkage (CHeReL). For this study, the CHeReL reported the quality of the record linkage9 as 3/1000 false-positive links. We were provided with anonymised data. Ethics approval for the study was obtained from the NSW Population and Health Services Research Ethics Committee.

Breastfeeding information at discharge has been collected in birth records since 2006. One or more of the following three options can be reported in tick-boxes: “breastfeeding”, “expressed breast milk” or “infant formula”.

The primary outcome was any breast milk feeding (any breast milk, with or without infant formula) at discharge from birth care. Consistent with other studies,7 the secondary outcome was exclusive breast milk feeding (only breast milk, either directly from the breast and/or as expressed breast milk) among those with any breast milk feeding.

The exposure of interest was cosmetic breast augmentation, which has a specific surgical procedure code (45528-00) in the Australian Classification of Health Interventions.4,8 This code is distinct from unilateral breast augmentation and breast augmentation following mastectomy. Hospital records from 2000 onwards were available for identification and date of surgery.

Other factors potentially predictive of breast milk feeding at discharge from maternity care that were available for analysis included: maternal age, country of birth, socioeconomic status according to the Australian Bureau of Statistics Index of Relative Socio-economic Disadvantage,10 marital status, urban or rural residence, private care, parity, multifetal pregnancy, antenatal care before 20 weeks’ gestation, smoking during pregnancy, morbid obesity, hypertensive disorders of pregnancy, diabetes (pregestational or gestational), labour analgesia, labour induction or augmentation, mode of birth, severe maternal morbidity,11 maternal postnatal length of stay, gestation, small for gestational age (< 10th birthweight for gestational age percentile), major congenital anomalies (eg, cleft lip or palate, spina bifida, tracheoesophageal fistula), neonatal intensive care unit admission, and perinatal mortality. These factors are known to be reliably reported.12

Statistical analysis

Descriptive statistics were used to summarise the distributions of maternal and pregnancy characteristics among all women with and without breast augmentation. Poisson regression modelling with robust standard errors13 was employed to determine the association of breast augmentation with (i) any breast milk feeding (compared with none) and (ii) exclusive breast milk feeding (compared with non-exclusive) among the “any breast milk feeding” group.

To avoid confounding by factors likely to be associated with reduced breastfeeding,14,15 regression analyses were limited to women who had a singleton infant with no major congenital anomalies and born at term (≥ 37 weeks). Crude and adjusted relative risks (RRs) with 95% confidence intervals were estimated for characteristics likely to be associated with breastfeeding.

Finally, among women with at least two births in the study period, we examined the primary and secondary breastfeeding outcomes across births in the following groups: no breast augmentation, breast augmentation between births, and breast augmentation before both births. The before-and-after effect of breast augmentation among women who had breast augmentation surgery between births was assessed using the McNemar test of paired data, with continuity correction.

Results

Of the 388 148 women who were eligible for the study, 902 had documentation of cosmetic breast augmentation surgery (Box 1). Breastfeeding information at discharge was missing in 9759 records (2.51%). Among the remaining 378 389 women, 892 (0.24%) had breast augmentation before a birth. The median age at the time of breast augmentation surgery was 28 years (range, 18–43 years), and the median interval between surgery and birth was 3.1 years (range, 1.0–10.1 years).

Maternal, pregnancy and birth characteristics for all women with and without breast augmentation are presented in Box 2. At discharge, 705 women (79.0%) with breast augmentation provided any breast milk to their infants, compared with 88.5% of women without breast augmentation.

Breast milk feeding outcomes were then assessed among 341 953 singleton infants with no major congenital anomalies born at term. Compared with women without, women with breast augmentation had reduced likelihood (adjusted RR, 0.90; 95% CI, 0.87–0.93) of feeding their infant with any breast milk at the time of discharge from birth care. Factors controlled for that were positively associated with breast milk feeding included: older maternal age, non-Australian-born, high socioeconomic status, nulliparity, non-smoker, no obstetric interventions, and longer hospitalisation after birth (Appendix 2). Women with breast augmentation in the 2 years preceding birth had similar rates of any breast milk feeding to women with a longer period since breast surgery (77% v 81%; P = 0.17).

For women whose infants received any breast milk, there was no association between breast augmentation and exclusive breast milk feeding. Among these, 593 women (94.0%) with breast augmentation exclusively breast milk fed. The adjusted RR for exclusive breast milk feeding among women with breast augmentation, compared to those without, was 0.99 (95% CI, 0.97–1.01).

Among the 106 835 women with two births during the study period, 106 593 had no record of breast augmentation, 167 had breast augmentation before both births, and 75 had breast augmentation between the two births. The rates of any breast milk feeding and exclusive breast milk feeding at the first and second births were compared for these three groups of women (Box 3). The rate of any breast milk feeding was the same for both births among women with no augmentation (87%). Among women with breast augmentation between the births, the rate declined from 87% in the first birth to 72% in the second birth (P = 0.02). There was no evidence of significant change among women with augmentation before both births (77.2% v 73.7%; P = 0.29; Box 3, A). However, among women who provided any breast milk, the rate of exclusive breast milk feeding was similar in first and second births for women with and without breast augmentation (Box 3, B).

Discussion

This is the first study to document the population prevalence of cosmetic breast augmentation in a maternity population, and the largest to compare breast milk feeding outcomes for women with and without cosmetic breast augmentation. We found that women with breast augmentation are less likely to provide their infants with any breast milk at the time of discharge. However, among women who provide breast milk, women with breast augmentation are no more or less likely to exclusively breast milk feed their infants. Both the main population analysis and the subgroup analysis of women with breast augmentation between births showed lower rates of any breast milk feeding following augmentation surgery. This consistency of findings strengthens the case that there is an effect, although possible mechanisms are unclear.

Uptake of breast augmentation surgery is increasing, with 8000 Australian, 10 000 British and 307 000 American women undergoing the procedure in 2011.4–6 We found that 79% of these women can be expected to breast milk feed at discharge, compared with 89% of women without surgery. As maternity care affects breastfeeding success,2 these findings underscore the importance of identifying, supporting and encouraging all women who are vulnerable to a lower likelihood of breastfeeding.

Underlying breast hypoplasia and insufficient lactogenesis have been suggested as a reason for reduced breastfeeding rates among women with breast augmentation.16 However, we found that among women who had breast augmentation between births, any breast milk feeding fell from 87% in the “before augmentation” birth to 72% in the “after augmentation” birth, while the rates in comparison groups remained stable. A demonstrated ability to provide breast milk before augmentation surgery suggests that hypoplasia is not the explanation for lower breastfeeding rates among women with breast augmentation. Similar to the one existing population-based study,17 we found no association between breast augmentation and adverse birth outcomes, including preterm birth, small for gestational age, congenital anomalies, neonatal intensive care unit admission or perinatal death.

Lower breastfeeding rates may reflect maternal and family attitudes and expectations, may be a consequence of surgery, or the breast implants may reduce the ability to lactate. Although a variety of health outcomes have been investigated among women who have silicone breast implants, and their breast milk fed infants, epidemiological studies have not substantiated links with adverse outcomes.18–21 Nevertheless, women with breast implants may fear transmitting silicone or other breast implant materials into breast milk. They may also fear, or have been told by their surgeon, that breastfeeding could undo a satisfactory augmentation result. Another explanation is that lactiferous ducts, glandular tissue or nerves of the breast are damaged during surgery, or by pressure from the implants on breast tissue.22 Furthermore, complications of the surgery including capsular contracture, haematoma formation, infection or pain may reduce the ability or desire to breastfeed.22 Future qualitative research is needed to better understand why women with prior breast augmentation are less likely to breastfeed.

Our findings of reduced rates of any breastfeeding are consistent with the only study that reported rates of any breastfeeding after augmentation among women who attempted breastfeeding.16 However, the latter study reported a stronger effect at 2 weeks postpartum (RR, 0.67; 95% CI, 0.50–0.91). In contrast, our findings differ from the systematic review of three small studies, which found women with breast implants who breast milk fed were less likely to exclusively breastfeed.7 We believe our whole-population findings are more robust. The previous studies had selected populations (eg, lactation referral clients) and variable end points (eg, exclusive breastfeeding, insufficient lactogenesis), used historical controls and made limited attempts to control for potential confounders.7 However, it is possible that differences in the rates of exclusive breastfeeding may become apparent after discharge, as follow-up in the three studies was longer (minimum 2 weeks postpartum).

A strength of our study is the use of recent, large, linked population health datasets that include a third of all births in Australia. Breastfeeding information is reported by a midwife, and previous validation studies show events occurring around birth or immediately postpartum are well reported.12 Longitudinal record linkage allowed the ascertainment of cosmetic breast augmentation surgery. Although a longer lookback period may have increased case ascertainment,23 some missed cases among a population of more than 300 000 women without breast augmentation are unlikely to change the findings. Similarly, women who have cosmetic surgery overseas or interstate are not captured in this study. Identification of breast augmentation surgery in routinely collected data has not been evaluated but, in general, surgical procedures are reliably identified in hospital discharge data, and other breast surgery, such as mastectomy, is accurately reported (sensitivity, 97%; positive predictive value, 97%).12,24

Another strength is that breastfeeding was assessed at the same time for both exposed and unexposed women, unlike prior studies.7 The 89% breastfeeding rate at discharge in our study is similar to the rate reported in the Australian National Infant Feeding Survey (90.2% for < 1 month).25

However, information on breastfeeding initiation was not available. If women with breast augmentation initiated breastfeeding but gave up before discharge, the rate of exclusive breastfeeding could be lower if these women were included in the “any breastfeeding” denominator. Another limitation of the study is that breastfeeding is only assessed at one time point (discharge). Breastfeeding rates decline steadily over the first months of infancy25 and it is unclear whether this decay would be the same for women with and without breast augmentation. Information was not available on intention to breast milk feed, paternal support for breastfeeding, nor on the details of the breast augmentation surgery, such as the incision type or the type and volume of the breast implant.

An absolute rate of one in five women with breast augmentation who subsequently give birth may be unable or unwilling to breast milk feed their infants. This information should be provided as part of informed decision making to women contemplating breast augmentation surgery.

1 Study population flowchart, 2006–2011

2 Maternal, pregnancy and birth characteristics for participants, by breast augmentation status

	Breast augmentation (n = 892), no. (%)	No breast augmentation (n = 377 497), no. (%)	P*

Mother’s age at birth (missing = 106)			< 0.001
< 20 years	3 (0.3%)	15 406 (4.1%)
20 to < 35 years	608 (68.2%)	276 043 (73.2%)
≥ 35 years	281 (31.5%)	85 942 (22.8%)
Region of birth (missing = 1489)			< 0.001
Australia or New Zealand	761 (85.5%)	264 041 (70.2%)
Asia	45 (5.1%)	58 811 (15.6%)
Other	84 (9.4%)	53 158 (14.1%)
Married or de facto	718 (80.5%)	308 709 (81.8%)	0.32
Socioeconomic status (missing = 6140)			< 0.001
Most disadvantaged	103 (11.6%)	79 232 (21.3%)
Disadvantaged	134 (15.1%)	71 517 (19.3%)
Average	159 (17.9%)	75 027 (20.2%)
Advantaged	210 (23.7%)	71 656 (19.3%)
Most advantaged	282 (31.8%)	73 929 (19.9%)
Urban residence at birth	653 (73.2%)	263 218 (69.7%)	0.02
Private care	370 (41.5%)	120 211 (31.8%)	< 0.001
Nulliparous	378 (42.4%)	206 078 (54.6%)	< 0.001
Multifetal pregnancy	18 (2.0%)	5282 (1.4%)	0.12
First antenatal visit < 20 weeks’ gestation	834 (93.5%)	344 892 (91.4%)	0.02
Smoking during pregnancy	85 (9.5%)	45 073 (11.9%)	0.03
Hypertensive disorders	70 (7.9%)	38 568 (10.2%)	0.02
Diabetes	32 (3.6%)	26 621 (7.1%)	< 0.001
Morbid obesity	0	1277 (0.3%)	0.08
Regional labour analgesia	284 (31.8%)	101 925 (27.0%)	0.001
Labour induction	256 (28.7%)	103 368 (27.4%)	0.38
Mode of birth (missing = 287)			0.62
Unassisted vaginal	485 (54.4%)	210 506 (55.8%)
Instrumental vaginal	130 (14.6%)	51 447 (13.6%)
Caesarean section	276 (31.0%)	115 258 (30.6%)
Severe maternal morbidity	12 (1.4%)	6102 (1.6%)	0.52
Mother’s postnatal length of hospital stay			0.78
1–2 days	327 (37.2%)	132 944 (35.7%)
3–4 days	359 (40.8%)	157 913 (42.4%)
5–6 days	168 (19.1%)	70 634 (19.0%)
≥ 7 days	25 (2.8%)	10 869 (2.9%)
Preterm birth (< 37 weeks’ gestation)	61 (6.8%)	21 871 (5.8%)	0.18
Small for gestational age	75 (8.4%)	35 722 (9.5%)	0.28
Neonatal intensive care unit admission	119 (13.3%)	53 510 (14.2%)	0.48
Major congenital anomalies	36 (4.0%)	13 842 (3.6%)	0.50
Perinatal mortality	0	8	0.89
Infant feeding at discharge
Any breast milk feeding	705 (79.0%)	334 250 (88.5%)	< 0.001
No breast milk feeding (formula only)	187 (21.0%)	43 247 (11.5%)
Exclusive breast milk feeding among women who provided any breast milk	653 (92.6%)	308 552 (92.3%)	0.76
Breast-related readmission within 6 weeks	13 (1.4%)	4471 (1.2%)	0.42

* χ2 test.

3 Breast milk feeding outcomes for women with two births, showing the before-and-after effect of breast augmentation, 2006–2011

When the light begins to fade

Being mortal is about the struggle to cope with the constraints of our biology, with the limits set by genes and cells and flesh and bone. Medical science has given us remarkable power to push against these limits, and the potential value of this power was a central reason I became a doctor. But again and again, I have seen the damage we in medicine do when we fail to acknowledge that such power is finite and always will be.

We’ve been wrong about what our job is in medicine. We think our job is to ensure health and survival. But really it is larger than that. It is to enable well-being. And well-being is about the reasons one wishes to be alive.

Atul Gawande. Being mortal. London: Profile Books, 2014.

Atul Gawande, 49, is a general surgeon at the Brigham and Women’s Hospital in Boston and holds academic positions in the Harvard Medical School and School of Public Health. He is also a writer. Being mortal uses both his clinical experiences and literary skills as they relate to the care we give, or should give, when our patients enter the twilight zone between responsiveness to medical and surgical interventions and when they move beyond the reach of our finest instruments and technical remedies.

The central question of the book is not new, nor are the answers. The question is this: given the inevitable decline in capacity as we age, especially when sick, how can medicine best assist us? The answer is that by determining what it is that patients perceive to be what gives their life purpose, we should work with them so that it can be achieved. Of course, that is not always possible but it is more likely than we may think.

The book is packed with cases that demonstrate that this is so. Each page is informed by Gawande’s clinical experiences, including interviews with hundreds of patients and family members about their experiences of illness and ageing, as well as discussions with health care workers including aged care, nursing, palliative care specialists, geriatricians and other experts in the field.

Gawande’s arguments derive from, and are informed by, the immediacy of his clinical practice. The book contains eight chapters: The Independent Self, Things Fall Apart, Dependence, Assistance, A Better Life, Letting Go, Hard Conversations, Courage, and a personal epilogue about Gawande’s experience as he shared his father’s final journey over years with cancer.

Gawande was born to Indian parents (both doctors) in Brooklyn and grew up in Athens, Ohio. He graduated first from Stanford in 1987 and was a Rhodes Scholar in philosophy, politics and economics at Balliol College, Oxford, in 1989. He completed his medical studies at Harvard in 1995. A political activist, he helped the first Clinton presidential campaign and worked on his ill fated health care reform efforts. He has published highly acclaimed articles with The New Yorker on medical matters (including the chapter in this book entitled Letting Go) and became a staff writer there in 1998. He gave the 2014 Reith Lectures on BBC Radio 4 on The Future of Medicine; his first presentation entitled Why do Doctors Fail? He wrote Being mortal at the Rockefeller villa in Bellagio, Italy.

The medical curriculum and mortality

Gawande observes:

I learned about a lot of things in medical school, but mortality wasn’t one of them . . . What worried us was knowledge . . . Yet within a few years, when I came to experience surgical training and practice, I encountered patients forced to confront the realities of decline and mortality, and it did not take long to realize how unready I was to help them.

His is not a book about palliative care and certainly not about euthanasia, although both feature in it: rather, it is about discerning together with patients and carers who have entered the twilight zone what their wishes are about the life that lies ahead of them, and then making judicious choices — sometimes heroic, sometimes quiet and low-key — in relation to therapy and support that are most likely to enable and enhance those choices.

People with serious illness have priorities besides simply prolonging their lives. Surveys find that their top concerns include avoiding suffering, strengthening relationships with family and friends, being mentally aware [even at the expense of foregoing narcotic analgesics], not being a burden on others, and achieving a sense that their life is complete . . . The question therefore is not how we can afford [our system of technological care]. It is how we can build a health care system that will actually help people achieve what’s important to them at the end of their lives.

Gawande heads up a health system innovation centre at Harvard.

Gawande’s book is a gallery of portraits of patients, carers and doctors and a document of conversations with them that reassure, challenge and inspire. Oliver Sacks, himself a great medical writer, says:

Being Mortal is not only wise and deeply moving, it is an essential and insightful book for our times, as one would expect from Atul Gawande, one of our finest physician writers.1

Late for Christmas? Consider giving yourself a copy for the new year!